His 103 World Civilizations I- Week 5 Final

http://www.cambridge.org/9780521827171 This page intentionally left blank History and Memory in the Carolingian World The writing and reading of history in the early middle ages form the key themes of this book. The primary focus is on the remarkable man- ifestations of historical writing in relation to historical memory in the Frankish kingdoms of the eighth and ninth centuries. The book consid- ers the audiences for history in the Frankish kingdoms, the recording of memory in new genres including narrative histories, cartularies and Libri memoriales, and thus particular perceptions of the Frankish and Christian past. It analyses both original manuscript material and key his- torical texts from the Carolingian period, a remarkably creative period in the history of European culture. Presentations of the past developed in this period were crucial in forming an historical understanding of the Greco-Roman and Judaeo-Christian past and, in subsequent centuries, of early medieval Europe. They also played an extraordinarily influential role in the formation of political ideologies and senses of identity within Europe.   is Professor of Medieval History in the University of Cambridge. Her previous publications include The Carolingians and the Written Word (1989), The Frankish Kings and Cul- ture in the Early Middle Ages (1995) and The New Cambridge Medieval History, II, c. 700–c.900 (1995). She has presented many conference papers and lectured extensively at universities in Britain, continental Europe, North America and Australia. She is a Fellow of the Royal Historical Society. History and Memory in the Carolingian World Rosamond McKitterick Professor of Medieval History, University of Cambridge and Fellow of Newnham College cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge cb2 2ru, UK First published in print format isbn-13 978-0-521-82717-1 isbn-13 978-0-521-53436-9 isbn-13 978-0-511-21108-9 © Rosamond McKitterick 2004 2004 Information on this title: www.cambridge.org/9780521827171 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. isbn-10 0-511-21285-2 isbn-10 0-521-82717-5 isbn-10 0-521-53436-4 Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Published in the United States of America by Cambridge University Press, New York www.cambridge.org hardback paperback paperback eBook (EBL) eBook (EBL) hardback http://www.cambridge.org http://www.cambridge.org/9780521827171 In loving memory of Claude Anthony Pierce, 21 November 1919–16 May 2001 Contents Preface page ix Acknowledgements xii List of abbreviations xiii 1. Introduction: History and memory in the Carolingian world 1 2. Carolingian history books 28 3. Paul the Deacon’s Historia langobardorum and the Franks 60 4. The Carolingians on their past 84 5. Politics and history 120 6. Kingship and the writing of history 133 7. Social memory, commemoration and the book 156 8. History and memory in early medieval Bavaria 174 9. The reading of history at Lorsch and St Amand 186 10. Texts, authority and the history of the church 218 11. Christianity as history 245 12. Conclusion: History and its audiences in the Carolingian world 265 Bibliography 284 Index of manuscripts 323 General index 327 vii Preface This book’s themes are the writing and reading of history in the early middle ages. The primary focus is on the many remarkable manifesta- tions of historical writing in relation to historical memory in the Frankish kingdoms of the eighth and ninth centuries. I consider the audiences for history in the Frankish kingdoms and the recording of memory in various new genres, including narrative histories, cartularies and Libri memori- ales, and thus particular perceptions of the Frankish and Christian past. I offer analyses of manuscript material and of key historical texts from the Carolingian period, a remarkably creative period in the history of European culture. Presentations of the past developed in the eighth and ninth centuries were crucial in the formation of an historical understand- ing of the Greco-Roman and Judaeo-Christian past, as well as for the his- tory of early medieval Europe in subsequent centuries. They also played an extraordinarily influential role in the formation of political ideologies and senses of identity within Europe. This book draws in part on material already published in articles or chapters in books over the past decade, but here presented in a completely revised and augmented form. I am grateful to the original publishers as listed below for their kind permission to make use of my work in this way. In Cambridge I am fortunate in being able to draw on the wonder- ful resources of the Cambridge University Library, and I should like to thank all the staff in Manuscripts and Rare Books, the Periodicals De- partment, the West Room, the Reading Room, the Anderson Room, the Map Room, and the departments of Accessions and Cataloguing for their unfailing helpfulness over the years. I am also greatly obliged to all the assistance given me as a reader by the staffs of the manuscripts depart- ments of Bamberg Staatsbibliothek; Berlin, Deutsche Staatsbibliothek; Brussels, Bibliothèque Royale; Copenhagen, Kongelige Bibliotek; Düsseldorf, Universitätsbibliothek; The Hague, Koninklijke Bibliotheek; Leiden, Universiteitsbibliotheek; London, British Library; Oxford, Bodleian Library; Paris, Bibliothèque nationale de France; Prague, Knihovna metropolitnı́ Kapituli; Rome, Biblioteca Apostolica Vaticana; ix x Preface St Gallen, Stiftsbibliothek; Valenciennes, Bibliothèque Municipale; Vienna, Österreichische Nationalbibliothek; Wolfenbüttel, Herzog August Bibliothek; and to many others for kindly meeting my requests for microfilms and photographs. Much of the material in this book, moreover, was first presented at conferences, as working papers at ‘workshops’, or as lectures in Aix, Auxerre, Bergen, Cambridge, Chapel Hill, Cividale, Copenhagen, Laurence (Kansas), Leeds, Lille, London, Oslo, Oxford, Paris, Perth (Western Australia), Rome, Sewanee, Utrecht, Vienna, Washington, DC, York and Zwettl. I have consequently benefited greatly from the com- ments, suggestions and reactions from the many friends, colleagues and students who heard them, especially Sverre Bagge, Lars Boje Mortensen, Claude Carozzi, Christine Carpenter, Mayke de Jong, Flavia de Rubeis, David Ganz, Carl Hammer, Wolfgang Haubrichs, Martin Heinzelmann, Yitzhak Hen, Michael Hoeflich, Matthew Innes, Dominique Iogna-Prat, William Klingshirn, Regine Le Jan, Niels Lund, John Morrill, Ruth Morse, Marco Mostert, Jinty Nelson, Thomas Noble, Michel Parisse, Richard Pfaff, Walter Pohl, Susan Rankin, Alastair Reid, Susan Ridyard, Anton Scharer, Jonathan Shepard, Terje Spurkland, Jonathan Steinberg, Huguette Taviani-Carozzi and Chris Wickham. Many of these friends were also kind enough to send me photocopies of rare editions of texts and offprints of their own work, which have been of immeasurable help. Most of the ideas explored in this book, moreover, were initially formulated in the context of lectures and classes for undergraduates and research students in Cambridge, who provide an unfailingly stimulating and de- manding audience. Despite all the efforts of government institutions to make working in a university in Britain an exhausting and demoralizing juggling act, it is the students who continue to make university teaching and research so enjoyable and worthwhile. I was especially fortunate to be elected to a Hugh Baldson Fellowship at the British School at Rome in 2002 and should like to thank all at the School who helped to make my stay in Rome so productive and enjoyable. For assistance with visits to France (in connection with my collaborative research project with Dominique Iogna-Prat) I am indebted to the British Academy and the CNRS. I am grateful, as ever, for the support offered by the Principal and Fellows of Newnham College. I am particularly indebted to my audiences in Oxford, Paris and York in spring 2003 who commented on the material offered in the introduction. My greatest debt, however, is to my current and former graduate students in early medieval history in Cambridge, and to the gatherings of graduate students in Utrecht, Vienna and Cambridge since 1997 for the ‘Texts and Identities’ workshops, for the constant stimulus of their criticism Preface xi and questions on all the topics discussed in this book. I should especially like to thank Helmut Reimitz, Max Diesenberger and Richard Corradini in Vienna for the work we have done together and the generous help they have given me. As usual, I am indebted to Cambridge University Press and the un- failingly professional assistance and support they provide for their au- thors, but I wish here, in the year of his retirement, to acknowledge my long, happy and productive association with the History Editor William Davies. I am very grateful to Liz Hosier of the Faculty of History in the University of Cambridge, who gave me invaluable help with typing. My daughter Lucy has contributed in many ways to this book, both in prac- tical assistance and with information and suggestions. But without my husband David the work for this book could not have been undertaken or completed; my lasting and most fervent thanks, as always, are to him. Cambridge, September 2003 Acknowledgements The cover picture is from the Psalterium Aureum (Golden Psalter) of St Gallen, Stiftsbibliothek MS 22, p. 140, produced in the second half of the ninth century (before 883). It depicts King David’s general Joab setting out on campaign against the Syrians and Ammonites from the illustration to Psalm 59 and is one of the few historical narrative pictures to be found in Carolingian manuscripts. I am grateful to the Bibliothekar Ernst Tremp, and the Stiftsbibliothek St Gallen, for kindly permitting me to reproduce this page. I am grateful to the following editors and publishers for kindly al- lowing me to make use in this book of work published in earlier ver- sions: Paolo Chiesa, Wolfgang Haubrichs, Jörg Jarnut, Flavia de Rubeis, Marco Mostert, Meta Niederkorn, Anton Scharer, Georg Scheibelreiter and Susan Ridyard; Blackwell (Early Medieval Europe); Brepols and Utrecht University’s Centre for Medieval Studies; Cambridge University Press; Centre d’études médiévales, Auxerre; Ecclesiastical History Soci- ety (Studies in Church History); Finnish Institute, Rome; Forschungsstelle für Geschichte des Mittelalters der Österreichischen Akademie der Wissenschaften, Vienna; Forum, Udine; Oxford University Press (English Historical Review); The University of the South (Sewanee Medieval Stud- ies); Institut zur interdisziplinären Erforschung des Mittelalters und seines Nachwirkens, Paderborn; Oldenbourg Verlag; The Royal Histori- cal Society of Great Britain (Transactions of the Royal Historical Society). xii Abbreviations Apart from the following, all works are cited in full at the first reference to them and susbsequently in short title form in each chapter. Full details may also be found in the Bibliography. BAV Rome, Biblioteca apostolica Vaticana Bischoff, Katalog Bernhard Bischoff, Katalog der festländischen Handschriften des neunten Jahrhunderts (mit Ausnahme der wisigotischen) Teil I: Aachen-Lambach (Stuttgart, 1998) Bischoff, Mittelalterliche Studien I, II, III, Bernhard Bischoff, Mittelalterliche Studien, 3 vols. (Stuttgart, 1966, 1967 and 1981) Bischoff, Schreibschulen Bernhard Bischoff, Die Südostdeutschen Schreibschulen und Bibliotheken in der Karolingerzeit, I: Die Bayerischen Diözesen, 3rd edn (Wiesbaden, 1974), II: Die vorwiegend Österreichischen Diözesen (Wiesbaden, 1980) BnF Bibliothèque nationale de France c., cc. capitulum, capitula CCCM Corpus Christianorum, Continuatio Mediavalis (Turnhout, 1966–) CCSL Corpus Christianorum, Series Latina (Turnhout, 1952–) ChLA Chartae Latinae Antiquiores, ed. Albert Bruckner, facsimile edition of the Latin charters prior to the ninth century 1– (Olten and Lausanne, 1954–) xiii xiv List of abbreviations CLA E. A. Lowe, Codices Latini Antiquiores: A Palaeographical Guide to Latin Manuscripts Prior to the Ninth Century I–XI plus Supplement, Oxford (1935–71) cod. codex CSEL Corpus Scriptorum Ecclesiasticorum Latinorum (Vienna, 1866–) DA Deutsches Archiv für Erforschung des Mittelalters EHR English Historical Review EME Early Medieval Europe fol. folio Innes and McKitterick ‘Writing of history’ M. Innes and R. McKitterick, ‘The writing of history’, in R. McKitterick (ed.), Carolingian culture: emulation and innovation (Cambridge 1994), pp. 193–222 McKitterick, Books, scribes and learning R. McKitterick, Books, scribes and learning in the Frankish kingdoms, 6th–9th centuries (Aldershot, 1994) McKitterick, Carolingians and the written word R. McKitterick, The Carolingians and the written word (Cambridge, 1989) McKitterick, Frankish kings and culture R. McKitterick, The Frankish kings and culture in the early middle ages (Aldershot, 1995) McKitterick, Migration R. McKitterick, The migration of ideas in the early middle ages (Cambridge, forthcoming) McKitterick (ed.), NCMH R. McKitterick (ed.), The new Cambridge medieval history II c. 700–c.900 (Cambridge, 1995) McKitterick, Uses of literacy R. McKitterick (ed.), The uses of literacy in early mediaeval Europe (Cambridge, 1990) MGH Monumenta germaniae historica AA Auctores antiquissimi, 15 vols. (Berlin, 1877–1919) List of abbreviations xv Cap. Capitularia, legum sectio II, Capitularia regum francorum, ed. A. Boretius and V. Krause, 2 vols. (Hannover, 1883–97) Conc. Concilia, legum sectio III, Concilia, II, ed. A. Werminghoff (Hannover, 1906–8), III, ed. W. Hartmann (Hannover, 1984) Epp. Epistulae III–VII (=Epistulae merovingici et karolini aevi, Hannover, 1892–1939) Epp. Sel. Epistulae Selectae in usum scholarum, 5 vols. (Hannover, 1887–91) Fontes Fontes iuris germanici antiqui in usum scholarum ex Monumentis Germaniae Historicis separatim editi, 13 vols. (Hannover, 1909–86) Leges nat. germ. Leges nationum germanicarum; ed. K. Zeumer (Lex visigothorum); L. R. de Salis (Leges burgundionum); F. Beyerle and R. Buchner (Lex ribuaria); K. A. Eckhardt (Pactus legis salicae and Lex salica); E. von Schwind (Lex baiwariorum); 6 vols. in 11 parts (Hannover, 1892–1969) Poetae Poetae latini aevi carolini, ed. E. Dummler, L. Traube, P. von Winterfeld and K. Strecker, 4 vols. (Hannover, 1881–99) SRG Scriptores rerum germanicarum in usum scholarum separatim editi, 63 vols. (Hannover, 1871–1987) SRL Scriptores regum langobardicarum et italicarum saec. VI–IX, ed. G. Waitz (Hannover, 1898) SRM Scriptores rerum merovingicarum, ed. B. Krusch and W. Levison, 7 vols. (Hannover, 1920) SS Scriptores in folio, 30 vols., Hannover (1826–1924) MIÖG Mitteilungen des Instituts für Österreichische Geschichtsforschung MS manuscript ÖNB Österreichische Nationalbibliothek PL Patrologiae cursus completus, series latina, ed. J.-P. Migne, 221 vols., Paris (1841–64) xvi List of abbreviations RB Revue Bénédictine Reynolds, Texts and transmission Reynolds, L. D. (ed.), Texts and transmission. A survey of the Latin classics (Oxford, 1983) Scharer and Scheibelreiter, Historiographie A. Scharer and G. Scheibelreiter (eds.), Historiographie im frühen Mittelalter (Munich and Vienna, 1994) Settimane Settimane di Studio del Centro italiano di studi sull’ alto medioevo (Spoleto, 1954–) TRHS Transactions of the Royal Historical Society 1 Introduction: History and memory in the Carolingian world History and memory in the Carolingian world, the title of both this book and this chapter, is at once a reflection of the current interest in the ways by which various medieval societies constructed and understood their pasts and an acknowledgement of the degree to which memory has become a much-explored and much-theorised topic. The book’s principal themes are the writing and reading of history in the early middle ages, with a pri- mary focus on the remarkable manifestations of historical writing in the Frankish kingdoms in the eighth and ninth centuries. Within this frame- work I consider what is meant by history books, and the Franks’ choice of historical texts, whether of Roman, Christian, ‘barbarian’ or Frankish history: where they appear and where, when and for whom they were made. Further questions concern the readership of these history books and how far the physical characteristics of the Carolingian manuscripts in which the texts survive reveal anything of what contemporaries may have thought about these texts and their wider cultural context. Historians of western, middle-eastern and oriental history have looked at the way a common past could inform what Eggert and Patzold in 1994, and in relation to Saxony in the early middle ages, called ‘Wir-Gefühl’, that is, a sense of ‘us-ness’.1 It has become a commonplace that ideas about the past could define societies and that the present plays a crucial role in moulding understanding of the past.2 The focus of the study of medieval historical writing in particular, moreover, has shifted in recent 1 W. Eggert and B. Patzold, Wir-Gefühl und Regnum Saxonum bei frühmittelalterlichen Geschichtsschreibern (Berlin, 1984). 2 I draw in part here on M. Innes, ‘Introduction’, in Y. Hen and M. Innes (eds.), The uses of the past in the early middle ages (Cambridge 2000), pp. 1–8, but I have in mind also the early Islamic historiographical tradition: see A. al-D. al-Duri, The rise of historical writing among the Arabs, ed. and trans. L. Conrad (Princeton, 1983); A. Noth, The early Arabic historical tradition: a source critical study, 2nd edn in collaboration with L. Conrad, trans. M. Bonner (Princeton, 1994); F. Donner, Narratives of Islamic origins. The beginnings of Islamic historical writing, Studies in Late Antiquity and Early Islam 14 (Princeton, 1998); C. F. Robinson, Islamic historiography (Cambridge, 2003); B. Lewis, History – remembered, recovered, invented (Princeton, 1975). On government-sponsored histories and the work of the T’ang ‘historiographical office’ in the eighth and ninth centuries in China 1 2 History and Memory in the Carolingian World years away from a preoccupation with sources of information and tex- tual affiliation. Now historical narratives are studied both as constructed texts and bearers of memory which were targeted for particular audiences, and as an important element in the promotion of the political culture and identity of particular groups. Studies such as those assembled by Guenée, Magdalino, Scharer and Scheibelreiter, or Hen and Innes,3 to cite only four among the best and most recent, have focussed above all on percep- tions and uses of the past. All take the so-called ‘linguistic turn’ and post-modernism into ac- count to a greater or lesser extent, but have resisted the reduction of historiography to literary history.4 Examination of authorial intention, of the audiences for history, and of manuscript traditions of the surviving texts to try to determine a text’s meaning in context offer what Geary has described as ‘escape routes’ out of the ‘prison house of language’ and sterile notions of intertextuality.5 But too zealous a pursuit of the escape routes threatens to undermine the potential value of historical narratives as representations of a contemporary memory of reality, as accounts of events and people and attitudes, or as powerful combinations of both ob- jective and subjective interpretations of the past, however difficult these may be for modern historians either to reconstruct or to distinguish. It is in this respect that memory and the imperative to record what is remembered have been invoked as well. Here the work on both the mechanisms and the rituals of remembering in the middle ages by such scholars as Carruthers, Coleman, Geary, Geuenich, Oexle, Schmid and Treitler, quite apart from the contributions of the anthropologists, have greatly enlarged our understanding of the sheer capacity for remembering as well as the creative forms mnemonic devices could take in the middle ages.6 see W. G. Beasley and E. B. Pulleybank, Historians of China and Japan (Oxford, 1969); D. McMullen, State and scholars in T’ang China (Cambridge, 1988); and D. Twitchett, The writing of official history under the T’ang (Cambridge, 1992). 3 B. Guenée, Le Métier de l’historien au moyen âge (Paris, 1977); P. Magdalino (ed.), The perception of the past in twelfth-century Europe (London, 1992); A. Scharer and G. Scheibelreiter (eds.), Historiographie im frühen Mittelalter (Vienna, 1994); and Hen and Innes (eds.), The uses of the past. 4 For a concise discussion of the ‘linguistic turn’ and its implications for the study of histori- cal writing see W. Pohl, ‘History in fragments: Montecassino’s politics of memory’, EME 10 (2001), pp. 343–74, especially pp. 343–54; and W. Pohl, Werkstätte der Erinnerung: Montecassino und die Gestaltung der langobardischen Vergangenheit (Vienna, 2001). 5 P. Geary, ‘Frühmittelalterliche Historiographie. Zusammenfassung’, in Scharer and Scheibelreiter, Historiographie, pp. 539–42. 6 M. Carruthers, The book of memory. A study of memory in medieval culture (Cambridge, 1990); J. Coleman, Ancient and medieval memories. Studies in the reconstruction of the past (Cambridge, 1992); P. Geary, Phantoms of remembrance. Memory and oblivion at the end History and memory in the Carolingian world 3 If historical narratives are statements about what people remember of the past as well as what they choose to forget, then the degree to which texts reflect collective memories needs to be further explored. Halbwachs’s notion of the part shared memory plays in the self-definition of a social group has been wholeheartedly accepted by most historians and underpins Fentress and Wickham’s rich account of what people in the past have done with respect to social memory.7 I myself have taken the notion of ‘shared memory’ in the chapters of this book as something established by communication, whether oral or written. I have consid- ered how recalled past experience and shared images of the past are the kinds of memories that have particular importance for the constitution of social groups.8 I have also looked at the ‘construction of the past’, that is, the creation of accounts of past events that drew on memory but selected from it in distinctive ways that became accepted and thereafter shared by a group.9 Nevertheless, the notion of ‘shared memory’ can only really be useful if tested, first of all, against the articulated memories of a specific group. Secondly, we should confront the problematic issues and method- ological difficulties it raises, not least how we might be able to document how, or even whether, such accounts were indeed disseminated, known and accepted. It is with these issues in particular that this introductory chapter is concerned. Because we are bound to concentrate on the surviving written evidence for the memory-keeping and historical composition of any group in the past, it may be helpful to look at an extract from the royal Frankish annals, first written at the end of the eighth century, in order to identify some of the main questions and methodological difficulties to which I have alluded. In the entry for the year 788 which recounts the downfall of Tassilo, duke of Bavaria, and the annexation of his realm by his cousin Charlemagne, the annalist tells us that Charlemagne called an assembly of the first millennium (Princeton, 1994); and idem, ‘Land, language and memory in Europe, 700–1100’, TRHS sixth series 9 (1999), pp. 169–84; K. Schmid and J. Wollasch (eds.), Memoria. Der geschichtliche Zeugniswert des liturgischen Gedenkens im Mittelalter, Münstersche Mittelalter-Schriften 48 (Munich, 1984); D. Geuenich and O.-G. Oexle (eds.), Memoria in der Gesellschaft des Mittelalters, Veröffentlichungen des Max Planck In- stituts für Geschichte 111 (Göttingen, 1994); and L. Treitler, ‘Homer and Gregory. The transmission of epic poetry and plainchant’, The Musical Quarterly 60 (1974), pp. 333– 72. For discussions of the role of gender see E. van Houts (ed.), Medieval memories: men, women and their past 700–1300 (London, 2001). See also M. Innes. ‘Memory, orality and literacy in an early medieval society’, Past and Present 158 (1998), pp. 3–36. 7 J. Fentress and C. Wickham, Social memory (Oxford, 1992); and see M. Halbwachs, Les Cadres sociaux de la mémoire (Paris, 1925) and La Mémoire collective (Paris 1950). 8 See especially chapters 7 and 8. 9 See especially chapters 3, 4, 5 and 6. 4 History and Memory in the Carolingian World at Ingelheim (near Mainz on the Rhine) and that Tassilo came there with his fideles. Bavarian fideles began to say that since giving his son and other hostages and taking oaths, Tassilo, incited by his wife Liutperga, had not maintained his faith (to Charlemagne) inviolate but had been seen to betray it. And Tassilo could not deny this but confessed that since then he had sent messengers to the Avars, had urged the vassals of Charlemagne to join him and had plotted their deaths. He has also ordered his homines to make mental reservations when they were swearing oaths and to swear deceitfully. What is more he confessed to having said that even if he had ten sons he would rather lose every one of them than accept that the agreements should remain as they were and that he would be better dead than living thus. And after he had been convicted of all these things, the Franks and the Bavarians, Lombards and Saxons, and those from every province gathered at that assembly, remembering his earlier evil deeds and how he had even deserted the lord king Pippin on campaign – which is called harisliz in German – saw fit to condemn him to death.10 (My emphasis.) My interest in this introductory chapter is not so much in the vivid picture of Tassilo’s despair and hopeless defiance, nor how the narrative works as an account of the triumph of Charlemagne and of justice over the hapless Tassilo, for these have been fully explored by Matthias Becher and by Stuart Airlie.11 What I wish to highlight are the implications of the phrase reminiscentes priorum malorum eius about remembering Tassilo’s evil deeds and an earlier desertion of Pippin. The Royal Frankish annals is the only text to refer to this remembrance on the part of the assembly. The revised version of the Royal Frankish annals, produced in the early years of the ninth century omits this state- ment.12 Other texts drawing on the Royal Frankish annals, but writing later into the ninth century, such as the annals of Fulda, also omitted it. Only the more nearly contemporary ‘Lorsch annal’ entry for 788 refers in a more general way to how the assembly recalled all the wicked deeds Tassilo had done.13 The allusion in the Royal Frankish annals is to the desertion from Pippin’s army in 763. Both Airlie and Becher commented on the 10 Annales regni francorum, ed. F. Kurze, MGH SRG, VI (Hannover, 1895), p. 80; English translation P. D. King, Charlemagne. Translated sources (Kendal, 1987), p. 86. 11 See S. Airlie, ‘Narratives of triumph and rituals of submission: Charlemagne’s mastering of Bavaria’, TRHS sixth series 9 (1999), pp. 93–120; and M. Becher, Eid und Herrschaft. Untersuchungen zum Herrscherethos Karls des Großen, Vorträge und Forschungen Sonder- band 39 (Sigmaringen, 1993). 12 Annales Einhardi, ed. F. Kurze, MGH SRG, VI (Hannover, 1895), p. 81. 13 Annales laureshamenses, ed. G. Pertz, MGH SS, I (Hannover, 1826), pp. 22–39, at p. 33. See also the facsimile edition of ÖNB cod. 515: F. Unterkircher (ed.), Das Wiener Frag- ment der Lorscher Annalen. Christus und die Samariterin. Katechese des Niceta von Reme- siana. Codex Vindobonensis 515 der Österreichischen Nationalbibliothek Facsimile Ausgabe, Codices Selecti 15 (Graz, 1967). See also below, pp. 104–10. History and memory in the Carolingian world 5 unlikelihood of either the Lombards or Saxons remembering something that had happened in 763, for neither people was even under Frankish rule, let alone part of the army. Neither Airlie nor Becher mentioned the possibility, however, that this memory was an allusion to something recorded in the annals themselves. Indeed, it is the annals which present the damning account of Tassilo’s relations with his uncle Pippin III and Charlemagne, how he was given Bavaria to rule in 748, swore oaths of fidelity to Pippin III in 757, deserted the Frankish army in 763, renewed his oath of fidelity to Pippin’s son Charlemagne in 781 and was forced to acknowledge his subordinate status in 787 with the handing over of hostages, including his own son. The following questions emerge, therefore. In what sense was the mem- ory described in the annals one that was part of collective memory? Who is included in the notion of ‘collective memory’? Is the memory based on knowledge of the annals or a real memory among those at Ingelheim in 788 of disgraceful behaviour in 763? Does the allusion then support an early date for the composition of the annals? Is the 788 entry perhaps a witness to the expectation that this text would be in circulation and form the basis of subsequent knowledge? Is this why the Franks, Bavarians, Lombards and Saxons are credited with retrospective knowledge in this way? Was their knowledge formed by this particular piece of historical writing? The entry also underlines one of the fundamental difficulties in charting memory, namely that we are bound to do it from the evidence of surviving written texts and that these in their turn raise the problem of how we can determine the impact and influence of such an historical text. To suggest knowledge of the contents of a text is to make assump- tions about the process of production, methods of circulation, speed of reception and the impact of the text itself, all of which must be tested. Thus we also need to consider the relevance of literacy to the extension and record of memory, and the degree to which literacy, and thus the recoverable indicators of memory, are the preserve of an elite. It is easy to label literacy as the ‘preserve of an elite’ in the early middle ages but it is much harder to define it. It has, after all, become a common- place, voiced by me as much as by everyone else, that the great majority of our sources are primarily those of the ‘social elites’ of early medieval Europe. These ‘social elites’ might be defined as the groups in society who had power or some measurable superiority in some sphere over other in- dividuals or groups of people.14 But what we all actually mean by saying that should be challenged. I have also maintained, for example, that the local charter evidence from centres all over western Europe and dating 14 See R. Le Jan (ed.), La Royauté et les élites dans l’Europe carolingienne (du début du IXe siècle aux environs de 920), Centre d’Histoire de l’Europe du Nord-Ouest 17 (Lille, 1998). 6 History and Memory in the Carolingian World from the period from the seventh to the tenth centuries is an indication of the exploitation of literate modes by landowners down a social hierarchy that may have included peasant farmers and small-scale landowners.15 Since 1989 a number of studies have explored the degree to which local lay communities used the written word for their legal transactions and played some role in their production or preservation in the early middle ages.16 Even so, how great a proportion of the ‘peasant’ population, esti- mated at 90 per cent of the total in this period, had any regular access or customary familiarity with the written word? I stress the words ‘regular’ and ‘customary’, for we should also re- member that this is a world in which even a freed slave is known by the word cartularius, that is, ‘charter man’.17 It is a world in which re- ligions of the book, Christianity, Islam or Judaism, predominate. In the Carolingian empire, above all, there is an insistence from the second half of the eighth century onwards on the central role of texts for the consol- idation and harmonization of the Christian religious faith and practice, for the transmission of knowledge, and for the exercise of justice and government. In the course of the ninth century, even a written musical notation is developed to provide a written supplement for the transmis- sion of melodies for the liturgical chant, hitherto passed on by cantors.18 Certainly, the possession of the skills of writing may have been the pre- serve of the specially trained. Similarly, those who could read complex material, as distinct from those who could simply recognise the letters of the alphabet, may have formed an intellectual elite. This intellectual elite, owing to the limited opportunities for education, could also have been a social elite, though it is important to remember that Ebbo, archbishop 15 McKitterick, Carolingians and the written word, pp. 77–134, in relation to the evidence of the St Gallen material. 16 For discussions of Lorsch and Bavaria see M. Innes, State and society in the early middle ages 400–1000 (Cambridge, 2000); and W. Brown, Unjust seizure: conflict, interest and authority in an early medieval society (Ithaca, 2001). For Italy see N. Everett, Literacy in Lombard Italy, c. 568–774 (Cambridge, 2003). See also R. Schieffer, Schriftkultur und Reichsver- waltung unter den Karolingern, Abhandlungen der Nordrhein-Westfälischen Akademie der Wissenschaften (Opladen, 1996); and the studies emanating from the Utrecht Uni- versity Centre for Medieval Studies Pionier Project on medieval communications: Marco Mostert (ed.), New approaches to medieval communication, Utrecht Studies in Medieval Literacy 1 (Turnhout, 1999); and K. Heidecker (ed.), Charters and the use of the written word in medieval society, Utrecht Studies in Medieval Literacy 5 (Turnhout, 2000). The new early medieval charter study project which younger scholars such as Warren Brown, Marios Costambeys, Adam Kosto and Matthew Innes are coordinating is likely to bring more valuable material to the fore. 17 A point stressed by J. L. Nelson, ‘Literacy in Carolingian government’, in McKitterick (ed.), Uses of literacy, pp. 258–96. 18 For the background see S. Rankin, ‘Carolingian music’, in R. McKitterick (ed.), Car- olingian culture: emulation and innovation (Cambridge, 1994), pp. 274–316. History and memory in the Carolingian world 7 of Rheims, among others, was of humble origin and thus the extent to which the acquisition and possession of literate skills may have enabled an individual to cross other social barriers.19 If so many men and women in Carolingian Europe had some kind of access to written culture, however restricted in scope, right the way down the social scale, and if literacy could offer a passport into an elite group, then of what value is the concept of an elite at all? Should we rather regard access to a written culture as defining an elite? Did the possession of literate skills make one eligible to be or to become a member of the elite in the Carolingian world? To focus exclusively on levels of literacy in any social or political group defined according to criteria which do not include the criterion of literacy, therefore, is to miss the point. We should focus instead on the extent to which the exploitation of written culture provided the means for contemporaries at the time to define themselves as an elite. What really matters, therefore, is what men and women in the Carolingian world hoped to achieve by exploiting the written word and why they chose that medium of communication. A more productive approach might be to look at the problem of the elites and written culture in the Carolingian world from the opposite direction. I propose to consider the idea of elites in relation to written culture as a phenomenon reflected in and thus defined by their use of texts, and especially historical texts. In particular I should like to pursue the possibility for the Carolingian period that there was an elite, lent cohesion by their particular use of the written word to provide that elite with a recorded memory, and thus identity, that would transcend other political or social divisions. If we consider how the past was understood by the Franks in the Carolingian period, that is, in the period from the eighth to the tenth centuries, and how a group placed itself in relation to that past, we may be able to see how that group becomes defined as an elite with special characteristics, and how the written texts act as an enabling mechanism in the expression of an elite memory and identity. In other words, the study of history and memory in the Carolingian world is a study both of the texts in which an elite defines itself and of the extant Carolingian manuscripts which provide the indications of the Franks’ understanding of the past. The understanding of history and the past in the Frankish kingdom embraced Roman, Christian and ‘Germanic’ (Lombard, Anglo-Saxon, Gothic and Frankish) history. Records of its past included historical 19 On Ebbo see S. Airlie, ‘Bonds of power and bonds of association in the court circle of Louis the Pious’, in P. Godman and R. Collins (eds.), Charlemagne’s heir: new perspectives on the reign of Louis the Pious (814–840) (Oxford, 1990), pp. 191–204. 8 History and Memory in the Carolingian World chronological syntheses and narratives, and records of other kinds.20 From the work recorded elsewhere in this book as well as still in progress, it seems clear that there was a concerted effort to acquire and copy his- tory books throughout the Carolingian realm from the late eighth century onwards. There is also the well-known phenomenon of the remarkable number of contemporary histories written in the late eighth and the ninth centuries.21 These two phenomena, that is, the copying of older history books and the writing of new history, are arguably related and interde- pendent. The possibility of a connection between the copying of older history, primarily in monastic scriptoria, and the composing of new his- tory, incidentally also raises the issue of the distinctiveness of the role of the monastery itself in Carolingian record keeping and history writing, and the role monasteries may have played in the formation of Frankish perceptions of the past, but I shall have to leave detailed consideration of these to another occasion. I have set out the evidence about the history books and history writ- ing of the Carolingian world in the remaining chapters of this book. It is clear that, for the Franks, an understanding of the past worked at sev- eral levels and was manifested to them in a number of different textual contexts. What these Carolingian history books reveal is the interplay be- tween memory, forms of historical record and the writing of history. It is this interplay which is an essential component of the process of defining an elite and a people. The books read and produced in Frankish cen- tres indicate the formation of a sense of the past – biblical, Roman and Christian – to which the Franks collectively belonged and which they had inherited. In their own history writing the Franks also show an impulse to forge an identity that explicitly placed the origin of the Franks in a far distant Roman and Trojan past. It is to the forging of this identity by means of the creation of a common memory in the form of a distinctive narrative that I now turn. There is only space to do so with specific reference to one example, the Liber historiae francorum, and the implications of one ninth-century copy of it now in Paris. With the Liber historiae francorum in this codex is the text of the Annales regni francorum. The association of complementary accounts of Frankish history and its implications will form the final section of this introductory chapter. 20 For fuller discussion see below, chapters 7, 9 and 12. 21 See Innes and McKitterick, ‘Writing of history’; and for the general context see A. Momigliano, ‘Pagan and Christian historiography in the fourth century A.D.’, in A. Momigliano (ed.), The conflict between paganism and Christianity in the fourth century (London, 1962), pp. 79–99. History and memory in the Carolingian world 9 Frankish historical writing and the Liber historiae francorum Although it provides evidence of the process of the reception of a com- plex cultural heritage, the interest in the Judaeo-Greco-Roman past needs to be used and developed in order to constitute evidence of the defini- tion of an elite identity and the enhancement of historical memory. The Liber historiae francorum was one of the new Carolingian historical works produced in the eighth century and copied extensively in the ninth cen- tury. It constructs a specific past for a particular group of people. First of all, it provides them with a group memory and identity. Secondly, it places them, both culturally and historically, within the wider history of the Roman empire and Christian Roman Gaul. The raison d’être of this text, indeed, could be described as the definition of a people by means of its history. The Liber historiae francorum is customarily dated 727 due to its reference at the end to the sixth year of Theudebert (IV). From the outset the text insists that it is about both kings and the people of the Franks: let us present the beginnings of the kings of the Franks, the ori- gins and deeds of the kings and those peoples (Principium regum francorum eorumque origine vel gentium illarum ac gesta proferamus). The Liber historiae francorum is a short but remarkable history which has suffered unreasonable neglect. More crucially it has been underes- timated as a piece of historical writing largely as a result of the way it is printed and presented in modern editions. Most of its earlier sections (apart from the first four chapters) are held to be so derivative from the sixth-century Gallo-Roman author Gregory of Tours as not to be worth mentioning. Attention has thus focussed more or less exclusively on the last eleven chapters, 43–53 (in which the Liber historiae francorum au- thor writes a completely independent account of the seventh and early eighth centuries) without a consideration of how these chapters fit into the structure and message of the text as a whole. Dependence on Gregory, however, is a misleading way in which to un- derstand the Liber historiae francorum author’s use of the earlier text. It would be far better to describe the process of composition as judicious selection, with some highly significant changes. Further, there are sub- stantial additions. The changes and additions are most notable at the beginning of his work. Certainly the judicious use of Gregory is impor- tant, but more as a witness to Gregory’s high status as a history book than as a symptom of derivative and impoverished history writers in the eighth century. The many small changes and larger insertions are even more significant, however. These, as we shall see, alter the emphasis in crucial ways which are entirely consistent with the opening chapters. 10 History and Memory in the Carolingian World Above all, one needs to read the whole narrative as it is presented in the early manuscripts of the Liber historiae francorum. Although the whole text is undoubtedly printed in Krusch, even he diminished the impact of the text by presenting it visually as a text full of borrowings signalled in a much smaller typeface.22 As a result he confuses the purpose of the original manuscripts. In the manuscripts, the Liber historiae francorum demands to be read on its own terms, for no such distinction is made. It should not be seen in terms merely of what is borrowed or new, but as a complete text with very distinctive emphases of its own. The themes of war and kingship and the elaborate account of the marriages and role of queens, of treasure, of devotion to particular saints and particular churches in Paris, are all reiterated throughout the text. The Liber historiae francorum rejected Gregory’s emphasis at the be- ginning of the work and Gregory’s picture of biblical Franks as the new chosen people, with its emphatically Judaeo-Christian chronology and framework.23 Instead, he or she (and it may well be the latter) provides a spirited alternative view of the Franks and their origins.24 Thus the Franks’ superiority in relation to other barbarian groups (Alans, Huns, Burgundians) is stressed and illustrated with the story of their origins. The text begins with a statement about the origins of the Franks. It lo- cates them to Troy and thereafter to the refuge a group of Trojans found north-west of the Black Sea. They thus have historical, rather than mythi- cal, origins. These historical origins are rooted in a past linked with Rome because of the association with Trojan origins. The author contrives nev- ertheless to convey a sense of Frankish superiority even over the early Romans.25 Thus Aeneas, who is provided with the significantly pejorative 22 Liber historiae francorum, ed. B. Krusch, MGH SRM, II (Hannover, 1888), pp. 241–328. 23 For a useful context see M. de Jong (ed.), The power of the word: the influence of the Bible on early medieval politics, special issue, EME 7 (1998), pp. 261–357. Compare M. Garrison, ‘The Franks as the new Israel? Education for an identity from Pippin to Charlemagne’, in Hen and Innes (eds.), The uses of the past, pp. 114–61. 24 R. Gerberding, The rise of the Carolingians and the Liber historiae francorum (Oxford, 1987), pp. 150–9, locates the author to Soissons but dismisses the notion that the text could have been produced at Notre-Dame, ‘simply because it was a nunnery’! For an alternative view see J. L. Nelson, ‘Gender and genre in women historians of the early middle ages’, in Nelson, The Frankish World 750–900 (London, 1996), pp. 183–97; and R. McKitterick, ‘Frauen und Schriftlichkeit im Frühmittelalter’, in H.-G. Goetz (ed.), Weibliche Lebensgestaltung im frühen Mittelalter (Cologne, Weimar and Vienna, 1991), pp. 65–118, and revised English version ‘Women and literacy in the early middle ages’, in McKitterick, Books, scribes and learning, chapter 13, pp. 1–43. 25 See M. Innes, ‘Teutons or Trojans? The Carolingians and the Germanic past’ in Hen and Innes (eds.), The uses of the past, pp. 227–49; F. Graus, ‘Troja und trojanische Herkunftssage im Mittelalter’, in W. Erzgraber (ed.), Kontinuität und Transformation der Antike im Mittelalter (Sigmaringen, 1989), pp. 25–43; I. N. Wood, ‘Defining the Franks: Frankish origins in early medieval historiography’, in S. Forde, L. Johnson and A. V. History and memory in the Carolingian world 11 epithet of tyrant (tyrannus), fled to Italy but other Trojan leaders, Priam and Antenor, fled to the Moaetian marshes on the Don river and built the city Sicambra. It is from this branch, by implication more worthy of praise, that the Franks’ history proceeds. As late as the fourth century their descendants are still called Trojans and have become allies of Rome to drive the Alans out of the Moaetian marshes (c. 2). Valentinian, the Roman Christian emperor, gives them the name of Franks (because they were hard-hearted and bold) but he taxed them once their ten-year agree- ment had lapsed and they thereupon revolted, with the declaration of the wish to be free. The Franks then were defeated by the Romans and fled to the Germanies (note that the phrase reflects the designation of Roman provinces) and lived under Marcomir, son of Priam, and Sumo, son of Antenor. It is a weird chronology to be sure, but the emphases are rather interesting; they provide an alternative Roman connection and one in which Franks, although often driven hard and sometimes behaving badly, succeed in maintaining their independence. It is that very independence and identity which is consolidated in the following chapters, in a variety of ways and not always with consistent approbation. In the early stages of their history, for example, the Roman emperor had kept his side of the bargain, namely, to remit taxes to the Franks if they assisted him against the Alans. As already remarked, the Franks rebelled when asked to pay taxes after those ten years had elapsed. Further on in the narrative, it is not just Fredegund who is criticized (though the author clearly also appreciates her wiliness, especially the godmother trick played on Queen Audovera and the successful plan to deceive the enemy army with a precursor of Birnam wood coming to Dunsinane);26 some of the activities of the Merovingian kings are also regretted. A full analysis of the whole import of the narrative of the Liber historiae francorum, especially for the portrayal of Merovingian kingship, is not possible here. For the moment, and in the context of my theme, I wish to stress a few points only. These relate most closely to the cultural identity of the Franks. First of all, in chapter 4, there is a brief account of the raising up of Faramund, Marcomir’s son, as a rex crinitus. On the death of Faramund the Franks raised up Chlodio, his son, also with long hair, and the author claims that from this time they began to have long-haired Murray (eds.), Concepts of national history in the middle ages (Leeds, 1995), pp. 47–59; and compare J. Barlow, ‘Gregory of Tours and the myth of the Trojan origin of the Franks’, Frühmittelalterliche Studien 29 (1995), pp. 86–95. 26 Liber historiae francorum, cc. 31 and 36, pp. 293 and 305; compare William Shakespeare, Macbeth, Act V, Scene 5. 12 History and Memory in the Carolingian World kings. We should register the significance of this being insisted upon in 727 as if to reinforce the legitimacy of the Merovingian line at a time when their actual power was beginning to be overshadowed by that of the Carolingian mayors of the palace. Throughout the text, moreover, there is a consistent acceptance of the importance of the election to the kingship and maintenance of a line of succession, most usually from father to son. The structure of the history reinforces this, in that each chapter tends to start with the death of another king by way of introducing the next phase of the story. Secondly, it is during the reign of Faramund that the author chooses to locate the introduction of laws, expounded by four prominent men, Wisowastus, Wisogastus, Arogastus and Salegastus, who lived in villas beyond the Rhine called Bethagm, Salechagm and Widechagm. These laws, although those of the Franks and expounded by four leading men, were by implication supported and acknowledged by the kings. The same claim for the introduction of laws to the Franks is of course also made in the prologue to the law of the Franks known as the Pactus legis salicae in some of its redactions (though not the earliest) and embroidered further in the Lex salica karolina.27 This portion of the story is indicative first of all of the close connection between historical and legal texts in general. Secondly it uncovers at least a tantalizing possibility of a link, involv- ing the Carolingian royal court, between the author of the Liber historiae francorum, or a copy of his/her text, and the redactor of the Lex salica pro- logue. The conventional understanding is that this prologue was added to the Lex salica at some stage in the sixth century, despite the fact that it does not appear in the earliest (A, B) redactions but only in Lex salica manuscript ‘C6’ (Paris, BnF lat. 18237 from western France, s.IX/2), and various copies of the D, E and K versions, the earliest of which are the three ‘E’ redactions attributed to the leges scriptorium associated with the court of Louis the Pious and usually dated to the first quarter of the ninth century.28 Quite apart from the difficulties the manuscript evidence presents for our understanding of the relationship between the various redactions of the Lex salica, the earliest manuscripts of the Liber historiae francorum also survive only from the late eighth and early ninth centuries. Thus the precise relationship or direction of influence between the Lex salica prologue and the Liber historiae francorum’s suggestion about the origin of the laws is difficult to determine. It is certainly very possible that the history inspired the author of the prologue rather than the other 27 Pactus legis salicae and Lex salica, ed. K. A. Eckhardt, MGH Leges nat. germ. IV, 1 (Hannover, 1962), p. 2, and IV, 2 (Hannover, 1969), pp. 4 and 5. 28 R. McKitterick, ‘Zur Herstellung von Kapitularien: die Arbeit des Leges-Skriptoriums’, MIÖG 101 (1993), pp. 3–16. History and memory in the Carolingian world 13 way about. Had the story been known to Gregory or to pseudo-Fredegar, would they not have used it? It is also notable that a geography still conceived within the framework of Roman imperial administration is maintained. The author is anxious to stress that King Chlodio, the first real long-haired king, was based in the Roman province of Germania, that the Romans were on the west of the Rhine as far south as the Loire, that the Goths occupied the re- gion beyond the Loire and that the Burgundians were clustered around Lyon. Chlodio subsequently settled in the region between Cambrai and the Somme river. Chlodio was there succeeded by his son Merovech, from whom the Merovingian line took its name. In keeping with the geo- graphical ideas incorporated into many early medieval histories, recently studied by Andrew Merrills, the Liber historiae francorum maintains a no- table insistence on political geography throughout the work.29 Finally, there is a signal prominence given to the role of the queens, especially Clothild (Chrotildis), wife of Clovis, first king of the Franks. Indeed, a veritable gesta Clothildis is inserted into the Liber historiae francorum. The gesta of Clothild occupy page after page in the manuscript copies. There are details of Clovis’ courtship of Clothild, her marriage, her advice to rulers, the baptism of the king, the baptism of her children, her endowment of churches, her management of her sons and thus of the succession, and an insistence on pious devotion to the church and the special royal association with the churches of St Victor, St Geneviève and St Peter in Paris. Clothild also becomes a role model for subsequent queens in the narrative, with the marriage, sons, death and burial of queens faithfully recorded. The queen as a source of legitimacy for the royal line and the queen as fount of Christian piety are thus twin preoccu- pations of the author. This is cleverly reinforced (by dramatic repetition of phrasing as well as the actual content of the text) by Queen Clothild’s unequivocal association with the baptism of the king, of the army, of his sisters and of the whole Frankish people. The manuscript transmission: the example of Paris, BnF lat. 10911 The copying of a text is one important indicator of its reception as well as for the process of the dissemination of the particular representation of a group. Some indication of the dissemination of copies of the Liber historiae francorum in the early Carolingian empire can be gained from the manuscript transmission. These codices provide some hint, indeed, 29 A. Merrills, History and geography in late antiquity (Cambridge, 2005). 14 History and Memory in the Carolingian World that the Carolingians, if not actually promoting the text, certainly ap- pear to have associated themselves with it and thus conceived their own history in direct relation to the story it contained. The extent to which the Liber historiae francorum provides an alternative, or even preferred, view of the Frankish past and of Frankish status to the ones presented in other contemporary Frankish histories needs to be addressed more fully than can be done here. I have remarked elsewhere, moreover, that so far insufficient attention has been paid to the Carolingian edition of Frede- gar,30 though Helmut Reimitz is now considering the implications of the Carolingian transmission of the six-book version of Gregory of Tours’ Historiae.31 Out of thirty-two extant manuscripts of the Liber historiae francorum, fifteen date from the later eighth, the ninth or the early tenth centuries. All come from the westerly and Rhineland areas of the Carolingian empire. A number of these contain the Liber historiae francorum text only, though in at least two instances the manuscripts would appear to be incomplete. The five codices in which the Liber historiae francorum text is combined with other Frankish histories, such as the Vita Karoli or with the Annales regni francorum or other Frankish annals, in particular, merit further at- tention. They need, furthermore, to be set beside the manuscript trans- mission of Gregory of Tours on the one hand and of Fredegar on the other, quite apart from the massive documentation of the manuscript transmission of Einhard’s Vita Karoli,32 and the dissemination of the Annales regni francorum in composite Frankish history books. The Lorsch 30 See chapter 2 below and the preliminary study by R. Collins, Fredegar, Authors of the Middle Ages and Religious Writers of the Latin West IV, no. 13 (Aldershot, 1996), who is preparing a new edition of Fredegar’s Chronicle. 31 Helmut Reimitz, ‘Social networks and identities in Frankish historiography. New aspects of the textual history of Gregory of Tours’ Historiae’, in R. Corradini, M. Diesenberger, and H. Reimitz (eds.), The construction of communities in the early middle ages: texts, resources and artefacts, The Transformation of the Roman World 12 (Leiden, 2003), pp. 229–268 and plates 1–3; and H. Reimitz, ‘Der Weg zum Königtum in historiographi- schen Kompendien der Karolingerzeit’, in J. Jarnut and M. Becher (eds.), Historiographie und Identität in den fränkischen Regna der Merowinger- und Karolingerzeit (forthcoming). I am very grateful to Helmut Reimitz for discussion of Paris, BnF lat. 10911, his kindness in letting me read his paper ‘Der Weg zum Königtum’ in advance of publi- cation, and stimulating discussions on matters to do with Frankish historiography in general. Paris, BnF lat. 10911 is one of the three codices being prepared in a digital edition by R. Corradini, K. Giesriegl and H. Reimitz, Drei fränkische Geschichtsbücher aus der Karolingerzeit: Paris Bibliothèque nationale lat. 10911, Vienna ÖNB lat. 473, St Petersburg, NLR lat. F.v.IV.4. See also M. Heinzelmann and P. Bourgain, ‘L’Œuvre de Grégoire de Tours: la diffusion des manuscrits’, in N. Gauthier and H. Galinié (eds.), Grégoire de Tours et l’espace gaulois. Actes du congrès international Tours, 2–5 novembre 1994, 13e supplément à la Revue Archéologique du Centre de la France (Tours, 1997), pp. 273–317. 32 See the comprehensive study provided by M. Tischler, Einharts Vita Karoli. Studien zur Entstehung, Überlieferung und Rezeption, 2 vols., MGH Schriften 48 (Hannover, 2001). History and memory in the Carolingian world 15 fragment of the Liber historiae francorum (Paris, BnF lat. 7906) combines it with sections from the Aeneid and Dares Phrygius and thus concentrates on the origin of the Franks and the wanderings of those who survived the Trojan war. This suggests that Virgil’s Aeneid was also a history book for the Franks.33 All these constitute important evidence for the exploitation of the writ- ten word, both to promote and to inculcate a particular sense of identity. I should like here to comment briefly, therefore, on only one of the exam- ples of a composite Frankish history book, namely, Paris, BnF lat. 10911. It was written by a single scribe, possibly in Paris, in sloping early caroline minuscule and is probably to be dated to the first third or second quarter of the ninth century.34 The consistency of the presentation of the text as well as the single hand responsible suggests that this is a very particular and deliberate design rather than a mere assemblage of related texts. The manuscript reminds the reader forcefully of the historical context in which this history was written, for it enhances the message of the Liber historiae francorum as a history of Frankish kings, Frankish queens and gens francorum by using it to preface the Annales regni francorum.35 The former text, after all, composed c. 727, was effectively written from the vantage point of the emergence and triumph of Charles Martel and at the point when Dagobert III, raised in the monastery of Chelles, is made king of the Franks in Neustria. To my mind this can with hindsight be seen as a premonition, if nothing more, of the rule of the Carolingian mayors of the palace and the short reign of the last Merovingian king of all. Childeric III was also taken from a monastery, and relegated to a monastery once more on his deposition in 751. Certainly the connection between the Liber historiae francorum and the triumph of the Carolingian mayors and kings is made in the early Carolingian period.36 That Charles Martel is regarded as the real hero of the earlier part of the story in the eyes of the scribe at least is also indicated by the text’s headings suddenly breaking out into red rustic capitals with Charles’s triumph over the Saracens. 33 R. A. Gerberding, ‘Paris, Bibliothèque Nationale latin 7906: an unnoticed very early fragment of the “Liber historiae francorum”, Traditio 43 (1987), pp. 381–6. For further on the Aeneid see McKitterick, Migration. 34 For some comments on this manuscript in relation to Einhard’s text see Tischler, Einharts Vita Karoli, pp. 1156–8. There, however, Tischler makes the palaeographically improbable suggestion that the manuscript is from Fleury and is to be dated to the late ninth century. He also does not consider what its relationship to Bern, Burgerbibliothek 599, definitely from Fleury, might be. 35 On the Annales regni francorum see below, chapters 4, 5 and 6. Compare the independent comments of R. Collins, ‘The “Reviser” revisited: another look at the alternative version of the Annales regni francorum’, in A. C. Murray (ed.), After Rome’s fall: narrators and sources of early medieval history. Essays Presented to Walter Goffart (Toronto, 1998), pp. 191– 213. 36 Gerberding, The rise of the Carolingians. 16 History and Memory in the Carolingian World Incidentally, another Carolingian copy of the Liber historiae francorum in Paris, BnF lat. 5596 (s.IX, possibly from St Germain-des-Prés) rein- forces the perception of the Liber historiae francorum as the pre-history of the Carolingian rulers by highlighting the names with enlarged initials and a line of red capitals or uncials. It treats the emergence of Pippin II and Charles Martel in the same way. In this codex the Liber historiae francorum is combined with the Vita sancti Remigii, which of course also gave prominence to the baptism of Clovis and the role of Clothild. It thus serves to reinforce the message in the Liber historiae francorum. Further, in the text of the latter, the naming of the Franks by the Emperor Valen- tinian, the names of the lawgivers in the reign of Faramund and the career of Clovis are highlighted, as is the emphasis on the ruler in each chapter. Notes, furthermore, have been inserted by a Carolingian reader anxious to get the lines of royal succession (and the names of the royal wives) straight. The manuscript thus contrives to place the history of the Franks as an extension of that of fifth-century Gaul by the simple expedient of visual markers and layout. It suggests that the perspective offered in BnF lat. 10911 was not an isolated one. Krusch’s edition obscures other significant elements of BnF lat. 10911’s arrangement of the text. The Liber historiae francorum text is actually divided into chapters 1–51 and ends with the initial triumph of Charles Martel over Ragamfred and Chilperic. Krusch had created fifty-three chapters by dividing chapter 5 into two parts at Ipse itaque Merovechus genuit filius nomine Childericum. He also divided chapter 48 into two (= 49 and 50). The narrative is then augmented by a rearrange- ment of a selection of chapters from the Continuations to Fredegar’s Chronicle, bringing the story to the death of Charles Martel. Thereafter in Paris, BnF lat. 10911, the Annales regni francorum continue the narra- tive with the first entry recording the death of Charles Martel. The Annales regni francorum 741 restarts at chapter 1 on fol. 56r. This is also the beginning of a new quire, with a large DCCXLI written in red uncial and the first letter of Carlus written in red, as a consciously new text. At DCCL the text pauses. Six lines are left blank and so is the following verso, so that this decade-long period before the assumption of the kingship by the Carolingian family forms a short epoch in itself, recording the competition presented to Pippin by Carloman and Grifo and the emergence of Pippin alone. This section is thus also distinguished visually from the rest by isolating it in the manuscript. Entries for 751 and 752 are absent, and as far as this scribe is concerned they were missing, for space was left for them as well as chapter numbers notionally assigned; the next entry, for 753, is numbered c. 13 and thus cc. 11 and 12 are absent. The text then runs to c. 73 for 813 and the entries for 814 History and memory in the Carolingian world 17 onwards are not numbered. This suggests that one exemplar may have offered the text either only as far as 814 or distinguished it in some way with special headings (as Vienna, ÖNB cod. 473 does).37 This can be set out schematically as follows: LHF I-LI LHF LII = Fred. Cont. 10 LHF LIII = Fred. Cont. 11 (final section), 12, 13 LHF LIIII = Fred. Cont. 14, 15, 16, 17, 18, 19 LHF LV = Fred. Cont. 20–24 (to 741) ARF I-LXXIII = Annales regni francorum 741–813 (that is Charles Martel, Pippin and Charlemagne) ARF 814–829 (Louis the Pious) The scribe distinguished between the sections by use of different scripts and the structure of the book. Not only do the Annales regni francorum start on a new quire but the sections taken from the Continuations of Fredegar’s Chronicle and conflated into a differently structured narra- tive are distinguished by means of capitals and elaborate headings. For chapters 52–4 there are headings in rustic capitals or half uncial, and sum- maries of the chapters. The famous and problematic calculation of years from the Continuations to Fredegar’s Chronicle, chapter 16, starts a new quire (fol. 49r). Originally this quire may have contained three and a half empty folios (fols. 52v, 53–55v) which were only filled in the late twelfth or early thirteenth century with a letter purporting to be from Alexander the Great to Aristotle.38 The Annales regni francorum section, therefore, which starts in the following quire, may originally have proceeded after a substantial gap in the manuscript. Although the sewing of the binding now makes it difficult to be certain, it is possible, however, that the quire containing most of the reorganized sections from the Continuations of Fredegar’s Chronicle in fact originally ended on fol. 52v, forming a quire of four leaves. The twelfth-century scribe may then have used the final blank leaf of the original quire and added three single leaves to accom- modate the rest of his own text. Whatever the case, it is clear that the Carolingian scribe distinguished codicologically between the Liber histo- riae francorum and added sections from the Continuations of Fredegar’s Chronicle on the one hand, and the Annales regni francorum text on the other. Yet he also preserved the sense of continuity by maintaining the same generous margins and spacious layout throughout the manuscript. Further, the Annales regni francorum replicates the structure of the Liber 37 See below, pp. 122 and 215. 38 For Bibliography and context see the entry for Alexander in the Lexikon des Mittelalters I (Stuttgart, 1999), col. 356. 18 History and Memory in the Carolingian World historiae francorum with its emphasis, in the first words of the initial series of chapters, on the ruler or else the main protagonist in each chapter. 1. Carlus maior domus defunctus 2. Quando Carlomannus et Pippinus 3. Tunc Carlomannus et Pippinus 4. Iterum Carlomannus et Pippinus 5. Tunc Carlomannus confessus est Pippino germano suo 6. Tunc Carlomannus Romam perrexit 7. Grifo fugivit in Saxoniam 8. Grifo de Saxonia . . . pervenit 9. Burghardus . . . Fulradus 10. Pippinus secundum morem francorum electus est ad regem The addition of the Annales regni francorum and compression of text from the Continuations of Fredegar to this manuscript also contrives, by this clever juxtapositioning, to present the Liber historiae francorum from the Carolingian perspective. It belittles the succession of Merovingian kings, Chilperic, Chlothar, Theoderic son of Dagobert, mentioned in c. 53 and makes no reference to the last Merovingian king Childeric III at all. The manuscript presentation thus makes the whole story of the ori- gin, kingship, cleverness of queens, law making and Christianity of the Franks into a prelude to the triumph of the Carolingian house, where these themes are taken up and elaborated. It is entirely consistent, for example, that it should be only in this copy of the Annales regni francorum that the deaths of Queen Hildegard and Queen Bertha are noted under 783. The compilation suggests that the Liber historiae francorum was in fact rather greater an inspiration for the royal Frankish annalist than has hitherto been imagined. In effect, therefore, BnF lat. 10911 presents a Carolingian history book which sets the rise and triumph of the Carolingian rulers Charles Martel, Pippin III and Charlemagne in the context of the early emergence of the Franks from a specifically Trojan, Roman and at least geographically germanic past. But it is the Franks who emerge triumphant once they oc- cupy Gaul. The Liber historiae francorum, and especially this Carolingian packaging of it, thus serves as a statement of confidence in the process of Frankish identity. It records, moreover, in its emphases and omissions as well as in the adjustments and additions to Gregory’s text, the kind of reception, selection and rejection of aspects of the Frankish past which were both a part of that process of a formation of identity and a reflec- tion of it. The Liber historiae francorum is an uncompromising statement of the group identity of the Franks and constructs a specific past that places them, both culturally and historically, within the wider history of the Roman empire and Christian Frankish Gaul. The combination of History and memory in the Carolingian world 19 the Liber historiae francorum with the triumphalist Annales regni francorum presents, in effect, a strong definition of an elite to contemporaries in the Carolingian world, and to posterity. So far so good, but we then have to consider the problem of the manuscripts themselves. Who might have read them? How effective was the dissemination of the texts they contained and who was responsible? Thus this codex containing the Liber historiae francorum and the An- nales regni francorum needs to be set within the context of the trans- mission of the latter and all the uncertainties and puzzles that arise in consequence. The major problem with the Annales regni francorum 741–829 is that not one of the surviving manuscripts can be regarded as the original. What we have are copies from the time of Charlemagne’s grandsons. Al- though there are a number of ninth-century copies, none survives from the late eighth century and the earliest copy, Paris, BnF lat. 10911 dis- cussed above, dates from c. 830. The Annales regni francorum purportedly survive in five versions, known as A, B, C, D and E, defined by their late nineteenth-century editor Kurze according to the groups of manuscripts with related portions of the texts (A up to 788, B up to 813, C and D up to 829 but with particular variants: thus D is distinguished from C (and A and B) because it includes the notes about the conspiracies of 785 and 792 against Charlemagne led by Hardrad and Pippin the Hunchback re- spectively. The revised version of the Annales regni francorum, that is, the ‘E’or 5th redaction, was once known as the Annales Einhardi. It actually only revises the text from 741–801. Kurze thought the E version from 802 to 829 was dependent on D. All the (many) manuscripts of E were thought by Kurze to go back to one archetype, ‘Ex’, in which the Annals were linked with the Vita Karoli of Einhard. The oldest of these manuscripts, Vienna, ÖNB lat. 510, is unfortu- nately late tenth century so the problem of the E recension in terms of its original dissemination is even worse than for all the other versions. Kurze tried to work out how the text was constructed, and thought in precisely these terms, of construction rather than composition. For him it was more a copying process and only towards the end of his discussion does he introduce the idea of the ‘original’ and authorship. His stemma is a clear visual presentation of the hypotheses and intermediaries he pro- posed in order to take account of the sequence of apparent borrowing, repetitions, errors in common and the like.39 39 Annales regni francorum, pp. V–XIX; and the fuller study F. Kurze, ‘Ueber die karolingis- chen Reichsannalen von 741–829 und ihre Ueberarbeitung. 1: Die handschriftliche Ueberlieferung’, Neues Archiv 19 (1894), pp. 295–339; and ‘Zur Ueberlieferung der karolingischen Reichsannalen und ihrer Ueberarbeitung’, Neues Archiv 28 (1903), pp. 619–69. 20 History and Memory in the Carolingian World The various recensions Kurze proposed in their turn have led to hy- potheses about the sequence of composition. Kurze interpreted the dates at which particular recensions stopped as an indication of original batches of annals – one to 788, another taking the story to 813, later continued to 829. The manuscripts themselves certainly reflect texts truncated in this way, but whether this is as part of a larger design in the compilation of historical miscellanies, or due to the original circulation of the text in a number of recensions to which other sections were gradually added, is difficult to determine. Further, the Annales regni francorum are regarded as ‘official’ court-based history. The problem with this of course is how can we work out how or whether the court might determine either content or distribution. The attraction of Kurze’s theoretical framework for the composition of the Annales regni francorum, of course, is that it envisages a major effort to record the early rise to prominence of Pippin and Charlemagne, and then a further reflection on Charlemagne’s reign at the end of it, around the time that his heir Louis the Pious was crowned. The degree to which the manuscripts may be able to throw light on the original conception, pro- cess of composition and dissemination of the annals nevertheless seems to me to be very limited. Despite, or because of this, it has prompted a great deal of speculation. It is important, however, to be clear about the basis on which we are speculating. An analysis based on the printed text alone, without taking the manuscripts into account, is simply inadequate. The manuscript dissemination may nevertheless be able to illuminate the perception of the Frankish past subsequently, in the reigns of Louis the Pious and his sons Lothar, Louis the German and Charles the Bald, as well as the subsequent circulation of Frankish history in the later ninth century and the course of the tenth century.40 What this leaves unresolved is the status of the text either as histor- ical record, or as a statement of a particular memory of the Frankish past. If the manuscript witnesses are regarded as reflections of particular preferences in memory and the selection of what was wanted from the past to hand on to posterity, then each manuscript has a great deal to disclose. Any effort at total reconciliation of accounts and perspectives will probably be fruitless, nor is it necessarily advantageous. What the Frankish annals manuscripts point to above all is a number of alternative histories and alternative memories of the Frankish past. The early history of the Franks and Carolingians is remarkably widely circulated, in many permutations and combinations, including the Annales mettenses priores,41 40 See below, pp. 101–13. 41 See Y. Hen, ‘The Annals of Metz and the Merovingian past’, in Hen and Innes (eds.), The uses of the past, pp. 175–90. History and memory in the Carolingian world 21 the Annals of Fulda,42 the Chronicon laurissense breve, the Continuations of Fredegar, the Annales regni francorum, and the dozens of so-called mi- nor annals that remain to be analysed.43 Such a rich dissemination is crucial for our assessment of the formation of an historical memory of the origin of the Franks and the rise and triumph of the Carolingian family. A further questions remains: If the Annales regni francorum is court- based, official history, is there any discernible role played by the court in the dissemination of this particular rendering of the Frankish past? Of all the manuscripts surveyed by Kurze and discussed by everyone else since, only two, St Petersburg Saltykov-Schedrin Library F.v.IV.4 and Vienna, ÖNB cod. 473, have any links with a Carolingian court (both, as it happens, with that of Charles the Bald) and both are part of composite presentation manuscripts.44 There is, however, one early fragment of the Annales regni francorum, not known to Kurze, and apparently ignored by, or unknown to, every subsequent commentator. This is Cologne fragment Sankt Maria in Kapi- tol AII/18, dated s.IX 1/3 and containing the E text entry for the year 824 but with some variants.45 It is a single leaf, desperately frail and thin, with the script made very difficult to decipher as a consequence. It witnesses to a court-associated copy of the revised text of the Frankish annals and may well be the archetype of E1–E8 whose existence was supposed by 42 It is too often forgotten that the Annals of Fulda start in 714 with the career of Charles Martel and present in their entirety an interestingly different perspective on the career of the early Carolingian rulers: ed. F. Kurze, Annales fuldenses sive annales regni francorum orientalis, MGH SRG VII (Hannover, 1891). 43 These fill G. Pertz (ed.), MGH SS, I and II (Hannover, 1826 and 1829) with others in MGH SS, XV (Hannover, 1887); and see R. McKitterick, The Frankish kingdoms under the Carolingians, 751–987 (London, 1983), pp. 3–5. J. Davis made a start on an analysis of thirty-nine of these ‘minor annals’ in ‘Conceptions of kingship under Charlemagne’ (unpublished M.Litt. dissertation, University of Cambridge, 1999), pp. 173–83. 44 For an account of the contents of St Petersburg F.v.IV.4, see Tischler, Einharts Vita Karoli, pp. 1163–76, and see note 31 above. For the Vienna codex see H. Reimitz, ‘Ein karolingisches Geschichtsbuch aus Saint-Amand und der Codex Vindobonensis palat. 473’, in C. Egger and H. Weigl (eds.), Text-Schrift-Codex. Quellenkundliche Arbeiten aus dem Institut für Österreichische Geschichtsforschung, MIÖG Ergänzungsband (Vienna and Munich, 2000), pp. 34–90, and below, chapter 5. 45 This fragment was first signalled by B. Bischoff, ‘Die Hofbibliothek Ludwig dem Frommen’, in J. J. G. Alexander and M. T. Gibson (eds.), Medieval learning and literature. Essays presented to Richard William Hunt (Oxford, 1976), pp. 3–22, and in English trans- lation (by M. Gorman), in B. Bischoff, Manuscripts and libraries in the age of Charlemagne, Cambridge Studies in Palaeography and Codicology 1 (Cambridge, 1994), pp. 76–92, at p. 90, there with the old signature of Capsula 34.I. It was also discussed by L. Kolarova, ‘The transmission and dissemination of Carolingian annals’ (unpublished M.Phil. dis- sertation, University of Cambridge, 1995). I am most grateful to Dr Joachim Oepen of the Historisches Archiv des Erzbistums Köln for his kindness in sending me information and photographs of this remarkable fragment. See also p. 130 below. 22 History and Memory in the Carolingian World Kurze. It raises the possibility, therefore, of the royal court being directly associated with the dissemination of at least one version of the Frankish past. I have suggested in this introductory chapter that Frankish history books, containing both new and old texts, reflect a particular exploitation of the potential of written culture in the Carolingian world. They supplied an arsenal of past precedent and knowledge and a statement about present identity. They imply a potential audience for these texts and provide the means for their dissemination. Taken together the texts and extant manuscripts constitute the self-defining action of an elite and the use of history texts to shape shared memory. In these Frankish historical texts we see the use of the written word to persuade contemporaries and posterity of the importance and status of the Franks. But in shaping memory, the Franks drew on a host of written versions and records of the past. The an- nals I have discussed are simply the most prominent among many variant compilations disseminated throughout the Frankish world. Fortunately we do not have to rely solely on these narratives for an understanding of their relation to reality, for there is a host of documentary, legal and epistolary material with which to compare them. Nevertheless, the great variety of reception and dissemination of history texts mirrored in the ninth-century manuscripts is an unequivocal indicator of the sheer ex- tent of the network of a shared memory of Frankish history in the early middle ages. It is to the documentation of this shared memory and his- tory in its extraordinary variety in the Carolingian world that the following chapters are devoted. In the next chapter, on Carolingian history books, I take up the issue of determining the purpose and subsequent audience for early medi- eval historical texts on the basis of the evidence provided by surviv- ing manuscripts and their dissemination, and develop therefore a major theme this introduction has also addressed. The manuscripts enable modern historians to assess the status of the text and to place them in an appropriate historical context. This chapter discusses the manuscript tra- dition of the Liber pontificalis, Einhard’s Vita Karoli, the Annales fuldenses and Fredegar’s Chronicle. It assesses the implications of the survival and record of copies in the early middle ages of Roman and early Christian historiography. Finally, it discusses the implications of three major his- torical miscellanies or history books, one in St Omer, one in Lucca and the third the so-called ‘Verona miscellany’, which is now divided into four volumes, two each in Berlin and St Petersburg. Here the agenda of the compiler is crucial for determining the purpose of the collection as a whole and suggesting its possible audience. I argue that all historical History and memory in the Carolingian world 23 writing can be transformed to serve a particular author’s or compiler’s specific purpose in relation to whatever audience or audiences the author or compiler may have had in mind. In the third chapter, I turn to the question of the particular audience ini- tially envisaged for a particular text in relation to Paul the Deacon’s Histo- ria langobardorum (History of the Lombards). The conquest of Lombardy in 774 is universally recognized, both by contemporaries and by modern historians, to have been one of the most momentous events in Charle- magne’s reign. Yet the process by which Charlemagne consolidated his rule in the Lombard kingdom of Italy, and the cultural and religious consequences of 774 for both Franks and Lombards, are far less easy to determine. In this chapter I argue that it is in the context of Paul the Deacon’s political involvement both before and after the Frankish conquest that all his writings, and especially the Historia langobardorum, should be seen. I look at both the content of the Historia langobardorum and the manuscript tradition and transmission of the work. I suggest that Paul’s history might be better understood as a very skilful piece of image- making about the Lombards’ past and identity on behalf of the Lombards for the Franks, either in Francia itself or for the Franks and Lombards at the court of Pippin of Italy. The History was designed to serve a particu- lar function, namely, to instruct the Franks about the Lombard past and provide some measure of legitimation of Carolingian rule. What can be put together about its reception suggests that it was successful. In short, the Historia langobardorum is a very active contribution to the shaping of Frankish and Lombard relations and the understanding of kingship in the aftermath of 774. The next three chapters form a group. They continue the discussion of the shaping of perceptions of kingship already touched on in chapter 3 but concentrate in particular on the forging of Frankish identity. In chapter 4, ‘The Carolingians on their past’, I argue that the Annales regni francorum forge a Frankish identity by constant reiteration and triumphal narrative. The ruler and the Franks are the achievers and together create the great realm. Consolidated within an historical and Christian framework, this is the message passed on to their contemporaries and to posterity. The insis- tence on precise chronology according to the year of the Incarnation is a deliberate device to enhance a very determined expression of the Franks’ identity and cultural affiliations. I am concerned with the construction of a past by the Franks, its coherence and consistency, and the degree to which such a construction constitutes the formation of the collective memory of the newly formed Frankish people under Carolingian rule. I focus on the Annales regni francorum as one highly influential historical narrative that constructed so powerful an image of Frankish society and 24 History and Memory in the Carolingian World its events, and evoked such a convincing sense of identity, that it is their version of the Frankish past that has been remembered, and believed, ever since. Chapter 5, ‘Politics and history’ serves as a companion piece to the pre- ceding chapter, for it takes up the issue of the dissemination of the Annales regni francorum within the Carolingian empire in the ninth century. It also takes up one of the challenges I offered in chapter 2 concerning historical compilations. It begins as a study of a single manuscript but it impli- cates all other Carolingian codices containing historical texts. It focusses in particular on a royal history book, Vienna, ÖNB cod. 473, probably written in 869. It included the Frankish annals with other historical texts (such as the Liber pontificalis, the Liber historiae francorum and Vita Karoli of Einhard) to create an extraordinarily significant and important book. Such a collection exposes the necessity to explore the implications of the dissemination of Carolingian historiography in terms not only of the spe- cific political impetus for the initial production of each text but also of their subsequent impact and use. In chapter 6, ‘Kingship and the writing of history’, I examine the cre- ation of a particular image of royal and Frankish power. I test the hypoth- esis that the writing of history in the Frankish world was not simply a matter of an observer recording, or even selecting judiciously, disingenu- ously or with deliberate intent to mislead, from events as they happened. I make the case for the need to register the formation of an historical sensitivity by means of the other texts which a particular author of an his- toriographical work might have encountered, as well as the formation of a collective memory, understanding and interpretation of what had hap- pened. My particular focus is the account in the royal Frankish annals of the usurpation of the Merovingian throne by the Carolingian mayor of the palace Pippin III. I argue that the claim that the Pope had sanc- tioned the deposition of the last Merovingian king in 751 is a piece of creative mythmaking on the part of the annalist for political and ideologi- cal reasons. This claim had an astonishing resonance in political thinking and ideas about the French constitution thereafter. I suggest, therefore, that the pope’s first involvement in the political affairs of the Frankish kingdom was not until 753/4 and that Pippin’s apologists created a very particular understanding of the making and early years of Carolingian royal power that has misled historians and political theorists ever since. The following two chapters then take up the wider themes of social memory and history and the different manifestations of their expression that can be observed in surviving Carolingian documents. Both in chap- ters 7 and 8, therefore, I discuss cartularies and Libri memoriales, where there is a conjunction of an historical sense of the past, attachment to History and memory in the Carolingian world 25 geographical place, commemoration, record and writing. Libri memo- riales, first written in the Carolingian period, survive from Brescia, Durham, Passau, Pfäfers, Reichenau, Remiremont, Salzburg, St Gallen and Winchester. I examine the continental books in particular, notably the Liber memorialis of Remiremont and the Liber vitae of Salzburg. These types of book constitute written forms other than historical narrative in which the past was remembered and commemorated in the early middle ages, and have much to tell us about cultural attitudes in the early middle ages. The Libri vitae and the cartularies augment oral modes and mem- ory as a form of communication in the early middle ages. They also func- tion as a symbol on many different social and spiritual levels. They illus- trate, moreover, the role of writing as a form of communication over time, and it is this which is absolutely crucial to our understanding of the uses and implications of literacy in the early middle ages as a whole. The Liber vitae of Salzburg, first compiled in 784, is a remarkable example of the use of writing in commemoration and the recording of social memory in the early middle ages. I argue that this book reflects not only more general cultural assumptions in early medieval Bavaria in the years surrounding Charlemagne’s annexation of Tassilo’s duchy of Bavaria, but also spe- cific political affiliations and social communities within Bavaria at the end of the eighth century. The Liber vitae of Salzburg certainly anchored the Salzburg community within its locality and immediate institutional memory. That memory itself, however, was not separated from a wider political realm represented both by Bavarian dukes and the Carolingian king, and by the links forged between the religious houses and sees of west Francia and Bavaria. It witnesses to the different strands of influence and affiliation in Agilolfing and Carolingian Bavaria. The prayers for the dead, the Libri vitae and the cartularies which formed the subject of the two preceding chapters provide an ostensibly immediate conjunction between past and present time, in that the dead are remembered in terms of the commemoration of their anniversaries in the present. Yet such texts also convey a very particular sense of the historical past in which chronology has a crucial role to play. In thinking about the writing of history and the keeping of historical records of many different kinds, we also need to think about the historical mindedness of the Franks and the formation of historical sensitivity by means of reading texts. How did an author’s historical ‘training’ and reading influence the presentation and perception of the events, issues and personalities he records? What are the implications of the presence of histories and what was made available by copying in the ninth and tenth centuries? In the remaining chapters of the book I turn to the question of histor- ical knowledge and the role of texts. As a case study in chapter 9, I draw 26 History and Memory in the Carolingian World in particular on the evidence, insofar as I have been able to reconstruct it from surviving manuscripts from this period, of what was read and used at the monasteries of Lorsch and St Amand in the ninth and tenth cen- turies. In chapter 10 I explore a particular aspect of the ‘longer past’ of the Franks, namely the history of the church. I aim to demonstrate that the relationship of the Carolingian Franks to the books in their libraries was not simply that of scholars to a repository of learning. The books in Frankish monastic and cathedral libraries, as part of a past which the Franks had assimilated to themselves, as discussed in chapters 7, 8 and 9, formed part of the Frankish sense of identity. In other words the Franks were not only a textual community in relation to the Bible, as has long been recognised; they were also a textual community in terms of their intellectual and textual inheritance. I use the phrase ‘textual community’ as defined by Brian Stock in his book: ‘What was essential for a textual community was not a written version of a text, though that was some- times present, but an individual who having mastered it, then utilized it for reforming a group’s thought and action’.46 In the case of the Franks in the Carolingian period, I am concerned both with the ruler’s advocacy of particular texts but also of the Frankish scholars who mastered a precise set of texts. I explore the formation and ramifications of the very distinc- tive understanding of books and of the Franks’ place in history within the Carolingian world and the Carolingian church. I investigate in particular the wider implications of Jerome-Gennadius’ De viris illustribus, Eusebius- Rufinus’ Historia ecclesiastica and Cassiodorus-Epiphanius’ Historia eccle- siastica tripartita within Carolingian Europe. These three works played a key role in creating a context for the Franks’ understanding not only of the history of the church, but also of the circumstances of the compo- sition and dissemination of Scripture and the work of the fathers of the early church. There are striking instances from the Carolingian period of the way in which the distinctive stress on texts, authors, authority and sacred places in the ecclesiastical histories, Jerome-Gennadius’ De viris illustribus and a number of related texts could be drawn on by readers, scribes and artists. In a companion piece to the preceding chapter, I focus in chapter 11, ‘Christianity as history’, on two examples in order to explore the way in which the written tradition of the church appears to have shaped per- ceptions and attitudes. I suggest how the particular image of the church and the Christian faith, presented by the historical tradition outlined in chapter 10 and disseminated so zealously by the Franks, manifests 46 B. Stock, The implications of literacy. Written language and models of interpretation in the eleventh and twelfth centuries (Princeton, 1983), pp. 89–92. History and memory in the Carolingian world 27 itself in the Carolingian period. My two examples are Vercelli, Biblioteca Capitolare cod. CLXV, a collection of canon law, and Munich, Bayerische Staatsbibliothek Clm 22053, the famous miscellany in a codex known as the Wessobrunner Gebet manuscript. Both demonstrate how im- ages were deployed to reinforce the importance of sacred places, relics, authors and books in Christian history. They also serve to underline the importance of codicological context for our own understanding of any one text and its significance for its intended audience. They thereby add to our own knowledge of the intellectual resources of particular groups in the Carolingian world, quite apart from enlarging our understand- ing of the reception of late antique Christian texts in early medieval Europe. The concluding chapter, ‘History and its audiences in the Carolingian world’, develops the principal theme of the preceding eleven chapters in focusing on contemporary memory and the writing of history in the eighth and ninth centuries. It takes the events of 817 and the contempo- rary accounts of them as a case study. It discusses how the Frankish writ- ers constructed their past in the early middle ages and how their sense of an immediate history related to the construction of a longer past. In order to pull together the issues discussed in the preceding chapters, I look again at some of the material already considered but here from the per- spective of their potential audiences. I address the following questions. What contributed to the Franks’ sense of place in historical time? What did they use to construct their past? How did their own immediate history relate to this longer past? I argue that the Franks’ sense of the past was a composite one. It had overlapping sequences. These comprised different local and institutional senses of identity which were expressed in terms of their own community’s foundations, property, associations, dead mem- bers, benefactors and others. They were remembered, moreover, with some association with a particular place. Simultaneously there was an understanding of the chronological progression of Jewish and Christian history; of being heirs of both imperial and Christian Rome; and of their own sense of achievement as Franks, expanding ever eastwards and im- posing their own composite culture on others. A sense of the past was deeply integrated into the sense of identity of the audiences for history the Carolingian world. How distinctive this Frankish sense and exposi- tion of the past is among their contemporaries, notably in the Byzantine, Islamic and Anglo-Saxon regions, forms the subject of my concluding paragraphs. 2 Carolingian history books The history of Alexander by Quintus Curtius Rufus was written in the first century AD. Its earliest surviving manuscript, Paris, BnF lat. 5716, was written in the Carolingian period for Count Conrad by the scribe Haimo in the Loire region in the second half of the ninth century. How may we account for this? Quintus Curtius Rufus is not an orthodox member of the Roman histo- riographical canon in the sense that he does not deal with contemporary Roman themes. Certainly his History of Alexander is sensational and emotional; it has much of the exotic and the remote in its narrative, and good character sketches. There are vivid depictions of the characteristics of different peoples – effeminate Persians, intelligent Scythians, volatile Egyptians, the Greeks, who are ‘political trimmers by temperament’, ‘time-serving Cretans’, the Sicilians’ penchant for flattery and the fickle- ness of the mob. The speeches abound and the morals are pointed. Ac- counting for its attractions, therefore, is perhaps not difficult, though it risks being a subjective exercise in relation to modern tastes. Nevertheless, other ninth-century copies of Quintus Curtius Rufus of different origin besides that of Count Conrad’s b