Language and History in the Early Germanic World 0521471346, 9780521471343

Table of contents :
Preface ix
List of abbreviations xi
Map xiv
General introduction 1
PART I: THE GERMANIC WORLD
Introduction to Part I 11
1. Religion 13
2. Law 30
3. Kinship 49
4. Warfare 67
5. People and army 84
6. Lordship 102
7. Kingship 121
PART II: CONTACT WITH THE NON-GERMANIC WORLD
Introduction to Part II 143
8. Contact with the Celts 145
9. The migration of the Goths 164
10. Germanic loanwords in Latin 182
11. Latin loanwords in Germanic 201
12. Trade and warfare with the Romans 219
13. Names of days of the week 236
14. The vocabulary of writing 254
PART III: CONTACT WITH CHRISTIANITY
Introduction to Part III 273
15. Problems of Christianisation 275
16. The influence of provincial Roman Christianity 291
17. The influence of Gothic 308
18. The influence of the Merovingian Franks 325
19. The influence of the Anglo-Saxons 341
20. Contrasts in Christian vocabulary 357
21. The vocabulary of ethics and fate 374
Bibliography 392
Index of words 419

Citation preview

This book offers a distinctive and accessible approach to the earli­ est encounters of the Germanic tribes of Northern Europe with classical antiquity and with early Christianity. It brings together lin­ guistic evidence ranging in date from before Caesar to about ad 900, and in geographical scope from Ireland to the Crimea and Visigothic Spain to the Eastern Baltic, to shed light on important aspects of Germanic culture and to show how they were affected by contact with the (mainly but not exclusively Latin) outside world. It shows how semantics and loanword studies, often regarded with trepidation by non-specialists, can form part of a fascinating jigsaw puzzle and provide important clues for historians and arch­ aeologists of the period. Likewise, it demonstrates that philologists and linguists ignore historical evidence at their peril.

LANGUAGE AND

H I S T O R Y IN T H E E A R L Y

G E R M A N IC W O R LD

Th e York Helmet: an Anglo-Saxon artefact o f the eighth century. Decorated with a Christian prayer in Latin, but also incorporating Germanic decorative elements and based on a Roman parade helmet, it embodies the cultural fusion which can also be traced in the linguistic evidence o f the period. By courtesy o f York Castle Museum.

LANGUAGE AND HISTORY IN THE EARLY GERMANIC WORLD D. H . G R E E N Schröder Professor Emeritus, University o f Cambridge Fellow o f Trinity College, Cambridge

CAMBRIDGE UNIVERSITY PRESS

PUB LIS HE D BY THE PRESS S Y N D IC A T E OF THE U N I V E R S I T Y OF C A M B R ID G E

T h e Pitt Building, Trumpington Street, Cam bridge CB2

irp,

United Kingdom

C A M B R ID G E U N I V E R S I T Y PRESS

T h e Edinburgh Building, Cam bridge CB2 2 RU, United Kingdom 40 West 20th Street, New York, n y 10011-4211, U S A 10 Stamford Road, Oakleigh, Melbourne 3166, Australia © Cam bridge University Press 1998 This book is in copyright. Subject to statutory exception and to the provisions o f relevant collective licensing agreements, no reproduction o f any part may take place without the written permission o f Cam bridge University Press. First published 1998 Printed in the United Kingdom at the University Press, Cam bridge Typeset in Baskerville

mt

11/12.5 [ se]

A catalogue recordfor this book is availablefrom the British Library Library of Congress Cataloguing in Publication Data Green, Dennis Howard, 1922Language and history in the early Germ anic world / D. H. Green, p. cm. Includes bibliographical references and index. ISBN o 521 47134 6 hardbac k I. Germanic languages - History. 2. Germ anic tribes - History. 3. Europe - History - 392-814. 4. Languages in contact - Europe History. 1. Title.

PD75*G74 43° 9-dc2i isbn

1998 97-9527 cn>

0 521 47134 6 hardback

Contents

Preface List o f abbreviations Map

page jx xi xiv

General introduction

PART

i: T H E

GERM ANIC

i

WORLD

Introduction to Part i

11

1.

Religion

13

2.

Law

30

3.

Kinship

49

4.

Warfare

67

5.

People and arm y

84

6.

Lordship

102

7.

Kingship

121

PART

II: CONTACT

W ITH

THE N O N -G ERM AN IC

Introduction to Part 11

WORLD

143

8.

Contact with the Celts

145

9.

T h e migration o f the Goths

164

10.

Germ anic loanwords in Latin

182

il

Latin loanwords in Germ anic

201

vii

viii

Contents

12.

Trade and warfare with the Romans

219

13.

Names o f days o f the week

236

14.

T he vocabulary o f writing

254

PART III: CONTACT WITH CHRISTIANITY

Introduction to Part hi

273

15.

Problems o f Christianisation

275

16.

The influence o f provincial Roman Christianity

291

17.

T he influence o f Gothic

308

18.

T he influence o f the Merovingian Franks

325

19.

The influence of the Anglo-Saxons

341

20.

Contrasts in Christian vocabulary

357

21.

The vocabulary of ethics and fate

374

Bibliography Index of words

392 41g

Preface

This book derives from a three-year cycle o f lectures given to Cam bridge undergraduates, revised and adapted over many years. It makes no claim to advance the frontiers o f knowledge, but is meant instead to provide students with a broad survey which they could not be expected to map out for themselves. W hat originality the book does possess lies in its arrangement o f the linguistic material, correlating it with historical find­ ings and seeing it as illustrating the encounter o f the early Germ anic tribes with Rome and Christianity. Compressing a three-year cycle into a book has led to a drastic reduction o f the material originally presented. M y wish to give as broad a survey as possible, however, means that each o f the words discussed has had to be treated briefly: difficulties are pointed out, conclusions are suggested, but in many cases the detailed argument is to be retrieved from the secondary literature cited. Even so, I am painfully aware o f many themes and examples which have had to be sacrificed. I owe a debt o f gratitude to former students who attended the lectures and who, now colleagues, have encouraged me to convert the material into a book. Foremost amongst these are K ate Brett o f the Cam bridge University Press and David Dumville. To the latter, as well as to Peter Johnson, I am further grateful for critical comments on individual chap­ ters in draft form. Financial support, making research stays in Germ any possible, from my University, my College and the British Academ y is gratefully acknowledged. I am especially grateful to Catherine Walters for her skill in transferring my handwriting to the computer and for her willingness to cope with the range o f linguistic difficulties which this involved. Stimulus and the opportunity to discuss shared problems across disci­ plinary boundaries have also come from meetings o f the Center for Interdisciplinary Research on Social Stress in San Marino. T h e Introduction is an abridgment o f a paper in After Empire. Towards an IX

X

Preface

ethnology of Europe’s barbarians, Woodbridge 1995. Chapter 9 is an adap­ tation o f a paper given at a C IR O S S conference in 1995, the proceed­ ings o f which are to appear in a volume, Franks and Alamans in the Merovingian period. An ethnographic perspective, in preparation. Chapter 18 is related to a paper given at a C IR O S S conference in 1996, the proceed­ ings o f which will appear in The Visigoths from the migration period to the seventh century. An ethnographic perspective, in preparation. M y last book I dedicated to my wife Margaret. This one I dedicate to her memory.

Abbreviations

ABäG AfdA A jK AjR Al. ANF A SE BAR B zN CG Du. EFr. EFri. EG Engl. FEW FM S Fr. Fri. Frk. FS FuF Gm c. Go. Gr. GR GRev. GRBS GRM HJLg

Amsterdamer Beiträge zur älteren Germanistik Anzeigerfiir deutsches Altertum Archivfiir Kulturgeschichte Archivfiir Religionswissenschaß Alem annic Arkivför nordisk Filologi Anglo-Saxon England British Archaeological Reports Beiträge zur Namenforschung Com m on Germ anic Dutch East Franconian East Frisian East Germ anic English Französisches Etymologisches Wörterbuch, ed. W. von Wartburg, Bonn 1928fr. Frühmittelalterliche Studien French Frisian Frankish Festschrift Forschungen und Fortschritte Germ anic Gothic Greek Germania Romana, second edition, vol. i, T. Frings, Halle 1966; vol. h , T. Frings and G. Müller, Halle 1968. Germanic Review Greek, Roman and Byzantine Studies Germanisch-Romanische Monatsschrift Hessisches Jahrbuchför Landesgeschichte XI

xii

Hung. HZ IE IF Ir. It. JIE S JR G ZM Lat. LFr. LG Lith. M Du. MFr. MGH AA M G H SS MHG M IÖG Mir. M Lat. M LG M LR MWe. Ndjb NG NoB Norw. NOW ELE NWG OC OE OFr. OFri. OHG Olr. OLFr. OLith. ON OPort. OPr.

L is t o f abbreviations

Hungarian Historische Zeitschrift Indo-European Indogermanische Forschungen Irish Italian Journal o f Indo-European Studies Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz Latin Low Franconian Low Germ an Lithuanian Middle Dutch Middle Franconian Monumenta Germ aniae Historica, Auctores Antiquissimi Monumenta Germ aniae Historica, Scriptores Middle High Germ an Mitteilungen des InstitutsJur österreichische Geschichtsforschung Middle Irish Medieval Latin Middle Low Germ an Modern Language Review Middle Welsh Niederdeutsches Jahrbuch North Germ anic Namn och Bygd Norwegian North- Western European Language Evolution North West Germ anic O ld Catalan O ld English O ld French O ld Frisian O ld High Germ an O ld Irish O ld Low Franconian O ld Lithuanian O ld Norse O ld Portuguese O ld Provençal

L is t o f abbreviations

O Pru. OS O S 1. O Sp. PBA PBB PG Pol. Port. RhM R hV RLAC RLGA

Rum. SI. Sp. SRhFr. Sw. UG VG VL We. WFrk. WG WuS ZjbLg ZfdA ZfdPh ZfdS ZJdW ZJM J ZJMR ZfrPh ZĶG ZJÍF ZRG(GA) ZSlPh ZV S

xiii

O ld Prussian O ld Saxon O ld Slavonic O ld Spanish Proceedings of the British Academy Paul und Braunes Beiträge Primitive Germ anie Polish Portuguese Rheinisches Museum Rheinische Vierteljahrsblätter ReallexikonJur Antike und Christentum, ed. T. Klauser, Stuttgart 1950fr. Reallexikon der germanischen Altertumskunde, first edition ed. J. Hoops, Strassburg, 191 iff.; second edition ed. H .Jankuhn et a i, Berlin, 1968fr Rumanian Slavonie Spanish South Rhenish Franconian Swedish Upper Germ an Vulgar Greek Vulgar Latin Welsh West Frankish West Germ anic Wörter und Sachen Zeitschriftßir bayerische Landesgeschichte ZeitschriftJur deutsches Altertum ZeitschriftJur deutsche Philologie Zeitschriftfür die deutsche Sprache Zeitschriftßir deutsche Wortforschung Zeitschriftßir Mundartforschung Zeitschriftßir Missiomwissenschaft und Religionswissenschaft Zeitschriftßir romanische Philologie ZeitschriftJur Kirchengeschichte Zeitschriftfür Namenforschung ZeitschriftJur Rechtsgeschichte (Germanistische Abteilung) Zeitschriftßir slavische Philologie Zeitschriftßir vergleichende Sprachforschung

M ap o f the area covered by the migrations o f Goths from northern Poland to the Black 5 (Chapter 9).

General introduction

T h e disciplines concerned with the encounter between the early Germ anic world and the Roman Empire as well as Christianity are archaeology, history and philology. Although they have not always been ready to make use o f each other’s findings recent work has been mark­ edly more interdisciplinary,1 which in turn has accentuated the problems which arise when one discipline, in the case o f this book philology, seeks assistance from the others. Archaeology finds it notoriously difficult to connect the finds which for it constitute a distinct culture or group with a particular tribe to which a name can be given, thereby associating them with written his­ torical sources or linguistic evidence.2 Ethnic groups which may share a similar material culture cannot always be distinguished from one another (e.g. Goths and Heruli).3 W here a particular class o f find is missing altogether (as with the absence o f military gear from Gothic grave goods) we are left even more in the lurch by archaeology and have no choice but to fall back on the evidence o f language, where even the restricted vocabulary o f Wulfila’s Gothic bible offers examples such as brunjo ‘breastplate’, mēki ‘sword’ and wepna ‘weapons’.4 This illustrates drastically the main disadvantage o f archaeological evidence for philol­ ogy: it opens the door on a past which is silent, because wordless, it records aspects o f the material basis o f an early society, but permits only inferences about the social, institutional and religious thought-world o f a society which left no written records.5 O n the other hand, archaeology disposes o f a finely calibrated and reliable dating scheme, covering a number o f distinct periods, with further subdivisions, which far surpasses the known inability o f philology 1 E.g. Wenskus (1961) or the indispensable volumes o f R IG A , many o f whose entries are divided into sections labelled archaeological, historical or linguistic. 2 Green (1995), 143. 3 Heather (1991), 4, 87. 4 Wolfram (1988), 9 9 i 5 Jankuhn (1967), 118; M urray (1983), 221; Roberts (1994), 185. I

2

General introduction

to establish more than a broadly based relative chronology.6 H ow helpful archaeology can be in resolving linguistic problems has been shown in the case o f the G oths’ migration from the Baltic coastline to the Black Sea, in the course o f which they made contacts, including linguistic exchanges, with a number o f other peoples.7 Thanks to archaeological dating methods it has been possible to reconstruct this migration chronologically and geographically, to relate these findings to contem­ porary historical records and thus to provide a background against which to set the question o f linguistic exchange.8*Even on a much smaller scale archaeology can also throw light on linguistic problems, either in the case o f Go. wein as an early loan from Lat. imum or, on the basis o f the Sutton Hoo helmet, in explaining details o f the description o f helmets in Beowulfy T h e gaps in our evidence on this early period, both linguistic and archaeological, are so frequent that neither discipline can afford to ignore what it might learn from the other.10 T h e same is true, both negatively and positively, when we attempt to correlate historical with linguistic evidence. For many o f the debates amongst historians the evidence is so sparse or ambiguous that it would be unwise to neglect the possible relevance o f information from a neigh­ bouring discipline - and the same applies equally to debates amongst philologists. Moreover, although historical sources have the advantage over archaeological evidence in speaking to us in words, the language o f these sources is overwhelmingly Latin (or Greek), not the barbarians’ vernacular. This raises the question how far a classical language was able to render the details o f a more primitive society at all accurately, how far these may have been distorted, linguistically or conceptually, by being subjected to an interpretatio Romana.11 If early barbarian laws, codified in Latin, found it necessary to employ vernacular terms for want o f a reli­ able Latin equivalent,12 this suggests that historical sources in Latin, whatever their nature, risk distorting the thought-world o f the barbar­ ians on whom they report. Less distortion, but still difficulties o f inter­ pretation, are involved when we approach the world o f the Germ ani from the language which they themselves used. H ow necessary it is for a linguistic discussion o f early G erm ania to take the historical dimension into account is shown by the occasional 6 8 9 10 12

Birkhan (1970), 97. For an example see Heather (1996), 19. 7 See below, p. 168. Bierbrauer (1994), 5iflf.; Heather (1996), i8ff., 35ÍT. (both with maps). Wein: see below, pp. 227ff; Beowulf. Cram p (1957), 6off.; Newton (1994), 33ff. Cram p (1957), 57. 11 Richter (1994), 99i. O n interpretatio Romana see below, pp. nf., 246fr. Schmidt-W iegand (1979), 56fr.

General introduction

3

oversimplified view o f the G erm anic world at large.13 T h e G erm ani described by Tacitus need not be the same as those described over a century earlier by Caesar, so that differences between these two sources, far from needing to be explained away, may reflect the result o f the encounter o f the G erm ani with the world o f Rom e.14 Similarly, to compare Arminius o f the Cherusci with the Goth Athanaric means ignoring the three and a half centuries separating them and the differ­ ent acculturation processes undergone by W G and E G tribes.15 Too often linguistic conclusions have been reached on the basis o f what can be termed a linguistic absolutism, ignoring the question o f their intrin­ sic historical possibility, as when a Gothic mission to Germ any (even purportedly affecting Scandinavia) has been put forward without any trace o f historical backing.16 Historical factors can at times account for a linguistic feature, as with the different employment o f the Gm c. stem *nk- in W G and EG. This word, denoting ‘powerful’, ‘authority’ or the area in which authority is exercised, was also used in the names o f rulers (especially in the east: e.g. Ermanaric, Theoderic), decking themselves out with the trappings o f extended authority by adopting this outlandish mode o f Celtic origin.17 Concentration o f power could run into opposition and be regarded as tyranny, in western G erm ania (Maroboduus, Arminius), but also, in view o f what Tacitus says o f the more centralised authority o f E G kingship, in the east.18 This is reflected in the negative use o f this word by W ulfila,19 whose objection to the despotic authority it denoted was not aimed at a Rom an Empire by now Christian, but at Erm anaric’s concentration o f power. (This Ostrogothic ruler is similarly criticised in other sources.)20 If, by contrast, W G uses the word positively (of the Frankish empire, but also o f Christ’s kingdom)21 the explanation m ay be sought in the adoption o f Rom an imperial ideology by the Franks and the Anglo-Saxons, which supplied them with a justification unavailable to the Goths in the fourth century.22 However profitable it may therefore be to correlate the findings o f 13 Green (1Q95), I43f. 14 WaUace-Hadrill (1971), 4. 15 Wolfram (1975), 23. 16 Schäferdiek (1978a), 52if. Cf. also Rotsaert (1977), i82f. 17 Herold (1941), 274F.; Ris (1971), 15fr. 18 Maroboduus and Arminius: Thom pson (1965), 66ff., 83f; Tacitus: Germania 43. 19 See below, p. 139. 20 Jordanes, Getica 129; Widsith 8f. (cf. Cham bers (1912), 190); Deor 2iff. Cf. also Schneider (1928), 246f.; Dronke (1969), 217, 220; Gillespie (1973), 37, 39. 21 See below, pp. i39f. 22 Franks: Levison (1946), i22f.; Erdmann (1951), 3ff.; Anglo-Saxons: Erdmann (1951), 10, 14; Drögereit (1952), 32; Yorke (1981), 175; Fanning (1991), 15, 17, 19. O n the argument o f this Introduction see also Green (1997), xxixff.

General introduction

4

these three disciplines, this must be done in full awareness o f a methodological danger. Archaeological findings, for example, are often based on conclusions which have been reached by historians, whilst the latter can also seek support from archaeological theories which can then be re-imported into archaeological discourse as if they were an inde­ pendent historical confirmation.23 T h e danger o f a vicious circle is man­ ifest, unless each discipline argues towards its conclusion only from its own material, looking across its boundaries only at the end. T h e argument o f this book concerns the encounter o f G erm ania with classical antiquity and Christianity. T he three parts into which it falls represent the three factors commonly regarded as accounting for the rise o f medieval Europe (in a context wider than this book we should need to extend G erm ania to embrace tribal Europe at large in order to include the Celts and the Slavs).24 Each o f these factors existed along­ side the others, sometimes in opposition and sometimes as a comple­ ment, but all gradually fusing together in the transition from late antiquity to the early Middle Ages.25 Against this wider historical back­ ground our linguistic evidence will have to be placed. Late antiquity is the period in which two o f these factors, the classical tradition and Christianity, enter upon a culturally fertile symbiosis.26 Even though this symbiosis found its institutional expression in the conversion o f a Rom an emperor, Constantine, and the Christianisation o f the Roman state in the fourth century, this decisive change was pre­ ceded and made possible by the gradual readiness o f Christian leaders to identify with the classical world in a number o f respects, to work in and through Rom an society, rather than in opposition to it. T h e conver­ sion o f the Rom an world to Christianity was therefore accompanied by a conversion o f Christianity to the Rom an world.27 Where, for example, the emperor Domitian, the object o f a pre-Christian cult o f emperorworship, was given the double tide Dominus et Deus, the gospel o f St John gave a polemical twist to this by having the aposde Thom as address the risen Christ as ‘my Lord and my G o d ’ (Joh. 20.29).28 Antiquity and G erm ania came together not merely along the imper23 Hachmann (1971), 11; Bierbrauer (1994), 52, 76. For an example see Wagner (1967), 109, fn. 17. 24 E.g. Hiibinger (1969), 203i.; Goetz and Welwei (1995), 2; Allen Brown (1996), 3,15; Davies (1996),

!525 Examples o f cultural fusion in written texts, but also in objects and buildings, have been discussed by Wood (1997), il iff. Whereas the former are for him Latin texts, our concern will be with ver­ nacular words. 26 Lane Fox (1988); Brown (1993) (1996). 27 Brown (1993), 82. 28 Heiler (1959), 548ff

General introduction

5

ial frontier which, in addition to its function as a barrier, acted as a zone in which the two worlds converged,29 but also by the readiness o f Germ anic leaders to heighten their standing by appropriating Rom an models o f authority for themselves. M any o f these leaders sought official approval o f their occupation o f Roman territory from the eastern capital, Constantinople. O doaker deposed Romulus Augustulus, but asked the emperor Zeno for the title o f patrician and expected to govern in Z eno’s name.30 After his victory over the Goths the Frankish ruler Chlodw ig received from the emperor Anastasius I a diploma conferring the tide o f consul. C lad in purple, chlamys and a diadem Chlodw ig rode through Tours, scattering gold and silver like a Rom an ruler, but the crowd’s shouts o f Augustus suggest, if not officially, that for some the polit­ ical claims m ay have gone further.31 T he burial o f Chlodw ig’s father, Childeric, shows a conjunction o f Germ anic features with the trappings o f Roman authority in a different way.32 As in other wealthy Germ anic graves the burial goods include equipment fittings, gold buckles and arm-rings, but also typically Frankish weapons (francisca and scramasax), gold-and-garnet jewellery and, possibly connected with Childeric’s own funeral, the remains o f (sacrificed?) horses. Also included, however, is Childeric’s seal-ring, a feature o f Roman, not Germ anic, civilisation (and inscribed with his name in Rom an letters). It portrays the Frankish king in Roman military dress and holding a lance. It is a typically com ­ posite picture o f sub-Roman power in an emergent Germ anic state.33 For all the contradictions between them Christianity and the Germ anic world could occasionally find common ground on which, in some regions o f Germ ania, the Church could concentrate to ease the task o f Christianisation. O n the other hand a Germ anic king could accept a new G od whose usefulness was demonstrated to him, especially on the battlefield, as was already the case with Constantine at the Milvian Bridge. In the Germ anic world this precedent was repeated by (or attrib­ uted to) Chlodwig, who staked his chance o f victory at Tolbiac on the new G od because, in the face o f what looked like certain defeat, his native gods apparently had let him down.34 Although from a Christian point o f view this cannot be called an inner conversion (for the Christian G od is seen in the same light as the pagan ones), this conjunction o f the 29 Brown (1996), 15,16. 30 Herrin (1989), 19. 31 Wallace-HadriU (1962), 65i., 175^; Herrin (1989), 35i.; M cCorm ick (1990), 335fr. 32 James (1991), 6iff.; Richter (1994), igf.; Brown (1996), 85t 33 For a Langobardic example, the visor o f king Agilulf, see M cC orm ick (1990), 289fr, Brown (I993), I28f. 34 Wallace-Hadrill (1962), 169.

6

General introduction

old with the new characterises the slow fusion o f cultural worlds in the period with which we are concerned. It is also true o f the Anglo-Saxon helmet from Benty Grange, where the image o f the boar (a pagan symbol o f ferocious fighting qualities, providing miraculous protection to the wearer)35 was not felt to be inconsistent with the new religion, for this helmet also has a silver cross (another protective symbol) on its noseguard.36 T he double insurance policy, pagan and Christian, which the owner o f this helmet appears to have taken out is o f a piece with what Bede reports o f king Redwald, setting up a heathen as well as a Christian altar in the same building, and with the similar evidence from G regory o f Tours (paying respects to heathen as well as Christian altars if one passes them).37 The richness o f this process o f cultural fusion over the centuries is demonstrated best, however, when not merely two, but all three cultural traditions, Roman, Christian and Germ anic, come together on particu­ lar occasions. Three examples from Anglo-Saxon England, whose repre­ sentative significance goes far beyond this country, m ay illustrate this.38 With the Ruthwell Cross the Germ anic past is adapted to Christian ends in the use o f runes to inscribe part o f the poem Dream o f the Rood?9 Christianity is also present in the cross itself, a poem dealing with the crucifixion, the inclusion o f passages from the Vulgate, and the themes o f some sculptural panels (e.g. Christ in Judgment, the Annunciation). The sculptures themselves are examples o f realistic work in the classical tradition, their programme is o f Byzantine origin, and the inhabited vine-scroll ornament is likewise inspired by classical models. These same three traditions also come together in the Franks Casket in what has been described as a ‘heady m ix’, even though the meaning o f their conjunction may still escape us.40 Germ anic tradition is present again in the runes recording vernacular (and occasionally Latin) words, but also in scenes depicting Weland the Smith and Egil. Christianity is repre­ sented by the scene o f the Adoration o f the M agi, but also by the con­ quest o f Jerusalem by Titus. Roman tradition could also be summed up in this last scene, but is present in any case in the theme o f Romulus and Remus. Even if no unifying theme suggests itself, the three traditions which converge in the Middle Ages are all at least present here. 35 Beck (1965), 14t. 36 Campbell (1991), 55; Webster and Backhouse (1991), 50t; Newton (1994), 42. 37 Bede: Historia ecclesiastica 1115; G regory o f Tours: Historiarum libri decern v 43. Cf. Wallace-Hadrill (1962), 172. 38 Green (1995), i52f. 39 Swanton (1970), gflf.; Wormald (1991), 91; Howlett (1992), 71fr. 40 Webster and Backhouse (1991), ioiff.

General introduction

7

A magnificent example is provided by the Anglo-Saxon helmet found in the Coppergate at York.41 Its Germ anic background is clear, since its type is known only in England (Benty Grange, Sutton Hoo) and Sweden (Vendel, Valsgärde) and a protective animal is carved between the eye­ brows. A late Rom an parade helmet o f a general (attested from Holland and from Burgh Castle)42 was the prototype for this Germ anic adapta­ tion, but convergence o f two traditions does not exclude Christianity which is even more obviously present here than in the cross on the Benty Grange helmet, for a metal strip on the York find, running from the back o f the helmet over the crown to the nose-guard, has a Christian Latin prayer inscribed upon it. These last three examples illustrate in visible, physical terms (in stone, whalebone, metal) what will emerge from the following pages with regard to the Germ anic languages in these early centuries: their vocab­ ulary, preserved in fossilised form in writing centuries later, likewise reflects aspects o f Germ anic society in contact with the Rom an world and Christianity. In language, too, we must avoid seeing these three constituent traditions in isolation from one another. Even though, for clarity o f exposition, these traditions are treated in three separate parts o f this book, cross-connections enough remain to underline the forces which drew these traditions together to constitute a new whole. For Romans as for Christians in the Empire the arrival o f Germ anic barbar­ ians on the horizon must have signified chaos and destruction, which is how the history o f these centuries, written by Rom an or Christian authors, was presented by contemporaries. Such a view may be justified, but only insofar as we recognise that it is only one point o f view. To it should be added, if only we could reconstruct it, the experience o f the barbarians themselves, shaken by their encounter with a vastly superior civilisation, but also enriched by it and attracted by the promises it held out to them.43 No written historical sources convey their point o f view to us, however, so that an analysis o f the linguistic evidence, however difficult, may allow us a ‘barbarocentric’ approach to these centuries, supplementing the traditional ones based on Roman and Christian per­ ceptions only. T h e linguistic evidence discussed in this book comes from all the G er­ manic languages with the exception o f Frisian, whose written tradition 41 Cam pbell (1991), 66Í; Webster and Backhouse (1991), 6ofT. 42 T h e Dutch example is reproduced in Webster and Brown (1997), plate 48. 43 Davies (1996), 214.

8

General introduction

is too late for the chosen timespan. T h e term ‘Com m on G erm anic’ is therefore to be understood with this reservation, but even so the geo­ graphical span covered extends from Ireland to the Crim ea, from Visigothic Spain to the eastern Baltic. T h e timespan which is taken into account ranges from about 300 b c through to about a d 900, occasion­ ally later in order to take account o f so important a Germ an source as Notker. Although English possesses the adjective ‘G erm anic’ (as distinct from ‘G erm an’) and although the geographical term ‘G erm ania’ (as distinct from ‘G erm any’) causes no difficulty, it lacks a noun, corresponding to Germ an ‘Germ anen’, for those who speak the languages or inhabit the area. To call them ‘Germ ans’, however, invites confusion444 5and it is quite unhistorical to use the term o f a period before the tribes involved regarded themselves as Germans and had evolved a word for this (deutsch)*5. 1 have therefore used the Latin word ‘G erm ani’ to fill this gap. O n the other side o f the linguistic frontier the word ‘Rom ania’ is used o f the area in which the Romance languages are spoken, as distinct from one specific example in eastern Europe, ‘Rum ania’. 44 E.g. the title o f Todd (1992), by contrast with his use o f ‘G erm ani’ in the text, or Allen Brown 0996), 34, where ‘East Germ an tribes’ clearly means ‘East G erm anic’. 45 Weisgerber (1949), 105fr, 117fr.

PART I

T h e Germanic world

Introduction to Part I

O u r earliest written evidence about the Germ anic tribes comes to us from outside, from their civilised neighbours o f the Mediterranean world, the Greeks and the Romans, as they gradually came into contact with northern Europe. This dependence on what Greek and Rom an geographers and historians tell us about G erm ania is not without its dangers, for it forces upon us a view which, because it is from outside, follows the conventions and the conceptual patterns o f classical ethnog­ raphy and historiography (going back to Herodotus), whether or not they do justice to the particular barbarians with whom contact was m ade.1 Seeing G erm ania in this way through classical eyes in foreign cat­ egories amounts to an interprétatif) Romana, a term used by Tacitus {Germania 43) when he equates two named Germ anic gods (Alci) with the Roman pair Castor and Pollux.2 This equation, like others which we may assume were made, rests on a feature shared by the Germ anic and Roman deities, but ignores the differences, so that any interpretation o f the unknown {Alci) by the better known is fraught with danger. T h e danger o f relying on an interprétatif) Romana goes much further than this. T h e information available to classical authors writing about G erm ania may have been incomplete, since they depended on chance reports from military and trading sources, isolated contacts with a still largely unexplored world.3 This information may also have been dis­ torted, first by the difficulty o f conveying the strangeness o f a barbarian world in a classical language with quite different concepts, but also by the tendency o f classical ethnography to shape its descriptions accord­ ing to its established conventions. Greek ethnography, for example, knew at first nothing o f the Germ ani, but did know that northern Europe was inhabited in the west by Celts and in the east by Scythians.4 Accordingly, earliest contacts with the Germ ani confuse them with Celts if the 1 Geary (1988), 3gf. 2 Wissowa (1918), iff. 4 Hachmann (1971), 33fr. I

1 Tim pe (1989), 307fr.

12

T h e Germ anic w orld

encounter is made in the west and with Scythians if in the east, but in either case the barbarian newcomers are understood and presented in terms o f what they are not.5 Another factor affecting our view o f G erm ania when classical sources first report on it is the fact that at this time the Germ anic tribes were already heavily influenced by the Mediterranean world, especially Rome, so that we are deprived o f any picture o f the Germ anic world in its ‘purity’. Indeed, those parts o f this world about which we are best informed are precisely those most exposed to the influence o f Rome (and later o f Christianity). In Parts n and hi we shall be preoccupied with detailed aspects o f this influence, but at this point we need to stress more the disturbing effects o f this encounter between a barbarian society and a higher civilisation, accentuating divisions between Germ anic tribes and within them.6 It was part o f Rom an policy to encourage such divi­ sions for its own ends: to promote pro-Roman factions, which often led also to anti-Roman parties and thus to a fragmentation o f tribal society from which Rome could only profit. Moreover, the progressive encounter with Rome led to a militarisation o f Germ anic society and to the formation o f military and ethnic confederations to deal with the challenge. W hat our earliest written sources present us is therefore a picture o f Germ anic society at one particular, highly critical point in time.7 T h ey rarely allow us to penetrate to a time earlier than this when traditional Germ anic society may have been dominated more by agri­ cultural than by military concerns.8 O nly in part can this deficiency be made good by archaeology which, in analysing the material remains o f an early society, can only venture hypotheses about its social, political and religious superstructure. For the rest, we are dependent on what conclusions may be drawn from later (early medieval) sources about what preceded them. For us the most important o f those sources is language, whose conservative ability to retain something o f the past in fossilised form in vocabulary recorded later provides a hope o f undoing the limitations imposed by our sources upon our attempt to reconstruct a picture o f early G erm anic society at the time o f its encounter with Rome and in its transformation to the early Middle Ages. 5 Ibid., p. 33; Wagner (1967), 92. 8 Wallace-Hadrill (1975), 21.

6 G eary (1988), 59.

7 Wenskus (1961), 426i.

CHAPTER I

Religion

Reports by classical historians on the Germ anic tribes containing evi­ dence about the barbarians’ religious practices range from the second century b c (dealing with the Cimbri) to the collapse o f the Rom an Empire in the west in the fifth century a d , but are then continued in what the Church had to say about their paganism in homiletic literature and in decisions taken by its councils. From this wide chronological span o f historical evidence, largely from a time when the Germ anic tribes, living in a preliterate culture, left no written evidence themselves, we can expect to be reasonably well informed about the Germ anic religion. Despite the considerable difficulties in interpreting this evidence, which also recur in both the other disciplines (archaeology and philology) from which information can be extracted, it is possible to sketch an outline o f Germ anic religious practice. We start with a feature central to any religion, namely the word for the divine powers worshipped.1 To judge by O N evidence (apposite here because Christianity came so much later and because o f the rich evi­ dence o f mythological poetry) the Germ anic languages disposed o f a variety o f terms for the pagan gods, only one o f which survived by proving adaptable to Christian ends. In their O N forms these terms are: týr (sing.) and tívar (plur.), rögn and regin (two variants), bond, höpt and goþ. These terms are also attested elsewhere in Germ anic, some in a recognisably religious function. Thus týr, which in its singular form in O N is more commonly used as the name o f the god Tyr, is attested in W G in the same function in the name o f a day o f the week (OE tiwesdœg, O H G ziestag)2 and has a further parallel in Lat. dens. T h e second O N pair o f words corresponds to Go. ragin ‘decision reached by a council’ and O S reganogiskapu ‘divine decision, decree o f fate’, which suggests, as 1 Cahen (1921); Kuhn (1978), 258fr.; M arold (1992), 705.

'3

2 See below, pp. 249F.

T h e Germ anic w orld

14

with Lat. numen, a semantic development ‘decision’ > ‘divine decision’ > ‘the gods who decide’. Both bond and höpt meant originally ‘bonds, fetters’ and are attested only in this general sense elsewhere in G erm anic.3 Goþ has parallels in every G erm anic language, all with a reli­ gious function only. It is on this word that we shall concentrate, asking why it was the only term to be used throughout Germ ania for the Christian God. Etymologically, its IE origin *ghu-tó-m was a neuter past participle, designating an abstract power and therefore neuter, like all the other O N terms apart from Týr/tívar. IE evidence for the verbid stem to which this word belongs suggests either the meaning ‘to call, appeal’ or ‘to pour (a sacrificial libation)’, where the latter is less likely since what is poured is the libation, not the god.4 All the O N terms listed above were com m only used in the plural, denoting the gods collectively. W hen týr occurred in the singular it was com m only not as an appellative meaning ‘god’, but as the name o f one particular god. A part from this their collective function is clear: regin denoted a legislative body or an assembly o f those wielding power, whilst the general meaning o f bond and höpt ‘bonds, fetters’ suggests that the gods could only be conceived as such by their combination, form ing a joint network. (Their rare occurrence in the singular as a poetic term for a goddess arose exceptionally from the need for a stylistic variation.) Goþ agrees with these other terms in being employed in a collective sense, especially when in mythological literature the gods are seen in joint battle against the equally combined forces o f the giants and monsters. It differs from the others, however, in that it was also used in the singular without any forcing o f the language. Moreover, as a neuter noun it could be used o f a god (Oðin as hanga-gop) and o f a goddess (Skaði as öndurgoþ).5 Goþ therefore resembles the other terms (apart from 7 yr/twar) in being a neuter and in being used as a collective plural. It differs from them in also being employed in the singular (Tyris not comparable, being a name, not an appellative). Gender and number are therefore key factors in determining the word’s further history in the context o f Christianity. Whereas O N gop was neuter as a pagan term, the Church in adopt­ ing this word converted it into a masculine.6 Traces o f the original neuter are still detectable, however, throughout Germ ania. In Gothic the nom. sing, ö-stem gup agrees with the neuter a-stem wawrd and differs from a masculine a-stem such as dags in having no final s, characteristic o f mas3 Dronke (1992), 657fr.

4 M eid (1992), 494.

5 Cahen (1921), 28.

5 Ibid., pp. 29fr

Religion

15

culines. The position is comparable in ON: goþ agrees in its lack of ending with orp, as opposed to dagr. Masculine and neuter a-stems in Germanic also differed in their nom. plur. endings: here, too, Go. guda agrees with waúrda, but differs from dagái. In OHG, even though the plural form gota with its masculine ending is used of the pagan gods, the plural form helligot (where the lack of an ending is characteristic of neuters) occurs twice as a rendering for the pagan manes ‘souls of the dead’.7 Moreover, where the pagan implications cannot be missed, as with abgot ‘idol’, the neuter is clear beyond doubt: daz abgot (never the masculine der in OHG) and plural forms such as abgot (with no ending, as with neuters) or abgotir/abgutir (cf. the neuter lembir or wihtir).8 Although OE can sometimes inflect the word as a masculine plur. (godas) it also uses the neuter form godu for dei, daemonia. In ON, although the traces are restricted, the neuter is still to be found in afgoþ (loaned from Germany) and in certain oath-formulas where the pre-Christian form has been conservatively retained.9 The change of this word’s gender to masculine is true of all the Germanic languages after the conversion as a means of distinguishing the personal God of the Christians from the impersonal divine powers acknowledged by the pagans.10 Although the morphology of the word can remain characteristic of a neuter noun, its syntactical context often reveals its change of gender. For example, in Wulfila’s phrase gup meins the noun inflects like a neuter (no final -s), whereas the possessive adjec­ tive shows the -s of a masculine. Similarly, in ON almáttigr gup the noun is inflectionally neuter, but syntactically masculine. Other examples reveal the same change. In OHG the Isidor translators’ wording ih bim eino got shows that the noun is conceived as a masculine since the numeral eino is a weak masculine, and the same is true in OE: god āna. In OS the phrase alomahtigna god (lobôn) applies the masculine acc. sing, adjective to the noun. Throughout Germania Christianity therefore personalised the earlier conception of divine power by this change of gender. This process is taken one step further in OHG when this word for what is now seen as a personal God is used almost as if it were a name for Him. Where in this language masculine a-stems had the same form for nom. and acc. (tag), with names the acc. case added -an (Hartmuat, but Hartrnuatan). This ending was also used of ‘Christ’, seen as a name (Kristan), but also of epithets applied to Him (truhtinan, gotan). With that 7 Karg-Gasterstädt (1945), 422. 8 Ibid., pp. 426f. 9 Cahen (1921), 33fr.; Baetke (1948), 351fr.; Kuhn (1971a), 325.

10 Cahen (1921), 35fr.

i6

T h e Germ anic w orld

the OHG word for God has become as much a name as was the case with the ON god 7yr, but whereas this name is attested alongside the plural tivar and other plural terms for the gods, the OHG Christian term remains in singular isolation (the plural form gota occurs in learned trans­ lations of Lat. dii and refers to what no Christian could regard as divine). The other factor in the history of the word got was its number (singu­ lar or plural). The pre-Christian use of this word was generally in the plural, to be expected in a polytheistic religion, but the singular, if not common in ON, was possible, especially in mythological poetry of the late pagan period.11 The plural quality of this pagan word (and of the other terms for divine powers) was not stricdy plural, but rather collec­ tive in function. The collective use of tivar was imposed on it by its sin­ gular form having become the name for one particular god. The collective idea of the gods in legal assembly called up by regin meant that it could not be readily used of one god. The general usage of bond and höpt as bonds and fetters, conceived as ties or shackles, hindered their easy adaptation to singular usage. Only gop, if we can generalise from ON usage, was free for use in the singular without linguistic forcing. Our next step is to proceed from the Germanic gods to the good fortune (heil) which they were held to bestow on their worshippers or withhold from them, but also to the adjective heilag, applied to this supernatural gift of good fortune, to those intermediaries who transmitted it and to places, objects or persons which embodied it. These two words have been dealt with in detail in their Germanic ramifications by Baetke, whose argument I largely follow.12 Heilag is attested in every Germanic language, for even if the Gothic form was not used by Wulfila, it is found in the runic inscription on the ring from Pietroassa. Although the problem which for long preoccupied scholarship (did this adjective derive from another adjective or from a noun?) has been shown to be ultimately irrelevant, it would be as well to proceed from these two possibilities. The adjective from which heilag could derive is attested in Go., OHG, OS and OE and meant ‘hale, healthy’, ‘whole, intact’ (and in a Christian context ‘saved’), whilst the noun is found in OHG, OE and ON and meant ‘good luck, good fortune’ (and ‘salvation’ as a Christian term).13 The Germanic adjective represented by OHG heil is cognate with OS1. célü, likewise meaning both ‘healthy’ and ‘whole’. If we accept this 11 Baetke (1948), 367.

12 Baetke

(1942), 57fr.

13 Ibid., pp. 57t

Religion

17

as the original meaning in Germanic this still tells us nothing about the origin of the religious implications of the substantive heil or of the derived adjective heilag. Attempts to bridge this gap have been criticised for jumping from a profane category (‘whole’) to a religious one (‘holy’) and for illicitly smuggling in an unwarranted intermediate step.14 Difficulties of this kind disappear if instead we derive heilag from the noun heil: since the noun possesses a religious and magical meaning this would account for the religious function of the adjective. The OHG neuter noun heil (together with its cognates in OE and ON) translates such terms as omen, augurium and auspicium, whilst in OHG it also renders the idea of salvation (salus, salutare).15 The noun also occurs in ON as a feminine meaning ‘good fortune’, which prompts the question as to the relationship of these two meanings of ON heill. This can be answered by recalling that not merely the ON feminine, but also the OHG neuter noun could signify ‘good fortune, blessing’, as when it glosses ‘salvation’ or when Otfrid describes his religious under­ taking (i 1,113: nu w il ih scnban unser h eil) .16 The change of meaning from the abstract ‘good fortune’ to the concrete ‘omen’ is quite plausible, for in ON the feminine heill also occurs in the sense of ‘amulet’. By this shift from abstract to concrete (cf. ‘he is the pride of his family’) an omen (e.g. a bird or animal) could therefore be regarded as incorporating good fortune. The original shift produced the meaning ‘good omen’, but later ‘any kind of omen’. If the meaning ‘good fortune’ suggests an originally religious function, so too does the meaning ‘omen’. Tacitus makes it clear that the consultation of oracles was closely connected with Germanic religion: it was performed by someone with a priesdy func­ tion who prayed to the gods and looked up into the sky.17 An ON term for consulting the oracle (fella blótspán) suggests by the first element blótthat the action was accompanied by a sacrifice. In consulting the oracle it was therefore hoped that a sign bringing good fortune would be received from the gods. Sacrifice to them at the oracle was a means of gaining such fortune. All this implies that heil came from the gods, but could be withheld by them. H eil was ‘good fortune’ not in a profane or simply magical sense, but rather a blessing won by correct observance of religious ritual. It could attach itself to certain people (e.g. a particular king, leader or priest) or to certain objects (e.g. an amulet or weapon), but its origin was still supernatural. As in other archaic religions, the good or bad fortune 14 I b i d p p . 58fr.

lD I b i d pp. 6of.

16 Ibid., pp. 63i.

17 Tacitus, Germania 10.

T h e Germ anic w orld

i8

o f men was felt to depend on the favour or anger o f the gods. T h e O N feminine heill designated this abstract quality o f good fortune, while the neuter noun denoted the concrete sign in which it was made manifest. So far we have seen a clear semantic connection between the noun heil and the adjective heilag. both had a religious function. This still leaves the link between the noun heil and the adjective heil unaccounted for. T h e fact that the meaning ‘whole’, ‘healthy’ is attested for this adjec­ tive throughout G erm ania does not necessarily suggest that it was there­ fore the original one. In the first place, the abstract noun derived from this adjective (O H G heilî, O S hêli, O E hāl) m ay well be used in the sense o f ‘health’ (sanitas), but is primarily used to render salus, salutare ‘salva­ tion’ (e.g. Weißenburger Katechismus 32,88: ad aeternam salutem ci euuigeru heili).18 Secondly, a religious function is also implied for the adjective itself in various formulas o f greeting, as in Beowulf (407: was pū Hröðgdr hāl) or in Gabriel’s annunciation to M ary in Tatian (3,2: heil wis thugebono follu).19 To derive this usage from a meaning ‘healthy’ might explain some o f these examples, but not all: Gabriel is certainly wishing M ary much more than a pregnancy free from gynaecological complications. Similarly, if Wulfila (who uses háils in the context o f health) also employs it in a formula o f greeting for Gr. chaire, this is because the m eaning ‘good fortune’, hence ‘j o y ’, was still associated with his vernacular word. Thirdly, other usages show that this adjective could be used in the sense ‘salutary’, meaning not ‘health-giving’ in any profane function, but rather bringing, conveying, assuring nothing less than salvation, good fortune in the Christian sense.20Wulfila therefore used in láiseinái háilái in the sense o f ‘teaching which brings salvation’, O E hāl dön, like heil tuon in Tatian, conveys the religious sense o f salvare, whilst Otfrid employed the same adjective to express the state o f salvation achieved by baptism. These functions o f the adjective (‘conveying or granted good fortune’) can be derived easily from the meaning o f the noun (‘good fortune’), but not so easily from the other meaning o f the adjective (‘whole, healthy’). Instead, the position is likely to be the converse: if someone is blessed with heil, this will show itself in his good fortune, in his returning unin­ jured from battle or in his good health.21 This raises the possible objec­ tion that the verb heilen was used frequently in the sense o f sanare, curare, i.e. with reference to the apparently profane idea o f health, not to any religious concept o f heil. Alongside this, however, heilen is also attested in the (older?) meaning o f salvare, and the W G term for Christ as saviour 18 Baetke (1942), 69.

19 Ibid., pp. 69fr.

20 Ibid., p. 72.

21 Ibid., pp.

74fr.

Religion

l9

(OHG heilant) is best understood against the background of salvare. In OHG heilant therefore rendered salvator or salūtam, whilst in OS hêliand varied nerigend, which likewise implied salvare, not sanare. The develop­ ment of heilen from ‘to save’ (metaphysically) to ‘to cure’ is explicable by the religious and magical associations of primitive medicine, resting on the driving out of demons and evil spirits. To ‘heal’ someone meant res­ cuing him from their power by giving him back his heil. The semantic development of the verb heilen (from salvare to sanare) therefore corre­ sponds to that of the noun heilî (from salus to sanitas). In view of this close semantic connection between the adjective heil and the noun heil the initial question whether heilag derived from the adjective or from the noun loses its relevance. If heilag derived from the noun heil, it denoted a person or object standing in some sort of relation­ ship with heil, therefore ‘conveying heir or ‘imbued with heiF. Equally, if it derived from the adjective, it had the same function, since we have seen that the adjective heil is attested with the same meanings. (The suffix -ig/-ag acted as an intensifier, as with OHG ein: einac or reht: rihtig.) Heil therefore came from the gods, direcdy or through an intermediary, whilst heilag could be applied to the originator, the transmitter or the recipient of this gift from above. As regards the nature of this heil (what kind of person or object was involved?) Baetke has discussed a wide range of contexts in which it was used in Germania, but the Christian slant of our written evidence is such that we need to concentrate on pagan ON examples, even here only on early ones so as to avoid the risk of contamination with Christian ideas.22 In the ON Battle o f the Goths and Huns we learn of a ‘holy border district’ (1,8: á mörk enni helgu), where the force of the adjective can best be under­ stood from the parallel in the Landnámabók: a settler dedicates his land to the god Thor and therefore calls it þórsmörk. The territory in the lay may have been similarly entrusted to the care of the gods or it may have been regarded as hallowed by the presence of a king imbued with special heil. How someone of royal birth could be so regarded, especially in the context of victory in battle, is suggested by the description of Erpr in the Hamðismál as ‘hallowed, blessed in battle’ (28,8: gumi inngunnhelgi). As the son of a king, he is expected to have his share of heil in battle, to be victorious (as in the OHG Ludwigslied v. 57, where the ruler is designated as wīgsāliģ). Heil as a special royal prerogative is also suggested by the ON name Helgi (together with its feminine counterpart reflected in Russian 22 Ibid., pp. 130i., 136fr. (but also Dronke (1969), 175).

20

T h e Germ anic w orld

O leg and Olga), occurring earliest as a royal name in Scandinavia, long before the earliest contact with Christianity. T h e welfare o f a land (its prosperity and military success) depended on the gods, on the heil they transmitted through the person o f the king. To sum up. Like the Christian sonetus the Germ anic term heilag denoted a quality o f divine origin, but unlike sarietus the gift it repre­ sented was not to be enjoyed in the afterlife, but in the here and now. This only partial agreement was later to create difficulties in establish­ ing a Christian vocabulary in Germ anic. We pass now from the gods and the gifts they bestowed to the w ay in which their favours were won, by the act o f ritual worship, above all by sacrifices. Worship o f the gods is to be understood not as subjective reli­ gious experience, but as participation in an act o f publicly performed ritual. Accordingly, ritual rather than belief was the focus o f Germ anic religion: whereas O H G vocabulary possessed numerous terms for offerre inherited from the past, the position was quite different for credere, even if we do not accept the theory o f Gothic origin for gilouben 23 As far as the lexical evidence permits a conclusion, there was no general term for ritual worship as such, but instead separate terms for different aspects o f religious practice.2 24 3 O ne such term, attested in O H G and O E , is bigang (together with the verb bigangan).25 Its original meaning was ‘to go round, to process’ but since O H G bigang glosses ritus and cultus it must have been applied to acts o f religious worship involving a procession, such as are attested for the worship o f the goddess Nerthus and elsewhere in Germ ania. Such prac­ tices were at first anathema to the Church, but were also at times adapted to Christian use (Echternacher Springprozession), so that in modern Germ an one can still say die Felder begehen for going round the fields on Rogation days and das Fest begehen is to celebrate any festival. In O H G the term was similarly used o f any festival (Christian, Jewish or pagan), but the pre-Christian evidence for processional rites in G erm ania com ­ bines with their comparative infrequency in Christian worship to suggest a pagan origin. Rhythmic movement as part o f religious ritual m ay also be suggested by O E lāc, a word attested throughout G erm ania meaning, among other things, ‘dance’.26 Religious implications are suggested by the further meaning o f the O E form ‘sacrifice’, so that a sacrificial cer­ 23 Lange (1958), 282. 24 Wesche (1937), 26. 25 Jente (1921), 35t.; Philippson (1929), 194; Wesche (1937), 26t; Helm (1937), 59; (1953), 190, 210. 26 Jente (1921), 44fr.; Helm (1953), 192; Beck (1968a), iff; Düwel (1970), 235f.

R eligion

21

emony involving a dance would seem to lie behind O E usage. Since the meaning ‘sacrifice’ is attested only in O E , it might seem rash to postu­ late a Germ anic origin for this practice, were it not for the evidence o f the personal names O N Asleikr, Latinised O H G Ansleicus and O E Oslāc, suggesting someone who sacrificed to, or worshipped, the group o f Germ anic gods known to Jordanes as the Anses and in O N as the Æsir. T h e centre o f Germ anic ritual worship was the act o f sacrifice itself, including blood-sacrifices and bloodless ones, the former embracing animals and human beings, the latter, for example, the weapons and other gear o f the defeated enemy.27 Although historical testimony and archaeological finds can be confirmed by linguistic evidence, the terms to be adduced under this heading, both in O H G and O E , were slowly driven back and replaced by words like O H G opfarôn and O E offrian.28 T h e most important sacrifice was o f an animal some parts o f which after slaughter were offered up to the gods and others consumed ritually by the worshippers who thereby enjoyed communion with their gods.29 Although the historical evidence for this is slight (e.g. Tacitus, the m artyr­ dom o f Saba amongst the Goths, Christian reports on the W G pagans),30 the archaeological evidence occurs more frequently (bones o f sacrificed animals, particularly horses), as long as we feel enough confi­ dence in the archaeologists’ ability to distinguish natural death from sac­ rificial death and to maintain in the latter case that the flesh was consumed ritually. I f we can accept this it provides valuable support for the linguistic testimony. Hum an sacrifice, like animal sacrifice, can be assumed archaeologically when skeletons are found on what other evidence suggests were religious or sacrificial sites. Historical evidence supports this assumption, not merely in the context o f warfare (sacrifice o f prisoners o f war).31 Already from Tacitus we learn not merely that M ercury (=Wodan) was worshipped above all, but that human sacrifices were made to win his favour. T h e account o f the processional worship o f the goddess Nerthus also concludes with the ritual killing o f slaves who took part in the cere­ mony.32 Scandinavian examples mention human sacrifices to O ðin outside the context o f warfare.33 Since danger was to be averted by 27 Düwel (1970), 238t. 28 Philippson (1929), 195; Wesche (1937), 62, 71. 29 Helm (1937), 556; (1953). '97, >986; de Vries (1956a), 4o6f.; Jankuhn (1967), 119, 129, 143, 145; M uch (1967), 176fr.; Polomé (1986), 276. 30 Tacitus, Germania 9; Saba: Helm (1937), 55f.; Thom pson (1966), 64h; W G pagans: Helm (1953),

197fr 31 M ogk (1909), 6oiff.; Helm (1937), 56; (1953), 1996; de Vries (1956a), 408fr.; Jankuhn (1967), 120, 132; Beck (1970), 240fr. 32 Tacitus, Germania 9 and 40. 33 M ogk (1909), 614.

22

T h e Germ anic w orld

sacrificing someone else this practice may have come particularly to the fore in time of battle, but need not have been confined to it. The OHG word bluozan ‘to sacrifice’ (noun: bluostar) has cognates throughout Germania except in OS.34 It occurs in all OHG dialects, but must have died out in all of them at about the same time (probably because of the victorious advance of opfaron), since it is not attested after about 800. Although the word is probably cognate with Lat. fla(d)men ‘priest’ the etymology is too uncertain to help us with the original func­ tion of the Germanic word, although its loan into Finno-Lappish terms for ‘to sing’ could suggest a sacrificial rite accompanied by sung recital.35 In OHG the verb rendered Latin terms like sacrificare, immolare, libare and victimare, suggesting various forms of sacrifice (libation, burnt offering, animal), but this wide range may result more from the variety of Latin words which the vernacular term translated only approximately. Less questionable is the pagan flavour of the term, for it glosses idolatriae and occurs in the Franconian baptismal formula with regard to the pagan practices to be abjured. In OHG and OE some traces of (pagan) sacri­ fices are still preserved, but in Gothic the word blötan has been safely christianised by a change of construction: it no longer governs the dative (‘to sacrifice to a god’), but the accusative with a different meaning (‘to worship God’), so that gupbhstrds renders theosebēs ‘worshipping God, devout’.36 Wulfila therefore retained no trace of worship by means of sacrifice and has extracted the word from its pagan origins. A unique sacrificial term was christianised in the form of OE blētsian ‘to bless’.37 Its dialectal variants blēdsian and bloedsia suggest derivation from *bl5dison, a verb formed from blöd ‘blood’, which implies blood sac­ rifice (animal or human) and possibly consecration of a pagan altar by sprinkling it with the blood of the victim. The archaeological evidence for animal sacrifice suggests a division between those animals which could not be consumed in a ritual meal and those which could be, as part of the communion between gods and worshippers. Some of our linguistic terms support this idea of sacrificial food. The OHG zebar is attested in all dialects and its cognates are found in most Germanic languages.38 In OHG the word glosses sacrificium, but also hostia, victima, holocaustoma, where the latter terms imply the sacrifice 34 Wesche (1937), 39ff.; Helm (1953), 194; Diiwel (1970), 226F.; Polomé (1986), 287. 35 Wesche (1937), 4 4 36 Ibui’ P- 4337 Jente (1921). 4 1; de Vries ( '956a)> 4 ' 5 38 Jente (1921), 42fr.; PhUippson (1929), 195; Wesche (1937), 45fr.; Helm (1937), 58; (1953), 196; de Vries (1956a), 423; Diiwel (1976), 227.

Religion

23

o f animals. Again, this could be more a reflection o f various Latin terms imperfectly translated by one G erm an term, but other factors speak against this. T h e Germ anic word is cognate with Latin daps ‘sacrificial m eal’, but also with Arm enian tvar ‘herd o f cattle’ or ‘ram ’ . T h e word was also loaned into Finnish: teuras ‘animal for slaughter’. Furthermore, the idea o f an animal fit for communion consumption is emphatically suggested by the negative o f the O H G form, preserved in M H G unztfer, ungezibere, modern G erm an Ungeziefer ‘vermin’. O n ly what could be sac­ rificed for a ritual meal was regarded as zebar, everything else (unclean and inedible) was ungezibere. Another term with similar implications, Gothic hunsl, is also found in O N (the runic inscription o f Rök) and in O E , but not in O S or O H G .39 Wulfila employed the term in the general sense ‘sacrificial offering’, but also in a context where the N ew Testament meant specifically ‘sacrificial food’, so that the idea o f animal sacrifice was presumably still present in Gothic. This possibility is confirmed by the O E counterpart hūsel, orig­ inally conveying the sense ‘sacrifice’, as in Gothic, but by its Christianisation to ‘communion’ implying that this had been combined with a ritual meal. If the O N term later also acquires this Christian meaning, this is likely to be under O E influence, but by means o f loan­ meaning, not as a loanword, since the example from Rök shows that the word was known in Scandinavia from early on. A last example o f animals sacrificed for religious purposes is provided by Gothic sáuþs ‘sacrifice’.40 Its conjunction with hunsl, but especially with alabrunsts ‘burnt offering’ suggests an animal sacrifice, while the par­ allel o f O N sauðr .‘sheep’ could mean that this was the animal in ques­ tion. However, the equation o f sauðr with blótnaut ‘sacrificial cattle’ could imply a more general usage, but still the sacrifice o f an animal. O ther G erm anic languages retain no trace o f this word as a religious term. Instead, they use the related verb *seuðan as a general term for ‘to seethe, boil’, suggesting the manner in which the beast may have been prepared for a ritual meal. From worship o f the gods and sacrifice to them we move to the location where this was performed, to the place o f worship. Tacitus reported o f the Germ ani that their holy places were in the open, in woods or groves, since they did not regard it as consistent with the majesty o f the gods 39 Philippson (1929), 195; Helm (1937), 57; (1953), 195; Düwel (1970), 228t 40 D e Vries (1956a), 370; Düwel (1976), 227, 232Í

24

T h e Germ anic w orld

either to represent them in human form or to imprison them within the walls o f a building.41 Although it has been suggested that Tacitus made use here o f a topos o f Stoic philosophy (the superiority o f religious worship not tied to concrete forms in any way), this is a case where an interpretatio Romana need not simply have been imposed upon Germ ania, but could have been used to understand a given feature o f Germ anic religion. In other passages where he was not concerned to make this general point about their worship, but was reporting on specific aspects o f a variety o f Germ anic tribes as learned from different informants, Tacitus made numerous allusions to the religious awe in which the Germ ani held woods and groves.42 There was a sacred grove for the wor­ shippers o f Nerthus, the Suebi gathered for their sacrifices in a hallowed wood, the figures and emblems which they took into batde were kept in their sacred groves, Civilis summoned tribal chieftains for consultation in a hallowed grove, and near the Weser there was a forest consecrated to the god whom Tacitus equated with Hercules. O n ly on the improb­ able assumption that on all these scattered occasions Tacitus was impos­ ing a topos o f classical thought on his Germ anic material can we ignore this variegated evidence, the more so since it is confirmed from quite a different angle by later Christian references to pagan worship at similar sites.43 Some o f these references may admittedly be conventional (apply­ ing to all pagans at any time or place what had been observed only in one case), but their frequent mention o f the religious importance o f trees and woods to Germ anic pagans, agreeing with Tacitus, is borne out by some o f our linguistic evidence. Confirmation is also provided by the finds o f archaeology, as far as these go.44 T hey support Tacitus to the extent that temples (buildings for religious purposes) do not appear to have existed, whilst the focus o f worship on woods and groves is borne out by excavations o f sacrificial sites, fenced o ff against profane use, in moorland (and presumably also in woodland). Richer information on pagan places o f worship is forthcoming from early place-names which survived the coming o f Christianity. Placenames which can assist us are o f two types, each composed o f two ele­ ments: the first type contains the name o f a god as its first element (theophoric place-names), the second has a term for a place o f worship as its second element.45 (We may deal now with the first type only, since 41 Tacitus, Germania 9; Tim pe (1992), 452, 483. 42 Tacitus, Germania 7, 39 and 40; Historiae iv 14; Annales 11 12; de Vries (1956a), 351i. 43 Helm (1953), 167fr. 44 Jankuhn (1966), 421. 45 Helm (1953), 174.

R eligion

25

the second presupposes words for a place of worship which we have yet to consider.) Most of the evidence, but by no means all, comes from Scandinavia, which raises the often unanswerable question as to how far these NG examples can be generalised for Germania at large. Theophoric place-names may contain a general term for the gods at large or the name of an individual god.46 Scandinavian place-names of the type Gudhjem, Gudme, Gudum (< Gudhem) derive from an appellative meaning ‘home of the gods’, extended to mean the place where the gods dwell and are the object of a special rite, then designating places where such worship took place. In addition to this collective word for the gods, the term *ansu- (one group of gods within the Germanic pantheon) can also be used, as in Sw. Asum (< -hem), Norw. Oslo and German Aßlar. Much more common are place-names which specify an individual god, throughout Scandinavia and especially with reference to the gods Freyr, Oðinn, þórr (but also Ullr, Freyja and 7yr). This category is also represented in WG evidence: Wodan in Germany (Godesberg < Wudinisberg) and in England ( Wednesbury, Wensbw), Donar in Germany {Donnersberg, whilst in England Thursley shows Viking influence), Tiu in Germany (Ciesburg for Augsburg) and in England {Tuesley)A1 For all such place-names of theo­ phoric nature (generic or individual deities) we may assume the presence of a religious site in or near the places in question. The problems of interpretation raised by such place-names concern both the second element (how far is a place of worship really involved instead of a term with no religious function?) and the first element (is a deity necessarily referred to?). As an example of the latter difficulty we may take what have been regarded as non-Scandinavian parallels to such a theophoric place-name as Norw. Freysaker, where similar names such as Óðinsakr, Ullinsahr, frjarðarakr suggest the presence of an individual god.48 The position is by no means so clearcut outside Scandinavia, however, with such place-names as Fri. Franeker and Du. Vroonloo. Not merely would regarding these latter names as theophoric involve assum­ ing the existence of a WG god Frô for whom we have no other reliable evidence, it would also ignore the persistence of the non-religious meaning of the wordfr ô (‘lord, ruler’) which in Scandinavia came to be used as the name of a god. In view of these reservations the WG exam­ ples probably refer not to a pagan god, but to the property of a Germanic or even feudal lord. Apart from the qualifications which are called for, onomastic evidence of this kind can enlighten us on the location of 46 Andersson (1992), 524fr., 527fr

47 Helm (1953), 175fr

48 Green (1965), igf.

26

T h e Germ anic w orld

places o f religious practice and on the worship o f particular gods, even though most o f the conclusions are confined to Scandinavia. In addition to place-names the linguistic evidence includes a number o f appellatives denoting a place o f worship. Some o f these occur as the second element o f one o f the place-name types just considered and most o f them imply that the site was an open-air one or in woodland. Without being dependent on Tacitus the argument can confirm him. T h e O H G word harug has cognates in O E and O N .49 In O H G it occurs in the earliest sources and persists until the tenth century, trans­ lating such Latin terms as ara, delubrum, fanum, lucus and nemus, i.e. ranging from ‘altar’ through ‘sanctuary, chapel’ to ‘holy grove’ . None o f the examples suggests a large place o f worship, but rather a small grove, particularly when harug renders lucus or nemus or when in one gloss it occurs as a translation variant o f forst. N o G erm an example shows harug as a translation o f silva (with a non-religious function), but the position is different in O E , where hearg can denote ‘w ood’ as well as ‘holy grove’, but in addition a pagan ‘sanctuary’ or ‘idol’. In O N hörgr is attested as ‘heap o f stones’, but also ‘place o f worship’ (with stones as a primitive altar?) and ‘mountain’. If we accept the etymological connection with OIr. cam ‘heap o f stones’ and Lat. career ‘(stone) enclosure, prison’, the semantic starting-point would seem to have been ‘altar’, then, as with the Latin terms above rendered into O H G , moving to ‘sanctuary, chapel’ and finally ‘holy grove’. T h e word can also occur as (or in) a place-name (e.g. Harrow or Harrowden in England), but the double func­ tion o f the appellative, religious or non-religious, means that we cannot be certain whether the place-name refers to an original place o f worship. T h e position is largely comparable with O H G lôh, O E leak.50 These words are cognate with Lat. lucus ‘open clearing, holy grove’ and in agreement with Latin the Germ an form means ‘clearing, holy grove’, whilst English proceeds from ‘open ground’ to ‘m eadow’ . Although infrequently, the religious function is still attested in O H G , but with the evidence from place-names it is again difficult to tell whether they refer to an earlier religious site (perhaps most probably in the case o f Heiligenloh). T h e word nimid is attested only once, in the Indiculus superstitionum where it refers to religious rites performed in the woods {De sacris silvarum, 49 Jente (1921), gff.; Philippson (1929), i86f.; Wesche (1932), 17fr.; de Vries (1956a), 376i., 380; Andersson (1992), 528i.; Holm berg (1992), 541i. 50 Wesche (1932), 25i.; Helm (1953), 169; de Vries (1956a), 352.

Religion

27

quae nimidas vocant).51 Although the Indiculus has been suspected o f simply transferring features o f late antique paganism to northern Europe it is now considered more probable that it was geared specifically to the conversion o f the Saxons and that nimidas is in fact an O S word.52 There are no other Germ anic cognates, but in Celtic nemetos meant ‘holy’ and nemeton ‘sanctuary’, so that we are possibly dealing with a Celtic loan­ word into O S .53 T h e word’s ultimate relationship with Lat. nemus con­ firms the placing o f these religious rites in a wood. T h e O H G word baro translates Latin ara in a sacrificial context, while its O E counterpart bearo, although normally used for ‘wood’ in a non­ religious context, occurs once with the meaning ‘sanctuary’.54 Moreover, this religious meaning is suggested for the O E word when Christian buildings are said to be erected -mm-), yielded M H G sam(e)ztag. How far this Gallic explanation for this G erm an word needs to be com ­ bined with a Gothic explanation o f the same form in Bavarian is quite uncertain, if only because a derivation from eastern Europe is not the only way o f accounting for sambatum instead o f sabbatum as a base.75 (Gothic remains true to its terminology o f Greek origin in this case, too: sabbato means both ‘Sabbath’ and ‘week’,76 for which there are no par­ allels in the western vernaculars.) Although England and north-western G erm any borrowed Saturni dies, they also share a term which, by contrast, is specifically clerical in origin: O E sunnanifen, M H G sun(nen)âbent.77 This corresponds to Lat. dominica vespera and, in meaning the evening before Sunday, is a Christian equiv­ alent o f the Jewish use o f the word for ‘evening’ to mean ‘Friday’, the 71 73 74 75 77

See below, p. 310. 72 See below, pp. 310i. Frings and Niessen (1927), 292i., ßoif.; GR 2, 444ÍF.; Avedisian (1963), 246i. Frings and Niessen (1927), 292f.; GR 2, 444fr.; F E W 2, 3i.; Avedisian (1963), 232fr See below, pp. 3i2f. 76 O nce only: Le. 18.12. Kretschmer (1918), 460fr.; Avedisian (1963), 242fr.; GR 2, 445h

N am es o f days o f the week

253

eve o f the Sabbath. T h e word was not originally used as a substitute either for sœterndœg or for sambaztag. It denoted the last part (following vespers) o f the day before Sunday, not the whole o f the last working day: in this sense sunnûn âband is used by Otfrid (v 4,9). In this restricted meaning the word could be employed in areas where either o f the two terms for ‘Saturday’ was known. W here these terms were not in use, however, sunnenâbent came to stand for the whole day preceding Sunday, first when the Sunday in question was a feastday (so that mass was cele­ brated in the morning o f the preceding day), then applied to any Saturday. T h e origin o f this last term for Saturday, in which Germ any m ay well have followed the English impetus, is thus clearly Christian. However varied the range o f linguistic possibilities for naming days o f the week may have been in Germany, a distinctive pattern has begun to emerge. In the first place, many o f these names suggest a borrowing from northern Gaul and the area o f the lower Rhine (Sunday, Tuesday, Wednesday, Thursday, Saturday). As distinct from this, another group­ ing is discernible in south-eastern G erm any or in the south at large. W hat the examples o f this group all have in common is that, by one method or another, they avoid any reference to the pagan gods whose names occurred in the group from the north-west.78 78 O n the implications o f this contrast see below, pp. 359i., 368f.

C H A P T E R 14

T h e vocabulary o f writing

In this chapter the three constituent factors o f the Middle Ages which provide the framework for this book come together. In their adaptation to the needs o f a barbarian society the runes can be regarded as a typically Germ anic form o f writing, but in the acquisition o f literacy by G erm ania they have to be seen in connection with two other factors: with Rome (whose use o f writing made a further contribution to European literacy) and with Christianity (whose status as a religion o f the book brought it into collision with the background o f pagan magic to runic practice). Writing was therefore introduced to G erm ania from three different sources (all o f Mediterranean origin)1 and it is the interplay between these three, as reflected in the linguistic evidence, which will now engage us. W hen precisely the runes, a mode o f writing derived from a northern Italic script, were first adopted is uncertain: some have argued for a date around 100 b c , others have suggested ad ioo .2 In either case, they were taken over at about the time when classical sources begin to provide us with our first detailed information about the Germ ani. This does not mean that Germ ania, in anything other than a technical sense, can be regarded as a literate society when we begin to learn more about it, for runic practice is not to be seen on a par with writing as practised in the Mediterranean world. T h e runes were not used for purposes o f every­ day communication as was the alphabetic script by the Romans, for they were employed in magic and divination, and were therefore tied up with the Germ anic religion, they were largely confined to short, terse inscrip­ tions and knowledge o f them was restricted to a small, closed circle o f rune-masters.3 1 Green (1994), 35. 2 Rosenfeld (1952), 194; Klingenberg (1973), 56. 3 Klingenberg (1973), 56, 96fr., n iff.; Ebenbauer (1981a), 36; Nielsen (1985), 75fr.; Düwel (1988), 70fr.; M üller (1988), h iff. Even Page (1995a), i05ff., despite his scepticism about Anglo-Saxon runes and magic, has to admit some cases.

254

T h e vocabulary o f w riting

255

It is difficult to determine what words were early used o f runic writing in Germ ania, but the best approach is to begin with terms used in runic inscriptions themselves, supplementing them with what other (nonrunic) sources m ay tell us.4 T h e widest range o f such terms comes to us from Scandinavia, where the persistence o f paganism and the large number o f runic monuments amply account for this. However, despite the disparity between Scandinavia and the rest o f Germ ania, it is safer to regard as common Germ anic only those runic terms used both in Scandinavia and elsewhere in the Germ anic world. W ith this restriction we are faced with an unquestioned basic list o f four words, for each o f which a general, non-technical usage needs to be distinguished from a specialised runic function. T h e first word, *rûna,b has a general meaning attested throughout G erm ania with the following range: Go. rūna ‘(closed) consultation, deci­ sion’, O N run ‘mystery, secret wisdom’, O E run ‘deliberation, mystery’, O S runa ‘council, discussion’ and girûni ‘secret’, O H G rûnôn ‘to whisper’ . In its specialist usage the word occurs in O N , O E and O H G , meaning a ‘rune’ or ‘runic writing’. Outside Gm c. the word was borrowed into Finnish, where it denoted first a ‘magic song, incantation’, then a poem o f any kind, whilst Celtic cognates (We. and OIr.) mean either ‘secret’ at large or, more specifically, a secret which in its written form is not access­ ible to all.6 T h e attestation o f the Gm c. word in its technical usage, applied to runic writing, outside Scandinavia in two W G languages sug­ gests a C G ancestry for runic use. As a term belonging to the vocabulary o f writing it was only used o f the runic script, never o f the Rom an alphabet. Outside this context, however, the word could later be used safely by the Church in a general sense: Go. rūna rendered ‘mystery’ (of the kingdom o f God) in Me. 4.11, while O H G girûni translated sacramentum.1 Semantically this word seems therefore to have developed from ‘low-voiced, secretive speech’ (cf. modern Germ an raunen) to ‘secret, secret knowledge, m agic’ and finally to a ‘mysterious sign, character’ used in magic practice. T h e second word, P G *stab-,8 occurs in the earliest written texts in O N , O E and O H G with the general meaning ‘staff, but in the special­ ised sense ‘stave, runic sign (carved on a stave)’ only in O N (already about 620 in a runic inscription from Gummarp). This restriction o f the technical usage to Scandinavia only apparently argues against a C G 4 Kuhn (1938), 55. 5 Ibid., pp. 55,58, 67; Arntz (1944), 294t; Ebel (1963), 82fr. 6 Ebel (1963), 82Í 7 Starck and Wells (1990), 218. 8 Kuhn (1938), 55,58; Arntz (1944), 282; Ebel (1963), 86fF.

256

Contact w ith the non-G erm anic w orld

employment o f the word for runes, however. T h e word stcefis applied in O E to the letters o f a Mediterranean script (Greek or Latin, designating book learning at large),9 which suggests that before the introduction o f the Rom an script to Anglo-Saxon England this word had been in use, as in Scandinavia, for the earlier system o f writing. Even more telling is the distinction made in O E between the compounds rmstœf1,r unic sign’ and bðcstæf1,letter o f the Roman alphabet’.10 This distinction became neces­ sary only after the Latin script was introduced alongside the runic one, but presupposes the earlier use o f the word stcefmeaning a ‘sign used in writing’. This O E distinction is also found in O H G rûnstab and buohstab which, given the absence o f the simplex in Germ any to denote a ‘letter’ (by contrast with England), suggests imitation o f the O E precedent in the context o f the Anglo-Saxons’ mission to G erm any and the influence o f their scriptoria on the continent.11 A comparable distinction, also based on the O E model, is found in Scandinavia, once writing in Rom an letters was introduced there in connection with the Anglo-Saxon mission to the north. Here the variants bókstqfr and latinustqfr point to a contrast between writing in Latin in Roman letters (for ecclesiastical purposes) and writing in the vernacular in runic characters (mnastafi). (The forms böcstæf, buohstab, bókstqfr have no counterpart in Gothic, presumably because they were coined late, after the transition o f the first element o f the compound from meaning ‘writing tablet’ to ‘parchment codex’.)12 With the abandonment o f writing in the runic script O H G buohstab came to designate littera at large, it stood now for the only written char­ acter in use and replaced the simplex stab from which it had derived. Although Gothic shows no counterpart to buohstab, it is dubious whether we can assume the meaning littera for Go. stafs, used to render Gr. stoichmn, Lat. elementum,13 It has been suggested that the Greek word sometimes indicated the element o f a sentence, hence a ‘letter’, so that Wulfila’s translation by stafs would show that this word must have been a Gothic term for a written character, originally a rune. Against this it can be argued that stoichmn denoted a simple sound o f the voice, the first element o f speech, rather than necessarily a letter as an element o f writing. Furthermore, another meaning o f stoichmn was a ‘small upright pole’, close enough to the general (non-runic) meaning o f the Gm c. word for Wulfila to be able to imitate Greek by a loan-meaning. It is also 9 10 11 13

Bos worth and Toller (1954), 907. Kuhn (1938), 6of.; Ilkow (1968), 67i.; Hoops (1913b), 349^; Ebenbauer (1981b), 87i. See below, pp. 341fr. 12 See below, pp. 26of. Kuhn (1938), 56; Ebel (1963), 89; Ebenbauer (1981b), 88.

T h e vocabulary o f writing

257

revealing that he nowhere translates grâmma ‘letter’ by stafs, but uses bāka instead. None o f this suggests that stafs in Gothic designated a written character. O u r third term for consideration, *writan,H is used with the general meaning ‘to tear, cut, scratch’ in O E , O S and O H G (with the cognate rista attested in ON), but with the specialised meaning ‘to cut, incise runes’ in runic inscriptions in O N and O H G . In addition, it occurs as an overall term for ‘to write’ in O E , both in the runic script and also, over­ whelmingly, with letters o f the Roman alphabet. This term is the second one used in its technical runic sense outside Scandinavia, where it was employed in runic inscriptions o f the sixth and seventh centuries, even though it was later ousted by rista. In O H G the word is used in the runic inscription on the fibula o f Freilaubersheim (second half o f the sixth century), but outside that context the verb or a derivative noun occurs only rarely in the sense o f writing, e.g. to gloss characteres litterarum or notece meaning ‘written characters’.14 15 Otfrid employs the verb twice (iii 17,36: in erdu tho mit themofingare reiz =Joh. 8.6; iii 17,42 = Joh. 8.8) in the context o f Christ writing with His fingers on the ground. From this restricted use we may conclude that the idea o f writing by scratching or making an indentation (as at first with runes carved on wood, bone or metal) may still have attached to the w ord.16 In Gothic the noun writs occurs (Lc. 16.17), but only in the restricted metaphorical sense ‘tittle, iota’ (deriving from the m eaning ‘dash, point’). T he word thus clearly existed in Gothic, but was not used by Wulfila as a specialised term o f writing. To sum up: only in O E did the term wntan, formerly used o f runes incised into a material, come to be used at all regularly o f writing at large, including the use o f the Latin script on parchment. In thus expanding the applica­ tion o f this term the Anglo-Saxons repeat what they did with staf and what is also the case with the next term. This fourth term, PG *reðan,17 is attested in its general sense in all the Gm c. languages with a range o f meanings from ‘to determine, order, take care o f ’ through ‘to advise’ (modern Germ an raten) to ‘to interpret’ (modern Germ an erraten). In its technical runic sense the word occurs only in Scandinavia, where O N ráða means not merely ‘to interpret’, but ‘to read’ runes: ráða mnar, no doubt assisted by alliteration, occurs on a number o f rune-stones.18 Outside Scandinavia, however, the position is similar to what we saw with *writan. Although there is no evidence for 14 Kuhn (1938), 55, 58, 71fr.; Arntz (1944), 284; Ebel (1963), 14fr. 15 Arntz (1944), 284, fn. 5; Starck and Wells (1990), 489. 16 Schröder (1924b), 58. 17 Kuhn (1938), 56, 58, 68; Arntz (1944), 294; Ebel (1963), 73fr. 18 Ebel (1963), 74E

258

Contact w ith the non-G erm anic w orld

the use o f O E rædan in connection with the runic script, the fact that it is regularly employed o f reading the Rom an script implies that it had earlier been used for reading runes by the Anglo-Saxons. Even without further evidence from other Gm c. languages it is probable that this was another technical runic term, earlier in use beyond the north. T h e conclusions to be drawn from these four technical terms are clear. To start with, the position o f O E is worthy o f attention, since it is the only Gm c. language to apply three o f these runic terms regularly to using the Rom an alphabet (stœf, wntan, rædan). O n the other hand, O H G uses rîzan very rarely and stab only in two compounds which it adopted from O E , so that it largely, if not entirely, failed to adapt runic terms to Latin texts, as O E had done. In this O H G is eventually followed by O N .19 A t the opposite extreme to O E Gothic stands alone, for none o f these terms is attested for Gothic runic inscriptions (which may be because o f their rarity), but none was used by Wulfila for other forms o f writing, including that devised by him for the needs o f Gothic, where an opportunity did present itself.20 We come now to what m ay be called the post-runic vocabulary o f reading and writing in Germ ania. This can be divided into two separ­ ate parts, one antedating the com ing o f Christianity to northern Europe and the other closely connected with the new religion. O n ly one pre-Christian example has been seriously entertained, but one central to the rise o f literacy. Although it has been claimed that the word ‘book’ is yet another example o f a runic term adapted to a new writing practice, the general consensus now sees it rather in connection with the Rom ans’ use o f writing, brought with them to G erm ania either by their arm y or by merchants. T h e Rom an arm y was dependent on a regular flow o f written information and on the transmission o f orders over long distances, there is evidence for a correlation between rank and literacy in the army, with certain offices in the legions requir­ ing educated, i.e. literate soldiers.21 For Rom an merchants literacy and written records becam e necessary with the growing com plexity o f trade, especially extending over more than just local transactions. In such circumstances writing could take various forms: names or other details scratched on an amphora, the use o f a writing tablet for short messages or even longer communications.22 It was the simple and 19 Kuhn (1938), 64. 20 Ibid., pp. 64f. 21 Rosenfeld (1952), 20iff.; (1956a), 262, fn. 3; Harris (1989), i66f., 253fr., 293F. 22 Rosenfeld (1952), 201, 205; Harris (1989), 18, 202, 286.

T h e vocabulary o f w riting

259

adaptable writing tablet which influenced the vocabulary o f writing in Germ ania. T h e word bāka, employed at first usually in the plural, has the meaning ‘book’ or ‘written text’ in all the Gm c. languages, which o f itself points to the early use o f this word in the context o f writing.23 In Gothic alone, however, the word is also used to denote one single letter in a written text (II Cor. 3.6 and Rom. 7.6). It is generally agreed that there is a connec­ tion between this word and the related Gm c. term for ‘beech’ (O N boh, O E hoc, O H G buocha), so that the name o f the particular wood used served to denote an object employed as a surface for writing. This has its parallels elsewhere in Gm c. (OE asc ‘ash tree’ and ‘spear’, lind ‘lime tree’ and ‘shield’),24 but also, in connection with writing, in the classical lan­ guages (Gr. bibbs ‘inner bark o f the papyrus’ and ‘book’, with the same two meanings attested for Lat. liber).25 O u r problem with Gm c. bāka is to decide what kind o f object made o f beech-wood was meant. Some have argued that this word, when used as a term for writing, originally denoted a beech-stave with a rune carved on it for casting lots (as described by Tacitus, Germania 10), then the runic sign itself.26 Others see the word as denoting a flat board carved from this wood and used as the writing tablet employed by Roman officers or merchants.27 Accepting the first explanation means adding bāka to the runic vocabulary we have been considering, whilst according to the second explanation the word would be o f post-runic, but still pre-Christian, origin. In deciding between these two possibilities a difficulty has always lain in the anomaly o f Wulfila’s use, unique within G m c., o f the singular bāka meaning a single written character (Gr. gramma, Lat. littera), a usage which is confirmed by the similar function o f SI. buky, thought to be loaned from G othic.28 T he first explanation just referred to saw it as an obvious development for Wulfila then to use the plural bākās to mean a consecutive text (Gr. grámmata, Lat. litterae) or book. Conversely, so it was argued, if the plural bākās had meant ‘writing tablets’ or a ‘book’ or longer communication made up o f several tablets, it would be difficult to explain how a singular word meaning ‘writing tablet’ (or ‘beech tree’) could be used for a single letter. From this the conclusion was drawn that Wulfila made use o f a word originally denoting a beech-stave with a rune carved on it, then a runic sign, to translate in the singular gramma or 23 Hoops (1913a), 338F.; Kuhn (1938), 59fr.; Rosenfeld (1952), 193fr.; Ebel (1963), 95^ Ebenbauer (1981a), 35fr. 24 Hoops (1913a), 339; Ebenbauer (1981a), 35. See above, pp. 6gf. 25 Hoops (1913a), 337; Ebenbauer (1981a), 35. 26 Ebenbauer (1981a), 36. 27 Rosenfeld (1952), 193fr. 28 Kuhn (1938), 59; Rosenfeld (1952), 196.

26o

Contact w ith the non-G erm anic w orld

littera, but in the plural to denote a consecutive text or book, and that the word in this latter meaning was transmitted to W G as a term o f the Gothic mission.29 About this explanation doubts are justified. I f Wulfila did make use o f a native term from runic vocabulary in the context o f a new script which he evolved from Mediterranean models for a Christian purpose, this would be the only example o f a policy which Gothic, in sharp distinction from O E , otherwise shunned. In addition, we shall see that the influence o f the Christian vocabulary o f Gothic m ay have reached Bavarian (geographically nearest to the Christian Goths in the south-east), but other Germ an dialects only rarely and O E not at all.30 T h e wide spread o f the word ‘book’ in W G thus presents a major difficulty for a theory o f its Gothic origin. By concentrating on Wulfila’s use o f the singular bāka to denote an individual letter this explanation ignores another singular use o f the word which is even more widespread in Gm c. and hence provides a firmer starting-point. There is evidence to suggest that the object made from beech-wood, easily split into thin slices, which was termed bāka was not a rune-stave, but rather a writing tablet (Lat. tabella or pugillaris) covered in wax to provide a writing surface. T h e word denoting this common Roman implement conveyed in the singular the idea o f a short message or ‘document’, in the plural it denoted a number o f tablets bound together for a longer communication. W ith the introduction o f Christian writing practice, using parchment pages and a bound codex, the word bāka was then used o f the book or codex: consecutive pages replaced a group o f tablets, and the bound covers o f the book (still often o f beech-wood) resembled the outer covers o f two or more tablets put together for sending. T h at the word bāka could have originated from a term to render a writing tablet is suggested at three different points in Germ ania. In the O S Heliand the word bôk (w. 232, 235) is used in the scene where the Vulgate (Lc. 1.63) has Zachariah, struck dumb, write the name o f his son John on a writing tablet (pugillaris). In O E the transition o f this word’s meaning from tablet to parchment is already registered when bāc (in addition to its meaning ‘book’) also denotes a shorter document or ‘charter’ (bācland is therefore land held by written charter).31 Finally, the Ostrogothic D eed o f Arezzo contains the word frabaúhtaböka ‘document o f sale’, where the word is used in a sense comparable to O E bāc.2,2 T h e 29 Gothic mission: Kuhn (1938), 65; Arntz (1944), 281. 30 See below, pp. 308fr. 31 Bosworth and Toller (1954), 113, 114. 32 Streitberg (1919), 480.

T h e vocabulary o f w riting

26i

spread o f these examples through Germ ania suggests that bāka ‘writing tablet’, then ‘short communication’ (short enough for one tablet or, after the transition to a different material for writing, one parchment docu­ ment) was C G . T h e Ostrogothic testimony heightens the possibility that Wulfila’s use o f the word for one individual letter (gramma, littera) was peculiar to his ecclesiastical vocabulary and cannot be taken as the start­ ing-point for what is found elsewhere in Germ ania. Two linguistic details lend support to this view.33 First, the earlier prac­ tice with writing tablets is reflected in a common construction for reading or writing books. In the case o f a bound codex we might expect either verb to be followed by the preposition ‘in’ (‘to read in a book’), whereas a more common construction with ‘on’ refers back to consult­ ing a tablet: O N ríta á bókum, O H G lesan ana buohhum, Go. gakunnan ana bākām. Secondly, the use o f this word for ‘book’ in the plural, as these examples indicate (only later was the O H G neuter plural, diu buoch, replaced by a singular), reflects the fact that at an earlier stage in the history o f literacy any lengthy communication in writing was transposed to a number o f writing tablets (bākās in the plural) bound together. These two grammatical details therefore survived the technical change from one medium to another. T h e use o f the plural for ‘book’ also throws light on Wulfila’s anom ­ alous singular with a different meaning (‘letter’). His plural employment o f bākās agrees with the position elsewhere in Germ ania, but it also agrees with the use o f the plural o f the word for ‘letter’ applied to a lengthy written text in both the classical languages on which his Gothic bible was based: Gr. grámmata, graphái, Lat. litterae.u In other words, if the Gothic plural bākās corresponded to the Greek plural grámmata, then it is understandable that Wulfila used the singular bāka to render the Greek singular grámma, especially since the Goths now occupied territory beyond the limits in which the beech grew, so that any semantic confu­ sion in using the singular was avoided.35 Far from reflecting past prac­ tice with native runes, Wulfila’s singular bāka instead illustrates his encounter with the terminology o f writing in the classical world. G erm ania therefore learned o f Roman writing tablets and designated them with a word common to Gothic and to W G, but this raises the question how this word spread throughout G erm ania in this sense. If we fall back on polygenesis and assume that W G acquired knowledge o f 33 Hoops (1913a), 33gf. 34 Ebenbauer (1981a), 36. 35 Rosenfeld (1952), 196. See also Hoops (1913b), 342fr. (with map); Behre (1981), 58ÍI

2 Ô2

Contact w ith the non-Germ anic w orld

Rom an writing practice across the Rhine frontier, while the Goths made their contact across the lower Danube, we confront the remarkable, indeed incredible coincidence that at both geographical extremities o f the imperial frontier the same vernacular word (not even a loanword) was independendy chosen to designate this novelty.36 However, if we postulate monogenesis (the Gm c. word was coined at only one point o f contact and then spread to both W G and EG), we must dismiss, as we saw, the possibility o f a spread from Gothic to W G. T h e converse possibility, from W G to Gothic, is only feasible while the Goths were still in northern Poland, which would tie up with other terms, including Latin loanwords into Gm c., where the same assumption must be made, especially in the sphere o f trade.37 To explain the adaptation o f a native word like bāka to new ends, however, it is not enough to refer simply to Roman trade with northern Europe in the first centuries a d , which o f itself could explain at the most Latin loanwords adopted in Gm c. Instead, we need to assume Germanic-speakers, already making use o f bāka, accompanying Rom an traders as interpreters and guides or acting as trading middlemen themselves. Wulfila’s acquisition o f a meaning for the singular bāka which was anom ­ alous within G erm ania (it resulted from his contact with two classical languages and the need to render gramma or littera in the bible) m ay serve as a bridge to the second aspect o f post-runic vocabulary. U nder this heading we deal with technical terms in ecclesiastical literature, applied either to the Gothic script devised by Wulfila or to the Rom an script employed elsewhere in Germ ania. Some o f these terms are loanwords from Greek or Latin, others are loan-translations, so that we are now concerned, as was the case with bāka, with terminology evolved after the Germ anic tribes had made closer acquaintance with M editerranean writing practice, but now in a Christian context. We may start with the two complementary essentials o f any literate culture, writing and reading. In Gothic Wulfila needed to find a vernac­ ular term for the new technology o f writing (with pen, ink and parch­ ment in contrast to runes carved on wood, bone or metal).38 This need can only partly account for his avoidance o f runic ivrítan, for this was applied in O E to the new technology with no difficulty. Wulfila’s choice o f word for reading, where no novel technology was involved, also avoided a runic term such as O E radon, so that we may suspect that, in 36 Kuhn (1938), 60; Rosenfeld (1952), 195.

37 See above, pp. 2o6ff.

38 Rosenfeld (1952), 207.

T h e vocabulary o f w riting

263

addition to finding terms for a new technology, Wulfila m ay also have wished to avoid runic vocabulary. For writing Wulfila employed the verb mēljan for Gr. gráphein and the plural noun mēļa for grámmata and graphe.39 O nly in Gothic are these words used in the context o f writing, their cognates in O H G are mâlm ‘to mark, stain, paint’ and mâl ‘a mark, stain, point’, with further developments in O E and O N . It is conceivable that Wulfila used words with the general meaning ‘mark’ and ‘to make a mark’ which had not previously been applied to writing in order to designate the novel type o f writing found in the Mediterranean world. This is confirmed by etymological cognates o f the Gothic stem: its counterparts in Gaulish and Latvian mean ‘black’, in Lithuanian ‘dark blue dye’, while Lat. mulleus means ‘dark red’ and Gr. mêlas ‘black’.40 These cognates suggest that Go. mēljan denoted writing with a dark ink on parchment, as opposed to incising runes on other materials or scratching on the w ax o f a writing tablet. W hy Wulfila chose a word meaning ‘to make a black or dark mark’ is explained by his word for ‘ink’, which like his Mediterranean models suggests the colour black: Go. swartizl, Gr. mélan, Lat. atramentum. (It is even possible that the Greek model suggested the phonetically similar Gothic stem mil-, differing only in the length o f the vowel.) A Greek model also lies behind the Gothic word for ‘to read’, likewise unique within Germania: gakunnan renders anagignöskein, two words which correspond closely in their formation.41 T h e Greek verb derives from one meaning ‘to know’ and conveys the idea ‘to get to know, to recognise’, therefore ‘to recognise the meaning, to read’. Similarly, Go. gakunnan is a perfective formation from kunnan ‘to know’ and was used to mean both ‘to recognise’ and ‘to read’. This parallel, as well as the unique position o f the Gothic word within Germ ania, makes it likely that Wulfila followed the Greek model, which is suggested even more forcibly when the Gothic prefix ga- is replaced by another which is the same as in Greek: anakunnan as a variant rendering o f anagignöskein. For the same two literate activities O H G similarly based its choice on the precedent o f classical literacy, turning to Latin rather than to Greek. For writing W G uses a word, as in O H G sknban, which was loaned from Lat. scriberef Within W G there is a semantic differentiation: in O H G the verb means ‘to write’, but in O E it means ‘to impose penance (after 39 Kuhn (1938), 6if.; Arntz (1944), 288; Rosenfeld (1952), 207f. 40 Lehmann (1986), 25of. 41 Kuhn (1938), 62. Cf. Ebei (1963), 76Í 42 Zim m er (1892), 145fr.; Kuhn (1938), 63.

264

Contact w ith the non-G erm anic w orld

confession), to shrive’. O S and OFri. seem to have agreed first with O E , but then been influenced by O H G . In O N skrifa means ‘to write’, but skript ‘confession, penance, shrift’. This double function (O E knows nothing o f ‘to write’, O H G knows only this) suggests a twofold loanprocess from Lat. scribere in England and in Germany, independently o f each other. This implies a late loan (no longer common W G), borne out by the Christian usage o f the O E term in penitential practice. In this field the Anglo-Saxons influenced the continent in developing a closely organised system o f confession and penance, based on interrogationes and canones, drawn up in Latin texts {scripta, cf. O E scriflböc ‘penitential’).43 T h e Latin term therefore found its way into O E in this specialised context, then from O E into those Gm c. languages on the continent most closely exposed to the Anglo-Saxon mission (OFri., O S, then O N ).44 T h at the adoption o f scribere in G erm any (OS and O H G ) was also rel­ atively late can be seen from two points. W hen in the Heliand (231fr.) Zachariah writes on a writing tablet (bôk) his activity is summed up in two words: wntan, in agreement with O E , and gisknban, in agreement with O H G .45 This could suggest that there was a period on the conti­ nent when both words for writing, one runic in origin and the other introduced with Rom an writing practice, were in use. H ow they could be differentiated is shown in O H G by Otfrid, who uses the new verb sknban (of Latin and Christian texts) with reference to classical authors, the evangelists and his own activity as an author, but the other verb rîzan in describing Christ writing with His fingers on the ground.46 Implied by this is a reluctance to see the new word as applicable to anything but parchment, whereas the runic term could be used o f making a mean­ ingful incision in wood, metal or earth. (The physical similarity between incising a rune and using a stilus on a w ax tablet m ay also justify the use o f wrîtan in the passage on Zachariah in the Heliand, as also when Notker uses it o f writing Latin letters with a stilus on a tablet.)47 I f sknban in G erm any was closely connected with writing on parchment and was introduced with it, it was as much an ecclesiastical term as its O E counterpart, even though not used in penitential discipline. For the act o f reading O H G employs the verb lesan, probably passing it on in this sense to O S and O N , as was the case with sknban.48 O H G lesan is also used in the sense ‘to gather, collect’, as is likewise the case in O S and O N , whilst the meaning ‘to gather’ is the only one attested for 43 Zim m er (1892), 146fr. 44 Ibid., p. 148. 45 Schröder (1924b), 57. 47 Notker, Boethius, De Consolatione 223, 31; 372,15. 48 Kuhn (1938), 626; Arntz (1944), 293; Ebel (1963), 78.

46 Ibid., pp. 57t.

T h e vocabulary o f w riting

265

Gothic and O E , where the semantic gap for ‘to read’ was filled from else­ where (gakunnan, rœdan). In the case o f lesan, too, a runic origin has been suggested, combining the ideas o f ‘reading’ with ‘gathering’ on the basis o f the description o f the casting o f lots given by Tacitus (Germania 10): the priest picks up a wooden stave with a rune carved on it in order to read its meaning. However, this suggestion operates with a meaning for lesan ‘to choose, pick out - by picking up’ which is not attested. Just how distinct this verb originally meaning ‘to gather’ (cf. Heliand 2568f., where lesan is used in conjunction with tesamne ‘together’) must be kept from the idea o f selection is shown in a passage from Wulfila. W hen rendering M e. 13.27 (Christ gathers together His elect) he makes a clear distinction between galisan ‘to gather’ and gawaljan ‘to choose’.49 I f the verb arlesan occurs very occasionally in O H G in the sense ‘to choose’ (and also very seldom in the case o f O E ālesan, O S alesan), this meaning was probably loaned from Lat. eligere.50 It is in fact Latin which explains how O H G developed the meaning ‘to read’ from a verb meaning ‘to gather’. Not merely was this secondary O H G meaning unique within Germ ania, it must also be set against the fact that Lat. legere meant ‘to read’ as well as ‘to gather’.51 O H G lesan is therefore as much a loan-meaning from Latin as Go. gakunnan was a loan-translation from Greek. T h e Gothic and O H G words for ‘to read’ are both unique within G erm ania because each is a copy o f a non-Germ anic model. T h e exposure o f O H G to the Latin vocabulary o f literacy in the context o f Christianity went much further than a loanword for ‘to write’ and a model for ‘to read’. To illustrate this we m ay start by turning back to the vernacular word, O H G buoh, for the Roman writing tablet, tabella. T h e later development o f this G erm an word, under the impact o f liter­ acy in the hands o f Christian clerics, first to ‘document’ and then to ‘book’, gave rise to a lexical gap for the original meaning, which still needed to be expressed, since the writing-tablet remained in use through the M iddle Ages.52 Although Lat. tabula/tabella had been taken over in G erm an as a loanword (O H G zabal), this is attested only in the meaning ‘gaming board’,53 as still in M H G schâchzabel ‘chessboard’, so that the lexical gap still persisted. It was filled by borrowing the same Latin word a second time, producing now O H G tavala, meaning primarily ‘writing tablet’ alongside occasional ‘gaming board’ .54 From the phonological difference between zabal and tavala, it is clear that the second form was 49 Kuhn (1938), 63. 50 Ibid. 53 Starck and Wells (1990), 752.

51 This against Arntz (1944), 293. 54 Ibid., p. 623.

52 BischofF(i979), 26fr.

266

Contact w ith the non-G erm anic w orld

borrowed later: it shows V L -v- in place o f Classical Latin -b- and it must have been loaned after the completion o f the Second Sound Shift. O ther words belonging to the vocabulary o f writing in O H G point to the same conclusion: they are all loanwords from Latin which must have been taken over late. We m ay divide them into various aspects o f writing practice, the first o f which, as with tabula/tabella, concerns the object or material on which writing was done. T h e Rom an writing tablet, as we have seen, was important enough to call forth various attempts to name it in Gmc.: boka as a counterpart formed from native linguistic material, then zabal as an early loanword (but outside the context o f writing) and finally tavala. Although the Rom an object could be made ambitiously from ivory, the more frequent form was o f wood, which accounts for its wide spread and persistence as a cheap and practical writing instrument. For more pretentious writing purposes papyrus, largely an Egyptian monopoly, was used in classical antiquity in the form o f scrolls and then by Christians from the second century in the shape o f folios folded into a codex.55*It continued to be used well into the Middle Ages, so that documents from the Merovingian royal chancery, for example, still used papyrus towards the end o f the seventh century. T h e Latin word is attested in O H G , even if once only, in the form p a p īri T h e Latin word charta, used o f a sheet for writing made from papyrus, has found an equally faint reflection in O H G /carta (one example only).57 T h e dis­ advantage o f papyrus (its durability in the European climate is question­ able) m ay be one o f the reasons why writing in late antiquity made the transition not merely from scroll to codex, but also from papyrus to parchment. T h e new material, made out o f sheepskin, was more durable, even if more expensive because o f the careful preparation required (the C odex Amiatinus, produced in Northumbria, demanded as many as 500 sheepskins).58 T h e Latin word for this new material, pergamenum, was in origin the adjective from the place-name Pergamon, in Asia Minor. It was borrowed into O H G , where it is attested reasonably frequently in the forms pergamin /pergamente.59 T h e different surfaces used for writing demanded different instruments for inscribing the letters. T he wax surface o f the writing tablet had to have shallow incisions made on it, which could then subsequently be erased so that the tablet could be used again. T he instrument for performing both operations was the stilus, with one end sharp for incisions and the other 55 Bischoff (1979), 19fr. 58 Bischoff (1979), 2i f f

56 GR 2, 358. 57 Starck and Wells (1990), 323. 59 Starck and Wells (1990), 458.

T h e vocabulary o f writing

267

blunt for erasures (it could also be used for faint entries on parchment).60 In the biblical passage (Lc. 1.63) where Zachariah wrote on a writing tablet {pugHlaris, rendered as bôk in the Heliand) the Tatian translators refer not to this, but to the writing instrument he used: scribsahs (4,12), literally ‘writing knife’.61 Elsewhere, the word more commonly used for this tool was grijfil, probably derived from another Latin word o f Greek origin, graphium (cf. O H G grafin this sense), with assimilation to grîfan ‘to grip’ by popular etymology and the addition o f the suffix -il, used for tools.62* Although the stilus could also be used on parchment, the result was too faint to be easily legible for practical purposes, so that the use o f ink demanded another form o f writing instrument. Although classical antiq­ uity had used the calamus or reed pen (which continued to be used in the Mediterranean world in the early Middle Ages), the more widespread instrument for conveying ink on to parchment in the Middle Ages was the quill pen, made from a bird’s feather. T he word for this was Lat. penna (or pinna?),63 meaning originally no more than ‘feather’, but then ‘pen’, which found its way into Germ ania not as a loanword (Engl, ‘pen’ is a later bor­ rowing from OFr.), but as a loan-translation: O E feðer, O H G fed era l As regards the use o f ink, although Wulfila devised a native word for it on the model o f Greek and Latin, preference is given to loanwords from Latin elsewhere in Germ ania. Each o f the three most common Latin terms, reflecting different ways o f making ink, is attested in Germany. Atramentum, conveying the idea o f a black liquid which we saw in Go. swartizl, was loaned into O H G as atrament and, showing a closer assimilation to O H G phonology, atraminza.65 Another word, deriving from Lat. encaustum (originally the purple ink used by the Rom an emperor for his signature), occupies a more restricted area in northern W G , as in M D u. inket, inkt, ink and Rhenish enk, ink. This word also found its w ay into OFr. as enque (cf. modern Fr. encre).66 Both these forms were unable always to hold their own against a third form: Lat. (aqua) tincta ‘coloured liquid’, attested in O H G as tincta/tinte/timpte and glossing both atramentum and encaustum.67 Another word, restricted in its scope on the continent, points to the possibility o f O E influence, this time by means o f a loan-translation. Just as Wulfila coined swartizl for the new technol­ ogy, so was atramentum copied in O E by bbec, an adjective used as a noun to mean ‘black ink’. This form has counterparts on the continent in O S 60 62 64 66

Bischoff (1979), 32. 61 Cf. also Starck and Wells (1990), 548. Ibid., p. 239; Kluge and Seebold (1989), 278. 63 F E W 8, 526fr. Bosworth and Toller (1954), 285; Starck and Wells (1990), 143. 65 GR 2, 103fr. Ibid.; F E W 3, 224C 67 GR 2, 105i.

268

Contact w ith the non-G erm anic world.

blac, M D u. black, EFri. blak (with scattered outriders in Franconian and U G , not always understood).68 T h at we are justified in seeing these con­ tinental forms as under the influence o f O E is suggested by palaeographic evidence: Irish and Anglo-Saxon scriptoria used a characteristic dark brown ink, the employment o f which can be traced on the conti­ nent in those areas which, as with the word for ink, were particularly exposed to the influence o f the Anglo-Saxon mission and scriptoria.69 It looks therefore as if the O E word for ‘ink’ accompanied the particular type o f ink the Anglo-Saxons used. A practical advantage with the writing tablet was the possibility, after using it once, o f wiping the w ax smooth and clean in order to use it again. This was not possible when using the stilus on parchment, but it was feasible (and desirable in view o f the expense o f parchment) if pen and ink had been used: the surface was washed or scraped clean o f the ink already used for the first text, and then used again for a new one, a ‘palimpsest’.70 T he word used in Latin for this procedure was the verb delere, meaning originally ‘to destroy’, but also acquiring a technical meaning in the context o f writing: ‘to delete’ in order to use the surface again. With or without the addition o f a prefix the word is found in O E (ā)dilegian (with -g- functioning as a glide), above all in clerical transla­ tions, especially o f the Psalms, where it renders Lat. delere both in the general sense ‘to destroy’ and with the specialist meaning ‘to delete, erase’.71 W hereas in O E the loanword remains tied to ecclesiastical usage, it passes beyond this in Germany. Here it can also be used in the context o f the Psalms both with O S fardiligon and O H G firdiligôn/firtiligôn (nomen delere ‘to delete a name from the Book o f Life’), but also with a much wider range (‘to abolish, consume, exterminate’), especially with Notker.72 If we accept the suggestion that the Latin word was first bor­ rowed into O E in clerical scriptoria and then taken to the continent by the Anglo-Saxon mission,73 then the path o f this word southwards in G erm any would be marked by the variation in O H G between -d- and -t-, where the latter represents an attempt to give an U G form to a word from the north. In this connection it m ay be suggestive, but certainly not conclusive, that Lat. pumex ‘pum ice’, used in scraping parchment clean, was borrowed into O E (pumic) and O H G (pumez).n However, the fre­ quency o f this practice in scriptoria could well mean that the word was loaned into the two languages independently. 68 Ibid., pp. KMf. 69 Rosenfeld (1952), 208; Bischoflfiigyg), 30. 70 Bischoff (1979), 23fr. 71 Schröder (1923), 246ÍT. 72 GR 2, 226fr. 73 Schröder (1923), 248. 74 Bosworth and Toller (1954), 779; Starck and Wells (1990), 465.

T h e vocabulary o f w riting

269

For the finished product o f this scribal activity, the actual written communication, Latin used libellus brevis, which could then be abbrevi­ ated to libellus or to brevis. T h e first, a diminutive o f liber ‘book’, is attested in O H G in the forms libel/livol, where the latter, so far from denoting only a short book, as the Latin diminutive might suggest, is even applied to one o f the longest works in O H G literature, O tfrid’s Evangelienbuch,75 This form differs from libel (which points to a learned borrowing direct from written Latin libellus) in presupposing an origin in V L livellus. Brevis, on the other hand, is reflected in O H G brief ‘letter, document, sheet’ and may thus be said to have retained the force o f a shorter communication (cf. also brieftabula ‘writing tablet’).76 In the O S Heliand 230fr. bref is used o f a short communication (one name only) and is varied by bôk ‘writing tablet’. T h e O H G diphthong presupposes an earlier long e, as in V L in this position in an open syllable by contrast with the short e in Classical Latin. A stage later than Classical Latin is also implied by the final f in O H G , indicating that in its Latin model -v- was no longer pronounced as a semi-vowel, but as a fricative. This last word has brought us to the question o f dating these loan­ words by phonological means, where we are helped by the evidence from Germ an sound changes. By chance many o f the words acquired from Latin in this lexical field contain consonants which would have been subject to the Second Sound Shift, but it cannot be fortuitous that none o f them shows its effects: tavala (as opposed to zabal, not used in the context o f writing), papîr, karta, pergamín/pergamente, griffil, atrament, firdiligôn. Two exceptions are more likely to be only apparent, repre­ senting assimilations to U G phonology: firtiligôn alongside firdiligôn (cf. for a later stage in Germ an, OFr. danser > M D u. dansen > M H G tanzen) and pergaminza alongside pergamente (cf. Lat. pigmentam > O H G ptminta, but also piminza).17 ( O H G paffur as a rendering for papyrus, with one p shifted and the other unshifted, remains a problem.)78 If none o f these words was therefore borrowed early enough to be affected by the Sound Shift and if they are attested once a written tradition in O H G began this suggests a probable loan in the seventh and eighth centuries (in some cases even later).79T h ey represent a stage in the introduction o f writing to Germ any considerably later than the runic script or the Romans’ use o f the writing tablet, as is confirmed by the technical fact that the wide use o f papyrus, parchment and ink, together with all the other appurtenances o f a 75 Starck and Wells (1990), 372; Otfrid: hi 1, 1 (cf. also Kelle (1881), 369). 76 Starck and Wells (1990), 77. 77 GR 1 , 191; 2, 386t. 78 GR 2,358. 79 Hüpper (1986b), 112.

270

Contact w ith the non-G erm anic w orld

growing literacy, were brought to northern Europe by the Church. W hether we have to locate this writing activity o f the Church in the monasteries, whose schools and scriptoria were focal points o f literacy, or in the Merovingian chanceries is immaterial besides the fact that in either case this activity and the employment o f such technical terms as these were in the hands o f clerics.80 In quite different ways this chapter and the preceding one have had to some extent to take account o f Christian influence. With the days o f the week what might be called a Christian censorship o f pagan planetary names played a more successful rôle in eastern Europe (including Gothic) than in the west, but was not completely absent here, especially with names which were regarded as holy within the Christian week. With the vocabulary o f writing the wish to avoid any association with pagan runic magic combined with the rôle o f Christianity as a religion o f the book to make this vocabulary a growing concern o f the Church. To the religious implications o f these points we shall have to return later. Taken together, these last two chapters prepare the way for Part hi, in which we consider the impact o f Christianity on Germania. 80 Monasteries: BischofT (1979), 58; Ebenbauer (1981a), 36; Hüpper (1986b), 112, 122. Chanceries: Schröder (1924b), 58.

P A RT III

Contact with Christianity

Introduction to Part I I I

In Part n we were mainly concerned with the growing contact o f the Germ anic tribes with the Roman Empire, above all in the case o f the Franks on the lower Rhine and the Goths on the lower Danube. By the fourth century, however, Christianity had become an important part o f the Empire and o f classical civilisation, but had not developed any policy o f a planned, systematic mission. It was therefore not the case that the new religion was brought to Germ anic barbarians by missionaries, but rather that the migrations o f these Germ ani, in taking them to the fron­ tier zone o f the Empire, and then across it, also brought them into contact with Christianity.1 As long as they remained beyond the imper­ ial frontiers these tribes were pagan, but they were converted to Christianity, in one form or another, within a generation or so o f arriv­ ing in the Empire.2 If not in the sense in which he meant it, the apolo­ gia made by Orosius can be said to be correct: G od is to be praised because the barbarians were granted entry on to Roman territory so that the churches o f Christ in east and west should be filled with believers, Huns, Suevi, Vandals and Burgundians.3 Contact with Rome was there­ fore tantamount to contact with Christianity. T h e theme o f Part m is therefore a particular aspect o f the subject of Part it, but one o f such historical importance that it has to be treated separately. Moreover, the Church, having assimilated many Roman traditions itself, was the primary means by which these traditions were conveyed to medieval society: in episcopal organisation, administration, literacy and learning. It was the crucial factor which assisted in the amalgamation o f classical and barbarian traditions, in bringing about the threefold conjunction which is the subject o f this book. In looking at the contact o f Germ ania with Christianity we shall have to narrow our focus somewhat by concentrating principally on the 1 Schmidt (1954), 155.

2 Thom pson (1966), 127t.

273

3 Schäferdiek (1978a), 508.

274

Contact w ith Christianity

evidence from O H G . This is forced upon us for the practical reason that the enormous scope o f this theme, chronologically as well as geograph­ ically, makes selectivity imperative. Even so, to focus on O H G still allows a reasonably wide treatment, both because o f the variety' o f external Christian influences on G erm any (Gothic, Merovingian Frankish, Anglo-Saxon) and ultimately because o f O H G influence beyond its own boundaries (on O S and indirectly on ON).

C H A P T E R 15

Problem s o f Christianisation

In the earliest period Christian communities lived very much as small groups turned inwards and anxious, especially in the age o f persecu­ tions, to give no offence to the pagan society in which they lived and not concerned to convert pagans at all costs.1 T h e gaining o f new believers was an act o f individual conversion, confined at the most to contact between small groups and far removed from political concerns o f state. Proselytising arose from chance encounters with non-Christians who may have shown interest, often with women and slaves, the under­ privileged o f antiquity especially open to the revolutionary prospects o f the new religion. Although Christianity had its apostles and evangelists the early Church dispensed with public missionary preaching, if only to avoid the attention o f a hostile pagan state. Instead, whoever showed interest in the new religion (and the impetus had to come from that direction) was allowed to attend the service o f a local parish, where the sermon to the congregation also served as a missionary sermon to the unconverted interested enough to attend. If the early Christians were forced by circumstances largely to dispense with verbal propaganda, they were able to make this good by a propaganda o f the deed. Pagan outsiders were therefore drawn to the new religion (as in the case o f Tertullian) by observing the Christians’ readiness for self-sacrifice and their high moral standards, their care for the poor and the weak in society (it m ay be no accident that one o f the earliest Christian terms to find its w ay into Germ anic concerned such activity: Almosen ‘alms’).2 In this sense every Christian who attracted favourable attention by living up to such ethical demands was his own missionary. T h e same was even more true o f the Christian ready to embrace martyrdom. T h e lack o f an active missionary policy, together with the fact that the impetus to conversion had to come from those outsiders who were ' Frend (1974), 32fr.; Holl (1974), 3fr.; K ah l (1978), 45fr., 6off.

275

2 See below, p. 306.

276

Contact w ith Christianity

interested, meant that particular importance was attached to the institu­ tion o f the catechumenate, the instruction o f the candidate for conver­ sion in the tenets and demands o f the new religion as a precondition o f baptism. For the Church to impose this requirement, however, means that it gave pride o f place to the would-be convert’s individual freedom o f choice, to his informed awareness o f what he was about to undertake. This emphasis ties up with the rejection o f force as a means o f conver­ sion (even Augustine, despite his attitude to the Donatist heretics, seems not to have contemplated the use o f force against non-Christians) - quite apart from the fact that in the earliest centuries the Church was more the victim than the wielder o f compulsion. T h e small-scale nature o f the earliest spread o f Christianity, from individuell to individual or at the most from group to group, and its divorce from the rôle o f the state meant that the widest limits which the early Church could conceive for itself were those o f the Rom an Empire. I f Christianity was taken beyond these boundaries in the earliest centuries (to Arm enia or to the Goths while they were still outside the Empire) these were not the outcome o f a consciously adopted universal mission. In almost all these respects the early medieval Church, as it arose after Constantine, Chlodw ig and G regory the Great, adopted a different atti­ tude towards the spread o f its religion. From Constantine’s victory and conversion there date the great increase o f Christians within the Empire and new types o f missionary activity, including missions organised by the emperor himself (in which religious and political factors are intertwined) or by bishops to restore peace to their dioceses (so that an element o f compulsion is involved), but also isolated proselytising by individuals amongst the barbarians beyond the imperial frontiers.3 T h e connection o f conversion with the state and politics has a twofold explanation. O n the one hand, this would be a natural consideration for possible converts in Germ ania, where the pagan tribal religion was inti­ mately associated with the social and political life o f the tribe. O n the other hand, the missionary entering pagan territory under the protec­ tion o f a neighbouring Christian ruler would be regarded above all as the political emissary o f an expansive power. N or could the missionary, setting foot on pagan soil beyond the Roman frontier, dispense with such protection (as with the Frankish support given to Colum banus or Boniface), even if that meant entering a ruler’s service.4 Concom itant with this political dimension and with princely support was the phe3 Frend (1974), 38.

4 Baetke (1944a), 82fr.; Holl (1974), i2Íf.

Problem s o f Christianisation

277

nomenon o f mass conversions, o f whole peoples following the example o f their ruler, perhaps the most pronounced difference from the early Church.5 T h e conversion o f a ruler generally meant the adhesion to Christianity o f many o f his followers, however superficial the process m ay have been, so that the real task o f religious education remained to be accomplished afterwards. This is a feature recurrent in the Christianisation o f the Germ anic tribes and is especially well attested for Anglo-Saxon England. Religious education taken in hand only after a mass baptism testifies to a drastic reduction o f the rôle o f the catechumenate which had been an indispensable preliminary to baptism in the early Church.6 Whereas hitherto Christ’s twofold command to the dis­ ciples (Mt. 28.19: ‘Teach all nations, baptising them’) had been followed in that sequence, now it was reversed. If baptism represents the ritual borderline between paganism and Christianity it was often crossed at what we can only regard as an astonishingly early point in a missionary process in which the emphasis fell primarily on the power o f the Christian G od and the need to acknowledge this by renouncing the tribal gods, thus leaving the main task o f dogmatic instruction to sub­ sequent pastoral supervision. Although Gregory the Great remained opposed to forcible conversion he accepted the principle that it might be necessary to prepare the way for preaching the gospel to pagans outside the Empire by first sub­ jugating them militarily and thus providing a protective framework for peaceful preaching.7 He was sanguine enough to see this use o f military force as proceeding not from a wish for bloodshed, but from a desire to bring the gospel tó the subjugated peoples (per subditas gentes). Indirecdy Gregory therefore advocated the possibility o f missionary warfare, a policy later carried to extremes by Charles the Great against the conti­ nental Saxons, calling forth the criticism o f Alcuin. If the use o f military force was made necessary by the wish to spread the gospel beyond the frontiers, this points to a final difference from the position in the early Church which had not regarded it as its task to be active outside the Empire. T he historic change in policy is represented again by Pope G regory whose pontificate was marked from the beginning by a concern with the conversion o f pagans (Langobards, Sardinians, Anglo-Saxons) and whose mission to Kent was a decisive step in inaugurating an organ­ ised universal mission to northern Europe.8 5 Baetke (1944a), 123; Levison (1946), 47i. 6 Baetke (1942), 22if.; K ahl (1978), 47fr.; Schneider (1978), 242. 7 K ahl (1978), 63i., 68; Schneider (1978), 244. 8 Frend (1974), 4iff.

278

Contact w ith Christianity

T h e different attitudes shown by the Church to the need to convert pagans in the periods before and after the victory o f Constantine and the institution o f Christianity as the religion o f the Rom an state are therefore considerable, so much so that they make it difficult to employ a number o f key-ter ms in a w ay which does justice to such widely differing conditions. With regard to ‘mission’ we need to be more cautious than those philologists who have attributed Christian terms in O H G to the influ­ ence o f a variety o f mission movements bringing them to Germany. This makes no distinction between the occasional, interpersonal proselytising o f the early Church and the more organised form o f universal mission which later developed.9 How relevant this distinction is to G erm ania can be seen in the traces o f the vocabulary o f the provincial Rom an Christians o f the Rhineland or in the conversion o f the earliest Gothic Christians by Cappadocian prisoners o f war. In neither case was any organised mission responsible for the formation o f an early Christian vocabulary, as was later the case with the Gregorian mission to England or with the Anglo-Saxons in Germany. Instead, these terms result from the personal contact o f Germ anic pagans with Christian communities still surviving in the Rhineland, they need not even indicate that the Franks who first used the word for ‘church’ to designate the buildings they saw were themselves Christian.10 To associate a group o f Christian terms in O H G with a fully fledged mission to G erm any begs the ques­ tion in another way if we recall that to the probable presence o f some Germ an terms o f Gothic origin there corresponds no known Gothic mission to Germany, whilst conversely the historically attested presence o f Irish missionaries (even if their rôle has been exaggerated) has yielded no reliable examples o f their linguistic influence on Germ an. Doubts about the term ‘mission’ are also called for by the attempt o f the Church in sixth- and seventh-century Francia to regain ground lost as a result o f the barbarian invasions. This may well have involved the conversion o f individual pagans, but from the point o f view o f the Church it was more an ecclesiastical restoration or a Reconquista, still operating well within the boundaries o f the Empire. Given such reservations it is best to avoid using ‘mission’ as a blanket term altogether, certainly in the linguistic context, and to speak instead more generally o f the creation o f a Christian vocabulary in G erm an from various sources. Even the term ‘conversion’ has its pitfalls. Strictly speaking, it involves 9 Fritze (1969), 78fr.

10 See below, pp. 299, 300.

Problem s o f Christianisation

279

conversion from sinfulness by means o f repentance (Acts 2.38: ‘Repent and be baptised every one o f you in the name o f Jesus Christ for the remission o f sins’), but the difficulty in Germ ania lay in conveying the religious dimension o f repentance and sin, for pagans may well have regretted some o f their actions or even regarded them as a níðingsverk, but not in the context o f their religious practices. Until they understood this religious dimension their renunciation o f the old religion could only fall short o f what the Church meant by conversion. This shortfall is heightened by the contrast between individual conversion and a mass process involving a whole tribe. There may be no cause to doubt the gen­ uineness o f the conversion o f individual Visigoths, won over by the example set by Christian prisoners they had taken and ready to accept martyrdom in repeated persecutions, but can it be meaningfully said that a whole tribe was converted together in the same sense?11 Even if religious motives (alongside political ones) m ay have informed the indi­ vidual conversion o f Constantine or Chlodwig, can it be said that their followers, who accepted baptism with their leaders, were likewise prompted? In what sense can the term ‘conversion’ be applied to the Saxons, forced to baptism by Charles the Great? W hat is involved here is at the most a change o f cult or o f religious observances far removed from the religious sense o f metánoia, a superficial change o f ritual alle­ giance to which a neutral term such as ‘adhesion’ to the new religion or even ‘Christianisation’ is better suited than the more loaded ‘conver­ sion’. H ow far short o f the Christian sense o f conversion the change o f many Germ anic tribes actually fell is suggested by the need to describe their newly found Christianity as syncretistic, combining the old with the new. O ne cause o f this syncretism was the encounter between a tribal religion (whose gods guaranteed fertility and victory, therewith the sur­ vival o f the tribe) and a universal G od who was claimed to have the power to intervene in any tribe’s life, for the common reaction o f a tribal religion to such a situation is to incorporate the new divinity into its exist­ ing pantheon. Widukind therefore reported o f the Danes that, although nominally Christian, they also worshipped their ancestral gods and even held them to be superior to Christ.12 T h e religious psychology behind this is understandable, for it must have been difficult for worshippers in a polytheistic religion to accept that the new G od was the only one and that all their tribal gods were false, an argument which has been used in 11 Baetke (1944a), 84,123.

12 Kahl (1978), 34f., 51, fn. 89.

28o

Contact w ith Christianity

the case o f Constantine himself.13 N or need the Church always reject such mixed beliefs, for they provided opportunities to missionaries willing to adopt a policy o f accommodation by working with these half­ beliefs in external form, but hoping to outplay them .14 Apart from such cases o f a short-term policy o f accommodation for the sake o f long-term advantages these syncretistic beliefs could only be opposed by the Church, but their very presence testifies to the superficial conversion o f those nominal Christians to whom they are attributed.15 Long after their nominal conversion the Franks are supposed to have performed human sacrifices on the Italian campaign o f Theudebert I, whilst from Iceland it is reported that Helgi the Lean had a mixed belief: he believed in Christ, but in difficult undertakings he appealed to the god Thor. Examples like these are many and run the gamut from a subsidi­ ary role granted to the new religion to its grudgingly tactful toleration o f traces o f the old.16 T hey are an indication o f how drastically novel the idea o f one universal, ‘supertribal’ G od must have been and suggest that this new G od must have been regarded for some time by many as a primus inter pares. T h e interaction between a tribal and a universal religion also under­ lies the idea that the Church had a universal mission to all pagans. A tribal religion does not proselytise, since its adherents all belong to the tribe, just as the benefits conferred by the gods are confined to that tribe, even though there may be other gods for other peoples.17 Acceptance o f the Christian G od could only take place when His power to intervene in tribal affairs was shown, but for this the new G od had to be brought to the Germ anic tribes, i.e. the Church had to develop a missionary policy going beyond the Empire. T he idea that Christianity as a universal relig­ ion automatically developed a universal mission is historically by no means self-evident,18 for the early Church relied for proselytising on per­ sonal encounters, on chance contacts, rather than planned ones, it dis­ pensed with missionaries and developed no mission programme. For Pope Leo the Great the universality o f the Empire, embracing so many peoples, argued against the need for a mission beyond its frontiers.19 However, the fall o f the Empire and its invasion by pagan tribes destroyed the basis o f this argument and granted renewed relevance to Christ’s command to the disciples to go and teach all peoples (Mt. 28.19: Ite, docete omnes gentes). W hereas the early Church had understood gentes, 13 Lane Fox (1988), 330, 6igf. 14 Flint (1991), 305t. 15 Schäferdiek (1978a), 526,538; Baetke (1944b), 165Í'. (3 and 4). 16 Kuhn (1978), 316Í' 17 Thom pson (1966), 62; K ahl (1978), 30fr. 18 Fritze (1969), 123. 19 Ibid., pp. 124t.

Problem s o f Christianisation

281

in imitation o f Hebrew goyim, to refer to all Gentiles or pagans, seeing them as individuals to be proselytised, the barbarian invasions granted urgency to the ethnic sense o f gentes, now conceived as the pagan tribes in and beyond the Empire.20 T h e decisive step in this development o f a universal mission was the mission o f Augustine, sent by Pope Gregory to convert the AngloSaxons.21 W here the disciples had been enjoined to preach the gospel to every creature (Me. 16.15: omnis creatura), Gregory interpreted this in ethnic terms (omnis natio gentium), a widening o f scope from earlier indi­ vidual proselytising to pagan peoples at large, especially those outside Rom e’s frontiers.22 Later the compliment was to be returned when the christianised Anglo-Saxons themselves sent missionaries abroad, to some Germ anic tribes with whom they felt a degree o f kinship (Frisians, Saxons), but also, if we can trust the list given by Bede, to such distant peoples as the Slavs and Avars.23 Geography may have made the latter goal unreachable, but their intention, together with what they did achieve, shows that G regory’s universell mission had borne fruit and was now continued by those whose Christianisation he had himself inaugu­ rated. T h e problems o f christianising pagans were accompanied by the lin­ guistic problems posed by Christianity. Words had to be devised every time a language boundary was crossed to express features o f the new religion, so that each language had to be ‘baptised’ as much as its speak­ ers.24 T h e universalism o f Christianity gave rise to linguistic problems from the start, for with the gift o f tongues bestowed upon the apostles (Acts 2.iff.) a range o f languages from the eastern Mediterranean and beyond is listed so that ‘every man heard them speak in his own lan­ guage’.25 This linguistic problem has two aspects. T h e first is how far the early Church was ready to make use o f various languages in translating the bible and in the liturgy, whilst the second concerns the ways in which terms o f the new religion were rendered in the various languages.26 T h e first o f these aspects soon presented itself as a problem with various languages within the sphere o f influence o f the classical oecumene (e.g. Syriac in Asia Minor, Coptic in Egypt). Speakers o f these languages, coming into contact with Christianity, each time faced the problem o f devising a vocabulary for its novel message. After the first 20 Ibid., pp. 88f, 127i. 21 Schäferdiek (1978a), 534; Fritze (1969), io6ff. 22 Fritze (1969), no. 23 Ibid., pp. 78ÍT. 24 Bartelink (1974), 397fr. 25 M ohrm ann (1957), i if. 26 Ibid., pp. 12if.

282

Contact w ith Christianity

transposition (from Aram aic into Greek) the Greek language rapidly became the standard language o f Christianity, both in the west and the east, but in Italy and Gaul Latin came to the fore as well, so that Christendom was divided into two linguistic areas, characterised by two different linguistic policies.27 In the Greek east there was never a central­ ising factor like Rome in the west: the philological tradition o f transla­ tion at Alexandria, which had earlier brought foreign literature into Greek, now returned the compliment by translating Christian ideas from Greek into other languages o f the east (e.g. Syriac, Coptic, Armenian). First the bible (or parts o f it), then liturgical prayers, but also lives o f saints and martyrs now found their way into a number o f ver­ naculars within the orbit o f the Greek Church.28 These include Gothic, so that Wulfila’s translation o f the bible is a result o f the linguistic decentralisation o f the Greek Church. In this respect the position was different in the west, where Rome represented a centralising force by which vernacular languages in Gaul and Iberia succumbed to Latin as the language o f Empire, a process paralleled by the complete Latinisation o f the western Church.29 This dominance o f Latin as the language o f the Church was not confined to those regions o f the west whose vernaculars had given way to Latin, for it was also used as an instrument in extending Christianity to regions where native languages still survived (Celtic and Germanic). Here, too, Latin retained its monopoly as the language o f the bible and the liturgy, in contrast to the east with Gothic, Slavonic, Arm enian and Georgian. With the second aspect o f the linguistic problems raised by the spread o f Christianity we touch on a number o f points where the difficulties and solutions found in Germ an recur in other Germ anic languages and can be observed at earlier stages in the development o f a Christian vocabulary for Greek and Latin. T h e Christian conviction o f ‘newness’ led to a rejection in Christian Greek o f terms which had earlier belonged to the religious sphere and might therefore have lent themselves to Christianisation. Instead, words were chosen which lay outside the reli­ gious sphere o f paganism altogether or completely new words were coined: by either means the danger o f contamination could be avoided.30 In the early Christian period Latin showed a similar attitude. This is reflected in the large number o f Christian Greek loanwords present in Latin, not just from the time when Greek was the language o f 27 Ibid., p. 15. 28 Ibid., pp. i6ff.; Herrin (1989), 92. 30 Ibid., pp. 20, 23t.

29 M ohrm ann (1957), i6f.

Problem s o f Christianisation

283

the oecumene, but a conscious choice o f loanwords from without (already hallowed by their Christian function) as a way o f avoiding suspect association with religious ideas o f pre-Christian Latin.31 After the Edict o f Toleration, Latin revealed a change o f attitude, a greater willingness to adapt pre-Christian terms to Christian ends. Terms which had previously been rejected because o f their pagan associations were accepted as Christianity became more established and less threatened by paganism.32 Behind these two linguistic attitudes o f rejection or adoption o f preChristian religious vocabulary there lay two different options available for the expansion o f Christianity. T he Church could proceed either polemically, presenting its doctrine in all its shocking newness, milking no concessions to the pagan world around it and therefore ready to accept only slow gains in proselytising, or it could adopt a more eirenic stance, prepared to accommodate itself to the beliefs o f pagans whom it could thereby hope to win over more easily, even if this at first jeopar­ dised the authenticity o f the Christian message and opened the w ay for a religious syncretism which could only be remedied subsequently by long-term pastoral instruction.33 Bearing in mind these two options (polemic and eirenic) and their effects on the choice o f Christian terms in any language which is to be opened up to missionary activity, we m ay now consider the various classes o f linguistic loan traffic involved in the transmission o f Christian terms from Greek to Latin and then to O H G . Although we are dealing with three such classes (the third o f which falls into two subcategories), their basic division is twofold, corresponding to the two options available to missionary expansion.34 T h e first class, loanwords, illustrates the historical continuity o f Christianity. Greek therefore made use o f loanwords from Hebrew (e.g. pascha), just as Latin in turn borrowed from Greek (e.g. heremita, ecclesia). Continuity was preserved into O H G when this language borrowed from Latin either a term which ultimately went back to Greek (e.g. evangelio, alamuosan, ‘alms’) or one o f Latin origin (e.g. signum ‘sign o f the cross’ > segan ‘blessing’, cantico ‘hymn’). T he advantage o f this method is clear: it is an easy linguistic process, making for an international terminology fit for a universal religion and offering quick linguistic returns. Its disadvantage 31 Ibid., pp. 23, 28, 29i. 32 Ibid., pp. 33i.; Bartelink (1974), 400, 416F. 33 Flint (1991), 71,77. T h e so-called ‘Germanisation o f Christianity’ has been criticised by Rathofer (1962), 51fr. and Schäferdiek (1984a), 521fr., but unconvincingly proposed again by Russell (1994). 34 Betz (1974), 135fr.; K ahl (1956), 94.

284

Contact voith Christianity

is equally clear, however. This method was only effective in the case o f material objects and persons, where the meaning o f the strange word was conveyed by what could be heard (cantico) or seen (evangelic in the form o f a gospel-book, segan as the sign o f the cross). In the case o f other concepts (e.g. angelus > engil), however, a loanword remained largely foreign to those who were to be converted, so that the price o f a nominal, uncom­ prehending Christianisation had to be paid for the easy creation o f a vocabulary by this means. T h e second class, loan-translations, consists o f cases where a new word was created in the recipient (hitherto pagan) language by a part-for-part translation o f the giving (Christian) language. Greek dikawsis yielded Latin iustificatio, or Latin evangelium produced O H G cuatchundida (whether by someone who understood the meaning o f the Greek original or was acquainted with Isidore o f Seville’s explanation: bona adnuntiatio).35 T h e advantage o f this method is that the use o f native word-material was less artificial than a preference for loanwords and held out some promise that the new Christian term would be effectively understood (e.g. cuatchundida by contrast with evangelio). T he drawback is that this linguistic process was still to some extent artificial, since the result o f this part-for-part translation was a word which had never existed before and must there­ fore have had a flavour o f novelty. Moreover, although these coinages were often readily understandable, this was by no means always the case, as with gewizzenî as a strictly derivative rendering for conscientia. T h e third class, loan-meanings (a word in the recipient language acquired a wider (Christian) semantic range under the influence o f a word in the giving language), falls into two subcategories, religious and secular. To the words in this class we can apply what has been said o f the conversion o f Greek eirēnē ‘peace’ to Christian ends under the influence o f Hebrew shalom: that it is a Greek ship laden with a strange cargo.36 Since the need was for a religious vocabulary the temptation was to appropriate established religious terms for new, Christian purposes. This had already happened in Latin {dominus was christianised under the influence o f Greek kýrios, salus followed the precedent o f sötería),37 and a similar pattern was repeated in O H G (heil fell under the influence o f salus, sêla under that o f anima).38 T h e advantage o f this method is that it shows the least degree o f artificiality, producing a term which was readily understood and making the transition to Christianity largely 35 See below, p. 331. 36 M ohrm ann (1957), 24. 37 Heiler (1959), 549i.; Bartelink (1974), 413^ 38 O n sêla see Weisweiler (1940), 25fr.; Eggers (1957), iff.; La Farge (1991).

Problem s o f Christianisation

285

painless. However, this success was achieved at the cost o f deleting the essential difference between paganism and Christianity, so that the initial gain concealed the fact o f a deferred payment. O ne solution to this difficulty was held out by secular loan-meanings adapted to a Christian function, for these avoided associations with the pagan religion by reject­ ing religious terms altogether. Examples o f this method in Christian Latin are found in redemptio (originally ‘buying back’ or ‘ransoming from captivity’) and gentes (earlier used o f other peoples or barbarians, as opposed to the populus Romānus, and therefore capable o f transferring its negative connotation to gentiles or pagans, as distinct from the populus D ei).39 In these cases the novelty lay not in the word itself, but in its reli­ gious function acquired only from the new religion. T h e same is true in O H G o f such words as huldi, used to render gratia, or truhtin as a transla­ tion o f dominus.40 But these secular loan-meanings avoid the drawback o f their religious counterparts only by opening the way to other dis­ advantages. I f the words had previously had only a non-religious func­ tion it would take some time to bring home to native speakers precisely what Christianity had done to them. M ore importantly, the non-reli­ gious terms chosen could often introduce secular associations which threatened to distort the picture o f Christianity, as with O H G huldi and truhtin as terms at home in the war-band which could involve a reinter­ pretation o f Christianity not as intended by the translator, but as mis­ understood by those whom he addressed. Each o f these methods traditionally used in creating a Christian vocabulary in the language o f a freshly converted people had therefore advantages and disadvantages. A decision in favour o f one advantage almost inevitably entailed a corresponding loss o f effectiveness or distor­ tion o f the Christian meaning. W hen we find one Germ anic language repeatedly choosing or giving priority to one method, whereas another language adopts another, this contrast implies not just a different lin­ guistic choice, but a different attitude to Christianity and a different policy o f Christianisation. Another problem posed by linguistic loan traffic is whether we are dealing with prim ary or secondary loans. In Gothic, for example, although the majority o f Christian loanwords were taken as primary loans from Greek, this does not exclude the possible influence o f Christian Latin (armahaírtiþa may be connected with misericordia), even in the case o f words which Latin had itself acquired from Greek. 39 Bartelink (1974), 399.

40 Green (1965), 140fr., 322fr.

286

Contact w ith Christianity

Candidates for such secondary loans (Greek > Latin > Gothic) have been seen in aiwaggelista and aikklesjo, but even more strikingly in Iesus Xristus.4] In such cases Greek words enter Gothic as secondary loans via Latin, alongside the more com m only attested direct route. Secondary loans like these, reflecting the expansion o f the new religion, occur fre­ quently elsewhere.4 142 T h e mobility suggested by these examples makes it imperative to inquire into the dynamics o f such loanword traffic, to ask what histori­ cal forces carry specific words so far. H ow necessary this is, even if an answer is not always possible, can be shown by a negative example. T h e proposal that Gothic Christian terms could have reached O E through the Frankish interpreters taken by Augustine on his mission to K ent has been rejected on the grounds that the O E attestations occur in isolation in a few glosses o f the tenth century which cannot have miraculously preserved traces o f these interpreters’ linguistic practice centuries earlier.43 H ow it is sometimes possible to answer this kind o f question can be illustrated from Gothic, this time with regard to classical loanwords in Wulfila’s text. As far as Greek is concerned the evidence consists o f bib­ lical place-names (and matching ethnic names) which fall into two classes: on the one hand names which have adopted Gothic flectional endings and phonology (e.g. e > i; ia > ja) and have been ‘Gothicised’, and on the other hand names which have retained their Greek inflection and phonology (e.g. e = at), thus representing Hellenised forms.44 Behind this linguistic distinction a wider difference is detectable. W hereas the second group is largely made up o f names from the H oly Land with which the Goths became acquainted only with the com ing o f Christianity (e.g. Galeilaia, Iudaia, bradâtes), the first group includes names o f the Empire known to the Goths in the Pontic area, Asia M inor and the Balkans from their earliest raids in this region before any mis­ sionary activity amongst them (e.g. Galatja, Makidonja, Makidoneis).45 W hereas Wulfila’s Hellenising Gothic and his linguistic loyalty to the bible meant that the one group retained Greek features, the other group was Gothicised by having been adopted into the language o f the barbar­ ian raiders before they heard o f the H oly Land from Christianity. 41 Jellinek (1923), 443fr.; C orazza (1969), 92. 42 E.g. Greek > Gothic > O H G (phqffo), Greek > Latin > O H G ( O E > O H G (goispel), even Greek > Latin > Irish > O E > O H G (einchoran). 43 Gneuss (1992), 122i. 44 Schulze (1907), 165fr.; Gaebeler (1911), iff. 45 Rappaport (1899); Thom pson (1966), index; Wolfram (1988), 52fr.; Schwarcz (1992), 47fr

Problem s o f Christianisation

287

N ot all the classical loanwords o f Christian content in Gothic came from Greek, however, for a few stemmed from Latin. If these suggest Latin- as well as Greek-speaking Christians amongst the Goths before the days o f Wulfila, they can best be sought in Dacia, where archaeolog­ ical finds o f a Christian nature from the first half o f the fourth century, confined to what was then the Latin-speaking area, imply Christians amongst these Latin-speakers. Even later, in Wulfila’s day, this Latin presence was still important: he enjoyed imperial backing, the emperors in question (Constantius II, Valens) were Latin-speaking and had Latin­ speaking clerics in their households.46 For these Latin terms as for the Greek ones in Wulfila’s Christian vocabulary it is not sufficient to base our argument on language alone when the historical evidence can so strengthen our case. T h e expansion o f Christianity into new language-areas thus pre­ sented a number o f linguistic difficulties, both for those who had to devise a vocabulary from scratch in a barbarian vernacular and for the new converts. These difficulties come to a head in the choice between two methods o f translation which the Church inherited from classical antiquity.47 Translation could be either literal (word for word) or loose (sense for sense). As far as the bible is concerned, the first method rests on the conviction that in a sacred text every word had its meaning and that, as Jerome maintained, even the word-order concealed a meaning. It is a method which was followed by Wulfila, who tried to remain as true to the Greek text as with his Hellenised place-names, but which pro­ duced a Gothic text which must have been to a considerable extent unin­ telligible to Goths without Greek.48 T h e second (freer) method follows the meaning rather than the word and is best represented in O H G by Notker’s translation o f the psalms where, by contrast to the Windberg and Trier renderings with their regular equations o f Spiritus with geist, the St. Gallen author used a variety o f additional vernacular terms for Spir­ itus, dictated by context and ease o f understanding.49 Both in theory and in practice the newness o f Christianity produced problems in the creation o f a religious vocabulary: for Greek, for Latin and for the barbarian vernaculars. These difficulties are also reflected in the problems facing modern philology in interpreting these early 46 Jellinek (1923), 446i.; Thom pson (1966), 82; Gschwantler (1976), 176; Schäferdiek (1978a), 498. 47 Beck (1981a), 579; Copeland (1991), 9fr. 48 K ahl (1956), 95. Similarly, Gregory the Great lamented the difficulties caused by a literal trans­ lation. Cf. Copeland (1991), 51. 49 Betz (1957), 5if.

288

Contact w ith Christianity

endeavours, which can be illustrated by an example central to the Christian message: the choice o f a verb for ‘to believe’. T h e pagan religions o f antiquity have been described as essentially concerned with cult acts: pagans performed rites, but held no creed or doctrine.50 This attitude dictated the way in which the pagan Rom an state dealt with Christians in its midst: they were required to perform a nominal act o f worship to the gods without attention to what their beliefs might be. T h e pagan religion therefore had no fear o f heresy, no need to impose orthodoxy as long as the required ritual acts were performed. W hat was true o f the Rom an religion was also applicable to the Germ anic one. T h e Gothic Christian Saba was expelled from his com ­ munity not because o f the unacceptability o f his Christian opinions as such, but because he offended against the tribal gods by not taking part in a ritual meal.51 In O N the verb tma ‘to believe’ (in the gods) could be seen in terms o f biota ‘to sacrifice’, whilst the epithet goðlauss was used o f someone no longer willing to sacrifice to the gods.52 Perhaps because o f this association o f paganism with ritual cults the G erm anic word *blotan ‘to sacrifice’ was not christianised in W G and N G (it was applied only to pagan practices). T h e Christian use o f the same verb in Gothic, however, is only an apparent exception, for its meaning was generalised to ‘to serve, worship’ and it was used o f fasting, prayer and good works, whilst the meaning ‘to sacrifice’ (by pagans or Jews) was transferred instead to hunsljan.53 By contrast with this ritual nature o f paganism the concept o f belief is central to Christianity both in the sense o f trust in G od, entrusting oneself to Him, and in the sense o f holding tenets o f the new religion to be true.54 This concept created the need to find a word which could be christianised, a need which was met by the same term in Gothic (galáubjan) and W G (O H G gilouben, O S gilôbian, O E getyfan), whilst N G chose a different solution. T h e verb galáubjan, together with its modern G erm an cognates erlauben, loben, lieb, suggests a word-field whose central semantic function embraced such ideas as love (cf. Go. Hufs), praise (cf. O H G lobôn) and trust or hope (cf. Go. lubáins). Galáubjan therefore meant something like ‘trust’ before it was adopted for Christian purposes, whilst usláubjan meant ‘to entrust (oneself)’. It is probable that the earliest construction was galáubjan + dative (commonly attested with Wulfila), implying trust or belief in someone. O ther constructions were probably influenced by 50 Lane Fox (1988), 31, 253; Bartelink (1974), 406. 51 Thom pson (1966), 68. 52 Ström (1951), 23; Ljungberg (1947), 171. 53 Kuhn (1978), 313. 54 Ibid., pp. 31of. ; M eid (1992), 5051!

Problem s o f Christianisation

289

Christian Greek or Latin. Following the model o f Greek with the preposition eis Wulfila also used du and rarely in, whilst the Latin construction credere in is echoed by O H G in, O S an, O E on (these last two also using te and to). T h e difference between these alternative construc­ tions (with a simple dative or with a preposition) is still detectable in the contrast between ‘to believe someone’ and ‘to believe in G od ’. As far as their gods were concerned the question o f believing in their truth was irrelevant to Germ anic pagans, just as they had little cause to doubt the existence o f the Christian God. W hat counted was their trust in the gods’ power to assist the tribe, so that the missionaries’ task was to call this into question and to awaken trust in the superior power o f the Christian G od.55 Initially, therefore, galáubjan remained close to the concept o f trust: Wulfila used it to render a verb like pisteúein (but not a verb meaning ‘to think, be o f the opinion’), just as the O H G Tatian regu­ larly translated credere by gilouben, but putare and arbitrait by wânen. H ow gilouben could come to imply belief in the truth o f something as well as trust in someone can be illustrated in O H G versions o f the Creed, where a formula like Ich kelouben an got may well concern trust in the power o f the new G od (alemáctigen), while the subsequent construction (Huh kelouben. . . daz thie drîgenenneda ein got ist) refers to acceptance o f the truth o f an article o f religious belief.56 A comparable semantic extension can be demonstrated with another verb meaning ‘trust, confide’ which also lent itself to Christianisation and thus also came to mean ‘believe’: O E (ge)truwian, O S (gi)trûên. H ow this extension could come about we can see in the O S Heliand. O n one occasion Christ’s miracles and the help He brings are summed up in their effect on beholders (2350: that sie gitrûodin thiu bel), which could amount to their trust in His superior power. This half-line is then varied by another (2351: gilôbdin an is lera) which equates gitrûên with gilôbian, but also, by referring to Christ’s teaching, suggests belief in the truth o f what He says. Taken together, these two half-lines move from trust in Christ to holding His sayings to be true, in much the same way as in the Creed formulas. O n another occasion, even without the equation with gilôbian, a similar extension is shown: Christ’s miracles may again awaken trust in Him (3114: that sie gitrûodin thiu bet), but the introduction o f the follow­ ing line with the conjunction ‘that’ (3115: that he selbo uuas sunu drohtines) shows that this trust is subsumed under belief in the dogmatic truth that Christ is the Son o f God. 55 Kuhn (1978), pp. 312,314; Ljungberg (1947), 170.

56 Kuhn (1978), 3i2f.; Steinmeyer (1916), 341.

290

Contact w ith Christianity

T h e use o f (ge)truwian and (gi)trûên as Christian variants o f gelÿfan and giloUan in O E and O S explains the religious employment o f the cognate tma in O N ,57 where the satisfactory use o f a verb corresponding to galáubjan was excluded by the regular reduction o f prefixes. In O N tma could therefore be used with the dative to mean ‘to trust’ in a non-reli­ gious context, but also in a pagan religious context (with reference to Oðin), possibly under Christian influence. Finally, the construction tma + á (where the preposition corresponds to W G an or on) occurs in a reli­ gious context with Christian poets. Like the religious use o f galáubjan in Gothic and W G, the similar use o f tma in O N appears to be the outcome o f Christian influence, where the model for Scandinavia was O E or O S (from both regions missionaries were active in Scandinavia), more prob­ ably O S .58 Several problems remain unresolved with galáubjan. T h e suggestion that the W G forms are the result o f a Gothic mission59 runs up against doubts about this theory and also the difficulty o f explaining how it may have reached far enough to account for O S and O E . If the choice o f this verb was an obvious one in Germ anic and if the prepositional construc­ tions were based on Greek and Latin models, the influence o f these lan­ guages could have independently produced similar results in different Germ anic languages. This possibility o f polygenesis is likely for the O E and O H G forms (there is no evidence that Anglo-Saxon missionaries were responsible for introducing their word to Germany), whereas it is quite conceivable that the O S form was influenced by either O E or OHG. 37 Ljungberg (1947), 151fr.

58 Ibid., p. 168.

59 Kuhn (1978), 310.

CHAPTER

l6

T h e influence o f provincial Rom an Christianity

T h e earliest contacts o f the G erm ani with the Romans, and therewith with Christians am ong them, took place along the Rhine and Danube frontiers o f the Empire from the second century, but cannot have been very extensive before Constantine’s conversion in 312 and the declara­ tion o f Christianity as the state religion. Christian communities on the Rom an frontier must have kept a low profile during the pagan period and were hardly obtrusive enough to attract the attention o f G erm anic outsiders. Even extending the period to be surveyed down to the conversion o f Chlodw ig and his followers does not result in a large number o f terms taken over into G erm an from provincial Rom an Christianity because widespread Christianisation set in only after the fifth century.1 T h e historical evidence for early Christianity in the frontier region o f the Empire in what is now Germ any presents a picture markedly differ­ ent in the south (with the Alem anni and the Bavarians) and in the Rhineland (concerning the Franks). In south G erm any centres o f Rom an Christianity are known above all in towns: in the area later occu­ pied by the Alem anni in Basel, Konstanz, Chur and Augsburg, in what was known later as Bavaria in Lorch, Salzburg, Passau and possibly Regensburg.2*This situation in the south, permitting even an early epis­ copal organisation, was drastically changed with the Alemannic penetration to the foothills o f the Alps and with Bavarian ethnogenesis in the first half o f the sixth century. T h e disturbances which these barbarian invasions brought to the incipient Christian communities are reflected in the evidence for episcopal seats being pulled back from threatened areas in the sixth century. Despite the large-scale disruption which this suggests there is evidence for the survival o f scattered pockets 1 Eggers (1978), 473t 2 Büttner (1951), 8fF.; M ilojčič (1966), 231fr.; Schäferdiek (1976), 181; Ewig and Schäferdiek (1978), 117L Prinz («985)> 329 -

291

292

Contact w ith Christianity

o f (Romance-speaking) Christian groups in these regions now under Germ anic occupation. T h e position in the Rhineland was different in two respects. First, although the extreme north-west presents a picture o f episcopal retreat similar to the south, the sector o f the Middle Rhine and Mosel showed greater continuity and thus presented more opportunities for the Germ ani to make their first acquaintance with Christianity, however external.3 Secondly, the Rhineland was more important for the Romans than south Germ any for a number o f reasons, which means that Latin loanword traffic (and therewith the possibility o f Christian terms) is better attested from Gaul to Germ any than over the Alps.4 It was through Gaul that the Romans’ most important line o f communication ran: up the Rhone to Lyon, thence to Trier and the Rhine frontier with such bases as Xanten, Cologne, Bonn and M ainz. T h e geographical importance o f Trier on this route was heightened by the presence o f the emperor’s court there in the fourth century. All this brought trade, administration, agriculture and a greater military presence in its train, amounting to a more thorough Romanisation o f this area than in south Germany, reflected in Latin loanwords which entered G erm any from Gaul. Christianity penetrated Germ any from the same direction, radiating from Gaul to the Roman provinces G erm ania I and II which belonged politically to Gaul. Particularly important was Lyon with its earliest organised Christian community in Gaul (attested from a persecution in 177) linked to a Greek colony from Asia Minor, especially amongst the merchants.5 O f the martyrs o f the persecution whose names are known a good half were Greeks. Above all in the frontier region the officials o f the pagan Roman state had every reason to proceed as ruthlessly against what they regarded as a disloyal religion as in the hinterland o f Lyon. T h e towns in which Christianity first gained a foothold in the Empire were also fewer in the Rhineland than in Gaul and in any case under the watchful eye o f Roman civil and military administration. There is there­ fore no evidence, written or archaeological, for a Christian presence in these provinces before the second half o f the third century. Indeed, Germ ania I and II and Belgica I, together with Raetia and N oricum in the south, underwent sparse Christian influence before the conversion 3 Büttner (1951), gf.; Schäferdiek (1976), 181; Ewig and Schäferdiek (1978), n6f.; Prinz (1985), 329. 4 Kuhn (1951a), 57f., 60; Hatt (1963), 63ff. 5 Hegel (1962), 94; Dem ougeot (1963), 23fr; Petrikovits (1978a), 252; (1978b), 576f.; Schäferdiek (1987), 150t

T h e influence o f provincial Rom an Christianity

293

o f Constantine. W hat evidence there is suggests that Trier was influ­ enced by Lyon and in turn passed on some Christianity, however spo­ radic, to the Rhineland.6 We are dealing up to the end o f the third century with a small number o f Christians whose sparse communities were not organised under episcopal leadership. A further factor making for a reduced presence o f Christianity was the persecutions before the Edict o f Toleration in 313. It produced its crop o f Christian martyrs whose memory, albeit with pious accretions, is preserved with what is presumed to be a historical nucleus in legends o f martyrs (e.g. the Theban legion, St Ursula) or o f bishoprics (e.g. concerning Maternus, Helena). T h e legend o f the Theban legion suggests the possibility that some o f these early martyrs may have come from the ranks o f the Roman army after their conversion had brought them into conflict with the authorities.7 T h e fourth century presents a totally different picture: the victory o f Christianity. After the Edict o f Toleration it was possible to practise the new cult in public and Christianity could develop freely in the Rhineland with some measure o f organisation. N ot merely did the number o f Christians increase significantly, they also made their victorious presence more obvious, even ostentatiously, as with the programme o f Constantinian church building, reaching Trier (cathedral and the Liebfrauenkirche) and Cologne (St. Gereon). Christian symbolism also appeared on coins struck by Constantine, as did the Christian labarum on the battle-standards o f his troops.8 How different the position now was from the earlier martyrdom o f Roman soldiers because o f their Christian allegiance is illustrated by a complex o f five military grave­ stones o f the last third o f the fourth century from Cologne. Their inscriptions commemorate soldiers from the imperial retinue: two o f them were certainly Christian, one probably.9 A n indication o f the more thoroughgoing organisation o f the Church is the creation o f bishoprics in the Rhineland by the middle o f the fourth century.10 In this an early lead was given by Trier, where the bishopric was founded in the middle o f the third century with responsibility for both Germ anic provinces before they received their own bishops. In these provinces Cologne provided the best conditions for the rise o f a large Christian community, yet the first bishop known there, Maternus, was active only from the beginning o f the fourth century. He seems to 6 Schäferdiek (1987), 151; Demougeot (1963), 39fr. 7 Hegel (1962), 93ÍT.; Petrikovits (1978b), 582fr. 8 Schäferdiek (1987), 152. 9 Petrikovits (1978a), 255. 10 Hegel (1962), 96fr.; Petrikovits (1978a), 253fr.; (1978b), 593fr.

294

Contact w ith Christianity

have been responsible for creating the bishopric o f Tongeren by 342, when Servatius is attested as a bishop attending a council o f the Church. T h e evidence for M ainz follows, whose first named bishop, Mārtiņus, took part in another council in 346, where bishops are also named for the first time for Worms, Speyer and Strassburg. In addition to this incipient episcopal organisation o f the Rhineland the ecclesiastical hier­ archy is attested by occasional references to other clerical ranks: a deacon accompanied Maternus o f Cologne to a synod at Arles, we hear o f a subdeacon from Trier, a lector is known from what is probably Koblenz, and a presbyter is attested for Bingen.11 These scraps o f evi­ dence conform to the general picture o f early Christianity in all refer­ ring to clerics in towns. T he presence o f Christianity in the countryside was certainly much weaker. T h e early Christian communities in Trier and Cologne resembled what was true o f the community at Lyon from which they had received their religion: they all show traces o f an eastern, specifically Hellenistic presence.12 Artistically this is true o f sarcophagi found at Trier and o f glassware at Cologne, tombstones at Trier had inscriptions in Greek or bore Greek names, just as the name o f the Trier bishop Eucharius sug­ gests a Greek. T h e same is true at Cologne with bishop Euphrates. However decisive the victory o f Christianity in the fourth century may have been, the timespan available for the Christianisation and ecclesiastical organisation o f the Rhineland after the Edict o f Toleration should not be exaggerated, since already by the fifth century this area witnessed a repaganisation.13 T h e reasons for this include the per­ sistence o f pagan Rom an cults: even after the conversion o f Constantine there was at first no official policy o f destroying pagan sanctuaries, least o f all in the countryside where Christianity had scarcely made headway. Pagan practices continued even in the neighbourhood o f Christian centres, as with a temple o f Isis outside the townwalls o f Cologne (on the site o f the later church St. Gereon) or with two Mithras temples. Such pagan sanctuaries were still frequented by worshippers as late as the middle o f the fourth century or even until its close, so that Christianity faced considerable opposition from Roman and oriental pagan religions even in the period o f its victory. T h e balance was decisively tipped against the progress made hitherto by Christianity with the Franks’ onslaughts on the Roman frontier dis­ 11 Petrikovits (1978a), 257; (1978b), 603t. 12 Hegel (1962), 98; Dem ougeot (1963), 43; Frend (1964), 125fr. 13 Petrikovits (i960), 140; (1978a), 260; (1978b), 610.

T h e influence o f provincial Rom an Christianity

295

trict which began early in the fifth century. In 420 they took temporary possession o f Trier, by about 440 Cologne was occupied by Franks and M ainz largely destroyed, while other sectors o f the Rhine frontier were subjected to attacks by other tribes (Alans, Vandals, Burgundians, Huns), so that by the middle o f the century the whole o f the west bank was in Germ anic hands. This meant a considerable setback for the Church, the more so since its recent gains had been confined to towns destroyed or seized by the barbarians. T h e gravity o f this setback is effectively sug­ gested by the episcopal lists recorded for the Rhineland.14 Whereas Reims and Trier, further removed from the zone o f greatest danger, record no interruptions in episcopal succession from the third century, the lists for Cologne, Tongeren, M ainz and Worms show significant gaps in the approximate period 460 to 550. In this period the Church must have suffered considerable disorganisation as a result o f the invasions, not merely as the victim o f active pagan hostility, but simply because economic dislocation and impoverishment made it financially impossi­ ble for the native population to build (or rebuild) churches and sustain an ambitious ecclesiastical organisation.15 It is not surprising that pagan cults persisted in this period: those connected with the Germ anic relig­ ion introduced by the invaders, but also those surviving from the Roman and pre-Roman period, especially among the rural population. T h e long-term result was that when a missionary movement began work in this area (in the sixth, but mainly in the seventh century) it had to face two types o f paganism. In this it was to some extent able to build on scat­ tered Christian communities which had managed to weather the storms o f the invasions. O ne last detail concerns that aspect o f Christianity which presented itself most obviously to the incoming Germani: its church buildings, especially in the demonstrative form encouraged by Constantine as an outward token o f the victory o f Christianity. For obvious reasons there is no evidence for such buildings in the age o f persecutions, but after this an approximate dating has been suggested for various centres.16 A t Trier a place o f worship was present towards the end o f the third century, but the twin building (cathedral and Liebfrauenkirche) was begun in 326 and was constructed on a striking scale. A t Cologne St. Severin was begun towards 320, St. Gereon was possibly influenced or encouraged by Constantine, whilst St. Ursula was standing by 355, when it was 14 Ibid. (1978a), 292; (1978b), 623. 15 Ibid. (1978b), 623i 16 Hegel (1962), ioif.; Dem ougeot (1963), 41fr.; Gaudem et (1963), 8; Hatt (1963), 58, 63.

296

Contact w ith Christianity

destroyed by the Franks. A cathedral at Bonn was built in the fourth century. There is no comparable evidence for the building o f churches in the period o f disruption brought about by the barbarian invasions. A new beginning was made possible only after the conversion o f Chlodw ig and the mission o f the sixth and seventh centuries.17 T h e linguistic aspect o f our survey is confined to four Christian loan­ words which probably entered Germ an from Gaul in this period. T h e first, the word for ‘church’, ties up with what we have seen o f church buildings in the Rhineland and with the presence o f Greek speakers along this western route. In the earliest days o f Christianity the private houses o f individual believers served as places for meeting and worship, so that there was at first no technical term for the place o f worship: its unobtrusiveness was adequate to what were still small groups o f Christians and was in any case advisable in the age o f persecutions. This began to change in the third century with the spread o f the new relig­ ion and with the measure o f toleration with which it came to be regarded by the Roman state, so that only now do we encounter techni­ cal terms for the Christian church building.18 T h e terms designating this building in most European languages go back to three Greek words from the third century: ekklēsia, kyriakón and basilike.19 T he first was taken via Latin into all the western Rom ance lan­ guages and Celtic, the second was loaned into Germ anic (thence also into Slavonic), whilst the third survives in eastern Rom ance (but also in isolated pockets in the west). T h e questions which this distribution poses are: how was it that Germ anic did not take over the ecclesia o f the western Church, when and where did it make contact with the second o f these Greek terms? Originally, Greek ekklēsia had a political function before Christianity, meaning ‘summons’, then ‘assembly o f citizens’, but from the third century it is attested in a Christian context meaning ‘congregation’. This meaning o f the word was extended to the building where the assembly was held {pikos ekklēsias). T h e word established itself quickly as a techni­ cal Christian term in Greek and was borrowed in the third century into Church Latin with both the meanings now present in Greek: ‘Christian community’ and ‘church building’ . From Latin it passed into western Romance (e.g. OFr. glise, OPr. gleize) and Celtic (OIr. eclis, We. eglwys).20 17 Petrikovits (1978a), 292, 297. 18 Kretschmer (1906), 539Í; M ohrm ann (1977), 211, 212. 19 Kretschmer (1906), 539fr.; D ölger (1950), i6iff.; M ohrm ann (1977), 21 iff.; Masser (1966), 17fr. 20 Kretschmer (1906), 54.of.; M ohrm ann (1977), 215i; F E W 3, 203.

T h e influence o f provincial Rom an Christianity

297

Barely later than ekktēsia the second term, kyriakón, occurs for the first time in the third century. Its formation (‘that which belongs to the Lord, the house o f the Lord’) made it more acceptable in this sense than ekklēsia, for there are suggestions that this word’s extension to ‘building’ was felt to be incorrect. Nonetheless, although kyriakón was in use throughout the fourth century, it eventually succumbed in Greek to ekklēsia and died out by the fifth century, so that it must have made contact with Germ anic during the century dominated by Constantine.21 Although this Greek word was not adopted as a loanword into Latin, it left its mark indirecdy in a loan-translation, dominicum, which shadows it also chronologically: attested first in the third century, it was in use in the fourth and died out in the fifth century. Dominicum appears not to have been a native Latin formation, but dependent on its Greek model, whose eclipse it accompanied.22 Basilikē was taken into Latin as a secular word denoting a large public hall used for trade or meetings, but acquired a Christian function mainly as a result o f Constantine’s programme o f church-building, particularly the Basilica Constantini at Jerusalem (the H oly Sepulchre). O nly in this period were churches first erected o f a size imposing enough to merit such a term (originally ‘royal’). T hat this word’s Christian adaptation was later than with dominicum is suggested by a passage in the fourthcentury Itinerarium Burdigalense, where the word basilica is employed, but has to be explained by the better known dominicum. Like dominicum, basil­ ica differed from ecclesia in denoting only the church-building.23 Apart from its survival in eastern Rom ance (e.g. Rum. biserica) the word sur­ vived only peripherally in Gaul, where OFr. basoche was applied to the basilica o f St Martin at Tours and occurs as a place-name in western, northern and eastern France.24 From the distribution o f these three terms kyriakón emerges, especially as regards the Latin west, as a foreign element in the competition between ecclesia and basilica. Early intensive Christianisation established ecclesia in Italy and southern Gaul, whereas basilica was brought as a newer term in the fourth century to areas evangelised only later. Given the conflict between these two words, we have to ask how kyriakón fits into this picture and how it found its way into Germ anic. I f the Greek word was used mainly in the fourth century, but hardly later, then it must have crossed the linguistic frontier during this short 21 Kretschmer (1906), 541; M ohrm ann (1977), 222; Masser (1966), 19. 22 Kretschmer (1906), 543; Dölger (1950), 175ÍT., 189; M ohrm ann (1977), 223. 23 Kretschmer (1906), 543; Dölger (1950), 172fr.; M ohrm ann (1977), 228. 24 F E W 1, 270.

298

Contact w ith Christianity

timespan, but how and where? T hat there is a connection between kyriakón and a Germ anic word which on W G evidence must have been orig­ inally *kirika there can be no doubt, even apart from minor discrepancies. T h e least troubling is the abandonment o f accentuation o f the final syllable in Greek in favour o f regular initial accentuation in Germ anic. T h e loss o f Greek medial -a- in the Germ anic form has par­ allels in Vulgar Gr. kyrikós alongside kyria/cós.25 T h e change o f gender to Germ anic feminine is best explained by the w ay in which this word is inserted into the conflict between ecclesia and basilica, from which it takes over their gender.26 Finally, the change in the stem vowel fromjy (= Ü) to i can be accounted for either from Greek (assimilation o f y to following i) or from Latin (replacement o f Greek y by i, criticised already in the third/fourth-century Appendix Probi).27 W ith these qualifications we can explain how the Greek word issued in a Germ anic form *kirika, attested in O H G , O S, OFri. and O E (already from the late seventh century) and later carried to O N . In Germ anic the word became so firmly established that it remained unaffected by the subsequent victory o f ecclesia in western Romania, just as other Latin loanwords crossed the linguistic frontier into Germ an before they were replaced by newer terms in Rom ance alone.28 A t first O H G kirihha, denoting the place o f worship, did not correspond fully to ecclesia, denoting also the Christian commu­ nity, as is clear from the Abrogans, where ecclesia in the first sense is ren­ dered by chirihha, but the abstract meanings by ladunga ‘summons’ and samanunga ‘assembly’. In the course o f time kirihha came to be so much the regular equivalent o f ecclesia that its meaning was extended under the influence o f the latter.29 T h e earliest suggestion (going back to Walahfrid Strabo) for the route by which this word found its way into G erm an invoked transmission by the Goths, an explanation which seemed persuasive with a Christian term o f Greek origin, present as a loanword in Germ anic, but not in Latin, especially if we take into account the conversion o f the Goths in the fourth century and the undoubted influence o f Greek on Gothic Christian vocabulary at large.30 This earlier view still survives unques­ tioned with some non-Germanists,31 although it has been decisively questioned on linguistic grounds. In the first place, a word correspond­ ing to W G *kirika is nowhere attested in Gothic. Wulfila had no call for 25 27 28 31

Kretschmer (1906), 542; Masser (1966), 19f. 26 GR 2,118. Keil (1864), 4, 147 {gyrus non girus). Cf. Benediktinerregel (54), p. 37 (kirieleison). See above, p. 244. 29 Masser (1966), 28f. 30 Kluge (1909), i25f. Dölger (1950), i94f.; M ohrm ann (1977), 222.

T h e influence o f provincial Rom an Christianity

299

a word ‘church’ in his bible translation, but he used aikktēflo in the sense o f ‘assembly, congregation’, whilst the Gothic Calendar used it meaning ‘house o f G od ’.32 Both meanings o f Greco-Latin ecclesia are therefore attested in Gothic, but with no suggestion o f a loanword from kyriakón, so that it is risky to base the spread o f a word through G erm ania on an unattested starting-point. N or is the assumption that the Slavonic term points to a Gothic one from which it was borrowed any more reliable as long as Slavonic could equally well have acquired its term from G erm an.33 In addition, whereas Gothic Christian terms which found their way into Germ an from the south-east reveal an understandable centre o f gravity in Bavaria, only rarely spreading beyond this, there is no suggestion o f this with kirihha. Instead, the earliest attestations o f the word, occurring in place-names from the beginning o f the eighth century, but pointing to an earlier period when a place could be characterised by the by no means self-evident presence o f a church, show no such concentration. In the eighth century some o f these placenames m ay be from Bavaria, but others are in Alsace, Francia and Frisia: an early distribution in the west which we should not expect with a word transmitted by the Goths.34 Finally, doubts are called for by another western weighting o f the evidence, the O E form cirice, for we have grown more sceptical about the ability o f Gothic words to travel ‘up the Danube and down the R hine’ to reach the Anglo-Saxons so early, without any evidence o f a mission to carry them. Attention has turned instead to the possibility o f a western transmis­ sion o f the Greek word into W G. In principle, it is unsafe to assume that a Greek word must have entered Germ ania from the east, for this ignores the presence o f Greek trading communities, including Christians, in the west: their main base was Marseilles, from where a trade route ran to Lyon and onwards to Trier and the Rhineland.35 Christianity as well as trade m ay have been carried along this route, but this tells us nothing about this specific word. Nor is it enough, although it is certainly signif­ icant, that Constantine’s programme o f ostentatious church-building in the fourth century throughout the Empire also left its traces in Trier and Cologne.36 These were buildings o f a size and importance to attract attention, even on the part o f pagan Frankish mercenaries in the Roman arm y who did not have to be Christians themselves to feel the need o f a word with which to designate such buildings. 32 Masser (1966), 23Í 33 Ibid., pp. 24i'. 34 Ibid., pp. 32fr. 36 GR 2, u6ff; Krautheim er (1965), i8f., Herrin (1989), 114.

35 Ibid., pp. 2iff.

300

Contact w ith Christianity

This raises the question whether Gr. kyriakón was in fact taken over in this context, whether its presence can be assumed in the west. T h e first indication is too specific to bear much weight, but it is suggestive that Athanasius o f Alexandria, in exile at Trier during the fourth century, used the Greek word in his Christian writing, for with that the question whether the word found its way to G erm ania in the short span o f its life is answered, on however narrow a basis.37 Secondly, the presence o f the Greek word is suggested on a wider footing by its adoption as a loantranslation in Lat. dominicum, especially in Gaul, where it is attested with Hilarius o f Poitiers, in the Itinerarium Burdigalense, and in the loan o f the word into Irish domhnach, used as an element in place-names.38 T h e use o f a neuter adjective as a substantive to denote a place o f worship, common to the Greek and Latin forms, is not so obvious as to exclude the dependence o f the Latin word on the Greek model, which it further matches in its brief chronological span.39 Thirdly, W G *kirika as a loan­ word from Greek shows that even in the west kyriakón was not completely ousted by dominicum and that it survived long enough to cross the lin­ guistic frontier. We must assume that this loan into Frankish took place in the fourth century, while the Greek model was still in active use, presumably in Trier, hence passing to Germ anic mercenaries serving there. T h at would be far too narrow a basis to explain the word’s expansion, for while these mercenaries may have needed a word to describe what they saw around them they had no occasion to pass the word further into the G erm anic hinterland where there was no call for the word. This need must have grown in the fifth century, however, with the Frankish occupation o f Cologne and Trier, so that we must distinguish between the first encounter with the word (fourth century, Germ anic mercenaries) and its wider adoption (fifth century, Frankish occupation). Neither o f these stages presupposes christianised Franks, merely speakers requiring a word for this aspect o f Roman life. By contrast, Christianity lies behind the later spread o f the new vernacular word to other G erm an dialects, part o f the subsequent expansion o f Frankish power east o f the Rhine.40 In connection with this expansion o f the word from its point o f entry in the lower Rhine/M osel area a problem presents itself: how did it find its way to O E? Pagan Franks had no interest in passing on such a term beyond the area where it was relevant, any occasional Frankish converts 37 Dölger (1950), 193; Petrikovits (1978a), 258. 38 Dölger (1950), 17 2 ,1 8 1 ,187; M ohrm ann (1977), 223; Schäferdiek (1984b), 50, fn. 17. 39 Schäferdiek (1984b), 48. 40 Ibid., p. 50.

T h e influence o f provincial Rom an Christianity

301

on Rom an soil were not active as missionaries in Germ ania, whilst Latin loanwords which m ay have reached the Anglo-Saxons while still in Schleswig-Holstein were trading words, i.e. the word accompanied the thing, precisely what could not happen with a word for a church build­ ing.41 T h e first opportunity the Anglo-Saxons had o f adopting ‘static’ loanwords from Latin as opposed to the mobile ones o f trade was when, moving along the North Sea coastline towards northern France, they settled (temporarily) on former imperial territory where they encoun­ tered such novelties and, like the Franks, felt the need to name them. Yet precisely this part o f Roman territory had been rarely touched by Christianity and had few church buildings o f any size or number, so that there is little chance that the word reached the Anglo-Saxons in this area.42 Instead, it is more likely that the word entered O E once they had reached Britain, as a loan from Frankish. T h e occasion could have been the marriage o f the Frankish princess Bertha to Æthelberht o f Kent, for in the last third o f the sixth century she worshipped at a church just outside Canterbury, dedicated to St M artin o f Tours.43 O r the loan took place slightly later with Augustine’s mission to K ent ca. 600 (in which he made use o f Frankish interpreters), for he converted pagan temples into churches, re-used Romano-British churches and founded a monastery at Canterbury.44 Either occasion called forth the need for an O E word for ‘church’ and Frankish affiliations in either case explain how the Frankish term reached England, just as it was spread further in Germany. In some respects the position is very similar with the second example, the word ‘bishop’. Its origin was likewise Greek: eptskopos ‘overseer’, applied at first to someone in charge o f a Christian community, but already in the second century to one in authority over presbýteroi ‘priests’.45 T h e word was adopted in Church Latin and is also found in Germ anic (Gothic, O H G , O S, OFri., O E and later in ON). Within Germ any the word was present before 800 over the whole o f the Upper and Middle Germ an area. Like episkopos, the Germ anic derivatives meant ‘bishop’, even if they can sometimes be equated in O H G and O E with ‘priest’.46 T h e Goths have been invoked here, too, as intermediaries between Greek and W G, although the case for this is even weaker.47 In the first 41 43 44 46 47

See above, pp. 2i6f. 42 Schäferdiek (1978a), 535; Ewig and Schäferdiek (1978), 116. Schäferdiek (1978c), 152; Richter (1982), 132; Prinz (1984), 315fr. Schäferdiek (1978c), 154, 156, 158; Masser (1966), 22, fn. 30. 45 Waag (1932), 2of. Ibid., pp. 23fr.; Rotsaert (1977), 198, 201. Kluge (1909), i35f. O n this see W aag (1932), 29 and Rotsaert (1977), i84f.

302

Contact w ith Christianity

place, Wulfila used aípiskaúpus, a learned ecclesiastical form which differs markedly from W G *biskop. O ne o f the most telling facts that suggested Gothic transmission o f kyriakón to W G (the absence o f this Greek word from Church Latin) does not apply to ‘bishop’, where Latin has episcopus, so that there is less compulsion to look for an eastern transmission. It is an oversimplification to assume that a Christian term o f Greek origin (even when there is no Latin parallel) could only enter Germ ania in the east and that the Goths must therefore have been involved. Lasdy, what constituted a difficulty for ‘church’ is also a problem for ‘bishop’: how to explain the presence in O E o f a Christian term supposedly trans­ mitted by a Gothic mission which, even if it existed, barely penetrated G erm any beyond Bavaria. If the case for Gothic intermediaries is a shaky one, it can be weak­ ened still further by an argument directed against the further suggestion that the W G forms derived direcdy from Greek or from Latin. Whereas Greek, Latin and Gothic all share the recognisable form episkopos, the W G equivalent *biskop has two features (voicing o f intervocalic -p- to -b-} then loss o f unaccented e-) which show that it could not derive direcdy from Greek or Church Latin, but must have come via Vulgar Latin.48 We encounter difficulties, however, when we turn to the V L evidence for this word in western Romania, for the different vernaculars point to a variety o f alternative forms as their starting-points. Thus, forms in northern France and the Grisons derive from *episcu (cf. OFr. (e)vesque), in south­ ern France and Iberia from *epispu (cf. OPr. bisbe, Sp. obispo), in southern Italy from *viscopu, in northern Italy from *viscovu, and in Sardinia from *piscopuý9 Although this last V L form most closely resembles the W G basis, this must be reckoned as chance, unless we assume that Sardinia happened to retain a variant which was originally more widespread. O u r problem lies therefore with the relationship not between W G and Gothic, but rather between W G and Latin, between *biskop and episcopus. T h e main phonological problem lies in the twofold presence o f an intervocalic -p- in the Latin word, the first o f which has under­ gone V L voicing, whereas the second seems not to have. In addition, difficulties were created by O E bisceop as long as it was assumed that its agreement with continental W G indicated an early loan into W G , before the Anglo-Saxons came to Britain in about the middle o f the fifth century. O ther possibilities exist, however: G erm any and England could have acquired the word from the same source independently and 4 9 48 Rotsaert (1977), 186.

49 W aag (1932), 30; GR 2, 236.

T h e influence o f provincial Rom an Christianity

303

at different times or the continental form could have been taken to England at a later date.50 T h e phonological difficulty can be divided into various stages. First must come the voicing o f intervocalic -p- (more particularly o f the first -p- in episcopus), for this is a regular feature o f V L in western Romania, which for that reason must be regarded as the source o f the W G forms. O n ly after this voicing could the loss o f the initial e- have taken place in the V L form. This was assisted by the threat to this unaccented syllable, com ing immediately before the word-stress on -i- (cf. ecclesia and OFr. glise, OPr. gleiza, whereas modern French église results from subsequent assimilation to Church Latin).51 In addition, we m ay reckon with a faulty division o f the construction definite article + substantive (*illebescobo) producing a substantive in which the initial vowel has now been lost (cf. It. I’ospedale > lo spedale).52 This still leaves unexplained the apparent discrepancy between the voicing in what has now becom e the initial b- o f the Germ anic word and the retention o f the second -p- o f the Latin word (changed to - f only in O H G by the Second Sound Shift). T h e most convincing explanation o f this discrepancy sees in it an example o f phonetic substitution, to which loans from one language into another are especially prone.53 As regards voiced labial consonants W G possessed the stop b- in the initial position, but not in some dialects medially, which knew only the fricative -h-. W ith the loss o f e- in what had becom e V L *ebiscobus the initial b- therefore remained unaffected in W G , whereas there was no equivalent available in some dialects for the medial -b-. In the intervocalic position o f this -b- these W G dialects could therefore employ only -b- or -p-, where the latter, although voice­ less, had the advantage o f corresponding more closely as a stop to the V L model. This explanation has been used to throw light on the date and area where the V L word found its w ay into Germ anic. D ating this loan is assisted by the fact that Rom ance introduces a further change into the development o f intervocalic labial consonants: after -p- had been shifted to -b- this was later changed to the fricative -h-. From this it is clear that the word was transmitted to G erm anic after the first change in Rom ance, but before the second (even though some o f the forms attested in O H G m ay have been adopted after the second change was completed).54 It would be helpful if Rom ance philologists were more in 50 W aag (1932), 29. 51 K napp (1973), 189; F E W ^ ^ i t . 52 Rotsaert (1977), 190, 195; K napp (1973), 186. 53 Pogatscher (1888), 195; Gysseling (1961), 24; Rotsaert (1977), i88f.

54 K napp (1973), i8off.

304

Contact w ith Christianity

agreement about the chronology o f these changes, but whereas there is consensus about the voicing o f -p- to -b- in the fifth century, the sub­ sequent change {-b- to -5 -) is dated either soon afterwards in that same century or as late as the seventh century. This suggests the fifth or sixth century as a time for the loan into G erm anic, possibly already in the fifth century if we are to allow enough time for the word’s spread and use in a place-name like Bischheim, attested for Alsace as early as about 530.55 As regards the location o f this loan we receive com plem entary assistance from the phonology o f W G , for the absence o f a medial -band hence the need to replace it by -p- are characteristic o f O E , Low Franconian and M iddle Franconian. We are therefore probably dealing with radiation o f the word from Gallo-Rom an to the area o f Trier and Cologne.56 Still to be accounted for is the early presence o f the word in O E where, as with ‘church’, the loan was probably from Frankish rather than G allo-Rom an. A loan o f this word into continental W G in the fifth century (let alone in the sixth) leaves hardly enough time for inter­ nal traffic within G erm ania to carry it as far as the Anglo-Saxons while still on the continent, even assuming that any G erm anic m ediator or recipient was interested in a Christian bishop. T h e continental litus Saxonicum was moreover weakly christianised, so that if the AngloSaxons were struck by no churches there before crossing to Britain they are hardly likely to have encountered a bishop or paid m uch attention to him. Am ongst the earliest Christian loanwords in O E there are some which go back to the beginning o f Christianisation or even slighdy before, and ‘bishop’, like ‘church’, probably belongs to them.57 Before any official evangelisation o f K ent the Frankish princess Bertha was allowed to practise her Christian religion and to that end had taken a bishop, Liudhard, with her from Francia, so that in a narrow circle in still pagan Canterbury there was call for a term to denote his office. I f that is too isolated a basis to account for the w ord’s history in English, the field is widened shordy afterwards with the com ing o f the papal mission: Augustine was entrusted with the organisation o f bishoprics for the English Church, had the pallium conferred on him by the pope, and brought with him Frankish inter­ preters and therewith their Christian terminology. W hereas the terms for ‘bishop’ and ‘church’ entered W G in the context o f provincial 35 Rotsaert (1977), i8gf.; Wollmann (1993), 24; W aag (1932), 22. d6 Gysseling (1961), 24, fn. 7; Rotsaert (1977), 210. 57 Wollmann (1993), 3f.

T h e influence o f provincial Rom an Christianity

305

Rom an Christianity, the same words were later brought to England under Frankish auspices.58 O u r third example, O H G offrôn, derives from Latin offerre>used in the religious sense ‘to sacrifice’, and stands in contrast to another verb, O H G opharôn, from Latin operand This contrast is primarily geo­ graphic: whereas the second G erm an form was at home in the south (SRhFr., Alem annic, Bavarian) the first was concentrated in the north­ west (OS, Low Franconian, Middle Franconian, Isidor). This north­ western centre o f gravity for offrôn is confirmed by the presence o f offrian in O E which, in conjunction with Celtic examples (e.g. offerendum: OIr. oiffrend), points to radiation o f this word from northern Gaul, reaching as far as Trier and northwards. From the geographical spread o f another word in G erm any it is assumed that operari also entered this same region from northern Gaul, not in the religious sense which underlies opharôn in U pper G erm an, but in the secular sense ‘to work’ (cf. OFr. ovrer and such north-west G erm an dialect forms as Oppermann ‘workman, handym an’).60 W hereas this north-western form was brought to G erm any by the Romans (as part o f their building pro­ gramme?), the southern import was reinterpreted as an ecclesiastical term (‘work, service’ > ‘religious service’). Because this word retained its original secular meaning in the north-west, the w ay remained open for offrôn as a Christian term.61 W hereas operari cam e to the Low er Rhine as an early Rom an loan­ word, there is no reason to assume that offerre reached this area so early as a Christian term. In dating this latter w ord’s entry into the Christian vocabulary o f G erm an we best proceed from the fact that the Latin word acquired popularity as a Christian term m eaning ‘to offer, dedi­ cate, sacrifice to G o d ’ especially in the fourth and fifth century.62 This is also the time when early attempts at evangelising Celtic Britain were undertaken from Gaul, which would account for the presence o f the word in Insular Celtic. I f offerre was therefore established as a Christian term in northern G aul in the fourth century and in that usage in Rhenish Franconian {Isidor) by the end o f the eighth century, we are left with a wide margin in which to date the loan. However, this is likely to have taken place towards the early part o f this timespan, if only because northern G aul, the reservoir from which so many Latin loan­ words were drawn, was capable o f active radiation to north-western 58 Schäferdiek (1978c), 152, 157. 59 Müller (1966), 152fr.; GR 2, 340fr.; F E W 7, 333. 60 GR I, io6f. 61 GR 1 , 107; Wesche (1937), 70. 62 F E W 7,333.

3o 6

Contact w ith Christianity

G erm any only as long as the Empire provided the wider political and military framework within which even isolated Christian terms could be transmitted. It is even possible that ojjron entered G erm an at about the same time and in the same context as kirihha and biscof. O E qffrian, for the same reasons as with cirice and bisceop, can hardly have been taken by the still pagan Anglo-Saxons from the continent to Britain. O n ce again, Augustine’s mission is a more convincing candidate, introducing to K ent either the Latin term itself or, through his inter­ preters, its Frankish form. In all these cases the linguistic agreem ent between O E and continental W G should not blind us to the chronological difference in the entry o f a loanword into these lan­ guage groups. O ur last example concerns Christian almsgiving. T h e Greek word eleēmosjmē ‘mercy, compassion’ was also used in the Christian sense o f ‘alms’ and as such was borrowed into Latin as eleemosyna. From Latin the word was taken into Germ anic (O H G , O S, OFri., M D u. and OE) and within Germ an it is found in Franconian, Alemannic and Bavarian (from the eighth century).63 Although Germ an shows occasional forms like elimosina or elimyosa, best regarded as learned, bookish forms under the influence o f Church Latin, more commonly a form such as alamuosan occurs, whose initial a-, in place o f e-, is also found in the other Germ anic dialects. This points to radiation from Gaul, where the same initial a- is also found (OFr. almosne, OPr. almosna), in contrast to the rest o f Rom ania (It. limosina, Sp. limosna, Port. esmola).M Since Insular Celtic also shows the same loanword (e.g. OIr. almsan) with the same feature, radiation from northern Gaul seems to be assured, as with ojjron (and at the same time?). Here too a special position is taken by O E , whose term ælmesse has been derived on phonological grounds either from a Latin form *alimosina or from a Frankish *alamosina (itself a conflation, such as we also find in O H G , between alamosan and elemosina). O n the basis o f the complex sound changes to which this loanword was subjected in O E (unaccented -sin- > -sn- > -ss-, but before this o (before i) > ē, then short­ ened before ss to e) it has been concluded, with specific reference to the umlaut, that this word could not have been taken into O E much before 6oo.65 In other words, O E acquired this word from the continent, in Latin or in Frankish form, at about the same time as the other words we have considered. 63 GR 2, 88ff.

64 GR 1 ,48; F E W 3,212.

65 Pogatscher (1888), 137; Wollmann (1993), 3f.

T h e influence o f provincial Rom an Christianity

307

From this early group o f four Christian loanwords from G aul into G erm any it is clear that they all indicate only elementary externals o f Church life and practice which even the pagan outsider could not fail to notice: the church building, the bishop, the liturgical act o f sacrifice, almsgiving. None o f these terms necessarily suggests that the loan o f the word was tantamount to adoption o f Christianity, rather than first superficial contacts with it.

C H A P T E R 17

T h e influence o f G othic

T h e first beginnings o f Christianity amongst the Goths can be traced back to the third century.1 Captives made in the Black Sea area on raids to Asia M inor included some Christians (among them Wulfila’s mater­ nal grandparents) and a linguistic reflection o f this earliest Gothic Christianity is seen in Christian loanwords in Gothic which antedate Wulfila’s bible translation o f the fourth century. This translation, prob­ ably accompanied by a vernacular liturgy, provided the linguistic basis for the spread o f the new religion to other E G tribes, for which Gothic served as a lingua franca.2 T h e missionary activity o f the Visigoths in spreading their Arian Christianity was known to Jordanes (‘they invited all people o f their speech everywhere to attach themselves to this sect’) and is borne out by historical testimony.3 T h ey converted the Burgundians and Vandals by the end o f the fourth century, and the Ostrogoths, Sciri, Gepids, Rugii and Langobards during the course o f the fifth century. From the fourth to the sixth century the Goths there­ fore carried their new religion to a variety o f Germ anic tribes setded, like themselves, on the frontiers o f the Rom an Empire and within it, but we have no evidence o f their likewise being active as missionaries am ong the tribes further west (Franks, Alemanni, Bavarians). C an the words o f Jordanes be applied to Germanic-speakers in what is now G erm any and not simply to E G tribes? T h e first detailed attempt to fill this gap in historical evidence by lin­ guistic means was made by K luge,4 who postulated a Danubian point o f entry in south-eastern G erm any (Bavaria) and Gothic missionary activ­ ity as a vehicle. W ith rare exceptions this thesis is still accepted in general principle, even though some words have been subtracted from his list, but others added to it. Because o f disagreements which K lu ge’s thesis 1 Schäferdiek (1978a), 497fr. 2 Gschwantler (1976), 175fr.; Schäferdiek (1978a), 498, 500. 3 Jordanes, Geiica 133; Gschwantler (1976), 177fr.; Schäferdiek (1978b), 79fr. 4 Kluge (1909), 124fr. C f already von Raum er (1848), 401fr.

308

T h e influence o f G othic

309

has called forth we can best approach the problem he raised by dis­ cussing some o f his examples which have withstood the test o f time, then some which have since been withdrawn from his list, and finally some o f the newest suggestions for adding to it. T h e word about whose transmission from Gothic the greatest consensus exists is O H G phaffo ‘priest’, even though alternative derivations have to be excluded before we can consider the Gothic explanation.5 This word, about whose derivation from Gr. papas there can be no doubt, is attested in O H G texts only from the tenth century, but must be considerably earlier than this because o f place-name evidence dating from 739 (e.g. Pfaffenhofen) and because it shows the effects o f the Second Sound Shift. Both the textual and the place-name evidence show a common distrib­ ution, with the earliest and most frequent examples occurring in Bavarian, then in Alemannic, followed by SRhFr. and the other Franconian dialects, together with O S. Such a distribution in time and space suggests a south-eastern origin for the word’s career in O H G : it spread first in Bavaria as a sub-literary term (to judge by its absence from written texts for so long), was used at first in place-names, then in texts. O nce adopted into the ecclesiastical language it spread northwards and westwards. A direct loan from Greek is out o f the question (how was contact pos­ sible between Greek and Bavarian and how do we explain phaffo, instead o f *phaffes?). Although the Greek word found its w ay into Latin, the form pāpa is equally excluded because o f its long ā (whereas O H G pre­ supposes a short vowel) and because the Latin word meant not ‘cleric’ (as in Greek and O H G ), but ‘supreme cleric, pope’. By the same token V L cannot come into question: it had the same meaning ‘pope’, but also underwent voicing o f the intervocalic -p-, which then became a fricative (cf. in both respects O S pāvos, which must have come from Romance). Finally, a loan from OIr. popa is just as improbable: it meant not ‘cleric’, but ‘pope, monk, ascetic’, there are doubts about the linguistic traces o f the Irish mission to Germany, which in any case would have barely left enough time for the Second Sound Shift. T h e Gothic theory faces the difficulty that Wulfila regularly used gudja ‘priest’. However, papa occurs in the Gothic Calendar (ca. 400) and in a Gothic document from Naples (with the same meaning as in O H G ), so that its absence from Wulfila’s bible and occurrence soon after suggests 5 Kluge (1909), 126fr.; Waag (1932), 5fr.; Schäferdiek (1982), 242fr.

310

Contact w ith Christianity

that the Gothic word, like the O H G , was at first sub-literary. This is also confirmed for Greek, where the earliest evidence for Christian papäs is a little later than the Gothic Calendar. T h e Goths must therefore have acquired this word from the spoken vocabulary o f Greek Christians and used it first in writing a little earlier than its official recognition in Greek. Gr. papäs should have produced Go. papas (cf. Gr. Satanäs: Go. Salarias), but since in this latter equation the Gothic word passed over into the weak declension in inflected cases this allowed the back-formation o f a weak nominative Satana. Similarly with Go. papa, and hence O H G phaffo. T h e peculiarities o f the Germ an word can therefore be explained by Greek (‘priest’), but only if we postulate Gothic transmission (sub-liter­ ary function at first, Second Sound Shift, weak declension, initial focus in Bavaria). We come next to three Bavarian names for days o f the week: ertag (with variants) ‘Tuesday’, phingtag ‘Thursday’ and pherintag ‘Friday’.6 These words occur only in Bavarian and presuppose a Greek back­ ground, visible in the first part o f each compound even though an orig­ inally Greek word has been phonologically Gothicised or Bavarianised. T h e restriction o f these words to south-eastern G erm any suggests entry from the east (hence Gothic transmission). Although phingtag occurs for the first time in the twelfth century (the names for days o f the week occur very rarely in the whole O H G corpus) the double effect o f the Second Sound Shift shows that it must be o f considerable antiquity.7 Before this phing- must have been *pint~, a regular Germ anic adaptation o f Gr.pémptë(hëméra) ‘fifth day, Thursday’. Wulfila m ay not attest Go. *pintēdags, but neither did he use papa, and if phingtag had not occurred in M H G we should have had to deny it for O H G , too. T h e south-eastern centre o f gravity suggested by the sequence G reek-G othic-B avarian is confirmed by parallels from eastern Europe (SI. pçtükü, Hung, péntek ‘Friday’ because Monday, not Sunday, was regarded as the first day o f the week) and by their absence from the west. T he case for Gothic transmission is a strong one. All the evidence for pherintag is confined to Bavaria and although it occurs in glosses o f the eighth and ninth centuries the effect o f the Second Sound Shift shows, as with phingtag, that it must be older still.8 In the glosses pherintag renders Lat. parasceue, itself a loan from Greek, where the word meant ‘preparation, preparation for the sabbath, therefore 6 Kranzm ayer (1929), 70fr.; Stutz (1980), 214fr. 7 Kluge (1909), 138t; Stutz (1980), 214t.; Wiesinger (1985), 159. 8 Kluge (1909), 139t; Stutz (1980), 214t; Wiesinger (1985), 160.

T h e influence o f G othic

311

Friday’. This Greek word was also loaned into Gothic, where Wulfila used paraskatwe in the sense ‘(day of) preparation’. T h e O H G word reveals in its first element a contraction o f the original word to *pare (to which -tag has been added by analogy with other days), which could have taken place either in Gothic or in Bavarian, but in any case is paralleled by other abbreviations, especially where foreign words were assimilated (cf. for O H G sarcophagus > sore, discipulus > disco). In this case we have therefore a juxtaposition o f a learned, ecclesiastical term (paraskatwe) with a word from popular speech (*parē), such as we found with papa and will meet again with sabbato. To this derivation from Greek via Gothic a rival suggestion has been opposed, seeing in pherintag a loan from Lat. perindie ‘second day before a fixed date’, therefore ‘Friday’ (by reference to Sunday).9 Acceptance o f this learned monastic coinage from Latin into Bavarian would make the sequence G reek-G othic-Bavarian irrelevant, but doubts are in order. It is significant that the O H G examples gloss Lat. parasceue, not perindie, so that some awareness o f their association was still felt. T h e Latin theory also suffers the disadvantage o f not being able to account for the restric­ tion o f this loan to Bavarian. Thirdly, the relevance o f the Second Sound Shift means that the Bavarian word must have been coined at the latest early in the seventh century, at a time when there were no monasteries in Bavaria where such a learned adoption was conceivable. We may con­ clude that although the group phinztag, pherintag and (as we shall see) sambaztag, all suggesting Gothic transmission from Greek to Bavarian, may not o f itself be compelling, it is highly suggestive. With ertag we come again to a Bavarian word which is attested only much later in M H G .10Although the Sound Shift provides no help for an early dating, this word agrees with the others in permitting a regular phonological derivation from a Greek word for ‘Tuesday’. Gr. Areos hëméra, corresponding to Lat. Mortis dies and likewise including the name o f the pagan god o f war, is attested in Greek sources o f the third and fourth centuries.11 Even though there is no reflection o f it in Wulfila’s Gothic the intermediate position o f Gothic is suggested by the regular phonological development which can be reliably reconstructed: from Areös hëméra to Go. *arjaus dags, then to O H G *arwtag (cf. G o. flip u s/-aus: O H G fridu/-$), and finally to M H G ergetag (cf. O H G fario: M H G ferge), one o f the variant forms o f this word. We need not be disturbed by the 9 Knobloch (i960), 430fr. 11 Thumb (1901), 171.

10 Kluge (1909), 140fr.; Stutz (1980), 2i4f.; Wiesinger (1985), I58f.

312

Contact w ith Christianity

pre-Christian origin o f this Greek name (incorporating the name o f a planetary deity o f pagan antiquity) as opposed to the Christian nature o f other names for days o f the week passed on from Greek to Gothic, for the double origin o f this nomenclature is reflected elsewhere in Germ ania (cf. mittawecha and Woensdag). Even if this Bavarian term differs from the others in deriving from a pre-Christian Greek form this does not mean that it had a pagan overtone for the Goths (let alone the Bavarians): for them it was simply a phonetic loan, otherwise m eaning­ less in the new linguistic context, and it cannot be taken as an indication that they knew o f Ares as the god o f war or used his name in place o f a Germ anic deity.12 N or can the suggestion that the Goths accepted this word because it reminded them o f Arius, the supposed founder o f their Christian sect, be seriously maintained any longer.13 We are left there­ fore with a Greek and a G erm an term whose connection with each other can be explained by Gothic mediation. This is no longer entirely the case with another name for a day o f the week: O H G sambaztag ‘Saturday’.14 T he word is attested from the ninth century (Otfrid, the Tatian translation, Notker), but this wider geograph­ ical spread, making it a word characteristic o f southern G erm any at large, is something new in the days we have considered and prompts the question whether this now tells against a possible Gothic influence. T h e Christian origin o f this word (‘sabbath’) is clear, but why should it have come to Germ any from Greek rather than Latin, which also took over the word from Greek? T hat the source o f the O H G word with its characteristic -m- is to be sought in the east is clear from a distinction running through Romania between western forms without -m- (It. sabbato, Provençal sabde, Sp., Port, sabado) and eastern forms with -m- (e.g. Rum. sâmbâta), supported by other eastern, but non-Romance languages (e.g. O S 1. sqbota, Hung, szombat). Although ecclesiastical Greek as well as Wulfila’s Gothic know only forms without -m-, these east European forms allow us to reconstruct a V G variant *sámbaton and hence *sambato for popular speech in Gothic, too. With that reconstruction we have a distinction between ecclesiastical and popular speech, such as we saw with papa and *parē, but also the means o f explaining O H G sambaztag, the end-result o f a line o f influence from Greek via Gothic. T hat is only part o f the picture, however, for there is another explana­ tion for Germ an forms with -m- in southern G erm any at large and in 12 Kluge (1909), 144; Wiesinger (1985), 159. 13 Schäferdiek (1982), 241, fn. 11. 14 Kluge (1909), 137t; Meyer (1894), 326fr.; Avedisian (1963), 231fr.; Stutz (1980), 215F.

T h e influence o f G othic

313

the west, namely the possibility that these forms entered from the west, as is suggested by French forms with -m- (.samedi in the whole o f north­ ern France). As an explanation o f these French forms various sugges­ tions have been made: Greek Christian communities along the Rhone valley and in Lyon (as with kyriakóri), Visigoths in south-western France at a later stage o f their migrations, or V L speakers (under the influence o f V G ).15 These two centres o f gravity, suggesting two points o f entry for sambaztag into Germany, one in the east and one in the west, would account for its wider spread than with the other terms.16 O r should we place greater weight on the absence o f Bavarian examples from O H G and reckon with influence from the west alone, as with kirihha? In that case this word would belong to the next stage in our argument (subtrac­ tions from K luge’s list). Those who question Gothic influence on O H G accept it in some cases (ertag, phinztag), but rescue their thesis by seeing these names for days o f the week in the context o f trading contacts, which devalues them as examples o f Gothic religious vocabulary.17 This argument operates with a distinction between secular and religious which is much too dis­ junctive, and it conflicts with a striking feature o f classical loanwords in Gothic. Whereas those in the sphere o f trade and warfare derived from Latin,18 those for Christianity came overwhelmingly from Greek (even if sometimes via Latin), so that the Greek source for these names suggests a Christian context for their adoption in Gothic. N or does the attempt to reinforce this trading argument by reference to *mûta ‘customs’ (regarded as a loanword from Gothic to Bavarian in the context o f trade) carry much weight: given the absence o f any effect o f the Second Sound Shift, this loan must belong to a later period.19 If we base ourselves on the examples so far considered the following criteria suggest themselves for Gothic influence on O H G .20 First, a peculiarity about the Germ an word which suggests an origin in south­ eastern Europe and therewith Gothic transmission (for the most part a word o f Greek origin with no counterpart in the Latin o f the western Church). Secondly, a parallel, semantic or phonological, between the G erm an word and its Gothic counterpart. Thirdly, a suggestion that the Germ an word had its centre o f gravity in the south-east (Bavarian) or in the south (Bavarian and Alemannic), even though this need not exclude 15 16 18 20

Meyer-Lübke (1923), 9; von Wartburg (1953), 296fr.; GR 2, 4446; Avedisian (1963), 2326 Stutz (1980), 216; Wiesinger (1985), 190. 17 Knobloch (i960), 430. See above, pp. 202ff. 19 Schäferdiek (1982), 242. Against this Wiesinger (1985), 193fr. Cf. also Stutz (1980), 219.

314

Contact w ith Christianity

its later spread elsewhere by internal G erm an loanword traffic (as with phaffo). Fourthly, a dating in the sixth century or earlier (here the rele­ vance o f the Second Sound Shift is decisive). N ot all these criteria are present in every case, so that room for doubt remains, and some criteria are not always applicable (a Greek word could enter G erm an sometimes from the west, so that Gothic transmission was not necessarily involved). So far, all the examples discussed from K lu ge’s list have come ulti­ mately from Greek, which has been o f assistance in recognising them as loanwords and in establishing Gothic transmission. However, the influ­ ence claimed for Gothic is not confined to Greek terms unknown to Latin Christianity, although the borrowing o f native Gothic words into O H G is more difficult to establish than with obviously non-Germ anic ones. For this purpose we need further criteria. First, a parallel between the Germ an word and the Gothic one (either the same Germ anic word or different words, but with the same function). Secondly, the word must be used as a Christian, not just secular term in G erm an and in Gothic. Thirdly, we need to ask whether the O H G word, o f south-eastern origin, was in opposition to other Christian terms in Germany, deriving from elsewhere. These additional criteria increase our philological difficulties, so that there is more room for doubt. These uncertainties explain why subtractions from K luge’s list have been made as well as additions to it. With regard to a number o f O H G words to which K luge attributed a Gothic origin the present consensus places them in a different context. Kirihha and biscof, far from entering G erm any in the east, probably derived from the provincial Roman vocabulary o f Christian Gaul. We shall later argue that pharra ‘parish’ was probably an import o f the Frankish mission, just as heilag and ostarun acquired their Christian func­ tion under the influence o f the Anglo-Saxon mission. These terms there­ fore now have to be dropped from K luge’s Gothic list. A part from them we may now turn to an example for which K luge claimed a G reek model. With this example (Go. dáupjan, O H G toufen ‘to baptise’)21 K luge pro­ ceeded from the assumption that it was unlikely that this Christian concept was conveyed by the same word twice in G erm anic inde­ pendently, especially in view o f the range o f alternatives found in Germania, and concluded that the Gothic choice had influenced 21 Kluge (1909), 131Í; Knobloch (i960), 435t; Loi (1969), 67IT.; Stutz (1980), 212t; Schäferdiek (1982), 244t

T h e influence o f G othic

315

Germ an. Go. dáupjan resembles Gr. baptizein in several respects. T h e Greek word is a causative verb, derived from the adjective baptós ‘dipped, dyed’ and therefore meaning ‘to dip, wash’, just as the Gothic is a caus­ ative verb, derived from the adjective diups ‘deep’ and hence meaning ‘to dip, immerse’. Just as Greek retained this general meaning o f its verb after its Christian specialisation o f meaning, so did Wulfila use his Gothic verb in the sense ‘to dip, wash’ alongside the meaning ‘to baptise’. From this agreement it was concluded that the Gothic word had been christianised in imitation o f Greek and in full understanding o f the linguistic composition o f the Greek verb, hardly surprising in view o f the exposure o f the Goths, and o f Wulfila especially, to the Greek lan­ guage. It was also concluded that such knowledge o f Greek was impos­ sible in early Germ any (where Christian Latin baptizare would not suggest toufen, as its Greek model had for the Goths), so that the Christian use o f toufen derived from Gothic. Although we are dealing with a ver­ nacular term, the linguistic route followed (G reek-G othic-O H G ) was the same as in the previous cases. This argument rests, however, on the assumption that knowledge o f Greek was not possible in Germ any at this time, and about this, on however elementary a level, doubts may be entertained.22 W hat has been said o f Anglo-Saxon England in this connection is equally applic­ able to Germany: understanding o f a considerable number o f Greek words was available from a variety o f sources, from G reek-Latin glos­ saries, Isidore’s Etymologiae and patristic writings. For the O H G Benediktinerregel an acquaintance with Greek terms which bypasses their Latin equivalents has been suggested (it renders kirieleison directly by truhtin kinade uns) and knowledge o f central terms o f Greek Christianity was occasionally passed on in glosses (e.g. soter grece salvator latine heilant). Even apart from this it is still possible to account for O H G toufen by a Latin model, thereby avoiding any need to involve the Goths. T h e Latin verb (in)tingere originally meant ‘to dip’ (and ‘to dye by dipping’), but was then used in application to Christian baptism, especially by Tertullian about 200. I f it seems a far cry from Tertullian’s Africa to Germ any we may recall that the verb was also used in this sense by such writers as Augustine, G regory the Great, Isidore and Bede. W here tingere was used alongside baptizare in Christian Latin the former denoted the ritual act o f immersion, the latter the sacramental act o f baptism.23 O f particular 22 Gneuss (1992), 119; Ibach (i960), 403fr.; Benediktinerregel (54), p. 37; Steinmeyer and Sievers (i879ff ),111 182,12. 23 Loi (1969), 73, 8if.

3 i6

Contact w ith Christianity

interest is the evidence from Isidore [Baptisma graece, latine tinctio interpretatur), for his use in clerical schooling ensured that a Latin word meaning ‘to dip’ could be imitated in a Germ anic vernacular to denote baptism without any dependence on Greek. We can go a step further, for even without any knowledge o f the Greek or Latin term the practice o f baptism by immersion, known from the bible with Christ’s baptism in the Jordan and in use by the western Church, could lead to the adoption o f a Germ anic verb meaning liter­ ally ‘to dip’ as a term to describe a ritual act visible to pagans and new converts alike.24 O nly if baptism by immersion had been unknown in the west would we be driven to assume that toufen was borrowed from Gothic. Instead, if we regard the derivation o f Go. dáupjan from Greek as unconnected with the dependence o f O H G toufen on Latin (or on visual observation o f baptismal practice) we face once more the proba­ bility o f polygenesis. Even though O S dôpian m ay conceivably be a loan­ word from O H G (as part o f the Frankish conversion o f the Saxons), the argument for polygenesis must still be used for O E dēpan/dyppan. These verbs occur in England rarely as Christian terms in tenth-century glosses which cannot be adduced as witnesses to continental W G (or even Gothic) usage centuries earlier.25 T h ey probably owed their chance formation as much as their O H G parallels to knowledge such as that conveyed by Isidore or to contemporary Church practice. T h e case for a Gothic background to O H G toufen must therefore be regarded now as extremely weak.26 We turn now to additions which have been made to K luge’s list. T h ey are examples o f vernacular origin, the first o f which suggests the possibility o f Greek influence via Gothic in yet another way. T h e semantic field o f O H G (ir)fullen ‘to fill, fulfil’27 covers three usages: physical (a space is filled), metaphorical (a person is filled with jo y or time is fulfilled) and abstract (a prophecy or commandment is ful­ filled). O f these the second and third represent biblical usage so that the question arises, if they are Christian extensions o f the first usage, where this adaptation m ay have taken place. T h e linguistic reflection o f this question is seen in Otfrid (or even more so in Notker), who prefers (gijfullen for the physical usage, but irfullen for the biblical, thus making a 24 Ibid., p. 77; Schäferdiek (1982), 244. 25 Gneuss (1992), 122. 26 Another suggestion by Kluge to be included here is Go. armahairts, O H G armherz, O E earmheort ‘merciful’ . O n this see Kluge (1909), I48f.; Wahmann (1937), 134fr.; Betz (1949), 73; Beck (1979), 109fr.; Gneuss (1992), 122i. 27 H inderling (1971), iff.

T h e influence o f G othic

317

distinction by means o f the prefix ir-. T h e variant irfullen is attested early (eighth century), especially in U G . By contrast, further north Tatian uses (gijfullen and the Heliand (gijfullian in both the physical and biblical sense, with no differentiation. In this northern Germ any agrees with O E (ge)jyllan. In U G the examples o f irfullen come chiefly from religious texts and the verb is used predominantly in the biblical context o f prophecies, commandments or time being fulfilled, so that it is a standing theolog­ ical term. This O H G differentiation between (gijfullen and irfullen could not have been prompted by Latin, which used a range o f variants (replere, implere, adimplere, complete) without any semantic distinction or correspondence to the G erm an pair. We do better to turn to Gothic for a parallel with O H G , since Wulfila used fiilljan/Jullnan predominantly in the physical sense, but usfulljan/usjullnan (whose prefix corresponds to O H G ir-) some­ times in the physical or metaphorical sense, but much more frequently in the abstract (of a prophecy or commandment).282 9Although Wulfila based his Gothic translation on the Latin as well as the Greek model, Latin can have given him no lead in this. By contrast, the Greek bible did differentiate between the physical meaning (for which it used a variety o f verbs: pimplánai, gémein, gemizein) and the metaphorical or abstract (plēroūn). Wulfila’s text may lack the variety o f physical verbs, but more important is the point that, following the Greek model, he differentiated at all. T h e fact that U G usage with its verbal distinction had no parallel further north (Tatian, O S, OE), but agreed with Gothic practice (behind which lies Greek biblical usage) suggests, as was more obviously the case with phajfo, Gothic influence. N o such Greek background can be detected for the equation o f O H G abgot with Go. afgupsP T h e formation o f the Germ an word for ‘idol’ (with its derogatory prefix expressing scorn for the pagan gods) must be Christian, since for Germ anic paganism there was no antithesis between gods and false gods. This relatively late birth o f abgot is confirmed by its distribution within Germ anic, confined to continental W G (OH G , O S, OFri.), for whenever a similar distinction had to be made in O E and O N the pagan gods were designated in another way.30 Gothic, too, went its own w ay as regards the substantive by using galiugagup (‘false god’), but Wulfila also employed an adjective afgups to render asebēs ‘impious’. There are, however, difficulties in using the similarity between Gothic 28 In each o f these Gothic pairs the second verb is an inchoative, doing duty for the passive. 29 Karg-Gasterstädt (1945), 42off. 30 Cahen (1921), 73 and 49fr.

3i8

Contact w ith Christianity

and continental W G to suggest a common origin in Germanic: one form is an adjective and the others are substantives, one form describes man’s relationship to G od while the others denote the pagan god himself, and it is difficult to imagine an early, Germ anic origin for a specifically Christian concept. A useful pointer is provided by word-formation, where the Gothic adjective constitutes no problem, whilst abgot stands alone am ong noun compounds with ab- and is clearly anomalous. This prefix occurs in verbal abstracts (e.g. ablâg: ablâzan), but also in forming compound adjec­ tives (e.g. abtrunni, also Go. qfgups) or adjectival abstracts (e.g. abtrünnig). This shows up O H G abgot as exceptional, whereas Go. qfgups belongs to an established category, so that we may suspect the possibility o f a loan accounting for the Germ an anomaly. T he Gothic adjective occurs twice: once to denote the impious, those who do not acknowledge G od by their behaviour, and once with reference to heretics, meaning therefore not ‘godless’, but having a false conception o f God. If we can apply a context like this to O H G , this would suggest that abgot belonged to the missionary period when Christianity was struggling with the old relig­ ion. T he transfer o f the word from one language to another would explain how the Gothic adjective came to be an O H G substantive (probably also under the influence o f got), in other words how the anomaly in word-formation came about. T h e flectional variants in Bavarian and Alemannic, in particular, suggest that the loan from Gothic to O H G is likely to have taken place in southern Germany.31 A Bavarian centre o f gravity and the possible influence o f Gothic are also found with O H G stuatago ‘judgm ent day’ .32 T h e first element o f this compound belongs etymologically to an U G group o f verbs with several variants: stouwon (Bavarian, ‘to complain, criticise, accuse’), stouwôn (Bavarian, Alemannic, SRhFr., ‘to complain, criticise’), stuowen (Bavarian, ‘to criticise’), stuowôn (Bavarian, Alemannic, ‘to quarrel’), stuoen (Bavarian, Alemannic, ‘to make amends, suffer punishment’). T h e last o f these variants, emphatically Bavarian, was dying out already at the beginning o f the written tradition and is connected with forms more actively represented in Gothic: staua fern, ‘trial’, staua masc. ‘ju d ge’, stojan ‘to pass judgm ent’, stauastols ‘tribunal’. Both in O H G and in Gothic these are legal terms, but view the process from different angles: the accused ‘suffers punishment’ (OHG), but the judge ‘passes judgm ent’ 31 Karg-Gasterstädt (1945), 433. 32 Freudenthal (1949), 89; M üller (1957), 3o8ff.; Mastrelli (1975), 75ff.

T h e influence o f G othic

319

(Gothic). In view o f this contrast there are no grounds for assuming Gothic influence on this Germ an verb. T h e position is different, however, with stuatago, attested once only, in Bavarian (Muspilli v. 55). From O H G usage o f the verb this should mean ‘day o f punishment’, but in treating the ‘day o f judgm ent’ Muspilli presupposes the Gothic meaning. T h e relationship between the two forms suggests a loan from Gothic to Bavarian, for whereas the Gothic words are semantically coherent, stuatago is anomalous in two respects: it occurs only in Bavarian and in a meaning isolated within G erm an.33 O n e final example is Gothic in origin, but in a different way. O H G kûski ‘virtuous, pure’ as a specifically Christian term shows a wide range o f word-formations: the adjective kûski, the substantives kûski and kuskida (all three in positive and negative forms) and the negative verb biunkûsken.34 T h e range o f this word within W G points to an origin within O H G : in contrast to the frequency and range o f formations in O H G the word occurs only twice in O S (once in the Heliand, once in the Genesis), which suggests a loan from the south, whilst the one attestation in O E (in the Genesis, translated from OS) points to onward transmission from O S to O E . If the W G evidence implies radiation from O H G the range o f evidence within O H G is similarly informative: the word’s centre o f gravity from the eighth to tenth century was markedly U G , especially Alemannic or more specifically the monastery o f Reichenau. By contrast, it was relatively unproductive in Bavarian and came to Franconian only later. Such a specific point o f radiation (Reichenau) within O H G and then within W G suggests a specific cause and argues against the word being a common Germ anic inheritance. It is in fact a loanword from Latin conscius ‘conscious, aware’, capable o f indicating ethical awareness (especially o f guilt) in pre-Christian Latin and hence eminently adaptable to Christian ends, as in conscientia. A part from OPr. and O C cusc (with a later spread to OFr.) Lat. conscius has left no traces in Romance. Even these isolated forms cannot have come directly from Latin, for whilst 0 + n before s could have produced 0 (cf. It. Costantino) this cannot account for u, and there is the further irregularity that c before i has not been palatalised. These irregularities can be accounted for, however, if we assume that the Latin word entered 33 O ther suggestions under this heading include Go. weihs, O H G wîh ‘holy’ (Baetke (1942), 85fr., 165fr.; Stutz (1980), 210, fn. 15), Go. ansts, O H G ansi ’grace’ (Wahmann (1937), I2iff.; Green (1965), 239fr.; Stutz (1980), 210) and Go. dulþs ‘festival’, Bavarian tuld ‘fair’ (Wesche (1937), 94fr.; Bohnenberger (1948), 468fr.). 34 Sperber (1915), 55fr.; F E W 16, 430; Frings and M üller (1951), 109fr

320

Contact ivith Christianity

OPr. and O C not directly, but via Gothic (even though the word is not found in this language). Already as attested by Wulfila, Latin 5 could become ū in Gothic (e.g. Rumoneis) and c before i could be represented by k (e.g.fascia > fasķja). W hat is irregular as a direct development from Latin to Romance can therefore be accounted for by Gothic transmission, a fact rendered more significant by the way in which Catalan and Provençal together match the area o f the Visigothic kingdom. A n iso­ lated form within Romania and a specific point o f radiation within Germ ania suggest a definite point o f contact, transmission o f the Gothic form (or its Romance derivative) from the south-west, and learned, monastic rather than missionary contact. O H G kûski exemplifies Gothic linguistic influence, but not a Gothic mission.35 Kluge avoided the question o f the historical context for Gothic loan­ words in O H G , leaving it to historians to answer,36 but he still talked o f Gothic words coming up the Danube to Germany, in which he has been followed by others who extend the range o f this traffic to England and Scandinavia. Dissatisfaction with this and the lack o f evidence for a Gothic mission to Germ any have led at times to a rejection o f Gothic influence on the Christian vocabulary o f G erm any at all.37 Such dis­ satisfaction on the part o f historians is likely to be increased by three suggestions made by philologists as to how Gothic influence entered Germ any (from the east, the south and the west), for this could be symp­ tomatic o f philological uncertainty. O n the other hand, this diversity could be merely the result o f the changing locations o f the Goths on their migrations (Black Sea, Moesia, Pannonia, northern Italy, southern France, northern Spain). We must now consider these three possibilities, bearing in mind that different geographical stages in the Goths’ m igra­ tions imply different datings o f a possible linguistic influence,38 that Gothic words could be transmitted indirectly, not necessarily by the Goths themselves, and not as part o f any missionary activity. Entry o f Gothic words from the east (the Danubian theory) would imply Balkan radiation o f the Arian Christianity o f the Gothi minores while still settled north o f the Danube in Gothia or on imperial territory south o f the river and in the period between Wulfila and the Avars’ inva­ sion o f Pannonia, i.e. in the fifth century.39 To explain the adoption o f 35 Betz (1949), 49fr, has pointed to Go. (ga)timreins and O H G (Reichenau) limbar ‘religious edifica­ tion’ in a similar monastic context. 36 Kluge (1909), 160. 37 E.g. Schäferdiek (1982), 239fr. 38 Stutz (1980), 221; Wiesinger (1985), 183i. 39 Stutz (1980), 220; Schäferdiek (1982), 248i.

T h e influence o f G othic

321

these words in south-eastern G erm any we have to take into account the ethnogenesis o f the Bavarians in the first half o f the sixth century in Raetia Secunda (now Bavaria) and western Noricum Ripense (Upper Austria, south o f the Danube).40 This newly formed tribe incorporated various E G tribal groups (Sciri, perhaps the Rugii, Heruli) and could have entered its new territory from Bohemia, Pannonia or RaetiaNoricum. T h e Ostrogoths had been Arians while still in Pannonia, the Rugii had an Arian royal house while in Lower Austria (and the Sciri, too?), and the Langobards were Arians during their period in Lower Austria and Pannonia. In short, in the course o f their ethnogenesis the Bavarians were surrounded by Arian tribes to the south and east, they included within their own formation elements which had broken loose from these tribes, already nominally converted to Christianity. Knowledge o f Christianity brought to the Bavarians in this way in the course o f their tribal formation was bound to be confined to scattered groups and to be external and rudimentary, hence the later need for a more thoroughgoing Christianisation o f the Bavarians in the seventh and eighth centuries.41 This route from the east was probable when behind the parallel between Gothic and O H G there stood a Greek model not adopted by Latin and possibly with reflections in eastern Europe (Rumanian, Slavonic, Hungarian). It was also probable when a Gothic term was attested only in Bavarian or at the most in U G .42 T h e rudimentary acquaintance with Christianity which this route provided is confirmed by the nature o f Gothic loanwords initially or entirely confined to Bavarian (phaffo, names for certain days o f the week), for these suggest no more than an external contact with Christianity, not necessarily conversion, let alone a Christian mission to the Bavarians. This amounts to a modified version o f the Danubian theory as put forward by Kluge. Gothic linguistic influence still remains, but transmitted indirectly by scattered groups from E G tribes converted by the Goths, and then involved in the ethnogenesis o f the Bavarians. A ny missionary activity was therefore conducted earlier among these E G tribes, not with the newly formed Bavarians. Although we are dealing directly neither with Goths nor with a mission, this does not preclude Christian terms o f Gothic origin in Bavarian. T h e transmission o f Gothic terms from the south (the Alpine theory) 40 Wiesinger (1985), 185fr.; Schäferdiek (1982), 249. 42 Stutz (1980), 219.

41 Wiesinger (1985), 189, 190.

322

Contact w ith Christianity

embraces two possibilities, with slightly different dating: they came either from the Ostrogoths in northern Italy (in the period 489-553) or from the Langobards, their successors in that region (in the late sixth or early seventh century).43 T h e obvious route for this traffic would have been over the Brenner or from Aquileia to Juvavum, while the trans­ mitters involved were either the Ostrogoths under Theoderic or the Langobards, converted to Arianism by the Goths before becom ing Catholic. T he indications o f heretical Christianity and syncretism to be found in Bavaria from the second half o f the sixth century could tie up with Arian influence from the Goths and still from the Langobards.44 For each o f these possibilities a wider framework has been suggested. For the Ostrogoths it is above all Theoderic’s political alliance with the Thuringians against the combined threat from the Franks and Byzantium, leading the Gothic ruler to extend his authority to southern G erm any as far as the Danube, the frontier with the Thuringians.45 For more than a generation southern G erm any was therefore under Ostrogothic influence. It has been questioned whether this was long enough for Gothic religious influence to make itself felt in G erm any in the face o f Frankish penetration, but precisely this policy o f Theoderic has been seen as the context for the transmission o f Gothic loanwords to G erm an.46 Admittedly, there is no evidence that Theoderic’s policy o f alliances had a religious dimension, but given the political-religious nature o f his relationship with Chlodw ig this is by no means unlikely. Alternatively, for the Langobards a strong cultural influence has been argued across the Alps on Bavaria before the Franks, so that religious ter­ minology, ultimately o f Gothic origin, could have been carried by them.47 It has not proved easy to suggest specific words which m ay be attrib­ uted to this route, although tuld is a possible candidate. Its presence in Bavarian, but also in Alemannic suggests two possible entries: either from fifth-century Pannonia to the ‘pre-Bavarians’ and then to the Alem anni or from sixth-century Italy across the Alps to Bavaria or independendy to the Alem anni (a later entry still reconcilable with the operation o f the Second Sound Shift).48 Either explanation would involve only indirect Gothic influence: other E G tribes by the eastern route, Langobards by the southern.49 43 46 47 49

Ibid., pp. 22of.; Wiesinger (1985), 184. 44 Wiesinger (1985), 192. 45 Beyerle (1955), 65fr. Schäferdiek (1982), 250; Beyerle (1955), 81. 48 Bohnenberger (1948), 476. Baesecke (1930), 148fr., 158; Kranzm ayer (i960), 30fr. Cf. Eggers (1954), 136, fn. 4.

T h e influence o f G othic

323

Penetration o f Gothic terms from the west (the Aquitanian theory) is connected with the Visigoths and their kingdom o f Toulouse or in northern Spain.50 In the former case the period in question would be 418-507, although this might have to be extended to the seventh century if we postulate Gothic influence on Romance (as in OPr. cusc) and then transmission to G erm any by the strong Aquitanian component in the Frankish mission east o f the Rhine. In the latter case the period would be from 484 to the end o f the sixth century, when the Visigoths in Spain were eventually Romanised linguistically. T h e argument for a western entry for Gothic terms has lately been conducted as part o f a wider thesis suggesting a Frankish mission east­ wards, mainly to the Alem anni and Bavarians, in the sixth century.51 Knobloch assumes that Gothic elements entered the Frankish ecclesias­ tical vocabulary as a result o f the Frankish conquest o f the Visigothic kingdom (507-11), but regards 596, the date o f the mission to the AngloSaxons, as a terminus ante quem because he believes that some o f the Gothic terms were carried to England by this mission. His examples are not argued in detail and some remain very uncertain (e.g. Welt, glauben). Moreover, by restricting this influence to the sixth century in order to accommodate what are regarded as Gothic terms in O E this thesis runs into a dangerously tight chronology.52 T h e Spanish Visigoths were con­ verted to Catholicism in 589 and the mission to England began in 596, so that in the space o f merely seven years Christian terms o f a hitherto heretical confession had to be adopted officially by the Frankish Church and taken by Frankish interpreters to England. Frankish Christian terms may have reached O E in this way, but we have no evidence that Gothic ones were amongst them. I f we abandon this hypothesis o f Gothic words in O E this has the result o f providing a longer period for the Catholicised Goths to be linguistically active in the Frankish Church, a period extending now to the seventh century, when the more active Frankish mission to G erm any could have taken words with it which were ultimately o f Gothic origin. However uncertain the historical context suggested for the Aquitan­ ian theory m ay be, we are left with the western associations o f O H G kûski and with Gothic connections with Reichenau, as in the case o f zimbar. T h e latter example suggests an individual monastic contact, pos­ sibly in the context o f Aquitanian recruits to the Frankish ecclesiastical 50 Stutz (1980), 221, fn. 64; Wiesinger (1985), 184: Schäferdiek (1982), 256; Thom pson (1969), 214. 51 Knobloch (i960), 427fr.; (1966), 22if.; (1967), 300fr. Cf. also Stutz (1980), 21 iff. 52 Schäferdiek (1982), 255fr

324

Contact w ith Christianity

penetration o f the area east o f the Rhine. In this connection Pirmin, the founder o f the monastery at Reichenau, has been suggested as the point o f contact, together with monks in Reichenau confraternity lists whose names are Gothic. This would be more persuasive if we were certain that Pirmin was in fact a Goth and that the monks’ names were G othic.53 Again, we remain with the fact o f Gothic linguistic influence on O H G , but with no trace o f a Gothic mission. It is this mission which, as a historian, Schäferdiek has rightly rejected on the grounds that there is no historical evidence to support it. From this he proceeds, equally rightly, to reject the suggestion that there were linguistic effects o f such a mission in southern Germany. From this point, however, he argues incorrectly that there are no loanwords o f Gothic origin in O H G . This last step ignores two points. It pays no attention to the possibility that these loanwords could be transmitted indirectly, not by Goths themselves. It also fails to consider that the loanwords need not have been spread in the context o f missionary activity. T h e case for Gothic loanwords in the Christian vocabulary o f Germ an does not stand or fall with that for a Gothic mission. 53 Betz (1949), 5if., ioif.; Schäferdiek (1982), 256, fn. 94.

CHAPTER

l8

T h e influence o f the M erovingian Franks

U nder this heading we have to consider two topics: what has been regarded as an Irish mission in Francia and in southern Germ any (where effects on O H G vocabulary have been proposed) and a Frankish mission east o f the Rhine. T h e former is to be questioned and stress will be placed instead on the latter. These topics have to be seen against the background o f the loss o f Christian territory resulting from the barbarian invasions.1 In the Reconquista the conversion o f Chlodwig was a vital first step, but no more, since between his death (511) and the arrival o f Columbanus (590) Christianity made litde headway, apart from at the Merovingian court itself.2Although there was a first hesitant phase o f Christianisation in the sixth century (e.g. the episcopal lists start again in the Rhineland, first with Gallo-Rom an, then with G erm anic names), this was confined to the Franks, built on still surviving Christian pockets, and aimed at no more than the status quo ante. O nly in the seventh century did this change with a thoroughgoing re-Christianisation o f the former frontier area, but also with a missionary advance into areas east o f the Rhine. Behind this advance lay the alliance between the monastic movement o f the Irishman Columbanus (mainly centred on his foundation at Luxeuil) and Merovingian kings and nobles expanding Frankish power amongst the Alem anni and Bavarians.3 Early in the seventh century attempts were made to extend the authority o f the Church to north-eastern Gaul by figures such as Amandus and Eligius, with connections with Frankish royalty and with Luxeuil, aiming at areas still largely pagan. Luxeuil also stood behind another missionary impetus: to the Alemannic area, opened up to Christianisation initially by Columbanus himself, sent there by 1 Büttner (1951), 9; Schäferdiek and Schäferdiek (1976), i8of.; Ewig and Schäferdiek (1978), ii6ff. 2 Ewig (1970), 3of.; Schäferdiek (1976), i8 if; (1978a), 537. 3 Schäferdiek (1978a), 541; Hillgarth (1987), 323. 325

326

Contact w ith Christianity

Theudebert II after his expulsion from Luxeuil, then by others with similar affiliations. T he influence o f the Vosges monastery reached as far as Bavaria, where Eustasius, successor o f Columbanus at his foundation, and other monks from Luxeuil made up a systematic mission over some period o f time for which a connection with D agobert’s authority over the eastern territories seems likely A n eastward extension o f Christianity in conjunction with Frankish political expansion can also be assumed for the sector east o f the Middle Rhine in this century.4 T h e turn from the sixth to the seventh century thus saw the beginning o f missionary activity on a broad front carried into new territory. T h e impetus lay with Frankish rulers, but also with the Irish monastic ideal o f Columbanus, who sought assistance and protection from rulers in founding monasteries in Francia and whose form o f Christianity helped to legitimise their expansive authority. By the end o f the seventh century the Anglo-Saxon mission largely took over, so that any Irish linguistic influence will have to be sought in this century. Its location will have to be attempted in areas where we have evidence for Irish activity in Germany, i.e. mainly amongst the Alemanni, but also the eastern Franks and Bavarians. However, Irish traces are not the same thing as Irishmen, let alone their influence on the Germ an language. In historical research a reaction has set in against an earlier equation o f seventh-century Christianisation simply with an ‘Irish mission’.5 Instead, the Irish are now seen as providing the impetus to a much wider movement, both Gallo-Roman (e.g. the combination o f the rules o f Columbanus and Benedict) and also Frankish (e.g. the Frank Waldebert, the second suc­ cessor o f Columbanus as abbot at Luxeuil). From the Irish impetus there arose a joint movement which Prinz has defined as ‘Hiberno-Frankish’ . A t least three historical factors justify this greater stress on the Frankish contribution. In the first place, the Irish impetus would have achieved little without Frankish leadership and support.6 O n the frontiers o f Christendom a mission could only thrive under the protection o f a local king or noble­ man, so that Columbanus, on arriving in Francia, turned to the royal court, thereby linking his monastic and missionary efforts with the power o f the Merovingian state. A second qualification o f Irish influence is called for by Prinz’s distinction between Irishmen active in G erm any 4 Büttner (1951), 13fr.; Schäferdiek (1976), i8iff.; Prinz (1978), 133, 458; (1981), 81. 5 Büttner (1951), i2f.; Prinz (1982), 202. 6 Ewig (1970), 48f.; Ewig and Schäferdiek (1978), 126; Prinz (1978), 447^ 452h; (1981), 77h, 80; (1982), 2o8ff.; HiUgarth (1987), 314.

T h e influence o f the M erovingian Franks

327

(Columbanus himself or Kilian, Colm án and Totnán, martyred in W ürzburg in 689) and missionaries under Irish influence, but o f Frankish, Franco-Burgundian or Gallo-Rom an origin.7 This distinction makes it clear, for example, that no Irish foundation can be placed in Bavaria, by contrast to Francia, and that the number o f known Irish mis­ sionaries in G erm any is extremely small. Thirdly, the indirect nature o f Irish influence can be shown in another way.8A part from the short spell o f Columbanus in the Lake Constance area and a few o f his compatri­ ots known to have been in Germany, the influence o f Luxeuil was only indirectly Irish: continental monks won for Irish monasticism, but as missionaries incapable o f transmitting Irish terms to the nascent Christian vocabulary o f Germ an. Taken together, these three considera­ tions qualify the extent to which we can talk o f a purely Irish mission in G erm any and justify the alternative term ‘Hiberno-Frankish’. O n ly one Irish loanword into Germ an has been suggested: O H G glocka ‘bell’, but even here there are difficulties.9 T h e use o f a bell in a church service is attested from the fourth century with Lat. signum for northern Gaul and from the sixth with campana for southern Gaul and Italy. However, in the seventh century clocca was used in place o f signum in northern France (cf. Fr. cloche) and spread over part o f the campana area in northern Italy. This new word also gained entry into W G and ulti­ mately N G. In G erm any two variants occur: clocka in the Low and Middle Germ an dialects, glocka in the south. For these variants two differ­ ent entries into G erm any have been suggested, behind each o f which an Irish word has been thought to stand: clocka thus reached northern G erm any via the lower Rhine and northern France (where the Irish are said to have brought the word), whereas glocka is held to be an import to southern Germ any from Italy (the Irish foundation Bobbio has been sug­ gested as a point o f origin). For the Irish origin o f the word the evidence is no more than circum­ stantial. A vernacular form is attested in Middle Irish (tenth century), but before this the earliest evidence is Hiberno-Latin clocca in the late seventh century (in Adom nán’s Vita Columbae) and then in Anglo-Saxon circles on the continent with Irish connections. M uch more hypothetical is the 7 Prinz (1965), 347,351; (1981), 81; (1982), 210; Schäferdiek (1976), i86f. Franks: Ewig (1970), 32Í, 37; Prinz (1978), 459. Franco-Burgundians: Prinz (1978), 453, 455; (1981), 78. Gallo-Romans: Ewig (l 97°), 30fr., 44; Ewig and Schäferdiek (1978), 128. 8 Cf. Weisgerber (1952), 14E, 32. 9 Förster (1921), 37; Weisgerber (1952), 22; Eggers (1978), 492; Strasser (1982), 402fr.; Herren (1984), 20if.; F E W 2 ,149fr., 790fr.; GR 1 , 131; 2 ,190fr

328

Contact w ith Christianity

suggestion o f an Irish origin simply because o f the word’s geographical spread on the continent (northern France, Germany, northern Italy) in areas where Irish missionary activity is attested, for Lat. clocca is so widely distributed that it need not be tied down to Irish transmission. Even less convincing is the proposal that the word was felt to be Irish because o f a reference to what ‘in Irish is called the excommunication bell (clog) o f Blaithmac’, for this evidence is late and attests the word in Irish, but not as specifically Irish. A n Irish origin would be quite excluded if we were to accept the suggestion o f a Latin or Romance derivation (from a verb ‘to strike’ to a noun meaning ‘the stroke o f a bell’ or denoting the instru­ ment itself, ‘bell’).10 This proposal is based on the argument that an Irish origin would be too late to explain the palatalisation found in Italian and Iberian dialect forms. O ur evidence suggests that the Latin word was the source o f the continental forms, but the fact that the first attestation was Hiberno-Latin is not enough to establish an Irish origin for the Latin word. Even if we were to accept an Irish origin for this word, that is not the same as accepting its transmission to G erm any by Irish missionaries. T hat is particularly clear with the north Germ an clocka: although the Irish could conceivably be associated with the beginning o f the word’s route (France and even the lower Rhine), none was active in northern Germany, so that other forces must have brought it this far. N or can we fall back in this case on a transmission o f the word first from Irish into O E and only then to O S (as part o f the Anglo-Saxon mission to Germany). T he O E form o f this word is clugge, so that if the AngloSaxons brought the word at all it cannot have been in this form, but as a Latin term in use in the circle around Boniface.11 T h e south G erm an form glocka also presents difficulties for the Irish theory. We know o f no Irish mission from Bobbio to Germany, so that Irish impetus is as uncer­ tain in this case as with its north Germ an counterpart. Whereas an importation o f glocka from Italy cannot explain initial g- in place o f c(the north Italian dialect forms suggest only the latter), this could be regarded as a native U G change attested from the tenth/eleventh century.12 If that is accepted, there is no need to postulate a loan from Italy or to bring in Bobbio or an Irish impetus at all. This could imply a common entry for both the north and south Germ an forms, opening up the possibility o f an Irish influence in the south which was excluded for the north. But this is no more than a possibility for which we lack proof. 10 Meier (1975), 283ff.

11 Haubrichs (1987), 402, fn. 53.

12 Strasser (1982), 405fr.

T h e influence o f the M erovingian Franks

329

Irish influence on Germ an o f another type has been proposed for Lat. loanwords which, it is claimed, entered Germ an via Irish. For this we need specific pointers which for different reasons are lacking. G erm an terms for ‘psalm, psalter’ (O H G salm, salteri alongside occa­ sional psalm, psalteri) have been attributed to the influence o f Ir. salm, saltair,1314This is highly improbable, not merely because the consonant group ps- was unusual in Germ an and could therefore have been sim­ plified to S-. T h e same is true in O E (sealm) and already in V L (all the Rom ance derivatives likewise drop the p- and only restore it later, as occasionally in O H G , under the influence o f Church Latin). Given the wide range o f languages where this simplification o f an unusual sound group occurred we need a positive reason for assuming Irish influence in the case o f Germ an. Another proposal concerns Lat. versus ‘verse’, O H G fers (alongside vers) and Ir. fers.H Two arguments tell against Irish influence on Germ an here. Although Lat. v- can become f - in Ir. (e.g. verbum :ferb), this substitution can also occur elsewhere. T h e alternation between v and f i s also an internal O H G feature: the Ludwigslied has fand. alongside uolgon (where u stands orthographically for v) and Otfrid uses our word in both forms (fers and vers). With this alternation within Germ an it is difficult to argue that fers must betray external influence from Irish. Secondly, the form fers also occurs in O E where it may well be the result o f Irish influence on England (cf. also Firgilius). This English evidence now radically changes the position for O H G fers, for if we were to attribute this after all to external influence it is more likely to have come from the Anglo-Saxons (whose linguistic loans to Germ any are well attested) than from the Irish (where these still have to be established). T h e remaining suggestions for Latin-Irish—Germ an loanword traffic m ay be taken together: they all contain a Latin long ē.15 This regularly produced a diphthong in direct Latin loans into Germ an (e.g. speculum: spiagal), but a long i in Latin loans into Irish (e.g. census, V L cēsus: cis), so that Irish transmission has been put forward for some Germ an forms with i. A m ong these, however, we find sēta: sida ‘silk’ and crēta: krida ‘chalk’, trade words unlikely to be connected with an Irish mission and, with their intervocalic -d-, suggesting Rom ance transmission. Another consideration implies that a long ī need not point to the Irish.16 A t the time o f its earliest contacts with Romania Germ anic possessed the long vowels 5 and ē which were open, not closed, so that Latin words with 13 Weisgerber (1952), 25; Eggers (1978), 492i.; F E W 9, 500; GR 2, 421fr. 14 Jellinek (1925), n iff.; Weisgerber (1952), 23 (and fn. 60); Strasser (1982), 402. 15 Weisgerber (1952), 23fr.; Eggers (1978), 492. 16 Frings (1939), 107.

330

Contact w ith Christianity

closed long vowels were taken over as long ū and ī instead. Both these vowels were affected and not just in G erm an alone, which argues against restricting the G erm an examples with long ī to a specifically Irish influ­ ence. O ther factors bear this out for the words in question. As a term o f monastic discipline Lat. poena > pma ‘punishment’ was loaned into G erm an as pina, with Ir. pm (long and closed) as a suggested intermedi­ ary.17 Q uite apart from the argument o f the last paragraph, the Latin word was also loaned into O E (pīnian) so that, whether or not the O E form was under Irish influence, there is no certainty that the Germ an derived from Irish rather than English. With clēricus: clîrih it is uncertain whether the vowel was long or short in Latin and in Germ anic, but apart from that the word is also attested in Ir. clérech and O E cliroc.18 T h e fact that in G erm an as in English these forms are isolated hapax legomma con­ stitutes a classic case for loan-traffic and weakens further the suggestion o f direct influence from Irish to Germ an. This is weakened even more radically in the remaining cases. W hereas there is no reason to doubt the loan o f Lat. delete as a term o f the scriptorium into O E (ā)dilgian and O H G diligðn/tiligôn ‘erase, destroy’ it is by no means certain that We. dileu and Ir. dilegaid (only from Middle Irish) are loanwords rather than native formations.19 Even if they were to go back to Latin, the O E testimony and the importance o f Anglo-Saxon book culture on the continent mean that we lack evidence for direct Irish influence on Germ an. This is excluded for the last two cases suggested as examples o f Irish transmis­ sion: Lat. firia: O H G fir(r)a ‘ecclesiastical feastday, church service’ and expensa > expësa / spësa: spîsa ‘monastic provisions, food’.20 In neither case is the word attested in Irish or Celtic at all (OIr. has only féróil weo/7gallo-roman *(e)bescobo, *(e)bescobd/\dä. episcopus\ Sprachwissenschaft 2, i8iff.

B ibliography

411

Russell, F. H. (1975) The ju st war in the M iddle Ages, Cambridge Russell, J. C. (1994) The Germanization o f early medieval Christianity. A sociohistorical approach to religious transformation, New York Scardigli, P. (1964) Lingua e storia dei Gori, Florence (1995) ‘Fremde Einflüsse im Germanischen’, R LG A 9, Berlin, pp. 552fr. Schabram, H. (1987) ‘Altenglisch sāp: ein altes germanisches Wort für “Bernstein”?’ In: R. Bergmann, H. Tiefenbach, L. Votz (ed.), Althochdeutsch, Heidelberg, pp. i2off. Schäferdiek, K. (1976) ‘Bekehrung und Bekehrungsgeschichte.il. Deutschland und Nachbarländer’, R LG A 2, Berlin, pp. i8off. (1978a) ‘Germanenmission’, R L A C 10, 492fr. (1978b) ‘Die geschichtliche Stellung des sogenannten germanischen Arianismus’. In: H. Frohnes (ed.), Kirchengeschichte als Missionsgeschichte. Vol. h I, ed. K. Schäferdiek, D ie Kirche desfiüheren Mittelalters, Munich, pp. 79fr (1978c) ‘Die Grundlegung der angelsächsischen Kirche im Spannungsfeld insular-keltischen und kontinental-römischen Christentums’. In: H. Frohnes (ed.), Kirchengeschichte als Missionsgeschichte. Vol. 11 1, ed. K. Schäferdiek, D ie Kirche desfiüheren Mittelalters, Munich, pp. 149fr (1981) ‘Christentum der Bekehrungszeit’, R LG A 4, Berlin, pp. 501fr (1982) ‘Gab es eine gotisch-arianische Mission im süddeutschen Raum?’, ^

4 5 , 2 3 9 fr

(1984a) ‘Germanisierung des Christentums’, Theologische Realenzyklopädie 12, 52lff(1984b) ‘Kirihha - *cyrica - kuriakón. Zum geschichdichen Hintergrund einer Etymologie’, P B B 106, 46fr (1987) ‘Zur Frage früher christlicher Einwirkungen auf den west­ germanischen Raum’, Z K G 98,149fr. Scheller, M. (1950) Vedisch priyá- und die Wortsippe frei, freien, Freund. Eine bedeut­ ungsgeschichtliche Studie, Göttingen Scherer, A. (1955) ‘Die keltisch-germanischen Namengleichungen’. In: FS for F. Sommer, Wiesbaden, pp. 199fr. Schirokauer, A. (1946) ‘Wortgeschichte von H erd, GRev. 21, 55fr Schlesinger, W. (1963a) Beiträge zur deutschen Verfassungsgeschichte des Mittelalters. Bd. L: Germanen, Franken, Deutsche, Göttingen (1963b) ‘Über germanisches Heerkönigtum’. In: D as Königtum. Seine geistigen und rechtlichen Grundlagen. Mainauvorträge 1354 (Vorträge und Forschungen 3 ),

Darmstadt, pp. 105fr. Schmid. W. E (1973) ‘Aisten’, R LG A 1, Berlin, pp. n6ff. Schmidt, K. D. (1948) Germanischer Glaube und Christentum. Einzeldarstellungen aus dem Umbruch der deutschen Frühgeschichte, Göttingen (1954) Grundriß der Kirchengeschichte, Göttingen Schmidt, K.H. (1984) ‘Keltisch und Germanisch’. In: J. Untermann and B. Brogyanyi (ed.), D as Germanische und die Rekonstruktion der indogermanischen Grundsprache, Amsterdam, pp. 113fr (1986) ‘Keltisch-germanische Isoglossen und ihre sprachgeschichtlichen

4 12

B ibliography

Implikationen’. In: H. Beck (ed.), Germanenprobleme in heutiger Sicht, Berlin, pp. 23lff. (1987) ‘Handwerk und Handwerker im Keltischen und Germanischen. Beiträge zu einem historischen Vergleich’. In: W. Meid (ed.), Studien zum indogermanischen Wortschatz, Innsbruck, pp. 265fr. (1991) ‘The Celts and the ethnogenesis of the Germanic people’, Historische Sprachforschung 104,129fr. Schmidt-Wiegand, R. (1967) ‘Alach. Zur Bedeutung eines rechtstopograph­ ischen Begriffs der fränkischen Zeit’, B z N (N.F.) 2, 2iff. (1974) ‘Fränkisch druht und druhtin. Zur historischen Terminologie im Bereich der Sozialgeschichte’. In: FS for W. Schlesinger, Cologne, pp. 524ff (1978) ‘ Wargus. Eine Bezeichnung für den Unrechtstäter in ihrem wort­ geschichtlichen Zusammenhang’. In: H. Jankuhn (ed.), ‘Zum Grabfrevel’, Abhandlungen der Akademie der Wissenschaften 113, Göttingen, pp. i88ff. (1979) ‘Die volkssprachlichen Wörter der Leges barbarorum’. In: FS for K. S. Bader, Zürich, pp. 4i9ff. (1987) iReht und ewa. Die Epoche des Althochdeutschen in ihrer Bedeutung für die Geschichte der deutschen Rechtssprache’. In: R. Bergmann, H. Tiefenbach, L. Votz (ed.), Althochdeutsch, Heidelberg, pp. 937fr. (1994) ‘Ewa’, R IG A 8, Berlin, pp. 35fr. Schneider, H. (1928) Germanische Heldensage. I. Band. Einleitung: Ursprung und Wesen der Heldensage. I. Buch: Deutsche Heldensage, Berlin (1938) D ie Götter der Germanen, Tübingen Schneider, R. (1978) ‘Karl der Große - politisches Sendungsbewußtsein und Mission’. In: H. Frohnes (ed.), Kirchengeschichte als Missionsgeschichte. Vol. 111, ed. K. Schäferdiek, D ie Kirche desfrüheren Mittelalters, Munich, pp. 227fr Schnetz, J. (1938) ‘Unnamen beim Geographen von Ravenna’, ffr fF i^ , 85fr (1951/2) ‘“Baias” und der Baiernname’, ffr L g 16, iff. Schönfeld, M. (1911) Wörterbuch der altgermanischen Personen- und Völkernamen, Heidelberg Schramm, G. (1957) Namenschatz und Dichtersprache. Studien zu den zweigliedrigen Personennamen der Germanen, Göttingen (1974) ‘Die nordöstlichen Eroberungen der Rußlandgoten (Merens, Mordens und andere Völkernamen beijordanes, Getica xxm 116)’, F M S S , iff. (1975) ‘Hunnen, Pannonier, Germanen. Sprachliche Spuren von Völkerbeziehungen im 5. Jh. n. Chr.’, ZeitschriftJur Balkanologie 11,2, 7iff. Schröbler, I. (1958) Notker I I I von St. Gallen als Übersetzer und Kommentator von Boethius, D e Consolatione Philosophiae, Tübingen Schröder, E. (1900) ‘Zu Genesis und Heliand’. ZfdA 44, 223fr (1918) ‘Studien zu den deutschen Münznamen’, 48, 241fr (1923) ‘Tilgen’, fJd A 60, 246fr (1924a) ‘“Herzog” und “Fürst”: Über Aufkommen und Bedeutung zweier Rechtswörter’, Zf tGfGA) 44, iff. (1924b) ‘Bunte Lese 11. 6. Writan und skribari, ZfdA 61, 57i.

B ibliography

413

(1925) ‘Got. kintus’ , £ V S 53,8off. (1932) ‘Herzog’, Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen. Philologisch- historische Klasse, pp. 182fr. Schröder, W. (1967) Review of G. Becker, Geist und Seele im Altsächsischen und im Althochdeutschen, AfdA 78, 79fr. Schultze, A. (1941) ‘Das Eherecht in den älteren angelsächsischen Königs­ rechten’, Berichte über die Verhandlungen der sächsischen Akademie der Wissen­ schaften 93, 5 (1943) ‘Über westgotisch-spanisches Eherecht’, Berichte über die Verhandlungen der sächsischen Akademie der Wissenschaften 95, 4 Schulze, W. (1907) ‘Gotica’, £ V S 41,165fr. Schürer, E. (1905) ‘Die siebentägige Woche im Gebrauche der chrisdichen Kirche der erstenJahrhunderte’, Zeitschriftfür neutestamentliche Wissenschaft 6, iff. Schütte, G. (1902) Review of O. Bremer, Ethnographie der germanischen Stämme, AfdA 28, 4fr Schützeichel, R. (1957) ‘Bezeichnungen für “Forst” und “Wald” im frühen Mittelalter’, ZfdA 87,105fr. Schwab, U. (1972) arbeo laosa. Philologische Studien zum Hildebrandslied, Bern Schwarcz, A. (1992) ‘Die gotischen Seezüge des 3. Jahrhunderts’. In: R. Pillinger, A. Pütz, H. Vetters (ed.), D ie Schwarzmeerküste in der Spätantike und im frühen Mittelalter, Vienna, pp. 47fr. Schwarz, E. (1951) Goten, Nordgermanen, Angelsachsen. Studien zur Ausgliederung derger­ manischen Sprachen, Bern (I953a) ‘Die Urheimat der Goten und ihre Wanderungen ins Weichselland und nach Südrußland’, Saeculum 4 ,13fr. (1953b) ‘Die Krimgoten’, Saeculum 4, 156fr. (1956) Germanische Stammeskunde, Heidelberg See, K. von (1964) Altnordische Rechtswörter. Philologische Studien zur Rechtsauffassung und Rechtsgesinnung der Germanen, Tübingen (1972) Kontinuitätstheorie und Sakraltheorie, Frankfurt (1981) ‘Berserker’, 2fd .W 17, 129fr. Sehrt, E. H. (1962) Notker-Glossar. Ein althochdeutsch-lateinisch-neuhochdeutsches Wörterbuch zu Notkers des Deutschen Schriften, Tübingen (1966) Vollständiges Wörterbuch zum Heliand und zur altsächsischen Genesis, Göttingen Seneca, D e ira, ed.J. W. Basore, Seneca. M oral essays, Cambridge, Mass., 1994, pp. io6ff. Senn, A. (1925) Germanische Lehnwortstudien, Heidelberg Sieg, G. (i960) ‘Zu den Merseburger Zaubersprüchen’, P B B 82, 364fr Sigurðarkmða in skamma, ed. G. Neckel, Edda. D ie Lieder des Codex Regius nebst ver­ wandten Denkmälern, Heidelberg 1914, pp. 202fr. Skok, P. (1923) ‘Gibt es altgermanische Bestandteile im Rumänischen?’, ffr P h 43, 187fr.

414

B ibliography

(1925) ‘La semaine slave’, Revue des Etudes Slaves 5 , 14fr.

(1930) ‘Gibt es altgermanische Bestandteile im Rumänischen?’, ŽfrPh 50,257ff. Sonderegger. S. (1965) ‘Die ältesten Schichten einer germanischen Rechtssprache’. In: FS for K.S. Bader, Zürich, pp. 4igff. Specht, F. (1939) ‘Greutungi - Graudenz?’, f V S 66, 224fr. Sperber, H. (1915) ‘Beiträge zur germanischen Wortkunde. D. keusch’, W uS 6, 5 5 ff-

Sprandel, R. (1978) Verfassung und Gesellschaft im M ittelalter, Paderborn Stancliffe, C. E. (1980) ‘Kings and conversion: some comparisons between the Roman mission to England and Patrick’s to Ireland’, F M S 14, 59fr Starck, T. and Wells, J. C. (1990) Althochdeutsches Glossenwörterbuch (mit Stellennachweis zu sämtlichen gedruckten althochdeutschen und verwandten Glossen),

Heidelberg Stearns, M. (1978) Crimean Gothic. Analysis and etymology o f the corpus, Saratoga, Calif. (1989) ‘Das Krimgotische’. In: H. Beck (ed.), Germanische Rest- und Trümmersprachen, Berlin, pp. 175fr. Steinhäuser, W. (1928) ‘Eintritt der Stimmhaftigkeit bei den westgermanischen Reibelauten f , þ , s, X, - In: FS for M. H. Jellinek, Vienna, pp. 139fr. Steinmeyer, E. von (1916) D ie kleineren althochdeutschen Sprachdenkmäler, Berlin Steinmeyer, E. von and Sievers, E. (1879fr.) D ie althochdeutschen Glossen, Berlin Steuer, H. (1994) ‘Fahren und Reiten. 6: Karolinger- und Wikingerzeit’, R IG A 8, Berlin, pp. 165fr. Stevenson, J. (1994) ‘Christianity in sixth- and seventh-century Southumbria’. In: M. O. H. Carver (ed.), The Age o f Sutton Hoo. The seventh century in north­ western Europe, Woodbridge, pp. 175fr. Straka, G. (1951) ‘Observations sur la chronologie et les dates de quelques mod­ ifications phonétiques en roman et en français prélittéraire’, Revue des langues romanes 71, 247fr. (1956) ‘La dislocation linguistique de la Romania et la formation des langues romanes à la lumière de la chronologie relative des changements phoné­ tiques’, Revue de linguistique romane 20, 249fr. Strassburger Eide, ed. E. von Steinmeyer, D ie kleineren althochdeutschen Sprachdenkmäler, Berlin 1916, pp. 82fr. Strasser, I. (1982) ‘Irisches im Althochdeutschen?’. In: H. Löwe (ed.), D ie Iren und Europa im früheren Mittelalter, Stuttgart, pp. 399fr. Streitberg, W. (1894) ‘Ost- Westgoten’, I F 4, 300fr. (1919) D ie gotische Bibel, Heidelberg Ström, F. (1951) ‘Tro och blot. Tvâ grundbegrepp i fornnordisk religion’, Arv 7, 23fr.

Struve, K. W. (1991) ‘Zur Ethnogenese der Slawen’. In: M. Müller-Wille (ed.), Starigard/ Oldenburg. E in slawischer Herrschersitz des frühen Mittelalters in Ostholstein, Neumünster, pp. gff. Stutz, E. (1966) Gotische Literaturdenkmäler, Stuttgart

(1980) ‘Die germanistische These vom “Donauweg” gotisch-arianischer

B ibliography

415

Missionare im 5. und 6. Jahrhundert’, Denkschriften der Österreichischen 145, pp. 207fr. Suerbaum, W. (1977) Vom antiken zum frühmittelalterlichen Staatsbegriff. Über Akademie der Wissenschaften

Verwendung und Bedeutung von Respublica, Regnum, Imperium und Status von Cicero bis Jordanes, Münster Swanton, M. (1970) The Dream o f the Rood, Manchester Swiggers, P. (1985) ‘Gothic siponeis’, ^ V S 98, 109i.

Szemerényi, O. (1952) ‘The development of the Indo-European mediae aspiratae in Latin and Italic’, Arckivum Linguisticum 4, 99fr. (i960) ‘Gothic auhuma and the so-called comparatives in -uma', P B B 82, iff. (1979) Germanica I (1-5)’, Z yS 95> 103fr. Tacitus, Agricola, ed. M. Hutton, revised H. M. Ogilvie, Cambridge, Mass., 1970 Annales, ed. E. Koestermann, Leipzig 1952 Germania, ed. J. G. C. Anderson, Oxford 1938 Historiae, ed. E. Koestermann, Leipzig 1950 Tatian (Old High German), ed. E. Sievers, Paderborn 1872 Thompson, E. A. (1965) The early Germans, Oxford (1966) The Visigoths in the time o f Ulftla, Oxford (1969) The Goths in Spain, Oxford Thumb, A. (1901) ‘Die Namen der Wochentage im Griechischen’, ftfd W 1,163fr. Thurneysen, R. (1901) ‘Die Namen der Wochentage in den keltischen Dialecten’, ff d W 1, i86ff. Tiefenbach, H. (1995) ‘Friede. Sprachliches’, R IG A 9, Berlin, pp. 594fr Timpe, D. (1989) ‘Entdeckungsgeschichte’, R LG A 7, Berlin, pp. 307fr (1992) ‘Tacitus’ Germania als religionsgeschichdiche Quelle’. In: H. Beck, D. Ellmers, K. Schier (ed.), Germanische Religionsgeschichte. Quellen und Quellenprobleme, Berlin, pp. 434fr Todd, M. (1992) The early Germans, Oxford Toth, K. (1995) ‘Fremde Einflüsse im Germanischen. Das Beispiel des Altenglischen’, R IG A 9, Berlin, pp. 563fr Ullmann, W. (1961) Principles o f government and politics in the M iddle Ages, London Unruh, G. C. von (1957) ‘ Wargus. Friedlosigkeit und magisch-kultische Vorstellungen bei den Germanen’, ZR G (G A ) 74, iff. Untermann, J. (1954) ‘Über die historischen Voraussetzungen für die Entlehnung von got. alew\ P B B 76, 390fr. Usener, H. (1911) D as Weihnachtsfest, Bonn. Second edition Väänänen, V (1967) Introduction au latin vulgaire, Paris Vernadsky, G. (1951) ‘Der sarmatische Hintergrund der germanischen Völkerwanderung’, Saeculum 8, 340fr Völundarkviða, ed. G. Neckel, Edda. D ie Lieder des Codex Regius nebst verwandten Denkmälern, Heidelberg 1914, pp. 112fr Vollmann-Profe, G. (1987) Otfrid von Weißenburg, Evangelienbuch. Auswahl, Stuttgart Vries, J. de (1941) Altnordische Literaturgeschichte 1, Berlin (1956a) Altgermanische Religionsgeschichte 1, Berlin

4i 6

B ibliography

(1956b) ‘Das Königtum bei den Germanen’, Saeculum 7, 289fr. (i960) Kelten und Germanen, Bern Waag, A. (1932) ‘Die Bezeichnungen des Geistlichen im Althoch- und Altniederdeutschen. Eine wortgeschichtliche und wortgeographische Untersuchung’, Teuthonista 8, iff. Wagner, N. (1967) Getica. Untersuchungen zum Leben des Jordanes und zur frühen Geschichte der Goten, Berlin (1975) ‘ Cheisuringu gitan. Zu v. 33~35a des Hildebrandsliedes’, EfdA 104, 179fr. (1979) ‘Ostgermanisch-alanisch-hunnische Beziehungen bei Personernaraen’. In: R. Schützeichel (ed.), Studien zur deutschen Literatur des Mittelalters, Bonn, pp. uff. Wahmann, P. (1937) Gnade. Der althochdeutsche Wortschatz im Bereich der Gnade, Gunst und Liebe, Berlin Walde, A. (1906) Lateinisches etymologisches Wörterbuch, Heidelberg Wallace-Hadrill,J. M. (1962) The long-haired kings and other studies in Frankish history, London (1971) Early Germanic kingship in England and on the continent, Oxford (1975) Early medieval history, Oxford Wartburg, W. von (1950) D ie Ausgliederung der romanischen Sprachräume, Bern (1953) ‘Sabbatum: Samstag’. In: FS forJ. Orr, Manchester, pp. 296fr. (1956) Von Sprache und Mensch, Bern (1968) ‘Ursprung und Geschichte des Wortes bâtir'. In: FS for E. Gamillscheg, Munich, pp. 639fr Webster, L. and Backhouse, J. (1991) Eds., The making o f England. Anglo-Saxon art and culture A D 6 oo-g oo, London Webster, L. and Brown, M. (1997) Eds., The transformation o f the Roman world, A D 40 0 -goo, London Weisgerber,J. L. (1949) Der Sinn des Wortes ‘Deutsch', Göttingen (1952) ‘Die Spuren der irischen Mission in der Entwicklung der deutschen Sprache’, Rh V 17, 8ff. (1953) Deutsch als Volksname. Ursprung und Bedeutung, Stuttgart Weisweiler, J. (1923) ‘Beiträge zur Bedeutungsentwicklung germanischer Wörter für sitdiche Begriffe. Zweiter Teil’, I F 41, 304fr (1929) ‘Bedeutungsgeschichte, Linguistik und Philologie. Geschichte des ahd. Wortes euud. In: FS for W. Streitberg, Heidelberg, pp. 419fr (1940) ‘Seele und See. Ein etymologischer Versuch’, I F 57, 25fr Weisweiler, J. and Betz, W. (1974) ‘Deutsche Frühzeit’. In: F. Maurer and H. Rupp (ed.), Deutsche Wortgeschichte, Berlin, pp. 55fr Wenskus, R. (1961) Stammesbildung und Verfassung. D as Werden derfrühmittelalterlichen Gentes, Cologne (1973a) ‘Alanen’, R LG A 1, Berlin, pp. 122fr. (1973b) ‘Attila. Historisches’, R LG A 1, Berlin, pp. 467fr. (1978) ‘Boihaemum’, R LG A 3, Berlin, pp. 2076 (1984) ‘Ding. Historisches’, R IG A 5, Berlin, pp. 444fr

Bibliography

417

(1986a) ‘druht’, R IG A 6, Berlin, pp. 202i. (1986b) ‘dux. II. dux in altgerm. Zeit’, R IG A 6, Berlin, pp. 30iff. Werner, J. (1950) ‘Römische Trinkgefäße in germanischen Gräbern der Kaiserzeit’. In: FS for E. Wahle, Heidelberg, pp. i68ff. Wesche, H. (1932) D as Heidentum in der althochdeutschen Sprache. I. Teil: die Kultstätte, dissertation Göttingen (1937) ‘Beiträge zu einer Geschichte des deutschen Heidentums’, P B B 61, iff. White, L. T. (1962) M edieval technology and social change, Oxford White, S. D. (1989) ‘Kinship and lordship in early medieval England: the story of Sigeberht, Cynewulf, and Cyneheard’, Viator 20, iff. Whitelock, D. (1979) English historical documents, c. 50 0 -1042, London Widsith, ed. K. Malone, London 1935 Wiesinger, P. (1985) ‘Gotische Lehnwörter im Bairischen. Ein Beitrag zur sprach­ lichen Frühgeschichte des Bairischen’. In: H. Beumann and W. Schröder (ed.), Frühmittelalterliche Ethnogenese im Alpenraum, Sigmaringen, pp. 153fr. Wiesner, J. (1963) D ie Thraker, Stuttgart (1968) D ie Kulturen derfrühen Reitervölker, Frankfurt Wissowa, G. (1918) ‘Interpretatio Romana. Römische Götter im Barbaren­ lande’, ArchivJur Religionswissenschaft 19, iff. Wolfram, H. (1968) ‘Methodische Fragen zur Kritik am “sakralen” Königtum germanischer Stämme’. In: FS for O. Höfler, Vienna, pp. 473fr. (1975) ‘Gotische Studien I/IF, M IÖ G 83, iff, 289fr (1976) ‘Gotische Studien IIP, M IÖ G 84, 239fr (1988) History o f the Goths, Berkeley (1989) ‘Ermanarich. Historisches’, R IG A 7, Berlin, pp. 510fr (1991) ‘L’itinéraire des Goths de la Scandinavie à la Crimée’. In: A. Rousseau (ed.), Sur les traces de Busbecq et du gotique, Lille, pp. 135fr Wollmann, A. (1990a) Untersuchungen zu den frühen lateinischen Lehnwörtern im Altenglischen. Phonologie und Datierung, Munich (1990b) ‘Lateinisch-Altenglische Lehnbeziehungen im 5.und 6.Jahrhundert’. In: A. Bammesberger and A. Wollmann (ed.), Britain 400-600: language and history, Heidelberg, pp. 373fr (1990c) ‘Zu den frühesten lateinischen Lehnwörtern im Altfriesischen’, A B äG 31/32, 506fr. (1993) ‘Early Latin loan-words in Old English’ A S E 22, iff. Wood, I. N. (1977) ‘Kings, kingdoms and consent’. In: P. H. Sawyer and I. N. Wood (ed.), Early medieval kingship, I^eeds, pp. 6ff. (1987) ‘Pagans and holy men, 600-800’. In: P. Ní Chatháin and M. Richter (ed.), Irland und die Christenheit, Stuttgart, pp. 347fr. (1994a) ‘Frankish hegemony in England’. In: M. O. H. Carver (ed.), The Age o f Sutton Hoo. The seventh century in north-western Europe, Woodbridge, pp. 2 3 5 ff-

(1994b) The Merovingian kingdoms 4 50 -751, London (1995) ‘Pagan religions and superstitions east of the Rhine from the fifth to

4i 8

Bibliography

the ninth century’. In: G. Ausenda (ed.), After Empire. Towards an ethnology o f Woodbridge, pp. 253fr. (1997) ‘The transmission of ideas’. In: L. Webster and M. Brown (ed.), The transformation o f the Roman world A D 40 0 -goo, London, pp. 11iff. Wormald, P. (1991) ‘The age of Bede and Aethelbald’. In:J. Campbell (ed.), The Angb-Saxons, Harmondsworth, pp. 70fr Wulfila, D ie gotische Bibel, ed. W. Streitberg, Heidelberg 1919 Würzburger Markbeschreibung, ed. E. von Steinmeyer, D ie kleineren althochdeutschen Sprachdenkmäler, Berlin 1916, pp. 115fr Yeandle, D. N. (1989) ‘The Ludwigslied: King, Church and Context’. In: J. L. Flood and D. N. Yeandle (ed.), ‘M it regulu bithuungan’. Neue Arbeiten zur althochdeutschen Poesie und Sprache, Göppingen, pp. i8ff. Yorke, B. A. E. (1981) ‘The vocabulary of Anglo-Saxon overlordship’. In: D. Brown,J. Campbell, S. C. Hawkes (ed.), Angb-Saxon studies in archaeology and history 2 (BAR British Series), Oxford, pp. 171fr Zeiss, H. (1932) ‘Herzogsname und Herzogsamt’, Wiener Prähisbrische Zeitschrift 19. !45ffZimmer, H. (1892) ‘Aus der Bedeutungsgeschichte von schreiben und schrifi, 2fd A Europe’s barbarians,

36, 145fr.

Zwölfer, T. (1929) Sankt Peter. Apostelfurst und Himmelspförtner. Seine Verehrung bei den Angelsachsen und Franken, Stuttgart

In dex o f words

Greek and Latin are included in this index as the languages o f our earliest sources on Germanic society and as the languages providing concepts and models for adoption in the barbarian languages. N o distinction is made between Classical and Vulgar Latin, and Latinised forms o f Germ anic terms are included under Latin. Although the text o f this book frequently quotes words from only one or other Germ anic language, this index lists the cognate forms in the other languages to facilitate access from whatever particular language with which the reader may be concerned. Variant dialectal forms, especially in O ld High Germ an, are not listed. GREEK Aland 170 anachorètes 347 anagignöskein 263 álsos 27 Aphrodite 238 archisynágögos 180 À\reös hëméra 244, 250, 311, 359 Arès 238 aretë 379f. Arkýnia 159E asebes 317 baitè 177 baptizein 314fr. baptós 315 basileús 124, 135 basilikè 296fr. bíblos 259 broútis 184 chaire 18 chilíarchos 180 despotēs 103, 180 dikaíosis 284 drachmē 205 droüngos 185 dýnamis 379f. eirèriè 52, 284 ekktèsia 296fr eleèmosynè 306

G R E E K (cont.) eleútheros 96 episkopos 301fr. éthnos 125 gemein 317 gemízein 317 geneá 62 génos 62, 63 gérontes 115 gerousia 115 grámma 257, 259, 261, 263 gráphai 261 gráphein 263 graphe 263 hekatontárches 180 Helios 238 Hermès 238 kairós 35 kánnabis 165 kárkaron 203 Karpátes 164 koíranos 124 Krónos 238 kyriake 237, 241, 245, 249 kyriakón 296fr, 302, 313 kyri(a)kós 298 kýrios 104, 284, 357 laós 125 legem 84

420

In d ex o f w ords

G R E E K (cont.) melon 263 mêlas 263 men 150 monachós 339 monastmon 338 óchlos 125 oikodespotēs 102, 103 Orkÿnios 159i. orphanós 153 papās 3ogf. paraskeue 237, 239, 252, 310, 360 paroUda 335 pároikos 335 páscha 283 philos 55, 58 phylë 62 pimplánai 317 pímplemi 90 pisteúein 289 plëroün 317 plethos 90 póntos 179 porphýra 205 pósis 180 présbeis 115 presbýteros 115, 335 prosábbaton 237 pýrgos 157,158 sábanon 205 sábbata/sábbaton 237, 238, 239 sákkos 205 *sámbaton 312 Selēnē 238 sótería 284 soudárion 204 stoichdon 256, 382i. stratēģos 234i. stratía 84 syngéneia 62 syngenes 58 theosebēs 22 thorax 155 238

L A T IN accumbere 203 acetum 202, 207 adimplere 317

L A T IN (cow/.) 54, 61 aßnitas 54, 61 agmen 85 alba 352i. allodium 198 omārus 189 ambactus 149 amicus 57, 58, 59 amp(h)ora 229 anachoreta 347i. angélus 284 anima 284 annona 203, 221, 238 antrustiones n o arbitrari 289 wraz 205, 229 architriclinus n i, 112 asellus/asinus 171, 204, 213, 229i. assarius 205 astracum/astricum 21 o atramentum 263, 267 audire 377 aurora 352 aoMJ 54 Baibari 161 Baioarii 161 bandum 234 baptizare 315Í, 333 basilica 297 bellum 87 Boihaimum/Boiokaemum 16 if. Æozï i6off. 146i., 190 brevis 213i., 269 brutis 184i, 195 Caesar 151 calamus 267 calcare 212 calcatorium/calcatura 212 calx 21of. cambiare 189 caminus 189 campana 327 campio 74i., 234 campus 74, 234 canonicus 339 canticum 283 Cantuarii 161 *capillare 203 carcar/career 26, 203, 209, 233 carrago 185,191, 230 carrus 185

In d ex o f words L A T IN (cant.) caseus 212 castra 217 catena 210, 233 caterva 91, 92 catillus/catinus 172, 205, 229 caulis 211 caupo(nari) 224fr. causa 209 cautio 2o8f. cellarium 210, 214 census 209 cententionalis 205 centum 180 character 257 charta 266 *christian(iz)are 333 cilicium 205 cista 214, 229 claustrum/clostnim 209F., 233 clericus 330 clibanus 214 dims 59,149 clocca 327i. coenobium 338 cognatio 62 cohort 90, 91,92 collis 157 columna 157 comitatus/comités 107, io8f., n o communis 103 companio 194, 230 compater 334

compterez concibo 194 coniux 194 consacramentales 50 conscientia 284, 319 conscius 319i. consotatio 349 consors 194 contio 94 coquere 212 coquina/cocina 212 coquus/cocus 212 credere 289 329 >33° era* 213, 214, 348 cubitus 203 cuneusji, 90, 92 custor/custos 336F. daps 23 268, 330 deus 13

421

L A T IN (eoni.) discipulus 311 dominica 241, 243 dominica vespera 252i. dominicum 297, 300 dominicus 106, 241, 242, 243, 244, 245, 249 dominus 103, 105, n o , 115, 118, 124, 128, 284, 285, 358, 361fr. domus 337i. drachma 205 draco 234 dructis n o , 112 drungus 185, 191, 230 ducere 235 dux 122,137,138, 234!. ecclesia 283, 296, 303 electrum 187 eleeomosyna 306 elementum 256, 382!. eligere 265 emandpatdo 390 encaustum 267 episcopus 301fr. équités 85 equus 148, 179 evangelium 284, 331, 346!. ewa 31 exerdtus 85, 90, 92 expensa/expesa/spesa 330 faida/faidus 75 fascia 205 fatum 381, 386 feria 330 ferialis 330 feudum 190,198 feum/fevum 193 fideles 101 fiscus 103, 105 flamm 22 foederare/foederatus 61 foedus 61 fons 333f foresta/forestis 199 framea 69, 185F., 230 furca 211 ganta 186, 191, 193, 220, 227 gardingus 130 *gardinus 199 genimina 62 gens 62, 63, 126, 131 genus 125,137, 285 genus 62, 63, 125

422 L A T IN (coni.) Gepidoios 167 gigno 62 glaesum/glesum 187!, 220, 227 Gothiscandza 166 grammatica 339 grapkium 267 gratia 285, 331 Greutungi i68f. harebanruis/ haribannum 197 hasta 186 Hercules 247, 248 Hercynia/Hercynins 159F,161, 162 heremita 283 homo/homines ioi horreum 211 hortinus 199 hostis 85 humilitas 51, 350 implere 317 Iovis dies 240, 241, 248, 251! indicate 44 iudex 44 indicium 44 úmtör 340 Iupiter 245, 248 iuratores 50 iustificatio 284 Latiņus 217 legere 265, 359 Âgà>9i leudis/leudes 96, ioi libellas/liuellus 269 /tfor (adjective) 96 Äfor (noun) 259, 269 libertas arbitrii 390 /tfra 152, 225, 244 /i/fera 256, 257, 259, 261 lucerna 204, 207, 216, 223 Luna 245 Lara* 240, 241, 243, 245, 249 mam 115 mallum 30,197 mango 224, 225 manipulas 90, 91 mannitus 30 manus 197 mariscalcus 196 Mars 245, 247 Mariis dies 240, 241, 248, 250, 311, 381 martyrium 336

In d ex o f w ords L A T IN (coni.) media hebdomas 242, 251 mercans 225 *mercatans 225 Mercurii dies 240, 241, 248, 251 Mercurius 245, 247 meretrix 214 milia (passuum) 233 militare 202, 233 milites 85, 233 militia 85 miscere 228 misericordia 285 monachus/monicus 339 monarchus 113 monasterium 338F. malleus 263 mains 204, 230 munburdum 197 mara 211 mastam 212 Kali# 126 navis 214 27, 28 nota* 257 numm 14 nundinum 240 obediens 377 obedientia 375fr. oMtre 377 offene 20, 305F o/ium 212 operari 305F optimales 118 orarium 204F, 208 orbus 153 osa 190 pacifiais 61 pacta/pactum 209 paganos 211 papa 309 papyrus 266, 269 paranymphus i n , 112 parasceue 310 parca(e) 381, 384 par(r)ochia 335F pater 334 paterfamilias 120 patrinus 334 pax 43F, 61 pedes 85 penales 130

In d ex o f words L A T IN (cont.) penna/pinna 267 pergamentam 266 perindie 311 persietim 211 pigmentum 269 *piluccare 211, 227 pilum 69 pistor 212, 214 planta 211, 212 plebsgo, 125, 126 pieņus 90 pluma 186, 227 poena/peņa 330 pondo 207, 225, 229, 244 /ww 179 populus 9 0 ,95> 97f-> !25 180 praepositus/propositus 339i. presbyter 113, 334i primates 102 primi 102 princeps 102, 103, 105, 106, 107, 109, n o , 113, 117, 118, 122, 127 priores 102 progenies 61, 62 propinquitas 54, 61 providentia 331 prunum 211 psalmus 329 psalterium 329 publiais 103 pugillaris 260, 267 pulvinus 227 /w m x 268 purpura 205 /wtor* 289 puteus 210, 214 quercus 159 radix 211 recumbere 203 redemptio 285 regnare gyf. regales 133,138 regula 211 régufoi 133,134,138 replere Vļ respublica 104, 105, 119 rex I2if., 127, 128, 131, 138, 150 sabanum 205, 207 sabbata/sabbatum 237, 241, 244, 245, 252 .weettf 205, 229

L A T IN (corc/.) sacellum 336 sakramentales 50 sagena 214 284 salvator 355 sambatum 252, 312i. sonetus 353i. sanguis 62 180, 193, 220 sarcophagus 311 satisfactio 52 Saturni dies 240, 241, 244, 246, 252 Saturnus 244 scancio 197 scribere 263L scripta 264 securus 209 Wíftö 61 senātus 115 senescaleus/siniscalcus 197 113 j£7ztor 113fr., 340 seta 329, 330 233i. ■ rçgnum233Í, 283, 327 siliqua 230 jimzTû 212 Sol 245 solarium 210 240, 241, 244, 245, 246, 248i., 368 speculator 203 speculum 329 speruarius 199 Spiritus 287, 354f. strata 214, 233, 235 97, 100, 101 subiecti 100 sucinum 187 sucus 187 sudarium 204 suffragines 146 jtfwzw 30 Surus/Syrus 204 tabella/tabula 260, 265i. 211, 212 teloneum/toloneum 209 Tervingi 169 Thingsus 34f., 250 267 tingere 315i. 336 traiectorium 212 tribūnus 33, 124, 130, 180

423

424

In d ex o f w ords

L A T IN (cont) tribus 63, 94 trustis n o ^234 turba 91 tyrannus 117

G O T H I C (cont.) assarjus 205 aúrali 204i., 208 aúrtigards 173 *ausahriggs 173 awo 54

UlîTierugi 166 uncia 202, 229

biháitja 74 biuþs 173

vallum 213 vanga 192 vates 79 Veneris dies 240, 241, 248, 252 Venus 245, 248 versus 329 vertere 384 vicia 214 vinum 2, 204, 207, 213 vir 378! virtus 378fr. Volcae 162i. vulgus IOI wacta 197 wadium/ wadius 197 waldus 199 wangia 192 werra 87

G O T H IC afarsabbato 239

Mtoz 22, 288 bāka 173, 257, 259fr., 359 irun> I, 71, I55f., 173, 198 brunna 210 155 brúþfaþs 180 bruþs 184i. dags 352 dáupjan 314fr, 333 daúrawards 191 diups 315 dömjan 44fr. doms 44fr. drakma 205, 208 draúht- 68, io6ff. draúhtinassus 110, 363 draúhtinön n o , 203, 363 *draúhtins io6ff., 358, 362!., 370 draúhtiwitoþ n o , 363 driugan i n dulþs 319 eisarn 153fr.

oféoAr 3'7f0*^354 ahma 354 aikklésjo 286, 299 aipiskaûpus 301fr. 158 aiwaggelista 286 akeit 173, 207, 208, 213 alabrunsts 23 alëw 158, 204, 212, 223, 227 aZfa 27 anakumbjan 203, 208 anakunnan 263 andbahts 149! a/rno 203, 208, 221, 233 ansts 319 arbi 153 arbja 153 *arjausdags 244 arka 205, 208 armahaírtiþa 285 armahaírts 316 asilus 173, 204, 207, 208, 213, 229i.

faírguni 159 fasķja 205 frrabaúhtaböka 260i. frram 186 frastisibja 60 fráuja I02ÍT, 357, 361fr, 372 freihals 41, 56 freis 39ff- ,56, 152 frijön 39fr, 42, 55, 57, 248 frifinds 52, 55fr., 59 *fripus 42fr, 52, 56 Julljan 317 Jullnan 317 gadáila 194 gadraúhts io6fT, n o , 363 gagawaírþjan 53 gahláiba 194 gakunnan 263, 265, 359 Galatja 286 galáubjan 20, 288!. Galeilaia 286

In d ex o f words G O T H I C (cont.) galga 348 galiugaguþ 317 gardüwaldands 102 gards 102, 130, 173 gasibjön 60, 61 gaskapjan 387i., 388 *gastfaþs 173 gawaírþi 52i. gawaljan 265 gudja 33f., 130, 133, 309 guþ 14fr. guþblöstreis 22, 29 Gutþiuda 125 háilags i6ff., 353, 361, 362, 365 háiljan 358 háiLs i6ff. háims 103, 161 haírda 173 haírus 70 hansa 173 harjis 68, 84fr. heiwqfráaja 102, 103 hilms 71, 171, 173, 190 hláifs 171, 173 hundafaþs i8of. hunsl 23 hunsljan 23, 288 hweláuþs 96 Iesus Xristus 286 Israelites 286 Iudaia 286 jiuleis 353 juggaláuþs 96 kapiltón 203, 208, 233 karkara 203, 208, 209 katiLs/katilus 171, 173, 205, 207, 208, 213 káupön 173, 204, 208, 213, 225i. kawtsjo 2o8f., 214 kelikn 157i. kindins n o , 122, 124 kintus 173, 205 kubitus 203, 208 kunawida 132 kuni6 2Í, 125 karman 263 2of. laíktjo 209 lēkeis 173 lisan 264i. liudan 96

G O T H I C (cont.) Hufs 288 lubáins 288 lukarn 158, 204, 207, 208, 213, 227 Makidoneis 286 Makidonja 286 managei 125 marqan 189 máfa’ I, 70, 173, 178 mēļa 263, 359 mēljan 263, 359 150 militön 202, 208, 233 möto 171, 173 nasjan 358 19, 355, 357 Naúbaímbaír 207 þáida 177i. /KZ/NZ309, 311,312 paraskaiuuē 311 pasha 351, 360 paúrpaúra/paúpura 205 plinsjan 177 pund 207, 208, 213 ragin 13 (ga-)rēdan 257i. 139i., i50f. reikinön 151 reiks 139i., i5of. Rumöneis 157 nwa 255, 387 205, 207, 213 sabbato 252, 311 173, 205, 207, 208, 213 *sambato 312 sarwa 171 204 23 Jí'é/a 52,54, 59fr. sinista 115 sipöneis 156i. skaban 203 skilliggs 173, 230 skildus 70 spaíkulatur 203 stqfs/stabeis 256i., 382f, 388 staua(stöls) 318 stikls 171, 173 j/07'arc 45, 318

swartizl 263, 267 swnagogafaþs 180

425

426 G O T H I C (cont.) *tibr 22f. (ga)timreins 320 tiuhan 235 peihs 35 þiuda 79, n o , 124fr. þiudangardi 126, 173 þiudariön 126 þiudans 79, n o , 123, 124fr, 134fr, 150 þiudinassus 126 þvsundifaþs i8of. unkja 229 unsibjis 60 usjulljan 317 usfitllnan 317 usgáisiþs 355 usgeisnan 355 usháitan 73i. usláubjan 288 wadi 197 wairþ(s) 53 wardja i9if., 203 wflTgy 173 warjan 161 weihs 319, 353, 360, 361, 365 Hfcw 2, 173, 204, 207, 208, 213, 229 weinagards 173 wepna 1, 156 wöþs 79 writs 257

Index o f w ords M ID D L E H IG H G E R M A N Bêheim 161 gebeize 199 hraft 80 künec 131 Āft 151i. 209 phersich 211 phlücken 211 sam(e)ztag 252 sun(nen)ābent 252i. ungezibere/unzifer 23 vrîtuom 42 vrôndienest 105 vrôngerihte 105 vrônreht 105 vrônveste 105 vrônwalt 105 Wto£248, 249f., 251

M ID D L E L O W G E R M A N dïngesdag 250 gevadder 334 glar/glâr 187 gudensdach 248, 251

M ID D L E D U T C H aeImoes(e)ne 306

kerstenen 333 klocke 327i.

black 268

mag 58

ink(e)t/ink 267

pade 334

kerstenen 333

saterdach 252 sepe 188

om 54 pntfzč 58 pete(r) 334 saterdach 252 vridach 248 wôdensdach 248, 251

O L D E N G L IS H