Old English Runes: Interdisciplinary Perspectives on Approaches and Methodologies with a Concise and Selected Guide to Terminologies 9783110796834, 9783110796773

Citation preview

Old English Runes

Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde Herausgegeben von Sebastian Brather, Wilhelm Heizmann und Steffen Patzold

Band 134 Runische Schriftlichkeit in den germanischen Sprachen Herausgegeben von Edith Marold Im Auftrag der Akademie der Wissenschaften zu Göttingen

Band 4

Old English Runes

Interdisciplinary Perspectives on Approaches and Methodologies with a Concise and Selected Guide to Terminologies

Edited by Gaby Waxenberger, Kerstin Kazzazi and John Hines with the Assistance of Kerstin Majewski

Collected in this volume are papers arising from the Module 1 conference in the long-term project “Runic Writing in the Germanic Languages/Runische Schriftlichkeit in den germanischen Sprachen (RuneS)”, held in Eichstätt, 15–16 March 2012.

ISBN 978-3-11-079677-3 e-ISBN (PDF) 978-3-11-079683-4 e-ISBN (EPUB) 978-3-11-079690-2 ISSN 1866-7678 Library of Congress Control Number: 2022944560 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the internet at http://dnb.dnb.de. © 2023 Walter de Gruyter GmbH, Berlin/Boston Typesetting: Meta Systems Publishing & Printservices GmbH, Wustermark Printing and binding: CPI books GmbH, Leck www.degruyter.com

In Memory of

Ray I. Page 25 September 1924 – 10 March 2012

Klaus Düwel 10 December 1935 – 31 December 2020

Hans Frede Nielsen 20 May 1943 – 9 January 2021

Contents List of Abbreviations List of Figures

xiii

List of Tables

xvii

List of Maps

ix

xix

Gaby Waxenberger, Kerstin Kazzazi, and John Hines 1 Introduction Alfred Bammesberger The Ruthwell Crucifixion Poem Revisited

21

Catherine Hills Early Anglo-Saxon Runic Pots at Spong Hill, Norfolk, England

45

John Hines The Archaeological Dating of the Early Finds: How Certain are the Results? Lilla Kopár Inscriptions on Stone Monuments: Methodological Concerns

59

81

Tineke Looijenga Frisian Runes Revisited and an Update on the Bergakker Runic Item Regine Marth The Mysterious Gandersheim Casket: Are There Any Hard Facts?

103

127

Tim Pestell Runic Finds from the Kingdom of East Anglia and Their Archaeological 137 Contexts Christopher Scull The adventus Saxonum from an Archaeological Point of View: How Many Phases 179 Were There? Theo Vennemann The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

199

viii

Contents

Kerstin Kazzazi Introductory Remarks on the Contributions by Leslie Webster and Gaby 221 Waxenberger on the Franks Casket Leslie Webster A. The Dating and Provenance of the Franks Casket: An Art-Historical 225 Perspective Gaby Waxenberger B. The Franks Casket and its Inscriptions

253

Gaby Waxenberger C. The Dating and Provenance of the Franks Casket: A Linguistic and Runological 267 Perspective Gaby Waxenberger, Kerstin Kazzazi, and John Hines 299 A Concise and Selected Guide to Terminologies Index of Objects

331

Index of Persons

333

Index of Place-Names

335

List of Abbreviations Dictionaries, Grammars, and Editions BT

Chron Gall 452

DEB DOE Online

DOEW

HE IK RGA SB Wars

Bosworth/Toller: An Anglo-Saxon Dictionary. Eds. Joseph Bosworth and T. Northcote Toller. Oxford: Clarendon, 1898 [many reprints]. An Anglo-Saxon Dictionary: Supplement. Ed. T. Northcote Toller. Oxford: Clarendon, 1921 [many reprints]. “Gallic Chronicle of 452.” Ed. and transl. Alexander Callander Murray, 1999. From Roman to Merovingian Gaul: A Reader. Ed. Alexander Callander Murray. Toronto: University of Toronto Press. 76‒85. Gildas: The Ruin of Britain and Other Documents. Ed. and transl. Michael Winterbottom. London/Chichester: Phillimore, 1978. 13‒79; 87‒142. Dictionary of Old English: A to I Online. Eds. Angus Cameron, Ashley Crandell Amos, Antonette diPaolo Healey et al. Toronto: Dictionary of Old English Project, 2018. URL: ‹www.doe.utoronto.ca/›. The Dictionary of Old English Web Corpus. Compiled by Antonette diPaolo Healey with John Price Wilkin and Xin Xiang. Toronto: Dictionary of Old English Project, 2009. URL: ‹www.doe.utoronto.ca/›. Bede’s Ecclesiastical History of the English People. Eds. and transl. Bertram Colgrave and Roger A. B. Mynors. Oxford: Oxford University Press, 1969. Ikonographischer Katalog [individual volumes are listed in the references of the respective contributions]. Reallexikon der Germanischen Altertumskunde [individual volumes are listed in the references of the respective contributions]. Brunner, Karl. 1965. Altenglische Grammatik: Nach der angelsächsischen Grammatik von Eduard Sievers. Dritte, neubearbeitete Auflage. Tübingen: Niemeyer. Procopius, with an English Translation by H. B. Dewing. 7 vols. Ed. and transl. Henry Bronson Dewing. Cambridge, MA: Harvard University Press. Vols. 1‒5, 1914‒1940.

Languages and Dialects Angl. Gmc. Goth. IE Merc. ModG Nhb. OHG OE OFris. ON OS Pre-OE W WS

Anglian Germanic Gothic Indo-European Mercian Modern German Northumbrian Old High German Old English Old Frisian Old Norse Old Saxon Pre-Old English WestWest Saxon

https://doi.org/10.1515/9783110796834-201

x

List of Abbreviations

Others AD AE apn. b. BC BE BM c., ca. c., cent. cat. cf. Chap. d. dsm. ed(s). e.g. EpGl f f., ff. FC Fig(s). fn. fol. forthc. gpn. gsm. HER i.e. Ill. inv. m NCM no. nsf. nsm. OERC p., pl. PAS pers. Pl. Pre-OERC pret. RC RC SE RC SW RC NE RC NW s., sg.

anno Domini American English accusative plural neuter born before Christ British English British Museum circa century catalogue confer; compare Chapter died dative singular masculine editor(s) exempli gratia; for example Épinal/Erfurt Glossaries feminine following Franks Casket figures footnote folio forthcoming genitive plural neuter genitive singular masculine Norfolk Historic Environment Record id est; this means Illustration inventory metre Norwich Castle Museum and Art Gallery number nominative singular feminine nominative singular masculine Old English Runes Corpus plural Portable Antiquities Scheme person plate Pre-Old English Runes Corpus preterite Ruthwell Cross Ruthwell Cross South-East Ruthwell Cross South-West Ruthwell Cross North-East Ruthwell Cross North-West singular

Symbols

str. s. v. transl. wk. vb. vol(s).

strong sub voce; see under translated, translation weak verb volume(s)

Symbols * → > < ‹› // []

1. reconstructed form 2. non-existent form see/confer developed into developed from grapheme phoneme allophone

xi

List of Figures The Ruthwell Crucifixion Poem Revisited Figure 1: The Ruthwell Cross in Stephens (1866, plate facing 405) 22 Figure 2: The runes on the Ruthwell Cross’s lower stone in Brown (1921, 204) Figure 3: The runes on the Ruthwell Cross’s lower stone in Dickins/Ross (1963) Figure 4: Line 7 of The Ruthwell Crucifixion Poem with shaded area representing a gallows 37

26 27

Early Anglo-Saxon Runic Pots at Spong Hill, Norfolk, England Figure 1: Location map of Spong Hill; © Cambridge Archaeological Unit 53 Figure 2: Runic pots C1224, C1564; drawings: C. Hills 54 Figure 3: Runic stamp on C1224; © Norwich Castle Museum; photo: M. Dabski 54 Figure 4: Runic stamp on C1564; © Norwich Castle Museum; photo: M. Dabski 54 Figure 5: Distribution of runic and related pots at Spong Hill; © Cambridge Archaeological Unit 55 Figure 6: Pot with bossed t, C2009; drawing: K. J. Penn 55 Figure 7: Pot with linear t, C2451; drawing: K. J. Penn 56 The Archaeological Dating of the Early Finds: How Certain are the Results? Figure 1: The production of a radiocarbon date from a radiocarbon age by calibration 63 Inscriptions on Stone Monuments: Methodological Concerns Figure 1: (Great) Urswick 1A; © Wardens of Urswick Church; photo: L. Kopár 95 Figure 2: (Great) Urswick 1A, detail; © Wardens of Urswick Church; photo: L. Kopár Figure 3: (Great) Urswick 1C; © Wardens of Urswick Church; photo: L. Kopár 98

96

Frisian Runes Revisited and an Update on the Bergakker Runic Item Figure 1: The Bergakker Scabbard Mount, front; © Collection Valkhof Museum. Photograph by Thijn van de Ven 107 Figure 2: Inscription on the back of the Bergakker Scabbard Mount. The Bergakker Scabbard Mount is part of the Collection Valkhof Museum, Nijmegen; © Collection Valkhof Museum. Photograph by Stephan Weiss-König 107 Figure 3: Grave 833 from the Donderberg, Rhenen (Wagner/Ypey 2011, 602) 113 Figure 4: Schweindorf solidus. The runes run from right to left and read: welad; © Ostfriesisches Landesmuseum Emden; photo: C. Kohnen, Ostfriesische Landschaft Aurich 118 The Mysterious Gandersheim Casket: Are There any Hard Facts? Figure 1: The Gandersheim Casket, Braunschweig, Herzog Anton Ulrich-Museum, inv. no. MA 58; © Herzog Anton Ulrich-Museum, Braunschweig 128 Figure 2: The Gandersheim Casket, front; © Herzog Anton Ulrich-Museum, Braunschweig 128 Figure 3: The Gandersheim Casket, back; © Herzog Anton Ulrich-Museum, Braunschweig 128 Figure 4: The Gandersheim Casket, bottom with inscription; © Herzog Anton Ulrich-Museum, Braunschweig 132 Figure 5: The Gandersheim Casket, bottom frame from the inside; © Herzog Anton Ulrich-Museum, Braunschweig 132 https://doi.org/10.1515/9783110796834-202

xiv

List of Figures

Figure 6:

The Gandersheim Casket, right side; © Herzog Anton Ulrich-Museum, 134 Braunschweig The Gandersheim Casket, left side; © Herzog Anton Ulrich-Museum, 134 Braunschweig

Figure 7:

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts Figure 1a: Caistor St Edmund roe-deer astragalus; © Norwich Castle Museum and Art Gallery 140 141 Figure 1b: Lackford 2987 and Ingham 1038 urns, after Myres (1977) Figure 1c: Spong Hill pot stamp; © Norwich Castle Museum and Art Gallery 142 Figure 2a: The Harford Farm (= Caistor-by-Norwich) brooch; © Norwich Castle Museum and Art Gallery 145 Figure 2b: Close-up of the runic inscription on the Harford Farm brooch reverse 145 Figure 3a: The Binham bracteate; © Norwich Castle Museum and Art Gallery 147 Figure 3b: Detail of the runic legend on the Binham bracteate; © Norwich Castle Museum and Art Gallery 147 Figure 4a: The Undley bracteate; © The Trustees of the British Museum 149 Figure 4b: Line drawing of the Congham sword pommel by Jason Gibbons; © Norfolk Historic Environment Service 149 151 Figure 5: The Tiluwald sceatta; © Norwich Castle Museum and Art Gallery Figure 6a: Runic-inscribed tweezers from Heacham; © Norwich Castle Museum and Art Gallery 152 Figure 6b: The Baconsthorpe tweezers; © Norwich Castle Museum and Art Gallery 154 Figure 7: The Blythburgh tablet, back; © The Trustees of the British Museum 155 Figure 8a: The Shropham runic sheet, obverse; © Norwich Castle Museum and Art 157 Gallery Figure 8b: The Shropham runic sheet, reverse; © Norwich Castle Museum and Art 158 Gallery 161 Figure 9a: The ‘March’ lead sheet, obverse; © Norwich Castle Museum and Art Gallery 161 Figure 9b: The ‘March’ lead sheet, reverse; © Norwich Castle Museum and Art Gallery Figure 10a: The St Benet’s Abbey lead sheet, obverse; © Norwich Castle Museum and Art 163 Gallery Figure 10b: The St Benet’s Abbey lead sheet, reverse; © Norwich Castle Museum and Art 163 Gallery 167 Figure 11: The Keswick runic disc; © Norfolk Historic Environment Service 168 Figure 12: The ‘Reedham’ runic spindle whorl; photo: P. Murawski The Runic Inscription skanomodu: Frisian or Anglo-Frisian? Figure 1: The skanomodu solidus; © The Trustees of the British Museum, London The Franks Casket, Part A: The Dating and Provenance of the Franks Casket: An Art-Historical Perspective Figure 1a: Decoration on the entrance porch of St Peter’s Church, Monkwearmouth; 226 photo: J. Hines Figure 1b: Decoration on the entrance porch of St Peter’s Church, Monkwearmouth; 226 © G. Speake Figure 2: St. Cuthbert’s coffin, symbols of Sts Matthew, and Mark (Battiscombe 1956, 226 Plate VII) Figure 3: The Cuthbert Gospel of St John; © London, British Library, Add. MS 89000, 227 Front Cover

200

List of Figures

xv

Figure 4:

St John image, Lindisfarne Gospels, London, British Library, Cotton MS Nero D. IV f.209v; © The British Library Board 228 Figure 5: Depiction of the Tabernacle, Codex Amiatinus, Florence, Biblioteca Medicea Laurenziana, MS Amiatino 1, f. fols. IIv‒IIIr; reproduced with permission of MiBACT; further reproduction by any means is prohibited. 229 Figure 6: Jerome’s Prologue, Codex Amiatinus, Florence, Biblioteca Medicea Laurenziana, MS Amiatino 1, f. fol. IVr; reproduced with permission of MiBACT; further reproduction by any means is prohibited. 230 Figure 7: Decorated initial with the image of Gregory the Great, The St Petersburg Bede, MS Lat. Q. v. I. 18 fol. 26; © The National Library of Russia, St. Petersburg 231 Figure 8: The Ruthwell Cross; © Historic Environment Scotland; photo: K. Majewski 233 Figure 9: Carpet page, Book of Durrow, Trinity College Dublin Library MS 57, fol. 192v; © The Board of Trinity College Dublin 235 Figure 10: The Brescia Casket (Lipsanotheca), Museo di Santa Giulia, Brescia; © Photographic Archive of the Brescia Museums, Fotostudio Rapuzzi 236 Figure 11: (a) Carpet page animal, Book of Durrow, Trinity College Dublin Library MS 57, fol. 192v; (b) animal from initial, Corpus-London Gospels, Corpus Christi College Cambridge, MS 197B, fol. 2; (c) animal from initial, Durham Gospels, Durham Cathedral Library, MS A. II.17, fol. 2r; drawing: M. O. Miller 238 Figure 12: Symbol of St Luke, Echternach Gospels, Paris, Bibliothèque nationale de France, MS Latin 9389, f. 115r; © Bibliothèque nationale de France, Paris, Département des Manuscrits 240 Figure 13: Die Stamp from Fen Drayton, Cambridgeshire; © Portable Antiquities Scheme: Record, ID: NLM-468D41 (CC BY-SA 4.0) 241 Figure 14: Cross-shaft from Aberlady, Midlothian; © Trustees National Museums Scotland; photo: A. Maldonado 243 Figure 15: Decorated initial, St Petersburg Bede, MS Lat. Q. v. I. 18. f. 3v; © The National Library of Russia, St. Petersburg 244 Figure 16: Decorated initial, Salaberga Psalter: Staatsbibliothek, Berlin, MS Hamilton 553, fol. 2; © bpk, Staatsbibliothek zu Berlin, C. Seifert 247 The Franks Casket, Part B: The Franks Casket and its Inscriptions Figure 1: Full view of the Franks Casket; © The Trustees of the British Museum 253 Figure 2: The fragmentary Lid of the Franks Casket; © The Trustees of the British Museum 254 Figure 3: The Front Panel of the Franks Casket; © The Trustees of the British Museum 255 Figure 4: Top line of Back Panel; © The Trustees of the British Museum 256 Figure 5: Top line of Front Panel; © The Trustees of the British Museum 257 Figure 6: The Left Panel of the Franks Casket; © The Trustees of the British Museum 261 Figure 7: The Back Panel of the Franks Casket; © The Trustees of the British Museum 262 Figure 8a: The Right Panel of the Franks Casket; © The Trustees of the British Museum 263 Figure 8b: The original fragment of the Right Panel in the Museo Nazionale del Bargello, Florence, Italy; © Museo Nazionale del Bargello, Florence 264 The Franks Casket, Part C: The Date and the Provenance of the Franks Casket: A Linguistic and Runological Perspective Figure 1: Derby(shire) Bone Plate; © The Trustees of the British Museum 290 A Concise and Selected Guide to Terminologies Figure 1: The Pre-Old English vowel chart 302 Figure 2: The oral cavity (after Davis 1998, 29) 304

xvi

List of Figures

Figure 3: Figure 4: Figure 5:

The Common Germanic fuþark and Pre-Old English innovations The Old English fuþorc in ca. AD 650 and 750 307 The Germanic language family 309

306

List of Tables The Archaeological Dating of the Early Finds: How Certain are the Results? Table 1: Female graves or artefact-types with runic inscriptions dated by the national chronological scheme, cited at 95 % probability 67 Table 2: Male graves or artefact-types with runic inscriptions dated by the national chronological scheme, cited at 95 % probability 69 The Franks Casket, Part C: The Date and the Provenance of the Franks Casket: A Linguistic and Runological Perspective Table 1: Napier’s criteria for dating the casket 269 Table 2: The distribution of unstressed -æ and -i vs. -e 269 Table 3: Attested forms of -eo- (WGmc. *eu) in the OERC 271 Table 4: Attested distribution of the runic graphemes ‹f› and ‹b› representing the allophones [f], [v], and [b] 272 Table 5: Napier’s (1901, 379 f.) dialect criteria 274 Table 6: Anglian smoothing of eo/ēo 275 Table 7: Lack of diphthongisation after palatals: æ vs. ea 276 Table 8: Parasite vowels in the OERC 277 Table 9: Gmc. *a + nasal > early Nhb. ‹a›, ‹o› 280 Table 10: The clusters [xt] and [çt] within the OERC 281 Table 11: Rune ġiefu vs. rune gār 284 Table 12: Rune ġiefu ᚷ vs. the yew-rune ᛇ 286 Table 13: Occurrence of u-umlaut of e > eo vs. lack of o/a-umlaut of e > eo 286 Table 14: Lack of u-umlaut 287 Table 15: Rune no. 28 )ea ᛠ 289 Table 16: The Franks Casket and the Ruthwell Cross 290 Table 17: Runes o, u, y, w, and c on the Whitby Comb and the Franks Casket; Whitby Literary and Philosophical Society, Whitby Museum 291 A Concise and Selected Guide to Terminologies Table 1: Transliteration symbols used by Waxenberger (forthc.)

https://doi.org/10.1515/9783110796834-203

319

List of Maps Introduction Map 1: Findspots of the Pre-Old English inscriptions; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München 3 Map 2: Findspots of the Old English inscriptions; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München 4 Frisian Runes Revisited and an Update on the Bergakker Runic Item Map 1: Frisia in the seventh/eighth century, from Gerrets (2010, 3); © Daniël A. Gerrets 105 Map 2: The ‘Frisian’ and ‘Frankish’ runic items, part of a Continental North Sea group (taken from Looijenga 2021, Fig. 13.1, p. 377) 109 Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts Map 1: The kingdom of East Anglia 139 Map 2: Caistor St Edmund showing evidence for Anglo-Saxon cemeteries and settlement between the fifth and ninth centuries 144 The Franks Casket, Part A: The Dating and Provenance of the Franks Casket: An Art-Historical Perspective Map 1: Northumbrian monastic centres mentioned in the text; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München 232 The Franks Casket, Part C: The Date and the Provenance of the Franks Casket: A Linguistic and Runological Perspective Map 1: Runic inscriptions in Northumbria (late 7th–10th/11th cent.); map base: © EuroGeographics, linguistic details: RuneS research unit EichstättMünchen 268 Map 2: Inscriptions applying the yew-rune in the Pre-OE and OERC (fricative, vowel, rune no. 13 as a fuþorc-unit) 282 Map 3: Rune h used in the cluster ‹ht› in the OERC; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München 283 Map 4: Attested distribution of the runes ġ(i)efu ᚷ and gār ᚸ to denote [ɤ] initially, medially and finally; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München 285 Map 5: Attested distribution of rune no. 6 ċēn ᚳ; and its variants ᛷ, in inscriptions proper for the plosive allophones [k] and [ḵ]; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München 293 Map 6: Attested distribution of ċēn ᚳ [k]; [ḵ], calc ᛣ [k]; [ḵ], and rune no. 31 ᛤ [ḵ]; map base: © EuroGeographics, linguistic details: RuneS research unit EichstättMünchen 294 A Concise and Selected Guide to Terminologies Map 1: Old English dialect areas; map base: © EuroGeographics, linguistic details based on SB (1965): RuneS research unit Eichstätt-München 314

https://doi.org/10.1515/9783110796834-204

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Introduction This volume presents contributions to the conference Old English Runes Workshop,1 organised by the Eichstätt-München Research Unit of the Academy project Runische Schriftlichkeit in den germanischen Sprachen (RuneS) and held at the Catholic University of Eichstätt-Ingolstadt in March 2012. The conference brought together experts working in an area broadly referred to as Runology. Scholars working with runic objects come from several different fields of specialisation – often characterised as separate scholarly “disciplines” – and the aim was to provide more mutual insight into the different methodologies and theoretical paradigms used in these different approaches to the study of runes or, in the present instance more specifically, runic inscriptions generally assigned to the English and/or the Frisian runic corpora. Success in that aim should automatically bring with it the reciprocal benefit of improving access to and understanding of the runic evidence, expanding and enhancing insights gained within such closely connected areas of study of the EarlyMedieval past.

1 Methodology Gaby Waxenberger As a linguistic runologist specialised in English Runology, I realised the limits of historical linguistics very soon and it has been my aspiration to work in an interdisciplinary context with colleagues from different fields. Admittedly, historical linguistics can contribute to deciphering and interpreting the inscriptions, it can give us an insight into sound changes and language change; its methodology and approaches made it possible for me to realise that the corpus formerly known as English inscriptions consists, in fact, of two corpora, the Pre-Old English (Pre-OE) corpus (ca. AD 425–650) and the Old English (OE) proper. The Pre-OE inscriptions were still written in a rune-row closer to the Older → fuþark. In the course of the Pre-OE period, some of the vowels changed their qualities and also new vowels emerged, the result being an enlarged rune-row in ca. AD 650, the beginning of Old English. In this period, the transitory rune-row, the → Pre-fuþorc, developed to the Old English → fuþorc of 27 characters by ca. AD 650. However, the Old English fuþorc of 27 characters was again enlarged to 31 characters in the west of Northumbria by ca. AD 750. Historical → phonology and graphemics provided the tools

1 The contribution by Theo Vennemann was included additionally due to its thematic scope. https://doi.org/10.1515/9783110796834-001

2

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

for detecting a difference in the use of certain runes for the same phoneme(s) in the east and in the west of Northumbria. On the basis of historical phonology and graphemics, inscriptions can thus be assigned to certain dialect areas and certain periods (Waxenberger, this volume), but this is only possible if the linguistic data available is sufficient. For a more valid picture, as in the case of the Franks Casket, a second discipline, in this case, history of art, is needed (Webster, this volume; see also below). Only if a second, independent discipline corroborates the results of the first, can we be more confident regarding the reliability of our interpretations of the available data. Where the methods of individual disciplines reach their limits, a joint venture of more, sometimes neighbouring, disciplines can go beyond the last frontier. However, such a concerted action requires mutual understanding. This not only refers to the methodology, possibilities and limits of the other disciplines, but first and foremost, their frame of thinking and, more specifically, the unfamiliar terminology. For this reason, the editors have added A Concise and Selected Guide to Terminologies to this volume. We have attempted to explain some important terms in Linguistics, Archaeology and Art History in order to give the reader not at home in all of these disciplines the opportunity to look them up. A comprehensive guide written for the non-expert would be the first step towards fruitful interdisciplinary cooperation. Successful interdisciplinary work also requires a different format for future conferences: time should be invested in the explanation of methods and methodology to the non-expert. Hitherto there are no slots for such an undertaking, although it would be a good investment. Future research and autopsies in runology should be carried out by interdisciplinary teams so that all aspects of an object and its inscription could be taken into account, avoiding pitfalls and therefore leading to an overall and, more importantly, more reliable result. As a result of my research, we can provide two maps for the Old English Runes Corpus (OERC): one for the corpus of the Pre-OE inscriptions (see Map 1) and one for the corpus of the OE inscriptions proper (see Map 2). As a preliminary step in the direction of interdisciplinary work and above all, mutual understanding, it was our aim to focus on the methodological perspective in our workshop. Therefore, we issued the following basic questions or principles to the invited speakers of the conference: Your paper should consider/explain/make transparent the following questions: 1. Methodological approach should become clear to non-experts in your field. 2. The way to your result/conclusion should be understandable and transparent to non-experts in your field. 3. Your conclusion/result is valid within certain parameters.

Introduction

Map 1: Findspots of the Pre-Old English inscriptions; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München; unprovenanced: skanomodu solidus.

3

4

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Map 2: Findspots of the Old English inscriptions; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München.

Introduction

5

2 Runology – an “Interdiscipline”: Disciplinary and Methodological Perspectives Kerstin Kazzazi Taking up the description by Gaby Waxenberger in the preceding section on what working on runic inscriptions encompasses, in the following paragraphs, I would first like to briefly dwell on the term runology, taking a – non-exhaustive – look at what it has been understood to cover and why, as well as some implications of its possible definition also for further research. This will be followed by a short characterisation of the contributions concentrating on specific theoretical and/or methodological concepts used. What do scholars of runology deal with, how do they do it and what is their particular focus? In recent times, this complex question has increasingly been asked by runologists themselves in an attempt to define, first, the range of perspectives on the runic script as well as, second, the methodology their work should be based on. Michael Barnes (2013, 9–11) addressed these issues at some length in Oslo in 2010 and with reference to other runologists’ thoughts on it, such as Terje Spurkland, Judith Jesch, Ray Page and others. They were also dealt with at the “RunenWorkshop” in Schleswig in 2011 (Grimm/Pesch 2015, 11) and included in a contribution by Christiane Zimmermann (2017) as well as by Livia Kaiser in her dissertation on the so-called Frisian Runic Corpus (Kaiser 2021). It seems no coincidence that this question is becoming increasingly prominent after decades of academic work on the runes, as the methodologies in different disciplines have changed and become more specialised (Grimm/Pesch 2015, 11; Kaiser 2021). In fact, the question of what runologists do carries the seeds of the inherently interdisciplinary nature of runology that make it reach out further than its positioning within linguistic subfields such as comparative Germanic philology (Vergleichende germanische Philologie), historical linguistics, semantics, epigraphy, grapholinguistics (in particular graphemics) or linguistic pragmatics (cf. Zimmermann 2017, 432). For the perspective on the runic inscriptions proceeding from the script itself means that not only the language, or more precisely the individual text underlying this script, is taken into focus, but also the material side of the inscription, i.e., the inscribed object and the writing tools, along with the whole setting in which the object was found, and the historical and cultural context (of writing) which this implies (Page 1999, 7). In practical research, this necessarily interdisciplinary perspective has not always been duly observed, as the following remark by the late Hans Frede Nielsen (1996, 128) regarding the so-called Frisian Runic Corpus shows: modern runic scholars have not even all been fully aware of the hotchpotch of geographical, archaeological, numismatic, runological and linguistic criteria underlying the purported Frisian runic corpus.

6

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

An integral view of the runic inscription explicitly taking this interdisciplinary stance by proceeding from the runic inscription as a specimen of language influenced by and embedded in a web of extralinguistic contextual features has been discussed in detail by Christiane Zimmermann (2017, with ample reference to the relevant literature). This leads to close involvement of disciplines other than purely linguistic ones. The following list of disciplines and research fields involved in runological research is not exhaustive and may be extended with respect to specific research questions or individual inscriptions; it is meant here as a backdrop for highlighting the different fields from which the contributions to this volume come: 1. (Historical) Linguistics (analysis of the language(s) written in this script and of the dynamic relationship, e.g., between script and sound through time and space) as exemplified in the contributions by Bammesberger, Vennemann and Waxenberger, in part also Looijenga, this volume. 2. Archaeology (in particular, for dating through contextualisation of individual finds or the general historical-cultural background of runic objects); cf. the contributions by Hines, Looijenga, Pestell, Scull, Hills, this volume, 3. Art History (for dating of the object and the analysis of possible (inter-)relations between text and iconographical elements), as illustrated by the complementary contributions Webster (art-history) and Waxenberger (historical linguistics) in this volume, discussing the same object, the Franks Casket; see also the contribution by Marth. 4. In particular a more recent joint perspective on → material culture and language (taking into explicit account the specific relationship between material aspects, inscription, and context of the inscribed object), as outlined by Kopár, this volume. Other fields, within linguistics and other disciplines, both more general and highly specialised, are: 5. Epigraphy (for comparison with other epigraphically attested “Trümmersprachen” such as various Continental Celtic ones). 6. History (for relating inscriptions to historical facts, either contextually or by textual reference; and for general historical contextualisation of the finds).2 7. Numismatics (e.g., for the contextualisation (dating) of runic finds that either are or derive from coins, such as the skanomodu solidus). 8. Literacy Studies (for looking at runic inscriptions as specimens of written language from the perspective of recent theoretical models of literacy and writing).

2 Calls for interdisciplinary work on runic inscriptions have also come from scholars in Medieval Studies, cf. Izzi (2014, 146) with respect to the graffiti in the catacombs in Rome.

Introduction

9.

7

Sociolinguistics (for placing the runic documents in their socio-cultural setting), also in historical contexts, as Looijenga attempts to do in her contribution in this volume.3

All the above-mentioned disciplines have long been working together on the runes, as the late Klaus Düwel states in his seminal introduction Runenkunde (2008, 4), but recently the need for explicit reflection on what this entails both at a theoretical and at a methodological level has been growing (cf. Grimm/Pesch 2015; Zimmermann 2017; Kaiser 2021). The increase of active interdisciplinary co-operation can be seen by calls for this kind of exchange (cf. Düwel 1986, 18, recently re-iterated as a quote in the preliminary remarks to the new edition of the South Germanic runic inscriptions in Düwel/Nedoma/Oehrl 2020, v; cf. Barnes 1994, 27) being superseded by positive recognition of the same actually taking place (e.g., Quak 2001, 331). What is still needed is a more concrete explication of what this co-operation means in practice, i.e., who can contribute what information with what kind of methods to the study of runic inscriptions. A first attempt at such a formulation can be found in Zimmermann (2017, 436 f.). The following complex working definition, bringing together narrow and broad definitions from the literature, may be helpful (cf. also Kaiser 2021): 1. Runology in a narrow sense is rooted in the discipline of linguistics or philology (cf. Barnes 2013, 9–10), but 2. an intrinsic characteristic is its interdisciplinary nature, which implies that, proceeding from the perspective of the script in its relation to the languages represented, more than purely linguistic knowledge has to be taken into account, meaning that linguistic runologists necessarily need to work with other, nonlinguistic disciplines.4 3. This leads to a broader conception of runology (cf. Düwel 2004, 140) and runologists as encompassing the entire scholarly community working together on understanding the complex issue of rune-inscribed objects. It was the intention to provide scope for precisely this kind of methodological and interdisciplinary reflection by bringing together scholars from the fields mentioned above to present and exchange their approaches to runic objects regarding the different methods and concepts used that lay at the heart of the Old English Runes Workshop. For a fundamental problem encountered, but rarely explicitly referred to in interdisciplinary co-operation (recently see Kaiser 2021), is that disciplines dif-

3 More intra-disciplinary co-operation within linguistics, e.g., with the sub-disciplines of typology, multilingualism research, language contact studies, sociolinguistics, has also been demanded by some (cf. Braunmüller 1995, 11). 4 This distinguishes runology from historical linguistics as such, where the script tends to be more of a vehicle for rather than an object of study.

8

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

fer not only in their approaches and methodologies, but also in the theoretical frameworks, concepts and terminologies used. The following short characterisations of the individual contributions highlight some of these concepts specifically for colleagues “from the other side”, i.e., the linguistic concepts are introduced not for linguists, but for archaeologists, art historians and historians, whereas the archaeological terminology and patterns of thought are set out for linguists. Those concepts and terms mentioned here that are further elucidated in a Concise and Selected Guide to Terminologies at the end of this volume are marked in bold and by an arrow →. An Index of Objects makes it possible to obtain a synoptic overview of the current and in some cases contrasting analyses and opinions about a specific object, e.g., the skanomodu solidus, discussed in several contributions.

3 Characterisation of the Contributions with a Focus on Theoretical and Methodological Concepts 3.1 Contributions Drawing on Linguistic Theoretical and Methodological Frameworks Kerstin Kazzazi In the case of Linguistics, it is the particular perspective on language as a system in relation to the speakers using this system that leads to terms from the general vocabulary being used in a far more restricted way. This may turn out to be difficult in two ways: 1. The same term, e.g., a language label, may be used for slightly or even widely differing concepts (for instance the language label → (Anglo-)Frisian, see Looijenga and Vennemann, this volume); 2. completely different, discipline-specific concepts may be used that are not immediately accessible to non-specialists, even if they outwardly deal with the same objects, here runic inscriptions; examples of such specialist terms from linguistics include the concepts of → phoneme, → grapheme, → soundchange, → reconstruction. The volume contains three linguistic contributions (Bammesberger, Vennemann, Waxenberger) as well as two which are in themselves more interdisciplinary, including both linguistic and non-linguistic perspectives (Looijenga, Kopár). In the following, an attempt will be made to highlight some of the main methods or theo-

Introduction

9

retical concepts drawn on in each. The contributions proceed from quite different questions and thus employ different linguistic methods: Bammesberger deals with one specific inscription, namely the Ruthwell Cross inscription, in a philological framework, comparing it to another, thematically related text, the late Old English poem The Dream of the Rood in the Vercelli Book. He comes to the conclusion that, contrary to the communis opinio, “the Ruthwell Crucifixion Poem may be considered as a complete poetic work. The Vercelli version is then likely to contain extensions and secondary additions” (p. 29). He derives this result from intertextual comparison of the content, metre, style and grammar of the two texts, thereby philologically reconstructing what may have been the original textual form of the Ruthwell Cross runic inscription. This method of philological → reconstruction differs from what is termed comparative reconstruction in historical phonology. The latter theoretical-methodological framework underlies the basic approach of the contribution by Vennemann. If we have a specific documented form of which we do not know the meaning, we may use our knowledge about certain → sound changes to reconstruct an earlier form of the word or morpheme, and compare this to possibly cognate forms in related languages with meanings compatible with the context the form is used in; the result of this method is called an → etymology. The difficulties involved in arriving at an indisputably correct etymology for a documented form are demonstrated by Vennemann, who discusses the notorious skanomodu solidus. While it seems clear that the sound represented by the new rune ᚪ a āc on the skanomodu solidus is long /a:/, the origin of this sound and therefore the language attribution of this inscription remain controversial, as the provenance of the object cannot be determined (cf. recently Waxenberger 2017). Due to this new rune occurring on the Harford Farm brooch and the Pre-Old Frisian inscriptions, the coin has been classified as both English and Frisian by various scholars, with preference mainly being given to a Frisian origin (cf. Kaiser 2021). Linguistically, both seem possible. The traditional etymology of the element skaninterprets the sound in question as the result of a specifically Pre-Old Frisian sound change */au/ > /a:/. Vennemann, on the other hand, proposes an alternative etymology, suggesting that not only regarding the sound shape, but also from the point of view of the word formation process underlying the form, the element skan- may just as well be English, because the /a:/ could also go back Pre-OE/Pre-OFris. */aɪ/, as this diphthong developed to /a:/ in both Frisian and English. The phonological and morphological levels thus remain inconclusive, and the semantic level, i.e., the meaning side of each etymology, does not help in the decision either: both ‘beautiful’ (Pre-OFris. *skaun-i) and ‘shining’ (Pre-OE/Pre-OFris. *skain-i-) would be fitting for the first part of a personal name. However: As no etymology can be found containing an original /a:/ at an earlier language stage, instead of an original diphthong, what does seem clear is that the word skanomodu represents either Pre-Old Frisian or Pre-Old English rather than what is traditionally assumed to have been

10

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

their common precursor Anglo-Frisian.5 So, linguistic analysis gives us information, on the one hand, on what language (stage) the form may be and, on the other, on what it very probably is not. This is all that is possible in a case where essential information such as find-spot and provenance are unknown and comparable objects are lacking. In other cases, there is sufficient information of both linguistic and extralinguistic nature to make a more precise hypothesis possible. The Franks Casket is analysed from two perspectives in this volume: Webster looks at the iconography from an art-historical angle (see below), whereas Waxenberger discusses the complex linguistic material of this object. The most difficult question regarding this object has always been the relation of find-spot and provenance. One way to approach this question is by trying to fit the language used in the inscriptions into the known linguistic landscape. Most of the casket turned up in France, with only one panel having been found in Italy, but the language of the inscriptions is clearly Old English (with some Latin interspersed). By listing a series of typically northern dialectal features such as Anglian smoothing, Waxenberger is able to narrow down the area of provenance not only to northern England but to the east coast of Northumbria (→ Old English dialects). This was possible by the comparison of further dialectal features (such as the development of Germanic *a + nasal consonant, Nhb. retraction of æ > a), the runic representation of the clusters [xt], [çt] on the Franks Casket and, contrastively, on the Ruthwell Cross in the north-west of Northumbria, as well as, for example, the rune-forms on the Whitby Comb and the additional linguistic comparison to a non-runic text with a north-eastern provenance (Cædmon’s Hymn). The casket was most probably made in the early eighth century somewhere on the east coast between Whitby and Wearmouth-Jarrow. This combination of purely linguistic analysis and analysis of the runic script as such (use of specific runes, rune-forms) thus yields quite a clear picture, which in turn is perfectly corroborated by Webster’s conclusions from the art-historical perspective (see below). However, it clearly shows the limits of a single discipline, because Webster’s arthistorical data allows for a larger area as a potential provenance, including the monastery of Lindisfarne, while linguistic evidence is too sparse for Lindisfarne. Thus, it highlights that a second discipline may corroborate the results of the first, even going beyond this point, shedding further light on this matter. Looijenga’s contribution takes up the linguistically-based Anglo-Frisian hypothesis (see above) and looks at it from the archaeological – in particular the material culture – and historical perspectives. Her contribution thus brings together findings and insights from different disciplines and attempts to construct from them a coherent picture supporting the hypothesis not only of a common language stage, but also

5 This is traditionally regarded as an intermediate stage sharing some sound changes such as the so-called Anglo-Frisian Compensatory Lengthening (for more information and an overview see Nielsen 1998, 73–74, and also Kaiser 2021): → Anglo-Frisian.

Introduction

11

of a common Anglo-Frisian runic corpus. For Looijenga claims that those runic inscriptions commonly grouped into the two separate corpora of Old English (or, in more historical terms, → Anglo-Saxon) and Frisian inscriptions may in fact represent a single corpus, which she calls “Anglo-Saxon-Frisian” (ASF). The traditional division of the runic corpus into an Old English one and a Frisian one would then be purely geographical, being due not to the differentiation of the languages through characteristic sound changes, but due to the migration of some groups of speakers of the same language to Britain. Drawing on the runic Bergakker Scabbard Mouthpiece as an example, Looijenga makes an attempt to show how such an “Anglo-Saxon-Frisian” culture would fit into the picture of a region still strongly Romanised and how, in particular, the use of runes could be explained in view of the Latin script being dominant, thereby drawing on the archaeological practice of the identification of cultures (see below). This touches on the sociolinguistic question of the social function a script may have. It is becoming increasingly obvious in research on writing that this function has to be taken into consideration not only in determining the meaning of a written utterance, but also in its linguistic analysis, e.g., in its dating. This point is taken up by Kopár, who also adopts a decidedly interdisciplinary perspective on a specific group of runic inscriptions, namely those on stone sculptures in Britain. She argues for a holistic view on the artefacts, which means that the language written on these objects is only one of their features and as such must be interpreted in relation to all others, such as material, iconography or layout. For instance, the function of an inscription as a memorial message is conveyed not by the linguistic material alone, which may consist only of one (the commemorated) or two (the commemorated plus the commemorator) names, but by the placement of these names on a type of stone monument that in this particular cultural context is identifiable as a memorial stone. This is a (text-) pragmatic approach to the interpretation of linguistic documents, focusing not on their form, but on their function. By regarding the linguistic inscription as one part of a whole, seemingly discrepant results may turn out to reflect an intentional message, such as – possibly – the deliberate use on the Great Urswick Stone of a form of language that is more archaic than the iconography on the object would suggest, in order to tie in with a certain commemorative tradition. Such an insight has a potential impact on the dating of object and inscription.

3.2 Contributions Drawing on Archaeological and Art Historical Frameworks John Hines In the case of Archaeology and Art History, the objects of analysis are usually quite familiar, and it may be suggested that it is not so much arcane or difficult, disciplinespecific methods that need explanation for readers from outside those fields but

12

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

rather their habitual terminology and the patterns of thought that terminology represents. The first section of my own paper (Hines, this volume) is headed “Terminology and Methods”, and it seeks to define Archaeology in respect of its distinct subjectmatter as a basis for an account of chronological methods and taxonomy. Quite rightly, the question of how old an inscribed object is, is both a proper and a natural one for the runologist to ask; what is particularly important to appreciate at the same time, however, is the equally valid question of “how precisely can this be dated?” This leads directly to consideration of the range of different, although usually complementary, dating methods available to the archaeologist. Hills’s paper perfectly exemplifies how long-established methods of the typological classification (→ typology) of artefacts and interpretation of the stratigraphical relationships (→ stratigraphical sequence) between archaeological deposits can be combined with new computerised techniques for exploring correlations between itemised and quantified data; specifically, here, the now widely used method of → correspondence analysis, which can be used to search for sequences of relationship and change within a data-set of objects made up of different combinations of an inventory of “variable” features. Even elaborate techniques which transform objects into lists of their components, numerical counts, and ultimately into coordinates in algebraic space, tend nonetheless to return the data for practical purposes into simple and even quite coarsely bounded groups. At Spong Hill (see Hills, this volume), the sequence of chronological development of the cremation urns is re-grouped as an (overlapping) sequence of Phases A, B, and C, while the national chronological project on Early Anglo-Saxon inhumation grave goods (see Hines, this volume) also resolved its sequences into models composed of four or five lettered phases, for male (M) and female (F) burials respectively. In essentially the same manner, albeit drawing on different techniques, Looijenga refers to Fundgruppen in the Continental chronology of the fifth century produced largely from years of expert study by Horst Wolfgang Böhme and other scholars. European archaeology of the first millennium AD benefits from a long tradition of meticulous chronological analysis that finds expression primarily in elaborate systems of periodisation involving series of phases which may be labelled by numbers, letters, names, or combinations of those. It is often challenging enough for specialists in the discipline to navigate these schemata, let alone for those working in different areas of expertise related to the same general era to be able to interpret and understand the forms of reference concerned. Even more frustratingly, the field is bedevilled by inconsistency in respect of some of the most fundamental descriptive terms. Thus, the “Early Medieval Period” of the fifth to eleventh centuries AD in Britain and Ireland is correspondingly das Frühmittelalter in German but the (prehistoric) Late Iron Age in Scandinavia and the “High” Middle Ages (Haut moyen âge; alto medioevo) in French and Italian archaeology; the Insular High Middle Ages

Introduction

13

primarily denotes the period from the late eleventh to the mid-fourteenth century (up to the impact of the Black Death), followed by the Late Middle Ages; in Germany, conversely that whole period is usually referred to as das Hochmittelalter. Even in British archaeological literature, meanwhile, the phrase “early medieval” is all too often casually used for material of the late eleventh to early thirteenth centuries in contrast to a preceding → “Anglo-Saxon” (or just “Saxon”) Period. One could give more and more examples. Crossovers with historical circumstances introduce further concepts of local application: for instance, a “Conversion Period” – around the seventh century AD – in England, or the “First Viking Age”, largely the ninth century, preceding the hiatus in significant Viking assaults from the 910s to the 980s, that is meaningful in contextualising evidence from East Anglia discussed by Pestell, this volume. There is a crying need for better integration across Europe, but as yet the achievement of greater consistency and clarity can only be a fardistant pipedream. The process of recognising and defining regularly recurrent combinations of variable components such as those which characterise specific periods or phrases has much broader application in Archaeology. Probably its most profound application, although by no means an unchallenged practice, is the identification of “Cultures”. The reality of consistency in the → material culture of particular populations, areas or periods is empirically undeniable; the degree of attention that should be afforded to the observation or enforcement of conventions as opposed to the range of divergency and individuality, however, can be hotly disputed, as is also the interpretative significance attributable to cultural similarity or the sharing of some material-cultural features. In the simplest terms, to what extent does the geographical expansion of specific material-cultural features represent the movement and settlement of people who customarily made use of such features, or other mechanisms of influence involving cross-cultural domination, emulation or trade? These issues lie at the very basis of not only Scull’s review of the concept of the adventus Saxonum as a historical event or sequence of events but also Pestell’s overview of the runic evidence from East Anglia in the context of models of settlement, social evolution, and economic and cultural change over several centuries. Taking an even broader view, yet still with direct and fundamental relevance to the questions explored within this volume, these are critical issues for any discussion which uses, compares and seeks to explain relationships between populations labelled differentially as “Anglo-Saxon” or “Frisian” (see above), and indeed they apply no less in the context of generalised references to Franks, Thuringians, and other contemporary groups. Whatever the causes and mechanisms for it (and they were undoubtedly many and complex), southern Britain underwent massive cultural change in the fifth and sixth centuries AD. The material culture of the Roman imperial provinces of Britannia nearly totally disappears in what can only be characterised as a catastrophic collapse. To the west, north and east alike, it is only a minority of sites that made

14

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

up the previous infrastructure of towns and villas, military forts, and temples that provide evidence of continuing activity, and if so, always at a massively reduced level and involving radical changes of material practice. Nearly three centuries later, in his Historia Ecclesiastica Gentis Anglorum, the Venerable Bede offered an account of when, why, and how Germanic invaders representing three different peoples, the Angles, Saxons and Jutes, came to conquer and settle areas of what would become England. From the mid-nineteenth century onwards, many archaeological scholars have been impressed and convinced by the extent of similarity between what have been called “provinces” of material culture in fifth-/sixth-century England, comparably sized regions on the Continent and in Scandinavia, and Bede’s outline description of areas of origin and settlement; others have simply refused to believe this. Where modern scholarship tends to differ most significantly from earlier interpretations of this evidence is in emphasising the continuously constructed nature of identity and the endless renegotiation of its significance. While, therefore, it does appear quite appropriate to discuss runic literacy in Anglo-Saxon England from the Early Period (fifth to seventh centuries AD) in terms of Anglian and Kentish/Isle of White (interestingly, not “Jutish”) zones in particular, to do so makes no assumptions about ‒ and indeed rarely makes any reference to ‒ obsolete concepts of inherited cultural identity and the determinative force of tradition. A convulsion involving profound systemic collapse and its replacement with new economic and social systems, the infrastructure of which evolved gradually, was by no means limited to the provinces of the Western Roman Empire; indeed the thoroughness of the change in Britannia is matched in no other part of the Empire of comparable size at that date even though all of the provinces of the Western Empire were to pass under the governance of Germanic kingdoms (we might justifiably use the term “warlords”: Berndt 2019) in the course of the fifth and sixth centuries. In the northern Netherlands, beyond the imperial Limes, archaeologists now argue consistently for total depopulation in the fourth century and gradual recolonisation up to a century later (cf. Knol/IJssennagger 2017). Peter Schrijver (2017) has argued persuasively that this sequence also saw the replacement of the formerly Celtic-speaking population with a Germanic-speaking one. Frisian language and material-cultural practices thus were transplanted into and evolved within a new territory very much in the same way and at the same time as the distinctively Anglo-Saxon practices and the “Pre-Old English” language were developing from Germanic roots in southern and eastern Britain. To what extent Frisia received migrant or expanding Germanic settlers and culture from further east in the North Sea zone absolutely in parallel with Britain or was under equal influence from early Anglo-Saxons to the west and their Continental relatives to the east remains a key but open question. It would be misleading, though, to stress only those aspects of the archaeological record of the fifth and sixth centuries that might seem to validate the now oldfashioned view of a Dark Age chaos which fell between the end of the Ancient world

Introduction

15

and the emergence of medieval Europe. In Scandinavia and adjacent areas of the Continental littoral, in the south-east of the North Sea zone and the Baltic southern coastlands, precious gold and silver flowed in abundance from the surviving parts of the Empire, and was collected, often reworked, and subsequently redistributed from centres of power. The → central place has unfortunately become a rather overworked concept in the characterisation of Germanic Europe in this period, but there certainly was a hierarchy not only in social ranking but also of locations that we can trace archaeologically. We have, perhaps, still adequately to assess how the transmission of runic literacy may have been correlated with those conditions. The work of Alexandra Pesch on Formular-Familien amongst the gold bracteates (Pesch 2008) – superseding Mogens B. Mackeprang’s regional groups in his corpus work De nordiske Guldbrakteater of 1952 – has been thorough and exemplary in interpreting the dissemination of bracteate designs in precisely these terms, and her results must certainly be drawn upon in assessing, for instance, the clustering of bracteate finds in East Anglia including Undley and the Binham hoard (see Pestell, this volume). But runic literacy on bracteates is in fact exceptionally obscure and often seems to represent only a poor version of this field of knowledge and skill. There is by no means a straightforward link between runes, ideology and power. All of this, however, adumbrates conditions at a very early stage in relation to the horizons from which the majority of our Old English and Frisian runic inscriptions come, with the eighth and ninth centuries AD being strongly represented by these finds. Circumstances then were very different. Politically, England had been subject to a process whereby smaller units progressively lost any real independence to a small number of larger, dominant kingdoms – with Mercia and Wessex in particular vying for pre-eminence in the south of England. Within this context, however, smaller kingdoms with long histories, in particular Kent and East Anglia, might have to accept submission to a Mercian or West Saxon overking and yet could cling to a sense of their separate identity and seek means of expressing it distinctly using language and/or script for that purpose. While reciprocal relationships of personal power and subordination, or of parity, between individuals or the most successful kin groups seem to explain much in terms of the distribution of objects and influence early in the Anglo-Saxon Period, a regular structure of more impersonal trading patterns grew: a structure which is particularly well represented by special-purpose ports (known, with some historical justification, as wīcs) around the borders of England and nearby Channel and North Sea territories, as well as by monetisation, represented firstly by the sceatt coinage of the late seventh century and eighth century, succeeded by a reformed penny/denier coinage in the later eighth century. Coins, we know well, are an important context for the use and display of script. The wīcs were international entrepôts the linguistic profile of which is an interesting question for both historical linguistics and runology (cf. Zimmermann/Jöns 2017). Pestell’s paper considers the “Middle Anglo-Saxon Period” (broadly the second half of the seventh to the middle-to-later ninth century) in East Anglia in some de-

16

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

tail. A major factor in the distinctive cultural features of this phase is the consolidation of the place of the Church in a now fully Christianised England. As noted, the presence of a church and associated graveyard is clearly a feature of the site of Brandon in the north-western corner of Suffolk, and the exact ecclesiastical status of the settlement remains a matter of considerable discussion. It is similarly suggested that the Blythburgh tablets (from the coastal side of Suffolk) represent not just literacy but specifically ecclesiastical literacy. Conventionally, the Middle AngloSaxon Period is bounded at its later end by the raiding, invasions and conquests of Scandinavian Vikings, giving way therefore to an Anglo-Scandinavian Period in the north and east of England and a Late (Anglo-)Saxon Period generally. It is to this cultural context, probably in fact late in the period, that the St Benet’s Holm lead plaque with its presumably apotropaeic nonsense inscription in Scandinavian Younger fuþark runes is to be placed. It now appears in fact that a series of inscriptions on lead sheet, certainly including the “near March” and Shropham plaques, should also be dated to this late phase of Anglo-Saxon runic practice (Hines 2019). Some of these represent, as Pestell notes, the remarkable and generally welcome impact of hobby metal-detecting and the Portable Antiquities Scheme of England and Wales on the growth of the corpus of known runic inscriptions. The very important Baconsthorpe find, and one from Stoke Quay in Ipswich, conversely testify also to the importance of developer-funded contract archaeology as a source of new finds. There has been a long and close relationship between Archaeology and Art History as specialised disciplines; not one that has been without its rows (intelligently reviewed and evaluated by Wicker 1999), but on the whole the complementarity of the fields of study is generally appreciated. If Archaeology is the study of material remains of the past and, a priori, art historians specialising in the Early Middle Ages can only study material forms of art, to see Art History as essentially a specialised sub-discipline of Archaeology is neither inappropriate nor condescending. Coming back to the topic with which this introduction began, for a long time in the study of modern Archaeology – certainly from the mid-nineteenth to the mid-twentieth centuries – Art History, and especially the definition and tracing of specific styles (→ style) of decoration, was seen as fundamental to archaeological chronology. The history of art from Pre-history through the Ancient and Medieval eras and all the way down to the present day is one of a succession of styles, and of fashions changing. The date of the Franks Casket is central to Webster’s discussion of this perennially intriguing artefact, along with the geographical dimensions of its closest cultural connections and so its likely or possible provenance, a question around which Regine Marth’s discussion revolves, tracing the path the Gandersheim Casket may have taken to the eponymous abbey both from its own material clues and the information on this object found in the archives. The isomorphism of the nested concepts of style and “repertoire” which Webster uses, and the broader archaeological concepts of culture and inventory is far from superficial. At the same time, in discussing symbolism in the context of the Franks Casket, Webster points to a range of dis-

Introduction

17

course and interpretation that lies beyond and above the purely factual aspects of material and technology, or the purely literal semantics of a text, to which indeed all branches of specialisation and their methods within the Humanities should be directed as a common goal.

4 The Present and the Future of Interdisciplinary Collaboration When we drew up the concept for the Old English Runes Workshop, it was our hope that it would initiate inter- and transdisciplinary contacts and research between scholars from different fields working on the runes. To our great delight, this kind of co-operation has been far sooner in developing than the volume on the initiating conference – to our regret – has materialised: Since 2012, there has been a series of larger conferences (Across the North Sea, Leeuwarden 2014, organised by Nelleke IJssennagger, the proceedings of which have since been published 6) and smaller hands-on workshops (Norwich 2015, organised by Tim Pestell).

5 Acknowledgments We would like to thank the Göttingen Academy of Sciences and Humanities for their support, without which our conference of Module 1 and its proceedings would not have been possible. We would also like to express our gratitude to all the contributors for their exceptional patience. What is more, these conference proceedings would not have been feasible without the editors Sebastian Brather, Wilhelm Heizmann, and Steffen Patzold of the series Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde, and, last but not least, Edith Marold, who accepted the volume to be part of the RuneS series, Runische Schriftlichkeit in den germanischen Sprachen. We are grateful to our student assistants Philipp Simon and Bernhard Hübner for their help in preparing the final manuscript. We also thank our Geography student assistant Magdalena Koschmieder for creating the maps. Finally, our thanks go to Kerstin Majewski, who expertly formatted the manuscript and supported us with proofreading and bibliographical research. We would like to end with a great thank-you to those who have continued the interdisciplinary perspective of the Old English Runes Workshop. Eichstätt-München and Cardiff, September 2022

The editors

6 Hines, John and Nelleke IJssennagger (eds.). 2017. Frisians and their North Sea Neighbours: From the Fifth Century to the Viking Age. Woodbridge: The Boydell Press.

18

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

References Barnes, Michael. 1994. The Runic Inscriptions of Maeshowe, Orkney. Runrön 8. Uppsala: Inst. för Nordiska Språk. Barnes, Michael. 2013. “What Is Runology, and Where Does It Stand Today?” Futhark 4: 7–30. Berndt, Guido M. 2019. “Raids, Invasions, and Migrations: Some Considerations about Gothic Warlords and their Warrior Groups”. Le migrazione nell’alto medioevo. Atti delle settimane di studio LXVI Congresso Internazionale di studi sull’Alto Medioevo (Spoleto, 5–11 aprile 2018), Spoleto 2019. Vol. I. Ed. Fondazione CISAM. 237–264. Braunmüller, Kurt. 1998. “Methodische Probleme in der Runologie”. Runeninschriften als Quelle interdisziplinärer Forschung: Abhandlungen des Vierten Internationalen Symposiums über Runen und Runeninschriften in Göttingen vom 4.–9. August 1995. Ed. Klaus Düwel in cooperation with Sean Nowak. Berlin/New York: De Gruyter. 3–23. Düwel, Klaus. 1986. “Zum Plan einer Edition der festländischen Runendenkmäler”. Nytt om Runer 1: 16‒21. Düwel, Klaus. 2008. Runenkunde. 4th revised and updated ed. Stuttgart: Metzler. Düwel, Klaus, Robert Nedoma, and Sigmund Oehrl. 2020. Die südgermanischen Runeninschriften. Teil 1: Einleitung und Edition. Runische Schriftlichkeit in den germanischen Sprachen 1. Berlin: De Gruyter. Grimm, Oliver and Alexandra Pesch (eds.). 2015. Archäologie und Runen. Fallstudien zu Inschriften im älteren Futhark. Kiel/Hamburg: Wachholtz. Hines, John. 2019. “Practical Runic Literacy in the Late Anglo-Saxon Period: Inscriptions on Lead Sheet”. Anglo-Saxon Micro-Texts. Eds. Ursula Lenker and Lucia Kornexl. Berlin: De Gruyter. 29–59. Izzi, Luisa. 2014. “Anglo-Saxons Underground: Early Medieval graffiti in the Catacombs of Rome”. England and Rome in the Early Middle Ages: Pilgrimage, Art and Politics. Ed. Francesca Tinti. Turnhout: Brepols. 141–178. Kaiser, Livia. 2021. Runes Across the North Sea from the Migration Period and Beyond: An Annotated Edition of the Old Frisian Runic Corpus. Runische Schriftlichkeit in den germanischen Sprachen 2. Berlin: De Gruyter. Knol, Egge and Nelleke IJssennagger. 2017. “Palaeogeography and People: Historical Frisians in an Archaeological Light”. Frisians and their North Sea Neighbours: From the Fifth Century to the Viking Age. Eds. John Hines and Nelleke IJssennagger. Woodbridge: The Boydell Press. 5‒24. Mackeprang, Mogens B. 1952. De nordiske Guldbrakteater: Brakteatstudiets Historie. Brakteattypernes Udvikling, geografiske Fordeling, Kronologi, Motiver og Prægningsteknik; med fundfortegnelse, 11 Kt. og 28 Pl. med Afb. af samtlige Brakteatstempler. The Gold Bracteates of Scandinavia. Aarhus: Universitetsforlaget. Nielsen, Hans Frede. 1996. “Development of Frisian Runology: A Discussion of Düwel & Tempel’s Corpus from 1970”. Frisian Runes and Neighbouring Traditions. Eds. Tineke Looijenga and Arend Quak. Amsterdam: Rodopi. 123‒130. Nielsen, Hans Frede. 1998. The Continental Backgrounds of English and its Insular Development until 1154. Odense: Odense University Press. Page, Raymond Ian. 1999. An Introduction to English Runes. 2nd ed. Woodbridge: The Boydell Press. Pesch, Alexandra. 2008. Die Goldbrakteaten der Völkerwanderungszeit – Thema und Variation: Die Formularfamilien der Bilddarstellungen. Berlin/New York: De Gruyter. Quak, Arend. 2001. “Review of: Pforzen und Bergakker. Neue Untersuchungen zu Runeninschriften. In redaktioneller Zusammenarbeit mit Gaby Waxenberger herausgegeben von Alfred Bammesberger”. Indogermanische Forschungen 106: 331‒336.

Introduction

19

Schrijver, Peter. 2017. “Frisian between the Roman and the Early Medieval Periods: Language Contact, Celts and Romans”. Frisians and their North Sea Neighbours. From the Fifth Century to the Viking Age. Eds. John Hines and Nelleke IJssennagger. Woodbridge: The Boydell Press. 43‒52. Waxenberger, Gaby. 2017. “How ‘English’ is the Early Frisian Runic Corpus? The Evidence of Sounds and Forms”. Frisians and their North Sea Neighbours: From the Fifth Century to the Viking Age. Eds. John Hines and Nelleke IJssennagger. Woodbridge: The Boydell Press. 93–124. Wicker, Nancy. 1999. “Archaeology and Art History: Common Ground for the New Millennium”. Medieval Archaeology 43: 161–171. Zimmermann, Christiane. 2017. “Interdisziplinäre Interpretation: Theoretische Grundlagen und methodische Ansätze”. Die Faszination des Verborgenen und seine Entschlüsselung – Rāði sāʀ kunni: Beiträge zur Runologie, skandinavistischen Mediävistik und germanischen Sprachwissenschaft. Eds. Jana Krüger, Vivian Busch, Katharina Seidel, Christiane Zimmermann, and Ute Zimmermann. Berlin: De Gruyter. 429‒448. Zimmermann, Christiane and Hauke Jöns. 2017. “Cultural Contacts between the Western Baltic, the North Sea Region and Scandinavia: Attributing Runic Finds to Runic Traditions and Corpora of the Early Viking Age”. Frisians and their North Sea Neighbours: From the Fifth Century to the Viking Age. Eds. John Hines and Nelleke IJssennagger. Woodbridge: The Boydell Press. 243‒272.

Alfred Bammesberger

The Ruthwell Crucifixion Poem Revisited Abstract: The east and the west face of the Ruthwell Cross contain fragments of an Old English poem written in runic script (sections 2.1; 2.3). The poem has been entitled The Ruthwell Crucifixion Poem (2.4). Parallels to the Vercelli Book poem The Dream of the Rood are obvious (3.1; 3.2). Due to the dilapidation of the cross, the text of the poem exhibits considerable gaps. An attempt is made to fill in some of the gaps and to come as close as possible to the original version of the text (4.2; 4.3). Part of the suggested reconstruction must necessarily remain speculative (5.1– 5.5), but the function of the text on the Ruthwell Cross can be described (6.1–6.6). The relationship of The Ruthwell Crucifixion Poem to the Vercelli Book’s The Dream of the Rood (and the inscription on the Brussels Cross) is briefly dealt with (7.1–7.4). The Ruthwell Cross preserves an original poem which may be entitled The Old English Golgotha Poem or The Old English Good Friday Poem (8–9).

1 Introduction 1.1 We do not know when the Ruthwell Cross in southern Scotland was first erected: some date in the late seventh, more probably the eighth or even the ninth century is possible.1 The further history of the monument is both sad and edifying. Considered as idolatrous, the monument was knocked down in the second quarter of the seventeenth century. It was demolished but not completely destroyed. It was reerected and now stands in the Church of Ruthwell within a specially built extension. Stephens (1866, plate facing 405; see also Stephens 1884, plate facing 130) offers an engraving of the Ruthwell Cross. The sides are mistakenly reversed, but altogether the drawing gives a good idea of the very impressive monument (see Fig. 1).

1 With regard to the runic text dealt with in the following paragraphs, Kemble (1840, 352) noted that “the inscription is in Anglo-Saxon runes, in that dialect which was spoken in Northumbria in the eighth and ninth centuries”, and this assessment is still acceptable (see Ó Carragáin 2005, 53). Ó Carragáin (2005, 7) notes that the Old English poem The Dream of the Rood was written down in the late tenth century and continues: “More than two centuries earlier (c. 730–60), and at the very other end of Anglo-Saxon England (on the northern shore of Solway Firth, which then formed the extreme north-west tip of Northumbria) a poetic crucifixion narrative, clearly related to the Dream, was incorporated into the design of the Ruthwell Cross. The shape of the Ruthwell narrative is so different that we are justified in seeing it as a separate poem, a special epigraphic edition of the narrative”. Ó Carragáin (2005) represents a milestone in the study of the Ruthwell Cross to which all further research is indebted. A late date for the Ruthwell inscription is argued by Conner (2007). https://doi.org/10.1515/9783110796834-002

22

Alfred Bammesberger

Fig. 1: The Ruthwell Cross in Stephens (1866, plate facing 405).

1.2 The whole monument consists of two parts, a lower obelisk and the attached cross. Henry Duncan, a minister of Ruthwell Church in the nineteenth century, pointed out that “the whole length of the pillar is seventeen feet six inches, of which the lower stone measures twelve feet six inches, and the upper five feet” (Ó Carragáin 2005,

The Ruthwell Crucifixion Poem Revisited

23

12). In the metrical system this amounts to about 5.25 m for the whole monument.2 It is impossible now to see the monument as a whole, or to take overall photographs of it, and it is certainly difficult to appreciate how tall the monument is.3

1.3 What I will be concerned with are the two sides of the lower obelisk, which now face east and west.4 The borders of these two sides carry runic inscriptions. The direction of the runic inscription is clear: starting in the upper left corner it runs horizontally, then vertically downwards the right border, resumes in the second line of the left border and goes again downwards. The lower parts are demolished. It is probable, but cannot be proved, that in parallel to the top lines, there were also lines along the low border from left to right. We can divide up the inscription into the following major segments: (A) [horizontal], (B) [vertical], (C) [vertical], (D) [horizontal?] (E) [horizontal], (F) [vertical], (G) [vertical], (H) [horizontal?] The further subdivisions (f′) and (g′) will be explained below in 4.2 and 5.4. Of (d) and (h) nothing is left, and we could only speculate what the readings in these two base lines may have been (see 4.2). Fig. 1 is based on Stephens (1866, plate facing 405) and indicates the areas in which the runic inscription to be dealt with below occurs.5

2 Swanton (1970, 11–12) notes, “[t]he Ruthwell Cross now stands 5.28 m. high, rectangular in section and tapering towards the top (71 x 46 cm. at ground level to 33 x 24 cm. below the crosshead)”. 3 Further analysis of the monument is provided by O’Neill (2005, 81). The main points of her thesis are summarised by her as follows: “I have argued that the original form of the monument was not a cross but a pillar, terminating at the upper edge of the lower stone. I have asserted that the written and pictorial material on the monument combines in a unified thematic presentation to a female audience, on the nature of Christ and the appropriate religious response to his incarnation. I have especially questioned that commonly held belief that the poem of Old English inscription is narrated by the Cross. I have suggested that the monument was commissioned by a noble Northumbrian abbess whose spiritual guidance was in the school of Cuthbert of Lindisfarne. I have proposed that the abbess employed more than one sculptor or team, and that those sculptors were not from mainstream Northumbrian tradition of stone sculpture, but rather had some Pictish influence. I have argued that their models were mostly in the form of portable objects such as ivory panels and manuscripts, which a noble abbess would normally possess. I have based my construction as much as possible on observation of the evidence, seeking to avoid assumptions.” 4 ‘East’ and ‘West’ concern the present position of the monument. Originally, the east side faced north, the west side faced south (see Ó Carragáin 2005, xxii–xxiii; xxvi–xxvii). 5 Detailed descriptions of the runes are offered by Page (1959, 109–139) and Waxenberger (2010, 92–99). The monument carries further inscriptions both in runic script and in Latin letters, but these will not be discussed in this paper.

24

Alfred Bammesberger

1.4 The Ruthwell Cross offers the longest continuous runic inscription in Old English. Over 320 runes either remain or are recorded. With regard to length, the Ruthwell Cross inscription is followed by the Franks Casket with over 260 runes (Page 1999, 14).

1.5 The arrangement may be somewhat difficult to read. Page (1999, 147) notes that the inscription looks absurd and is maddeningly hard to read. So odd does it appear that I incline to think it may not be part of the original design for the cross, and to wonder if these runes were added by a later carver who had less command over the space he had to fill.

I will try to show that the inscription is most likely to be part of the original plan. To my mind the inscription was carved before the monument was finalised and erected. Carving the inscription may have been the final step in the production of the monument.

2 Readings 2.1 My first aim in this paper is to present an outline of the text as far as it can be read now. Brown’s readings (1921, 204) are given in Fig. 2, those offered by Dickins/Ross (1963) are represented in Fig. 3. As a starting point for the further discussion, the transliteration offered by Page (1999, 147) is given below: a) ‘[+ .nd] geredæ hinæ ḡod almeittig þa he walde on ḡalḡu gistiḡa modig f[ ore .] men ((buḡ)) [.]’ ̅ b) ‘[.] ic riicnæ kyniŋc heafunæs h[l]afard hælda ic ni dorstæ ̅ [b]ismær[ad]u uŋket men ba æt[ḡ]ad[re i]c ((wæs)) [m]iþ blodæ bist[e]mi[d] bi[.]’ c) ‘[+] kris[t] wæs on rodi hwæþræ þer fus[æ] fearran kw[o]mu [æ]þþilæ til anum ic þæt al bi[h]((eald)) sa((r.)) ic w[æ]s mi[þ] s[or]ḡu[m] gidrœ[f.]d h[n]aḡ [.]’ d) ‘[m]iþ s[t]re[l]um giwundad alegdun hiæ hinæ limwœrignæ gistoddu[n] him [.li]cæs ((hea))f((du))m ((bih))ea((l))[d]u ((h))i((æ)) [þ]e((r))[.]’

The Ruthwell Crucifixion Poem Revisited

25

2.2 On the basis of the inscriptional runic text, an arrangement according to Old English metrical alliterating lines can be attempted (→ alliterative verse). Below is the text as presented by Dobbie (1942): I [..]geredæ hinæ ḡod almeʒttig, þa he walde on ḡalḡu gistiḡa, [.]odig f[.......] men. [.]ug[..............................]

II [....] ic riicnæ k̅yniŋc, hêafunæs hlafard, hælda ic ni dorstæ. Bismærædu uŋk̅et men ba ætgad[..]; ic [...] miþ blodæ [.]istemi[.], bi[........................................]

III Krist wæs on rodi. Hweþræ þer fusæ fêarran kwomu æþþilæ til anum. Ic þæt al bi[....]. S[…] ic w[.]s mi[.] so[.]ḡum gidrœ[..]d , h[.]aḡ [..................]

IV miþ strelum giwundad. Alegdun hiæ hinæ limwœrignæ, gistoddu[.] him [......]icæs [..]f[..]m; [...]êa[.]du[..]i[.] þe[.......... ..............................]

26

Alfred Bammesberger

Fig. 2: The runes on the Ruthwell Cross’s lower stone in Brown (1921, 204).

The Ruthwell Crucifixion Poem Revisited

Fig. 3: The runes on the Ruthwell Cross’s lower stone in Dickins/Ross (1963).

27

28

Alfred Bammesberger

2.3 The following arrangement is essentially traditional and does not present any innovations. The text is split up into alliterating half-lines according to metrical indications. The result of this effort is as follows:6 East face (A)ġere(B)dæ hinæ god almeχttiġ modiġ fore allæ men (C) ic riicnæ ќyniŋc h)eafunæs hlafard bismæradu uŋќet men ba ætgadre

þa he walde on galgu ġistiga uġ hælda ic ni dorstæ ic wæs miþ blodi bistemid

West face (E) +krist wæs on (F) rodi hweþræ þer fusæ æþþilæ til anum saræ ic wæs miþ sorgum ġidrœfid (G) miþ strelum ġiwundad aleġdun hiæ hinæ limwœriġnæ bih)ealdun hiæ þer

f)earran kwomu ic þæt al bih)eald hnag ġistoddun him æt his licæs h)eafdum

2.4 This is the skeleton of the text. How further parts can be restored and/or reconstructed will be dealt with in the following pages. The title for the runic text was proposed by Howlett (1976) as The Ruthwell Crucifixion Poem (see further Howlett 1978; 2008).

3 The Ruthwell Crucifixion Poem and The Dream of the Rood 3.1 From the time of John M. Kemble onwards, the relationship obtaining between the runic text on the Ruthwell Cross and the manuscript version in the Vercelli Book 6 Velar [g] is left unmarked, but palatal [g] is written ‹ġ›. ‹k› is velar [k], ‹c› is the palatal sound, whereas ‹ќ› denotes [k] followed by a front vowel. The yew-rune is written ‹χ›. The insertions (A),

The Ruthwell Crucifixion Poem Revisited

29

has occupied the attention of scholars.7 Understandably, it was always the Vercelli version that stood in the foreground, it was normally considered as helpful and needed for interpreting the runic text. The main approach consisted in considering the two texts as being very closely related, so closely that mostly the Vercelli version was thought to represent the archetype from which Ruthwell descended and more or less represented an abridged version. I will try to show that the Ruthwell Crucifixion Poem may be considered as a complete poetic work. The Vercelli version is then likely to contain extensions and secondary additions.

3.2 In this respect, I differ from the majority of scholars, who consider the Ruthwell Crucifixion Poem as a shorter version of the Dream of the Rood. Kopár (2012, 127) notes: The Old English poem The Dream of the Rood, as well as its earlier and shorter version in runic script on the Ruthwell cross, indicate not only the widespread cult and veneration of the cross among the Anglo-Saxons, but also a view of the cross as a living being (with its own voice in both texts).

And similarly: The Dream of the Rood (and its earlier and shorter runic version on the Ruthwell cross), where the cross appears as a retainer, Christ as a heroic warrior, and the Crucifixion as a heroic act (implying the conflict of sacrifice and murder), and where faith is described in terms of loyalty and appears in accordance with the heroic code of the society (2012, 159).

I will try to argue in favour of The Ruthwell Crucifixion Poem being a separate poem8 with some overlapping with The Dream of the Rood.9

(B), (C), (E), (F), (G) indicate the spot in the panel where the respective inscription begins (see 1.3 above). 7 When Kemble wrote his 1840 paper on the Ruthwell Cross, he could not be aware of the connection obtaining between the runic text and The Dream of the Rood because the texts in the Vercelli Book were not yet available to him. But a few years later Kemble made the major additional discovery that Thorpe’s edition of the Vercelli Book poems offers a text with remarkable parallels: “The Dream of the Holy Rood contained all that had been recovered of the Ruthwell inscription, together with much more that must have perished; correcting in some respects, and confirming in others, the conclusions at which I had arrived by laborious comparison of the half-worn lines on the stone” (Kemble 1844, 33). 8 Basic for the further research history of the Ruthwell Cross was Dietrich (1865). 9 For the interpretation of The Dream of the Rood, Bütow (1935) and Schwab (1978) should be referred to in particular.

30

Alfred Bammesberger

4 Reconstructing The Ruthwell Crucifixion Poem 4.1 On the basis of the text given in 2.3 and using The Dream of the Rood as a source for supplementing gaps, Howlett (2008, 256) reconstructs the Crucifixion Poem as follows: + Ondgeredæ Hinæ Ḡod Almeʒttig. Þa He walde on ḡalḡu gistiḡa modig fore allæ men buḡa ic ni dorstæ ac ic scêalde fæstæ standa. Ahof ic riicnæ kyniŋc. Hêafunæs hlafard hælda ic ni dorstæ. ic wæs miþ blodi bistemid, Bismæradu uŋket men ba ætḡadre; biḡoten of þæs Ḡuman sida siþþan He His ḡast asendæ. +Krist wæs on rodi. Hweþræ þer fusæ fêarran kwomu æþþilæ til anum; ic þæt al bihêald. Saræ ic wæs miþ sorḡum gidrœfid; hnaḡ ic þam secgum til handa. Miþ strelum giwundad alegdun hiæ Hinæ limwœrignæ; gistoddun him æt His licæs hêafdum; bihêaldun hiæ þer Hêafunæs Dryhtin; ond He Hinæ þer hwilæ restæ. What is particularly striking in Howlett’s reconstruction is the relatively high number of incomplete verse lines.10 Proceeding on the assumption that the poem would originally have regularly alliterating verse lines, I have rearranged the text to a certain extent.

10 The apparently irregular verse lines in The Ruthwell Crucifixion Poem have repeatedly been commented on. Swanton (1970, 41) writes: “Some, like lines 39, 45 or 48, are clearly intended to represent conventional Old English alliterative verse, while others, like 40, 56, or 62, although adequate in syntax and sense, are metrically incomplete without alliterative continuation. Had the verse been composed especially for the monument it is inconceivable that it should have contained such defective lines. And while a beginning and end may well be missing, what remains of the inscription hardly represents the substance of the poem itself. Rather it has all the appearance of reference to or quotation from some familiar text. A priori this seems the most reasonable connection between the two versions.” Orchard (2009, 240) offers a similar note: “The Ruthwell poem, in so far as it can be reconstructed with any certainty, is unusual in a number of respects, not least of which is that not all of its lines are made up of paired and alliterating half-lines, as are all but one of the lines in The Dream of the Rood.”

The Ruthwell Crucifixion Poem Revisited

31

4.2 What is put down in italics may be considered as supplements based on the Vercelli Book readings; (A), (B), (C), (E), (F), (G), (F’), and (G’) indicate the positions on the stone where the respective inscriptions begin. 1 2 3 4 5 6

(A) +ondġere(B)dæ hinæ god almeχttiġ modiġ fore allæ men rod wæs ic arærid h)eafunæs hlafard bismæradu uŋќet men ba ætgadre bigoten of þæs gumu sida

þa he walde on galgu ġistiga nu ġemynte he mankynd lesan (C) ahof ic riicnæ ќyniŋc hælda ic ni dorstæ ic wæs miþ blodi bistemid siþþan he his gast onsendæ11

7 8 9 10 11 12

(E) +krist wæs on (F) rodi (stemni) (F’) hweþræ þer fusæ æþþilæ til anum saræ ic wæs miþ sorgum ġidrœfid (Gʼ) aleġdun hiæ hinæ limwœriġnæ bih)ealdun hiæ þer h)eafunæs dryctin

(G) miþ strelum ġiwundad f)earran kwomu ic þæt al bih)eald hnag ic þam secgum to handa ġistoddun him æt his licæs h)eafdum ond he hinæ þer hwilæ restæ12

4.3 The text can be translated as follows: 1 2 3 4 5 6

God almighty stripped himself courageous in the sight of all men; As a rood was I raised up, the lord of heaven; They scorned the two of us together; poured over by the blood from the man’s side

when he chose to climb the gallows, he intended to redeem mankind. I lifted up a powerful king, I did not dare to bend. I was made wet, when he had given up his spirit.

7 8 9 10 11 12

Christ was on the Cross However, readily noble persons together; I was sorely troubled with sorrows; They laid him down weary of limb; They beheld the lord of the heaven;

wounded with a lance; from afar there came I beheld all that, I inclined within reach towards the men. they placed themselves at his head. and he rested for a while.

11 The final part of line 6 may have been written in runes from left to right in the area marked ‘(d)’ in Fig. 1 (see 1.3 and 2.3 above). 12 The final part of line 12 may have been written in runes from left to right in the area marked ‘(h)’ in Fig. 1 (see 1.3 and 2.3 above).

32

Alfred Bammesberger

4.4 It will be noted that the version given in 4.2 and 4.3 differs in a number of details from the reconstruction offered by Howlett (2008, 256). I will now briefly comment on some of the major differences and try to justify my reconstruction. Notes to grammatical details are also appended.

5 Details in the Text of The Ruthwell Crucifixion Poem 5.1 With regard to line 1, the suggestion has been made before that we are concerned with hypermetric verses. In his edition of The Dream of the Rood, Pope (1981, 66) notes that at 39–40 the corresponding passage on the Cross consists of a single pair of hypermetric verses, running as follows in normalised spelling: Ongierede hine God ælmihtig

þa he wolde on gealgan gestigan

The greater regularity of the form suggests that the inscription may at this point be giving us an earlier reading rather than an abridgement.13

This assessment is convincing, less so is Pope’s immediate continuation: but it is certainly inferior, for the geong hæleþ of the Vercelli Book text brings out much more imaginatively the heroic aspect of the action, an aspect which the poem is all along at pains to emphasize as proper to Christ in his divine nature.

It seems to me that the Ruthwell text is poetically very strong because it stresses precisely Christ’s divine aspect. This is also to be recognised in the predicate walde, which is perhaps best translated as ‘chose (out of his divine will)’: Christ stripped himself of his human nature. His divine nature remained intact.

5.2 For the reconstruction of line 2b we have only the rune for u. It is likely that there is the rune for the palatal ġ, but there is no justification for reading a velar g. There-

13 The metrical aspects are discussed by Fulk (1992, 342–343).

The Ruthwell Crucifixion Poem Revisited

33

fore, the traditional assumption that we should have a text corresponding to ne dorste ic hwæðre bugan to eorðan (The Dream of the Rood, 42b) is highly unlikely.14 Ball (1991, 118–119) writes with reference to buga: The first thing to note is how much of this reading is speculative. Only ug can be seen today: earlier reliable witnesses, of whom Cardonnell is the most important, have reported no more than these two runes. The reading buga is suggested by the Dream of the Rood 42b: ne dorste ic hwæðre bugan to eorðan.

Ball is probably right in rejecting the alleged parallel to line 42b of The Dream of the Rood. It is actually likely that the notion of ‘I did not dare to bow’ is found on the Ruthwell Cross, but rather in the verb hælda (line 4b). Given the fact that The Ruthwell Crucifixon Poem is economical in its expression, it is unbelievable that ‘I did not dare to bend’ should occur twice in the short text. I think Ball’s speculative reconstruction is acceptable.

5.3 Line 3a is taken over completely from The Dream of the Rood, line 44a. It has the major advantage of providing an alliterating first half-line answering to (C) ahof ic riicnæ ќyniŋc (see also 6.3). In the following line, h)eafunæs hlafard (4a) is likely to represent a variation of riicnæ ќyniŋc. The half-line hælda ic ni dorstæ (4b) is a separate clause and means ‘I did not dare to bend’.15 The assumption that hlafard should be the object of hælda seems unlikely to me.16

5.4 Line 7 is perhaps the most drastic deviation from the conventional way the text has been read. The traditional reading is represented in Ó Carragáin’s translation: “Wounded with arrows, they laid him down limb-spent, and took their stand at the head and feet of his corpse, there they looked on {the Lord of heaven}” (2005, 181). This interpretation is unlikely for grammatical reasons: As ġiwundad would be ex-

14 It is undeniable that, in general, The Dream of the Rood has a good deal of repetition that does not immediately result from the stylistic device of variation. The notion of ‘bending’ recurs a few lines earlier: Þær ic þa ne dorste ofer dryhtnes word / bugan oððe berstan (The Dream of the Rood, 35–36a). In contrast, The Ruthwell Crucifixion Poem “is less repetitious, as if it sought only to get the central part of the Vercelli text” (Stanley 1987, 398). 15 The translation of lines 3b–4 given by d’Ardenne (1939, 148) seems correct: “I [lifted up] a mighty King, the Lord of Heaven. I dared not bow.” 16 Ó Carragáin (2005, 80) translates line 4 as “[t]he lord of heaven I dared not tilt”, but a transitive use of hælda is unlikely.

34

Alfred Bammesberger

pected to agree in case and number with limwœriġnæ, the form ought to be accusative singular. Equally or perhaps even more important is the observation that “[i]n St John’s gospel, the piercing of Christ’s side takes place directly after Christ had given up his spirit” (Ó Carragáin 2005, 81). If we rearrange the text in the suggested way, the sequence is logically sound: “The idea that Christ’s death involved the kenotic pouring out of his blood is central to the structure of the Ruthwell poem: it provides an important thematic link between the two tituli” (Ó Carragáin 2005, 182).17

5.5 If some arguments can be offered in favour of the suggested reading of line 7, one major problem still remains and cannot easily be resolved: the line lacks alliteration.18 The second half-line, miþ strelum ġiwundad, requires alliteration on st-.19 Therefore, at least one of the stresses in the first half-line should also exhibit initial st-: krist wæs on rodi has no word with initial st-, and we can only guess whether krist or rodi may have taken the place of a word with initial st-. Since krist is clearly required by the context, the only possibility that remains would be that rodi has been substituted for a word beginning with st-. The OE word for ‘stem’ would be a possibility. With all due caution, I submit that the original reading of line 7 was kríst wæs on stémni ‘Christ was on the stem (of wood)20

miþ strélum ġiwúndad wounded with the spear’.21

5.6 In lines 5b–6a bistemid and bigoten are in parallel construction. Ó Carragáin (2005, 80–81) translates ic wæs miþ blodi bistemid bigoten of þæs gumu sida as “I was drenched with blood po[ured from the man’s side…]”. This analysis, although

17 With regard to limwœriġnæ, Ó Carragáin notes, “because of the loss of blood from the wounds in his limbs as, in Beowulf, warriors are ‘wearied’ in battle, that is, ‘lifeless’” (Ó Carragáin 2005, 182). 18 It may be noted that in the rearranged text as given in 4.2 all lines have regular alliteration. It is therefore likely that originally line 7 also exhibited regular alliteration. 19 The Dream of the Rood does offer a form corresponding to strelum, but in a completely different context: eall ic wæs mid strælum giwundad (62b) refers to the Cross (and not to Christ). Evidently the passage has been considerably remodelled. 20 The form stemni would exhibit the final -i regular with masculine a-stems in the instrumental/ locative (see Dahl 1938, 61); it is therefore conceivable that rod was substituted for stemn, but the inflexional ending of stemni was taken over in the form rodi. 21 The instrumental strelum is likely to be a dual in form meaning ‘by means of the spear’; a corresponding dual form in -um is available in h)eafdum (line 11b); see 5.8.

The Ruthwell Crucifixion Poem Revisited

35

repeatedly found in translations, cannot be correct because OE begoten does not mean ‘poured (from)’. The correct analysis is ‘I was drenched, poured over with blood from the man’s side’.22

5.7 In the half-line hnag ic þam secgum to handa (10b) only the predicate is still available in the inscription, the remainder of my suggested reading is basically taken over from The Dream of the Rood. The preposition til is possible, but certainly to can also be allowed. Handa is dative singular in form, therefore ‘to the men to hand’ is basically correct, whereas ‘to the men’s hands’ (Swanton 1970, 121) would seem to indicate a plural form. Perhaps it is best to render the half-line as ‘I bowed down to within reach of the men’.

5.8 Line 11b, ġistoddun him æt his licæs h)eafdum, means ‘they placed themselves around his body’ (for the wake). OE h} eafdum is the dual form of hēafod ‘head’ and means literally ‘two heads’, but this is used elliptically to mean ‘at head and feet’.

5.9 The preterital stem bih)eald occurs twice in the text of The Ruthwell Crucifixion Poem. In ic þæt al bih)eald (9b), the verb is likely to have the basic meaning of ‘behold, see’. Theoretically the same meaning is acceptable in bih)ealdun hiæ þer h)eafunæs dryctin (12a). But it is quite conceivable that bih)ealdun here means ‘they received the Lord’ in a metaphorical way, because the passage may refer to the Eucharist. The b-half-line ond he hinæ þer hwilæ restæ (12b) may refer to Christ resting after having been taken down from the Cross and before ultimately returning to the Father on Ascension Day, but it is also thinkable that the communion bread is meant in the sense that it remained for a while in the mouth of the communicant. The relevance of the Easter ceremonies of baptism and communion to the iconography of the Ruthwell Cross is the pervading theme of Ó Carragáin’s 2005 monograph.

22 The translation offered by d’Ardenne (1939, 148) in her paper on the Brussels Cross seems basically correct: “I was all drenched and wetted with blood from the man’s side.”

36

Alfred Bammesberger

6 The Function of the Runic Inscription on the Ruthwell Cross 6.1 Admittedly, a good deal of the text in 4.2 is due to reconstruction and speculation, and therefore, to a certain extent guesswork. I would claim, however, that the guesswork is controlled by the reality of the stone monument. If the above reconstruction of the text is roughly correct, then we have every reason to consider the function of the runic inscription in its totality. Above all, line 7 is of basic importance (see 5.4 and 5.5). To my mind, it is the immediate continuation of line 6, which, unfortunately, is lacking, but is likely to have been more or less along the lines suggested by Howlett. The ‘blood’ is the most important aspect of this passage. The cross forms a unity with Christ because it is drenched by His blood, but immediately afterwards we learn how the blood issued from His body: the lance opened His side.23

6.2 The arrangement may seem strange and surprising at first sight. But ultimately this is perhaps less accidental than it seems. Fig. 4 shades the area which offers the text of line 7. This shaded area can be viewed as schematically representing a gallows, and this may well be intended as a pictorial representation of the crucifixion. It must be noted that the Anglo-Saxons were not acquainted with crucifixion, they used hanging as the capital punishment. Swanton (1970, 104) offers the following note: The Anglo-Saxons were unacquainted with crucifixion at first hand. There is no vernacular equivalent for Latin crucifigere and Gospel glossators are obliged to resort to only very approximate synonyms like cwylman, or þrowigan scolde. Most commonly, however, they simply use hon ‘hang’.

23 The unity of Christ and Cross is described by Ó Carragáin (2005, 2) as follows: “Christ requires the Cross to transcend the basic vow, the praecipuum sacramentum, of Germanic military culture, to defend and protect one’s lord even at the cost of one’s life, in order to cooperate with a Lord who wills to give his life for humankind.”

The Ruthwell Crucifixion Poem Revisited

37

Fig. 4: Line 7 of The Ruthwell Crucifixion Poem with shaded area representing a gallows.

6.3 A second instance of pictorial representation may be suspected. What has been set up as line 3b, ahof ic riicnæ ќyniŋc, is on the Ruthwell Cross a sequence of four rows of three (icr – æќy – niŋ) or four runes (iicn) each, and if we assume that ahof stood in a sequence of four runes on top of icr,24 then we have altogether five short horizontal lines which easily shape up to a vertical line symbolizing a tree. For the totally destroyed lower part of the right-hand margin, rod wæs ic arærid, may be reconstructed as line 3a of the text. Iconographically, the sequence rod wæs ic arærid ahof ic riicnæ ќyniŋc may be interpreted as the root and the top of the tree. The final -c of ќyniŋc is in the following line and is separated from h)ea- by a point. All this is hardly accidental, but I would certainly not press my interpretation.

24 If ahof was written in two lines of two runes each, which is less likely, the number of lines on top of (icr – iicn – æќy – niŋ) is increased by one.

38

Alfred Bammesberger

6.4 This second iconographic interpretation is probably less convincing than the first. But it may be useful to view the two suggestions in conjunction: The tree arises out of the earth, the root lies at its bottom. This may have been intended as the main message of the final section of (B) and the main part of (C) on the west face.

6.5 With regard to pictorial representation, Ó Carragáin’s (2005, 41) commentary may be quoted: We know that, in the context of prayers for the living and dead, particular Anglo-Saxon monasteries experimented with a variety of visual ways of recalling the scriptural Book of Life, of which liturgical diptychs were a symbolic image.

It must further be noted in particular that in Anglo-Saxon sculpture colours were used: “The traces of original painting on Anglo-Saxon sculpture are few and faint, but there can be little doubt much surviving sculpture was originally painted in bright colours, designed both to entice the eye and to glorify the message” (Webster 2012, 26). With reference to the rune-carver who is responsible for the Ruthwell Cross inscription, Swanton’s (1970, 29) note is worth quoting: “His medium is equally communicative and ornamental, and a fitting occupation of the space available will have been an important consideration.”

6.6 In the way the text of the runic poem has been arranged, it is undeniable that the whole structure is the result of careful planning. The east side basically describes the Crucifixion. The west side can be said to have the Foundation of the Church as its theme.

7 The Ruthwell Crucifixion Poem and its Relationship to The Dream of the Rood 7.1 The definitely right step in ascertaining the Ruthwell Cross text is certainly due to David Howlett, who refers to this text as the Ruthwell Crucifixion Poem. That The Dream of the Rood and The Ruthwell Crucifixion Poem exhibit parallels is not in

The Ruthwell Crucifixion Poem Revisited

39

doubt. But the usual assumption is that The Dream of the Rood represents so-tospeak the predecessor, of which The Ruthwell Crucifixion Poem offers just a few lines (3.2). The following lines will argue in favour of a different relationship, which will also take into consideration the possible chronological sequence.

7.2 The archetype may for the moment be entitled The Crucifixion Poem (see 9. below). About the age of this poem we cannot be specific, but it may well go back to the period of Bede. If Cædmon’s Hymn, the Creation Poem, was nine lines long (4 + 2 + 3), then the chances are that the twelve lines of The Ruthwell Crucifixion Poem come close to the original structure of The Crucifixion Poem. The inscription on the Brussels Cross is very short and has a clear affinity with The Ruthwell Crucifixion Poem and The Dream of the Rood (for details see Swanton 1970, 48–49; Ó Carragáin 2005, 339–354). Ó Carragáin (2005, 349), however, notes expressly that the version of the poem which the author of the Brussels text knew “was not necessarily identical with the surviving Dream”. This is definitely as far as we can reasonably go.

7.3 In the following tentative stemma the position of the Brussels version is therefore left indeterminate: The Crucifixion Poem The Ruthwell Crucifixion Poem … I = Cross

+ I¹ = Dreamer I² = Cross

The Dream of the Rood ‹Brussels Cross›

7.4 If this account is acceptable, then the text of The Crucifixion Poem must have undergone considerable reworking. In The Ruthwell Crucifixion Poem we have only one speaker ‘I’ (OE ic), namely the Cross. The most important innovation with regard to

40

Alfred Bammesberger

The Dream of the Rood is the inclusion of the ‘dreamer’, who is called ‘I1’ in the stemma, and is differentiated from ‘I2’, the Cross. If we think of the account given of Cædmon’s miracle, then it is likely that this part, namely the dreamer ‘I1’, could have been added quite early. Cædmon’s Hymn is a perfect poetic text as it stands – but in the prose account given by Bede we are told that the unlearned herd Cædmon was shy, normally he left communal gatherings when it was time to sing, but in a dream the text was revealed to him. The parallel to the dreamer ‘I1’ in The Dream of the Rood is quite close.

8 The Ruthwell Cross from Various Angles 8.1 The cross was knocked down but erected again. This is both sad and edifying. Ministers at Ruthwell and a veritable host of scholars have been interested in the fate, the history, and the interpretation of the monument. Obviously, the Ruthwell Cross requires interdisciplinary approaches if we wish to have a chance of understanding its meanings and its functions. Only a few points will be made in this final section of my – of necessity preliminary – paper.

8.2 It is rather obvious that the cross has primarily a liturgical function in a wide sense. The cross almost certainly stood out in the open air, and one can easily imagine that both clerics and secular persons gathered around it. The main function of the runic inscription has been indicated by Ó Carragáin (1987–1988, 10) as follows: part of the reason for inscribing the poem in runic letters was that it would have required readers to ruminate on the text at some length before it would yield its secrets. But monks or nuns who had (as St Benedict prescribed in his Rule) taken a vow of ‘stabilitas’ had, to put it mildly, time on their hands. The kind of reader who might not have found the text maddeningly difficult, in short, is precisely the kind of reader for whom the text is likely to have been intended: a permanent member of a community at Ruthwell who knew the monument from years of use, and who probably knew the poem off by heart from hearing it from other members of the community, from hearing it sung, or from reading some version of the poem in a manuscript.

8.3 The Ruthwell Cross originally offered the complete text of a twelve line poem on the Crucifixion and the Foundation of the Church. In this respect, the Cross stands unique because otherwise no monument is available that offers a text as extended

The Ruthwell Crucifixion Poem Revisited

41

as this and with such a clear message. The Franks Casket comes close as second, but the trouble is that so many aspects of the inscriptions on the Casket so far defy definitive explanation.

8.4 For the Ruthwell Cross I would claim that the inscriptions on the east and west faces of the lower stone, dilapidated as they are, invite us to reconstruct a full poem with a clearly delimited intention and a full religious message. This poem may be viewed as an ancestor of, rather than somehow deriving from, The Dream of the Rood: in no way is The Ruthwell Crucifixion Poem a minor version of The Dream of the Rood.25 On the contrary, The Ruthwell Crucifixion Poem is a work in its own right with, as I have tried to show, a clear intention and a full religious message.

9 Concluding Remarks 9.1 Howlett (1978, 170) sums up his interpretation of the Crucifixion Poem and its position on the Ruthwell Cross as follows: Each part of the poem relates a specific period of the Passion: I from Jesus’ disrobing to the raising of the Cross, II from the Crucifixion to Jesus’ death, III from Jesus’ death to the Deposition, and IV from the Deposition to the Sepulture. The verses do – pace Mr Page – define a particular carving on the Cross. The inner parts of the poem, II and III, alluding specifically to the Crucifixion, run down the south borders of the Cross, where they make a frame for the Crucifixion Panel at the base of the south face. What more could the carver have done to demonstrate his ‘command over the space he had to fill’?26

25 Swanton (Swanton 1970, 41) is unwilling to consider the inscription as a fully developed text: “[W]hile a beginning and end may well be missing, what remains of the inscription hardly represents the substance of the poem itself. Rather it has all the appearance of reference to or quotation from some familiar text. A priori this seems the most reasonable connection between the two versions. The Ruthwell artist was an habitual quoter, not only in the allusive nature of the material he portrays, but also in his use of the Vulgate to illustrate the theological content of his work. Just as he had used familiar verses from the Bible, some quoted direct, others appropriately modified or paraphrased, to identify the figural panels of his artistic scheme, so no doubt he chose appropriate parts from a singularly pertinent and masterly poem to fill the margins of the more universal motifs along the sides of the shaft – the central words of a poem that celebrates the relationship between Christ and cross.” 26 The sequence “where they make a frame …” (Howlett 1978, 170) may well be taken as an indication that, originally, (D) and (H) also offered runes containing parts of the Crucifixion Poem (see 1.3 above).

42

Alfred Bammesberger

9.2 In one respect, I would modify Howlett’s title of the poem. To my mind we are not just concerned with the Crucifixion (east face), but intimately linked to this is the Foundation of the Church (west face). If we take both aspects into consideration jointly, then perhaps The Old English Golgotha Poem would be a suitable title for the poetic text; still better, to my mind, is The Old English Good Friday Poem.

References Ball, Christopher J. E. 1988. “Problems in Early Northumbrian Phonology”. Luick Revisited: Papers Read at the Luick-Symposium at Schloss Liechtenstein, 15.–18. 9. 1985. Eds. Dieter Kastovsky and Gero Bauer, in collaboration with Jacek Fisiak. Tübingen: Narr. 109–117. Ball, Christopher J. E. 1991. “Inconsistencies in the Main Runic Inscriptions on the Ruthwell Cross”. Old English Runes and their Continental Background. Anglistische Forschungen 217. Ed. Alfred Bammesberger. Heidelberg: Winter. 125–136. Brown, Gerald B. 1921. The Arts in Early England, Vol. 5: The Ruthwell and Bewcastle Crosses, the Gospels of Lindisfarne, and other Christian Monuments of Northumbria, with Philological Chapters by A. Blyth Webster. London: Murray. Bütow, Hans. 1935. Das altenglische “Traumgesicht vom Kreuz”: Textkritisches, Literaturgeschichtliches, Kunstgeschichtliches. Anglistische Forschungen 78. Heidelberg: Winter. Conner, Patrick W. 2007. “The Ruthwell Monument Runic Inscription in a Tenth-Century Context”. The Review of English Studies 59: 25–51. Dahl, Ivar. 1938. Substantival Inflexion in Early Old English. Lund: Gleerup. D’Ardenne, Simonne Rosalie Thérèse Odile. 1939. “The Old English Inscription on the Brussels Cross”. English Studies 21: 145–64. Dickins, Bruce and Alan S. C. Ross (eds.). 1963. The Dream of the Rood. London: Methuen. Dietrich, Franz-Eduard-Christoph. 1865. De Cruce Ruthwellensi et de Auctore Versuum in illa inscriptorum qui ad Passionem Domini Pertinent. Marburg: Elwert. Dobbie, Elliott Van Kirk (ed.). 1942. The Anglo-Saxon Minor Poems. New York: Columbia University Press. Fulk, Robert D. 1992. A History of Old English Meter. Philadelphia: University of Pennsylvania Press. Howlett, David R. 1976. “A Reconstruction of the Ruthwell Crucifixion Poem”. Studia Neophilologica 48: 54–58. Howlett, David R. 1978. “Two Notes on ‘The Dream of the Rood’”. Studia Neophilologica 50: 167–173. Howlett, David R. 2008. “A Corrected Form of the Reconstructed Ruthwell Crucifixion Poem”. Studia Neophilologica 80: 255–257. Kemble, John Mitchell. 1840. “On Anglo-Saxon Runes”. Archaeologia 28: 327–372. Kemble, John Mitchell. 1844. “Additional Observations on the Runic Obelisk at Ruthwell; the Poem of the Dream of the Holy Rood; and a Runic Copper Dish found at Chertsey”. Archaeologia 30: 31–46. Kopár, Lilla. 2012. Gods and Settlers: The Iconography of Norse Mythology in Anglo-Scandinavian Sculpture. Turnhout: Brepols. Ó Carragáin, Éamonn. 1987/1988. “The Ruthwell Crucifixion Poem in its Iconographic and Liturgical Contexts”. Peritia 6/7: 1–71. Ó Carragáin, Éamonn. 2005. Ritual and the Rood: Liturgical Images and the Old English Poems of the Dream of the Rood Tradition. London: The British Library.

The Ruthwell Crucifixion Poem Revisited

43

O’Neill, Pamela. 2005. ‘A Pillar Curiously Engraven; with Some Inscription Upon It’: What is the Ruthwell Cross? London: BAR British Series 397. Page, Raymond Ian. 1959. “The Inscriptions of the Anglo-Saxon Rune-Stones”. Unpublished doctoral thesis, University of Nottingham. Page, Raymond Ian. 1999. An Introduction to English Runes. 2nd ed. Woodbridge: The Boydell Press. Pope, John C. (ed.). 1981. Seven Old English Poems: Edited with Commentary and Glossary. 2nd ed. New York: Norton. Schwab, Ute. 1978. “Das Traumgesicht vom Kreuz: Ein ikonologischer Interpretationsansatz zu dem ags. ‘Dream of the Rood’”. Philologische Studien: Gedenkschrift für Richard Kienast. Eds. Ute Schwab and Elfriede Stutz. Heidelberg: Winter. 131–192. Stanley, Eric G. 1987. “The Ruthwell Cross Inscription: Some Linguistic and Literary Implications of Paul Meyvaert’s Paper ‘An Apocalypse Panel on the Ruthwell Cross’”. A Collection of Papers with Emphasis on Old English Literature. Ed. Eric G. Stanley. Toronto: University of Toronto Press. 384–399. Stephens, George. 1866. The Old Northern Runic Monuments of Scandinavia and England. Volume I. London: John Russell Smith. Stephens, George. 1884. Handbook of the Old-Northern Runic Monuments of Scandinavia and England. London: Williams and Norgate. [Reprinted by Llanerch Reprint Publishers, 1993. Felinfach]. Swanton, Michael (ed.). 1970. The Dream of the Rood. Manchester: University Press. Waxenberger, Gaby. Forthc. A Phonology of Old English Runic Inscriptions with a Concise Edition and Analysis of the Graphemes. RGA-E. Berlin/Boston, MA: De Gruyter. Webster, Leslie. 2012. Anglo-Saxon Art: A New History. London: The British Museum Press.

Catherine Hills

Early Anglo-Saxon Runic Pots at Spong Hill, Norfolk, England Abstract: At Spong Hill, North Elmham, Norfolk a large early Anglo-Saxon cemetery was excavated in its entirety in ten summer seasons, 1972–1981 (Fig. 1). More than 2500 cremations and 57 inhumations were excavated. Three of the pots used as cremation urns, C1224, C1564 and C2167, carried impressions of the same stamp, a rectangle carrying three runic letters (Figs. 2–4). The first pot to be excavated, C1224, was published soon after its discovery (Hills 1974) and all three were included in the relevant cemetery catalogue volumes (Hills 1977, Fig. 58; Hills/Penn 1981, Fig. 69). The pots and their cemetery context were discussed in a paper given at Eichstätt in 1989 (Hills 1991). The inscription has been discussed by Page (1999, 92) and Pieper (1987). The initial interpretation was that the intention was to write the name of the god Tyr or Tiw, but Pieper’s generally accepted reinterpretation suggests instead a mirrored version of alu, a sequence known from other early runic texts. The purpose of this paper is not to revisit the interpretation of the runes but to discuss the runic pots within the context of the Spong Hill cemetery and to outline the recently established chronology of the site, which provides a date for the runic pots in the middle of the 5th century AD, earlier than had previously been suggested.

Spong Hill lies near the southern edge of the parish of North Elmham in central Norfolk, eastern England (Fig. 1). The Anglo-Saxon cemetery occupied the same site as an earlier Roman farm, and two kilometers to the east at Billingford there was a Roman settlement with an associated cemetery at the crossing of a Roman road and a small river, the Blackwater. Other local Anglo-Saxon cemeteries appear to have come into use only during the later phases of Spong Hill, which may have been a central burying place for the region in its earlier phases, in the 5th century AD. In recent centuries the site has been farmed, either as pasture or arable. The discovery of Anglo-Saxon cremations there is recorded on several occasions from the start of the 18th century, leading to more extensive digging in the 1950s, and the large-scale excavations of the 1970s as the threat from deep ploughing increased. Part of an Anglo-Saxon settlement was excavated in 1984. The excavations were funded by English Heritage and its predecessor organisations, and the project was initially directed by Peter Wade-Martins and Robert Carr, then by Catherine Hills from 1975 onwards. The burials were published in a series of catalogues together with site reports, in the series East Anglian Archaeology, and a final volume of analysis and synthesis has now been completed (Hills/Lucy 2013). This has shown that the cemetery was in use from the first half of the 5th century AD to the early 6th century AD. The runic pots belong to the middle of that period, the central decades of the 5th https://doi.org/10.1515/9783110796834-003

46

Catherine Hills

century. The basis for that conclusion will be presented in outline in this paper, drawing on the detailed analysis set out in the volume. The cremations were contained in pots, apart from a small number where cremated bone had been deposited directly in a pit, or in a bag which had not survived, beside a pot containing more cremated bone. The pots were handmade, not wheelthrown, and not fired in a kiln. Many had been crushed by the weight of soil, or by small piles of stones which originally marked some of the burials. Others had been damaged by later activities: agriculture, rabbits, tree roots and earlier excavations. The bones and grave-goods were usually contained in the lower half of the pot, which was found in situ in the majority of burials. The burial assemblages in the majority of cases were therefore either intact, even if crushed in situ, or were incomplete because the upper part of the pot had been cut off by later ploughing but were still clearly separate deposits. A number had been badly damaged and survived only as fragments, and there were disturbed areas of the site where smashed pots, artefacts and bones had been redeposited with consequently unreliable associations. About a third of the cremations were buried very close to one or more other cremations, in ‘burial groups’. Many of the burial groups were pairs of pots which included one decorated and one undecorated pot. Some of these were defined as “animal accessories” (McKinley 1994, 93) because they contained mostly cremated animal bones, buried next to a pot which contained mostly human bones. Both bones and associated artefacts within these paired cremations could in some cases be shown to be parts of the same assemblage, buried in two pots (Hills 1994). Nearly half of all cremations contained some animal bones, mainly domestic: sheep, cattle and pig, probably deposited as items of food, with a few wild species. The largest category however was horse, with 227 individuals identified, probably representing the deposition of whole animals, not joints of meat (Bond 1994). This is surprising as horse inhumation burials are rare in Anglo-Saxon cemeteries and usually associated with elite male human burials. About two thirds of the cremations were associated with artefacts, in some cases the burnt remains of objects put on the pyre with the body, in others unburnt objects deposited with the collected bones in the pot before burial. The burnt objects included melted glass beads and brooches, suggesting women had been laid out for cremation much as for inhumation, dressed with jewelry and other dress fasteners and attachments. There were also burnt glass and copper alloy vessels, but very few weapons. The most common grave-goods were antler combs and miniature iron shears and tweezers, the latter not apparently burnt. These categories occurred with both male and female cremations. McKinley identified individuals of all ages from neonate to older adult, and both sexes, although she found more female than male individuals (McKinley 1994, 68). The majority of the pots had linear, plastic or stamped decoration, in contrast to contemporary settlement sites where the majority of the potsherds are undecorated. This suggests the decoration had meaning in the context of burial, and was chosen as

Early Anglo-Saxon Runic Pots at Spong Hill

47

appropriate to the individuals buried. This is supported by analysis of the relationship between pottery form, size and decoration and the age and sex of the buried individuals, which demonstrates some consistent patterns of association (Richards 1987; Hills/Lucy 2013, chap. 4). Groups of pots decorated with the same stamps were identified, usually including pots similar in overall decorative style although they may include simple and complex versions of similar designs. Larger groups of stylistically similar pots were also identified (Penn 2013). Both stamp-linked groups and style groups (→ style) represent broadly contemporary pots, possibly made by one person, and the spatial clustering within the cemetery of many such groups probably indicates family burial areas (Fig. 5, Stamp Group 1). The chronology of the cemetery has been established through a combination of methods (see Hills/Lucy 2013, Chapter 3 for details). The burial groups provide stratigraphic relationships (→ stratigraphical sequence), mainly showing contemporary burial in the same pit, but also in some cases sequential burial with intercutting burials, and the stamp/style groups also provide series of contemporary burials. Correspondence analysis was carried out on the pots, based on combinations of decorative features, and also on grave-goods. This provided a relative sequence which was tested against the burial and style groups to distinguish three phases, A, B, and C. The first two phases are not spatially distinct: it appears that most of the cemetery area was in use from the start. The final phase C, with both inhumations and cremations, occupies a smaller area in the north-east of the cemetery. Comparative study of the artefact types, especially brooches, miniatures and combs, mostly drawn from northern Germany and southern Scandinavia, provided broad absolute dates within the 5th century. The latest phase, C, included both cremations and inhumations and could be set within the chronology previously established for East Anglian inhumations (Penn/Brugmann 2007). Spong Hill phase C equates to the first phase of the female burial sequence in this analysis, FA, dated ca. AD 480– 530/50 by Penn/Brugmann (2007, 69). This means that Spong Hill phases A and B, which are relatively earlier than C, date to the 5th century, as the Continental parallels indicate. The majority of the cremations, around two thousand, representing several generations, belong to phases A and B, so the cemetery must have come into use well before the mid fifth-century date given by Bede for the first arrival of Angles, Saxons and Jutes in England (Ecclesiastical History I. 15). Similar burials in comparable cemeteries can be found in a specific region of England, around the Wash, focusing on Norfolk and Lincolnshire. However, a review of cemeteries in other parts of England suggests a different story, with much less evidence of → AngloSaxon → material culture before the late 5th century. The debate as to the chronology and scale of Anglo-Saxon settlement in England should be refocused to take account of strong regional variation. There was significant impact in the region around the Wash from the first half of the 5th century, where communities included significant numbers of immigrants with strong links across the North Sea, while to the west and south the local British population did not become culturally Anglo-

48

Catherine Hills

Saxon until later, possibly more as a result of political takeover and acculturation than population replacement. The three pots with the runic stamp will be discussed in the context of the other pottery from the site, including some of the other stamped pots and pots with runelike decoration. Two of the pots and impressions of the runic stamp are illustrated on Figs. 2–4 and the location of the three runic pots and other pots discussed within the cemetery is shown on Fig. 5. Cremation 1224 was buried singly near the middle of the cemetery. It was in situ although the rim and part of the upper body of the pot are missing from plough damage. It contained the cremated bones of a young mature adult, possibly female, cremated animal bones including horse, also iron tweezers, the remains of two or three melted glass beads and four antler playing pieces. The pot is relatively large for Spong Hill: 19 cm in height, 27.5 cm in diameter. The decoration covers the upper half of the pot. Around the neck are at least eight horizontal lines separated by at least three bands of oval indentations, and around the middle of the pot is a band of alternately sloping lines and grooves. In between is a zone separated into four panels, divided by shallow slashed vertical cordons and vertical rows of cross in circle stamps. In each panel is a small round boss surrounded by small round indentations and three impressions of the same runic stamp. The stamp is a rectangle with three indentations across one narrow end, containing three runes. Cremation 1564 was buried ten metres to the south of 1224, still within the central part of the cemetery. It was close to another cremation, C1566, and crushed in situ. It contained the bones of an older mature adult, probably male, animal bone unidentifiable to species, also copper alloy sheet fragments, an iron nail, remains of six blue glass beads, and ivory fragments. The pot is of a similar size and shape to C1224 but the decoration is different. Around the neck are eight horizontal lines, above a row of rosette stamps, then another line. Below this sloping lines and bands of circle stamps form triangular panels within each of which is a single impression of the runic stamp. Cremation 1566 was buried close to C1564. They were possibly contemporary burials in the same pit but this was not clearly identifiable during excavation. The pot is decorated only with horizontal grooves at the neck. It contained the cremated bones of a mature unsexed adult and a piece of worked antler (Hills 1977, Fig. 21). Cremation 2167 was buried 30 metres to the west of the other two, near the western edge of the cemetery. It was buried singly, crushed in situ. It contained the cremated bones of a subadult human and sheep, also iron miniature tweezers, iron miniature shears and an iron bar or needle. The pot was originally of a similar shape to the other two, but slightly smaller. Around the neck lines and rows of oval indentations are like C1224, but below are triangular panels with single runic stamp impressions like C1564 (Hills/Penn 1981, Fig. 69). A fourth pot, C1672, has been included in the same Style Group, 26, as the runic pots because it has a similar decoration to C2167, horizontal lines and rows of dots

Early Anglo-Saxon Runic Pots at Spong Hill

49

around the neck, but it has no stamped decoration. It was buried some distance away from and between C2167 and C1224. This pot contained the cremated bones of two adults, one probably male, and an unusual collection of grave-goods: a griffinheaded razor, triangular comb, honestone, glass beads and a bone bead (Hills 1977, Fig. 77). Apart from C1564 and C1566 these cremations were not buried close to each other but were within a fairly narrow band across the middle of the cemetery (Fig. 5). The clay the pots were made from probably all derived from local boulder clay, with a variety of inclusions. There is no reason to identify any as an import. There is some pattern in their associations, in that three of the pots discussed contained tweezers, three had animal bones and all, apart from 1566 which may not be related, are relatively well equipped with grave-goods. However, they contained individuals of both sexes and different ages. Although there are only three pots with impressions of the runic stamp, they can be related to a number of other pots at Spong Hill, also decorated with unusual, often large and detailed, stamp motifs. The closest connection with the runic pots is with Stamp Group I which includes five pots: C1029, C1046, C1055, C1056, C1199 (Hills 1977, Figs. 57–58). These were all buried quite close together near the western edge of the cemetery (Fig. 5). They all contained the cremated bones of unsexed adults, and three also contained cremated bones of horse and other animals, and grave-goods including fragments of glass vessels, playing pieces and remains of bronze bowls, therefore relatively well equipped. One of these pots, C1046, has a decorative scheme close to that of C1224. Both have a zone on the upper half defined below by a horizontal line and divided into four panels defined by narrow slashed bosses with stamp rows on either side. Within each panel of both pots is a small round boss, surrounded either by dots or stamp impressions, with a small number of larger stamps in each panel. The latter are the runic stamp on C1224, on C1046 instead they are swastikas. The semi-circular stamp used on C1046 is seen again on the other four pots of this group, one of which also shares the swastika stamp. Apart from C1046 the pots in this group are closely similar in size, shape, and decoration, which consists of horizontal lines and stamp rows. Stamp Group 2 includes six excavated pots and sherds of two more. These pots also have large stamps with swastika, semi-circle and triangular motifs, arranged in horizontal rows and pendant triangles (Hills et al. 1994, Tbl. 1). Stamp groups 1 and 2, together with several other pots, form Style Group 1 (Penn 2013). There are other rectangular stamps, most of them on the pots belonging to Stamp Group 4. These include several versions of a narrow rectangular stamp which includes rows of X and swastika symbols, also rune-like symbols (Hills 1977, Pl. VI; Hills et al. 1994, Tbl. 2). These stamps are arranged in horizontal bands, possibly imitating Frankish roulette decoration. Several further pots have stamps in the form of animals or birds, combined with swastika motifs, some in narrow rectangular bands like the runic stamp (Hills 1983). All of these contrast with the more common-

50

Catherine Hills

ly found smaller stamps with simple patterns such as circles, crosses, or rosettes. This suggests that these motifs were chosen deliberately because they had more than decorative significance. Most of the other stamp-linked groups were found clustered together in the same part of the cemetery, suggesting family burial plots. The pots discussed above, however, apart from Stamp Group 1, were widely scattered over the cemetery, which might indicate the stamps used had more than family significance. The pots of Stamp Group 1 were all buried close together on the western edge of the cemetery (Fig. 5). All of the pots discussed so far have been dated to phase B at Spong Hill, the central decades of the 5th century. When these pots were buried the onlookers would have seen them from above, thus looking at concentric circular bands of decoration which can be compared to the decoration on circular brooches (Richards 1992, Fig. 27). Looking in this way at the pots from Stamp Group 1 and other pots with rows of stamps around the neck a parallel might be drawn with the concentric bands of stamped decoration seen on some Scandinavian gold bracteates. A well known B bracteate from Gudme, for example, Gudme II–B (IK 51,3) has concentric bands of semi-circular and triangular stamps with internal beads or pellets, the same motifs and the same arrangement as those used for the pot stamps on C1199 and C1029. Other bracteates with stamped borders often include triangular and semi-circular motifs as well as diagonal crosses and S-shaped stamps, which also occur on Spong Hill pots (Axboe 1981, Figs. 47– 53). More immediately relevant are two recent finds of gold bracteates in Norfolk (Behr 2010). One, from Brinton, is classed as an A-bracteate, with a central image which includes human, animal and bird elements, and a border of small stamped semi-circles (Behr 2010, Fig. 11). Most striking is the B-bracteate from Binham (Behr 2010, Fig. 12). Here the central image is of a standing figure between two beasts, his hand in the jaws of one. There is also a runic inscription. The border has concentric bands of triangle with circle and diagonal cross stamps, both motifs found on pots at Spong Hill. Bracteates and pots each have concentric bands of stamps, including some of the same motifs. In the middle of the bracteates are human and animal figures while in the pots are cremated human and animal bones. We cannot say they had exactly the same meaning but pots and bracteates belong to the same thought world, use the same symbols and are at least partly contemporary, since Axboe (2004, 275) dates the production of bracteates from the mid-fifth to the midsixth century AD. Outside Spong Hill runic stamps are rare. Teresa Briscoe listed sixteen in category L, runic, in her classification of Anglo-Saxon pot stamps, including the Spong Hill examples (Briscoe 1983, 70). Currently Diana Briscoe has 76 examples in the Archive of Anglo-Saxon Pottery Stamps (AASPS) of stamps with motifs definitely or probably representing the t-rune. These are quite small, essentially a multiple triangle with a central stem, and it is difficult to be sure how many are intentionally runic. The AASPS category L4ai, rectangular with two or more runes or rune-like symbols, so far occurs only at Spong Hill, but there are five examples outside Spong

Early Anglo-Saxon Runic Pots at Spong Hill

51

Hill of category L4bi, rectangular stamps with rune-like symbols (pers. comm. D. Briscoe). Two are from Pensthorpe and Grimston in Norfolk, both very similar to the Spong Hill Stamp Group 4 stamps, the other three are smaller, from Loveden Hill and Croft in Lincolnshire. Briscoe has also listed thirty-six examples of pots with stamps in the form of animals or birds, including those from Spong Hill (Eagles/ Briscoe 1999) and many examples of swastika stamps. All are found almost exclusively in eastern England, especially East Anglia, which reflects the distribution of large cremation cemeteries like Spong Hill. At Spong Hill and elsewhere there are many pots with linear or bossed designs which might include runes (Myres 1977, Figs. 368–369). It is not easy to draw the line between purely geometric use of patterns such as a diagonal cross or X, or vertical combined with zig-zag lines, and deliberate use of such motifs as runes. There are two categories of design worth considering in this context. First are those patterns which include linear patterns which are not simple geometric motifs, or where the pattern is interrupted or not regularly repeated. For example C1832 (Hills/ Penn 1981, Fig. 52) is decorated with horizontal bands of bosses, dots and vertical and curved lines, but also a row of short vertical and sloping lines which look like an attempt at runes – less clear but comparable to those on the Loveden Hill runic pot (Bammesberger 1991, Pl. 1). C2628 is decorated with vertical bosses, lines and dots, but in one panel as well as dots there is a swastika and a double square (Hills et al. 1987, Pl. Vc). A number of pots include linear swastikas, some repeated in a band around the neck, others singly, in one case on the base of the pot. C2436 has vertical slashed cordons which define panels occupied by irregular linear patterns which might include complex swastikas (Hills et al. 1987, Pl. Vf). These pots do not fall into clear groups, except for having irregular linear/bossed patterns. They belong either to phase A or to phase B and, like the stamped pots discussed above, contain individuals of both sexes and all ages, and a range of animal species. Gravegoods include most categories of find, apart from brooches. The second category forms a more coherent group, carrying what appear to be deliberate t ᛏ runes (Figs. 6–7). Some have bosses in the form of a single-arm t (Hills/Penn 1981, Fig. 65; Penn 2013, Style groups 31 and 32). There is also a group of nine similar pots with multiple-arm linear t (rune-like) motifs: C1345, C2218, C1333, C2451, C2504, C2523, C2528, C2583 and C2585. These cremations all belong to phase A, the initial period of the cemetery’s use in the first half of the 5th century. They contained individuals of both sexes and a range of ages, and most had animal bones including sheep and horse. Five had tweezers and/or miniature shears, as well as a range of other grave-goods: glass beads, antler comb, and ivory fragments. None contained brooches. The linear t pots cluster in the southeast corner of the cemetery (except 1345), while the bossed t spread across the eastern edge (Fig. 5). In conclusion, the three pots with the runic stamp from Spong Hill can be seen to be local products made and buried in central Norfolk in the 5th century, probably the middle decades of that century. The stamped decoration can be related to other

52

Catherine Hills

stamped pots which are also relatively well made, decorated with relatively large and complex stamped motifs such as animals or swastikas. Some of the latter can be paralleled at other Anglo-Saxon cremation cemeteries such as Caistor by Norwich or Cleatham, but the runic stamps so far appear as peculiar to Spong Hill. These pots belong to a period when local variation had developed. The pots of the earliest phase at Spong Hill have relatively few stamps, and those are not complex motifs. However, the pots of phase A do display a range of linear and bossed designs which include runes or rune-like symbols. The concept of runes arrived from North Germany and was used sparingly on cremation pots – and the astragalus from Caistor-byNorwich. The idea was developed further at Spong Hill within the context of developing local and familial traditions, when the pottery was decorated with motifs which had meaning to those who made the pots and buried their dead in them.1

Figures 1. 2. 3. 4. 5. 6. 7.

Location map of Spong Hill Runic pots C1224, C1564; drawings: C. Hills Runic stamp on C1224; photo: M. Dabski Runic stamp on C1564; photo: M. Dabski Distribution of runic and related pots at Spong Hill Pot with bossed t, C2009; drawing: K. J. Penn Pot with linear t, C2451; drawing: K. J. Penn

1 I would like to thank Tim Pestell of Norfolk Museums Service and David Gurney of Norfolk County Council’s Historic Environment Service for permission to use drawings and photographs from the Spong Hill archive, Diana Briscoe for information about Anglo-Saxon pot stamps outside Spong Hill, Andrew Hall of the Cambridge Archaeological Unit for the plans and Sam Lucy, Newnham College Cambridge, for reading and commenting on the text.

Fig. 1: Location map of Spong Hill; © Cambridge Archaeological Unit.

Early Anglo-Saxon Runic Pots at Spong Hill

53

54

Catherine Hills

Fig. 2: Runic pots C1224, C1564; drawings: C. Hills.

Figs. 3 and 4: Runic stamps on C1224 (left) and C1564 (right); © Norwich Castle Museum; photos: M. Dabski.

Early Anglo-Saxon Runic Pots at Spong Hill

Fig. 5: Distribution of runic and related pots at Spong Hill; © Cambridge Archaeological Unit.

Fig. 6: Pot with bossed t, C2009; drawing: K. J. Penn.

55

56

Catherine Hills

Fig. 7: Pot with linear t, C2451; drawing: K. J. Penn.

References Axboe, Morten. 1981. “The Scandinavian Gold Bracteates. Studies on their Manufacture and Regional Variation”. Acta Archaeologica 52: 1–100. Axboe, Morten. 2004. Die Goldbrakteaten der Völkerwanderungszeit – Herstellungsprobleme und Chronologie. Berlin/New York: De Gruyter. Bammesberger, Alfred (ed.). 1991. Old English Runes and their Continental Background. Anglistische Forschungen 217. Heidelberg: Winter. Bede, the Venerable. 2008. The Ecclesiastical History of the English People. The Greater Chronicle. Bede’s Letter to Egbert. Eds. Judith McClure and Roger Collins. Oxford: Oxford University Press. [1st ed. 1969]. Behr, Charlotte. 2010. “New Bracteate Finds from Early Anglo-Saxon England”. Medieval Archaeology 54: 34–88. Bond, Julie. 1994. “The Cremated Animal Bone”. Spong Hill VIII: The Cremations. Ed. Jaqueline McKinley. East Anglian Archaeology 69. Dereham: Norfolk Museums Service. Briscoe, Teresa. 1983. “A Classification of Anglo-Saxon Pot Stamp Motifs and Proposed Terminology”. Studien zur Sachsenforschung 4: 57–71. Eagles, Bruce, and Diana Briscoe. 1999. “Animal and Bird Stamps on Early Anglo-Saxon Pottery in England”. Studien zur Sachsenforschung 13: 99–111. Hills, Catherine M. 1974. “A Runic Pot from Spong Hill, North Elmham, Norfolk”. The Antiquaries Journal 54 (1): 87–91. Hills, Catherine M. 1977. The Anglo-Saxon Cemetery at Spong Hill, North Elmham, Part I: Catalogue of Cremations, Nos. 20–64 and 1000–1690. East Anglian Archaeology 6. Dereham: Norfolk Museums Service. Hills, Catherine M. 1983. “Animal Stamps on Anglo-Saxon Pottery in East Anglia”. Studien zur Sachsenforschung 4: 93–110.

Early Anglo-Saxon Runic Pots at Spong Hill

57

Hills, Catherine M. 1991. “The Archaeological Context of Runic Finds”. Old English Runes and their Continental Background. Anglistische Forschungen 217. Ed. Alfred Bammesberger. Heidelberg: Winter. 41–59. Hills, Catherine M. 1994. “The Chronology of the Anglo-Saxon Cemetery at Spong Hill, Norfolk”. Prehistoric Graves as a Source of Information: Symposium at Kastlosa, Oland, May 21–23, 1992. Ed. Berta Stjernquist. Stockholm: Almqvist & Wiksell. 41–50. Hills, Catherine M., Kenneth Penn, and Jacqueline McKinley. 1981. The Anglo-Saxon Cemetery at Spong Hill, North Elmham, Part II: Catalogue of Cremations, Nos. 22, 41 and 1691–2285. East Anglian Archaeology 11. Dereham: Norfolk Museums Service. Hills, Catherine M., Kenneth Penn, and Robert Rickett (eds.). 1987. The Anglo-Saxon Cemetery at Spong Hill, North Elmham, Part IV: Catalogue of Cremations, Nos. 30–32, 42, 44A, 46, 65–66, 2286–2799, 2224 and 3325. East Anglian Archaeology 34. Dereham: Norfolk Museums Service. Hills, Catherine M., Kenneth Penn, and Robert Rickett (eds.). 1994. The Anglo-Saxon Cemetery at Spong Hill, North Elmham, Part V: Catalogue of Cremations, Nos. 2800–3334. East Anglian Archaeology 67. Dereham: Norfolk Museums Service. Hills, Catherine M. and Sam Lucy. 2013. Spong Hill Part IX: Chronology and Synthesis. Cambridge: McDonald Institute for Archaeological Research. McKinley, Jacqueline. 1994. Spong Hill Part VIII: The Cremations. East Anglian Archaeology 69. Dereham: Norfolk Museums Service. Myres, John N. L. 1977. A Corpus of Anglo-Saxon Pottery of the Pagan Period. Oxford: Oxford University Press. Page, Raymond Ian. 1999. An Introduction to English Runes, 2nd ed. Woodbridge: The Boydell Press. Penn, Kenneth J. 2013. “Style Groups”. Spong Hill Part IX. Eds. Catherine M. Hills and Sam J. Lucy. Cambridge: McDonald Institute. 409–437. Penn, Kenneth, Birte Brugmann, and Karen Høilund Nielsen. 2007. Aspects of Anglo-Saxon Inhumation Burial: Morning Thorpe, Spong Hill, Bergh Apton and Westgarth Gardens. East Anglian Archaeology 119. Dereham: Norfolk Museums Service. Pieper, Peter. 2005. “Spong Hill: § 2 Spiegelrunen”. Reallexikon der germanischen Altertumskunde. Vol. 29. Eds. Heinrich Beck et al. Berlin: De Gruyter. 380–382. Richards, Julian D. 1987. The Significance of Form and Decoration of Anglo-Saxon Cremation Urns. British Archaeological Reports 166. Oxford: BAR Publishing. Richards, Julian D. 1992. “Anglo-Saxon Symbolism”. The Age of Sutton Hoo: The Seventh Century in North-Western Europe. Ed. Martin Carver. Woodbridge: The Boydell Press. 131–147.

John Hines

The Archaeological Dating of the Early Finds: How Certain are the Results? Abstract: This paper provides an introduction to the methods and terminology of archaeological chronology for those who are interested in runic inscriptions but have not studied Archaeology. In particular, it discusses how archaeological dating methods and results relate to the study of the Early Anglo-Saxon and pre-Old English inscriptions.

1 Terminology and Methods The primary and particular subject-matter of Archaeology is the material life and material remains of the past. It is important to emphasise that “artefacts” form just one category within that domain. We may rationally regard complex constructions and assemblages such as buildings and whole sites as extensions of that same basic category of “things people have made”. However, a distinction between any particular object and a context is a crucial one in archaeological procedures, and equally important are the relationships between objects and contexts. Beyond the composite “site”, in fact, Archaeology is also concerned with even larger-scale entities: regions, periods and cultures reflected in material life. Those are generally discerned by placing sites in their contexts. All of these levels of inter-relationship and of reference impinge upon archaeological chronology. All the same, we may assign artefacts and sites – the things we typically find and can excavate – to a fundamental level within the discipline. The questions of “what is it?” and “how old is it?” are equally fundamental in Archaeology, not least because both are essential to making informed associations between artefacts and sites as the basis for a higher or a more general level of interpretation. The concern of the present discussion is with dating: Archaeology can offer answers to the question of how old something is using a variety of forms of evidence, and thus produces a range of different forms of dating as a result. When a runic inscription is involved, runologists will rightly usually want to know first of all how closely we can determine when the inscription was made. But they should also be equally interested in what evidence there might be for how long and also in what circumstances that inscription could have been seen and read. Runic inscriptions are themselves artefacts, in a strict sense. They are also in the great majority of cases found upon objects that in turn are functional artefacts – e.g., dress-accessories, weapons, tools or monuments – to which the runic inscription in itself is therefore one additional attribute, usually one amongst several attributes https://doi.org/10.1515/9783110796834-004

60

John Hines

that the whole artefact has (Clarke/Chapman 1978, 149‒204). In terms of chronological reasoning, we may be able to date an archaeological context such as a grave, or a layer on a settlement site, on the basis of one or more of the artefacts that it contains; reciprocally we may be able to date an artefact by reference to the context in which it occurs. These relationships underlie a crucial series of different forms of date in Archaeology. For an artefact, consequently, there must be a date of production/manufacture. That is in fact something which it is often hard for Archaeology to determine very precisely. Inter alia, we rarely see the remains of the process of production itself for the relevant artefacts in the period in question here. There is then the date of deposition, assuming that the item has, as the majority of archaeological finds, been recovered from the ground. Ideally, archaeological “small finds”, as we commonly refer to individual artefacts, are retrieved from “closed contexts”: layers and assemblages that themselves were accumulated and deposited within a specific period in the past, and which have not been significantly interfered with since then. Those conditions imply a degree of contemporaneity between the finds within the single context – although of course that may have broad or narrow parameters depending upon the past processes by which the assemblage was brought together. In between the two dates (of production and deposition) lies the life-span of the object: the period in which it was used, seen, and possibly altered. The Harford Farm (Caistor-by-Norwich) disc brooch is a perfect example of an item that not only was modified during its life, but where the runic inscription on the back can be argued to pertain to that time of alteration (Penn 2000, 45‒50 and 81‒82; cf. Bammesberger 2003; Hines 2020). We might in fact stress here that relatively few of the objects with runic inscriptions that we have from Anglo-Saxon England appear to have been made with the specific intention that they should carry such a text. A special class of counter-examples are the bracteates, where in some cases (a minority, but a substantial minority) the design cut into the die from which the pendants were pressed includes a runic text; furthermore, there are cases where both the contents and the positioning of the text in the design appear intended to contribute actively and dramatically to the function of the item (Heizmann/ Axboe 2011, 297‒601). It is also perfectly obvious that, say, the Loveden Hill cremation urn must have been inscribed before the clay had even dried, let alone been fired, while the Spong Hill runic pot-stamps must have been ready-made before the cremation urns on which their impressions have been found were produced. The premise that we can date artefacts from the contexts in which they can occur and contexts from the artefacts that occur in them plainly involves a certain degree of circularity – but this in fact is one that can produce an important and reliable, purely archaeological form of chronological sequence: the → relative chronology, wherein we can set phenomena into relationships of “earlier than” or “later than”, and sometimes “contemporary with”. In this case “contemporaneity” may mean that, say, types B and C both are known to be later than a type A and earlier

The Archaeological Dating of the Early Finds

61

than a type D, but no more specific relationship between them can be defined. In other words, we may regard phenomena such as types B and C in this illustration as “contemporary” purely in the sense that we cannot tell them apart in date, or even just because any difference in date between them is irrelevant to the particular circumstances in which they are being considered. One form of relative chronology is archaeologically fundamental and totally secure: this is the → stratigraphical sequence between layers and/or features, whereby, for instance, a layer of earth that underlies another layer, or is cut by pits, post-holes or ditches, must be earlier than any of those overlying layers and features (Harris 1979). Even without direct “vertical” stratigraphical contiguity, it can be shown that features in a “horizontal” stratigraphical relationship to one another – for instance, graves in a regular row, or a group of pits or a ditch which block the doorway of a building – may also have determinable relative-chronological relationships (Holst 1999). Much of the detail in our archaeological chronologies lies in deduced or inferred relative-chronological sequences which can at certain points be associated with absolute chronological or calendrical anchor-points or horizons. The occurrence of historically dated artefacts in Anglo-Saxon contexts is rare, but not unknown, and indeed was crucial to the initial identification of the burials of Early Anglo-Saxon Kent as belonging to this period rather than being those of the earlier “Ancient Britons” by the Rev James Douglas in the 1790s (Douglas 1793; Hines/Bayliss 2013, 13). For us, such historically dated artefacts exclusively take the form of coins issued under the authority of a known ruler: usually a king whose reign is datable from historical sources; sometimes an archbishop or bishop. Such coins were in fact quite frequently forged and imitated even after the reign of the ruler in question, but nonetheless, in the simplest terms, the historically attested date of accession of the ruler at least provides a terminus post quem for any closed context in which such a coin occurs. Douglas (1793) was able to identify coins, made into pendants and deposited in Kentish graves as female dress jewellery, as the issues of sixth-century Merovingian Frankish kings. The precise study of variation in the form amongst closely related artefact-types, and its interpretation in terms of chronology, has been much debated amongst archaeologists (Malmer 1963). Nonetheless, artefact typology is undoubtedly one of the most basic archaeological procedures: at heart it is, very simply, the primary taxonomy by which we order and think about archaeological data. While one can order artefact-types in scales according to particular parameters, it is by no means obvious which direction development through time should have followed in relation to those scales: from simplicity to increasing complexity, for instance, or through ever greater simplification from the prototype stage. At this level of study, it is rarely utilitarian and functional aspects of the artefacts being studied that are most informative but rather aspects of style. Indeed, the typological classification of artstyles and their assignation to a historical sequence has been a delicate instrument for dating the artefacts which in effect they decorate in a non-utilitarian way for a

62

John Hines

very long time indeed, and remains so (Salin 1904; Leeds 1936; Bakka 1959). Amongst art-styles there is in fact a form of generic life-cycle that follows a more predictable route from phases of intense creativity through periods of greater stylisation and a simplification of the elements and motifs of the style, before a new burst of creativity introduces a new style. Nonetheless, we are well aware that the history of Germanic and Anglo-Saxon art-styles has been refined and corrected greatly over the past century and more, and will continue to be so (Høilund Nielsen/ Kristoffersen 2002). An especially important mode of relative chronology is that created by → seriation. The Victorian archaeologist Flinders Petrie is usually credited with defining these processes in the context of his study of Egyptian cemeteries in the 1890s (Petrie 1893). If we find two discrete contexts with assemblages of the same artefact-types in more or less the same relative quantities, we should be obliged to regard those as at least “broadly contemporary” in the sense that we have no basis on which to distinguish them chronologically. If, conversely, we can detect gradual differences between assemblages – e.g., one with types A, B, C and D; another with B, C, D and E; another with C, D, E and F; and so on – we may postulate that the difference is chronological: type A became obsolete, to be replaced by type E; then type B by type F, and so on. Obviously, there are common-sense tests on such inferences: are artefacts A and E both types of brooch, for instance, or is there no discernible reason why the obsolescence of one type should be coordinated with the introduction of the other? Gradual variation of this kind may, of course, track a scale of increasing wealth, or of social rank, or reflect nuances of regional background: the explanation will not necessarily be chronological. In connection with independent dating evidence, however, such sequencing or ordination by seriation is an extremely powerful chronological instrument, not least for a material culture such as that of the Early Anglo-Saxons and their contemporary Continental Germanic neighbours, for whom we have very large numbers of burials furnished with grave goods. In British archaeology, the “Early Anglo-Saxon Period” is precisely defined as the period in which artefacts were regularly included in the burials of the dead as “grave goods” (Hines/Bayliss 2013, 13, 27–30). From the middle of the twentieth century a range of additional, laboratory-based and mathematically more complex dating techniques have been developed, which essentially operate in tandem with the approaches outlined above to refine and support the archaeological chronology. The best known of these is → radiocarbon dating. In essence, this technique attempts to measure the age of a sample – which means the time that has elapsed since a living organism died – by comparing the levels of stable and unstable/decaying radioactive carbon isotopes within it (Bowman 1990; Hines/Bayliss 2013, 35‒60). The radiocarbon age thus calculated is expressed in terms of ‘n ± y’ years BP (= ‘Before Present’), where “Present” was set by convention at AD 1950, and the expression ‘± y’ represents 1 standard deviation (1σ in conventional mathematical notation) or c. 68 % of probability: in other words, there is

The Archaeological Dating of the Early Finds

63

The radiocarbon age is represented by the bell-shaped curve on the vertical axis. Its median line lies at 1469 BP and the probability of the true age is distributed symmetrically and equally either side of that beneath what is statistically known as a “normal curve”. 1σ is 18 years, so that 68 % of the probability lies in the range 1451–1487 BP and 95 % of the probability in the range 1433–1505 BP. This regular distribution of probability then has to be diffracted via the irregular calibration curve (the jagged diagonal band running from top left to bottom right), which produces a radiocarbon date with the range of probability shown against the horizontal axis, most of which manifestly lies within the range cal AD 525–650 but a small quantity of which spreads back into the 5th century cal AD. The calibration curve used for this simulation is IntCal13. Fig. 1: The production of a radiocarbon date from a radiocarbon age by calibration.

still just under a 1 in 3 chance (32 %) that the actual age lies outside this range: i.e., approximately a 16 % possibility of it in fact being earlier, and the same 16 % possibility of it being later. The 2σ range (c. 95 % probability) can easily be calculated, and most archaeologists refer to that range of higher confidence when using radiocarbon dates. The radiocarbon age, which in fact is in units of “radiocarbon years” that are not identical with calendar years, then has to be calibrated, because we know that the levels of radioactive carbon in the atmosphere are not constant and have fluctuated in the past. This process produces distributions of probability that can be expressed in a simplified term either as ‘cal BC/AD DATE ± x years’, or ‘cal BC/AD DATE–DATE (95 % probability)’, or more precisely represented purely in graphic form (see Fig. 1). The international standard “calibration curve” is under continuous review and revision, generally leading to ever greater precision. An important new standard, IntCal20, with particularly significant consequences for radiocarbon dating in the period of the fifth to eighth centuries AD, was issued in August 2020 (Reimer et al. 2020). The calibration and refinement of the radiocarbon curve are in fact possible because another technique, tree-ring dating or dendrochronology, allows us to measure and compare the levels of the radioactive carbon isotope C14 in known-age wood, which may be of very great age. Because growing conditions vary from year to year, the relative sizes of growth-rings in trees from a specific area can be com-

64

John Hines

pared to provide matches rather like a bar-code (the measurements being compared are processed statistically to ensure a sufficiently reliable fit), and so to produce long chronological sequences, which not only allow us to date other samples of the same species of wood from the same area, but also show us a detailed chart of climatic history (Baillie 1999). Dendrochronology can be astonishingly precise: if the bark on the wood is still present, and with evidence of insect pupation, it can date an event not only to a specific year but even to within a few weeks as much as three to four thousand years ago. Unfortunately, surviving timber of sufficient size is rarely available from Anglo-Saxon England, although we do have dendrochronological dates to support the chronological sequence for comparable Merovingian graves on the Continent; this technique has also proved invaluable in giving us precise dates for the famous Norwegian Viking ship burials. There are further laboratory-based dating methods, including thermoluminescence, archaeomagnetism, and so on, but these do not at present yield results of the same level of precision for our material.

2 A New Chronological Framework for Early Anglo-Saxon England What we can now use for the Early Anglo-Saxon Period are the results of a project that ran for some 15 years and published its final report in 2013 (Hines and Bayliss 2013). The research project “Anglo-Saxon England c. AD 570–720: The Chronological Basis” was funded by English Heritage. Its programme combined and compared the results of new and thorough work in a series of chronological techniques, both traditional and new. The typology of artefact-types from Anglo-Saxon graves of the sixth and seventh centuries AD was thoroughly reviewed and in significant ways revised. The varying combinations of artefact-types in grave-assemblages were then placed in a sequence by seriation using a computerised technique for exploring patterns in data-matrices called → correspondence analysis (an introductory account of the principles and mathematical nature of this technique can be found in Hines and Bayliss 2013, 60‒73). From the series of grave-assemblages thus produced (which had to be kept separate for male and female sequences respectively) certain burials were carefully selected for high-precision radiocarbon dating of samples of the human (and in one case equine) bone within them. Finally, the results were subjected to Bayesian modelling. Bayesian modelling is a statistically based process which on the one hand evaluates a prior chronological hypothesis – for instance, that the seriation of the grave-assemblages had identified a true chronological trend in the use, obsolescence and replacement of defined artefact-types – and then suggests refinements to the calendrical dates produced by the radiocarbon-dating method if and when the model is validated by the first test. It is of primary importance to note

The Archaeological Dating of the Early Finds

65

that this chronological project used a project specific calibration curve, generated to produce reference data of greater precision than those otherwise available at that time. These data were subsequently incorporated into an international standard curve, IntCal13, from 2013. Details on how this has now been superseded and improved upon by IntCal20 are provided and discussed in a major review paper by the author (Hines 2021). The full report on this new chronological framework runs to well over a quarter of a million words, with more than six-hundred figures and tables, so it is not easy to summarise the results briefly. However, for a period from c. AD 520 to the late seventh century, when regular burial with grave goods ceased, separate but comparable sequences of phases of male and female burials have been defined and dated. This means that if and when we encounter additional grave-assemblages with the appropriate combinations of artefact-types, we can assign them to a phase, and that phase has an associated date-range. Bayesian modelling using radiocarbon dates in fact defines a probability range (at 1σ and 2σ again: 68 % and 95 % probability respectively) for the beginning and end of each of these phases.1 With less archaeological reliability, that could be converted into a single, wider range of probability for any grave-assemblage assigned to that phase. More than 150 artefact-types are also assigned to particular phases or, more often, combinations of two or more sequential phases. This also, then, provides date-ranges for these artefact-types. It is important to highlight at this point that there is one particular issue in the chronology that remains unresolved. It is a significant unanswered question, but its terms in themselves stand as a measure of the precision and success of the project. Firstly, we suspect that a further adjustment to the calibration of the radiocarbon ages may be appropriate to take account of the consumption of fish by the people whose bones have been dated. Fish typically have ingested older atmospheric carbon than terrestrial animals, and this affects the organic chemistry of the skeletons of fish-eaters (see especially Jarman et al. 2018). At present it is not possible to make a consistent and reliable calibration of this effect; however, the best calculations we can make at present indicate that at most the results produced with a fully terrestrial calibration may set the dates somewhere between 5 and 10 years too early. This is highly relevant to the unresolved problem. Amongst the latest female furnished graves from England are some which contain early Anglo-Saxon coins: pale-gold shillings (tremisses, sometimes known as “thrymsas”) of the Pada and Vanimundus types, and silver “Primary sceattas” of Series A and B (Metcalf 1993‒ 1994). At present, the conventional numismatic chronology assigns the introduction of those types to c. AD 670 and c. AD 680 respectively, and essentially that is some

1 A date-estimate produced by Bayesian modelling is by convention expressed in italics. Thus the expression “95 % probability” is correct for the measure of probability implicit in a Bayesian model of the chronology incorporating the radiocarbon date, rather than the probability that can be calculated for the radiocarbon date on its own.

66

John Hines

10–20 years later than our chronological model would put them (Hines/Bayliss 2013, 492‒515). Adjusting the calibration for the sort of dietary effects just described reduces the discrepancy but does not remove it. Whether a further adjustment to the calibration is needed, or the numismatic chronology is a little too late, is not something to be debated here. The answer might indeed be that both adjustments are appropriate. Considering the state of Anglo-Saxon artefact and burial chronology for the seventh century AD when this project began, this proximity between two quite separate dating methods is a massive vindication of the dependability and value of its results. Nonetheless, there are important new questions now to be answered as a consequence of the research so far, and the explanation of this discrepancy with the numismatists’ preferred chronology is one of the most urgent of those. This issue apart, the results of the chronological project are revolutionary in terms of the range of the new scheme and the degree of confidence one can hold in it; these aspects are to be emphasised rather than trying to advertise the results as showing things to have been strikingly different than was thought before – even though bringing forward the date of a general end of furnished burial in AngloSaxon England from the early decades of the eighth century to the late decades of the seventh century is certainly a dramatic shift of that nature (cf. Geake 1997). Let us illustrate this by summarily reviewing the Early Anglo-Saxon or Pre-Old English runic corpus (for the relationship between these categories, which are respectively archaeological and linguistic, see Hines 2019, 30). This collection of material has expanded only a little since I reviewed it in the proceedings of the 1988 Eichstätt conference, “Britain 400–600: Language and History” (Hines 1990). The largest category in this corpus is that of the female artefact-types, nearly all from known and closed grave contexts. Robert Nedoma (2011) has published the results of our modelling for four of the eight burial contexts (see Table 1). Of these, the burial of a woman with an inscribed cruciform brooch in West Heslerton grave 177 is itself radiocarbon-dated, albeit to standard precision, not high precision (Haughton/Powlesland 1999, II, 310‒12). With limited information in 1988, I assigned the manufacture of the brooch to the first half of the sixth century (c. AD 500–550); we can now (using the remodelling and recalibration conditions presented in Hines 2021) say that there is 95 % probability that it was buried in the period c. AD 505– 565, and these two conclusions fit together as neatly as one could possibly wish. Likewise we can now date Chessell Down grave 45, with its inscribed copper-alloy pail, on the basis of the form of Kentish disc brooch that was buried on the costume of the woman in this grave, to our phase AS-FB, a phase which the national chronological project could estimate started cal AD 510‒545 (95 % probability) and ended cal AD 555‒585 (95 % probability).2 While seeming marginally less precise than the 2 Because of the focus of the case-study presented in Hines 2021, the estimated dates of the phaseboundaries of this phase have not simply been recalculated using IntCal20. It can, though, be reported that the boundary between phases AS-FB and AS-FC may be moved a decade later to cal AD 565–595 (95 % probability).

The Archaeological Dating of the Early Finds

67

Tab. 1: Female graves or artefact-types with runic inscriptions dated by the national chronological scheme, cited at 95 % probability. The dates from 1990 for the inscribed bowls in Chessell Down grave 45 and Cleatham grave 20 are deposition dates based upon associated brooch-types. Object

Date assigned in Hines (1990)

Dating evidence

Date for PRODUCTION unless marked with asterisk

DEPOSITION Date assigned in Hines/Bayliss (2013) Date assigned (from)

Date assigned (up to)

Buckland Dover grave 126, disc brooch

c. AD 575–625

Brooch-type (BR2-b4)

cal AD 525–550 (95 % probability)

cal AD 625–650 (95 % probability)

Harford Farm grave 11, disc brooch

–

Graveassemblage (AS-FE)

cal AD 625–650 (95 % probability)

cal AD 660–680 (95 % probability)

Wakerley grave 80, great squareheaded brooch

c. AD 525–560

Bead-types Start date in gravenot defined assemblage in the scheme. (BE-ConCyl and BE-ConSeg: Brugmann 2004, Group A2)

cal AD 555–585 (95 % probability)

West Heslerton c. AD 500–550 grave 177, cruciform brooch

Graveassemblage (AS-FB) and radiocarbon date

cal AD 520–565 (95 % probability)

Chessell Down grave 45, pail

c. AD 520–570*

Brooch-types (Kentish squareheaded and disc brooch BR2-b2)

cal AD 510–545 (95 % probability)

cal AD 555–585 (95 % probability

Cleatham grave 20, hanging bowl

‘late 6th or 7th century’*

Brooch-type (BR3-b)

cal AD 555–585 (95 % probability)

cal AD 660–680 (95 % probability)

Broughton Lodge, Willoughbyon-the-Wolds, grave 5

“late 5th or 6th century”

Brooch-type (small long) in graveassemblage

–

–

Welbeck Hill, grave 14

“well within the 6th century but not later than circa 570”

Bracteate-type

–

–

Boarley, disc brooch

–

Brooch-type (BR2-b4)

cal AD 525–550 (95 % probability)

cal AD 625–650 (95 % probability)

68

John Hines

c. AD 520‒570 I previously suggested, the congruency is highly satisfying and the mathematical robustness of the new results of the greatest importance. The burial with an inscribed hanging-bowl at Cleatham, North Lincolnshire, grave 20, is dated by an annular brooch of a type that first appears in phase AS-FB, but is most characteristic – especially in the same area of South Humberside in fact – in phase AS-FC, which began in cal AD 555–575 (68 % probability) or cal AD 555–585 (95 % probability): as was noted before, this burial might be late sixthcentury, but a seventh-century date seems more probable (Hines 1990, 444‒5; Leahy 2007). As of yet, we cannot incorporate the women’s graves at Broughton Lodge (Willoughby-on-the-Wolds), Nottinghamshire, grave 5, or Welbeck Hill, South Humberside/North Lincolnshire, grave 14, directly with the sequence; in the former case because it lies beyond its range, and in the latter case because the available data are insufficient, but future developments of the framework may rectify that. The Boarley disc brooch from Kent, meanwhile (Parsons 1992; 1999, 46‒7), which is not from a known burial context, like the Buckland Dover grave 126 brooch, is a type which may have been introduced a little earlier than previously thought, although it must be emphasised that the date “from” cited is a terminus post quem, not a date of introduction. This brooch-type was principally current in the late sixth and first of quarter of the seventh century according to our scheme. The Undley bracteate is also a female dress-accessory of a type that lies beyond the scope of the scheme at present, but Morten Axboe’s detailed work on bracteate chronology has strongly corroborated the dating within the second half of the fifth century that has been assigned to this bracteate since it was found around 40 years ago (Axboe 2004; Hines 1984, 204‒9). In that case, however, confidence in the dating does not decisively answer the runologically crucial question of where the bracteate was made.3 There are fewer runic inscriptions on male-associated equipment such as weaponry. This can partly be explained by the fact that for a period of about fifty years in the seventh century burial customs diverged, and the female graves with grave goods massively outnumber male ones, by a factor of around 6:1. That, however, probably is not the sole determining factor, and we do seem to have a real genderdifference in the distribution of runic inscriptions as artefact-attributes from the second half of the sixth century onwards (see Table 2). Four male-associated objects and/or burials with runic inscriptions belong to our phases AS-MA and AS-MB, the latter ending in cal AD 550–560 (68 % probability) or cal AD 545–565 (95 % probability). The boundary between AS-MA and AS-MB is set at cal AD 525–545 (68 % probability) or cal AD 525–550 (98 % probability); we have not defined a starting

3 The continuing finds of die-linked inscribed bracteates from Binham in Norfolk add new evidence of direct relevance to this question (see Pestell, this vol.; Behr/Pestell/Hines 2014). This is not the place to discuss their implications, but I would note that the additional evidence does not, in my judgment, conclusively settle the question of the origin of the Undley bracteate.

The Archaeological Dating of the Early Finds

69

Tab. 2: Male graves or artefact-types with runic inscriptions dated by the national chronological scheme, cited at 95 % probability. The dates from 1990 for the inscribed purse-mount in Watchfield grave 45 is a deposition date based upon associated types of belt-fitting. Object

Date assigned in Hines 1990

Dating evidence

Date for PRODUCTION unless marked with asterisk

DEPOSITION Date assigned in Hines and Bayliss 2013 Date assigned (from)

Date assigned (up to)

Chessell Down, grave 76, scabbard mount

c. AD 475–525

Scabbard mount-type (SW6-c)

Start date not defined in the scheme.

cal AD 525–550 (95 % probability)

Sarre, grave 91, sword pommel

–

Sword pommel-type (SW2-a)

Start date not defined in the scheme.

cal AD 525–550 (95 % probability)

Watchfield, grave 67, purse-mount

c. AD 520–570*

Types of beltfitting in graveassemblage (BU2-d and BU2-h)

cal AD 525–550 (95 % probability)

cal AD 545–565 (95 % probability)

Gilton (Ash), sword pommel

–

Sword pommeltype (SW2-b)

cal AD 525–550 (95 % probability)

cal AD 545–565 (95 % probability)

boundary for AS-MA.4 Phase AS-MB thus barely covers a human generation. By type, the scabbard mouthpiece from Chessell Down, Isle of Wight, grave 76 and the sword-pommel from Sarre, Kent, grave 91, belong to AS-MA, while the shield boss and belt-fittings associated with a purse-mount in Watchfield, Oxfordshire, grave 67, date this burial to AS-MB, the phase to which the type of sword-pommel represented by the Ash/Gilton inscription also belongs. In a paper on early runic inscriptions from Kent, I argued for moving the dates of these pommel-types rather later than in Wilfried Menghin’s general study Das Schwert im frühen Mittelalter of 1983 (Hines 2006). While I stand by the validity of the arguments I used there, it is clear that the results of our dating project imply that I was nonetheless suggesting alternative dates which may have been 20‒30 years too late. I am glad to have this opportunity to rescind that on the basis of the results of the research underlying the introduction of the new, stronger and larger, framework.

4 Again (see n. 3), the conditions of remodelling presented in Hines 2021 do not allow for simple new dating estimates to be presented; not least because there is now a defined and modelled phase AS-MA. It can, though, helpfully be reported that the boundary between phases AS-MB and AS-MC may be broadened a little to cal AD 540–570 (95 % probability), and that the inference of an apparently short duration of AS-MB remains unchanged.

70

John Hines

For good and justified scientific and practical reasons, the project was initially designed to cover the period from the late sixth century to the early eighth. In fact it proved both feasible and necessary to extend the scheme earlier, to c. AD 520/530, and we can now regard it as possible to apply essentially the same techniques to finds back to the Roman/Anglo-Saxon transition in the fifth century. In itself, that will be another major project. The availability of graves going well back into this period at, for instance, Mucking in Essex and Ringlemere in Kent provides one part of the necessary evidence; this work must also, however, meet the challenges of dating cremation burials – using typology, radiocarbon dating, and perhaps a new “rehydroxylation” process for pottery itself. Meanwhile the analysis and final publication of the major cremation cemetery at Spong Hill has meant another great step forward (Hills/Lucy 2013). We can now assign the three cremation urns marked with a common runic stamp from Spong Hill to Phase B at that site, around the middle of the fifth century (Hills, this volume), while the urn N59 at Caistor by Norwich, with the inscribed roe deer astragalus, is decorated in a style typical of Phase A at Spong Hill, and so to be assigned to around the second quarter of that century. The inscribed cremation urn from Loveden Hill (A11/251) is regrettably too plain to be datable by these means.

3 Dating the Ruthwell Cross A relevant and helpful way to round off discussion in this paper should be to review and discuss the problem of dating one especially important object with an Old English runic inscription: the Ruthwell Cross. Especially if we include runic coins, the quantity of runic material from England dating from the late seventh to mid-eighth centuries is relatively large. This sample represents a period immediately post-dating both what archaeologists recognise as the Early Anglo-Saxon Period – the period of regular deposition of artefacts with human burials – and what runologically and linguistically is now recognised as the “Pre-Old English” Period: a period in which the definitive sound-changes that distinguished Old English from the other West Germanic languages were underway but before they were “phonologized” (Waxenberger 2017; forthc.). Within that sample, however, four items stand out as particularly noteworthy, variously because of their material significance or because of the extensive and informative character of the runic texts upon them. St Cuthbert’s coffin is one of these, and this should be precisely datable because of the historically recorded translation of Cuthbert’s remains in AD 698. On that object, however, runes are used to write Latin words and a Greek monogram, and their evidence can shed no direct light on Old English linguistic developments (Bonner et al. 1989, esp. 231‒301). The other items referred to here include the Franks, or Auzon, Casket, discussed by both Webster and Waxenberger in the present volume, and the Bewcastle and Ruthwell

The Archaeological Dating of the Early Finds

71

Crosses. The Bewcastle Cross unfortunately has such a badly weathered runic inscription that little can be made of it in runological and philological discussions (Page 1960; Bailey/Cramp 1988, 61‒72). The Ruthwell Cross, by contrast, although incomplete, provides extensive legible runic texts, and the dating of those inscriptions is vital to a reliable mapping of the historical development of Old English. In terms of the procedures discussed above in the present paper, the Ruthwell Cross neatly encapsulates the difficulty of studying an object with at best a weak archaeological context. To a degree, this problem is characteristic of stone sculpture. Carved stone can occur, fragmentarily, in stratified archaeological layers; it may alternatively either be associated with or re-used as building material within a datable structure; but often it is not (Cramp 1984, xlvii‒xlviii). In many cases it has somehow survived as a monument, either re-located, or in situ but with the context surrounding it having changed entirely over the centuries, completely effacing the historical and material circumstances to which it originally belonged. In the case of this carved cross, its destruction and burial by Puritan iconoclasts in the midseventeenth century are well recorded, as are the stages of its subsequent exhumation, reconstruction and exhibition. It now stands incomplete and damaged, and questions have been raised as to whether it has been correctly put back together. Such concerns, however, are minor issues in relation to the linguistic and runological value of the Cross and the assessment of its probable date (see most recently Ó Carragáin 2005; Orton/Wood 2007). A more serious issue is the question of whether in this case the runic inscriptions were planned as part of the scheme of the cross from the beginning or were added later. Meticulous consideration of this question recently by Éamonn Ó Carragáin (2005) and Fred Orton and Ian Wood (2007) appears to have resolved the question in favour of the texts being linked to the design and carving of the monument ab initio, but it remains important to be aware of this question and of the arguments relevant to it. We can nonetheless proceed on the basis that whatever evidence and arguments can be brought to bear to date the cross itself can be postulated to date the inscription as well. Determining the date of the Ruthwell Cross apart from its vernacular texts in runes is not, however, an easy matter. We can be sure of a historical terminus post quem defined by a horizon in which Northumbria was converted to Christianity, and its Anglo-Saxon culture and language extended to the relevant area north of the Solway Firth, now Dumfriesshire. Both of those developments can be associated with the reign of King Edwin, c. AD 616/17–633/4 (Bede, HE, II.5, 9 and 16; Kirby 1991, 77‒88), most probably within the last few years of that period. For a long time, in fact, a late seventh-century date was considered plausible for both the Bewcastle and Ruthwell Crosses, because of their similarity to one another, and because the Bewcastle inscriptions were read as referring to a woman named Cyneburh and a man Alhfriþ: the known names of a king and his queen of the third quarter of the seventh century (Brown 1921, 197‒202, 245‒72 and 305‒17). Ray Page, however (in

72

John Hines

Bailey/Cramp 1988, 61‒2 and 65), showed how uncertain the reading of this inscription is, and in any event the personal name and name-elements identifiable are sufficiently common that direct association with that historical couple is by no means necessary. An assessment of the probable date of the Ruthwell Cross (it is noted again, independent of any judgments as to the probable age of the inscriptions) has therefore to be based largely on the sculptural and decorative styles represented by it. As has been helpfully and cogently surveyed and summarised by Richard Bailey (1996, 23‒57), a quantity of sculpted material securely datable to the second half of the seventh century is known, primarily from Northumbria, and, while this takes diverse forms, there is little in it to suggest that the Ruthwell and Bewcastle Crosses belonged to exactly the same milieu. Admittedly, the evidence might still allow for a specialised local development of large free-standing crosses such as those at Ruthwell and Bewcastle in the north-western corner of Northumbria at the same early date, but there is simply a more realistic space, and indeed more convincing associations, for these crosses in the eighth century. In northern England, the art of Christian sculpture and in particular the carving and raising of stone crosses extends from this period across the period of the Viking raids, conquests and settlements into an “Anglo-Scandinavian Period” largely datable to the tenth century (Bailey 1980). Although we can be sure of the overall chronological extent and the internal continuity of this tradition, it is practically impossible to put precise calendrical termini post or ante quos to any individual object or typological development within the series. The expulsion of the Viking Norse settlers from Dublin in AD 903 tends to be clung to as a handhold and a terminus post quem for the Norse colonisation of the north-west of England and thus for the rich Anglo-Scandinavian sculptural phase here, but the completely determinative significance of that event is not beyond question. We are forced, therefore, to turn from the convenience and reassurance of the calendrical figures of absolute chronology to the relative-chronological terms of sequences of typological development, both in the forms of the sculptural monuments and in their decorative arts-styles. In terms of the sculptural form, there are abundant parallels to confirm that the rectangular cross-section of the shaft of the Ruthwell Cross, and the curvilinear edging of the remaining fragments of the cross-head, are characteristic of the pre-AngloScandinavian phase of Northumbrian sculpture. This relative dating is confirmed by differences in decorative style, with motifs and forms appearing on the sculpture that not only fall into sequences of development and stylisation amongst themselves, affected in due course by the introduction of distinctly Scandinavian style-elements, but which also find parallels, with datable contexts or associations in some cases, in other media: primarily manuscript art in the case of the Ruthwell Cross, but also in metalwork and in various forms of carved bone material (Bailey/Cramp 1988, 12‒23). This is not the place to review all of those analogues in great detail: the most significant details in the present case are close parallels between panels of double-strand

The Archaeological Dating of the Early Finds

73

interlace in the Durham Cassiodorus manuscript and on the Bewcastle Cross, and in types of leaf used decoratively on both the Ruthwell and Bewcastle Crosses and in the Saint Petersburg Bede manuscript. The two manuscripts referred to here can in turn be linked to one another as apparently very close in date by a form of writing which is not only palaeographically of one style but in some areas is plausibly the writing of a single scribe (Wilson 1984, 63). The St Petersburg (formerly Leningrad) Bede is one of two manuscripts of Bede’s Historia Ecclesiastica Gentis Anglorum that have been assigned to a date within fourteen years of Bede’s death in AD 732 on the basis of marginal annotations that calculate how many years previously events Bede refers to took place. Although not completely consistent, these point to a date of AD 734‒7 for the Moore Bede and AD 746 for the St Petersburg manuscript. Some palaeographers regard these dates as suspiciously early for the style of writing in the manuscripts, and the principal argument for not accepting them as evidence for a secure absolute date is the fact such memoranda could demonstrably have been copied, unaltered, at a later date than when they were originally calculated. In the case of the St Petersburg Bede, David Dumville (2007) has argued strongly that one entry here which adds 132 years to an event dated to the year 626 gives a terminus post quem of AD 758 to the codex, and Richard Gameson (2015; see Webster, this volume) has accepted this reasoning. Those two scholars represent weighty authority and expertise that I could not possibly challenge, except that on this issue the question is forensic rather than palaeographical. The figure in question (written in Roman numerals as cxxxii) is a later addition over an erasure, presumably either to correct an error or to clean up a smudged entry and has also been rewritten by a later “restorer” (Dumville 2007, 70 n. 50). Of the other entries, two are sums which give AD 745 while eight yield AD 746. Four entries are damaged and cannot give precise results, but all have termini post quos a little before 746: AD 729, 739, 741 and 745 respectively. One entry produces the egregious sum of AD 861. The difference of twelve years is not of huge significance in the current context, but it does not appear that we should be too dismissive of the considerable body of palaeographical and art historical opinion that a date immediately before the middle of the eighth century is entirely credible for the St Petersburg Bede (Arngart 1973; O’Donnell 2005, 169‒73; cf. also Kiernan 1990). Even if we do allow ourselves to continue to use this date as a reference point, however, these parallels between the two crosses and the two illuminated manuscripts fall straightforwardly into the category of “broad contemporaneity” as it was defined above: there is no reason to assume that the motifs which link them could only have moved from manuscript illumination to sculpture, or vice versa, or that the surviving evidence gives us anything more than the most fragmentary glimpses of the original distribution and range of examples of such artwork. All the same, it is easy to see how the specific calendar date with which the St Petersburg Bede is associated justifies a transferred date of “circa AD 750” for the Ruthwell Cross (Waxenberger 2006, 285‒7). Waxenberger in fact refers to a then forthcoming study

74

John Hines

of other runic inscriptions found in Dumfriesshire by Ray Page, in which it appears he planned to explain more fully his final opinion that this was the plausible date for the cross and its inscription. If that was his intention, however, unfortunately it was not realised (Page 2006). From the perspective of archaeological chronology, the crucial question is just what does that “circa” imply? What are the realistic outer limits of the range of uncertainty around the middle of the eighth century that it denotes? As already noted, the corpus of Northumbrian sculpture datable to the seventh century does not leave room for any serious likelihood that the Ruthwell and Bewcastle Crosses could be dated so early. A striking feature of the symbolic decoration of these crosses is the vine-scroll: a motif that was used in markedly similar forms in two major churches in the Near East, in Jerusalem and Damascus respectively, in the period AD 691‒705. Much emphasis has also been placed on the iconographical meaning of the Ruthwell Cross and its probable immediate background. Éamonn Ó Carragáin has identified plausible dependency on a modification of the mass under Pope Sergius (ca. AD 650‒701), the end of whose papacy was also marked by the discovery of a relic of the True Cross in the Vatican: an event Bede attached much importance to (Ó Carragáin 1999; 2003). In this context, we might choose to be strictly critical of the argument made by David Howlett (1992) that a congruency between the iconographic scheme on the south face of the Ruthwell Cross and Bede’s Commentary on Luke’s gospel of AD 709‒16 implies that that datable text pre-dates the representation of the same ideas on the Cross rather than both separately representing the same currents in Christian understanding. Ó Carragáin, for instance, prefers a liturgical to an exegetical basis for the collocation of these scenes (1999, 196‒7; 2003, esp. 140‒55). In what we know of the general context, it is also difficult to ignore the tradition, finally recorded at the beginning of the twelfth century by Symeon of Durham, that Æthelwald, Bishop of Lindisfarne from c. AD 724‒740, had a special carved cross raised at Lindisfarne – presumably during his episcopacy, although the use of the pluperfect fecerat in the Latin text following a reference to his earlier period as a monk at Lindisfarne could indicate that this was believed to have been done earlier in the eighth century (Symeon, LE, I.12). Attention is also caught by Bede’s note that Whithorn had been elevated to the status of episcopal see within the Northumbrian church beyond the Solway “recently” – i.e., shortly before AD 732 (Bede, HE, III,4 and V,23; Mac Lean 1992). That event at least renders it plausible that other centres of the Church in this region would have seen constructive investment around the same time. The sum import of all of the above is that any date from c. AD 720 onwards could certainly be considered entirely credible for the Ruthwell Cross; indeed AD 720 itself is by no means a terminus ante quem non for the carving. But the carving of the Ruthwell Cross might equally well be later than AD 750, and here in fact it is much more difficult to identify fixed points that define the probable boundaries with a calendrical date. We can, however, at least place the Ruthwell and

The Archaeological Dating of the Early Finds

75

Bewcastle Crosses at the first phase of a stylistic sequence of Northumbrian carved crosses, along with Acca’s cross at Hexham, which has details paralleled in a dated Italian building of AD 737x756 (Cramp 1984, 74‒6; Bailey 1996, 45). The Rothbury cross-fragments exemplify the development that added a Classical figure style which was adopted and promoted in Carolingian art to this sculptural tradition around the end of the eighth century (Cramp 1984, 27‒8 and 217‒21). The eighthcentury sequence of cross-fragments known from Hexham implies that the Ruthwell Cross would be increasingly anachronistic in style as the second half of the eighth century progressed were it indeed carved at such a date. The problem with an expression like “circa AD 750” for an object such as the Ruthwell Cross is that the single calendrical date given takes on a rather more fixed value than is justified: the expression does not convey how great a margin of variability we should allow for around that. All of the evidence that has been reviewed here means that we can express the date of the Ruthwell Cross as something that we can only assign with true confidence to “some time in the eighth century” or “AD 700–800”. However, we can explain why the outer limits of that particular range, at the very beginning and end of the century, are quite improbable even if still possible. Can we devise any satisfactory way of reducing an expression of a reliable dating of the Ruthwell Cross to some intermediary point between the overprecision of the single approximate date in the middle of the eighth century and the wide lassitude of simply referring the cross to somewhere within the century as a whole? Ó Carragáin has increasingly preferred a date-range of c. AD 730‒760 for the cross (e.g., Ó Carragáin 2003), based upon the establishment of the see of Whithorn as a relevant historical context, and the gradual adoption of a Roman Marian liturgy in England between the deaths of Aldhelm in AD 709 and Bede in AD 732 (cf. Clayton 1990, 20‒30). In both cases, Bede’s death in fact operates as a terminus ad quem. While this three-decade range is better than the single date of “circa AD 750” which it encompasses, it is still necessarily a subjective judgment of probability – without the mathematical measurement of reliability that the statistics of radiocarbon dating and Bayesian modelling provide. The year 730 is in fact the date by which Ó Carragáin is sure all the preconditions for the raising of the cross have been met: it is not a terminus post quem before which the cross could not have been raised. And the year 760 can only be a “guesstimate” of an appropriate date after which one would expect the waves of intellectual and spiritual interest and creativity, and of material investment, represented by the cross to have waned. To say as much is emphatically not to reject Ó Carragáin’s preferred dating as implausible. It is rather to insist, perhaps rather pedantically, upon the strict classification and evaluation of the chronological terms he uses from an archaeological perspective. In that light, we might propose reformulating the dating somewhat – but really primarily in terms of its expression or even style rather than very much in substance. The absolute termini post and ante quos we can argue for in the case of the Ruthwell Cross are too broad to be useful: e.g., ca. AD 700 and 800 respective-

76

John Hines

ly, as explained above. In between, we have associable fixed dates at shortly before AD 732 and at (or around) AD 746, but those are not termini post or ante quos either. Where close absolute dates are lacking, palaeographers are in the habit of dividing centuries into convenient blocks: half-centuries, sometimes quarter-centuries, and even decades in cases of extreme self-confidence. In this case, we could take the middle of that range as the practical units to refer to. Thus, we can confidently suggest that the greatest concentration of probability lies in the quarter-century 725‒750, and the next greatest in the quarter-century 750‒775. We cannot go as far as thinking in terms of a level of confidence strictly comparable to the 1σ range (68 % probability) for the former, and up to 2σ (95 % probability) for the whole period 725‒775, but the analogy may not be completely misleading, and indeed it may be conceptually helpful to reflect upon the situation in comparable terms. If, for instance, we were to judge that it is at least as likely that the Ruthwell Cross dates from the period 725‒750 as either before or after those years, we would assign at least 50 % of the probability to the second quarter of the eighth century. If we consider the quarter-century 750‒775 to be half as likely as the second quarter, and the first and last quarters of the century to have an equal share of the remaining probability, then we have distributed the probability in these proportions: 700‒725: 725‒750: 750‒775: 775‒800:

12.5 % 50 % 25 % 12.5 %

It is emphasised that these figures are used illustratively, and not seriously proposed as a scheme to be treated as an arithmetically secure conclusion. As an illustration, however, this can be claimed to be both entirely reasonable and indeed really quite cautious. In the present author’s opinion, for instance, it is more likely that the Ruthwell Cross dates to a little before AD 725 than to a little after AD 775, but one could only fiddle with what are already artificial numbers to try to reflect that in these terms. Altogether though, this may be defended as a suitable explanation of a very summary dating of the Ruthwell Cross, on art historical and archaeological grounds, as “probably second quarter of the eighth century”. Within archaeology, there are circumstances in which one can simply not worry about the calendrical date-scale, and can discuss relative chronologies in terms of a sequence of phases. Something similar is also possible in language history, with pre-Old English, Early Old English, Late Old English etc., and runologically, with our older, younger and medieval fuþarks. The convenient (albeit not the only possible) currency for coordinating these, not only with one another, but also with any other historical phenomena, is of course calendrical dating. That, however, is demonstrably a difficult and risky area: difficult, not just because we often do not have sufficient evidence to bring us to the correct answer, but also because what

The Archaeological Dating of the Early Finds

77

clues we do have do not necessarily generate straightforward chronological expressions to reflect the distribution of probability and its limits. Let us therefore note again, finally, that this case-study of the Ruthwell Cross and the problems of its dating has been undertaken without any reference to the runic and linguistic evidence itself. There is an obvious reason for that, when we approach the source as runologists and philologists wanting an independent dating of the graphic and linguistic features it bears. Of course we must not simply ignore the fact that there are linguistic features which we can consider as markedly “early” rather than “late”, particularly the regular preservation of unstressed i in the prefix gi- and the particle ni. It is valid to combine these with the balance of the archaeological, art historical and historical evidence above to prefer a relatively early date for the Ruthwell Cross. One might even argue that the addition of this evidence into a whole assessment of the monument moves the distribution of probability over the second quarter and second and third quarters of the eighth century together closer to the 1σ and 2σ analogy just noted. But it does not at present help to resolve the problem of determining a precise calendrical date-range and the distribution of probability within it. In the case of runic inscriptions and material culture, it can be anticipated that the future accumulation of the newly uncovered, datable evidence will gradually provide us with further parallels that will gradually refine and narrow the probable date-range of the Ruthwell Cross. Specialists in the disciplines of archaeology and runology alike are familiar with the need to evaluate a diversity of probabilistic inferences concerning the objects of their studies around a certain core of definite facts – at the same time, usually, as non-specialists are asking them for a definite answer to questions such as “what is it?”, “how old is it?”, and “what does it say?”. Both groups of specialists should, therefore, have understanding for the complexities of each other’s field, and that is probably as good a basis as any for the future collaboration that will be essential for the strongest possible progress from this point onwards.

References Arngart, O. S. 1973. “On the Dating of the Early Bede Manuscripts”. Studia Neophilologica 45: 47‒52. Axboe, Morten. 2004. Die Goldbrakteaten der Völkerwanderungszeit ‒ Herstellungsprobleme und Chronologie. RGA-E 38. Berlin: De Gruyter. Bailey, Richard N. 1980. Viking Age Sculpture in Northern England. London: Batsford. Bailey, Richard N. 1996. England’s Earliest Sculptors. Toronto: Pontifical Institute of Mediaeval Studies. Bailey, Richard N. and Rosemary J. Cramp (eds.). 1988. The Corpus of Anglo-Saxon Stone Sculpture. Volume II: Cumberland, Westmorland and Lancashire North-of-the-Sands. Oxford: Oxford University Press for the British Academy.

78

John Hines

Baillie, Mike. 1999. From Exodus to Arthur: Catastrophic Encounters with Comets. London: Batsford. Bakka, Egil. 1959. “On the Beginnings of Salin’s Style I in England”. Universitetet i Bergen Årbok 1958, Historisk-antikvarisk rekke 3: 1‒83. Bammesberger, Alfred. 2003. “The Harford Farm Brooch Runic Inscription”. Neophilologus 87: 133‒135. Bede, HE: Historia Ecclesiastica Gentis Anglorum = Colgrove, Bertram and Roger A. B. Mynors (eds.). 1969. Bede’s Ecclesiastical History of the English People. Oxford: Clarendon. Bonner, Gerald, David Rollason, and Clare Stancliffe (eds.). 1989. St Cuthbert, His Cult and Community to AD 1200. Woodbridge: The Boydell Press. Behr, Charlotte, Tim Pestell, and John Hines. 2014. “The Bracteate Hoard from Binham – an Early Anglo-Saxon Central Place?” Medieval Archeology 58: 44‒77. Bowman, Sheridan. 1990. Radiocarbon Dating. London: British Museum Press. Brown, Gerard Baldwin. 1921. The Arts in Early England. Volume 5: The Ruthwell and Bewcastle Crosses, The Gospels of Lindisfarne, and Other Christian Monuments of Northumbria. London: John Murray. Clarke, David L., rev. Bob Chapman. 1978. Analytical Archaeology. London: Methuen. Clayton, Mary. 1990. The Cult of the Virgin Mary in Anglo-Saxon England. Cambridge: Cambridge University Press. Cramp, Rosemary J. (ed.). 1984. The Corpus of Anglo-Saxon Stone Sculpture. Vol. I: County Durham and Northumberland. Oxford: Oxford University Press for the British Academy. Douglas, Rev. James. 1793. Nenia Britannica. London: John Nichols. Dumville, David N. 2007. “The Two Earliest Manuscripts of Bede’s Ecclestiastical History?” AngloSaxon 1: 55–108. Gameson, Richard. 2015. “Materials, Text, Layout and Script”. The St Cuthbert Gospel: Studies on the Insular Manuscript of the Gospel of John. Eds. Claire Breay and Bernard Meehan. London: British Library. 13–39. Geake, Helen. 1997. The Use of Grave Goods in Conversion-Period England c600‒c850. Oxford: British Archaeological Reports. Harris, Edward C. 1979. Principles of Archaeological Stratigraphy. London: Academic Press. Haughton, Christine and Dominic Powlesland. 1999. West Heslerton: The Anglian Cemetery. 2 vols. Yedingham: The Landscape Research Centre. Heizmann, Wilhelm and Morten Axboe (eds.). 2011. Die Goldbrakteaten der Völkerwanderungszeit ‒ Auswertung und Neufunde. RGA-E 40. Berlin: De Gruyter. Hills, Catherine M. and Sam Lucy. 2013. The Anglo-Saxon Cemetery at Spong Hill, North Elmham, Volume IX: Chronology and Synthesis. Cambridge: McDonald Institute for Archaeological Research. Hines, John. 1984. The Scandinavian Character of Anglian England in the Pre-Viking Period. Oxford: British Archaeological Reports. Hines, John. 1990. “The Runic Inscriptions of Early Anglo-Saxon England”. Britain 400‒600: Language and History. Eds. Alfred Bammesberger and Alfred Wollmann. Heidelberg: Winter. 437‒455. Hines, John. 2006. “The Early Runic Inscriptions from Kent and the Problem of Legibility”. Das fuþark und seine einzelsprachlichen Weiterentwicklungen. Eds. Alfred Bammesberger and Gaby Waxenberger. Berlin: De Gruyter. 188‒208. Hines, John. 2019. “Practical Runic Literacy in the Late Anglo-Saxon Period: Inscriptions on Lead Sheet”. Anglo-Saxon Micro-Texts. Eds. Ursula Lenker and Lucia Kornexl. Berlin: De Gruyter. 29–60. Hines, John. 2020. “New Insights into Early Old English from Recent Anglo-Saxon Runic Finds”. NOWELE 73.1: 69–90.

The Archaeological Dating of the Early Finds

79

Hines, John. 2021. “The Chronological Framework of Early Anglo-Saxon Graves and Grave Goods: New Radiocarbon Data from RAF Lakenheath, Eriswell, Suffolk, and a New Calibration Curve (IntCal20)”. The Antiquaries Journal 101: 1–37. Hines, John and Alex Bayliss (eds.). 2013. Anglo-Saxon Graves and Grave Goods of the 6th and 7th Centuries AD: A Chronological Framework. Leeds: Society for Medieval Archaeology. Høilund Nielsen, Karen and Siv Kristoffersen. 2002. “Germansk dyrestil (Salins stil I‒III): et historisk perspektiv”. Hikuin 29: 15‒74. Holst, Mads K. 1999. “The Dynamic of the Iron-Age Village: A Technique for the RelativeChronological Analysis of Area-Excavated Iron-Age Settlements”. Journal of Danish Archaeology 13: 95‒119. Howlett, David. 1992. “Inscriptions and Design of the Ruthwell Cross”. The Ruthwell Cross. Ed. Brendan Cassidy. Princeton NJ: Department of Art and Archaeology, Princeton University. 71‒93. Jarman, Catrine L., Martin Biddle, Tom Higham, and Christopher Bronk Ramsey. 2018. “The Viking Great Army in England: New Dates from the Repton Charnel”. Antiquity 92: 183‒99. Kiernan, Kevin S. 1990. “Old English Manuscripts: The Scribal Deconstruction of ‘Early’ Northumbrian”. American Notes and Queries 3: 48‒55. Kirby, David P. 1991. The Earliest English Kings. London: Unwin Hyman. Leahy, Kevin. 2007. ‘Interrupting the Pots’: The Excavation of Cleatham Anglo-Saxon Cemetery. York: Council for British Archaeology. Leeds, Edward Thurlow. 1936. Early Anglo-Saxon Art and Archaeology. Oxford: Clarendon. Mac Lean, Douglas. 1992. “The Date of the Ruthwell Cross”. The Ruthwell Cross. Ed. Brendan Cassidy. Princeton, NJ: Department of Art and Archaeology, Princeton University. 49‒70. Malmer, Mats P. 1963. Metodproblem innom järnålderens kunsthistoria. Lund: Gleerups. Metcalf, D. Michael. 1993‒4. Thrymas and Sceattas in the Ashmolean Museum Oxford. 3 vols. Oxford: Royal Numismatic Society and the Ashmolean Museum. Nedoma, Robert. 2011. “Personennamen in älteren Runeninschriften auf Fibeln”. Language and Literacy in Early Scandinavia and Beyond. Eds. Michael Schulte and Robert Nedoma. NOWELE 61/62. Odense: University Press of Southern Denmark. 31‒89. Ó Carragáin, Éamonn. 1999. “The Necessary Distance: Imitatio Romae and the Ruthwell Cross”. Northumbria’s Golden Age. Eds. Jane Hawkes and Susan Mills. Stroud: Sutton. 191‒203. Ó Carragáin, Éamonn. 2003. “Between Annunciation and Visitation: Spiritual Birth and the Cycles of the Sun on the Ruthwell Cross: A Response to Fred Orton”. Theorizing Anglo-Saxon Stone Sculpture. Eds. Catherine E. Karkov and Fred Orton. Morgantown, WV: West Virginia University Press. 131‒187. Ó Carragáin, Éamonn. 2005. Ritual and the Rood: Liturgical Images and the Old English Poems of the Dream of the Rood Tradition. London/Toronto: The British Library and University of Toronto Press. O’Donnell, Daniel. 2005. Cædmon’s Hymn: A Multi-Media Study, Edition and Archive. Cambridge: D. S. Brewer. Orton, Fred, Ian Wood, and Claire Lees. 2007. Fragments of History: Rethinking the Ruthwell and Bewcastle Monuments. Manchester: Manchester University Press. Page, Raymond Ian. 1960. “The Bewcastle Cross”. Nottingham Mediaeval Studies 4: 36‒57. Page, Raymond Ian. 2006. “Bone Fragment with Runic Inscription: 2252”. The Mote of Mark: A Dark Age Hillfort in South-West Scotland. Eds. Lloyd Laing and David Longley. Oxford: Oxbow. 92‒93. Parsons, David N. 1992. “German Runes in Kent?” Nytt om Runer 7: 7‒8. Parsons, David N. 1999. Recasting the Runes: The Reform of the Anglo-Saxon Futhorc. Uppsala: Institutionen för nordiska språk.

80

John Hines

Penn, Kenneth. 2000. “Excavations on the Norwich Southern Bypass, Part 2: The Anglo-Saxon Cemetery at Harford Farm, Caister St Edmund, Norfolk”. East Anglian Archaeology 92. Gressenhall: Norfolk Museums Service. Petrie, W. M. Flinders. 1893. “Sequences in Prehistoric Remains”. Journal of the Royal Anthropological Institute of Great Britain and Ireland 29: 295‒301. Reimer, Paula et al. [41 other cited authors]. 2020. “The IntCal20 Northern Hemisphere Radiocarbon Age Calibration Curve (0–55 cal kBP)”. Radiocarbon 62: 725–757. Salin, Bernhard. 1904. Die altgermanische Thierornamentik. Stockholm: Wahlström and Widstrand. Symeon of Durham: Libellus de Exordio atque Procursu istius, hoc est Dunhelmensis, Ecclesie. Ed. and transl. David Rollason. 2000. Oxford: Clarendon. Waxenberger, Gaby. 2006. “The Representation of Vowels in Unstressed Syllables in the Old English Runic Corpus”. Das fuþark und seine einzelsprachlichen Weiterentwicklungen. Eds. Alfred Bammesberger and Gaby Waxenberger. Berlin: De Gruyter. 272‒314. Waxenberger, Gaby. 2017. “How ‘English’ is the Early Frisian Runic Corpus? The Evidence of Sounds and Forms”. Frisians and their North Sea Neighbours from the Fifth Century to the Viking Age. Eds. John Hines and Nelleke IJssennagger. Woodbridge: The Boydell Press. 93–124. Waxenberger, Gaby. Forthc. A Phonology of Old English Runic Inscriptions with a Concise Edition and Analysis of the Graphemes. RGA-E. Berlin/Boston, MA: De Gruyter. Wilson, David M. 1984. Anglo-Saxon Art from the Seventh Century to the Norman Conquest. London: Thames and Hudson.

Lilla Kopár

Inscriptions on Stone Monuments: Methodological Concerns About one third of the surviving corpus of → Old English runic inscriptions has been preserved in the medium of stone.1 Most inscriptions are part of the original design of monumental sculpture, while others appear to be later additions. There are over 35 stone monuments with inscriptions in Old English runes, or traces thereof, in the British Isles, and further evidence of runic graffiti comes from Italy (Monte Gargano and Rome). The recent discovery of the graffiti-like inscription at Kirby Misperton in Yorkshire has further increased this number.2 All monumental inscriptions belong to the post-650 period, in contrast to the absence of inscriptions on stone in the pre-650 runic corpus3 (Page 1999, 27‒29, and Figs. 7‒8), and all of them show obvious links to an ecclesiastical context.4 In geographical terms, rune-inscribed stones tend to have a northern provenance (from the Wirral to Lincolnshire, through northern England, to southwest Scotland and the Isle of Man), with the exception of the Dover and Sandwich/Richborough stones, the only two southern examples. This may be evidence of a regional epigraphic tradition in northern England, in particular in a commemorative context, and a shift in runic literacy from the late seventh century onwards with growing ecclesiastical support and affiliation. Whether the relatively high number of inscriptions on stone is a reflection of contemporary production or the result of a higher survival rate of stone monuments compared to portable objects made of precious or perishable material is hard to tell. Nonetheless, the surviving evidence points to a strong tradition of inscribed memorial monuments in an ecclesiastical context in the north of England from the late seventh century onwards. Inscriptions on stone provide valuable information not only about runic literacy and linguistic development, but as part of complex artifacts, also about commemorative practices, and about the function, social context, and patronage of sculpture.

1 This count excludes manuscript evidence and coinage. 2 I am grateful to Martin Findell and Joanna Story for discussing the inscription with me and to Martin Findell for sharing a draft of their forthcoming publication (by Martin Findell, Joanna Story, and Dominic Powlesland). 3 The Sandwich/Richborough stone might be an exception with its early, but rather uncertain dating to the fifth to eighth centuries in Tweddle et al. (1995, 168‒170); seventh to eighth centuries in Parsons (1994, 317); and the early seventh century in Elliott (1989, 106) (similarly Bailey 1996, 24). 4 Ray Page (1999, 131) noted two examples only where an ecclesiastical origin is uncertain: the Falstone stone with a Christian memorial inscription and the Sandwich/Richborough stone with an undeciphered text, possibly a name. Both stones are of unusual shape or monument type and were found outside of a direct church context, but an ecclesiastical origin cannot be excluded in either case. https://doi.org/10.1515/9783110796834-005

82

Lilla Kopár

In the context of sculptural studies, inscriptions are just as valuable to archaeologists, art historians, and historians as they are to runologists, epigraphers, and historical linguists. But while the latter focus on the texts in their own right and only note their context and physical existence on the periphery, the former tend to look at the carriers of the inscriptions first, the objects in their entirety, and consider the inscribed texts as one of the many components of an artifact. The purpose of this essay is not to discuss the individual runic inscriptions on stones at length, or to evaluate them in epigraphic or linguistic terms but rather to offer a guide to the linguistically inclined runologist to the material context of the inscriptions on stone monuments. The artistic medium of stone sculpture adds a significant layer of meaning and information to the inscriptions and should thus be taken into consideration when interpreting the runic texts.

1 Stone Sculptures as Artifacts As in the case of many other objects traditionally grouped together by medium (as opposed to function or context, for example), the term ‘stone sculpture’ denotes a diverse group of artifacts that served a variety of purposes. In the context of early medieval England, the list includes elaborately carved monumental crosses, sarcophagi, and so-called hogbacks (recumbent monuments with a curved ridge) as well as architectural sculpture (capitals, baluster shafts, arches, friezes, sundials, etc.), stone furnishings, and simple grave markers.5 Many pre-Conquest stone monuments are elaborate artifacts, often more complex than other runic objects, such as coins, weapons, or bone fragments. Their dating and interpretation are dependent on the analysis of a combination of (in our case) inscriptions, decorative and figural carvings, shape and monument type, location and context, and state of preservation, which calls for different types of scholarly expertise, from art historical and archaeological to epigraphic and linguistic. When discussing the main characteristics of stone sculptures as artifacts, we cannot but start with stating the obvious: Most stone monuments are relatively large, heavy, and sturdy objects. These simple physical features have far-reaching consequences for identifying and understanding function, patronage, audience, and provenance. Although most of early medieval sculpture survives as fragments, stone carvings tend to be more permanent than portable objects made of perishable (or valuable) materials. The makers of stone monuments were certainly aware of this longevity, which should be taken into account when interpreting the function, audience, and intended permanence of inscriptions. Furthermore, compared to por-

5 For a classification of early medieval stone monuments, see Rosemary Cramp’s General Introduction to the Corpus of Anglo-Saxon Stone Sculpture (1984a).

Inscriptions on Stone Monuments: Methodological Concerns

83

table objects (coins, weapons, jewelry, books, liturgical objects, textile, etc.), stone sculptures are relatively immobile artifacts. Although very few monuments still stand in situ (the Bewcastle Cross is a famous exception to this rule), they have rarely been moved far from their original location. Many pre-Conquest carved stones were reused in the post-Conquest period in local building projects and are built into existing structures. Others that survived into the modern period found new homes in local churches and later in museums. The modern find spot and current location of most monuments can, therefore, only provide limited evidence for function and dating in most cases, but it can be assumed that most sculptures did not travel far from their original locations of manufacturing and/or display. If this is not the case, their recent moves are often reasonably well documented in antiquarian or museum records.6 This relative assurance of (at least regional) provenance is a significant advantage of inscriptions on stone for the study of dialects, local scribal and epigraphic traditions, and the context of literacy. As opposed to many high-profile objects of contemporary art and → material culture (e.g., jewelry, books) that were directed to a small and exclusive audience, most stone sculptures were public artifacts of a Christian cultural context,7 and their inscriptions, for the most part commemorative in nature, were intended for a wider audience than texts inscribed on personal objects. The three-dimensional nature of free-standing sculpture, often with more than one decorated face, creates both a sense of unity and division within the artifact. In order to comprehend the monument as a whole, the observer needs to walk around it, taking in only a limited view of the object at a time, yet always being aware of the existence (and significance) of the other sides and the monument as a whole. It is, therefore, important for the understanding of inscribed stones to examine each side, whether it contains an inscription or not. Non-inscribed surfaces are just as important parts of the whole as those bearing inscriptions. Inscriptions on stone monuments are sometimes accompanied by images (figural and decorative carvings, symbols, or other visual clues) that complement the verbal message of the text. These images are in dialogue with each other and so are the texts, even if they appear on different sides (faces) on the monument. As visual representations of verbal utterances, inscriptions themselves function as ‘images’ on the sculptures (and on other runic objects); their layout and design constitute part of the ‘messages’ they convey. As Leslie Webster (2012, 9) noted, inscriptions were “important images in their own right, meant to be seen as well as read”. The

6 Each catalogue entry of the Corpus of Anglo-Saxon Stone Sculpture volumes (Cramp, gen. ed. 1984‒) carefully records this information if available. 7 The Christian cultural context is suggested by a combination of the following features of sculpture: Christian iconography manifested in carved images, symbols, or the cruciform shape of the monument; recognisably Christian monument types; location of the monuments at monastic or ecclesiastical sites; and Christian memorial inscriptions or prayer formulas.

84

Lilla Kopár

relationship between inscriptions and other images on an object may differ greatly from object type to object type, and even within the same artifact. On the richly inscribed Ruthwell Cross, for example, the Latin captions placed above and around panels with figural scenes on the front and back of the cross serve as identifications of those specific images, while the runic poem on the sides under the cross-arms, recording the words of the Cross itself at the Crucifixion, relates to the cruciform monument as a whole in all its complexity and functions (see also Bammesberger, this volume). Consequently, inscriptions may be assigned a variety of different roles and functions not only according to monument type (e.g., grave marker, dedication stone, memorial cross) but also according to their position within the layout of the object. As visual records of utterances, inscriptions “give voice to objects” on which they appear (Karkov 2011, 135). Different inscriptions carry different voices: some recount the origins of the object itself, others name makers, patrons, or owners who are absent, again others provide a bridge between the visual language of the object and its audience (Karkov 2011, 135). It is important to consider whether an inscription was part of the original design or a later addition to the artifact, as a result of potential repurposing of the object. If the inscription was part of the design (as in the case of the Great Urswick stone, see below), we may assume an intended unity and correspondence between form, monument type, images, and text; if added later (as most likely on the Chester-le-Street stone, for example), there may be no immediate correspondence, although the original features and function of the monument were likely taken into consideration. The voices of inscriptions represent the voices of individuals behind each artifact. It is their language and literacy that is the focus of linguistic analysis, and of special interest to runologists. However, it is not easy to pinpoint exactly whose words are represented and what overall role they played in the production of the monument and in the commemoration process. The production of a carved stone monument usually involved numerous people, but at least a craftsman/carver and a patron. This begs a series of questions relevant for the analysis of runic inscriptions (regarding dialect, style, literacy, epigraphy, and more): Who carved the runes (physically)? Who decided on the text? Whose language and literacy are being documented (carver, patron, or a third individual)? While these individuals probably all belonged to the same geographical and cultural context and probably social community, further identifications are impossible to make. This may not matter much for linguistic analysis but opens up interesting questions as to the agents of the preservation of archaic linguistic forms (e.g., at Urswick) or of the introduction of innovative runic characters (e.g., at Ruthwell).

Inscriptions on Stone Monuments: Methodological Concerns

85

2 Methodological Concerns 2.1 Numbering Before we move on to methodological concerns of the study of what is conventionally referred to as → Anglo-Saxon sculpture, let me add a brief note on conventions of cataloguing and numbering. Each scholarly discipline develops its own catalogues of relevant objects with a numbering system they consider most fitting for their purposes. Inscribed stones have thus been subject to cataloguing by both runologists and epigraphers and by scholars of sculpture (primarily archaeologists and art historians by training). This has resulted in double numbering of sculptures with runic inscriptions, which can at times be confusing and frustrating. For the study of pre-Conquest sculpture, the standard systematic catalogue is the Corpus of Anglo-Saxon Stone Sculpture, a publication and research project housed at Durham University. To date, thirteen regionally focused Corpus volumes have been published.8 The principles of numbering and entry structure are explained in the General Introduction to the series by Rosemary Cramp.9 Sculptures of known provenance are listed under those place names (with church dedications if applicable), followed by catalogue numbers (in Arabic numerals) listing all pieces from the site in order of the following monument types: crosses; slabs, upright and recumbent; other types of funerary monuments such as sarcophagi or hogbacks; architectural sculpture and church furnishings; and fragments of unidentifiable monument types (Cramp 1984a, xiii). The faces of the monuments are indicated with capital letters, whereby A is the principal face (for upright monuments, the west face when in situ or one of the broad faces when not; for recumbent monuments,

8 Rosemary Cramp, senior ed. and Derek Craig, series ed. 1984‒. Volumes published to date: I. County Durham and Northumberland, by Rosemary Cramp, 1984; II. Cumberland, Westmorland and Lancashire North-of-the-Sands, by Richard N. Bailey and Rosemary Cramp, 1988; III. York and Eastern Yorkshire, by James Lang, 1991; IV. South-East England, by Dominic Tweddle, Martin Biddle, and Birthe Kjølbye-Biddle, 1995; V. Lincolnshire, by Paul Everson and David Stocker, 1999; VI. Northern Yorkshire, by James Lang, 2002; VII. South-West England, by Rosemary Cramp, 2006; VIII. Western Yorkshire, by Elizabeth Coatsworth, 2008; IX. Cheshire and Lancashire, by Richard N. Bailey, 2010; X. The Western Midlands: Gloucestershire, Herefordshire, Shropshire, Warwickshire and Worcestershire, by Richard Bryant with Michael Hare, 2012; XI. Early Cornish Sculpture, by Ann PrestonJones and Elisabeth Okasha, 2013; XII. Nottinghamshire, by Paul Everson and David Stocker, 2016; XIII. Derbyshire and Staffordshire, by Jane Hawkes and Philip C. Sidebottom, 2018. Volumes I–XII are also available online as a searchable catalogue at ‹www.ascorpus.ac.uk/catindex.php›, last accessed 11 December 2020. 9 Published separately as Rosemary Cramp, The Corpus of Anglo-Saxon Stone Sculpture: General Introduction (Oxford, 1984) and included as prefatory matter within Corpus Volumes I–II. Later reissued as Grammar of Anglo-Saxon Ornament: A General Introduction to the Corpus of Anglo-Saxon Stone Sculpture by Rosemary Cramp (Oxford, 1991, repr. 1995, 1999). Also available online at ‹www.ascorpus.ac.uk/asgrammar.php›, last accessed 11 December 2020.

86

Lilla Kopár

usually the top surface), followed by other faces (B, C, D) in anti-clockwise order. Top and bottom surfaces are labeled E and F respectively. Panel divisions on individual faces are numbered, from top to bottom, in Roman numerals (i, ii, iii, etc.). Thus, for example, the mixed runic-Roman maker inscription (M[Y]REDaH·MEH·wO) on the Alnmouth stone can be located as Alnmouth 1Cii, indicating that the text is on one of the broad faces in the second panel from the top. In runic studies, on the other hand, it is customary to refer to inscribed stones by provenance only (e.g., Bewcastle, Crowle, Bakewell); for sites with multiple inscribed stones, Roman numerals are used to distinguish separate monuments (e.g., Whithorn I and II; Lindisfarne I, II, III, etc.). Discrepancies in numbering arise at sites with multiple inscribed and non-inscribed stone monuments, so runic Lindisfarne I, for example, is known to sculpture scholars as Lindisfarne 24 (with face A inscribed); Thornhill I as Thornhill 2; or the Collingham stone as Collingham 2a–b, indicating that the cross shaft is in two broken pieces.

2.2 Dating and Preservation There are two main methodological concerns of sculptural studies that a runologist must be aware of: one is the methods (and uncertainties) of dating, the other the issues of preservation and differences between the original and the current state of the monuments. The dating of monuments is essential for interpreting their inscriptions; however, the dating of carved stones is notoriously difficult. There is no one reliable method of absolute dating, thus a combination of different (often uncertain) methods must be considered. Furthermore, sculptural production shows significant regional variation, so it is hard to suggest a general development pattern or overall chronology. Lastly, the date of inscribed stones is often established with the help of their inscriptions, thus those seeking an independent dating of an artifact in support of a proposed date for its inscription may run into a circular argument. Although an absolute chronology of sculpture is hard to establish, sculptural production in early medieval England is traditionally divided into two main periods: the pre-Conquest or Anglo-Saxon period (ca. seventh to mid-eleventh centuries), and the post-Conquest period (also referred to as Norman or Romanesque, depending on the context). Anglo-Saxon sculpture can be further divided into early, or Anglian sculpture of the pre-Viking period (up to the late ninth century or c. AD 900), and later, or → Viking Age sculpture, wherever this distinction is applicable in political and cultural terms (i.e., in northern England and parts of the Midlands, but not in the southern territories outside of Scandinavian control and settlement). The date of stone sculpture is established by a combination of dating methods. These are, in order of accuracy (according to Cramp 1984a, xlvii): inscriptions; associations with independently datable structures, sites, or archaeological layer; histor-

Inscriptions on Stone Monuments: Methodological Concerns

87

ical context provided by documented historical events or individuals; typology (monument types) and style (ornaments, decorative patterns, as well as iconography). The latter can in itself provide us only with relative chronologies, unless coupled with historical evidence, as in the case of the appearance of Viking art styles and iconography along with the documented chronology of Viking settlement of the region in question. As for inscriptions, their contents are of rather limited use for dating. With the exception of the dedication stones of St. Pauls’s Church at Jarrow (no. 17, Co. Durham; AD 685; Cramp 1984b, 113‒114), Odda’s Chapel at Deerhurst (no. 1, Gloucestershire; AD 1056; Bryant and Hare 2012, 190‒195), and St. George’s Minster at Kirkdale (no. 10, Yorkshire; 1055‒c.1065; Lang 1991, 163‒166), none of which are in Old English runes, no inscription includes dates or records names that can with certainty be identified with documented individuals. Dating of inscribed stones thus often relies on linguistic, paleographical, and epigraphic evidence; and this is where the argument easily becomes circular from the runologist’s and epigrapher’s point of view. Another aspect of pre-Conquest sculpture that is of significance for runologists and epigraphers is the notable discrepancy between the original and present state of the monuments. Like many other rune-inscribed objects, the majority of sculpture survives as fragments, often in a weathered state, thus general issues with fragmentary objects apply to sculpture as well: incomplete or/and damaged inscriptions; lack of immediate context (evidence for the position of the inscription on the object or for accompanying decorations or texts); and sometimes even lack of evidence for object type that would provide information about the content, function, authorship, and intended audience of inscriptions. Surviving evidence of pigment on sculptures suggests that several carved stones may have been originally painted. These polychrome monuments would have appeared quite different to the contemporary audience than they do today. The colors of paint could have added visual emphasis to carvings and inscriptions, suggested links between images and texts, and, most importantly, could have corrected, altered, or enhanced carved lettering, or possibly even added (non-incised) characters or words. For example, the now blank panels on face D of the Alnmouth cross-shaft (no. 1) may have displayed inscriptions similar to those on face B (Cramp 1984b, 161‒162, Plates 156‒7). Similarly, the otherwise carefully decorated Halton Cross with its famous Sigurd imagery (Halton St. Wilfrid 1, Lancashire; Bailey 2010, 177‒183, Ills. 464‒470) may have carried an inscription on face D that was lightly incised and enhanced by paint, or just painted.10 Alternatively, an inscription was intended

10 While the layout of the monument clearly calls for a “finishing touch” in the empty square which may have been an inscription, it seems unlikely that the Halton Cross would have included an inscription in Old English runes. Its tenth-century date and Anglo-Scandinavian iconographical connections (through its Sigurd iconography) put it outside of the attested time frame and cultural context for Old English runic inscriptions on stone.

88

Lilla Kopár

but never executed, and the monument is unfinished. At the same site, a further shaft fragment of earlier date (late 8th or 9th century), Halton St. Wilfrid 3bA (Bailey 2010, 185‒187, Ills. 481, 483), shows an angel of the Last Judgment holding a large rectangular object, an open book, which may have carried a painted inscription. Lastly, considering the current state of stone monuments, we must keep in mind the devastating impact of weathering on most of the surviving inscriptions, which should implore not only careful examination and documentation of surviving monuments but also their preservation. The recent discovery of the Kirby Misperton inscription (see note 2), which had been overlooked by the skilled eyes of James Lang when cataloguing the site for his Corpus of Anglo-Saxon Stone Sculpture volume of York and Eastern Yorkshire (1991), suggests that there may still be unnoticed inscriptions out there.

3 The Runic Corpus The known corpus of inscriptions containing Old English runic characters carved on stone comprises 38 inscriptions, including those on the now lost Leeds and Norham fragments, the illegible text of the Bingley font or cross-base, and the graffiti-like inscription at Kirby Misperton. Further evidence for runic graffiti on stone outside of the British Isles comes from Rome and from Monte S. Angelo in Gargano, Italy. While most of these monuments contain runic inscriptions only, a significant portion of the corpus is biscriptal, containing inscriptions in both runic and Roman lettering (e.g., Ruthwell, Lindisfarne I, II, III, V, VI, Monkwearmouth II). A few inscriptions are in mixed script, with both runic and Roman characters in the same word (e.g., Alnmouth, Chester-le-Street, Norham).11 On Hartlepool I the runic name is paired with characters intended to represent the Greek alpha and omega. The Hackness stone also contains cryptic hahal-runes, besides ordinary runes and Latin inscriptions in Roman capitals (Page 1991, 84‒86). In comparison to the corpus of Old English runic inscriptions (currently around 120 inscriptions), there are over 30 inscriptions in Norse runes from the Isle of Man alone, all in stone, another 27 from Orkney and Shetland, with an additional 30 graffiti inscriptions at Maeshowe in Orkney, 15 from Scotland, 20 or more from Ireland (mostly from Dublin), and another 16 from England (cf. Barnes/Page 2006; Holman 1996; Barnes et al. 1997; Barnes 1994). With the exception of the Dublin material, the majority of these appear on stone, although not necessarily on monumental sculpture. These Norse inscriptions not only use a different subset of runic characters, but those that appear on stone monuments are generally distinguished

11 The necessity to differentiate between biscriptal and mixed script was pointed out by Gaby Waxenberger (2013).

Inscriptions on Stone Monuments: Methodological Concerns

89

from the Old English inscriptions by a different layout, mainly in vertical, often double lines. As opposed to the Old English material, the Norse runic corpus of England has only one biscriptal example, the eleventh-century sundial of Skelton in Cleveland (Yorkshire), with an Old Norse or Norse-influenced Old English text in Roman script, accompanied by Norse runes on the border in vertical double lines (Lang 2001, 195‒197, Ill. 744). The runic text shows influence of the Roman spelling tradition. The lack of further examples suggests a difference in the status of the Old English versus Norse runes in relation to Roman writing, in particular in the context of sculpture. As noted above, in terms of geographical distribution, Old English runic stone monuments are typically confined to the north of England (with the northern Midlands, southwest Scotland, and the Isle of Man), except for the Dover and Sandwich/Richborough stones. This distribution overlaps largely with the distribution of inscriptions in Roman script that, as noted above, may appear alongside runic texts (cf. Okasha 1971; 1983; 1992; 2004a; and Parsons 1999, 111, Fig. 16 for a comparative map). The influence of Roman script practice on runic inscriptions (e.g., the Falstone stone’s use of two separate runes, o and e, for the i-mutated o, otherwise represented by the œþil-rune, cf. Parsons 1999, 80) provides further evidence for the close relationship of the two scripts in northern England. The majority of rune-inscribed stones are dated to a period from the late seventh to the late ninth centuries; thus, in sculptural-historical terms, they belong to the Anglian or pre-Viking period. In spite of the proliferation of stone sculpture in Viking Age northern England (an approximately fivefold increase in production, with regional differences; Bailey 1996, 79), there are only a few stones with Old English runes that have been dated to the tenth century or later, and dating is difficult in all cases. The Bingley font or cross-base (no. 2, West Yorkshire), with its now illegible and “doubtfully identifiable runes” (Page 1969, 34, in Coatsworth 2008, 102; Page 1999, 32) is, according to Richard Bailey (1980, 52) “our best hope of finding ornament accompanied by runic inscription which would date to the Viking period”. It is a frustrating piece with a clumsy and disproportionate decoration that is hard to parallel (and date) in art historical terms. Coatsworth (2008, 102) dates the stone to the tenth to eleventh century. Another runic monument with Viking Age associations is the cross-shaft of Crowle (no. 1) in Lincolnshire. It carries a fragmentary Old English inscription with a dialectal spelling of bæcun with ‹æ› (showing Anglian smoothing; Waxenberger forthc.). The word becun/bæcun often appears in a common memorial formula (see below) usually dated to the eighth and ninth centuries, but the layout of the inscription in a vertical, raised and curving band as well as the decoration of the stone show Scandinavian influence (Parsons in Everson/Stocker 1999, 148‒150). The inscription itself provides no clear evidence for dating, but suggests a continuity of the local vernacular epigraphic tradition. The curved rune-band may point to a date as late as c. 1000 based on Scandinavian parallels, and the Scandinavian influence on the iconography suggests a period of

90

Lilla Kopár

production after the settlement of the Vikings in Lincolnshire from the last quarter of the ninth century onwards. An early to mid-tenth century date seems to be a reasonable suggestion for an object that combines local tradition with Scandinavian-inspired innovation. We encounter a similar situation in the case of the Collingham shaft (no. 2a‒b, West Yorkshire), where there is potentially a discrepancy between the late Anglian (or early Anglo-Scandinavian) style of ornamentation and a (linguistically) earlier or archaic runic text, that may in fact have been added to the monument later (Coatsworth/Parsons in Coatsworth 2008, 120‒121). The now lost fragment of Leeds (no. 9, West Yorkshire) with its possible (though damaged) becun formula with a Scandinavian-influenced name (Onlaf) may also point to a period of Viking influence, suggesting a late ninth-century date at the earliest (cf. Parsons in Coatsworth 2008, 207), thus pushing the later end of the tradition of rune-inscribed memorial stones in northern England into the early period of Scandinavian settlement in the late ninth and tenth centuries. The case may be supported by further monuments from the northeast. The Chester-le-Street cross shaft (no. 1, Co. Durham) shows a combination of Anglian and Scandinavian influence. Its horseman, canine beast heads, and knot patterns all have parallels in Anglo-Scandinavian sculpture (Cramp 1984b, 53‒54), which advances its date to the late ninth and possibly into the early tenth century. Its inscription in mixed runic-Roman script (EADmUnD) seems to be a later addition. In its mixed script it is paralleled by the Alnmouth cross shaft (no. 1, Northumberland), a monument that has also been assigned a late ninth to early tenth century date based on the style of ornamental and figural carvings (Cramp 1984b, 161‒162). A third monument from the same area, Monkwearmouth I (no. 3, Co. Durham), displays an inscription in runes only. Its figural carving and cross shape suggest a date of the first half of the tenth century, which makes it the latest purely runic monument in the northeast (Cramp 1984a, 123). The latest rune-inscribed stone monument of late tenth to early eleventh century date comes from Whithorn in Dumfries and Galloway in Scotland, once part of the kingdom of Northumbria. The Whithorn memorial stone marks the end of the Old English runic tradition in the northwest, the region that had once produced the impressive and complex monuments of Ruthwell and Bewcastle (see below). It is also the latest known witness to Old English runic epigraphy in the British Isles at large (Findell 2014, 54–55).

4 The Inscriptions As Ray Page (1999, 155‒156) has noted, “the content of the Anglo-Saxon rune-stone inscriptions is pretty dull”. This is not entirely true, although the majority of the inscriptions reveal little information beyond names of (now unidentifiable) individuals and standardised commemorative formulas. A notable exception is the eighthcentury Ruthwell Cross with its vernacular poetic account of the Crucifixion present-

Inscriptions on Stone Monuments: Methodological Concerns

91

ed by the speaking cross itself. The rune-inscribed lines are famously paralleled by lines 39‒42, 44‒45, 48‒49, 56‒59, and 62‒64 of The Dream of the Rood poem of the late tenth-century Vercelli Book (Vercelli, Biblioteca Capitolare CXVII, fols. 104v‒ 106r; cf. Swanton 1987, 94‒97, and Bammesberger, this volume), and in an altered and much abbreviated form on the Brussels Cross, an early eleventh-century reliquary cross, accompanied by a dedication and a maker inscription written in Roman letters. The Ruthwell Cross is unique in the surviving runic corpus both in the content and layout of its runic inscription. The biscriptal and bilingual nature of the monument (with the vernacular poem in runes and most of the Latin captions in Roman script) provides evidence for the use and status of different scripts in a Christian (presumably monastic) cultural milieu in a (primarily) non-commemorative context. Runes were perceived as an archaic script that communicated and indicated a sense of shared history, identity, and cultural heritage that was embraced, transformed, and incorporated (along with other cultural influences) into the culture of Christian Northumbria. As Catherine Karkov has noted (2011, 144‒145), the use of runes for Ruthwell’s vernacular poem lends the words of the cross both a sense of age or antiquity and a sense of mystery – as if they were an old story recounted by a wise and aged man or woman. [… T]he voice of the [runic] inscriptions sets Old English in a different relationship to the monument from Latin, and embeds it anachronistically into the chronology of the past, yet here eternally present, moment of the Crucifixion […].

The Bewcastle Cross, of similar date and stature, displayed a commemorative inscription longer and more detailed than the rest of the surviving commemorative inscriptions (see below), but the text is no longer fully decipherable, as the monument has been standing in situ for over a millennium, exposed to the elements. The combination of the runic text with Christian iconography and with a classicising image of a secular patron or the commemorated conveys a similar sense of shared history and multi-faceted cultural identity incorporated into Christian salvation history. Before we turn to the classification of the rest of the surviving commemorative inscriptions on stone, let us note some methodological concerns regarding the treatment of inscriptions. Because of the formulaic nature of many of the surviving inscriptions (memorial, maker, and owner formulas), it is tempting to divide longer inscriptions on the same object into recognisable constituent parts for purposes of classification. The combination of different textual units, or formulas, on the same object should, however, be treated as a coherent statement, the message of which, in the context of the given object, including its form and imagery, may be more than the sum of its components. Further, formulaic inscriptions often display varying degrees of explicitness in the wording and tend to operate as codified signifiers of a content, e.g., a memorial message, especially in the hands of carvers who sometimes have limited understanding of the script or of the archaic text. Despite abbreviations and linguistic and orthographic mistakes, the message of a formulaic inscription remains intact because the text becomes more than a linguistic utterance. Its (verbal) explicitness is governed by the context of the monument (monument

92

Lilla Kopár

type, iconography, place of monument, audience) as much as by the available surface to carve. Here is where considering the full monument can provide important clues to interpretation. In the context of name stones, for example, the object type itself conveys the memorial function and signals a request for prayer (cf. Okasha 2004b, 97; Maddern 2013, 92), thus it is only the name of the deceased (and in the case of stones with two names, likely the name of the patron) that needs to be indicated. The commemorated and the commemorator are distinguished by the layout of the text on the stones, that is, by a visual clue, not by linguistic means. Crosses and cross slabs can be functionally more ambiguous; therefore, inscriptions need to reveal more of the patrons’ intentions and the function of the monument. As noted above, any inscription is an image, a visual manifestation of a verbal utterance. Therefore, it has to be regarded as visual information as much as a linguistic statement, especially if integrated into an artifact. As an image, an inscription behaves the same way images do: it is influenced by the availability of space, it can be stylised, adapted, simplified, misinterpreted, and miscopied. These changes provide extra-linguistic information about makers, commissioners, and users as well as the function of the text beyond its linguistic statement. The line between text and image is a very fine one in the case of many runic inscriptions (e.g., the Spong Hill alu-inscription, or the fuþorc of the Thames scramasax), and the blurring of the linguistic/symbolic and textual/visual/decorative functions of runic texts provides important information about the use of the script in general.

4.1 Inscription Types The majority of the surviving runic texts on stone monuments are considered to be commemorative inscriptions. They are manifested at varying degrees of complexity and detail, from single personal names to full poetic lines with standard memorial and prayer formulas and maker signatures. These inscribed monuments bear witness to a widespread commemorative tradition in northern England with up to 18 vernacular memorial inscriptions in stone, although not all in runes (Parsons 2008, 79; Kopár 2015). R. I. Page (1959, 75; 1999, 138‒156) and later Gaby Waxenberger (2013) have differentiated between four major types of inscriptions by identifying recurring elements or formulaic units that appear alone or in combination with others. Waxenberger’s main types (based largely on Page) are as follows: Type I inscriptions, or single-name memorial formulas, record only one name in the nominative, assumed to be that of the deceased, e.g., at Chester-le-Street, Dover, Hartlepool I and II, Lindisfarne I, Monkwearmouth I. Type II inscriptions record two names, often in two different scripts, runic and Roman (Lindisfarne I, II, III, V, VI, probably also IV; Monkwearmouth II). On Lindisfarne I and V, the same name appears in two different scripts. The two names are

Inscriptions on Stone Monuments: Methodological Concerns

93

generally assumed to be those of the deceased and the patron (who, if related, may have had the same name). If the two-name formula is seen as an abbreviation of a full memorial formula (“X set up this monument in memory of Y”; see below), the first name would indicate the patron and the second name the deceased. The two names may also refer to two different deceased individuals commemorated or even buried together (Page 1959, 76; Maddern 2013, 10), or reference the same individual with his/her name before and after conversion or joining a religious order (Karkov 2011, 50, 93; Maddern 2013, 10). The biscriptal nature of the inscription would then signal the cultural shift associated with the change in religious status. The tradition of two-name memorial stones seems to be limited to Lindisfarne and Monkwearmouth, although runic name-stones are also known from Hartlepool. Waxenberger’s Type III formula is a simple request for prayer: “Pray for the soul/ for X”. All but (possibly) one of the monuments with a Type III (prayer) formula also display other commemorative inscriptions as well (Great Urswick, Thornhill III, Falstone, and Overchurch, but maybe not Lancaster), which suggests that this formula was a dependent phrase of a memorial formula, not necessarily a separate inscription category. It is important to note that the prayer the inscription calls for is as much for the deceased as for the patron, whose act of commemoration is also self-referential: a public monument commemorates both the living and the dead. The next category, Type IV or maker formula (“X made me/this”), makes a similar point: it emphasises the role of the maker in the act of commemoration. It appears on the Alnmouth, Kirkheaton and Urswick stones and possibly at St Ninian’s Cave (Page 1959, 77). These basic types of formulaic elements are combined to form complex, poetic commemorative inscriptions in alliterating half-lines on Great Urswick, Thornhill III (no. 2), and Falstone: (Great) Urswick (no. 1, Lancashire; Bailey/Cramp 1988, 148‒149): +Tunwini setæ æfter Torohtredæ bekun æfter his bæurnæ. Gebidæs þer saulæ. Lyl þis w[.]. ‘+Tunwini set up (this) monument after Torhtred his warrior/hero/child; pray for the soul. Lyl (made) this.’

Thornhill III (no. 2, West Yorkshire; Coatsworth 2008, 258): +Gilsuiþ arærde æft[.] Berhtsuiþe bekun on bergi gebiddaþ þær saule. ‘+Gilswith raised (this) memorial on a mound/hill for Berhtswith; pray for the soul.’

Falstone (no. 2, Northumberland; Cramp 1984, 172‒173; two fragmentary inscriptions, one in runes, one in Roman letters):

94

Lilla Kopár

+[.] æftær roe[…]tæ […]cun æftær e[…] geb[…]æd þe[.] saule +EO[.]TA[.] AEFTAER HROETHBERHTÆ BECUN AEFTAER EOMAE GEBIDAED DER SAULE ‘+Eo[.] set up (this) monument after Hroethberht his uncle. Pray for the soul.’

David N. Parsons (2008) approached the question of memorial inscriptions the other way round, taking the complex memorial formula, surviving at its fullest at Great Urswick, Falstone, and Thornhill, as his starting point. This “full memorial text” had three standard elements: “the sponsor formula (‘X set up in memory of Y’), the monument formula (‘a beacon in memory of someone, or standing somewhere’), and the invitation to prayer” (Parsons 2008, 80). While this complete formula (with some variation in wording, word order, or even formulaic components) may have been present in full on some of the now fragmentary monuments, it was represented in abbreviated forms on others.12 In its most condensed form, represented by a single name or two names, the poetry of the full memorial message may have been easily invoked by the name of the commemorated and that of the sponsor at rituals involving the monument. The nature of abbreviation was determined by the availability of space, the location of the inscription on the monument and its relation to images and decorations, and by the monument type, which in itself could convey part of the commemorative message (e.g., name stones).

5 Case Study: The Great Urswick Stone How inscriptions are influenced by availability of surface space and the overall design of a monument, how they interact with images, and how they contribute to our understanding of the production, dating, and overall interpretation of a piece of sculpture, can be illustrated by a brief case study of the Great Urswick Stone. The Great Urswick Stone (Urswick 1, Lancashire; Bailey/Cramp 1988, 148‒151, Ills. 564‒566, 568‒569) is a 117-cm-tall fragmentary cross shaft in red sandstone. It was discovered in 1911, built into the fabric of the church (Bailey/Cramp 1988, 148). Now it stands on a windowsill at St. Michael’s Church in Urswick, Lancashire (see Fig. 1). Its damaged, curved shape indicates its post-Conquest reuse as a lintel above a window. While its main inscription panel on face A remained intact, part of a figural scene on the same face that included further runic lettering was chiseled away, together with the upper and lower halves of the cross shaft.

12 Out of the 18 northern vernacular memorial inscriptions, some in runes, some in Roman characters, Parsons (2008, 82) identifies three as examples of the full alliterative text (see above), together with another four now fragmentary inscriptions that may have included the full text, and the remaining eleven as its variants, rearrangements or abbreviations. For further discussion of vernacular memorial inscription, see Kopár (2015).

Inscriptions on Stone Monuments: Methodological Concerns

95

Fig. 1: (Great) Urswick 1A; © Wardens of Urswick Church; photo: L. Kopár.

The shaft was once decorated with carvings on all four sides. Side A, the principal broad face, now displays in the top panel an interlace pattern with flat strands and malformed loops (see Fig. 1 above). Below it is a framed panel with the runic inscription in five lines. Due to the rune-carver’s miscalculation of space, the inscription spills out of the frame and runs into the figural panel below. This (now damaged) figural scene shows two frontal men with their faces turned to each other. There is a tall and slender relief cross between them. Both figures are wearing secular clothing with a high neck, and have a characteristic hairstyle with long hair. The one on the left extends his right hand across the cross shaft to touch or gesture to the figure on the right. Their facial features are very similar, with lightly incised eyes, mouth, and ears. The left-hand figure has an incised scroll under his left shoulder that might be part of his clothing (or a staff?). The scene of two human figures flanking an empty cross is in itself not an unusual one (cf. Kirkby Wharfe, Lancaster St. Mary, Burton in Kendal, Halton, Lindisfarne, Hope) but we would normally expect to find St. Mary and St. John, both saintly figures with haloes. Here, however, we have two male figures in secular clothing. A late ninth-century date may explain the use of secular clothing for a biblical scene (Bailey/Cramp 1988, 150; Cramp 1982, 17‒18). Alternatively, the carving may represent Christ welcoming the deceased in afterlife; a conversion scene (of the deceased?); or possibly a scene from an (unidentified) saint’s life (Bailey/Cramp 1988, 150). Although the commemorative inscription offers no direct explanation for the figural scene, it is quite likely that text and image

96

Lilla Kopár

Fig. 2: (Great) Urswick 1A, detail; © Wardens of Urswick Church; photo: L. Kopár.

were intended to convey a similar message: the hope of salvation through faith, commemoration, and prayer (see also face C below in Fig. 3). The inscription (see above) is one of the three surviving examples of the vernacular poetic commemorative text type: +Tunwini set up [this] monument after Torhtred, his warrior/hero/child; pray for the soul. It consists of a memorial and a prayer formula, accompanied by a maker signature (Lyl [made] this). It reveals the name of the patron, the commemorated, and the maker of the monument, all of whom were likely English and Christians, judging by their names, by the language and script of the inscription, and by the typology and iconography of the monument. While the commemorative text (with the exception of the added maker signature) is in alliterative half-lines, the layout does not reflect this (as we are indeed often used to from manuscripts of vernacular poetry). The text is laid out in continuous lines with no respect for word division or morphology (except maybe in the first line; see Fig. 2). The memorial formula fills four and a half lines of the designated frame. The runes decrease in size with each line, most drastically in line five. The carver mismanaged the space available for the text, so despite smaller and more tightly spaced characters, he could not accommodate the full text in the text box. The prayer formula, which completes the second long-line and alliterates through the sound b, and thus forms an integral part of the poetic text, starts in the middle of the fifth line within the frame but invades the figural image below. By accident or by design, the word saule ‘soul’, together with the last character of þer, is inscribed in three of the four quadrants of the cross between the two figures, and thus physically embraces the very instrument of salvation evoked by the monument itself (transliteration by Page in Bailey/Cramp 1988, 148, modified by L. Kopár): +tunwinisetæ æftertoroh tredæbeku næfterhisb æurnægebidæsþe rs / au / læ

Inscriptions on Stone Monuments: Methodological Concerns

97

The memorial inscription could have ended here, but the carver felt it necessary to add his own formulaic maker’s signature, carved across the upper bodies of the two human figures: lylþi / sw[. The last missing word can be completed with certainty to worhtæ or worhte, meaning ‘Lyl made this’ (R. I. Page in Bailey/Cramp 1988, 149). The layout of the inscription reveals valuable information about the production process of the monument: it suggests that the figural scene, the interlace pattern, and the text box (an empty panel) were executed first, and the inscription was added later. Although an inscription was certainly part of the original design of the monument, the actual text itself may not have been firmly set at the time of the layout design. If the carver of the runes differed from the stone carver responsible for the figural and decorative carvings, he may have had little to no say in the overall design of the monument. It is interesting to note that despite the apparent lack of space, he felt it necessary to add a maker’s signature, emphasising the role of the creator of the monument in the commemoration process, and adding his name to the list of those to be remembered and prayed for. But who was this Lyl? If he was the stone carver, did he also carve the runes? Is it by accident or by design that his name is written across one of the two figures? And most importantly from the perspective of the runologist and linguist, what influence did he have on the text recorded? As an artist-craftsman, he was not the most highly skilled one, but he introduced a number of unusual features, possibly under the influence of metalwork and manuscripts (cf. Bailey/ Cramp 1988, 149‒150). Was he similarly innovative (yet not particularly skilled and careful) when it came to adding the runic text as the last step in the production process? This is certainly a possibility. While these questions cannot be answered, it is worth considering the options and being aware of the uncertainties when discussing textual production and transmission, and linguistic accuracy. Just like the carvings of the monument, the inscription is unusual and ‘unpolished’, and its linguistic dating is inconsistent. According to Ray Page (in Bailey/ Cramp 1988, 149), the unstressed final vowels in Tunwini, setæ, Torohtredæ, bæurnæ, and saulæ, together with the “unusual fracture diphthong” in bæurnæ all point to an early date. Page also suggests (ibid.) that the -er of æfter (twice) is inconsistent with the earlier patterns of unstressed end vowels noted above and would thus suggest a later date. This is, however, not necessarily the case, since IE *e remained in Germanic and in Old English before r in unaccented syllables, therefore it is no reliable evidence for dating (cf. Campbell 1959, § 331.2).13 Page assigns a broad eighth or ninth century date to the text, with the caveat that our knowledge

13 I owe special thanks to Gaby Waxenberger for clarifying this matter.

98

Lilla Kopár

Fig. 3: (Great) Urswick 1C; © Wardens of Urswick Church; photo: L. Kopár.

of the local northwestern dialect is limited. The linguistic features noted above may thus be dialectal rather than strictly chronological markers. How does this suggested linguistic dating compare to the art historical dating of the cross shaft and its decoration?14 In art historical terms, the dating of the monument is aided by its decorations that appear on all faces. Face A offers no conclusive evidence, but the figural carving fits most comfortably with a ninth-century date. Face B contains the remains of an unusual double twist pattern, otherwise only paralleled in Leeds and in the metalwork binding of the Brunswick (Gandersheim) Casket (Bailey/Cramp 1988, 149). This pattern, together with the interlace of face A, suggests late Anglian influence and a late pre-Viking period date, i.e., ninth century. Face D has been chipped away. Face C shows an inhabited vine scroll with two birds on the top, a profile male, with hair similar to those on face A, and a frontal female figure, both dressed (see Fig. 3). Below them are two curved reptilian-looking quadrupeds that vaguely resemble the earlier ‘Northumbrian beast’ type. The scroll seems to be sprouting from an urn or chalice. The clumsily designed and executed scroll conforms to no standard type. The iconography of the vine scroll references the cosmic world tree

14 There is no archaeological or documentary evidence available in this case to help with dating.

Inscriptions on Stone Monuments: Methodological Concerns

99

(with a possible allusion to the Eucharist through the chalice and the redemptive wine), with Adam and Eve as representatives of the human race (Bailey/Cramp 1988, 149). It offers no conclusive evidence for dating, but the figural style is consistent with that of face A, thus a slightly later date is preferred. Adding up the pieces of evidence, we notice a bit of an inconsistency in the linguistic and sculptural dating. In summary, it seems that we are dealing with an innovative, though not particularly skilled carver of the ninth century who adopted and adapted known patterns of interlace and figural iconography and was possibly influenced by metalwork. He and his patron had an appreciation for vernacular commemorative texts and may have retained archaic linguistic features of commemorative poetic lines known to them while adapting the text to their particular case of commemoration. They participated in a tradition of commemoration through text and image combined in a public monument (as opposed to earlier name stones) that most likely served as a focal point of local commemorative rituals and prayers. Both the patron and the carver felt it necessary to leave a visible trace of their roles in this commemoration process, emphasising the significance of commemoration as a spiritual act as much as a public statement of status, power, and legitimacy. Their message was manifested in a complex artifact with a voice, the inscribed text of which cannot be separated from its images and the monument as a whole.

References Bailey, Richard N. 1980. Viking Age Sculpture in Northern England. London: Collins. Bailey, Richard N. 1996. England’s Earliest Sculptors. Toronto: Pontifical Institute of Mediaeval Studies. Bailey, Richard N. and Rosemary J. Cramp (eds.). 1988. The Corpus of Anglo-Saxon Stone Sculpture. Vol. II: Cumberland, Westmorland, and Lancashire North-of-the-Sands. Oxford: Oxford University Press for the British Academy. Barnes, Michael P. and Raymond Ian Page. 2006. The Scandinavian Runic Inscriptions of Britain. Runrön 19. Uppsala: Uppsala University. Barnes, Michael P. 1994. The Runic Inscriptions of Maeshowe, Orkney. Runrön 8. Uppsala: Uppsala University. Barnes, Michael P., Jan Ragnar Hagland, and Raymond Ian Page. 1997. The Runic Inscriptions of Viking Age Dublin. Medieval Dublin Excavations 1962‒1981, ser. B, vol. 5. Dublin: Royal Irish Academy. Bryant, Richard and Hare Michael (eds.). 2012. The Corpus of Anglo-Saxon Stone Sculpture. Vol. X: The Western Midlands. Oxford: Oxford University Press for the British Academy. Campbell, Alastair. 1959. Old English Grammar. Oxford: Oxford University Press. Coatsworth, Elizabeth (ed.). 2008. The Corpus of Anglo-Saxon Stone Sculpture. Vol. VIII: Western Yorkshire. Oxford: Oxford University Press for the British Academy. Cramp, Rosemary J. 1982. “The Viking Image”. The Vikings. Ed. Robert T. Farrell. London and Chichester: Phillimore. 8‒19. Cramp, Rosemary J., senior ed. and Derek Craig, series ed. 1984‒. The Corpus of Anglo-Saxon Stone Sculpture. 13 vols. to date. Oxford: Oxford University Press for the British Academy. [Also available online at ‹www.ascorpus.ac.uk/catindex.php›, last accessed 15 April 2021].

100

Lilla Kopár

Cramp, Rosemary J. 1984a. The Corpus of Anglo-Saxon Stone Sculpture: General Introduction. Oxford: Oxford University Press. [Also included as prefatory matter in Cramp 1984b and Bailey and Cramp 1988; later reissued as Grammar of Anglo-Saxon Ornament: A General Introduction to the Corpus of Anglo-Saxon Stone Sculpture. Oxford: Oxford University Press, 1991, repr. 1995, 1999]. Cramp, Rosemary J. (ed.). 1984b. The Corpus of Anglo-Saxon Stone Sculpture. Vol. I: County Durham and Northumberland. 2 vols. Oxford: Oxford University Press for the British Academy. Elliott, Ralph W. V. 1989. Runes: An Introduction. 2nd ed. Manchester: Manchester University Press. Everson, Paul and David Stocker (eds.). 1999. The Corpus of Anglo-Saxon Stone Sculpture. Vol. V: Lincolnshire. Oxford: Oxford University Press for the British Academy. Findell, Martin. 2014. Runes. London: The British Museum Press. Holman, K. 1996. Scandinavian Runic Inscriptions in the British Isles: Their Historical Context. Senter for middelalderstudier, Skrifter 4. Trondheim: Akademika Publishing. Karkov, Catherine E. 2011. The Art of Anglo-Saxon England. Woodbridge: The Boydell Press. Kopár, Lilla. 2015. “Heroes on the Fringes of the Anglo-Saxon Poetic Corpus: Vernacular Memorial Inscriptions on Stone Sculpture”. Heroes and Saints: Studies in Honour of Katalin Halácsy. Eds. Zsuzsanna Simonkay and Andrea Nagy. Budapest: MondAt. 85‒120. Lang, James (ed.). 1991. The Corpus of Anglo-Saxon Stone Sculpture. Vol. III: York and Eastern Yorkshire. Oxford: Oxford University Press for the British Academy. Lang, James (ed.). 2001. The Corpus of Anglo-Saxon Stone Sculpture. Vol. VI: Northern Yorkshire. Oxford: Oxford University Press for the British Academy. Maddern, Christine. 2013. Raising the Dead: Early Medieval Name Stones in Northumbria. Turnhout: Brepols. Okasha, Elisabeth. 1971. Hand-List of Anglo-Saxon Non-Runic Inscriptions. Cambridge: Cambridge University Press. Okasha, Elisabeth. 1983. “A Supplement to Hand-List of Anglo-Saxon Non-Runic Inscriptions”. Anglo-Saxon England 11: 83‒118. Okasha, Elisabeth. 1992. “A Second Supplement to Hand-List of Anglo-Saxon Non-Runic Inscriptions”. Anglo-Saxon England 21: 37‒85. Okasha, Elisabeth. 2004a. “A Third Supplement to Hand-List of Anglo-Saxon Non-Runic Inscriptions”. Anglo-Saxon England 33: 225–281. Okasha, Elisabeth. 2004b. “Memorial Stones or Grave-Stones?” The Christian Tradition in AngloSaxon England: Approaches to Current Scholarship and Teaching. Ed. Paul Cavill. Cambridge: Brewer. 91‒101. Page, Raymond Ian. 1959. “The Inscriptions of the Anglo-Saxon Rune-Stones”. Unpublished doctoral thesis, University of Nottingham. Page, Raymond Ian. 1969. “Runes and Non-Runes”. Medieval Literature and Civilization: Studies in Memory of G. N. Garmonsway. Eds. D. A. Pearson and R. A. Waldron. London: The Athlone Press. 28‒54. [Reprinted in: Page 1995, 161‒179]. Page, Raymond Ian. 1995. Runes and Runic Inscriptions: Collected Essays on Anglo-Saxon and Viking Runes. Ed. David N. Parsons. Woodbridge: The Boydell Press. Page, Raymond Ian. 1999. An Introduction to English Runes. 2nd ed. Woodbridge: The Boydell Press. Parsons, David N. 1994. “Sandwich: The Oldest Scandinavian Rune-Stone in England?” Developments Around the Baltic Sea in the Viking Age. Eds. Björn Ambrosiani and Helen Clarke. Stockholm: Birka Project for Riksantikvarieämbetet and Statens Historiska Museer. Birka Studies 3. 310‒320. Parsons, David N. 1999. Recasting the Runes: The Reform of the Anglo-Saxon Futhorc. Runrön 14. Uppsala: Uppsala University.

Inscriptions on Stone Monuments: Methodological Concerns

101

Parsons, David N. 2008. “The Inscriptions”. The Corpus of Anglo-Saxon Stone Sculpture. Vol. VIII: Western Yorkshire. Ed. Elizabeth Coatsworth. Oxford: Oxford University Press for the British Academy. 79‒84. Swanton, Michael (ed.). 1987. The Dream of the Rood. Exeter: University of Exeter Press. Tweddle, Dominic, Martin Biddle, and Birthe Kjølbye-Biddle (eds.). 1995. The Corpus of AngloSaxon Stone Sculpture. Vol. IV: South-East England. Oxford: Oxford University Press for the British Academy. Waxenberger, Gaby. 2013. “Text Types and Formulas on Display: The Old English Rune Stone Monument in England”. vindærinne wunderbærer mære. Gedenkschrift Ute Schwab. Studia Medievalia Septentrionalia 24. Ed. Monika Schulz. Wien: Fassbaender. 495‒518. Waxenberger, Gaby. Forthc. A Phonology of Old English Runic Inscriptions with a Concise Edition and Analysis of the Graphemes. RGA-E. Berlin/Boston, MA: De Gruyter. Webster, Leslie. 2012. The Franks Casket. London: The British Museum Press.

Tineke Looijenga

Frisian Runes Revisited and an Update on the Bergakker Runic Item Abstract: In this paper it is argued that an independent or separate “Frisian” runic tradition did not exist; “Frisian” runes are part of the Anglo-Saxon corpus. This results from the observation that the Frisian terp-area was abandoned for over a century from the end of the third through the fourth century AD. The old Frisian population left, and a new population of immigrants, mainly from along the southeastern North Sea coast (the same area the Angles, Saxons and Jutes come from), settled on the old homesteads, called terpen. For reasons unknown, they adopted or received the name Frisians, probably because the name of the land, Frisia, had remained, kept alive by Frankish and Merovingian writers. Another part of this paper is dedicated to the further analysis of the fifth century runic find from Bergakker in the Betuwe. It appears that the object may have been made in a Gallo-Roman workshop in Northern Gaul. Together with the fifth century runic object from a veteran’s grave in Fallward, Landkreis Cuxhaven, both objects may be witnesses of the merger of Germanic and Roman culture and the integration into Roman society. The interesting question is how runic literacy is to be understood in such a context. A comparison between the Anglo-Saxon runic corpus and the Old Frisian one is made, with the conclusion that both corpora should be merged, since they form a common tradition.1

1 The Archaeological Background Archaeological research (1995–2013) has led to new views on the settlement history of Frisia2 in the Roman imperial period and the Migration Period. This research revealed a sharp decline in habitation followed by a habitation hiatus in the coastal area of the Low Countries, mainly during the fourth century AD. This development started in the third century. For some reasons (demographic, political;3 and drain1 I owe a great deal to the discussions with John Hines, Gaby Waxenberger, Kerstin Kazzazi and Hans Frede Nielsen, and their friendly and scholarly comments. All conclusions in this paper, however, and possible wrong assumptions are mine. 2 Frisia is the general term indicating the area north of the Rhine estuary, which included presentday North Holland, northern Utrecht, Flevoland/IJsselmeer, the island of Texel, and the provinces of Groningen and Friesland. 3 In the middle of the third century the northern border of the Roman Empire became unstable. Also, Roman troops had to be moved from the Rhine border to other regions to act against Barbarians threatening the eastern parts of the Empire. Moreover, Germanic tribes crossed the Rhine and harassed the Southern Netherlands and Belgium (cf. Taayke 2013). https://doi.org/10.1515/9783110796834-006

104

Tineke Looijenga

age problems rather than increased flooding), the population of Frisia left their homesteads on the platforms (terpen) – until large parts of the area were abandoned and remained so for about a century.4 This certainly concerned the present province of North Holland and Central Frisian Westergo, less so Oostergo and the present province of Groningen (see Map 1). Those regions were less isolated from the Pleistocene hinterland; Ezinge, for example, remained inhabited, perhaps due to continuous links with Northern Drenthe. The situation lasted until, in the fifth century, new inhabitants began to occupy the area; people coming along the coast from easterly regions: Denmark, Schleswig Holstein, and Northern Germany: the area between the estuaries of Elbe–Weser–Ems. They settled on the abandoned terpen – since these clearly were suited for habitation, better than the land and marshes in between. At the time Frisia was repopulated, groups from the same easterly regions invaded England in a series of movements that is termed adventus Saxonum. Both migrations belong to that same adventus in the sense that part of the people migrating westwards stayed behind in Frisia, while another part went on to Britain. It appears that there was a sort of reverse or secondary migration as well by Anglo-Saxons from England back to the Continent, in casu Frisia.5 Frisia was still known as ‘Frisia’ – the name was kept alive in classical sources known and studied by Merovingian and Frankish scholars. It was probably they who called the new inhabitants ‘Frisians’, after the land the newcomers occupied, but who initially were part of larger groups that went under the name of Saxones, indicating a variety of tribes and people in changing compositions. At the end of the third century, a string of fortifications was built by Emperor Constantius Chlorus: the famous Litus Saxonicum. From then onwards, these Saxones are mentioned in classical sources as living near the Franci and Chauci, next to the sea and acting as pirates. According to De Boone (1954, 16), it would seem highly improbable that the Frisians would have been able to retain an independent position between Saxones and Franci; their name is not mentioned anymore in the fourth and fifth centuries. Archaeologically, the → material culture in Frisia is deeply influenced by a ‘Saxon’ style in the fifth century, whereas the ‘Frisian’ pottery style already disappeared in the third century (Taayke et al. 2013, 163 f.). Consequently, the new Frisians who resettled the coastal area in the fifth century were not related to the people who left during the third and fourth centuries. The migrating people along the North Sea shores to Frisia and Britain became known in history as Anglo-Saxons. This would mean that those who were called Frisians after settling in Frisia were, in fact, Anglo-Saxons. Depopulation of the source regions on the south-eastern part of the North Sea coast (the German Bight) has hitherto only been demonstrated archaeologically. Any

4 This has been discussed in various publications, for instance Taayke (2003); Nicolay (2005); Gerrets (2010); Lanting/Van der Plicht (2009/10); Nieuwhof (2011). 5 I owe this information to personal communication by John Hines in an email of 12/10/2012.

Frisian Runes Revisited and an Update on the Bergakker Runic Item

Meaning of the numbers 1–3 in the boxes below: 1. Frisian core region in the seventh/eighth cent. 2. Pleistocene Hinterland 3. Coastal Holocene plain

Map 1: Frisia in the seventh/eighth century, from Gerrets (2010, 3); © Daniël A. Gerrets.

105

106

Tineke Looijenga

historical documents about these population movements are lacking. We only have Bede (Historia Ecclesiastica) and Gildas (De Excidio Britonum), who wrote their reports later than the events in question, i.e., three centuries (Bede) and one century (Gildas). It was probably not a mass emigration, but even with small amounts of invading people, important cultural changes can occur. The adventus Saxonum is part of the Migration Period, a complex affair in which many and various societal shifts took place, not only around the North Sea, but in the whole of Europe in the wake of the decline of the Roman Empire.

2 Spread of Runic Knowledge One of the typical features these Anglo-Saxons brought with them was the knowledge of runes. Even if only very few attestations have come down to us, the evidence is clear. Runic objects from the fifth century onwards are found in Frisia and the eastern and southern parts of England. As regards Frisia, all finds are without a context – the tiny objects are stray finds from the terpen, occasionally found while digging the mounds for fertile soil in the three decades before and after 1900. Some objects did not emerge from Frisian soil but are recorded from elsewhere: from England and what is now Ostfriesland in Germany. These objects are considered Frisian because of some characteristic traits, which we will discuss below. As is well-known, the Anglo-Saxon-Frisian (ASF) inscriptions are initially characterised by two graphic innovations: new runes for the sounds /o/ and /a/. The question of why this was necessary has recently been discussed by Gaby Waxenberger (2017), who assumes that several changes in the spoken language required new, or rather, adjusted, graphs. The changes in language and subsequently the runic graphemes took some time, because, as she supposes, these changes went through allophonic phases first. During these phases, it was not necessary to introduce a new graph. The change in pronunciation affected the way the language was written down. The act of designing ‘new’ runes (in fact, adjusting old runes with one or two extra strokes) must have taken place in a certain location by people who were aware of the need of a new grapheme for a new phoneme. This presupposes rather literate persons. One might wonder if these people had a connection with the Roman world, such as the important man who was buried with his Roman paraphernalia in a 5thcentury boat-grave in Fallward, on the east coast of the Weser mouth. I return to this question below. All in all, some twenty inscriptions are counted as ‘Frisian’. I do not include the bracteate from Hitsum. Five possibly ‘Frisian’ pieces are found outside Frisia: the Schweindorf solidus in Ostfriesland, the skanomodu solidus and the æniwulufu thrymsa in the UK, the Amay Comb in Belgium, and the Caistor-by-Norwich astragalus in England (cf. Waxenberger 2017).

Frisian Runes Revisited and an Update on the Bergakker Runic Item

107

Two items found elsewhere in the Netherlands may be considered Merovingian or Frankish: the Borgharen Belt Buckle bobo found near Maastricht, and the Bergakker Scabbard Mount from Bergakker near Tiel in the Rhine and Meuse estuary. The Bergakker find will be discussed in the following.

3 The Bergakker Scabbard Mount This early fifth century item was found in former Batavian territory (on the Limes) near the city of Tiel. It is neither ‘Frisian’ nor ‘Anglo-Saxon’, but nevertheless deserves our greatest attention: it is a silver-gilt scabbard mount with a late Roman ornamental pattern: half-circles and points, ridges and grooves. It was found in 1996, published by the present author and the archeologist Arjen Bosman in the same year (cf. Bosman/Looijenga 1996), and discussed more broadly in Bammesberger/Waxenberger (1999: Bammesberger, 180‒185; Looijenga 141‒151; Odenstedt 163‒173; Quak 174‒179; Seebold 157‒162; Vennemann 152‒156) and Looijenga (2003, 317‒322). In the context of this paper, it is important to look at the Bergakker Mount more closely again. It might contribute to our understanding of the spread of runic knowledge in the crucial fourth and fifth centuries: the Migration Period and the

Fig. 1: The Bergakker Scabbard Mount, front; © Collection Valkhof Museum. Photograph by Thijn van de Ven.

Fig. 2: Inscription on the back of the Bergakker Scabbard Mount. The Bergakker Scabbard Mount is part of the Collection Valkhof Museum, Nijmegen; © Collection Valkhof Museum. Photograph by Stephan Weiss-König.

108

Tineke Looijenga

Frankish expansion. It may be useful to strongly emphasise that the Bergakker item has nothing to do with any old or new Frisians – the former having disappeared from history and the latter not yet having arrived on stage. In this context, one other inscription must be discussed; it is from the same period: the Fallward (Wremen) footstool (near Feddersen Wierde on the right bank of the estuary of the river Weser; see Map 1) with runes, early fifth century, found in the grave of a Germanic-Roman veteran. The reason for treating these two objects together is because they both have a Roman military connection and they are dated to the first half of the fifth century, a period from which few runic objects outside Scandinavia are known. The Bergakker Mount legend may read: haleþewas: ann: kesjam: logens It displays one Germanic name, one Germanic verb form and two possibly Latin words (for different readings and interpretations, see Bammesberger/Waxenberger 1999). The Fallward inscription reads: ksamella lguskaþi This has one Germanic name and one Latin loanword. Neither Bergakker nor Fallward display ASF runic innovations. What is interesting, though, is the occurrence of double runes in both inscriptions: ann and ksamella (double spelling is unusual in runic writing). Both names, haleþewas and [a]lguskaþi, may point to North Germanic origin. (The reading haleþewas should be preferred despite the unknown graph in the inscription; to see that as representing a u-rune does not make sense, haluþuwas is “sprachlich nicht zufriedenstellend zu deuten” (Nedoma 2010, 33). Besides, a u-rune in this shape is unknown, whereas the rune representing an e-rune in the inscription of Engers, reading leub, (a bow-fibula, dated sixth century, but now lost, see Henning 1889, Fig. 19) resembles the double V-shaped rune in Bergakker pretty clearly). The area around Bergakker near Tiel on the ‘Batavian island’ (see Map 2) was still Romanised in the early fifth century; people were not only speaking but also writing in Latin, attested by the finds of hundreds of inscribed wooden tablets and sealboxes in the Rhine delta. These point to correspondences between soldiers and possibly their families (Derks/Roymans 2007, 131‒136). Roman and Germanic cultures met and merged in the Lower Rhine area along the Limes, at that time still the frontier area along the Rhine. I imagine that the correspondence between soldiers may have enhanced their interest in writing, perhaps also in runes, although the Roman culture was the prestigious and leading one. Writing in Latin was the fashion, runic writing belonged to quite another domain. Integration of ‘Barbarians’ into Roman society went smoothly. For instance, Frankish kings such as Mallobaudes, Fraomar, Merobaudes, Bauto, Arbogast and other leaders took a high-ranking position in the Roman army and their troops went

Frisian Runes Revisited and an Update on the Bergakker Runic Item

Map 2: The ‘Frisian’ and ‘Frankish’ runic items, part of a Continental North Sea group (taken from Looijenga 2021, Fig. 13.1, p. 377).

109

110

Tineke Looijenga

with them – as voluntarii (on the expansion of the Franks, see Dierkens/Périn 2003, 165‒193). They became official Roman soldiers of Germanic descent, auxilia, according to Böhme (1996, 101), not independent and freely operating warriors. In that position, these Franks settled with their families especially in Germania II and Belgica II, and their acculturation went well, although some of the women kept to their typical Tracht with fibulae. On the other hand, this settling of several Frankish tribes in Northern Gaul led to a “Barbarisierung” (Böhme 1996, 101), which changed ethnic and social structures – culminating in the establishment of a Frankish kingdom at the end of the fifth century. This Frankish self-image did not lead to a Germanic society including writing in runes – on the contrary, the Franks became Romanised. There are only a few ‘Frankish’ runic objects.6 The Franks knew runes, we can be certain of that (think of the famous phrase by the sixth-century bishop Venantius Fortunatus, mentioning the barbarous rune painted on a piece of beech wood). Bergakker contains an unknown graph which must reflect a vowel. Unfortunately, there is no common agreement on the transliteration, which is due to the fact that this graph is read as either e or u. However, one word in the inscription is agreed upon by everybody: ann, first or third pers. sing. pres. of the verb unnan ‘to grant, to give’ or also ‘to like, to desire’. The object has parallels in Roman-type girdle mounts and buckles (cf. Böhme 1996, 100: “einfache Gürtelgarnituren des mittleren 5. Jahrhunderts”), such as have been found in the gravefield of Rhenen (Donderberg); see Fig. 3 below.7 The extraordinarily large grave-field Donderberg was already discovered and excavated in the early fifties of the last century, but the publication of the finds took several decades. The west-side of this grave-field is the oldest part, and we find our parallels for the Bergakker Scabbard Mount in grave 833 of this western part (Wagner/Ypey 2011, 600‒604; drawing see Fig. 2). The grave can be dated to Fundgruppe B (Böhme 1996; Wagner/Ypey 2011, 32 f.). There are parallels with a man’s grave, no. 143A, in Vron, dép. Somme, France (Böhme 1996, 96), and from grave 6 from Samson, prov. Namur, Belgium (Böhme 1994, 78), all dated to the second third of the fifth century (Böhme 1996, 96), that is, between AD 435 and 465. The Bergakker find was part of a hoard or deposit, which was buried shortly after 500 since no object in that hoard can be dated later than 500. The Rhine estuary was still under Roman influence at that time, the Rhine border stayed intact at least until 455 (Lanting/Van der Plicht 2009/10, 83‒85). Roman military structure

6 In the Frankish territory in northern Gaul of the fifth and sixth century, we may count the disc brooch of Chéhéry with Latin capitalis and runes, the bulla of Arlon, the recently discovered ringsword from St Dizier alu (Fischer 2013), two other ring-swords from Grenay and Fréthun (Pas de Calais) with runes, and the Borgharen belt buckle bobo among the possibly ‘Frankish’ runic objects. The brooch with a fuþark and two names from Charnay-lès-Chalon, on the Saone, France, might be regarded an ‘outlier’. 7 Rhenen lies on the north bank of the Nederrijn, some 28 kilometers west of Arnhem and some 17 kilometers from Tiel.

Frisian Runes Revisited and an Update on the Bergakker Runic Item

111

was kept alive in the whole Middle Rhine area until 455/459 when the region came under Frankish rule with a king in Cologne. In Germania II, Frankish foederati had already been active for some time. The question is, who made the runic inscription on the Bergakker mount? Who could write runes in the early fifth century? If not the maker of the mount, who, by the way, was a fine craftsman,8 it might have been somebody local – although no runes were used in the wider surroundings.9 It must have been somebody who knew Latin and Germanic, like the person who wrote the runes on the Fallward footstool. Both the coastal area and the Rhine estuary were multi-cultural regions where all kinds of ethnic exchange took place. Both objects, Fallward and Bergakker, show that Germanic people were somehow integrated into Roman civilisation, but also retained their Germanic identity, which is typical of the period. The excavations in the Betuwe area (urged by the construction of a railway from Rotterdam to the Ruhr area in Germany) have shown that the Germanic and Celtic tribes who were living there became integrated into Roman culture. Their material culture changed deeply under Roman influence, but on the other hand, all kinds of local traditions were kept alive.10 The area of Fallward (Landkreis Cuxhaven) is outstanding for the amount of Roman finds, especially late Roman military girdle mounts and tutulus fibulae (Schön 2003, 35 f.). The personal equipment of Roman veterans was brought home by their owners after 25 years of service in the Roman army. They were buried with it. The Fallward and the Bergakker inscriptions were in all probability made by persons with a comparable background, partly Romanised but still securely rooted in their Germanic descent.

8 According to Böhme (1994, 78 f.), military equipment was made in workshops of Late Antiquity, probably in Northern Gaul. At the time this equipment was made and used, the Roman military organisation in Northern Gaul, under the supervision of Aëtius, was still totally intact. 9 There is the Liebenau disc from the fourth century, possibly part of a sword belt. The runes are very difficult to identify. Except for the Aalen neck ring, reading noru and dated first half of the fifth century, and the Fallward or Wremen footstool, all other Continental runic objects date from the sixth century or later. Exceptions are bracteates and some Gothic runic objects in East Europe. 10 Two altar stones and their inscriptions may underscore this assumption. On the Bergakker site, a Roman altar stone was found, dedicated to the Germanic goddess Hurstrga. A second Roman altar stone was found near Tiel as well, in Zennewijnen, a place on a filled-up stream called Zenne, a tributary to the Waal. This altar bears the text: “Deae / Seneucaege / Ulfenus beneficiarius tribuni / legionis Tricesimae Ulpiae Victricis Severiane / [Alexandriane….] / aram cum ede sua a se (or: a solo) / refecit. Votum solvit libens merito imperatore / domino nostro Severo / [Alexandro]” (Toorians 2007, 137‒143). The dating is between AD 222 and 235. In the context of this paper, it is interesting that a mix of three languages/cultures appears in the inscription: Latin, Germanic and Celtic. This agrees completely with the area in which this goddess was worshipped. Celtic was used centuries before the Romans came, and remained in use for some time, especially in matters of local religion (for instance the cult of Hercules Magusanus in the same river delta). The man who placed the altar had a Germanic name, Ulfenus, the goddess was Celtic, the inscription is in Latin.

112

Tineke Looijenga

4 Runic Literacy The question arises as to how runic literacy must be understood in these contexts, in the early fifth century, shortly before or during the migrations that brought runes from Scandinavia to southern and western regions. It may be that the spread of runic knowledge went through contacts between Germanic soldiers in the Roman army. Anyhow, the source region was present-day northern Germany or southern Denmark (roughly between the mouths of Weser and Eider). Therefore, I suggest looking especially to the Saxones. They are mentioned in the historical sources for the first time in 356 by Julianus in a panegyric for Constantius II. Also, they are mentioned as attackers in contemporary Noord Brabant in 370 by Hieronymus, Chronicon 373; Ammianus Marcellinus, XXVIII, 5, and XXX, 7, 8, and by Eunapius’ Brevarium, commissioned by Emperor Valens (364‒378). Their name is used for pirates in British and Gaulish coastal regions until 440, when Saxon auxilia in the British army revolted and claimed parts of Britannia to found small kingdoms of their own (Lanting/Van der Plicht 2009/10, 70). The Franks, or some of them, may originate from the area along the German North Sea coast (the area of the Chauci, cf. Taayke 2003, 1 f.). But the rarity of runic finds of the fourth and fifth centuries south and west of the North German/Danish source regions is not enough to establish either Saxones or Franks or whoever as diffusers of runic literacy. There are some runic objects in Britain dated to the fifth century, though: the Spong Hill urns, maybe the Loveden Hill urn, the Caistor-by-Norwich astragalus (in an urn). That runic objects are absent from the wide area between Northern Germany and Southern Germany until the sixth century might be due to burial customs (cf. Böhme 1999). Cremation was customary among the Saxons living in the area now called Westfalen. In Southern Germany, the custom of corpse burials in the sixth/seventh century yielded a notable amount of runic objects, all found in graves. The question is: should the ‘Frisian’ corpus of some twenty pieces be regarded as different from the early English runic corpus? In what way are their inscriptions different? And if not, can we regard the runic Anglo-Saxon-Frisian objects as one corpus? In that case we might be able to compose a more coherent study based on a series of connected features shared by the whole group of ASF runic texts. The runic tradition of the Anglo-Saxons-Frisians started as a common tradition and has its origin in the older North Germanic tradition. Bos/Brouwer (2005) and Nicolay (2005) consider the migrating Anglo-Saxons to have been initially still bound with rather strong links to their homelands during the first phase of the migrations. In a later phase, they became more and more independent and started to develop their own products and styles. This can be seen in pottery and metalwork, and, I should like to add, in their runic usage from the sixth century onwards. The first stream of colonists (ca. AD 400‒450) brought typical cruciform brooches with them, made in their homeland. These brooches are found in Eastern England and Frisia and they are very much alike. The brooches were, in all probability, made in Schleswig-Holstein and the area

Frisian Runes Revisited and an Update on the Bergakker Runic Item

Fig. 3: Grave 833 from the Donderberg, Rhenen (Wagner/Ypey 2011, 602).

113

114

Tineke Looijenga

between Elbe and Weser (Bos/Brouwer 2005, 18). After ca. AD 450, these brooches appear in a larger area: Frisia, England, Northern Germany, Denmark, Sweden and Norway. Also, Nicolay (2005, 73) argues that the strong similarities in material culture between England and Frisia point to a common homeland of the immigrants. Bede was right, he says, in mentioning Jutland, Schleswig-Holstein and the area between Elbe and Weser as the homelands of Jutes, Angles and Saxons. A second phase in the migrations is testified to by the import of bracteates from Scandinavia, mainly in the sixth century (the mysterious Undley bracteate is dated AD 450, cf. Hines 1990, 440, 1). In the sixth century Frisian relations with Jutland seem strongest. Seebold (2003, 29 f.) points to the remarkable fact that Frisians and Jutes were sometimes considered as similar in ancient sources: Venantius Fortunatus mentions Eucii and Saxones together as adversaries of the Franks. Seebold (2003, 30) is of the opinion that this only makes sense if the ‘Euten’ were Frisians. And, Im Beowulf 11 wurden die gleichen Leute teils Friesen, teils Euten genannt, und schließlich nennt der Frankenkönig Theudebert I12 in einem Brief an Kaiser Justinian eutische Sachsen (cum Saxonibus Euciis, qui se nobis voluntate propria tradiderunt).

Jutland may have been the source region of a second stream of colonists who partly settled in Frisia. The first stream of immigrants in the coastal area of former Frisia (ca. AD 440) came from Schleswig-Holstein and the area between Weser and Elbe, the second (after AD 500) from further north, especially from Jutland. Bracteates and other golden objects in Frisia are found in deposits, not in graves, and thus belong to the central South Scandinavian world, where there are ritual deposits in a settlement context (outside this central area, bracteates are found in graves, in England, Norway and the Continent). This means, according to Nicolay (2005, 85), that the relations with the homelands were both cultural and ideological. The first to question the authenticity of Frisian runic writing was Ray Page in his article of 1996, which he presented at the First International Symposium on Frisian Runes, in 1994. He used the expression “baffling” for the Frisian runes, and this is still appropriate. In his article he clearly favours the idea of Frisian runic objects being not typically Frisian. Earlier in his paper, he is still looking for a possibility to distinguish the Frisian from the English corpus, and, after examining the oft-listed diagnostic differences (the -u-ending and the ā < Gmc. * au), he rejects these features because they prove to be not typically Frisian. But he suggests, one way in which we could fruitfully probe the evidence of the English and Frisian runic inscriptions: to seek if they demonstrate similar practices in presenting their messages; if such practices differ from those of the rest of the early runic world (Page 1996, 147).

11 ‘In Beowulf the same people were sometimes called Frisians, sometimes Jutes, and finally the Frankish king Theudebert I mentions Jutish Saxons…’ (transl. KK). 12 Theudebert I lived in the first half of the sixth century, more than a century later than the first stream of colonists.

Frisian Runes Revisited and an Update on the Bergakker Runic Item

115

In his paper, he already treats some of the differences in practice between the OFris. and OE corpora, so I will not repeat them here. One conclusion, however, is striking: the reason that the runes became more popular among the Anglo-Saxons in England is that there was a genuine need in daily use for a vernacular script. If so, are the […] not very informative Frisian inscriptions an indication of the only desultory use of the script in that region, because for them it served no obvious commercial or practical purpose (Page 1996, 147).

Traditionally, the Old English and Old Frisian runic corpora are to be distinguished on the grounds of two linguistic differences. This concerns sound changes that could be associated with Old Frisian rather than Old English: the monophthongisation of au > ā and the ending -u,13 both found, for instance, in the legend skanomodu with ā < *au and -u < *-az/*-an. The first change is in fact ‘Ingvaeonic’; on the second change see below. Especially Nielsen has attempted to demonstrate in several publications that the ending -u is not a Frisian linguistic feature in any strict sense. In his 1996 paper he states: Formally, -u could be a reflex of the ō-stem nominative suffix, Germanic *-ō, the regular reflex of which in all North and West Germanic languages would be -u, cf., e.g., early runic (nsf.) ō-stem laþu ‘invitation’ (Darum Bracteate I).

He suggests further on that skanomodu might be a woman’s name since “Old Saxon feminine names in mōd outnumber masculine ones.” The final -u must then be taken as reflecting a Murmelvokal (Düwel/Tempel 1968, 382, 390) and not as a reflex of West Gmc. n/asm a-stem suffixes *-az/*-an (Nielsen 1996, 128 f.). Nielsen adds that “scholars would be wise to look for other sources.” I do agree: the final -u reflects an unstressed /ə/. Besides, Page (1996, 141) remarks that some undoubtedly English inscriptions [...] have ‘u’ in final and unexplained place […] ‘giuþeasu’ and ‘flodu’ on the Franks casket, […] and the element ‘benu’ on the early runic coins.

Nielsen (1999, 51; and 1984, 18, n. 1) claims that the development ā < au (cf. skanomodu with skān- < *skaun-) is not restricted to OFris. but is also found in OS (cf. Heliand). Even katæ (Southampton knucklebone, considered Frisian) might not be OFris. after all (katæ < *kautōn). Final -æ seems to point to England, where an abundance of this feature in runic legends occurs (Waxenberger 2006, 278 f.; see for instance her list A.2: dat. sg. of masc. and fem. nouns in -æ). Waxenberger (2006) refers to Campbell (1959, § 369): “æ, e, and i fell together in a sound written e in unaccented syllables. æ and i remain undisturbed only in very early texts.” Gaby Waxenberger (2006) lists a number of instances, which makes me wonder if the OE final -æ had its counterpart in the final -u of the OFris. inscriptions. In both OE and

13 Cf., for instance, Nielsen (1996, 123‒129), and Page in the same volume.

116

Tineke Looijenga

OFris. æ and u represent vowels in unstressed syllables, such as can be found in habuku (Oostum comb) and æniwulufu (Folkestone/Glasgow thrymsa – gold coin), both presenting final ‹-u› and both having an unstressed vowel in the middle (in æniwulufu it may be a parasite vowel). However, there is no consensus as to the form: are they nominatives or datives? A nominative or a dative is in both cases possible: æniwulufu is most likely a personal name (PN), perhaps a dative, a-stem *æniwulfaz. habuku as a PN can be nom. sg. fem. of an a-stem *habukaz or of an ō-stem *habuko; habuku may also be the indirect object of the sentence and is then a dative sg. fem. ō-stem ‘for habuku (Habækæ, Habeke)’. To illustrate this name form, I would like to point to the much later attestations of the 14th, 15th and 16th centuries. Here we find, in records about the estates of the ‘Groninger Jonkers’ (local rulers), an abundance of names ending in -eke /ǝkǝ/.14 According to Kortlandt (1999, 48), there are sound changes common to Frisian and Anglian: Anglian shared the development of Frisian on the Continent, in particular the raising of ǣ to ē which had been preceded by the Anglo-Frisian retraction of ǣ to ā before w. […] After the ‘Anglian’ migration, Frisian fronted ā (from *ai) to ǣ unless it was followed by a back vowel in the following syllable and monophthongized *au to ā.

Kortlandt posed the question “Did the Old English dialects first diverge in Britain or on the Continent?” His conclusion is that neither view is correct […] and that the early divergences are the result of a chronological difference between two waves of migration from the same dialectal area in northern Germany, an earlier ‘Saxon’ invasion in the fifth c. and a later ‘Anglian’ invasion around the middle of the sixth c.

According to John Hines (pers. comm. in an email of 12/10/2012), “archaeology really cannot support the idea of an earlier Saxon invasion and a later Anglian one”. Since the runic innovations are found in both Frisia and Britain, we can safely conclude they had at first 15 a common, uniform writing system. There may have been

14 These male and female names are taken from Formsma et al. De Ommelander Borgen en Steenhuizen 1987 (castles and towers in the region around Groningen). These diminutive names form a minority but are nevertheless illustrative of name-giving in Groningen. Instances are: Ripeke Aykema, Beteke Aykema, Dodeke Allersma, Emeke Asinga, Doedeke Boeltzertzema, Dodeke and Popeke ter Borch, Reneke Busch, Emeke Dodekema, Renneke Elama, Teteke Entens, Abeke (!) van Ewsum, Reneke Fraylema, Reneke Gaykinga, Edzeko to Garreweer, Dydeke toe Godlinze, Doke te Godlinze, Vrouweke to Godlinze, Dodeke (Doeko) Grevinge, Reneka Hankema, Popeke Herathema, Elteke ten Holte, Teteke Jarges, Reweke to Kantens, Ludeke Clant, Reneke Busch de Marees van Swinderen, Ludeca de Mepsche, Abeko te Mude, Abeke Onsta, Hiddeke van Oosterwijtwerd, Reindeke Reynsma, Unico Ripperda, Beteke Scheltkema, Elteke de Sighers (all hypocoristics, both masculine and feminine). 15 According to Gijsseling, Frisian as a specific language came into being in the eighth century. See: “Het oudste Fries”, It Beaken XXIV (1962).

Frisian Runes Revisited and an Update on the Bergakker Runic Item

117

differences in pronunciation, and there were probably dialects. Besides the shared fuþorc with two new runes for /o(:)/ and /a(:)/, the Frisians employed different, enigmatic, runic forms, such as can be found in the Westeremden B, Britsum and Wijnaldum inscriptions. These forms seem to point to Scandinavia, which is quite possible in view of the networks people participated in, and in view of the composition of the inhabitants of Frisia after the Migration Period, starting in the fourth century. Page, Nielsen and myself have observed that several runic traditions may be reflected in the Old Frisian corpus. This should not surprise us, Frisia being at a cross-roads on the coast of the North Sea between Scandinavia, Northern Germany and Britain. This cross-roads concept may be illustrated by a recent report by the British student Ellen McManus, who executed an investigation into “stable isotopes of strontium (Sr) in several skeletal remains (teeth and bones) from the cemetery of Oosterbeintum”, excavated in 1988 and 1989, and dated to the early Middle Ages. McManus (2010, 81 f.) finds that a 20 to 30-year-old man and two individuals of more than 45 years of age from a double grave originated from areas with a much higher 87Sr/86Sr relation than that of the other individuals and animals buried. This is an indication of very old rock formations, such as can be found in Scandinavia. This double grave is one of the earliest burials of the grave field, dated AD 440‒ 485. Some five other individuals also display high 87Sr/86Sr values and might originate from North Denmark. This, however, is speculative although in accordance with archaeological evidence. According to McManus et al. (2013, 273), [i]t is possible that some of these non-locals originated in the Anglo-Saxon ‘homelands’ of southern Denmark and northern Germany, as proposed by Van Giffen […]. However, the range of 87Sr/86Sr and δ18O at Oosterbeintum cannot be explained solely by this particular migration event. Individuals must have also migrated from more distant regions, such as England and Scandinavia. These results, therefore, have implications beyond Frisia itself. The distance and diversity of the geographical origins of the non-locals buried at the cemetery suggests that Oosterbeintum, and by inference the wider terp region, was part of a wide network of contact and movement that spanned the North Sea region.

5 The Frisian Runic Corpus Let us have a look at the corpus itself. There are some new views: the solidi from Harlingen and Schweindorf might be dated rather later than usually assumed; i.e., in the first half of the seventh century. Moreover, the Schweindorf legend clearly shows five runes instead of six: welad, with no final rune -u, which is the traditional view (see Looijenga 2013, 431). All solidi may be regarded as either English or Frisian, although England has a numismatic context for coins, which lacks in Frisia at this date. In fact skan- (in skanomodu) is probably the clearest piece of evidence for the ‘Frisianness’ of this group of coins, if it is a group and if they are coins,

concluded Page (1996, 141) after a long overview of the coins and their legends.

118

Tineke Looijenga

Fig. 4: Schweindorf solidus. The runes run from right to left and read: welad; © Ostfriesisches Landesmuseum Emden; photo: C. Kohnen, Ostfriesische Landschaft Aurich.

A random comparison between some more or less similar (form, length, type) inscriptions of the early OFris. (left column) and OE (right column) corpora dated from the fifth to the eighth century and consisting of personal names and names of objects may give us an impression of the similarity of both corpora (the more or less enigmatic ones excluded); the transliterations are mine (cf. Looijenga 2003, Chapters 8 and 9). A em ura aib kabu deda habuku katæ kobu, skanomodu, æniwulufu welad edæ:boda adugislu me gisuhldu katæ tuda, oka, hada, aha æpa

B em sigimer gægogæ maga medu; luda gibœtæ sigilæ sigilæ medu; benu sїþæbad æko:?œri; godaluwaluda hariboki:wusa raïhan luda, pada, desaiona epa, æpa

The texts above are taken from several inscriptions for the sake of comparison – line 1: A: Ferwerd, B: Ash/Gilton; line 2: A: Oostum, B: Undley and Harford Farm; line 3: A: Hamwih, B: Harford Farm; line 4: A: Toornwerd and two runic solidi, B: Undley, upper Thames Valley gold coin; line 5: A: Schweindorf, B: Loveden Hill; line 6: A: Arum, B: Chessell Down, Whitby; line 7: A: Westeremden, B: Watchfield; line 8: Hamwih, B: Caistor-by-Norwich; line 9: A: Bernsterburen, Rasquert, Harlingen solidus, Hantum, B: Harford Farm, Kent gold coin, Suffolk shilling; line 10: A: Midlum sceat, B: Kent sceattas.

Observations: The legends show no very specific differences in length, depth, degree of information. Resemblances seem to be there, nevertheless, although sometimes in the absence of meaningfulness. There are names of persons and objects, a few verbs, and

Frisian Runes Revisited and an Update on the Bergakker Runic Item

119

some words which we do not know what they mean. There are some short plain sentences: ‘Luda repaired brooch’ and ‘Aib comb Habuku made’. In fact, you could well exchange the one (OFris.) for the other (OE). Common to all older runic texts is their brevity, the laconic tone, which makes us wonder whether this specific script was designed for simple, plain messages like these? Or is what we see merely a small and insignificant part of something much grander and more elaborate? These astonished considerations seem to be made by every researcher of runes; however, the inscriptions might not be so peculiar, for if we compare this kind of inscriptions to contemporary Roman use, for instance, there is not so great a difference. Everyday Roman use is also restricted to names: makers’ formulae, naming the object or the owner etc. In later days pilgrims wrote their names in runes on the walls of Italian cemetery walls and Roman catacombs. This shows that runes were in fact in general use, at least from the eighth/ninth century onwards. Interesting similarities are to be noted in endings (personal names ending in -a and -u, names of objects ending in -u), as well in the domain of syntax (very short sentences with subject, verb and object). In two cases a rune appears that is not in the ‘standard’ ASF fuþark; the rune ᚴ in Chessell Down aco:?œri has the same form as the rune transliterated as o in Britsum borod.16 This rune form is reminiscent of a form found in some inscriptions near Lund, Skåne, for instance on the Skårby Stone, now in the garden of “Kulturen” in Lund. In Britsum, this graph has been taken to denote a vowel, perhaps o, for the sake of readability and because it might be a variant of the new ōs-rune. In Chessell Down it may indicate a consonant, though. The use of runes on the Continent lasted only about a century or two and, judging by the frequently occurring personal names, seems to be restricted to a personal and private area. This may not be so surprising if it is taken into account that during Christianisation in western Europe people were allowed to keep to their usual name giving: as can be seen from names such as Alcuin, Hraban, Angilbert, Theodulf, Einhart, and the many Germanic names on otherwise Latin-inscribed gravestones from the late Roman period (for instance Hariulfus, Unfachus, Aldualuhus, Modoaldus). If a personal, vernacular, sphere was allowed for, runes fitted in. Runes in Merovingian and Carolingian times were often used in a hidden way: on the back of brooches, weapons, and so on. Only in places where Christendom was still a far way off, open use of runes could take place in elaborate texts in the public domain. However, remarkably, in England runes did not disappear after Christianisation, on the contrary, they seem to have been widely used in the ecclesiastical sphere. On the Continent, however, runes disappeared rapidly after the Christian reform of the Carolingians.

16 The Britsum inscription reads: þoniaberetdud //n borodmi LIU.

120

Tineke Looijenga

6 Naming the Object What is remarkable is the explicit mentioning of the object the runes are carved in. We find combs, bones, brooches (sil, sigil; cf. Looijenga 2003, 278, 285). A new interpretation of the Vimose comb with the legend harja is suggested against the background of five other combs with the title ‘comb’ – taken that the meaning of harja in Finnish is ‘brush, comb’. In January 2018, a sixth comb with the legend kabaR was reported in an online newspaper, (DR), from Ribe, written in the younger futhark, dated to the ninth century. The comb from Frienstedt (Germany) has a runic inscription reading kaba (see Schmidt/Nedoma/Düwel 2010/11). Two combs from Frisia (now province Groningen) have resp. kabu (Oostum) and kobu (Toornwerd). A comb from Elisenhof has kabR. The legends thus show distinction between West Gmc. kaba and Old Norse kabR. The Frienstedt comb was found near Erfurt, but, judging from its type, it may originate from a large area, from the Rhine and Meuse area to Frisia and Saxony, England, even Switzerland (see Roes 1963, 10 f.; esp. Plates VI–X). There are more combs with runes; actually, it is surprising that runes should appear so often on a humble personal object, or is that not precisely what runes were made for? Some decades ago, a comb with runes was found near Belgorod, in Russia (Ščukin/Kazanski/Sharov 2006, 48).17 The runes may be read kunla, possibly a name. Furthermore, we have combs from Whitby, Setre, Amay (Belgium), in fact from all over the runic world, from the second to the ninth century. Behind and beside this unpretentious behaviour must be a large application of script. It is only that all we have actually found of early runic usage is mostly restricted to objects with one or two words or cryptic texts. This is mainly due to circumstances: archaeology and coincidence. One may ask why somebody would want to write the object’s name when everyone could see what the object was?

7 Conclusion As a conclusion to this paper, I suggest merging both OE and OFris. corpora, for joint research and comparison. I am convinced that the early OE and OFris. runic inscriptions belong to an inseparable and common tradition that should be called Anglo-Saxon. The conclusion that a Frisian runic corpus existed simply because runic objects have been found in Frisia is not satisfactory. It suggests that something existed that probably did not: a specific Frisian runic culture. And if the new runic

17 The Ukrainian archaeologist Maxim Levada kindly alerted me to this remarkable find by email.

Frisian Runes Revisited and an Update on the Bergakker Runic Item

121

innovations, the development of the āc- and ōs-runes, took place in England, then there is no reason for giving Frisian runes a status aparte.18 An additional conclusion seems to be that runic writing in the early period should be understood against the background of the relations with the Roman Empire – runic writing emerged in the Roman Imperial period: a time of progressive interaction between Germanic peoples and Romans. Instead of adapting the Roman script, there clearly was a role for an indigenous script to Germanic speaking people – if only we knew why?

Appendices 1 The Oldest Attestations of the Name of the Frisians Pliny and Tacitus both use the names Frisii and Frisia. Cassius Dio writes about Phreisioi, Phrisioi, Ptolemy mentions Phrisioi, Phrissioi, and Procopius has Phrissones.19 A third-century Roman inscription has Frisiones (Gerrets 2010, 130 f.). Pliny further mentions the Frisiavones, maybe not a Germanic form (Nielsen 1999, 35). According to Tacitus, Germania 34, we have the Frisii maiores and Frisii minores, and the Frisiavones. The first group lived in Central Frisia, the second in what is now Holland and the third in the Celtic-speaking area Helinium, initially the area between the three river branches of the Rhine mouth, the Oer-IJ, Utrechtse Vecht and Oude Rijn; later Helinium was used to indicate the mouth of the Maas and Schelde. Remarkably, Bede does not mention Frisians among the invaders of England. He only speaks of Angles, Saxons and Jutes. Frisians are not mentioned in any Roman or Frankish source between AD 293 and 553 (Procopius, History of the Wars, Book VIII).

2 What Happened to the Old Frisians? During the third century, mass emigration began in Westergo (Friesland), probably caused by four factors: strong population growth, scarcity of resources, drainage problems due to the shrinking of the peat moors in the hinterland, and drastic changes in the Limes area of the Roman Empire. Frisii and other tribesmen such as

18 We need to define more clearly the grounds on which to decide the ‘Frisianness’ or ‘AngloSaxonness’ or ‘Frankishness’ of runic inscriptions. Why is there a need to label an inscription according to ethnicity? Is there no better way to describe runic culture and runic usage? 19 In History of the Wars VIII (xx, 7‒8), Procopius writes that the island of Brittia is inhabited by Angloi, Frissones and Brittanes (Procopius 47‒58). Procopius is quite ambiguous about what he heard about the island Brittia since the stories are in his opinion on the verge of mythology (Procopius 42‒46).

122

Tineke Looijenga

Chauci, Chamavi, and Saxones are recorded as pirates along the coast. It is conceivable that Frisii, Chauci, and other coastal inhabitants were regarded as Franci, the latter being an umbrella term used by the Romans. Franci are first mentioned in 291 (Panegyrici Latini XI) and it seems that their name must be understood as an overall term for tribes such as Chauci, Chamavi, Frisii, Bructeri, Salii, and others (Lanting/ Van der Plicht 2009/10, 66 f.). Frisii are recorded from Northern Gaul in 297. During the period following, prosperity increased in that region, which may have attracted immigrants from northern regions. Late third, early fourth century earthenware found in Zelle in East Flanders points to the presence of Frisii in the Scheldt area. Again, it is conceivable that Frisii merged with the Franci and went with them to Northern Gaul. The old Frisians are mentioned one last time in a Roman source (Pan. Lat. VIIIO (V) 9.3 Eumenius, Pan. Constantio Caesari) of around 297, when a rather large group was captured by the Romans near the Schelde and set to work as ploughmen in Gaul as dediticii ‘serfs’. Another group may have been selected for service in the Roman army. A fact worth mentioning is that a contingent of Frisians was stationed as military at Hadrian’s Wall in the third century. We cannot estimate how many Frisians opted for a military career in the Roman army, but this may be one of the incentives for leaving their homeland and homesteads. For instance, of the Civitas Cananefates, a tribe living in or next to the Frisian area in North Holland, a thousand young men were recruited to serve in the Roman Army, which must have been a severe loss to the tribe (Lanting/Van der Plicht, 2009/10, 56). During the fourth century, only Saxons lived along the North Sea coast and it was probably Saxons who re-colonised the old Frisian terpen region, revitalising the very homesteads the Frisians left behind. Among their remnants of material culture, no traces of the old population are found (cf. Taayke et al. 2013). From ca. 440 onwards, pottery and brooches appear in Frisia which have their origin in Saxon regions (Lanting/Van der Plicht 2009/10, 76). Saxones seems to be a collective name for several tribes, just like the name Franci. Incidentally, the name Angli Saxones seems to have been used for the first time by Paulus Diaconus († 790/95) in his Historia Langobardum as an indication for ‘English Saxons’ to be distinguished from the Antiqui Saxones, who still lived in northern Germany (mainly Westfalen).

References Ammianus Marcelinus. Rerum gestarum libri qui supersunt. Ed. J. C. Rolfe. Cambridge, MA/ London: Loeb Classical Library. 1972‒1986. Bammesberger, Alfred. 1999. “Die Runeninschrift von Bergakker: Versuch einer Deutung”. Pforzen und Bergakker: Neue Untersuchungen zu Runeninschriften. Ed. Alfred Bammesberger in editorial cooperation with Gaby Waxenberger. Göttingen: Vandenhoeck & Ruprecht. 180‒185. Beda Venerabilis. Historia ecclesiastica gentis Anglorum. Ed. Charles Plummer. Oxford: Oxford University Press. 1946.

Frisian Runes Revisited and an Update on the Bergakker Runic Item

123

Böhme, Horst Wolfgang. 1994. “Der Frankenkönig Childerich zwischen Attila und Aëtius. Zu den Goldgriffspathen der Merowingerzeit”. Festschrift für Otto-Herman Frey zum 65. Geburtstag. Eds. Horst Wolfgang Böhme and Claus Dobiat. Marburg: Hitzeroth. 69‒110. Böhme, Horst Wolfgang. 1996. “Söldner und Siedler im spätantiken Nordgallien”. Die Franken, Wegbereiter Europas: vor 1500 Jahren: König Chlodwig und seine Erben. Eds. Alfried Wieczorek and Raymond Brulet. Mainz: von Zabern. 91‒101. Böhme, Horst Wolfgang. 1999. “Franken oder Sachsen? Beiträge zur Siedlungs- und Bevölkerungsgeschichte in Westfalen vom 4.–7. Jahrhundert”. Studien zur Sachsenforschung 12: 43‒73. Bos, Jan Maarten and E. W. Brouwer. 2005. “De kruisvormige fibulae van Friesland”. De Vrije Fries 85: 9‒36. Bosman, Arjen V. A. J. and Tineke A. Looijenga. “A Runic Inscription from Bergakker (Gelderland), The Netherlands”. Amsterdamer Beiträge zur älteren Germanistik 46: 9‒16. Campbell, Alistair. 1959. Old English Grammar. Oxford: Clarendon. Cassius Dio Cocceianus. Historiae Romanae. Ed. Earnest Cary. Cambridge, MA/London: Loeb Classical Library. 1955‒1982. De Boone, Willem Jan. 1954. De Franken, van hun eerste optreden tot de dood van Childerik. Dissertatie Rijksuniversiteit Groningen. Groningen: s.n. Derks, Ton and Nico Roymans. 2007. “Bronzen zegeldoosjes en Latijnse schrift-cultuur”. Een bataafse gemeenschap in de wereld van het Romeinse Rijk. Opgravingen te Tiel-Passewaai. Eds. Nico Roymans, Ton Derks, and Stijn Heeren. Utrecht: Matrijs. 131‒136. Dierkens, Alain and Patrick Périn. 2003. “The 5th-century Advance of the Franks in Belgica II”. Essays on the Early Franks. Groningen Archeological Studies 1. Groningen: Barkhuis. 165‒193. Fischer, Svante, Jean Soulat, and Teodora Linton Fischer. 2013. “Swordparts and their Depositional Contexts – Symbols in Migration and Merovingian Period Martial Society”. Fornvännen 108: 109‒122. Gerrets, Daniël A. 2010. Op de grens van land en water, dynamiek van landschap en samenleving in Frisia gedurende de Romeinse tijd en de Volksverhuizingstijd. Dissertatie Rijksuniversiteit Groningen. Groningen Archaeological Studies 13. Eelde: Barkhuis. Gijsseling, Maurits. 1962. “Het oudste Fries”. It Beaken XXIV: 1‒26. Henning, Rudolf. 1889. Die deutschen Runendenkmäler. Straßburg: Trübner. Hines, John. 1990. “The Runic Inscriptions of Early Anglo-Saxon England”. Britain 400–600: Language and History. Ed. Alfred Bammesberger. Heidelberg: Winter. 437–456. Hines, John. 1996. “Coins and Runes in England and Frisia in the Seventh Century”. Frisian Runes and Neighbouring Traditions. Amsterdamer Beiträge zur älteren Germanistik 45. Eds. Tineke Looijenga and Arend Quak. Amsterdam: Rodopi. 47‒62. Kortlandt, Frederik. 1999. “The Origin of the Old English Dialects Revisited”. Amsterdamer Beiträge zur älteren Germanistik 51: 45‒51. Lanting, J. N. and J. Van der Plicht. 2009/10. “De 14C chronologie van de Nederlandse Preen Protohistorie VI: Romeinse tijd en Merovingische periode, deel A: historische bronnen en chronologische schema’s”. Palaeohistoria 51/52: 27‒168. Looijenga, Tineke. 1999. “The Bergakker Find and its Context”. Pforzen und Bergakker: Neue Untersuchungen zu Runeninschriften. Ed. Alfred Bammesberger in editorial cooperation with Gaby Waxenberger. Göttingen: Vandenhoeck & Rupprecht. 141‒151. Looijenga, Tineke. 2003. “Texts and Contexts of the Oldest Runic Inscriptions”. The Northern World 4: 317‒322. Looijenga, Tineke. 2013. “Die goldenen Runensolidi aus Harlingen und Schweindorf”. Land der Entdeckungen: Die Archäologie des friesischen Küstenraums / Land van Ontdekkingen. De archeologie van het Friese kustgebied. Begleitband zu den Ausstellungen / Catalogus bij de tentoonstellingen 2013‒2014. Ed. Jan F. Kegler. Aurich: Ostfriesische Landschaft. 430‒431.

124

Tineke Looijenga

Looijenga, Tineke. 2021. “Runic Literacy in North-West Europe, with a Focus on Frisia”. Frisians of the Early Middle Ages. Eds. John Hines and Nelleke IJssennagger-van der Pluijm. Woodbridge: Boydell & Brewer. 375–400. McManus, Ellen T. 2010. “An Isotopic Investigation of the Early Medieval Cemetery of Oosterbeintum, Friesland, the Netherlands”. Master’s thesis Archaeological Sciences University of Bradford. McManus, Ellen T., Janet Montgomery, Jane Evans, Angela Lamb, Rhea Brettell, and Johan Jelsma. 2013. “To the Land or to the Sea: Diet and Mobility in Early Medieval Frisia”. The Journal of Island and Coastal Archaeology 8 (2): 255‒277. Nedoma, Robert. 2010. “Schrift und Sprache in den Ostgermanischen Runeninschriften”. The Gothic Language: A Symposium. NOWELE 58/59. Ed. Hans Frede Nielsen. Odense: University of Southern Denmark Press. 1–70. Nicolay, Johan. 2005. “Nieuwe bewoners van het terpengebied en hun rol bij de opkomst van Fries koningschap”. De Vrije Fries 85: 37‒103. Nielsen, Hans F. 1996. “Developments in Frisian Runology: A Discussion of Düwel & Tempel’s Runic Corpus from 1970”. Frisian Runes and Neighbouring Traditions. Amsterdamer Beiträge zur älteren Germanistik 45. Eds. Tineke Looijenga and Arend Quak. Amsterdam: Rodopi. 123‒130. Nielsen, Hans F. 1999. “Foar-Aldfrysk: In oersjoch”. In skiednis fan ‘e Fryske taalkunde. Eds. Anne Dykstra and Rolf Hendrik Bremmer. Leeuwarden: Fryske Akademy. 34‒74. Nieuwhof, Annet. 2011. “Discontinuity in the Northern-Netherlands Coastal Area at the End of the Roman Period”. Transformations in North-Western Europe (AD 300‒1000), Proceedings of the 60th Sachsensymposion 19.‒23. September 2009 Maastricht. Neue Studien zur Sachsenforschung 3. Ed. Titus A. S. M. Panhuysen. Stuttgart: Theiss. 55‒66. Odenstedt, B. 1999. “The Bergakker Inscription: Transliteration, Interpretation, Message: Some Suggestions”. Pforzen und Bergakker: Neue Untersuchungen zu Runeninschriften. Ed. Alfred Bammesberger in editorial cooperation with Gaby Waxenberger. Göttingen: Vandenhoeck & Ruprecht. 163‒173. Page, Raymond Ian. 1996. “On the Baffling Nature of Frisian Runes”. Frisian Runes and Neighbouring Traditions. Amsterdamer Beiträge zur älteren Germanistik 45. Eds. Tineke Looijenga and Arend Quak. Amsterdam: Rodopi. 131‒150. Panegyrici Latini. Ed. Emil Baehrens. Leipzig: Teubner. 1911. Paulus Diacones. Historia Langobardum. Transl. by William Dudley Foulke. University of Pennsylvania, Hist Dep. 1906. Plinius Caecilius Secundus, Gaius. Naturalis Historia. Ed. Detlef Detlefsen. Quellen und Forschungen zur alten Geschichte und Geographie 9. Berlin, 1904. Procopius. History of the Wars. Ed. Henry B. Dewing. Cambridge, MA/London: Loeb Classical Library. 1978. Ptolemeus, Claudius Alexandrinus. Geographica. Ed. Karl Friedrich August Nobbe. Leipzig. 1843‒1845. Quak, Arend. 1999. “Zu den Runenformen der Inschrift von Bergakker”. Pforzen und Bergakker: Neue Untersuchungen zu Runeninschriften. Ed. Alfred Bammesberger in editorial cooperation with Gaby Waxenberger. Göttingen: Vandenhoeck & Ruprecht. 174‒179. Roes, Anna. 1963. Bone and Antler Objects from the Frisian Terp-Mounds. Haarlem: H. D. Tjeenk Willink. Schmidt, Christoph G., Robert Nedoma, and Klaus Düwel. 2010/11. “Die Runeninschrift auf dem Kamm von Frienstedt, Stadt Erfurt”. Die Sprache 49 (2): 123‒186. Schön, Matthias D. 2003. “Sachsen – Nachbarn der frühen Franken: Überlegungen zu Bestattungssitten im 4./5. Jh.”. Essays on the Early Franks. Groningen Archaeological Studies 1. Eds. Ernst Taayke et al. Groningen: Barkhuis. 35‒61.

Frisian Runes Revisited and an Update on the Bergakker Runic Item

125

Ščukin, Mark B., Michel Kazanski, and Oleg Sharov. 2006. Des les Goths aux huns: Le nord de la mer noire aus Bas – empire et a l’epoque des grandes migrations. BAR International Series 1535. 48, ill. p. 317. Seebold, Elmar. 1999. “Die Runeninschrift von Bergakker”. Pforzen und Bergakker: Neue Untersuchungen zu Runeninschriften. Ed. Alfred Bammesberger in editorial cooperation with Gaby Waxenberger. Göttingen: Vandenhoeck & Ruprecht. 157‒162. Seebold, Elmar. 2003. “Die Herkunft der Franken, Friesen und Sachsen”. Essays on the Early Franks. Groningen Archaeological Studies 1. Ed. Ernst Taayke. Groningen: Barkhuis. 24‒34. Taayke, Ernst, Rolf Bärenfänger, and Jan F. Kegler. 2013. “Wir waren schon immer hier! Brüche und Kontinuitäten in der Besiedlung des Küstenraums / Wij waren er altijd al! Breuken en continuïteiten in de bewoning van het kustgebied”. Land der Entdeckungen: Die Archäologie des friesischen Küstenraums / Land van Ontdekkingen: De archeologie van het Friese kustgebied. Begleitband zu den Ausstellungen / Catalogus bij de tentoonstellingen 2013‒ 2014. Ed. Jan F. Kegler. Aurich: Ostfriesische Landschaftliche Verlags- und Vertriebsgesellschaft. 160‒170. Taayke, Ernst. 2003. “Wir nennen sie Franken und sie lebten nördlich des Rheins, 2.–5. Jh.”. Essays on the Early Franks. Groningen Archaeological Studies 1. Ed. Ernst Taayke. Groningen: Barkhuis. 1‒23. Tacitus, Publius Cornelius. Germania. Translated by H. Mattingly and S. A. Handford, Harmondsworth: Penguin. 1970. Toorians, Lauran. 2007. “Van SENEUCAEGA tot Zennewijnen: de talen van de Bataven”. Een bataafse gemeenschap in de wereld van het Romeinse Rijk. Opgravingen te Tiel-Passewaaij. Eds. Nico Roymans, Ton Derks, and Stijn Heeren. Utrecht: Matrijs. 137‒144. Vennemann, Th. 1999. “Note on the Runic Inscription of the Bergakker Scabbard Mount”. Pforzen und Bergakker: Neue Untersuchungen zu Runeninschriften. Ed. Alfred Bammesberger in editorial cooperation with Gaby Waxenberger. Göttingen: Vandenhoeck & Ruprecht. 152‒156. Wagner, Annette and Jaap Ypey. 2011. Das Gräberfeld auf dem Donderberg bei Rhenen: Katalog. Leiden: Sidestone Press. Waxenberger, Gaby. 2006. “The Representation of Vowels in Unstressed Syllables in the Old English Runic Corpus”. Das fuþark und seine einzelsprachlichen Weiterentwicklungen. RGA-E 51. Eds. Alfred Bammesberger and Gaby Waxenberger. Berlin/New York: De Gruyter. 272‒314. Waxenberger, Gaby. 2017. “How English is the Early Frisian Corpus? The Evidence of Sounds and Forms”. Frisians and their North Sea Neighbours. Eds. John Hines and Nelleke IJssenagger. Woodbridge: The Boydell Press. 93‒124.

Regine Marth

The Mysterious Gandersheim Casket: Are There Any Hard Facts? The so-called Gandersheim Casket (see Figs. 1‒3) is one of the most enigmatic objects in the medieval collection at the Herzog Anton Ulrich-Museum in Braunschweig. This is because, although this work is well known and widely published,1 fundamental questions about its history, the technical aspects, and the framework at the base with its runic inscription, have not so far been answered. At the Eichstätt conference which took place in March 2012, I was invited to present the casket and discuss some of the questions and queries arising from it. The present paper is a summary of that lecture.

1 Description and History of the Casket The house-shaped casket, 12.6 cm in height and width and 6.6 cm in depth, is AngloSaxon and can be dated to around 800. It is made of whalebone2 with metal fittings. Its panels are both ordered and consistent, with twelve fields on the front and six on the back. In these fields fantastic beasts and animals can be identified as animals of earth, air, and water. These are intertwined with each other or with trees. At the centre of the back is a hidden “chi” as a symbol for Christ. These features undoubt-

1 See the proceedings of the international conference Das Gandersheimer Runenkästchen, Internationales Kolloquium, ed. Regine Marth, Braunschweig 2000, with essays dealing with technique, material and mounting of the casket, with iconography and style, with the history of the Gandersheim abbey and the runic inscription. For more recent literature see below. 2 In preparation of the international conference in 1999 the casket was thoroughly examined in noninvasive ways and was not removed from its mounts. Concerning the panels, a close parallel was found in an object labelled as walrus tusk in the Naturhistorisches Museum in Berlin, cf. Gandersheimer Runenkästchen (fn. 1, p. 23, Ill. 15, 16). In 2015, however, following the most recent publication of the casket by Carol Neuman de Vegvar, the object in the Naturhistorisches Museum in Berlin was re-examined. It turned out that it had been wrongly labelled as walrus tusk; it is now correctly labelled as whalebone. Cf. Carol Neuman de Vegvar, Hronæs Ban: “Exotism and Prestige in AngloSaxon Material Culture”, in: The Maritime World of the Anglo-Saxons, ed. Stacy S. Klein et al., Tempe, 2014, pp. 323–335, esp. 323, Ill. 14.3. Note: I wish to thank Gaby Waxenberger for several intense discussions about the casket during the last twenty years; and I am very thankful to her and Kerstin Kazzazi for their invitation to the Eichstätt conference, and to Kerstin Majewski for her patience while editing my paper. A special debt of gratitude is owed to Marjorie Trusted, Honorary Senior Research Fellow, Victoria and Albert Museum London, for her corrections in my English text. https://doi.org/10.1515/9783110796834-007

128

Regine Marth

Fig. 1: The Gandersheim Casket, Braunschweig, Herzog Anton Ulrich-Museum, inv. no. MA 58; © Herzog Anton Ulrich-Museum, Braunschweig.

Figs. 2 and 3: The Gandersheim Casket, front and back; © Herzog Anton Ulrich-Museum, Braunschweig.

The Mysterious Gandersheim Casket: Are There Any Hard Facts?

129

edly indicate a liturgical use of the casket, presumably not as a reliquary but as a container for the host.3 Such function for house-shaped shrines can be assumed from analogous pieces, but to my knowledge no casket or shrine has survived with a comparably complex iconography.4 The casket stands on a frame at the base inscribed in runic characters. This frame was not, however, nailed but glued onto the casket, a striking and unusual method of fixing. In 1957 the glue had dried out to the extent that the frame had become separated from the body of the casket. Before that, the frame had never aroused any suspicion concerning its age or its runic inscription (see below under section 2). The casket was donated to the museum in 1815. A note survives written by the museum’s then director Johann Ferdinand Friedrich Emperius (1759‒1822).5 Under the headline “Bereicherungen des Museums im Jahr 1815” we read for October 1815 under no. 3: Ein elfenbeinernes Kästchen mit Verzierungen, und metallner, mit Runen bezeichneter Einfassung. Ehemals war es in Gandersheim. Es enthält zwey vermeint. Reliquien, eines ein Stück vom Gewand der heil. Jungfrau, u. ein Stück von ihrem Kleide. – Abgeliefert von H Hassel.

One Johann Georg Heinrich Hassel (1770‒1829) is known, who worked as a writer for statistics and geography.6 His unpublished collected material and other documents are preserved in the Archives in Wolfenbüttel, but so far nothing has emerged which could shed light on this episode in the history of the casket. The abbey of Gandersheim is not mentioned in Emperius’s note. The phrase “ehemals war es in Gandersheim” presumably points to the fact that the Gandersheim abbey and convent had been secularised in 1810. Concerning Johann Georg Heinrich Hassel – if 3 See Gandersheimer Runenkästchen (fn. 1), introduction by Regine Marth and papers by Carol Neuman de Vegvar, Leslie Webster, Egon Wamers and Victor H. Elbern, all dealing with the iconography and its implications concerning the function of the casket. 4 See exh. cat. The Work of Angels. Masterpieces of Celtic Metalwork, 6th‒9th centuries AD, ed. Susan Youngs, London, 1989, cat. nos. 128‒132; exh. cat. Making of England. Anglo Saxon Art and Culture AD 600‒900, ed. Leslie Webster and Janet Backhouse, London 1991, cat. nos. 136, 137. See more recently Leslie Webster, Anglo-Saxon Art. A New History, London, 2012, pp. 106, 108, 140, Figs. 72, 100. 5 Herzog Anton Ulrich-Museum, Altregistratur H 79, not paginated. Cf. Regine Marth, “Liturgische Geräte aus Gandersheim im Herzog Anton Ulrich-Museum Braunschweig”, in: Der Gandersheimer Schatz im Vergleich, ed. Hedwig Röckelein et al., Regensburg, 2013, pp. 135–140, esp. 138–140, Fig. 1, Pl. VIII. 6 Cf. Braunschweigisches Biographisches Lexikon. 19. und 20. Jahrhundert, ed. Horst-Rüdiger Jarck and Günter Scheel, Hannover 1996, s. v. Hassel (Gerhard Schildt). Hassel was born in Wolfenbüttel 1770 and died in Weimar 1829. He worked for public administrations in Helmstedt, Weimar and Kassel; from 1809 to 1813 he was employed in Kassel by the ministry of home affairs of the Kingdom of Westphalia as a councillor for education and for churches. Until 1815 he was concerned with matters dealing with the abolition of this kingdom. After 1815 he was based in Weimar once again, where he was self-employed. His collected papers are kept in the Niedersächsisches Landesarchiv, Abteilung Wolfenbüttel (sign. LB 1225).

130

Regine Marth

he is the person mentioned in the note – it is known that between 1809 and 1815 he lived in Kassel where he worked for the ministry of home affairs in the Kingdom of Westphalia. In that capacity he was also responsible for church affairs.7 We might therefore speculate that he seized the casket after the secularisation process. Before moving from Kassel to Weimar in 1815, he wanted to be rid of it. The elliptical statement “abgeliefert von H Hassel” might point to the fact that Hassel was known to Emperius. But this remains unproven. The “relics” which are mentioned in the note by Emperius are no longer preserved in the casket, nothing definite is known about them since they were removed at an unknown time. Nevertheless, a small piece of Byzantine silk from the first half of the tenth century is kept in the museum from which traditionally it is said that it belonged to the casket.8 A larger fragment of the same fabric is preserved in the Gandersheim treasury,9 the original function of which is unknown. But the existence of the fabric itself must surely be considered a strong link between the casket and the abbey. In Gandersheim abbey a relic from the Virgin’s robes is still preserved,10 packed as a small package, undisturbed since medieval times. No details are therefore known about this relic, though it is tempting to think it might be a piece from the same fabric. With this chain of arguments the casket is fairly well connected with Gandersheim abbey, but when did the casket arrive there? Clearly, being an Anglo-Saxon object from around 800, it could not have been commissioned for the foundation of the Gandersheim convent in the year 852. The first connection between AngloSaxons and the convent might have occurred during the reign of the Anglo-Saxon king Æthelstan (ca. 894‒939) on the occasion of the wedding between his half-sister Eadgyth (910‒946) and the Saxon duke Otto the Great (912‒973), who was to become Holy Roman Emperor in 962. Not much is known about this wedding, neither the place where Eadgyth and her Anglo-Saxon entourage met the Ottonian court for the first time, nor the exact date, which must have been around 929/930. Where the festivities took place is also unknown. Possibly Gandersheim, which was a favourite liturgical centre for the Ottonians, played a part in this, and we could assume an Anglo-Saxon presence there on this occasion. And the Gandersheim Casket itself points to an Anglo-Saxon presence in Gandersheim at some time.

7 See also Allgemeine Deutsche Biographie 10, p. 760. 8 See Leonie von Wilckens, Die mittelalterlichen Textilien. Katalog der Sammlung, Herzog Anton Ulrich-Museum, Braunschweig, 1994, cat. no. 51. 9 See exh. cat. Krone und Schleier. Kunst aus mittelalterlichen Frauenklöstern, ed. Kunst- und Ausstellungshalle der Bundesrepublik Deutschland, Bonn, und dem Ruhrlandmuseum Essen, 2005, cat. no. 160 (Martin Hoernes). 10 Gandersheim, Stiftskirche, Inv. Nr. 377‒55, see Christian Popp, Der Schatz der Kanonissen. Heilige und Reliquien im Frauenstift Gandersheim, Regensburg, 2010, p. 152 (mentioned in a list).

The Mysterious Gandersheim Casket: Are There Any Hard Facts?

131

And there is more evidence in form of the Gandersheim Gospels, now preserved in the Kunstsammlungen der Veste Coburg.11 This is a Carolingian manuscript with a precious book cover; on the very last page is an owner’s entry reading “eadgifu regina aethelstan rex angulsaxonum et mercianorum”. This manuscript was therefore once in the possession of the queen Eadgifu (ca. 902‒955), who was married to King Charles III of France (879‒929), and of her brother, the aforementioned King Æthelstan. It must have been given to Gandersheim before the beginning of the twelfth century, when an inventory of the Gandersheim treasury was added,12 indicating that this gospel book played a special role in the convent. Everything considered it is likely that both, the manuscript and the casket, were given by Æthelstan as very precious – though perhaps not modern – gifts to his brother-in-law, Otto the Great, who passed them on to the Gandersheim convent. If the casket were unambigiously identified in the inventory recorded in the Gospels, its early provenance would be more secure, but this is not the case. However, two very general entries mention caskets: one refers to “undecim scrinia plena reliquiis sanctorum” and a second to “II scrinia I cristallinum, aliud auro et ebore ornatum” on the other. Clearly these statements are not specific enough to enable us to identify the casket without doubt.13

2 The Frame at the Base and the Runic Inscription As already mentioned above, the frame at the base (Figs. 4 and 5), which is not secured to the casket, bears a runic inscription. Here the texts of the long and the short sides of the casket are each repeated. In 2003, Gaby Waxenberger presented a new reading of the runic inscription, translating the long sides as ‘composes (or ‘cuts’) this in the sign of the cross (or ‘Christ’) here (or ‘this year’) Aeli in the monasterio (or ‘mynstre’) ?? (two non-decipherable signs)’ and the short sides as ‘May Christ save Aeli according to the divine law’.14 Following Waxenberger’s reading, it 11 See exhibition catalogue Otto der Große und das Römische Reich. Kaisertum von der Antike zum Mittelalter, ed. Matthias Puhle and Gabriele Köster, Regensburg, 2012, cat. no. V.1 (Michael Peter). 12 See Klaus-Gereon Beuckers, “Das älteste Gandersheimer Schatzverzeichnis und der Gandersheimer Kirchenschatz des 10./11. Jahrhunderts”, in: Gandersheim und Essen. Vergleichende Untersuchungen zu sächsischen Frauenstiften, ed. Martin Hoernes and Hedwig Röckelein, Essen, 2006, pp. 97‒129, esp. Fig. 1 and p. 129. 13 The entry where a casket made of gold and ivory is mentioned was regularly related to a Carolingian ivory casket, also in the Herzog Anton Ulrich-Museum, inv. no. MA 59, also thought to be from Gandersheim, on which traces of colour and gold are still visible. Only Klaus-Gereon Beuckers in 2006 (see fn. 12) thought that this entry referred to the Gandersheim Casket, assuming that the “auro” refers to the originally gilded metal fittings. This is, however, unconving for a number of reasons, see Marth (fn. 5), p. 140. 14 Gaby Waxenberger, “The Intriguing Inscription of the Gandersheim Runic Casket Revisited”, in: Bookmarks from the Past. Studies in Early English Language and Literature in Honour of Helmut

132

Regine Marth

Fig. 4: The Gandersheim Casket, bottom with inscription; © Herzog Anton Ulrich-Museum, Braunschweig.

Fig. 5: The Gandersheim Casket, bottom frame from the inside; © Herzog Anton Ulrich-Museum, Braunschweig.

Gneuss, eds. Lucia Kornexl and Ursula Lenker, Frankfurt u. a., 2003, pp. 143‒176. See also Gandersheimer Runenkästchen (fn. 1), the essays by Gaby Waxenberger, Elmar Seebold and Tineke Looijenga and Theo Vennemann, pp. 91‒120.

The Mysterious Gandersheim Casket: Are There Any Hard Facts?

133

has been agreed that the inscription is authentic and no longer a point for discussion. However, there are inconsistencies and questions concerning the frame itself. 1. Since this inscription is on the underside of the casket, it is not normally seen and is only visible if the casket is lifted up. 2. The casket itself shows no sign of any nails by which a frame at the base could have been attached originally. The nails which are visible retain only the bottom of the casket (see also below, section 3).15 So it seems that the casket did not originally have a frame at the base, but was perhaps put into some sort of supporting frame. Alternatively the frame may have been fixed with glue or putty right from the beginning. The frame was not necessary for the stability of the casket. 3. The frame at the base differs significantly from the other metal fittings. The inscription and the ornaments are incised, whereas the ornaments on the other fittings are cast. Furthermore, the metal is significantly thinner, only 1.5 mm, whereas the other fittings are 1.9 mm thick. 4. All this would seem to point decisively to a nineteenth-century addition to the casket, but this is not confirmed by the analysis of the material of the frame. The base frame does not differ significantly from the other metal fittings; especially the element zinc, a tell-tale element in nineteenth-century alloys, was revealed to be present in similar amounts in all six test areas, ranging in the metal fittings from 15.03 % to 19.68 % and in the base frame from 15.06 % to 18.08 %. The overall analysis proved to be typical for Anglo-Saxon alloys of the ninth century.16 The four parts of the base frame are soldered together, as are other parts of the metal fittings. From this, we can only conclude that the base frame was not made at the same time as the casket. Nevertheless, it seems to be Anglo-Saxon, made perhaps slightly later than the casket, copying presumably a runic inscription, which was attached to the casket. The frame with the inscription is mentioned in the director Emperius’s note, so it must have been made before 1815.

15 See Gandersheimer Runenkästchen (fn. 1), pp. 11‒13, 25‒27 (photographs and illustrations of the casket). 16 See Hans-Werner Pape, “Das Gandersheimer Runenkästchen – Technische Analyse, Material und Montage”, in: Gandersheimer Runenkästchen (fn. 1), pp. 19‒34, esp. p. 23.

134

Regine Marth

Figs. 6 and 7: The Gandersheim Casket, right and left sides; © Herzog Anton Ulrich-Museum, Braunschweig.

3 The Making of the Casket The casket consists of single panels of whalebone17 fixed together by nails, made of iron or brass, and putty. A metal frame, unnecessary for the construction, was added by using decorative round-headed brass nails. As already noted, this construction is different on the base frame. The carved panels of the front, the back (see Figs. 2 and 3) and all four sides of the roof show a strict division into twelve and six fields respectively, corresponding to the hinge fittings and the clasp. This correspondence continues on the roof where we see three fields at the back and two at the front. All of these fields are framed by a carved twisted rope-like border. The smaller side panels of the roof are also framed in this way. In each of the twelve fields of the front, on the two side fields of the back of the roof and on the small side panels of the roof one animal is depicted, the other fields show two animals each.

17 See fn. 2 above.

The Mysterious Gandersheim Casket: Are There Any Hard Facts?

135

The side panels (see Figs. 6 and 7), however, are different, and on close inspection it is clear that they must have come from another context. They show a centrally placed tree-like plant in which four animals are intertwined. The characteristic twisted rope-like border is missing. These two panels had to be reworked so that they could be used for this casket. This can be seen at the top edge of each where parts of the figurative carving had to be cut away. Furthermore, the corners of the metal fittings at the top of the container itself, directly below the roof, are not strictly vertical. They already correspond with the angle of the roof and are slightly bent inwards. This is unusual and indicates that these metal fittings were reworked at some point. When the base frame with the inscription is removed it becomes clear that no holes exist from a former original frame. The holes and nails there only correspond to the bottom panel. This is obvious because the nails are still visible, even when the casket sits within its frame, since that frame does not cover the nails (see Figs. 2 and 3). Although these inconsistencies should be noted, the casket shows a precise stability and mathematical structure based on the geometrical figures of the circle, the square and the equilateral triangle. We must assume that the casket was constructed with a clear, architectural design in mind.

4 Summary and Conclusion 1.

2.

3. 4.

5. 6. 7.

8.

The casket, with all probability, belonged to the ladies’ convent of Gandersheim, from where it was brought to the Herzog Anton Ulrich-Museum – at this time named Herzogliches Museum – in 1815. We do not know since when the casket was in the convent in Gandersheim. Very likely, it entered the convent in the tenth century, by contacts between the Anglo-Saxon and the Ottonian courts through the marriage between Eadgyth, an Anglo-Saxon princess, and Otto the Great. The casket cannot be definitely identified, though, in the inventory of the treasure of the Gandersheim abbey from the beginning of the twelfth century. By 1815 we know that the metal frame with the runic inscription was attached to the casket, since Emperius wrote about “metallner, mit Runen bezeichneter Einfassung”. The alloy of the metal base frame shows no significant differences from the alloy used for the other metal fittings. The casket did not originally have a riveted metal frame at its base. The casket has a clear compositional format and an elaborate design, pointing to its liturgical use as a container for the host. We can see this from analogies with other similar house-shaped shrines. The panels of the casket must have come from two different sources. The side panels had to be adapted and do not fit properly, and the metal fitting at the top of the container had to be reworked.

136 9.

Regine Marth

Nevertheless, the casket exhibits a balanced construction and a harmonious decorative programme.

There can be no doubt that the panels and the metal fittings come from England during the Anglo-Saxon period, and should be dated to around 800. Perhaps the two side panels were later alterations when the original side panels had to be replaced. Consequently, the metal fitting at the top had to be slightly altered. Or the two different sets of carvings were in fact produced in the same workshop and were simply combined in this way. Since a base frame was not originally fitted to the casket it must have been intended to be placed in a supportive frame. Very likely the casket came to Gandersheim in the tenth century. The inscription must have been attached to the casket at that time, but presumably not by means of a base frame. The appearance of the existing base frame looks very much like a fabrication of the nineteenth century, made before 1815, although somewhat confusingly the analysis of the metallic elements does not support this. The inscription itself, according to all runologists who dealt with it, is difficult to read, but is not a fake. The casket might have been used in Gandersheim not for the host but as a reliquary, and as such it might be mentioned among those “undecim scrinia...” in the inventory. Alternatively, it may have been used as a vessel for the host although in later times this would not have been deemed appropriate, since precious metals, gold or silver, were by then insisted on for such containers. All these somewhat contradictory observations seem to imply that the casket is a pastiche, assembled at different times. But on the contrary it is rightly regarded as a genuine and outstanding example of Anglo-Saxon art – which is indeed what it looks like.

Tim Pestell

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts East Anglia has long been of central importance in understanding the Anglo-Saxon period, with its rich archaeology and prominent position on the North Sea littoral. In recent decades the rise of metal-detection as a hobby has added a new dimension to our archaeological knowledge with a rising number of metal small-finds appearing for identification. A side benefit has been the discovery of several new finds of runological interest. Although the overall numbers remain small, they represent an important increase on the picture presented by Page (1995; 1999, 28) which drew largely upon finds made through excavation. In many respects, such discoveries continue to demonstrate the relative rarity of runic objects. Archaeologists in Norfolk, the most productive county in Britain for metal-detecting, recorded over 10,000 finds alone in 2019, of which 564 were of Anglo-Saxon date.1 These mind-boggling figures owe much to the relationship local archaeologists have enjoyed with metal-detectorists in the county since the 1970s: while many archaeologists elsewhere in the UK were attempting to have the machines banned, local archaeologists in both Norfolk and Suffolk began working with metal-detector users. With the large size of both counties – together some 9,169 km² (or 3,540 square miles), with predominantly agricultural and largely arable land, the fields are regularly turned by the plough and favourable for metal-detection. The result has been a huge number of archaeological finds made and recorded within the region in the last forty years, revealing the kingdom of East Anglia to have perhaps the highest density of runic inscriptions in England. More problematic is the effectively unstratified nature of such runic finds. Being retrieved from the plough-soil, any archaeological context has been lost, making dating of an inscription reliant upon an accurate dating of the object from its stylistic attributes. As we shall see, for those inscriptions made on otherwise plain sheets

1 The figures are no freak as Norfolk is habitually the county from which most finds are recorded by the Portable Antiquities Scheme (hereafter PAS). Thus in 2018 some 9,113 finds were recorded, of which 441 were Anglo-Saxon. Indeed, before the full adoption of recording objects only via the PAS rather than on Norfolk’s Historic Environment Record (HER), even more objects were recorded, for instance 17,752 in 2010–2011; 17,483 in 2011–2012; and 13,447 in 2012–2013, of which 950, 1273 and 818 respectively were Anglo-Saxon. My thanks go to my colleagues Erica Darch, Mary ChesterKadwell, and Garry Crace for providing these statistics. For their assistance during the writing of this paper I would like to thank my Norfolk colleagues Steven Ashley, Andrew Rogerson, Helen Geake and Heather Hamilton, and for various runic discussions over the years, John Hines, David Parsons, Gaby Waxenberger and Kerstin Kazzazi. Needless to say, none are responsible for those errors that remain. https://doi.org/10.1515/9783110796834-008

138

Tim Pestell

of lead (for instance), this is especially difficult. Instead, it is the wider site context, and frequently viewing the object as part of an assemblage of artefacts from the site, that have to be used to provide some sort of chronological setting for the runic objects rather than one informed by traditional archaeological survey techniques. This paper seeks to examine those runic-inscribed objects that have been found in East Anglia, both old and new, and to try and provide this contextualisation of their find-sites. In so doing, I hope to provide both a useful gazetteer of the latest runic finds from the region, their find-spots for runologists, and to tease out some significance for the appearance and use of runes in the landscape for the archaeologist. I shall first examine my area of reference, East Anglia, in its Anglo-Saxon sense of the kingdom of the East Angles. East Anglia is that rounded landmass jutting out into the North Sea, which provided one of the first landfalls for those travelling from the Germanic homelands. In time the area developed into an autonomous kingdom, in part at least aided by its distinctive topographical setting. In particular, the region is surrounded by sea to its north and east, while to the west the Fens provided a marshy, semi-permeable, frontier. Indeed, it was principally the south and south-west that provided the easiest overland route into the region, over the clayland belt. This weakness had long been recognised to judge from the various linear earthworks placed across this zone from as early as the Iron Age, which faced ‘out’ towards the rest of England (cf. Malim et al. 1997). Although we may place the historical East Anglia largely within the modern counties of Norfolk and Suffolk (which themselves represent a subsequent subdivision into the ‘North’ and ‘South Folk’), it is equally clear the kingdom’s borders were mutable, and ebbed and flowed with the tide of political fortune. While Bede records in the eighth century that Ely was “a district of about 600 hides in the kingdom” (HE iv, 19), by the tenth it had effectively come into Mercian hands. I shall therefore be reviewing runic finds from the modern counties of Norfolk and Suffolk, but with some latitude to consider one find apparently made in recent times in this more disputed march area of the Cambridgeshire Fens. Naturally, the kingdom also saw various fluctuations in settlement dynamics. Using Migration-period Anglo-Saxon cemeteries as an indicator of settlement, occupation was widespread, albeit apparently favouring the lighter more tractable soils that avoided the central clayland (cf. Williamson 1993; Warner 1996). This was only a preference and, by the eighth century, occupation appears to have intensified throughout the region, albeit with Ipswich in south-east Suffolk acting as the principal trading centre or emporium for the kingdom. This same period also appears to witness other centres of economic activity, especially along the western Fen edge, as we shall see. By the tenth century, Ipswich’s importance had waned, at which point Thetford, on the border between Norfolk and Suffolk, became the economic focus for the region. The tenth century also saw Bury St Edmunds’s rise in wealth, based on the minster there housing the relics of the East Anglian king, Eadmund, martyred by the Vikings in 869. By the eleventh century Norwich had in turn

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

139

Map 1: The kingdom of East Anglia.

eclipsed Thetford to become the richest town in the region, and following the Norman Conquest, was chosen as the place to which the East Anglian see was relocated. In a classic case of Norman twinning, the new cathedral was joined by a castle as Norwich became the seat of regional royal power. East Anglia has yielded runic inscriptions from throughout the Anglo-Saxon period, in a corpus that provides us with a well-populated distribution map (see Map 1) and a good assemblage of objects to consider. I shall attempt to do so in a chronological way, but I want to conclude by drawing out some wider possible implications that the find-spots present us with. We begin, therefore, in the Migration period with a number of early runic objects, most of which are well-known. Of these, perhaps the most celebrated is the roe-deer astragalus found within urn N59, one of 376 cremations and 60 inhumations from a cemetery in Caistor St Edmund (see Fig. 1a).2 The urn itself was not

2 The astragalus is now in Norwich Castle Museum and Art Gallery (NCM), accession number 1939.77.N59.

140

Tim Pestell

Fig. 1a: The Caistor St Edmund roe-deer astragalus; © Norwich Castle Museum and Art Gallery.

particularly distinguished except for the number of gaming pieces it contained, including 33 of plano-convex type and in excess of 35 astragali, some of sheep and some of roe deer, like that with the inscription. The urn in which the astragalus was found was dated by Myres to the second half of the fourth century, making it one of the earliest in the cemetery (Myres/Green 1973, 46). Needless to say, its date of deposition need not be contemporary with the urn’s manufacture, but we seem to have an inscription of ca. AD 425‒450 (as suggested by Bayesian Modelling: Hines 2019a, 58) reading raïhan, which Page suggested might be intended to mean ‘roedeer’ (1973, 114‒17; 1999, 179‒180; see also Bammesberger 2006, 178; Waxenberger 2020a, esp. 54). Myres also drew attention to the possibly runic markings found on a number of cremation urns from Caistor, such as R13 with its arrangement of lines “probably intended to suggest the ᛏ rune” (Myres/Green 1973, 61 and 66‒68). Such inscribed lines are very hard to claim a definite runic origin for and other cemeteries have likewise yielded such markings, for instance urns from the cemeteries at Lackford and Ingham in Suffolk (see Fig. 1b) (Page 1999, 92‒93; Myres 1977, 65‒67 and 356‒ 358, and Figs. 368 No. 1038 and 2987, and 369 No. 985).3 Indeed, in such a category 3 The Lackford urns are in Cambridge, Museum of Archaeology and Anthropology 50.101 and 49.36. The Ingham urn is held by the Society of Antiquaries of London.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

141

Fig. 1b: The Lackford 2987 and Ingham 1038 urns, after Myres (1977).

of possible or ‘pseudo’ runes one may cite the swastika brooch of sixth-century date from a cemetery excavated 1900‒1902 at Hunstanton.4 The only clear example of runic text applied to cremation urns within the kingdom is from Spong Hill in North Elmham, where three pots, 1224, 1564 and 2167, were decorated with a runic stamp (see Fig. 1c).5 The stamp was originally interpreted as a clumsy attempt to read tiy, perhaps for Tiw or Týr, but is now more widely accepted as using Spiegelrunen to double the reading alu or ‘ale’.6 Such a magical word, common in Scandinavian inscriptions, need not be surprising to find on a cremation urn associated with death and possibly the desire to protect the individuals buried in these urns. Indeed, while cautioning against magical interpretations to inscriptions being accepted too readily, Page has acknowledged that the actual act of adding the runic inscription on the Loveden Hill urn may have been the important element, explaining why it was apparently cut without care (Page 1964; 1999, 114; see also Waxenberger 2016, 354‒355). A similar explanation might lie behind the pseudo- or possibly runic inscription on the aforementioned urn from Lackford in Suffolk (Lethbridge 1951, 20; Page 1999, 89‒90). That cremation urns could have special associations seems clear from the appearance of the ‘wolf and ship’ motif from another urn, R9/10, from Caistor, or the enigmatic ‘Spong Man’ pot lid from Spong Hill (Myres/Green 1973, 118; Hills 2014). All remind us that there are quite probably more symbolic or meaningful elements in the decoration of other urns that are as yet unrecognised.

4 The cemetery, excavated by Prof. Mckenny Hughes, is still unpublished, but see Proc. Soc. Ant 18: 310‒321, and 19: 172–174; and Norfolk Archaeology 27: 222‒223. The brooch is NCM 1950.2.7 and the site is Norfolk HER 1142. 5 Now NCM 1994.192. 6 Cf. Hills (1974); Pieper (1987); Parsons (1994, 139‒143); Page (1999, 93).

142

Tim Pestell

Fig. 1c: The Spong Hill pot stamp, as seen on urn 1564; © Norwich Castle Museum and Art Gallery.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

143

Indeed, one might question why more urns do not seem to have such obvious symbolic devices on a container for the dead.7 Let us instead examine the location of both our runic-producing sites further. Spong Hill is one of the largest excavated Anglo-Saxon cemeteries in England, with over 2,400 cremations and an attached inhumation cemetery, dating from the early fifth to sixth centuries (see Hills, this volume). Given its size, it is perhaps unsurprising that it has examples of many different kinds of artefact or urn-type represented, and in so doing draws on elements found on the Continent in many areas, such as miniature combs and shears from south Jutland, stamped pots from the north German coast and brooch types found across north Germany and Denmark (Hills/Lucy 2013). However, its size also clearly indicates that it was a cemetery serving more than a single, local, settlement; rather, it would appear to have acted as a wider focal point in the landscape. What form this took is unclear. Williams (2002) has argued that some cremation cemeteries held special significance, acting as ‘central places’, and Spong Hill seems one such example. Williamson (1993, 65‒66) has shown that the cemetery served a wider territory, possibly corresponding to the eastern half of the Launditch Hundred in which it lies. The fact that the central watershed dividing Norfolk into eastward- and westward-draining areas occurs directly north of North Elmham parish is also seen as potentially important by Williamson, implying that it served, or drew upon, communities on both sides of this important natural feature. If Spong Hill did indeed serve as an early medieval central place for burial, with its associated implications of social and ritual significance, it is surely more than coincidence that North Elmham was also the parish in which Bishop Bedwin settled his bishopric following the division of the East Anglian see by Archbishop Theodore in 673 (HE iv, 5) (see also Whitelock 1972, 5). It may be complete coincidence that runic urns come from such an important cemetery, or simply that they had a greater chance of deriving from one with so many cremations. Yet, it is also of some interest to note that another unequivocally runicinscribed cremation urn comes from Loveden Hill in Lincolnshire (Page 1999, 114‒115 and 180‒181). Like Spong Hill, this prominent hilltop cemetery is of exceptionally large size with over 1,800 burials excavated and which thus served a wide area. Williams (2002, 355) has suggested Loveden Hill acted as a → central place, providing a focus for “a range of social, political and religious functions and gatherings”. Again, it appears to have maintained itself as a place of significance. Not only did the hill give its name to the local Wapentake or Hundred which met there, but Domesday Book reveals that the wider parish in which it was contained, Hough-on-the-Hill, was a soke centre and home to a probable Late Anglo-Saxon minster church, substantial remains of which still survive (GDB fol. 347v; Taylor/Taylor 1965, 320‒324).

7 One may note Richards’ investigation of possible symbolism on cremation urns (1992, 141‒142), and Price’s view that urns constitute “an extraordinary resource […] for the symbolic repertoire of early English mortuary behaviour” (Price 2010, xvi).

144

Tim Pestell

Map 2: Caistor St Edmund showing evidence for Anglo-Saxon cemeteries and settlement between the fifth and ninth centuries.

Let us return to Caistor St Edmund, a parish lying to the south-east of Norwich. It was of ancient importance, as the tribal capital Venta Icenorum of the local Iron Age Iceni tribe. This was subsequently developed as the local Roman administrative centre and civitas capital, walled probably in the early 270s and occupied until the early years of the fifth century (cf. Davies 2001). Like many similar Roman towns, there is no evidence for Early Anglo-Saxon occupation within the walls although only limited modern excavation that might detect this has been carried out. However, settlement in the immediate area is demonstrated by the presence of a number of Migration period cemeteries immediately surrounding the town (see Map 2). It was in one of these, 350 m to the east, that the runic astragalus cremation was found.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

145

Fig. 2a: The Harford Farm (= Caistor-by-Norwich) brooch; © Norwich Castle Museum and Art Gallery.

Fig. 2b: Close-up of the runic inscription on the Harford Farm brooch reverse; © Norwich Castle Museum and Art Gallery.

Despite the lack of evidence for early Anglo-Saxon settlement here, the presence of so many surrounding cemeteries implies the Roman town continued to act as some form of focus. This clearly continued into the seventh century when another new cemetery was founded, at the top of an overlooking hill to the north-west, reusing a Bronze Age barrow site (cf. Penn 2000). This so-called ‘Final Phase’ cemetery contained 46 inhumation graves, mostly unfurnished or with bodies accompanied by only an iron knife or buckle, but grave 11 contained a wealthy female buried in 690‒700 with the gold composite disc brooch known for its runic inscription luda:gibœtæsigilæ ‘Luda repaired the brooch’ (Hines 2000) (see Fig. 2).8 That there

8 Now NCM 1994.5.78.

146

Tim Pestell

was by now occupation focused around the town is demonstrable by the spread of contemporary Ipswich ware pottery and over 40 early silver pennies, or sceattas, to the west of the town, apparently indicating a local market. Two series B sceattas of 690‒710 were found within grave 18 in the cemetery, showing its contemporary use (Pestell 2011, 571‒573). It is of further interest to note the position of the parish church, set into the south-east corner of the Roman walled town, a position typical for a seventh-century, Conversion-period, foundation (Hoggett 2010, 64‒67).9 The implication is that a ruling family established the Harford Farm cemetery in a position overlooking the power centre they controlled, generating their wealth through a trading site outside the walls. This control continued into the Christian period and the church may even have had its origins in an early minster foundation. Indeed, this wealth and importance may have subsequently led the church to be granted to Bury St Edmunds abbey and was probably the reason for the church’s dedication to St Edmund and its gift to the monastery, even though an outlier to its other land-holdings. Such a wealthy, indeed perhaps aristocratic, centre provides a good explanation for why at least one of the objects with a runic inscription, the brooch, should be found here. It also suggests that Caistor may have maintained its position as a focal point within the landscape even through the less visible fifth and sixth centuries, before emerging again as an economic and political centre in the seventh century. If it is difficult to see the presence of such central places in especially the fifth or early sixth century, six more runic-inscribed objects, from Binham in north Norfolk, arguably provide us with a similar picture. The discovery of a gold B-type bracteate bearing a runic inscription waat or wææt, perhaps meaning ‘wet’ in the sense of ‘liquid’ or ‘drink’ (Behr/Pestell/Hines 2014, 49‒52), was made by metal-detecting in 2004 (see Fig. 3a). Subsequent detecting on the same field has yielded five more bracteates, two die-duplicates of the first B-bracteate. The other three are A-type bracteates, one of which is chopped and folded, another of which is the largest and, at 27.35g, heaviest bracteate yet discovered in England (see Fig. 3b).10 All three bear a die-identical central device of a left-facing bust with a snake-like animal in front of the face. Running around the upper edge of this is a poorly-struck inscription, unusually, using a combination of Roman and runic lettering. This possibly reads MASRA …þrua..þ which, if read correctly, is unintelligible (J. Hines pers. comm.).

9 Excavation within the churchyard in November 2020 has yielded an enamelled mount of seventhor eighth-century date, probably from a hanging bowl, perhaps strengthening the case for an early church here (G. Emery pers. comm.). 10 The bracteates are catalogued according to the Ikonographischer Katalog (IK) of bracteates: Axboe/Düwel/Hauck (1985‒1989), and Axboe/Behr/Düwel (2011). The Binham examples are A-type, IK 630, 1‒3, and B-type, IK 604, 1‒2. All have been acquired by NCM, accessions 2005.756; 2011.755; 2013.67.3 and .4, and 2017.61.1 and .2.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

147

Fig. 3a: The first B-type bracteate discovered at Binham, IK 604,1 (NCM 2005.756); © Norwich Castle Museum and Art Gallery.

Fig. 3b: Detail of the runic legend on the Binham bracteate, IK 604,1; © Norwich Castle Museum and Art Gallery.

The six bracteates clearly derive from a hoard. The field has now been well-detected but has produced nothing else of early Anglo-Saxon date except two bracelets or armbands, one of gold, the other heavily-gilt copper-alloy, and a silver equal-arm brooch, all clearly from the hoard (Behr/Pestell/Hines 2014). Several points are pertinent. First, the presence of the bracteates as a hoard is significant in showing that the objects were apparently treated in a similar way as in Scandinavia, as expressions of ritual activity rather than just jewellery. By extension, some of the isolated bracteate finds made elsewhere in England by metal-detecting might also potentially represent ritual activity in deliberate, single object, deposits rather than stray losses. More crucially, we may look at the wider context of the bracteate hoard’s burial in Binham. There are at present only eleven gold bracteates known from Norfolk (a fragment of a twelfth is uncertainly located), two of the other non-Binham finds coming from Brinton and Blakeney, thus forming a cluster within a 9 km (6 mile) radius. In addition, a bracteate die is known from Field Dalling, the parish neighbouring

148

Tim Pestell

Binham.11 This is not only a significant concentration of bracteates locally; the hoard itself represents the largest deposit of Anglo-Saxon gold between the mid-fifth century coin hoard from Patching, West Sussex, and the Crondall, Hampshire, and Sutton Hoo Mound One hoards (cf. White et al. 1999; Kent 1975; Sutherland 1948). Together, these bracteates suggest that in the fifth century we may already see a gateway community of the kind suggested by Hodges (1982, 50‒52), with a leading family in control of sufficient wealth to bury such a hoard of bracteates. Moreover, such a community clearly had the cultural connections with Scandinavia, and in particular Denmark or north Germany, to import bracteates or to have exemplars and the raw materials to enable their production locally: the runic B-bracteates are of a design known from a hoard probably from Schleswig-Holstein (known in the literature as ‘Hamburg’-B (IK 71)), although another similar die is represented by a bracteate from the cemetery of Derenburg, Saxony-Anhalt (IK 599; Müller 2002, 78‒ 79). The widespread appearance of bracteates with similar motifs that could not have been designed independently has been explained by Pesch (2007). She proposes the diverse geographical distribution of what she calls Formularfamilien through their production in elite residences. Here, such elites were able to utilise their ongoing contacts and shared ideologies with other elites – and their access to the material resources, craftsmen and knowledge of designs – to create objects with near-identical imagery across political and ethnic boundaries (Pesch 2007, 44‒49 and 353‒359; 2012, 674‒678). We should also note that it was likewise at such elite residences that knowledge of scripts and literacy would have been more likely. Bracteate hoards and clusters have also been associated with elite residences or → central places on the Continent, for example the cluster around Gudme in south-east Fyn, Denmark, and Uppåkra in Skåne, Sweden (see the various papers in Nielsen et al. 1994; Pesch 2011). Once again, it is surely no coincidence that Binham appears subsequently to have acted as a central place. One potential indication of this is Domesday Book’s record of Binham as both a large estate in its own right, and one to which Edgefield and Wells-next-the-Sea (not adjoining parishes) belonged. Equally suggestive is the foundation within the parish of a full Benedictine monastery at an early date, 1093, combined with the curious decision of the founder and community to appropriate an existing parish church for the priory. The latter ensured that the villagers continued to have full parochial use of the monastic church’s nave, an arrangement that strongly suggests the monastery was founded on an existing Anglo-Saxon minster church. The parochial of this could well have been coterminous with, or based upon, an early multiple-estate. Indeed, enough hints exist to suggest that the land area between the Stiffkey and Glaven rivers may have formed a single land unit in the Middle Anglo-Saxon period, before becoming split up later. It therefore seems possible that the concentration of bracteates in this

11 Norfolk HER 25251.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

149

Fig. 4a: The Undley bracteate; © The Trustees of the British Museum.

Fig. 4b: Line drawing of the Congham sword pommel by Jason Gibbons; © Norfolk Historic Environment Service.

same area indicates the presence of an early Anglo-Saxon elite here, and with it the possibility of some continuity of power or land tenure. What then of another gold-bracteate from the East Anglian kingdom, the famous runic-inscribed A-bracteate from Undley (IK 374) (see Fig. 4)?12 Although not part of a hoard like the B-bracteates from Binham, recent excavation has shown

12 Now held by the British Museum (BM) as 1984, 1101.1.

150

Tim Pestell

that the Undley bracteate is not isolated. A silver D-bracteate was excavated from a grave in a cemetery at nearby West Stow, while more recent excavations have found a further five in three graves from Eriswell in Lakenheath, three of silver (two dieidentical) and two die-identical of copper-alloy.13 Not only does this form another cluster, the fact that Undley’s runic inscription appears also to include the element medu ‘mead’ potentially places it within a continuum of other East Anglian runic texts, as we have already seen (Page 1999, 183‒5; Hines/Odenstedt 1987; Hills 1991; see also Waxenberger 2018; Hines 2019a, 31‒32 and also in this volume). Undley and the Lakenheath area more generally may yet lack good evidence for being the site of a central place, beyond the clustering of bracteates here, but this area of the Lark valley has produced some of the earliest evidence for Anglo-Saxon settlement; the complex of cemeteries in Lakenheath/Eriswell serves to emphasise this. I shall conclude this round-up of early Anglo-Saxon runic objects from East Anglia by looking at those from Congham and Eye. That from Congham is a silver sword pommel of Menghin’s Type Bifrons-Gilton which he dates to the late sixth or early seventh century (Menghin 1983, 312; TAR 1998‒1999, 33 No. 59) (Fig. 4b).14 Its simple inscription of a ᛏ t-rune, symbol of Tiw the god of war, is an apposite choice to be placed on a sword and is paralleled by the application of runes on other swords (Hawkes/Page 1967; Page 1999, 91‒92). Although the Congham pommel provides an example in East Anglia of a predominantly Kentish artefact type, we should not put too much stress on this, as an increasing number of silver and gold sword pommels and other fittings have emerged in recent years. More important is that Congham is a parish which in recent years has yielded artefactual evidence from metal-detecting suggesting it to have been a high-status settlement (Rogerson 2003, 115‒116; Davies 2011, 167‒197 and 310‒315). Even if a specifically runic-inscribed object could not be anticipated, it is certainly the type of site on which a high-status artefact such as a sword pommel is unsurprising. Eye’s runic object is similarly intriguing. It consists of a copper-alloy strip with three characters reading guþ or possibly gub (Caruth 2013, 3). Deriving from the excavation of an Anglo-Saxon settlement at Hartismere School in Eye, the site is notable for having revealed what appears to be a Continental-style longhouse, the first of its kind found in Britain. The site seems to have been occupied between the fifth and seventh centuries, principally in the fifth to sixth century, although not so densely as seen at West Stow. Over 200 small-finds were recovered, including a silver pendant and two gold objects, along with evidence for craft and industrial activity including iron smelting, non-ferrous metalworking, bone- and antler-working and textile production. Seen in these terms, the site cannot be labelled as high status, but it is certainly atypical, not least because prestige metalwork was also found. In conclusion, most of the runic objects described would appear to have

13 The West Stow bracteate is IK 565; those from Eriswell are IK 293, 633 and 634. 14 Now in Kings Lynn Museum, 2000.3.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

151

some correspondence to sites that could be considered in some sense above average and wealthy, or even as ‘central places’.

Middle Anglo-Saxon Runic Objects Moving on to discuss those objects of Middle Anglo-Saxon date, that is, of broadly mid-seventh to mid-ninth century date, we again encounter a mix of find-spots. Before considering them, it is worth noting quickly that this is also the period in which we see the emergence of coinage. As the coinage produced within the East Anglian kingdom cannot yet be localised to their mint-sites, and since multiplestrikings of the same design were made, this artefact type cannot be discussed as easily in landscape or site-based terms. However, it is perhaps important to note, as did Page (1999, 117‒129), that it is in East Anglia where we see a popularity in rune-use and coinage production generally, and many of the last uses of runes on coins in particular. To the examples that Page (1999) noted, a new runic type has now emerged, issued by the moneyer Tiluwald (see Fig. 5): a sceat of ‘eclectic’ type – that is, not conforming to any of the standard series of sceatta coins. Both known examples

Fig. 5: The Tiluwald sceat; © Norwich Castle Museum and Art Gallery.

152

Tim Pestell

Fig. 6a: The runic-inscribed tweezers from Heacham; © Norwich Castle Museum and Art Gallery.

have been found in Norfolk, suggesting that they are East Anglian in origin.15 The design features a derivative motif of a standard on the reverse, with the moneyer’s name on the obverse, appearing in a radial arrangement. Although similar to the layout of the proto-pennies issued by King Beonna of East Anglia, c. 749, Tiluwald’s coins are probably early in the secondary series of sceattas, that is, of about AD 710‒720.16 From these less-easily localised items, we move on to more secure Middle Anglo-Saxon runic finds and their sites. Two of the sites have little to be said about them, except that the runic objects from them are both tweezers. Those from Heacham are of a form dating to the eighth and early ninth centuries, having a pair of animal heads at the bow, the eyes inlaid with tiny blue glass chips, two of which still survive (see Fig. 6a).17 The field in which the runic tweezers were found (Norfolk HER 33630) has produced a wide range of objects, all revealed by metal-detection, but none that would suggest a site of particular importance or interest. They include a Roman brooch, a Late Anglo-Saxon finger-ring, another ring of either Early or Middle Anglo-Saxon date and an object of presumed Anglo-Saxon date but unknown type. In fact, the vast majority of the objects from the field are unrelated, being medieval or post-medieval in date. In contrast, Heacham itself is actually suggestive of having been some form of estate centre, at least by the eleventh century. Domesday Book of 1086 records the parish in a single entry, the manor being held by William de Warenne, a Norman baron who had fought at Hastings alongside Duke William. At the conquest the

15 Only one has a certain provenance, being found in East Walton (HER 34888 and EMC 2013.0046); it is now NCM 2013.145. The other was recorded by the Fitzwilliam Museum, EMC 2009.0366, described only as ‘Norwich’. 16 I am grateful to Dr Martin Allen and Dr Rory Naismith for their opinions on this new coin type. 17 NCM 2006.533.2.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

153

manor had been held by one Toki, and a smaller part by Alnoth. Most important for us, there were 35 freemen attached here (LDB fol. 163). These were individuals who had fewer feudal obligations to the lord of the manor and larger numbers of them are usually found at estate centres. This fact, along with the -ham place name ending, has been suggested by Tom Williamson as possibly indicating archaic estate centres. Archaeologically, Heacham is unexceptional, the most notable other find from the parish being a purse hoard of four Middle Anglo-Saxon styca coins, the debased, mostly copper-alloy, currency deriving from Northumbria and only intermittently found in East Anglia. On present evidence, the tweezers would therefore seem to be a stray find, not readily identifiable with any specific Middle AngloSaxon site or centre and certainly not found with any other Middle Anglo-Saxon finds suggesting above-average wealth. Much the same may be said for the find-spot of the second pair of runic tweezers, Baconsthorpe (HER 35206; see Fig. 6b). The tweezers were found in 2009 through metal-detecting during an archaeological excavation. In fact, the archaeology revealed related mostly to the Roman period, with ditches, pits, a kiln, and postholes and sill-beam features indicating two buildings. There was additionally some post-medieval occupation, but nothing of Anglo-Saxon date. While the tweezers themselves are of interest, given their suggested highly literate reading of rēdæ sē þe cuinne ‘read whoso may’ and Bēaw þās rūnæ āwrāt ‘Beaw inscribed these runes’ (cf. Hines 2011), the same cannot be said of Baconsthorpe parish more generally. Largely devoid of evidence of Anglo-Saxon date, the tweezers appear to be a stray loss that cannot even be connected with an early medieval settlement. This is perhaps most surprising given the nature of the inscription which might otherwise be expected at a minster or high-status settlement. This contrasts with the third pair of runic-inscribed tweezers found in East Anglia, from the excavations at Staunch Meadow, Brandon (Carr et al. 1988; Tester et al. 2014). Reading the masculine personal name aldred ‘Aldred’, its silver and niello manufacture immediately sets it beyond normal contemporary tweezers as an expensive commission and the serifed runes are neatly cut, as though formed by someone used to Roman script (Parsons 1991, 9; Webster/ Backhouse 1991, 85, Cat. No. 660). Brandon has also yielded two further runic objects, a silver-gilt pin with the first sixteen characters of the futhorc, and an antler tine crudely incised with wohswildumdeo or wohswildumde[..]}an, possibly a riddle saying something like ‘[I] grew on a wild beast’ (Parsons 1991, 9‒10; Page 2014, 260‒263). Most important, these three runic items derived from one site, which represents an exceptional number and is illustrative of its special nature. The excavations at Staunch Meadow uncovered about 1.2 ha of a sand island beside the river Ouse, comprising a high-status settlement dating almost exclusively to the Middle Anglo-Saxon period of AD 600‒900. A religious element is demonstrated by the remains of a timber church and cemetery, while wider settlement is shown by 35 buildings as well as evidence for craft and industry. Interpreting what

154

Tim Pestell

Fig. 6b: The Baconsthorpe tweezers; © Norwich Castle Museum and Art Gallery.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

155

Fig. 7: The Blythburgh tablet, back; © The Trustees of the British Museum.

type of a site Brandon was is contested, with views including a monastery and a high-status secular settlement. A wider culture of literacy at the site is evinced by the stunning gold plaque depicting a zoo-anthropomorphic symbol of St John with the inscription SCS EVANGELISTA IOHANNIS, three styli and glass and antler inkwells (Tester et al. 2014). Equally interesting is Page’s (2014, 261) observation that the futhorc inscription on the pin has in the twelfth place a distinctive variant of the rune j (ᚼ), usually appearing in manuscript versions. A similarly high-status, and probably ecclesiastical, origin lies behind the whalebone writing tablet from Blythburgh in Suffolk (see Fig. 7).18 The tablet is not inscribed with runes in a normal inscription (although three are on the raised surface); they were instead cut through the original wax surface of the tablet and into the recessed back. That text which survives does not appear to make any sense, but remains as an unintentional survival of what had been written into the wax. Of significance is that two sequences of runes, -uat and sunt, suggest that the writer was attempting Latin verbal forms in runes (Webster/Backhouse 1991, Cat. No. 65; Page 1999, 217). As a site, Blythburgh is of great interest, having good evidence for being both a royal vill and the site of a minster community. At the Conquest it was both an extensive royal demesne and gave its name to the surrounding Blything Hundred (Warner 1996, 120‒121). The church was wealthy with two other churches

18 BM 1902, 0315.1.

156

Tim Pestell

attached to it, and an endowment of two carucates of land attached (a nominal 240 acres or 97 ha). Moreover, the twelfth-century Ely document Liber Eliensis claimed the body of King Anna of East Anglia was still being venerated there in the twelfth century (Liber Eliensis i, 7). As Anna was killed in 654 by King Penda of Mercia at the Battle of Bulcamp – located within Blythburgh parish – the implication is that a minster at Blythburgh had become an established cult centre from the mid-seventh century. That this was probably also used as a royal family mausoleum is suggested by the remains of his (otherwise unknown) son Jurminus also being culted here before being translated to Bury St Edmunds Abbey by at least 1095 (Pestell 2004, 91‒92 and fn. 147). If Brandon and Blythburgh are sites very clearly in the upper echelons of AngloSaxon settlement in East Anglia, another runic-inscribed object seems to derive from a settlement associated with the age of the great estates. Discovered by metaldetection in Sedgeford (HER 31814) in February 2017, the find appears to be a handle some 89 mm long and 11 mm wide, from some form of implement, perhaps a fork or spoon (Rogerson/Ashley 2018, 118 and Fig. 7 No. 31; Hines 2019b). This strip bears the runic inscription of a name, Byrnferþ in West Saxon, Biornferþ if Northumbrian or Beornferþ if Mercian (Waxenberger 2017, 638), but the second rune is unparalleled and neither the inscription nor the object is able to help provide a date any closer than between the eighth to tenth century. The find was made within a parish that has been the subject of intensive excavation over the last twenty-five years by the Sedgeford Archaeological and Historical Research Project, which aims to investigate the origins and evolution of settlement within the village. A cemetery containing 291 individuals but representing only a fraction of the original burial population and dating from ca. 650–825, has been excavated at the bottom of the river Heacham valley, with evidence for buildings higher up on the south slope. These would appear to date to the eighth to eleventh centuries, while more recently a series of malting kilns of similar date has been located nearby.19 The discovery of the runic object in the field immediately south of the settlement and adjacent to the malting kilns undoubtedly links it with the occupation. The question is what type of settlement does Sedgeford represent? While the number of finds suggests intense occupation, according with the size of the cemetery, the range and quality of material culture does not seem to indicate a wealthy site of the type seen at Brandon. Indeed, the handle is perhaps one of the most important small finds from the site. The recent evidence for malting suggests that Sedgeford may have been one of the outlying production centres that were crucial to the machinery of estate exploitation. Quite which centre it originally belonged to is less clear, but nearby possibilities include Heacham, already noted, or the royal estates

19 Much of the detail remains unpublished, although an overview is SHARP 2014. There are interim site reports available through the SHARP website: ‹www.sharp.org.uk/publications›, last accessed 30 November 2021.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

157

at Southmere in Docking and Snettisham (Williamson 1993, 100–102). That there was literate ability at Sedgeford is further indicated by the presence of a copperalloy stylus, recovered in the current excavations.

Five Lead Sheet Inscriptions Five runic inscriptions made on lead sheets place us in more uncertain territory, not least as their dating is problematic. The first of these is a lead sheet with an extensive runic inscription found at Shropham (HER 17722) (see Figs. 8).20 The find-spot has produced a number of items ranging from Roman to medieval pottery and multiperiod metalwork. This includes much of Anglo-Saxon date including both Ipswich and Thetford ware pottery, the two type-ceramics in East Anglia for sites of Middle and Late Anglo-Saxon date (i.e., AD 700‒1100). The metalwork includes Middle

Fig. 8a: The Shropham runic sheet, obverse; © Norwich Castle Museum and Art Gallery.

20 NCM 2004.37.

158

Tim Pestell

Fig. 8b: The Shropham runic sheet, reverse; © Norwich Castle Museum and Art Gallery.

Anglo-Saxon pins, strap-ends, hooked tags, an ansate brooch, a ninth-century strapend, and Late Anglo-Saxon disc brooches, hooked tags, strap-ends, a buckle and a trefoil brooch. Among the coins are an eighth-century penny of Offa of Mercia, a ninth-century Northumbrian styca, an early tenth-century St Edmund memorial penny and an eleventh-century York-mint penny of Cnut. The field to the south has additionally produced two ninth-century hooked tags and an eleventh-century penny of Edward the Confessor. On the face of it, this assemblage may seem interesting yet in fact it is not especially significant in Norfolk; there are now a great number of sites that have been similarly metal-detected and which have yielded similar numbers and types of finds. What it does demonstrate is that there was long-lived occupation on this site, earlier evidence from the parish including finds of Migration Period date, including strap fittings, a girdle hanger, bird brooch and wrist clasp. The numbers

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

159

are small but may suggest a small cemetery which is being ploughed out.21 In conclusion the material being recovered from the site is of a wide range and can add little to help date the most likely period for the plaque’s inscription or use. Likewise, despite the longevity of occupation evidence, there is little to mark out Shropham as particularly unusual or notable. The tenurial evidence, albeit later, does not contradict this. While the parish gave its name to the Hundred, at Domesday Shropham was held between various manors, with jurisdiction or soke for its occupants resting with the King in nearby Buckenham (LDB: Fols. 152; 183; 239v). It therefore seems to have been – or at least to have become – a minor land unit with little specifically to recommend it as having the kind of settlement like Brandon or Blythburgh, in which literacy was being practiced, to have produced an inscribed runic text. The function of the lead sheet or plaque is still of interest. Originally rectangular, the sheet has had both ends folded back behind it to produce a nearly-square sheet some 36 × 44 mm. The bottom is now ragged and a fourth line of text appears to have been lost. Only one of the ends folded behind is now visible, although partly obscured as two corners are bent over the text. Sadly for such an extensive and promising text, the inscription has so far defied translation, although sequences suggest personal names including the soul of an individual.22 This suggests an ecclesiastical background and it may be that the hole in one bent-over corner on the reverse, and another hole in the centre at the top of the front indicate where the sheet had formerly been nailed to something. By analogy with similar lead sheets, a coffin-plate is one possibility and it may be suggestive that the find-site and settlement are close to the present-day parish church. Unfortunately, the date of both the sheet and the foundation of the church are unknown, although the latter has nothing to suggest an Anglo-Saxon date from its fabric. The proximity of Anglo-Saxon occupation to the church as revealed by the metaldetection might suggest that it had its origins in the pre-Conquest period, but in itself, this would not be surprising, as many English parish churches have their origins in the tenth and eleventh centuries (cf. Blair 1988; 2005). The issue is thus whether a lead sheet like this is likely to have been used in what would appear to have been a ‘normal’ parish church – assuming it does derive from Shropham’s

21 Interestingly, this site also yielded two runic sceattas, one Kentish of 670‒700, the other East Anglian of 710‒750. 22 For the interpretation and translation of the text see Waxenberger (2016, 364‒369). Hines (2019a, 47) comes to a similar interpretation and translation: 1 […] ārǣrde þis bǣcn Gode tō lofe 2 [for –]des saule ond for hi Ōs[–] … 3 [–]e for Alhmundes saule alra … 4 f … eafas[–]ea … ‘[…] raised this sign in praise of God, […] for […]d’s soul and for her Os[…], […] for Alhmund’s soul of all, […]’.

160

Tim Pestell

church and that the latter was a pre-Conquest foundation. In so far as we can see, there is nothing to mark out Shropham as a particularly important place in the Anglo-Saxon period. The second sheet’s findspot bears many similarities to Shropham. It was found in Toftrees (HER 60948) in September 2015 and recorded by the PAS as NMS-63179C. This small rectangular sheet measures 24 × 29 mm and has a small hole in one corner. The inscription is of great interest in appearing to be a charm, reading dædisdwerg, “impeccable Old English for ‘Dead is the/a dwarf’” (Hines 2019a). As a plain lead sheet, the object is not stylistically datable, and nor are the runes and Old English, although an eighth- or ninth-century date seems most favourable. The reading of ‘dwarf’, or alternatively ‘fever’, suggests the sheet is a charm used in a quasi-magical way that reflects the perception of dwarves as potentially harmful sprites and seen in the charm wiþ dweorh, ‘against a dwarf’, in the late tenth- or early-eleventh century compilation Lacnunga. Like Shropham, the site has evidence of multi-period occupation, yielding both pottery and metalwork of Roman to post-medieval periods and, most relevant, including a number of Anglo-Saxon date. While limited, these include two hooked tags and a strap-end of Middle to Late Anglo-Saxon date, a ninth- to tenth-century ansate brooch of ‘East Harling’ type, and an eighth- to ninth-century gilt and chipcarved object with the arm of a circular cross. Along with a number of Thetford ware sherds, this small assemblage would appear to focus upon ca. AD 800–900. The site, immediately east of the parish church, is at one level intriguing in relation to the runic sheet and the lack of contradiction between the use of charms and the role of the Church; at another, the occupation adjacent to the church accords with the increasing evidence for the establishment of parochial churches accompanying settlements by this date. Toftrees always seems to have been a small parish and was almost certainly a dependent settlement of a larger nearby estate, perhaps the royal manor of Fakenham, or a possible former estate at Raynham to the west. As such, Toftrees does not present itself as an obvious location for the discovery of a literate inscription. Our third lead sheet has an even more uncertain background and probably takes us to the borders of the East Anglian kingdom. Now in Norwich Castle Museum’s collections, the ‘near March’ sheet has a chequered background, with a findspot that is now lost. The object is some 30 × 40 mm and although damaged, has text on both sides, each probably written by a different hand (see Figs. 9). The sheet has been published by Hines (2019a), the fully legible Side A of which he reads and translates as: 1 2 3 4

scs̅)matt)eus scs̅marcusscslucus sc̅sgiohonnislib)era )m)ea}malo

‘S[an]c[tu]s Mattheus, S[an]c[tu]s Marcus, S[an]c[tu]s Lucas, S[an]c[tu]s Iohannes, libera me a malo.’

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

Fig. 9a: The ‘March’ lead sheet, obverse; © Norwich Castle Museum and Art Gallery.

Fig. 9b: The ‘March’ lead sheet, reverse; © Norwich Castle Museum and Art Gallery.

161

162

Tim Pestell

Hines reads and interprets Side B as: 1 2 3

omnigrminimscrumsuworumdo minobeuramln

Omni gr[atia] minim[o] s[an]c[to]rum suorum domino Beurn am[e]n.

Frustratingly, the sheet came to our attention when listed for sale on eBay. All that could be determined subsequently from the seller was that the find had been bought by him at a metal-detecting ‘rally’ among the contents of a bag of ‘scrap’ finds made by their original finder. Some support for this was provided by the fact the runic lead sheet was being sold along with a fragment of a lead medieval pilgrim badge. The finds were possibly found in the area of March.23 March is in Cambridgeshire, just over the western border of Norfolk, although whether the sheet was indeed found in this area is anybody’s guess. I like to think it was, not least given the number of important Benedictine abbeys founded nearby in the tenth century, including Ely, Ramsey, Thorney, Crowland and Peterborough. Thorney, of course, is also the house that produced a manuscript illustrating runes and their names in 1110–1111, now preserved in St John’s College, Oxford, as MS 17, fol. 5v (Ker 1964, 189; Parsons 1998). More pertinently, the Fens were an area of widespread ecclesiastical ownership and influence from the Conversion period, which would provide one explanation for the literate ability to write such a piece, and one which appears to be of religious interest. Sadly, on the available evidence we can only see the ‘March’ plate as an intriguing stray-find. The gap in our knowledge with the ‘March’ inscription contrasts with another lead sheet that helps to illustrate how an accurate find-spot can help to contextualise an object. In this case, the sheet bears not Anglo-Saxon but Scandinavian runes. This is the lead sheet found in January 2003 in a molehill at St Benet’s Abbey, Horning (HER 5199) in the Broadland of east Norfolk (see Figs. 10).24 The sheet bears around 45 runes: they were incised with a sharp point and arranged in five rows set between framing lines. Except at the edges, where there has been some slight damage, the characters are mostly clear to see. The first four rows give an impression of competent literacy, with neatly made runes between 9 and 12 mm high.

23 The object is now NCM 2010.112. Our knowledge is perhaps best summed up in the eBay seller’s own words, from his email correspondence with me (9/9/2009): “Dear Dr Tim, I wish I could help you, but I found the Roman piece [sic] but the Runic piece was bought with a number of detecting finds at a metal detecting rally in Cambridgeshire, approx a year or so ago, it was in fact in the chaps scrap bag, I did not notice at first. Unfortunately all I know is that the chap said that he lived in a village near March and that he did his detecting around that area. I wish I could help you more, but that is all I can tell you. Sorry I could be not be of much help. Regards”. 24 Now NCM 2003.54.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

163

Fig. 10a: The St Benet’s Abbey lead sheet, obverse; © Norwich Castle Museum and Art Gallery.

Fig. 10b: The St Benet’s Abbey lead sheet, reverse; © Norwich Castle Museum and Art Gallery.

164

Tim Pestell

“A transliteration of the runes on the object reveals a text that appears to be gibberish,” according to Hines (2019a, 51).25 The runes are mostly forms of the ‘longbranch’ type of the Scandinavian Younger fuþark. The only exceptions are the first rune in line 3, which looks more like an Old English w-rune than the þ-rune as it otherwise appears on this object. In line 5, a single vertical line of only half height aligned with the top of the line of script could be identified as a typical Scandinavian ‘short-twig’ Younger fuþark s. In light of the speculative attribution of a Scandinavian long-twig k to the Scotterthorpe plaque it is of interest to note that this inscription includes three clear examples of the rune-form ᚴ. This rune was certainly known in England (cf. Hines 2017). From an archaeological perspective, the find-spot is of great interest. There are a number of runic parallels for the sheet in mainland Scandinavia which may help to provide a context for the object. These are generally assigned to the post-Viking period, and while a few possibly date from the later eleventh century, most seem to be of the twelfth century or after. A date as late as the twelfth century for the St Benet’s piece may be possible, though the absence of dotted runes might suggest that it should not be placed much beyond the last quarter of the eleventh. Lead sheets, crosses and amulets frequently contain combinations of biblical and incantatory Latin, recognisable medieval charm-language (abracadabra, agla, etc.), and apparent gibberish. It is therefore tempting to associate the St Benet’s find with these inscriptions, and to draw attention to the fact that the St Benet’s sheet was found neatly and tightly folded. This was probably done deliberately, another possible element of any amuletic function. An eleventh-century dating would also fit best for the local context to the piece. St Benet’s Abbey was founded around 1020, Cnut being one of the principal benefactors and celebrated as the founder (cf. Licence 2006). There are difficulties in dating the early history of the abbey, largely because few original charters survive from the house but, as Domesday Book makes clear, by 1066 the abbey had secured a good local endowment of land. Among its patrons were a number with Danish connections, for instance Grimolf the Dane, who gave lands at Caister by Yarmouth (Caister-on-Sea) (Oxenedes, 291). Given that Norfolk appears to have been well settled by Scandinavians in the wake of the First Viking Age, his moniker suggests he may have been a later incomer and perhaps one of Cnut’s men, following the patronage of his lord. Likewise the abbot, Ælfwold, who had been placed in charge of the defence of the coast by Harold Godwineson, fled to Denmark in the wake of Hastings along with one of his men, Ringolf of Oby (Oxenedes, 293; Stenton 1922,

25 1 2 3 4 5

ikkofrukriþ okinifuitr wartrsom irsornrsmþu – rshr

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

165

227). Another of St Benet’s patrons, Eadric ‘the Steersman’, likewise fled, Domesday Book recording him as being an outlaw in Denmark (LDB fol. 200). Given these Danish links with the abbey, the Scandinavian parallels to the runic sheet are of interest. If this was indeed an amuletic charm, it is quite possible that its creation was influenced by examples seen by those people associated with St Benet’s while they were in Denmark itself. Perhaps equally pertinent, David Parsons (pers. comm.) has seen the sheet as an English product, in favouring the use of English wynn as a rune-form within the inscription. The find-spot, a molehill next to the High Altar, places its deposition or loss at the spiritual heart of the monastery, a place reserved for the burial of the most honoured members of the community. Given the use of inscribed lead sheets as items to accompany the dead at this time, it is not impossible that the inspiration for this lead sheet, if not the actual object itself, was brought back and used by someone like Abbot Ælfwold or Grimolf themselves. Finally, an inscription on another lead sheet was found in July 2012 in excavations at Stoke Quay on the Ipswich waterfront (Hilts 2013; Brown/Shelley 2014). The sheet would be some 80 × 30 mm if straightened out, but has been cut down and folded to modify it into a plate to an iron belt buckle. Secured using iron rivets, the sheet also has a series of perforations around the strip which appear to be original to the sheet’s primary use, and therefore to have been used to fix it to something. The belt buckle would appear to be of eighth-century date stylistically, of Marzinzik’s Type II 24b-I (Marzinzik 2003, 51 and 220‒223, Plates 137‒143) and was additionally found in a domestic refuse pit with Ipswich ware pottery (dating to 720‒850: cf. Blinkhorn 2012). The inscription’s partial text is in Old English and appears to read (o)ueb…for, seeing the ‘left-hand’ piece as the start of a text running left to right, with the ‘right-hand’ piece appearing after the strip had been bent through the belt buckle loop and continuing (based on Waxenberger 2016; 2020b). The meaning of this text is still unclear since only the sequence for makes sense.26 The appearance on a belt buckle in this way is wholly unconventional and, apparently, unparalleled, but the use of runic lettering on a sheet with holes provides a striking similarity with the St Benet’s Abbey, Toftrees and especially Shropham examples, already noted. While the former function of such objects, such as perhaps being attached to caskets, is still unclear, they do suggest an epigraphic tradition being used in a far more visible manner than many of the better-known inscriptions appearing on the reverse of objects such as dress-pins or on tweezers. I want to end this survey by rounding up those last few inscriptions known from the region, and then offering some concluding thoughts. The first of these takes us back to Ipswich. Another recent find, it was made in excavations at Greyfri-

26 for can be interpreted as an OE noun fōr ‘action of going, journey, trip, voyage’; as a verb OE fōr (3rd pers. sg. pret.) ‘went, travelled’; as a preposition or conjunction OE for ‘before, in front of’ and as an intensive prefix OE for- ‘very, extremely, utterly x’; for more details see Waxenberger (2016, 357).

166

Tim Pestell

ars Road, within the fill of a shallow pit. The inscription was found on a rectangular whale bone mount, some 49 × 20 mm in size, associated with Late Anglo-Saxon pottery that appears to date from 870‒950. The inscription, which is fragmentary, is difficult to decipher. Indeed, it is not even clear how many runes are represented. However, those visible seem to read …]þs or …]þis. The date and function of the object is unclear. Waxenberger (2005, 162) points out that the s-allograph type was used between the seventh and ninth/tenth centuries, which could fit the ceramic dating evidence. The object itself is perforated on the upper (inscribed) surface by two blind holes which “are likely to be registration marks, designed to fit with another piece of whale bone, suggesting that the object was originally part of a twopiece whale bone mould” (Riddler et al., forthc.). The object is, however, stylistically undatable and its exact function (mould or not) is unclear.27 Despite these limitations, the presence of a second runic inscription from Ipswich is clearly interesting and recalls the appearance of others from similar emporia, such as that of 750‒850 inscribed tatberht, Tātberht, found on a sheep’s vertebra from Lundenwic (cf. Page 2004), or indeed the Brandon antler tine. While wics like Ipswich are clearly not the same as our smaller ‘productive’ sites like Brandon already discussed, it would be unsurprising to see some access to, and use of, literacy in such places – if nothing else to assist in the administration and exaction of market tolls and dues for the authorities running such centres. It is not unlikely to be linked to the presence of an iron stylus at the belt of an individual from Ipswich’s Buttermarket cemetery (Pestell 2009a, 246‒247), and of course such urban centres would also have seen a concentration of people, some of whom would have been likely to have literate ability. The presence of two runic inscriptions from Ipswich should certainly not occasion undue surprise. Next in this round-up is an enigmatic ‘brassy’ (probably copper-alloy) disc that was found in the river Yare at Keswick/Eaton near Norwich (HER 31652) some time around 1995 (see Fig. 11). Only seen very briefly for recording, the Old English runes have an unclear meaning, appearing to read tlimsudn (cf. Hines 1997). It was sold soon after its discovery and its whereabouts, let alone function or date, are still unknown, but the object and its inscription both seem to suggest a Middle AngloSaxon date of between the seventh and ninth centuries. The find-spot was nondescript and has nothing to provide any hint at dating, nor to suggest what might have caused it to be there. However, it is perhaps worthy of comment that the findspot is within a couple of miles of Caistor St Edmund. Apparently pseudo-runic, another disc-shaped object survives as only a fragment found in metal-detection at Roughton (HER 61282), some time between 2015– 2016.28 Of heavily gilt copper-alloy, the fragment seems to derive from a mount or

27 I am very grateful to Ian Riddler for providing me with details of this find prior its publication by him. 28 Recorded as PAS NMS-DC6C55 and now NCM 2016.468.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

167

Fig. 11: The Keswick runic disc; © Norfolk Historic Environment Service.

more probably a disc-headed pin originally about 30 mm in diameter. Such an interpretation seems to take good parallels in other such pin-heads of similar size and with runic inscriptions, for example that recovered from Surrey Docks on the London foreshore, 27 mm in diameter; one from Bardney in Lincolnshire, 28 mm diameter; and Cumwhitton in Cumbria, ca. 28 mm.29 The Roughton fragment appears to be a stray find as other items from the findspot are of Roman and medieval date, and so it reflects the wide circulation of Anglo-Saxon material in the East Anglian landscape. Another class of object that seems to be emerging as one on which runes might be inscribed is spindle whorls, used in the processing of wool. Well known as an artefact type, they have a conservative design making them difficult to date. Many are known to be decorated with patterned sides or upper edges, some of which appear pseudo-runic, but a clearly inscribed example was found in Dagworth near Haughley in 2015. 30 Cast from lead in a bi-concave shape and 29 mm in diameter, it bears five incised runes on both top and bottom of Anglo-Saxon forms, although datable only to perhaps the eighth or ninth century.31 Frustratingly, the runes have

29 Recorded as PAS LON-41A95B; DUR-79B856 and LANCUM-EEFFFB. 30 Other examples of spindle-whorls with runic inscriptions may be viewed on the Portable Antiquities Scheme database: www.finds.org.uk. Perhaps the best example is that from Saltfleetby St Clements, Lincolnshire, inscribed in Scandinavian runes probably of early eleventh-century date: LIN-D92A22. For the more clearly imitative or pseudo-runic see examples from Wishaw, Warwickshire (WMID-87ED62), and Tiverton, Cheshire (LVPL-0630E0). 31 Recorded as PAS SF-D8FC51 the find lies in the modern civil parish of Haughley.

168

Tim Pestell

Fig. 12: The ‘Reedham’ runic spindle whorl; photo: P. Murawski.

an uncertain reading despite the graphs being relatively straightforward to read.32 As we have seen on several occasions already, metal-detecting has revealed the findspot to have multi-period occupation stretching from the Iron Age to the postmedieval, but there is good evidence here for an interesting Anglo-Saxon past. The finds suggest occupation here from the fifth century through to the Conquest, with a slight emphasis on eighth- and ninth-century finds, including faceted pins, ninthcentury strap-ends and a ninth-century ‘cogwheel’ type brooch. More interesting is that while Dagworth is today only a hamlet, at Domesday it was held as a manor by Breme, a freeman of Edward the Confessor who had been killed at Hastings. It had at least two churches, one being recorded as half a church with 25 acres of land, another with no land and another half a church assessed with 30 acres of land and 1½ of meadow (LDB fols. 408b and 409b). Warner has pointed out the significance of worða place-names as generally dating after 750 and “without exception these are interesting places” (Warner 1996, 204), a few having monasteries or early Christian associations. The fact one church was probably endowed with 56 acres while another had nothing suggests a small minster community with a dependent chapel was established here some time after the mid-eighth century. With adjacent Haughley parish to the north having a motte and bailey castle built after the Conquest by landowner Hugh de Montfort, the site at Dagworth was likely to have been at, or close to, the centre of an important early estate which maintained itself into the twelfth century. A second cast lead spindle whorl is more enigmatic, being loaned temporarily to an antiquities dealer in Norwich by the finder (see Fig. 12). The whorl is said to have been found in Reedham and to be one of a pair found in the same place. but few are as rune-like as the present example. Sadly, the finder has not responded to

32 John Hines (PAS record) has read the top as d h s þ d and the bottom as s (retrograde) t d h g (possibly n).

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

169

pleas to come forward with this object to allow its find-spot to be recorded. It is frustrating that all we have to go on is the report conveyed by the dealer as an intermediary, as the supposed find-spot, Reedham, is a place of interest due to it being mentioned in a medieval Bury St Edmunds abbey source, Liber Albus, which attributes Felix, the first bishop of the East Angles, as having established an ecclesia in Redham. An above-average land endowment held by the church at Domesday, of some 40 acres worth 6s 8d (LDB fol. 224) might suggest a former importance. Of course, large eleventh-century landholdings are no certain guide to ancient status (Williamson 1993, 152) and the medieval tradition of this Anglo-Saxon foundation is of uncertain veracity (Pestell 1999, 352‒353). The final item is a small circular lead pendant from Quidenham (HER 21882), which appears to show a clumsy runic (or perhaps better pseudo-runic) legend running radially around a central boss.33 No reading is possible as the runes are so hard to make out, with the exception of a ᛏ t-rune at the top. Strengthening its runic credentials is a very similar lead pendant that was found in excavations at Coppergate, York, on which the runic text was slightly clearer (Mainman/Rogers 2000, 2475‒2476, No. 4148). As a cast item with a very similar example known from York, this pendant is probably best treated in a more general locational sense. Jewellery of this Anglo-Scandinavian period enjoys a wide distribution in eastern England, stretching from Suffolk to Yorkshire and as such, the pendant relates more to cultural and stylistic influences from the Scandinavian world, perhaps also influencing the use of runic characters.

Discussion Our understanding of rune-use as an indicator of literacy has most often been considered from a linguistic approach, examining the appearance of new letter forms or sound values. In this paper I have tried instead to provide a wider context to the discovery of these objects. While many items have no obvious significance attached to their find-spot, simply being stray losses of portable objects, it is increasingly likely that many in fact do. Likewise, the expanding corpus of runic objects is beginning to reveal possible patterns with certain artefact-types apparently more favoured for inscription than others. Personal pieces of jewellery is now an obvious category and disc-headed pins seem particularly popular. While pins, strap-ends or brooches have been found with such inscriptions in the past, it is clear from the sheer number of these items recovered that simply being a ‘personal’ item was no particular incentive for it to be marked with a personal name. However, it is notable that of the East Anglian cor-

33 NCM 2004.705.

170

Tim Pestell

pus, three tweezers have now been found with inscriptions, despite not being such common find types as the aforementioned jewellery. Why should this be so? One possibility I raised in discussion with John Hines, when looking at the Baconsthorpe tweezers, was that they may in fact have acted as page holders for manuscripts. This would certainly place them in a literate milieu, which makes some sense of Hines’ reading of the Baconsthorpe text: ‘Read whoso may. Bēaw inscribed the runes’ (Hines 2011, 289‒290). The page-holder idea certainly has parallels from medieval contexts, for instance that from a context of 1290‒1400 at the Greyfriars, Norwich (Emery 2007, 143‒144), and another from an early to mid-thirteenth-century context in Winchester (Biddle/Hinton 1990, 757, Fig. 215 No. 2326A). While the Heacham and Brandon tweezers are not of the same shape as the broad-armed example from Baconsthorpe, it may be that such personal, portable, items were found convenient to slip over manuscript leaves to help them stay open. In a similar way, I have drawn attention to the potentially mutable nature of styli of the same date: a pin could easily enough be used to write text into a wax tablet, while styli on occasion might have been worn as pins signalling their owner’s literate ability (Pestell 2004, 45‒47). Perhaps helping to support this is the recent discovery in Barkston, Lincs., of yet another apparent tweezer arm, made of silver with a runic inscription. In this case, the reading is very similar to three lines of the Old English poem Azarius, an incomplete vernacular retelling of the Book of Daniel found in the tenth-century Exeter Book.34 The particular element seen on the tweezers relates to the story of three youths in Nebuchadnezzar’s furnace, praising God’s creation (Bradley 1982, 284).35 The inscription is, therefore, another highly literate one. That such esoteric texts could be displayed in apparently humble artefacts is most remarkably seen in a small lead pendant found in metal-detection at Weasenham, Norfolk.36 Crudely made and inscribed, it nevertheless displays two verses of symmetrical Anglo-Latin poetry using Latin, Hebrew and Greek name forms and reading (Okasha/Youngs 2003; Howlett 2006): + The name of God is in Hebrew El, Lord and also Adonoi

Still more humble are the lead spindle-whorls that again seem to be a favoured runic object, that from Dagworth joining an increasing number of genuine as well as pseudo-runic companions. Exactly why is uncertain given the wide variety of other domestic or work items that could be chosen, even given the inevitable losses of organic materials or inscriptions on iron objects through corrosion. Spindle whorls might have carried gender associations with females (Stoodley 1999, 33), but

34 The tweezer is recorded as PAS-6F2DA2. 35 The biblical verses appear in Daniel 3:51–90. 36 The pendant is now NCM 2000.3.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

171

it is still a moot point what made these objects apparently so popular or suitable for inscription. These items all draw attention to an emerging feature from more recent runic finds – the evidence for wider literate ability. In particular, if we see literacy as a special or reserved skill which was likely to have been concentrated in those places where it was practiced, then it is equally likely that stray finds bearing such inscribed texts would more probably be found in or near such centres. Given the association of elites with both rune-use, and literacy more generally, their networks of exchange, whether material, cultural or intellectual, are similarly likely to have provided the vectors for the sharing or copying of material culture and text. As we have seen within East Anglia, there are regular associations between the corpus of runic objects and historic estate centres, whether secular or ecclesiastical.37 Moreover, the expanding corpus of finds bears witness to genuine high-level literacy existing behind even brief texts, whether by challenging the reader, invoking charms, writing in Latin or using serifed runes or a runic symbol more familiar from manuscripts. It is the boom in new finds from metal-detecting that is swelling the number of longer and later, Middle Anglo-Saxon, runic texts. Together, many draw attention to a bilingual approach to literacy and the continued use and adaptation of the runic script in the eighth and even ninth century, despite the advance of Latinity and its own alphabet. Perhaps another way to look at this developing use of text is by looking at an alternative index of literacy – the presence of styli, objects used in the creation of text that have the capacity to survive in the ground compared to feather quills. Some 126 Anglo-Saxon styli are now known, largely as a result of metal-detection. They enjoy a wide distribution with a concentration in the east of Britain, those same areas that have traditionally seen most cooperation between archaeologists and detectorists. There are now many isolated finds that would appear to reflect stray losses by literate people moving through the landscape. Much the same might be said of many runic objects. However, a significant number of styli have actually been found on a limited number of sites, demonstrating the concentration of literate practice in certain locations. For instance, 13 have been found at Little Carlton in Lincolnshire, seven at Bawsey in Norfolk, and five at Ryther in Yorkshire. Excavations have yielded twelve styli from Whitby, three from Brandon in Suffolk and 20 at Flixborough on Humberside, two-thirds of which are made of iron, a metal normally screened out by metal-detectorists to reduce the amount of junk retrieved from the plough-soil.38 This suggests that workmanlike writing implements may once have been far more common than is now represented in other metal-detected assemblages.

37 A similar argument has been advanced for the use of runic inscriptions relating to ‘the topography of power’ in Blekinge, Sweden (Carstens/Grimm 2015). 38 These figures are updated from those presented in Pestell (2009b).

172

Tim Pestell

If the picture presented by styli is applicable, it may be that our examples of runic-inscribed objects are also more likely to appear on or near sites associated with literacy, as is certainly the case at Brandon. While sites such as Brandon and Flixborough have attracted debate over their possible interpretation – whether formerly ‘monastic’ or possibly secular estate centres39 – perhaps the fundamental point is that in either case they drew their occupants from the upper echelons of society. These were precisely where such specialist knowledge, and its appreciation, was more likely to be reserved in an increasingly hierarchical world. If this is true for the Christian Anglo-Saxon world, what can we make of those pre-Conversion runic inscriptions? There has long been an implicit understanding that rune-writing was a special skill with a restricted range, but this has rarely been applied to those sites yielding such finds. In some cases it is likely not to have been, with stray losses in the same way as we have seen in the Middle Anglo-Saxon period. However, the Binham bracteate hoard in particular illustrates the way in which certain individuals in society appear to have held power and wealth, and exercised cultural and intellectual interactions even in the late fifth or early sixth centuries. The appearance of more runic finds in North Elmham, Caistor St Edmund, and Loveden Hill, Lincolnshire, let alone on sword fittings in the high-status graves of wealthy Kentish cemeteries, should only reinforce this viewpoint of an elite having control of literate and runic knowledge. This may in turn help us to identify a more nuanced understanding of landscapes of power or belief in the Migration period. It is clear that with metal-detecting the number of runic-inscribed items known from not only East Anglia, but Anglo-Saxon England more generally, will continue to grow. With it, we not only have the potential to increase our understanding of language change, but to frame better our questions about the development of literacy and the wider society producing and using such texts.

References Axboe, Morten, Charlotte Behr, and Klaus Düwel. 2011. “Katalog der Neufunde”. Die Goldbrakteaten der Völkerwanderungszeit ‒ Auswertung und Neufunde. RGA-E 40. Eds. Wilhelm Heizmann and Morten Axboe. Berlin: De Gruyter. 893–999. Axboe, Morten, Klaus Düwel, and Karl Hauck (eds.). 1985–1989. Die Goldbrakteaten der Völkerwanderungszeit: Ikonographischer Katalog 1–3. Münstersche Mittelalter-Schriften 24. Vols. 1–3. München: Fink. Bammesberger, Alfred. 2006. “Das Futhark und seine Weiterentwicklung in der anglo-friesischen Überlieferung”. Das fuþark und seine einzelsprachlichen Weiterentwicklungen, RGA-E 51. Eds. Alfred Bammesberger and Gaby Waxenberger. Berlin/New York: De Gruyter. 171–187. Behr, Charlotte, Tim Pestell, and John Hines. 2014. “The Bracteate Hoard from Binham: An Early Anglo-Saxon Central Place?” Medieval Archaeology 58: 44‒77.

39 Cf. Blair (2005, 204‒212); Loveluck (2001, 120‒121; 2007, 144‒163); Pestell (2004, 18‒64).

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

173

Biddle, Martin and David Hinton. 1990. “Book Clasps and Page Holders”. Object and Economy in Medieval Winchester. Ed. Martin Biddle. Winchester Studies 7.ii. Oxford: Clarendon. 755‒748. Blair, John. 1988. “Introduction: From Minster to Parish Church”. Minsters and Parish Churches: The Local Church in Transition 900‒1200. Ed. John Blair. Oxford: Oxford University Committee for Archaeology. 1‒19. Blair, John. 2005. The Church in Anglo-Saxon Society. Oxford: Oxford University Press. Blinkhorn, Paul. 2012. The Ipswich Ware Project: Ceramics, Trade and Society in Middle Saxon England. Medieval Pottery Research Group Occasional Paper 7. Dorchester: Medieval Pottery Research Group. Bradley, Sidney Arthur James. 1982. Anglo-Saxon Poetry. London: Dent. Brown, Richard and Andy Shelley. 2014. “From Kiln to Cemetery: New Data on Early Medieval and Medieval Ipswich from Excavations at Stoke Quay”. Medieval Archaeology 58: 373–376. Carr, Robert, Andrew Tester, and Peter Murphy. 1988. “The Middle Saxon Settlement at Staunch Meadow, Brandon”. Antiquity 62: 371‒377. Carstens, Lydia and Oliver Grimm. 2015. “Landscape and Lordship – An Archaeological-Historical Analysis of Blekinge’s Topography of Power (1st to 15th Century AD)”. Archäologie und Runen: Fallstudien zu Inschriften im älteren Futhark. Schriften des Archäologischen Landesmuseums 11. Eds. Oliver Grimm and Alexandra Pesch. Kiel/Hamburg: Wachholtz Verlag–Murmann Publishers. 195–227. Caruth, Jo. 2013. “Is This the First Anglo-Saxon Long-House to Be Discovered in England?” Saxon 56.1: 1‒3. Davies, Gareth. 2011. “Settlement, Economy and Lifestyle: The Changing Social Identities of the Coastal Settlement of West Norfolk, 450‒1100 AD”. Unpublished PhD Thesis. University of Nottingham. Davies, John. 2001. Venta Icenorum Caistor St Edmund Roman Town. Norwich: Norfolk Archaeological Trust. Emery, Philip. 2007. Norwich Greyfriars: Pre-Conquest Town and Medieval Friary. East Anglian Archaeology 120. Gressenhall: Norfolk Museums Service. GDB = Great Domesday Book: 2002. Domesday Book: A Complete Translation. Eds. Ann Williams and G. H. Martin. London: Penguin. Hawkes, Sonia and Raymond Ian Page. 1967. “Swords and Runes in South-East England”. Antiquaries’ Journal 47: 1‒26. HE = Bede’s Ecclesiastical History of the English People. 1969. Oxford Medieval Texts. Eds. Bertram Colgrave and Roger A. B. Mynors. Oxford: Clarendon. Hills, Catherine. 1974. “A Runic Pot from Spong Hill, North Elmham”. Antiquaries’ Journal 54: 87‒91. Hills, Catherine. 1991. “The Gold Bracteate from Undley, Suffolk: Some Further Thoughts”. Studien zur Sachsenforschung 7: 145‒151. Hills, Catherine and Sam Lucy. 2013. Spong Hill Part IX: Chronology and Synthesis. McDonald Institute Monographs. Cambridge: McDonald Institute for Archaeological Research. Hills, Catherine. 2014. “Spong Man in Context”. Landscapes and Artefacts: Studies in East Anglian Archaeology Presented to Andrew Rogerson. Eds. Steven Ashley and Adrian Marsden. Oxford: Archaeopress. 79–87. Hilts, Carly. 2013. “Re-Cut Runes”. Current Archaeology 276: 10. Hines, John. 1997. “An Inscribed Disc from the River Yare near Norwich”. Nytt om runor 12: 13‒15. Hines, John. 2000. “The Runic Inscription on the Composite Disc Brooch from Grave 11”. Norwich Southern Bypass, Part II: Anglo-Saxon Cemetery at Harford Farm, Caistor St Edmund. East Anglian Archaeology 92. Ed. Ken Penn. Gressenhall: Norfolk Museums Service. 81‒82. Hines, John. 2006. “An Inscribed Lead Plaque from Holm St Benets, Norfolk”. Nytt om runor 19: 14‒15.

174

Tim Pestell

Hines, John. 2011. “New Light on Literacy in Eighth-Century East Anglia: A Runic Inscription from Baconsthorpe, Norfolk”. Anglia 129 (3‒4): 281‒296. Hines, John. 2019a. “Practical Runic Literacy in the Late Anglo-Saxon Period: Inscriptions on Lead Sheet”. Anglo-Saxon Micro-Texts, Buchreihe der Anglia/Anglia Book Series 67. Eds. Ursula Lenker and Lucia Kornexl. Berlin/Boston, MA: De Gruyter. 29–59. Hines, John. 2019b. “Two Personal Names in Recently Found Anglo-Saxon Runic Inscriptions: Sedgeford (Norfolk) and Elsted (West Sussex)”. Anglia 137: 278–302. Hines, John and Bengt Odenstedt. 1987. “The Undley Bracteate and its Runic Inscription”. Studien zur Sachsenforschung 6: 73‒94. Hodges, Richard. 1982. Dark Age Economics: The Origins of Towns and Trade AD 600‒1000. London: Duckworth. Hoggett, Richard. 2010. The Archaeology of the East Anglian Conversion. Anglo-Saxon Studies 15. Woodbridge: The Boydell Press. Howlett, David. 2006. “An Inscribed Lead Pendant from Norfolk”. Antiquaries Journal 86: 320‒326. Kent, John Philip Cozens. 1975. “The Coins and the Date of the Burial”. The Sutton Hoo Ship Burial Vol. 1. Ed. Rupert Bruce-Mitford. London: British Museum Publications. 578‒647. Ker, Neil. 1964. Medieval Libraries of Great Britain: A List of Surviving Books. 2nd ed. London: Royal Historical Society. LDB = Little Domesday Book: Williams, Ann and G. H. Martin (eds.). 2002. Domesday Book: A Complete Translation. London: Penguin. LE = Liber Eliensis. Ed. E. O. Blake. Camden Society Third Series 92. London: Camden Society. 1962. Lethbridge, Tom. 1951. A Cemetery at Lackford, Suffolk: Report of the Excavation of a Cemetery of the Pagan Anglo-Saxon Period in 1947. Cambridge: Cambridge Antiquarian Society. Licence, Tom. 2006. “The Origins of the Monastic Community of St Benedict at Holme and Bury St Edmunds”. Revue Bénédictine 116: 42‒61. Loveluck, Chris. 2001. “Wealth, Waste and Conspicuous Consumption. Flixborough and Its Importance for Middle and Late Saxon Rural Settlement Studies”. Image and Power in the Archaeology of Early Medieval Britain: Essays in Honour of Rosemary Cramp. Eds. Helena Hamerow and Arthur MacGregor. Oxford: Oxbow. 78‒130. Loveluck, Chris. 2007. Rural Settlement, Lifestyles and Social Change in the Later First Millennium AD: Anglo-Saxon Flixborough in its Wider Context. Excavations at Flixborough 4. Oxford: Oxbow. Malim, Tim, Ken Penn, Ben Robinson, Gerry Wait, and Ken Walsh. 1997. “New Evidence on the Cambridgeshire Dykes and Worsted Street Roman Road”. Proceedings of the Cambridge Antiquarian Society 85: 27‒122. Menghin, Wilfried. 1983. Das Schwert im frühen Mittelalter. Stuttgart: Theiss. Müller, Ulrich. 2002. “Vorläufige Ergebnisse der Untersuchungen in Derenburg, Meerenstieg II, Ldkr. Wernigerode”. Jahresschrift für Mitteldeutsche Vorgeschichte 85: 77–89. Myres, J. Nowell. 1977. A Corpus of Anglo-Saxon Pottery of the Pagan Period. 2 vols. Cambridge: Cambridge University Press. Myres, J. Nowell and Barbara Green. 1973. The Anglo-Saxon Cemeteries of Caistor-by-Norwich and Markshall, Norfolk. Society of Antiquaries Research Report 30. London: Society of Antiquaries. Nielsen, Poul, Klaus Randsborg, and Henrik Thrane (eds.). 1994. The Archaeology of Gudme and Lundeborg. Arkæologiske Studier 10. Copenhagen: Akademisk Forlat, Universitetsforlaget i København. Okasha, Elisabeth and Sue Youngs. 2003. “A Late Saxon Inscribed Pendant from Norfolk”. AngloSaxon England 32: 225‒230. Oxenedes = Chronica Johannis de Oxenedes. 1859. Ed. Sir Henry Ellis. Rolls Series 13. London: HMSO.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

175

Page, Raymond Ian. 1973. “The Runic Inscription from N59”. The Anglo-Saxon Cemeteries of Caistor-by-Norwich and Markshall, Norfolk. Eds. J. Nowell Myres and Barbara Green. London: Society of Antiquaries. 114‒117. Page, Raymond Ian. 1995. “Quondam et Futurus”. Runes and Runic Inscriptions: Collected Essays on Anglo-Saxon and Viking Runes. Eds. David Parsons and Raymond Ian Page. Woodbridge: The Boydell Press. 1‒16. Page, Raymond Ian. 1999. An Introduction to English Runes. Woodbridge: The Boydell Press. Page, Raymond Ian. 2004. “The Runic Inscription”. Tatberht’s Lundenwic: Archaeological Excavations in Middle Saxon London. Pre-Construct Archaeology Monograph 2. Ed. Jim Leary. London: Pre-Construct Archaeology. 103‒104. Page, Raymond Ian. 2014. “Runic Inscriptions”. Staunch Meadow, Brandon, Suffolk. A High Status Middle Saxon Settlement on the Fen Edge. East Anglian Archaeology 151. Eds. Andrew Tester, Sue Anderson, Ian Riddler, and Robert Carr. Bury St Edmunds: Suffolk Archaeological Service. 260‒263. Parsons, David. 1991. “New Runic Finds from Brandon, Suffolk”. Nytt om Runer 6: 8‒11. Parsons, David. 1994. “Anglo-Saxon Runic Inscriptions on Portable Objects”. Unpublished PhD Thesis, University of Cambridge. Parsons, David. 1998. “Byrthferth and the Runes of Oxford, St John’s College, Manuscript 17”. Runeninschriften als Quellen interdisziplinärer Forschung. RGA-E 15. Ed. Klaus Düwel. Berlin/New York: De Gruyter. 441. Penn, Ken. 2000. Norwich Southern Bypass, Part II: Anglo-Saxon Cemetery at Harford Farm, Caistor St Edmund. East Anglian Archaeology 92. Gressenhall: Norfolk Museums Service. Pesch, Alexandra. 2007. Die Goldbrakteaten der Völkerwanderungszeit – Thema und Variation. RGA-E 36. Berlin. Pesch, Alexandra. 2011. “Netzwerk der Zentralplätze. Elitenkontakte und Zusammenarbeit frühmittelalterlicher Reichtumszentren im Spiegel der Goldbrakteaten”. Die Goldbrakteaten der Völkerwanderungszeit – Auswertung und Neufunde. RGA-E 40. Eds. Wilhelm Heizmann and Morten Axboe. Berlin/New York: De Gruyter. 213‒277. Pesch, Alexandra. 2012. “Fallstricke und Glatteis: Die germanische Tierornamentik”. Altertumskunde – Altertumswissenschaft – Kulturwissenschaft: Erträge und Perspektiven nach 40 Jahren Reallexikon der Germanischen Altertumskunde. RGA-E 77. Eds. Heinrich Beck, Dieter Geuenich, and Heiko Steuer. Berlin: De Gruyter. 633‒687. Pestell, Tim. 2004. Landscapes of Monastic Foundation: The Establishment of Religious Houses in East Anglia c. 650‒1200. Anglo-Saxon Studies 5. Woodbridge: The Boydell Press. Pestell, Tim. 2009a. “Stylus”. Early Medieval (Late 5th‒Early 8th Centuries AD) Cemeteries at Boss Hall and Buttermarket, Ipswich, Suffolk. Society for Medieval Archaeology Monograph 27. Ed. Christopher Scull. London: Society for Medieval Archaeology. 246‒247. Pestell, Tim. 2009b. “The Styli”. Life and Economy at Early Medieval Flixborough c. AD 600‒1000. The Artefact Evidence. Excavations at Flixborough 2. Eds. Chris Loveluck and Dave Evans. Oxford: Oxbow. 123‒137. Pestell, Tim. 2011. “Markets, Emporia, Wics, and ‘Productive’ Sites: Pre-Viking Trade Centres in Anglo-Saxon England”. The Oxford Handbook of Anglo-Saxon Archaeology. Eds. Helena Hamerow, David Hinton, and Sally Crawford. Oxford: Oxford University Press. 556‒579. Pieper, Peter. 1987. “Spiegelrunen”. Runor och Runinskrifter. Kungl. Vitterhets Historieboch Antikvitets Akademien. Konferenser 15. Stockholm. 67‒72. Price, Neil. 2010. “Heathen Songs and Devil’s Games”. Signals of Belief in Early England: AngloSaxon Paganism Revisited. Eds. Martin Carver, Alex Sanmark, and Sarah Semple. Oxford: Oxbow. xiii‒xvi. Riddler, Ian, Nicola Trzaska-Nartowski, and S. Hatton. Forthc. An Early Medieval Craft: Objects and Waste of Antler and Bone from Ipswich Excavations 1974–1994. East Anglian Archaeology.

176

Tim Pestell

Rogerson, Andrew. 2003. “Six Middle Anglo-Saxon Sites in West Norfolk”. Markets in Early Medieval Europe Trading and ‘Productive’ Sites, 650‒850. Eds. Tim Pestell and Katharina Ulmschneider. Bollington: Windgather Press. 110‒121. Rogerson, Andrew and Steven Ashley. 2018. “A Selection of Finds from Norfolk Recorded in 2018 and Earlier”. Norfolk Archaeology 48: 112–123. Stenton, Sir Frank. 1922. “St Benet of Holme and the Norman Conquest”. English Historical Review 37: 225‒235. Stoodley, Nick. 1999. The Spindle and the Spear: A Critical Enquiry into the Construction and Meaning of Gender in the Early Anglo-Saxon Burial Rite. British Archaeological Reports British Series 288. Oxford: British Archaeological Reports. Sutherland, C. Humphrey. 1948. Anglo-Saxon Coinage in the Light of the Crondall Hoard. Oxford: Oxford University Press. TAR 1998–1999 = Treasure Annual Report 1998–1999. 2000. Ed. Roger Bland. London: Department for Culture, Media and Sport. Taylor, Harold and Joan Taylor. 1965. Anglo-Saxon Architecture. 2 vols. Cambridge: Cambridge University Press. Tester, Andrew, Sue Anderson, Ian Riddler, and Robert Carr. 2014. Staunch Meadow, Brandon, Suffolk. A High Status Middle Saxon Settlement on the Fen Edge. East Anglian Archaeology 151. Bury St Edmunds: Suffolk Archaeological Service. Warner, Peter. 1996. The Origins of Suffolk. Manchester: Manchester University Press. Waxenberger, Gaby. 2005. “Towards a Phonology of the Old English Runic Inscriptions (Epigraphical Material): The Phonemic Inventory”. Englische Sprachwissenschaft und Mediävistik: Standtpunkte – Perspektiven – Neue Wege. Ed. Gabriele Knappe. Frankfurt a. Main: Peter Lang. 157–166. Waxenberger, Gaby. 2016. “Graphemes: (Re)construction and Interpretation”. Variation within and among Writing Systems: Concepts and Methods in the Analysis of Ancient Written Documents. LautSchriftSprache 1/ScriptandSound 1. Eds. Paola Cotticelli and Alfredo Rizza. Wiesbaden: Reichert. 353–370. Waxenberger, Gaby. 2017. “A New Character on the Sedgeford Runic Handle/Ladle: Sound Value Wanted”. Anglia 135: 627–640. Waxenberger, Gaby. 2018. “All Good Things Come in Threes: The Three Sequences on the Undley Bracteate”. Worte über Wörter. Eds. Kerstin Kazzazi, Thomas A. Fritz, Sabine Wahl, and Karin Luttermann. Tübingen: Stauffenburg. 491–533. Waxenberger, Gaby. 2020a. “Appearances are Deceiving: The Caistor-by-Norwich Astragalus (ca. AD 425–475)”. Ihr werdet die Wahrheit erkennen – Ye Shall Know the Truth. Eds. Hans Sauer and Rüdiger Pfeiffer-Rupp. Trier: WVT. 47–56. Waxenberger, Gaby. 2020b. “The Runic Inscription”. Excavations at Stoke Quay, Ipswich. East Anglian Archaeology 171. Eds. Richard Brown, Steven Teague, Louise Loe, Berni Sudds, and Elizabeth Popescu. Oxford: Oxbow. 199‒202. Waxenberger, Gaby. Forthc. A Phonology of Old English Runic Inscriptions with a Concise Edition and Analysis of the Graphemes. RGA-E. Berlin/Boston, MA: De Gruyter. Webster, Leslie and Janet Backhouse. 1991. The Making of England: Anglo-Saxon Art and Culture AD 600‒900. London: The British Museum. West, J. R. 1932. St Benet of Holme 1020‒1210. The Eleventh and Twelfth Century Sections of Cotton MS Galba E.ii. The Register of the Abbey of St Benet of Holme. Norfolk Records Society 2 and 3. Norwich. White, Sally, John Manley, Richard Jones, John Orna-Ornstein, Catherine Johns, and Leslie Webster. 1999. “A Mid-Fifth Century Hoard of Roman and Pseudo-Roman Material from Patching, West Sussex”. Britannia 30: 301‒315. Whitelock, Dorothy. 1972. “The Pre-Viking Age Church in East Anglia”. Anglo-Saxon England 1: 1‒22.

Runic Finds from the Kingdom of East Anglia and Their Archaeological Contexts

177

Williams, Howard. 2002. “Cemeteries as Central Places – Place and Identity in Migration Period Eastern England”. Central Places in the Migration and Merovingian Periods: Papers from the 52nd Sachsensymposium. Eds. Brigitta Hardh and Lars Larsson. Lund: Almqvist. 341‒362. Williamson, Tom. 1993. The Origins of Norfolk. Manchester: Manchester University Press.

Christopher Scull

The adventus Saxonum from an Archaeological Point of View: How Many Phases Were There? The withdrawal of the Roman Legions to take part in Gibbon’s Decline and Fall of the Roman Empire […] left Britain defenceless and subjected Europe to that long succession of Waves of which History is chiefly composed. While the Roman Empire was overrun […] Britain was attacked by waves […] of Angles, Saxons and Jutes who, landing at Thanet, soon over-ran the country with fire (and, of course, the sword). (Sellar/Yeatman 1930, 12‒13)

1 Introduction I have retained as the title of this contribution the question that the organisers asked me to address in my presentation to the Eichstätt workshop because it so clearly illustrates the need for closer inter-disciplinary dialogue. We all look to colleagues in other disciplines for clear and unequivocal answers when their specialisms impinge upon our own, but among historians and archaeologists there are widely differing opinions about the adventus Saxonum: character, timescale and duration are all contested, and there are schools of thought that would deny any sizeable movement of population. The material (archaeological) evidence is complex and open to multiple interpretations, and archaeological chronologies are contingent models that must be used critically and with an understanding of their limitations. It is difficult when dealing with this subject to disentangle archaeological interpretation from historical and linguistic scholarship, and from an inter-disciplinary perspective we might argue that it would be wrong to do so. Sound inter-disciplinary research, however, demands that we test the different strands of evidence or argument to determine whether they are individually robust or whether assumptions and inter-weavings lend an appearance of strength to weak constructions. Interdisciplinary perspectives offer more powerful interpretative models but their usefulness depends upon the integrity and sensitivity with which we deploy each other’s expertise. In UK archaeology (in contrast, it would seem, to Germany and Continental Europe) a common if not a majority position amongst students and professionals (though not necessarily amongst experts in the particular field) has been revisionist and would downplay or deny the movement of people in explanations of cultural change in the 5th century. This is, broadly speaking, a legacy of paradigm change in archaeological explanation in the 1960s and 1970s – the New Archaeology and https://doi.org/10.1515/9783110796834-009

180

Christopher Scull

the processual Social Archaeology that developed from it in the UK – which inter alia reacted against diffusionist and migrationist perspectives. There is, though, compelling archaeological and biomolecular evidence for the movement of people to Britain in the 5th century that is broadly consistent with the received historical narrative (below), and accepting this allows more complex, nuanced and useful explanations of cultural changes across the British Isles than insistence on endogenous processes of development would permit.1 There are very good reasons to be stringently source-critical in our approaches to the documentary sources for the 5th century in Britain (Dumville 1977; Yorke 1999; Hills 2003, 21–39), but we should also recognise a consistency in the fundamental narrative. There were literate people in Britain, Gaul, and Byzantium in the 5th and 6th centuries who believed – or were content to record – that parts of Britain had fallen under the control of Saxons in the 5th century and were ruled by Angles and Frisians in the 6th (DEB, 22‒26; Chron Gall, 452, Theodosius [II], 18‒19; Wars VIII.xx; Cameron 1985, 213‒215), and among learned and Latinate folk in the late 7th and early 8th centuries in England there was a consciousness or tradition of shared ancestry with people living in the North Sea coastal regions of what are now the Netherlands and north Germany (HE v. 9). If we seek to minimise the role of people from the Continent in the events of 5th-century Britain, we also need to provide a convincing answer to the question: why does origin myth – if this is how we should view all ‘early history’ – invoke a Continental Germanic rather than an imperial Roman past? We need to consider, too, the nature of archaeological evidence, which comprises the forensic traces of past human actions and which is open to multiple interpretations. There are levels of confidence (and agreement) in interpreting archaeological data. Thus, when faced with observed (excavated) phenomena we can usually agree what we are seeing and what human activities it represents. For example, a decorated pot buried containing calcined human bone and burnt dress accessories is a result of cremation mortuary practice. We can convincingly identify this as representing a cultural tradition or traditions, especially when it is found with hundreds or thousands of others in an extensive urn field which is itself widely replicated in the archaeological record. More difficult is applying a fine chronology to such observed phenomena (without which cause and effect become blurred), and more difficult and contentious still are the cultural or historical narratives we construct to explain patterning in the archaeological record on the grand scale. I owe to Edward James (with whom I was sharing a platform at Reading University in the 1990s,

1 For a range of archaeological perspectives on these issues see Arnold (1984, 6‒20); Hines (1990; 1994); Higham (1992); Scull (1995); Lucy (2000, 155‒186); Hills (2003); Chapman (1997); Halsall (2007, 386‒392, 417‒422); Kruse (2007); Brugmann (2011); Härke (2011). For biomolecular approaches see Weale et al. (2002); Thomas et al. (2006; 2008); Pattison (2008); Hills (2009); Hedges (2011); Leslie et al. (2015); Martiniano et al. (2016); Schiffels et al. (2016).

The adventus Saxonum from an Archaeological Point of View

181

and who was valiantly mediating sharp questioning coming my way from medieval historians who had difficulty accepting my archaeological perspective) the observation that conversations with different colleagues about the same archaeological material can give the impression of a kaleidoscope being twisted – the constituent elements remain the same but the patterns of association and perspectives shift. Archaeological interpretation is contingent and (within limits) different archaeologists can interpret the same material very differently, as the examples discussed below demonstrate.

2 The 5th-Century Archaeology of Britain The characteristic archaeology of late Roman Britain – the material correlates of late imperial society, economy and polity such as settlement and building forms, coins, pottery, and the rest of the suite of manufactured and traded consumer goods – disappears from the archaeological record in the earlier 5th century (Esmonde Cleary 1989; 2011; Gerrard 2013). It is superseded in south and eastern England from the second quarter of the 5th century by archaeologies that are very different in most respects and which include a highly visible suite of material culture types and cultural practices closely paralleled in, and clearly derived from, the North Sea coastal areas of mainland Europe and south Scandinavia (Åberg 1926; Hills 1979; 2003; Böhme 1986; Myres 1986; Hines 1990). This is seen most strongly and clearly in eastern England north of the River Thames in the practice of urned cremation burial with grave goods. The ceramic forms and accompanying grave goods, in particular female dress accessories, point to origins in the Elbe-Weser region and Schleswig-Holstein.2 These similarities have been recognised for more than 150 years (cf. Kemble 1856) and are conventionally held to support the documentary narrative of an adventus Saxonum. The novelty of these elements of the material culture record and cultural practice, the scale on which they occur and the changes they represent are so marked and so clearly have their parallels and antecedents on the Continent and in south Scandinavia that the phenomenon cannot be explained convincingly except by the transfer of cognitive structures (mental templates) as part of the cultural equipment of living human beings (Scull 1995, 74). Moreover, given that long-distance travel and relocation was more likely a group than an individual undertaking, these living human beings were moving and settling in structured groups – whether war-bands, age-sets, families, kin-groups or settlement communities or, more likely, all of these.

2 For overviews of material culture types and some key sites, see Genrich (1954); Böhme (1974); Reichstein (1975); Hines (1984); Häβler (1983; 1985; 1990); Knol (2011); Hills/Lucy (2013); Nieuwhof (2013).

182

Christopher Scull

By the last quarter of the 5th century there are characteristic settlement and funerary archaeologies across all of eastern England associated with material cultures that we term generically → ‘Anglo-Saxon’ and sometimes more specifically ‘Anglian’, ‘Saxon’ and – more rarely – ‘Jutish’. It should be stressed, though, that this does not necessarily presuppose Continental descent or contemporary awareness of some specific ethnic identity: these are modern taxonomic chronological and cultural labels. Burial practices are furnished inhumation (something that only appears to become common after the middle of the 5th century) and urned cremation: the characteristic elements of the suite of → material culture items from both burials and settlements are developments of Continental and south Scandinavian types. This archaeology persists into the middle of the 6th century and is characteristic of a large number of cemeteries that come into use between c. AD 475 and c. AC 520 (Hines 1990, 26‒28, Figs. 1‒3; cf. Penn/Brugmann 2007, 42‒75). It seems unlikely that this later phase of the archaeology must represent increased or uncontrolled migration, or simply the growth of an incoming population of Continental descent. More probably, it represents complex developments in social and cultural identity, and their expression in burial practice, in which cultural emulation, exchange and appropriation play a significant part. This is discussed further below. There are, however, very different archaeologies in western Britain, representing autonomous British societies and polities which maintained links with Gaul, Iberia, and the Mediterranean into the 6th century.3 In the frontier zone of north Britain a relatively late and unemphatic Anglo-Saxon archaeological signature is confined to eastern areas (Alcock 1981; Lucy 1999; Collins 2010). The Anglo-Saxon archaeology with which we are concerned thus constitutes only one geographicallycircumscribed element of the material record for post-Roman Britain in the 5th to 7th centuries, and in concentrating on the cultural practices and societies that it represents we should not lose sight of this broader context. It may be argued that there has in the past been something of a temptation and a tendency to squash together Roman and post-Roman Anglo-Saxon archaeological sequences in order to fill the perceived gap left by traditional historical dates for the End of Roman Britain in AD 410 and the Coming of the Saxons in AD 446/449. Thus J. N. L. Myres’ identification of hybrid Roman/Germanic ceramic traditions (Myres 1956; 1969, 66‒72) may be seen as stemming in part from a desire to bridge the gap by locating the deep origins of Anglo-Saxon settlement in 4th-century imperial frontier policy, and the identification of particular classes of late Roman official belt fittings as necessarily representing Germanic military settlers – as much by those who followed them uncritically as by Hawkes/Dunning (1961) themselves – represents a similar concern to fill an archaeological void in the early 5th century with recognisable Anglo-Saxons. However, as scholarship has advanced under-

3 For the archaeology of post-Roman societies in west and north Britain see Laing (1975); Alcock (1971; 1983); Thomas (1986); Dark (1994; 2014); Campbell (2007); Collins (2012).

The adventus Saxonum from an Archaeological Point of View

183

standing – not least through critical evaluation of such influential models as those advanced by Myres and Hawkes/Dunning – it has become increasingly possible to de-couple archaeological chronologies from dependence on crude cultural attribution and contentious historical dates, and to re-evaluate some important classes of material.4 In recent years the corpus of relevant material culture types in England has also been significantly enhanced by the reporting of metal-detecting finds through the Portable Antiquities Scheme (Worrall et al. 2010). As a result, we now have a better appreciation of the Continental contexts and the dating of the earliest evidence for 5th-century Germanic settlement in Britain, and a developing recognition of the archaeological signature of indigenous societies in the south and east of Britain in the 5th century. The refinement of Continental site and material culture chronologies against which the earliest Continental material in England, and its subsequent development, can be assessed suggests settlement from the North Sea coastal region from c. AD 420. It is now possible to identify late Roman material culture types and suites that can be plausibly dated well into the first half of the 5th century in some parts of England (Cool 2000; 2014; Whyman 2001; Gerrard 2004), and material conventionally considered Anglo-Saxon or Germanic federate, in particular late Roman official belt fittings and quoit-brooch-style metalwork, can now be seen as belonging to late Roman traditions and to 5th-century Roman military or indigenous elites in southern Britain (Inker 2000). The recognition that some coinage was coming in to southern Britain from the Continent during the first half and middle of the 5th century and the proposal that clipped siliquae represent an attempt to control coinage that continued in circulation during the same decades together hint at monetary and jurisdictional contexts (Guest 2005, 110‒115; Abdy 2006; Moorhead 2006; cf. Dark 1994, 200‒206). We are beginning, then, to discern in the archaeological record the societies that felt the impact of migration from beyond the former frontiers of the Empire, and to appreciate that they must be understood on their own terms, as dynamic, and in the context of long-term changes in society and culture in the provinces of Britain, and more widely in the Roman west, from the 3rd century (Halsall 2007, 63‒162; Cool 2014).

3 How Many Phases Were There? Which brings us to the question that I was asked to address: How many phases were there? Being flippant, this almost invites the response, “How many do you want?” More seriously, it brings us back to the nature of the archaeological record,

4 Major contributions to the archaeological chronology for the period include Böhme (1974; 1986); Hines et al. (1999); Hines/Bayliss (2013); Hills/Lucy (2013). Important re-evaluations of culturally and chronologically significant material culture types include Bode (1998); Weber (1998); Halsall (2000); Inker (2000); Martin (2015).

184

Christopher Scull

the degrees of certainty with which we can read in it specific human actions – individually and in aggregate – and at what scales, and the precision that archaeological chronologies allow. It also brings us back to the point that different scholars can see the archaeological evidence in very different ways, and that interpretation is influenced by expectation and experience. The notion of “vagues germaniques” (Musset 1965) may not be a useful model – and indeed comes uncomfortably close to parody (Sellars/Yeatman 1930, 12–13; Goscinny/Uderzo 1963, 45) – but it is an influential image of long standing and to some extent underpins the expectation that an adventus Saxonum might leave traces of discrete and identifiable phases in the archaeological record. Before offering an opinion on how the archaeology might be read today it may be useful to reinforce these points by reviewing very briefly three influential but rather different models advanced over the past 60 years by three major scholars: Vera Evison in The Fifth-Century Invasions South of the Thames (1965), John Nowell Linton Myres in Anglo-Saxon Pottery and the Settlement of England (1969) and Horst-Wolfgang Böhme in “Das Ende der Römerherrschaft in Britannien und die angelsächsische Besiedlung Englands im 5. Jahrhundert” (1986). Evison, writing when the Salerno, D-Day and Inchon landings of the Second World War and Korean War were recent memories, saw the archaeological burial evidence south of the River Thames as consistent with a three-pronged military assault across the Channel to Kent, Sussex and Hampshire in AD 449, undertaken by Frankish war-leaders with imperial military experience and followed by conquest and settlement. Weapon burials such as Brighthampton grave 31, Abingdon grave 42, and Petersfinger grave 21 are seen as the burials of these leaders – founder graves in new cemeteries serving new communities of conquerors and settlers. Myres, by contrast, offered a 5-phase model based on his understanding of cremation pottery (and so by implication most applicable to eastern England north of the Thames) and his belief that the earliest Anglo-Saxon settlement took place under late imperial oversight: 1) Overlap and Controlled Settlement AD 360‒410; 2) Transition AD 410‒450; 3) Invasion and Destruction AD 450‒500; 4) Reaction and British Recovery AD 500‒550; and 5) Consolidation AD 550 onwards. Irrespective of whether or not we accept it, this is in many ways a sophisticated and nuanced model but (with the exception of the proposed hybridisation of Roman and Germanic ceramic traditions adduced as evidence for the Phase of Overlap) it is based on extrapolation from the scanty documentary record and could almost have been constructed without any reference to the archaeology. It deals only with a single element of the material evidence, the pottery, and treats it as an illustrative proxy for political and military history – something for which it is not an obviously sensitive source. Böhme, extending to England his survey of 4th- and 5th-century grave finds between the Rivers Elbe and the Loire (Böhme 1974), identified three major groups of material which he interpreted as representing 1) Roman military authority which survived in Britain up to c. AD 450; 2) settlement in eastern England north of the

The adventus Saxonum from an Archaeological Point of View

185

Thames of groups from Saxony and Schleswig-Holstein in the second quarter of the 5th century; and 3) large-scale migration from Saxony and Schleswig-Holstein (seen in the increase in numbers of ‘Anglo-Saxon’ cemeteries) following a final breakdown of late Roman military authority c. AD 450. The congruence with the narratives of Bede and Gildas is clear, and Böhme quite explicitly endorsed a reading of the archaeology in line with the received account of a federate revolt. All three of these models have an eye on the traditional historical narrative, and none accommodates the Scandinavian character of Anglian England from the later 5th century as established by Hayo Vierck (1966) and John Hines (1984; 1992) – although in fairness only Böhme was really in any position to take this into account, and the emphasis of his study was on the earlier 5th century and the North Sea littoral of mainland Europe. All three follow a narrative of invasion or migration and settlement but there are considerable divergences of dating and interpretation within the broad parameters that this establishes. In particular, what Evison saw as evidence for Frankish warriors is interpreted by Böhme as representing an insular or imperial military authority in the first half of the 5th century. Evison herself subsequently linked some specific metalwork and glass types to the first two phases of Myres’ model (Evison 1981). Taking into account more recent work, and the expanded body of relevant finds, the current archaeological picture allows us to propose that there is evidence for the presence in south and east Britain in the first half of the 5th century of elite groups who shared customs, culture and contacts with counterparts in Gaul and who looked to the Roman Imperium. There is no reason to assume that these were necessarily barbarian incomers rather than indigenous elites although, as conditions in contemporary Gaul demonstrate, it may not always be easy or necessarily appropriate to draw too hard-and-fast a distinction between the two (cf. Halsall 2007, 35‒62; 320‒370). There is also evidence for two major episodes or processes of intense cultural contact across the Channel and the North Sea which are likely to have involved the movement and substantial settlement of people from outside the frontiers of the former Empire: 1) from and along the North Sea coastal regions of the Netherlands and Germany in the second quarter and middle years of the 5th century; and 2) from Norway and south Scandinavian the middle decades and third quarter of the 5th century. The evidence suggests that any substantial movement of people (though not continuing contacts) ended during the third quarter of the 5th century. This is the point at which English material culture begins to diverge significantly and consistently from Continental and Scandinavian parent types and, if we envisage a period of migration and settlement over one or two generations, it is the time at which people born in England will have had their closest kinship and social contacts in England rather than Continental homelands – the point at which we might expect the budding-off of autonomous social groups from their parent societies (Scull 1995, 78‒79). At the regional or inter-regional scale adventus may, therefore, stand for the beginning or perceived tipping-point in a longer-term process. At a local scale, every

186

Christopher Scull

individual experience of migration, whether incomer or indigene, would be a different and individual adventus. Long-term and large-scale social processes are the aggregates of individual human actions and experiences: reading archaeology as the forensic traces of human action requires us to adopt perspectives that will allow us to focus at both scales, and those in between.

4 Cultural Change in Eastern England in the 5th Century Adducing movements of people in archaeological explanation does not mean necessarily advocating simple population replacement models or invasion with fire and the sword. Migration can be only one element – albeit an important one – in explaining the cultural changes apparent in the archaeological record of the 5th century in eastern England. It does not explain the collapse of the late Roman system and the virtual disappearance of its markers from the archaeological record; nor in itself does it explain language change or the widespread adoption of what we term Anglo-Saxon material culture and cultural practice, let alone the development of the Anglo-Saxon kingdom structure which existed by the end of the 6th century. For this, we have to consider more complex models of social and economic development, take into account the impacts of incomers on indigenous societies and communities, and consider the possibility that military conflict and political competition in post-imperial Britain were between dynasts, warlords and local polities rather than along some simple ethnic fault line (Scull 1992, 14‒15). However, these things become very much more difficult to explain if we exclude migration from the Continent from the equation or consistently seek to minimise its impact. The early and middle 1990s saw something of a rehabilitation of migration as an explanation in archaeology, drawing on the study of documented population movements and the substantial literature in cultural anthropology and sociology to propose generalising models of mechanisms, processes and motivations to migration (Anthony 1990; 1997; Chapman 1997). There has been criticism of this approach as applied to 5th-century Britain (Halsall 2007, 417‒419), but while there are clearly pitfalls to be avoided in applying anachronistic or over-specific cross-cultural parallels it is difficult to see how a better general understanding of the human dynamics of migration (especially when they may have observable material correlates in the archaeological record) can do anything other than improve the rigour and usefulness of our models (Scull 1995, 73‒79; Burmeister 1998; Kruse 2007, 293‒294, 327‒329, 346‒ 348). For example, approaching the movement of Continental people to Britain in the 5th century as a likely example of Chain Migration (Tilley 1978; Anthony 1997) allows us to model social processes and variables, and their possible archaeological correlates, at a variety of scales. There are three specific phenomena (discussed in Scull

The adventus Saxonum from an Archaeological Point of View

187

1995 but worth recapitulating here) observed across a range of documented migrations which still seem highly relevant to the case of migration to Britain from the Continent and Scandinavia in the 5th century: scouting; information flow and information networks; and the processes by which cultural identities and practices become reinforced and reinvented as a result of migration. Migration on any scale does not happen into the blue. Typically young men are the first to go, establishing themselves and sending for kin, or returning with the information that enables the next trip. Subsequent migration takes place along these known routes. Information flow and information networks are embedded in the structures of kin, co-residence and sodality. This acts to direct migration from specific communities and localities to specific destinations, something that should in theory be detectable in the archaeological record given a sufficiently sensitive cultural signature (such as, for example, the specific use of characteristic material culture types in the cremation ritual). Migration can also act to create or amplify inequalities, first-comers enjoying advantages that might be translated into influence or power as apex groups or families. It is easy to see how these observations might be made to align with conventional views of Germanic society – in particular the institution of the retinue and fissile political structures – and with the traditional narrative of a British ruler’s invitation to the Saxons and their subsequent revolt and settlement. More importantly, though, they bring into focus the likelihood that migration and settlement was not a single short lived event or episode, but a longer process, perhaps occurring over decades and involving some movement back to the homelands from Britain, in which motivations and experiences may have varied greatly with time and place. We may consider, for example, that it was probably a different experience moving in AD 450 from Frisia to an established settlement or community in Norfolk than it was moving in AD 420 from Saxony as one of the first to settle. We also need to consider ‘cultural drift’ and migration as a transformative experience. The act of migration tends to reinforce perceptions of identity but – notwithstanding the information flows embedded in continuing contacts – perceptions and markers of an original common identity diverge with time and distance. This can be seen in modern diasporas, where constructions of identity might be anchored simultaneously both to values and markers adopted from the host society and to an emphatic consciousness of tradition. Comparable social strategies of distinction and legitimation may perhaps be seen in the material culture of eastern England from the middle of the 5th to the early 6th centuries. The development of distinctive insular forms, and regional variations, from the types that indicate the earlier presence of people from the Continental North Sea littoral and south Scandinavia suggests the construction of local and regional identities, differentiation being nested and expressed within broader networks of affiliation which encompassed continuing inter-regional contacts. In all this, it is important that migration is seen as a social and political strategy as well as a subsistence response such as the last resort of the flooded coastal com-

188

Christopher Scull

munity. Applying concepts of affordance and opportunity to the social and political environments may be useful here. If, as is now widely accepted, the Roman Iron Age societies of the North Sea Coastal area and south Scandinavia were structurally dependent on the Roman Empire for the tangibles that cemented their elite value systems, then the dislocations of the early 5th century would have been extremely disruptive politically and socially.5 Gebühr (1998), arguing in detail the case of Angeln, has shown how such a crisis might provoke political fission and external raiding leading to permanent movements of population. However, we need to be alert to nuance and human agency when trying to distinguish between ‘push’ and ‘pull’ factors: one person’s crisis can be another’s opportunity. The rupture of imperial authority in Britain can be seen as opportunity for potentates and warlords to assert authority. The crisis in Angeln, in promoting social and political fission, opened opportunities to seek power, status and prestige elsewhere. Referring back from this to the earlier point about the range of circumstances, motivations and experiences encompassed within the migrations to Britain, it may be legitimate to propose that there was a greater frequency of groups and expeditions organised for warfare in the earlier than the later years of the process. There is an influential strand of thought which emphasises that social and political identities are contingent social constructs, and that the 5th century was a time when new social and political identities were negotiated and constructed in response to the collapse of the Western Empire and the value systems (within and without the Empire) that were dependent on imperial sanction and resource (Geary 1983; 2002; Halsall 2000; 2007, 35‒62). There is much that is persuasive and attractive in this view, and it is worth exploring one particular aspect which is relevant to the archaeology of cultural change in 5th-century Britain. This is, that furnished inhumation, and in particular weapon burial, was not inherently culturally ‘Germanic’ but was a response by elites within the former territories of the Empire to the collapse of the imperial systems from which status and legitimisation flowed: at times of social stress and uncertainty, competitive display in burial was adopted as a strategy of asserting or affirming status and new identities. As already noted, the strongest evidence in both material culture and cultural practice for people from the Continent is found in cremation cemeteries in eastern England from the second quarter of the 5th century. Furnished inhumation, though an element of late Roman burial practice, was not common in post-Roman England before the third quarter of the 5th century and our evidence is that unfurnished inhumation was usual among the post-Roman population in western Britain (Philpott 1991; Watts 1996; Rahtz et al. 2000). Some of the few furnished inhumations in eastern England that can be convincingly dated to before the third quarter of the

5 For aspects of the archaeology of relationships between societies of the North Sea coastal area and south Scandinavia and the Roman Empire see Lund Hansen (1987); Hedeager (1978; 1992); Näsman (1999); Galestin (2010); Rau (2010); von Carnap-Bornheim (2015); Fischer/ Sánchez (2016).

The adventus Saxonum from an Archaeological Point of View

189

5th century (sometimes the earliest or among the earliest in Anglo-Saxon cemeteries which were in use through the rest of the 5th and 6th centuries) contain material culture assemblages which need not be interpreted as ‘Germanic’ or ‘Anglo-Saxon’ at all but can be seen as representing late imperial or immediately post-imperial burial traditions: what Böhme interpreted as representing a Roman military authority. Mucking (Essex) cemetery 1 grave 117 and Mucking cemetery 2 grave 979 are good examples (Hirst/Clark 2009, 209‒211; 306‒368). Freed from over-prescriptive culture-historical interpretation, these become evidence for elements of post-Roman British society whose social identities – in particular a militaristic elite masculine identity – were expressed in burial by weapons and symbols of late imperial official rank. This challenges us to view burials such as those at Mucking not diachronically – as elements of a developmental or teleological narrative of cultural change from Roman to Anglo-Saxon – but synchronically, de-coupled from their perceived status as the earliest phase of an ‘Anglo-Saxon’ cemetery, as evidence for how it was in the early and middle years of the 5th century regardless of what came afterwards. It may be argued, therefore, that the widespread adoption of furnished inhumation in eastern England from the later 5th century embodies both indigenous traditions (inhumation and – arguably – weapon burial) and material culture types and elements of symbolic vocabulary which clearly derived from Continental and south Scandinavian exemplars. (We should note here that cremation is not necessarily an indication of egalitarianism: the evidence is that the clothed corpse was laid out on the pyre in much the same way as it was laid out in the grave, and with many of the same material culture items). The frequency and wide distribution of furnished inhumation has been seen as evidence for untrammelled Germanic migration in the later 5th century but, as noted above, this seems too simple a reading, especially as the characteristic material culture types associated with this archaeology are insular developments of Continental and south Scandinavian types. It is perhaps more likely that this represents at least in part elements of an indigenous population that had previously practised unfurnished inhumation now made highly visible in the archaeological record through the adoption of a new burial rite. This would thus represent a complex two-way process of cultural accommodation and development (representing – dare we say it? – ethnogenesis) in which kindreds of Continental descent adopted elements of insular social display and insular groups adopted elements of dress, material culture and burial rite which clearly derive from Continental cultural practice. We are dealing here with acculturation, or more properly with elements of cultural transmission, appropriation and re-invention. This reading of the archaeology is strengthened by the fact that items now seen as representing post-Roman British material culture of the first half of the 5th century (in particular official belt furniture and quoit-brooch style objects) are found in Anglo-Saxon inhumations of AD 450 or later, as in Brighthampton grave 31 where the scabbard was a repaired or composite piece with a South Scandinavian mouthpiece and quoit-brooch style chape. They clearly had an enduring symbolic value beyond their initial period of use and currency.

190

Christopher Scull

Explanation, therefore, comes down to human behaviour, not to external or impersonal historical process. We need to consider motivations to change or to renew (or not) affiliations, and what this may have meant for different segments of the societies under consideration: in their publication of the cemetery at Wasperton in Warwickshire, which may have been in continuous use from the 4th to the later 6th centuries, Carver et al. (2009, 3‒4, 133‒140) offer a powerful discussion of these issues. Power relationships are likely to have been crucial: the evocation of an ethnic identity may have been more important to the recent incomer whose life was changing, or to the local potentate or warlord seeking legitimacy in an uncertain political climate, than to a local agricultural population – or it may just have been important in different ways. We must, though, keep a sense of perspective. It sometimes seems to be suggested that early medieval people were free to adopt whatever identity they pleased, or to negotiate identity at will, but we have to be careful about exaggerating social freedom. We need to bear in mind the powerful inertial social factors that would have acted against this: descent and the inheritance of tradition, which are the social glue of small-scale pre-industrial societies. But at the same time we need to be aware that traditional societies are not static, and to recognise the effect of structuration: that social structures and social practice may be viewed as epiphenomena, and that inherited tradition, identity and expression are renegotiated (or re-understood) generation by generation. At times of crisis there may be both motivation and freedom to embrace radical change.

5 Where Next? Where might archaeological research effort be directed to clarify and enhance understanding of cultural change in eastern Britain in the 5th-century? Visibility and recognition are major issues. Our understanding of what 5thcentury British society might look like in the archaeological record of south-east England is fragmentary at best, and any opportunity to enhance and refine cultural, economic, social and environmental signatures should be a high priority. Linked to this is the need to understand better what data are already out there through the systematic study and syntheses of new material bearing on the 5th century, in particular metal-detector finds and material reported through the Portable Antiquities Scheme, and so-called ‘grey literature’ generated by commercial archaeology. This will require research perspectives that look across the conventional culture-historical periodisation between Romano-British and Anglo-Saxon archaeologies. Much depends on robust archaeological chronologies and better chronological discrimination. There is real potential further to refine material culture dating by exploiting the data handling and statistical techniques enabled by modern Information Technology and by looking beyond the UK to link insular studies with Scandinavian and Continental schemes. Looking to scientific dating methods, the combi-

The adventus Saxonum from an Archaeological Point of View

191

nation of → radiocarbon and → Bayesian modelling has potential to deliver useful results for 5th-century cremations, especially where there is an observed → stratigraphic sequence (Hines/Bayliss 2013, 555) and rigorous evaluation of the potential of → rehydroxylation kinetics for the precise and accurate dating of pottery must also be a priority (Wilson et al. 2009; 2012). Above all, we must continue to develop chronological frameworks within which the archaeological evidence can be treated on its own merits rather than depending upon pre-suppositions about cultural identity or historical ‘events’. At the time of writing, it is fair to say that the contribution of biomolecular studies to the understanding of population and cultural dynamics in 5th-century Britain has not lived up to (perhaps unrealistic) initial expectations. Studies based on modern DNA face the difficulty of attributing present-day patterning convincingly to specific episodes of migration in the middle of the first millennium AD when the history of Europe before and since has seen constant movements of people (Hills 2009; Hedges 2011; Martiniano et al. 2016). Genomics, especially when allied to isotopic evidence for individual mobility, has great potential, but as yet ancient genomes have been sequenced from only a handful of early medieval individuals in England (Martiniao et al. 2016; Schiffels et al. 2016). The priority must be to generate data-sets that are large enough to permit valid interrogation at the level of community and population, but even then the biological information will address only part of the story. Descent is undoubtedly a significant factor in structuring social identity, but social identity is multi-facetted and not necessarily co-terminous with descent or biological relationships. Care must be taken to ensure that insights from biomolecular approaches are integrated with up-to-date understandings of archaeological chronology, material culture, and the materialisation of social identity. Biases resulting from environmental conditions and past cultural behaviour that have affected the survival of ancient DNA also need to be recognised. There are parts of eastern England where human bone does not survive well in early medieval inhumations (cf. Hills et al. 1984; Greene et al. 1987; Filmer-Sankey/Pestell 2001), and the prevalence of cremation seriously constrains biomolecular investigation of the 5th-century communities in eastern England whose burial practices and material culture show the closest links with the Continent (cf. Hansen et al. 2017). Landscape studies, at a variety of scales and within the broad definition established by the European Landscape Convention (Council of Europe 2000), also have considerable potential (cf. Powlesland 2003). However, if we are going to go beyond the sterile ‘continuity’ debates bequeathed by culture-historical periodisation we need to eschew teleology, adopt sensitive diachronic approaches, and build in concepts of inheritance and affordance, thereby reintroducing human agency and contingency. Last, but not least, if we are going to refine our archaeological models – which in the absence of new documentary sources looks like the best way to advance understanding – we need to deploy theoretical perspectives which assess rigorously from the very substantial literature in the social sciences how the mecha-

192

Christopher Scull

nisms and effects of migration and cultural interaction might be recognised and interpreted in the archaeology of the 5th century. I have emphasised above that in interpreting the 5th century we need to think of human actions, reactions and adaptions within a human timescale: what may have seemed catastrophic change to a mature adult in AD 417 would by AD 430 have seemed the normal way of the world to a young adult born in AD 413. This can, though, put stress on the boundary between forensic rigour and unwarranted speculation. In preparing this contribution I found myself thinking of two writers first encountered well before I began my archaeological career as a university student and who between them have helped shape my instinctive approach to this period. The first is Alfred Duggan, whose novel The Little Emperors (published in 1951) treats the end of Roman Britain from the perspective of imperial bureaucracy and whose Conscience of the King (also published in 1951) is a grimly compelling fantasia on the psychology of warrior and warlord in the 5th century. The second is the children’s novelist Rosemary Sutcliffe who for my money, in her tales The Lantern Bearers (published in 1959) and Sword at Sunset (published in 1963), endures as one of the most acute and influential historical imaginations bearing on 5thcentury Britain. I suspect that their influence is deeper and more widespread than we acknowledge.

Acknowledgements I should like to thank the organisers for the invitation to contribute to the RuneS Symposium at Eichstätt and to include my paper in this volume; colleagues at the Symposium for illuminating comment and discussion; and Alex Bayliss, Hilary Cool, Guy Halsall and John Hines whose comments on an earlier draft have improved this paper: responsibility for the opinions expressed, of course, remains mine alone. I should also like to record my gratitude to all colleagues, past and present, whose work I have cited or drawn on in any way. The original draft of my contribution was submitted in April 2013 and revised to take account of more recent literature in May 2018. When commentating on past publications it can sometimes be difficult not to appear judgemental or condescending but nothing could be further from my intention. We build on the scholarship of others and should not take advantage of hindsight to belittle their responses to the data, questions and circumstances of their times. After all, much of our own current thinking will almost certainly appear quaintly outmoded in many respects to future generations.

References Abdy, Richard. 2006. “After Patching: Imported and Recycled Coinage in Fifth- and Sixth-Century Britain”. Coinage and History in the North Sea World c. AD 500‒1250. Essays in Honour of Marion Archibald. Eds. Barrie Cook and Gareth Williams. Leiden: Brill. 75‒98.

The adventus Saxonum from an Archaeological Point of View

193

Åberg, Nils. 1926. The Anglo-Saxons in England During the Early Years After the Invasion. Uppsala: Almqvist & Wiksells. Alcock, Leslie. 1981. “Quantity or Quality: The Anglian Graves of Bernicia”. Angles, Saxons, and Jutes: Essays Presented to J. N. L. Myres. Ed. Vera Evison. Oxford: Oxford University Press. 168‒186. Alcock, Leslie. 1971. Arthur’s Britain. Harmondsworth: Allen Lane. Alcock, Leslie. 1983. “The Archaeology of Celtic Britain, Fifth to Twelfth Centuries AD”. 25 Years of Medieval Archaeology. Ed. David Hinton. Sheffield: University of Sheffield. 48‒66. Anthony, David. 1990. “Migration in Archaeology: The Baby and the Bathwater”. American Anthropologist 92 (4): 895‒914. Anthony, David. 1997. “Prehistoric Migration as Social Process”. Migrations and Invasions in Archaeological Explanation (British Archaeological Reports International Series 664). Eds. John Chapman and Helena Hamerow. Oxford: British Archaeological Reports. 21‒32. Arnold, Christopher. 1984. Roman Britain to Saxon England. London: Croom Helm. Bode, Martina-Johanna. 1998. Schmalstede: Ein Urnengräberfeld der Kaiser- und Völkerwanderungszeit. Neumünster: Offa-Büche, Neue Folge 78. Böhme, Horst-Wolfgang. 1974. Germanische Grabfunde des 4. bis 5. Jahrhunderts zwischen unterer Elbe und Loire. München: C. H. Beck. Böhme, Horst-Wolfgang. 1986. “Das Ende der Römerherrschaft in Britannien und die angelsächsische Besiedlung Englands im 5. Jahrhundert”. Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 33: 469‒574. Brugmann, Birte. 2011. “Migration and Endogenous Change”. The Oxford Handbook of AngloSaxon Archaeology. Eds. Helena Hamerow, David Hinton, and Sally Crawford. Oxford: Oxford University Press. 30‒45. Burmeister, Stefan. 1998. “Ursachen und Verlauf von Migrationen – Anregungen für die Untersuchung prähistorischer Wanderungen”. Studien zur Sachsenforschung 11: 19‒41. Cameron, Averil. 1985. Procopius and the Sixth Century. Berkeley/Los Angeles: University of California Press. Campbell, Ewen. 2007. Continental and Mediterranean Imports to Atlantic Britain and Ireland AD 400‒800. York: Council for British Archaeology Research Report 157. Carnap-Bornheim, Claus von. 2015. “The Germani and the German Provinces of Rome”. The Oxford Handbook of the Archaeology of Roman Germany. Eds. Simon James and Stefan Krmnicek. Oxford: Oxford Handbooks Online. Carver, Martin, Catherine Hills, and Jonathan Scheschkewitz (eds.). 2009. Wasperton: A Roman, British and Anglo-Saxon Community in Central England. Woodbridge: The Boydell Press. Chapman, John. 1997. “The Impact of Modern Invasions and Migrations on Archaeological Explanation”. Migrations and Invasions in Archaeological Explanation. Eds. John Chapman and Helena Hamerow. Oxford: British Archaeological Reports International Series 664. 11‒20. Collins, Robert. 2010. “Recent Discoveries of Early Anglian Material Culture in the North East”. Medieval Archaeology 54: 386‒390. Collins, Robert. 2012. Hadrian’s Wall and the End of Empire: The Roman Frontier in the 4th and 5th Centuries. London: Routledge. Cool, Hilary. 2000. “The Parts Left Over: Material Culture into the Fifth Century”. The Late Roman Transition in the North. Eds. Anthony Wilmott and Peter Wilson. Oxford: British Archaeological Reports British Series 299. 49‒79. Cool, Hilary. 2014. “Which ‘Romans’; What ‘Home’? The Myth of the ‘End’ of Roman Britain”. AD 410: The History and Archaeology of Late and Post-Roman Britain. Ed. Fiona Haarer. London: Society for the Promotion of Roman Studies. 13‒22. Council of Europe. 2000. The European Landscape Convention. URL: ‹www.coe.int/en/web/ landscape›, last accessed 12 December 2020.

194

Christopher Scull

Dark, Kenneth. 1994. Civitas to Kingdom: British Political Continuity, 300‒800. Leicester: Leicester University Press. Dark, Kenneth. 2014. “Western Britain in Late Antiquity”. AD 410: The History and Archaeology of Late and Post-Roman Britain. Ed. Fiona Haarer. London: Society for the Promotion of Roman Studies. 23‒35. Duggan, Alfred. 1951. The Little Emperors. London: Faber and Faber. Duggan, Alfred. 1951. The Conscience of the King. London: Faber and Faber. Dumville, David. 1977. “Sub-Roman Britain: History and Legend”. History 62: 173–192. Esmonde Cleary, Simon. 1989. The Ending of Roman Britain. London: Batsford. Esmonde Cleary, Simon. 2011. “The Ending(s) of Roman Britain”. The Oxford Handbook of AngloSaxon Archaeology. Eds. Helena Hamerow, David Hinton, and Sally Crawford. Oxford: Oxford University Press. 13‒29. Evison, Vera. 1965. The Fifth-Century Invasions South of the Thames. London: Athlone Press. Evison, Vera. 1981. “Distribution Maps and England in the First Two Phases”. Angles, Saxons, and Jutes: Essays Presented to J. N. L. Myres. Ed. Vera Evison. Oxford: Oxford University Press. 126‒167. Filmer-Sankey, William and Tim Pestell. 2001. Snape Anglo-Saxon Cemetery: Excavations and Surveys 1824–1992. Ipswich: East Anglian Archaeology Report 95. Fischer, Svante and Fernando López Sánchez. 2016. “Subsidies for the Roman West? The Flow of Constantinopolitan Solidi to the Western Empire and Barbaricum”. Opuscula 9: 249–269. Galestin, Marjan. 2010. “Roman Artefacts Beyond the Northern Frontier: Interpreting the Evidence from the Netherlands”. European Journal of Archaeology 13 (1): 64‒88. Geary, Patrick. 1983. “Ethnic Identity as a Situational Construct in the Early Middle Ages”. Mitteilungen der Anthropologischen Gesellschaft in Wien 113: 15‒26. Geary, Patrick. 2002. The Myth of Nations: The Medieval Origins of Europe. Princeton: Princeton University Press. Gebühr, Michael. 1998. „Angulus Desertus?“. Studien zur Sachsenforschung 11: 43‒85. Genrich, Albert. 1954. Formenkreise und Stammesgruppen in Schleswig-Holstein. Neumünster: Offa-Bücher, Neue Folge 10. Gerrard, James. 2004. “How Late is Late? Pottery and the Fifth Century in South-West Britain”. Debating Late Antiquity in Britain AD 300‒700. Eds. Robert Collins and James Gerrard. Oxford: British Archaeological Reports British Series 365. 65‒75. Gerrard, James. 2013. The Ruin of Roman Britain. Cambridge: Cambridge University Press. Green, Barbara, Andrew Rogerson, and Susan White. 1987. The Anglo-Saxon Cemetery at Morning Thorpe, Norfolk. Gressenhall: East Anglian Archaeology Report 36. Goscinny, René and Albert Uderzo. 1963. Astérix et les Goths. Paris: Dargaud. Guest, Peter. 2005. The Late Roman Gold and Silver Coins from the Hoxne Treasure. London: The British Museum Press. Hansen, Henrik, Peter Damgaard, Ashot Margaryan, Jesper Stenderup, Niels Lynnerup, Eske Willerslev, and Morten Allentoft. 2017. “Comparing Ancient DNA Preservation in Petrous Bone and Tooth Cementum”. PLoS ONE 12 (1): e0170940. Härke, Heinrich. 2011. “Anglo-Saxon Immigration and Ethnogenesis”. Medieval Archaeology 55: 1‒28. Häßler, Hans-Jürgen. 1983. Das sächsische Gräberfeld bei Liebenau, Kr. Nienburg (Weser). Teil 2. Beiträge zur Frühgeschichte Norddeutschlands. Studien zur Sachsenforschung 5.1. Hildesheim: August Lax. Häßler, Hans-Jürgen. 1985. Das sächsische Gräberfeld bei Liebenau, Kr. Nienburg (Weser). Teil 3. Beiträge zur Frühgeschichte Norddeutschlands. Studien zur Sachsenforschung 5.2. Hildesheim: August Lax.

The adventus Saxonum from an Archaeological Point of View

195

Häßler, Hans-Jürgen. 1990. Das sächsische Gräberfeld bei Liebenau, Kr. Nienburg (Weser). Teil 4. Beiträge zur Frühgeschichte Norddeutschlands. Studien zur Sachsenforschung 5.4. Hildesheim: August Lax. Halsall, Guy. 2000. “Archaeology and the Late Roman Frontier in Northern Gaul: The So-Called Föderatengräber Reconsidered”. Grenze und Differenz im früheren Mittelalter. Eds. Walter Pohl and Helmut Reimitz. Wien: Verlag der Österreichischen Akadamie der Wissenschaften. 167‒180. Halsall, Guy. 2007. Barbarian Migrations and the Roman West: 376‒568. Cambridge: Cambridge University Press. Hawkes, Sonia and Gerald Dunning. 1961. “Soldiers and Settlers in Britain, Fourth to Fifth Century”. Medieval Archaeology 5: 1‒70. Hedeager, Lotte. 1978. “A Quantitative Analysis of Roman Imports in Europe North of the Limes and the Question of Roman-Germanic Exchange”. New Directions in Scandinavian Archaeology. Eds. Kristian Kristiansen and Carsten Paludan-Müller. Copenhagen: Nationalmuseets Forlag. 191‒216. Hedeager, Lotte. 1992. Iron-Age Societies: From Tribe to State in Northern Europe, 500 BC to AD 700. Oxford: Blackwell. Hedges, Robert. 2011. “Anglo-Saxon Migration and the Molecular Evidence”. The Oxford Handbook of Anglo-Saxon Archaeology. Eds. Helena Hamerow, David Hinton, and Sally Crawford. Oxford: Oxford University Press. 79‒90. Higham, Nicholas. 1992. Rome, Britain and the Anglo-Saxons. London: Routledge. Hills, Catherine. 1979. “The Archaeology of Anglo-Saxon England in the Pagan Period: A Review”. Anglo-Saxon England 8: 297‒329. Hills, Catherine. 2003. Origins of the English. London: Duckworth. Hills, Catherine. 2009. “Anglo-Saxon DNA?”. Mortuary Practices and Social Identities in the Middle Ages: Essays in Burial Archaeology in Honour of Heinrich Härke. Eds. Duncan Sayer and Howard Williams. Exeter: University of Exeter Press. 125‒140. Hills, Catherine, Kenneth Penn, and Robert Rickett. 1984. The Anglo-Saxon Cemetery at Spong Hill, North Elmham, Part 3: Catalogue of Inhumations. Gressenhall: East Anglian Archaeology Report 21. Hills, Catherine and Samantha Lucy. 2013. The Anglo-Saxon Cemetery at Spong Hill, North Elmham. Volume 9. Cambridge: McDonald Institute for Archaeological Research. Hines, John. 1984. The Scandinavian Character of Anglian England in the Pre-Viking Period. Oxford: British Archaeological Reports British Series 124. Hines, John. 1990. “Philology, Archaeology and the adventus Saxonum vel Anglorum”. Britain 400‒600: Language and History. Eds. Alfred Bammersberger and Alfred Wollman. Heildelberg: Winter. 17‒36. Hines, John. 1992. “The Scandinavian Character of Anglian England: An Update”. The Age of Sutton Hoo: The Seventh Century in North-Western Europe. Ed. Martin Carver. Woodbridge: The Boydell Press. 315‒329. Hines, John. 1994. “The Becoming of the English: Identity, Material Culture and Language in Early Anglo-Saxon England”. Anglo-Saxon Studies in Archaeology and History 7: 49–59. Hines, John, Karen Høilund Nielsen, and Frank Siegmund (eds.). 1999. The Pace of Change: Studies in Early Medieval Chronology. Oxford: Oxbow. Hines, John and Alexandra Bayliss. 2013. Anglo-Saxon Graves and Grave Goods of the Sixth and Seventh Centuries AD: A Chronological Framework. Leeds: Society for Medieval Archaeology Monograph 33. Hirst, Sue and Dido Clark. 2009. Excavations at Mucking Volume 3: The Anglo-Saxon Cemeteries. London: Museum of London Archaeology. Inker, Peter. 2000. “Technology as Active Material Culture: The Quoit-brooch Style”. Medieval Archaeology 44: 25‒52.

196

Christopher Scull

Kemble, John Mitchell. 1856. “On Mortuary Urns Found at Stade-on-the-Elbe, and Other Parts of North Germany, Now in the Museum of the Historical Society of Hanover”. Archaeologia 36: 270‒283. Knol, Egge. 2011. “The First Early Medieval Cemeteries along the Northern Dutch Coasts and their Significance for Anglo-Saxon Migration”. Transformations in North-Western Europe (AD 300‒ 1000): Proceedings of the 60th Sachsensymposion 19‒23 September 2009 Maastricht. Ed. Titus Panhuysen. Hannover: Niedersächsisches Landesmuseum. 218‒227. Kruse, Pernille. 2007. “Jutes in Kent? On the Jutish Nature of Kent, Southern Hampshire and the Isle of Wight”. Probleme der Küstenforschung im südlichen Nordseegebiet 31: 243‒376. Laing, Lloyd. 1975. The Archaeology of Late Celtic Britain and Ireland c. AD 400‒1200. London: Methuen. Leslie, Stephen, Bruce Winney, Garrett Hellenthal, Dan Davison, Abdelhamid Boumertit, Tammy Day, Katarzyna Hutnik, Ellen Royrvik, Barry Cunliffe, Wellcome Trust Case Control Consortium 2, International Multiple Sclerosis Genetics Consortium, Daniel Lawson, Daniel Falush, Colin Freeman, Matti Pirinen, Simon Myers, Mark Robinson, Peter Donnelly, and Walter Bodmer. 2015. “The Fine-Scale Genetic Structure of the British Population”. Nature 519: 309–314. Lucy, Samantha. 1999. “Changing Burial Rites in Northumbria AD 500‒750”. Northumbria’s Golden Age. Eds. Jane Hawkes and Susan Mills. Stroud: Sutton. 12‒43. Lucy, Samantha. 2000. The Anglo-Saxon Way of Death. Stroud: Sutton. Lund Hansen, Ulla. 1987. Römischer Import im Norden: Warenaustausch zwischen dem Römischen Reich und dem freien Germanen während der Kaiserzeit unter besonderer Berücksichtigung Nordeuropas. Copenhagen: Nordiske Fortidsminder, Serie B, Band 10. Martin, Toby. 2015. The Cruciform Brooch and Anglo-Saxon England. Woodbridge: The Boydell Press. Martiniano, Rui, Anwen Caffell, Malin Holst, Kurt Hunter-Mann, Janet Montgomery, Gundula Müldner, Russell L. McLaughlin, Matthew D. Teasdale, Wouter van Rheenen, Jan H. Veldink, Leonard van den Berg, Orla Hardiman, Maureen Carroll, Steve Roskams, John Oxley, Colleen Morgan, Mark Thomas, Ian Barnes, Christine McDonnell, Matthew Collins, and Daniel Bradley. 2016. “Genomic Signals of Migration and Continuity in Britain before the AngloSaxons”. Nature Communications 7: 10326. Moorhead, Sam. 2006. “Roman Bronze Coinage in Sub-Roman and Early Anglo-Saxon England”. Coinage and History in the North Sea World c. AD 500‒1250: Essays in Honour of Marion Archibald. Eds. Barrie Cook and Gareth Williams. Leiden: Brill. 99‒109. Musset, Lucien. 1965. Les Invasions: Les Vagues Germaniques. Paris: Presses Universitaires de France, Nouvelle Clio 12. Myres, John Nowell Linton. 1956. “Romano-Saxon Pottery”. Dark Age Britain: Studies Presented to E. T. Leeds. Ed. Donald Harden. London: Methuen. 16‒39. Myres, John Nowell Linton. 1969. Anglo-Saxon Pottery and the Settlement of England. Oxford: Oxford University Press. Myres, John Nowell Linton. 1986. The English Settlements. Oxford: Oxford University Press. Näsman, Ulf. 1999. “The Ethnogenesis of the Danes and the Making of a Danish Kingdom”. The Making of Kingdoms. Eds. Tania Dickinson and David Griffiths. Oxford: Anglo-Saxon Studies in Archaeology and History 10. 1‒10. Nieuwhof, Annet. 2013. “Anglo-Saxon Immigration or Continuity? Ezinge and the Coastal Area of the Northern Netherlands in the Migration Period”. Journal of Archaeology in the Low Countries 5 (1): 53‒83. Pattison, John. 2008. “Is it Necessary to Assume an Apartheid-Like Social Structure in Early Anglo-Saxon England?” Proceedings of the Royal Society B: Biological Sciences 275 (1650): 2423–2429. Penn, Kenneth and Birte Brugmann. 2007. Aspects of Anglo-Saxon Inhumation Burial: Morning Thorpe, Spong Hill, Bergh Apton and Westgarth Gardens. Dereham: East Anglian Archaeology Report 119.

The adventus Saxonum from an Archaeological Point of View

197

Philpott, Robert. 1991. Burial Practices in Roman Britain: An Archaeological Survey of Grave Treatment and Furnishing AD 43‒410. Oxford: British Archaeological Reports British Series 219. Powlesland, Dominic. 2003. “The Heslerton Parish Project: 20 Years of Archaeological Research in the Vale of Pickering”. The Archaeology of Yorkshire: An Assessment at the Beginning of the 21st Century. Eds. Terry Manby, Stephen Moorhouse, and Patrick Ottaway. Leeds: Yorkshire Archaeological Society. 275‒292. Rahtz, Philip, Sue Hirst, and Susan Wright. 2000. Cannington Cemetery. London: Britannia Monograph Series 17. Rau, Andreas. 2010. Nydam Mose. Die personengebundenen Gegenstände. Grabungen 1989‒1999. Århus: Jysk Arkaeologisk Selskabs Skrifter 72. Reichstein, Joachim. 1975. Die kreuzförmige Fibel. Neumünster: Offa-Bücher, Neue Folge 34. Schiffels, Stephan, Wolfgang Haak, Pirita Paajanen, Bastien Llamas, Elizabeth Popescu, Louise Loe, Rachel Clarke, Alice Lyons, Richard Mortimer, Duncan Sayer, Chris Tyler-Smith, Alan Cooper, and Richard Durbin. 2016. “Iron Age and Anglo-Saxon Genomes from East England Reveal British Migration History”. Nature Communications 7: 10408. Scull, Christopher. 1992. “Before Sutton Hoo: Structures of Power and Society in Early East Anglia”. The Age of Sutton Hoo: The Seventh Century in North-Western Europe. Ed. Martin Carver. Woodbridge: The Boydell Press. 3‒23. Scull, Christopher. 1995. “Approaches to Material Culture and Social Dynamics of the Migration Period in Eastern England”. Europe Between Late Antiquity and the Middle Ages: Recent Archaeological and Historical Research in Western and Southern Europe. Eds. John Bintliffe and Helena Hamerow. Oxford: British Archaeological Reports International Series 617. 71‒83. Sellar, Walter Carruthers and Robert Julian Yeatman. 1930. 1066 and All That. London: Methuen. Sutcliffe, Rosemary. 1959. The Lantern Bearers. Oxford: Oxford University Press. Sutcliffe, Rosemary. 1963. Sword at Sunset. London: Hodder & Stoughton. Thomas, Charles. 1986. Celtic Britain. London: Thames & Hudson. Thomas, Mark, Michael Stumpf, and Heinrich Härke. 2006. “Evidence for an Apartheid-Like Social Structure in Early Anglo-Saxon England”. Proceedings of the Royal Society B: Biological Sciences 273 (1601): 2651–2657. Thomas, Mark, Michael Stumpf, and Heinrich Härke. 2008. “Integration Versus Apartheid in PostRoman Britain: A Response to Pattison”. Proceedings of the Royal Society B: Biological Sciences 275 (1650): 2491–2421. Tilley, Charles. 1978. “Migration in Modern European History”. Human Migration: Patterns and Policies. Eds. William McNeill and Ruth Adams. Bloomington: Indiana University Press. 48‒74. Vierck, Hayo. 1966. Some Leading Types of the Anglian Province of Material Culture, Fifth to Seventh Centuries AD, with their Overseas Connection. BLitt Thesis, Oxford University. Watts, Lorna. 1996. Henley Wood, Temples and Cemetery Excavations, 1962–69 by the Late Ernest Greenfield and Others. London: Council for British Archaeology Research Report 99. Weale, Michael, Deborah Weiss, Rolf Jager, Neil Bradman, and Mark Thomas. 2002. “Y Chromosome Evidence for Anglo-Saxon Mass Migration”. Molecular Biology and Evolution 19.7: 1008–1021. Weber, Martin. 1998. “Das Gräberfeld von Issendorf, Niedersachsen. Ausgangspunkt für Wanderungen nach Britannien?” Studien zur Sachsenforschung 11: 199‒212. Whyman, Mark. 2001. Late Roman Britain in Transition, AD 300‒500: A Ceramic Perspective from East Yorkshire. PhD thesis, University of York. Wilson, Moira, Margaret Carter, Christopher Hall, William Hoff, Ceren Ince, Shaun Savage, Bernard Mckay, and Ian Betts. 2009. “Dating Fired-Clay Ceramics Using Long-Term Power Low Rehydroxylation Kinetics”. Proceedings of the Royal Society Series A: Mathematical, Physical and Engineering Sciences 465: 2407‒2415.

198

Christopher Scull

Wilson, Moira, Andrea Hamilton, Ceren Ince, Margaret Carter, and Christopher Hall. 2012. “Rehydroxylation (RHX) Dating of Archaeological Pottery”. Proceedings of the Royal Society Series A, Mathematical, Physical and Engineering Sciences 468: 3476‒3493. Worrall, Sally, Egan, Geoff, Naylor, John, Leahy, Kevin, and Michael Lewis (eds.). 2010. A Decade of Discovery: Proceedings of the Portable Antiquities Scheme Conference 2007. Oxford: British Archaeological Reports British Series 520. Yorke, Barbara. 1999. “The Origins of Anglo-Saxon Kingdoms: The Contribution of the Written Sources”. The Making of Kingdoms. Eds. Tania Dickinson and David Griffiths. Oxford: AngloSaxon Studies in Archaeology and History 10. 25–29.

Theo Vennemann

The Runic Inscription skanomodu: Frisian or Anglo-Frisian? Abstract: The dominant though controversial interpretation of the → Anglo-Frisian runic inscription skanomodu on an early sixth-century gold solidus, reworked into a pendant or brooch, is that of a bahuvrihi → compound +skān-o+mōd-u1 ‘(having a) beautiful mind’, understood as a personal name. On this interpretation the first constituent would be derived from the Germanic a-grade class II deverbal adjective base +skaun- ‘beautiful’ by the monophthongisation +au > ā, which is unique to Frisian, with the consequence that the inscription itself has to be considered Frisian. In the present paper the possibility is investigated that the first constituent is derived from the Germanic a-grade class I deverbal adjective base +skain- ‘visible, shiny’ by the monophthongisation +ai > ā which is shared by Frisian and Old English, with the consequence that if this alternative analysis is accepted, the inscription can no longer be classified as exclusively Frisian by this criterion but only as Anglo-Frisian, i.e., Anglo-Saxon or Frisian. The second argument for the Frisianness of the skanomodu inscription, the final -u, rests on the assumption that this -u derives from the masculine nominative singular termination +-az, which only works for Frisian. However, a masculine nominative would be unexpected on a piece of jewellery. The conclusion reached is that skanomodu is a feminine personal name in the dative.2

1 Introduction: The Coin and the Inscription skanomodu is a runic inscription on an early medieval coin of unknown provenance, now kept in the British Museum, London (see Fig. 1). The solidus is described numismatically in Berghaus (1967, especially 9‒11 and 23). The following quotation reflects what a runologist, Ray Page (1999, 185), considers worth knowing about the coin: There is no find-spot, and the only known fact of its early history is that it was in George III’s cabinet. Comparable runic solidi come from the Low Countries, and there is no numismatic objection to a Frisian provenance for this one. The skanomodu piece is usually described as a copy of an issue of the late fourth-/early fifth-century Emperor Honorius, with an obverse

1 As in previous publications, I mark reconstructed forms by a raised cross +, incorrect forms by an asterisk *. In quoted material the asterisk may either indicate reconstructed or incorrect forms. 2 This paper would not have come into being without the help and advice of friends and colleagues. I beg the reader to take note of the final section, “Acknowledgements”. https://doi.org/10.1515/9783110796834-010

200

Theo Vennemann

Fig. 1: The skanomodu solidus; © The Trustees of the British Museum, London.

giving the imperial bust, and a reverse of the emperor, holding standard and Victoriola, trampling upon a captive. Its function is uncertain. On the whole it is unlikely it was intended as currency despite its appropriate weight, 4.35 grammes, for a solidus. Such a gold coin would be too heavy for most commercial uses in the late sixth/early seventh century. This specimen once had a loop to enable it to be worn as a pendant or brooch, so it is likely it was a copy of a Roman coin that had been converted into a jewel, as so many were in the Dark Ages. The legends of the original have not survived on this copy. They have degenerated into meaningless letter-like groups save on the reverse where part is replaced by the runic skanomodu.

This inscription is written in the Anglo-Frisian → futhorc (see Fig. 1).3 It can therefore reasonably be assumed to be either → Anglo-Saxon or → Anglo-Frisian.

2 Anglo-Saxon or Frisian? Klaus Düwel (2008, 85 f.) writes about the provenance of the skanomodu solidus: Als ältestes a[n]g[el]s[ächsisches] Runenzeugnis galt früher der geöste skanomodu-Solidus (Ende 5./Anfang 6. J[ahr]h[undert]4) von unbekanntem Fundort mit dem Männernamen ‘Schönmut’, in jüngster Zeit wird er dem kleinen fries[ischen] Corpus zugerechnet. [...] Den

3 Copied from Bammesberger (1990a, 457). Transliteration added. 4 AD 500‒525, according to Berghaus (1967, 22, 39).

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

201

skanomodu-Solidus als ags. Inschrift einzuordnen, hängt sicher auch damit zusammen, daß es fries. Inschriften erst seit dem Ausgang des 19. Jhs. gibt und eine Diskussion der unterscheidenden Merkmale erst seit dem Katalog von Düwel/Tempel (1970) einsetzte. ‘The eyeleted skanomodu solidus (end of 5th/beginning of 6th century5) of unknown finding site and inscribed with the man’s name ‘Beautiful-mind’ used to be regarded as the oldest Anglo-Saxon runic document; most recently it is counted as part of the small Frisian corpus. [...] Categorizing the skanomodu solidus as an Anglo-Saxon inscription is undoubtedly connected to the fact that Frisian inscriptions have only been known since the end of the 19th century and that a discussion of the differentiating features only set in with the catalogue of Düwel/ Tempel (1970).’

The reason why the solidus is now considered Frisian rests primarily on a linguistic consequence of the widely assumed interpretation of skanomodu as a bahuvrihi compound skān-o+mōd-u ‘beautiful-mind’ (‘having a beautiful mind’): skān- can reflect the Germanic adjectival base +skaun- ‘beauti-’ only in Frisian, viz. on account of the exclusively Frisian monophthongisation of +au into ā. This identification of skan- with the +skaun- of +skaun-i- ‘beautiful’ may, of course, be questioned, and it has been questioned, e.g., by Dietrich Hofmann (1976, 73), who wrote: 6 Als friesisch gilt vielen Forschern die Runeninschrift skanomodu auf einer Münze, die sich jedenfalls seit der Zeit von König Georg III. (1760‒1820) und vielleicht noch länger in England befindet.7 In der Tat sprechen einige Anzeichen dafür, vor allem das in friesischen Inschriften mehrfach bezeugte, sprachgeschichtlich rätselhafte -u. Wenig sicher ist allerdings die Form skano- mit scheinbar friesischer Monophthongierung von germ[anisch] au. Falls es überhaupt richtig ist, sie mit germ. *skauni- ‘schön’ zu verbinden, wäre -o- statt -inicht korrekt, und dann könnte auch -a- ungenaue Schreibung statt -au- sein, unabhängig von der friesischen Monophthongierung. Es ist unwahrscheinlich, dass diese schon durchgeführt war, als die Inschrift etwa im ersten Viertel des 6. Jahrhunderts entstand. 8 skanomodu on a coin ‘Many researchers consider Frisian the runic inscription which has been in England at least since the time of King George III (1760‒1820) and possibly longer.9 Indeed several signs speak in favor of this, above all the -u which occurs in several Frisian inscriptions but is enigmatic from the perspective of historical linguistics. The form skano- with apparently Frisian monophthongisation of Gmc. au is, however, not very certain. If it is at all correct to connect it with Gmc. *skauni- ‘beautiful’, -o- rather than -i- would not be accurate, and in that case -a- could be an imprecise rendering of -au-, too, independently of the Frisian monophthongisation. It is improbable that the latter had already been carried out when the incription originated, roughly in the first quarter of the 6th century.’

The question of whether skanomodu is Anglo-Saxon or Frisian may seem a matter of small importance to some people, but to others it is not. For some Frisians it is a

5 Cf. the preceding note. 6 The precisely linearised form of the inscription is owed to Düwel/Tempel (1970, Fig. 20). It is also used in Quak (1990, 361). 7 Cf. reference to Page (1973, 35 f., 186, 188) and Miedema (1974, 112 f., 117). 8 Cf. note 6 above. 9 Cf. note 7 above.

202

Theo Vennemann

matter of ethnic pride or identity, as is evident from the fact that one Frisian society of university students and teachers with a specialised interest in matters Frisian named itself with this “oldest known Frisian word”: The members of the first board had to think of a name for the newly born study association. During Oebele Vries’ lecture for the course Historical Sociolinguistics they came across the word Skanomodu. Skanomodu is the oldest known Frisian word. The word is inscribed on a coin which dates from about AD 800 and translates in Modern Frisian to skjinne moed, which in English translates to fair courage.10 The board members thought Skanomodu was a lovely name and so the name was given to the study association (from the Internet site “Skanomodu: Study Association for Minorities and Multilingualism”: last accessed 14 June 2014).

Some linguists, too, attribute a more-than-runological significance to the object: Bekanntlich trägt nun ein Goldsolidus aus dem frühen 6. Jahrhundert die Runeninschrift skanomodu. Die Inschrift wird als urfriesisch und das ā in skāno < germ[anisch] *skaun- ‘schön’ als der früheste Beleg für die Entwicklung einer eigenständigen friesischen Sprache betrachtet, so von Miedema (1977) und neuerdings auch von Hofstra (1989)11 (Århammer 1995, 74). ‘As is well known, a gold solidus from the early 6th century bears the runic inscription skanomodu. The inscription is considered Proto-Frisian, and the ā in skāno < Gmc. *skaun- ‘beautiful’ the earliest evidence for the development of an independent Frisian language, e.g., by Miedema (1977) und recently also by Hofstra (1989).’12

For linguists, the question may become the basis for an argument. Here are three examples:13 Aufgrund der Runeninschriften kann die Monophthongierung [der fallenden w[est]germ[anischen] Diphthonge /ai/ und /au/] schon für das 6. J[ahr]h[undert] angesetzt werden, z. B. die Form skanomodu mit urfr[iesisch] (6. Jh.; Quak 1990, 361) skān- < wgerm. *skaun- (Miedema 1974, 121) (Versloot 2001, 768). ‘On the evidence of the runic inscriptions the monophthongisation [of the falling West Germanic diphthongs /ai/ and /au/] can be dated as early as the 6th century, e.g. the form skanomodu with Proto-Frisian (6th century; Quak 1990, 361) skān- < West Gmc. *skaun- (Miedema 1974, 121).’

10 They did not get the date right, and the translation also differs somewhat from that of the specialists. See, e.g., the quotation from Düwel (2008) farther above. 11 The author himself, however, does not seem convinced because he continues: “Bezweifelt wurde die Aussagekraft der āc-Rune der Inschrift von Hofmann (1976, 73; 1989, 376).” (‘The relevance of the āc-rune of the inscription was cast into doubt by Hofmann (1976, 73; 1989, 376).’) Cf. the Hofmann quote farther above. 12 Cf. the preceding note. 13 A fourth example occurs in note 44 below, a further one in Nielsen (2000, 117).

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

(1)

203

Als einziger zuverlässiger früher Beleg für monophthongiertes /ā/ < /au/ in den friesischen Runeninschriften verbleibt sonach ska2nomodu14 = vor-afries. Skānomōdə15 m. auf dem in London aufbewahrten Goldsolidus unbekannter Herkunft, zu datieren 576–600/610 (Nedoma 2007, 310). ‘Consequently ska2nomodu16 = pre-Old Frisian Skānomōdə17 m[asculine] on the gold solidus of unknown provenance (kept in London, dated 576–600/610) is the only reliable early piece of evidence remaining for monophthongised /ā/ < /au/ in the Frisian runic inscriptions.’

(2)

Based on onomastic material, Maurits Gysseling has argued that Frisian monophthongization of *au to ā can be dated to the eighth century on the basis of place-name evidence.18 But runic evidence, notably the skanomodu inscription (< *skaun- ‘beautiful’), if it is Frisian, might suggest an even earlier date for monophthongization, around the sixth century” (Laker 2007, 178).

It follows that the matter is less trivial than it may appear at first sight and that the question of the Frisianness of the skanomodu inscription – namely its skan- and -u constituents – deserves attention.

3 More on the “Frisian” Features of scanomodu At the conference “Across the North Sea – North Sea Connections from AD 400 into the Viking Age: Second Interdisciplinary Symposium on Runes and Related topics in Frisia” (Fries Museum, Leeuwarden, 5‒8 June 2014), Gaby Waxenberger gave a paper, “The Frisian Runic Corpus: Sounds and Forms”, which was, meritoriously, accompanied by a fully fledged article (since published as Waxenberger 2017) on which the presentation was based. There Waxenberger includes the skanomodu coin among the 15 inscriptions “considered” in her analysis, i.e., more or less tentatively counted as Frisian. Although there occur places in her paper where skanomodu is listed under the heading ?Frisian inscription (with a question mark),

14 Nedoma writes a1, a2, a3 for the Frisian runes that other runologists indicate by their probable sound values æ, a, o (cf. Nedoma 2007, 300). 15 Nedoma puts a dot under the ə and writes: “Daß hier -u für altes *-az (bzw. *-an) steht, deutet auf durchgeführte Reduktionsprozesse, wobei als output-Lautung am ehesten an einen geschlossen realisierten Zentralvokal zu denken ist, phonetisch [...] ‘obermittelhoch’; der mittlere Zentralvokal [ə] wäre wohl durch -e- bezeichnet worden.” (‘That -u here stands for old *-az (or, respectively, *-an), indicates that reduction processes have been carried out, where the output sound first coming to mind is a close central vowel, phonetically [...] ‘upper middle high’; the middle central vowel [ə] probably would have been marked by -e-.’) 16 Cf. note 14 above. 17 Cf. note 15 above. 18 Here, a footnote refers to Gysseling (1962, 9–11).

204

Theo Vennemann

Waxenberger (2017, 107) seems to concur in essence with Alfred Bammesberger’s (1990a) assessment, which she summarises as follows: For Bammesberger (1990a, 462) it ‘would seem that scanomodu cannot readily be viewed as Old English. scān- would be fully in agreement with Frisian phonology’. Even though OE ēa is clearly the ‘end result’ of Gmc. *au, Bammesberger points out that we do not know the intermediate stages and continues: ‘On the whole it seems preferable to accept the view that the monophthongization of Gmc. *au is a typically Frisian feature.’ Bammesberger (1990a, 459) assumes that skanomodu is a bahuvrihi compound with an adjectival stem skano- and the second part -modu. Since -mod- can be brought together with the nominal stem Gmc. *mōda(> OE neutr. mōd; OHG muot), skanomodu may have meant ‘having a skano mind’ according to Bammesberger.

3.1 The -u Evidence One feature of the inscription that some scholars, as already Düwel/Tempel (1970, 383), view as pointing to a Frisian origin is its final -u, called “rätselhaft” [(enigmatic)] by Hofmann (1976, 73, cf. the above quotation).19 Wolfgang Krause (1968, § 34 n.) considers it possible (“vielleicht” [perhaps]) that it represents a reduced reflex [ə] of the Proto-Germanic nominative masculine singular suffix -az. He refers to the “Personennamen Ādugīslu (urgerm. *Audagīslaz)” on the weaving-slay of Westeremden. But he does not seem to consider this a certain example, because he adds “falls N. Sg. vorliegt” [if the form is nominative singular]. And indeed, Ray Page (1999, 100), transliterating adujislu:}mejisuhldu, comments, “though the noun endings cause concern, the two personal names are clear, and a reading ‘for Adugisl and 20 Gisuhild’ or something like that is likely”, implying that the final -u in adujislu marks an oblique case, presumably a dative (‘for ...’). Krause goes on to cite the forms kabu (for kambu) on the Frisian comb case from Oostum and kobu (for kombu) on the comb case from Toornwerd, referring to Düwel/Tempel (1970, 369) (which was not yet published at the time). Here, too, the nominal case is not certain, an oblique case appearing entirely likely; nor is the stem formation clear: Old High German had a weak (n-stem) noun kambo ‘comb’ in addition to the strong (a-stem) kamb, and this word may have been more wide-spread in West Germanic

19 The combination of the +au > ā argument and the final -u argument are regularly used in favor of claiming a Frisian origin of the skanomodu inscription. As a typical formulation one may cite Quak (1990, 359): “Diese gilt meistens als friesisch, da man den ersten Teil dieses Eigennamens (?) mit a[lt]fri[esisch] *skān- mit ā aus älterem *au identifizierte. Im Altenglischen müßte das *sceanlauten. Auch findet man in dieser Inschrift die Endung -u.” (‘This inscription usually counts as Frisian because the first part of this proper name (?) was identified with Old Frisian *skān-, with ā from older *au. In Old English this would have yielded *scean-. Furthermore, the ending -u is found in this inscription.’) 20 The bind-rune section me is usually amended to read meþ ‘with’.

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

205

at this early date. In any event, the assumption of a change +-az > -u does not seem warranted. Nielsen (2000, 92) reckons with “the shift *-ō to -u as reflected in the -u suffixes of the personal names skanomodu (London gold solidus), if seen as a nsf. [nominative singular feminine] ō-st[em] suffix as I am now inclined to interpret it (Nielsen 1996, 128‒129), and gisuh(i)ldu (Westeremden A), presumably a dsf. [dative singular feminine] (j)ō-st. form, cf. dsf. O[ld] S[axon] geƀu, O[ld] H[igh] G[erman] gebu ‘gift’.” But a nominative seems out of the question, because what meaning would it carry on a coin, or rather piece of jewellery? Here I differ from Elmar Seebold, who writes in his e-mail of 20 September 2014: “Vorausgesetzt, es ist ein Name – dann muss der im Nominativ Singular stehen (abgesehen von Besitzerinschriften, bei denen der Genetiv zu erwarten wäre).” [‘Presupposing it is a name – in which case it has to be a nominative singular (except for possessor inscriptions for which the genitive is to be expected).’] A name on a coin remade into a piece of jewellery, possibly an amulet – what can it express? The owner if it is a genitive. The recipient or dedicatee if it is a dative. But what if it is a nominative? If Nielsen’s idea of a singular feminine ō-stem skanomod- is accepted, then skanomodu, too, should be understood as a dative singular, ‘for Skanomod’. Another inscription considered Frisian, that on the Schweindorf solidus, which is rendered as weladu in nearly all earlier accounts I have seen, is now read welad by Tineke Looijenga (2013, 431).21 In Figure 1 above showing the skanomodu solidus, the final -u, if an -u it is, is much less distinctly inscribed than the initial eight runes. Perhaps then the significance of the “Frisian -u” has been overrated all along; -u may eventually prove to be useless as a criterion for “Frisianness”. Bammesberger (1990a, 462) long ago arrived at the conclusion that it “certainly can be accounted for within the linguistic system of a West Germanic language”, but “hardly yields any direct information as to the underlying dialect (Old English or Old Frisian)”.22

3.2 The skan-Evidence The derivation of the skan(o)- part of the skanomodu inscription from Gmc. + skaun(i)- can be found in all recent accounts and seems to be widely accepted.23 21 Nedoma (2014, note 34) disagrees with Looijenga’s reading. The excellent scan accompanying Looijenga (2013) to me suggests an initial þ (þelad) rather than w. Page (1995, 150) writes “þeladu or weladu”, as Tineke Looijenga kindly points out in an e-letter of 16 September 2014. 22 Also cited in Waxenberger (2017, 107). 23 To cite just one instance: Nielsen (2000, 117) writes: “The earliest item [among “several inscriptions belonging to the so-called Frisian runic corpus [...] which evidence the shift of Gmc. au to ā] is the sixth-century London gold solidus, which bears the runic legend skanomodu skānomōdu < skaunamōd- ‘Schönmut’.” Stephen Laker, in an e-letter of 16 June 2014, draws attention to the fact that Frisian has another adjective with the base skān-, which seems to have escaped the attention of runologists dealing with skanomodu. This is Gmc. +skauna- ‘schräg, schief (‘slanted, tilted,

206

Theo Vennemann

Rather than give a list of individual references I will simply (1) quote a rather cautious statement from Page’s Introduction, then (2) a recent unhedged assertion, and finally (3) the most recent authoritative summary:

(1)

The runes have usually been interpreted as a personal name, and I still think this explanation the most likely. The first element could derive from Gmc *skaun- which gives OE Scen- in the rare name Scenuulf but is more common in Continental Germanic names,24 and the second element looks like the common -mod. There are two linguistic features that favour Frisia as the region of striking: the first element shows the Germanic diphthong au developed to a simple vowel represented by a (contrasting with the Old English diphthong ea), and the ending is the unstressed vowel -u found in other Frisian runic texts. Had the coin not appeared in an English king’s collection, and been first published at a time when few Frisian runes were known, I do not think it would have been taken as Anglo-Saxon. (Page 1999, 185).25

(2)

Die friesische runenepigraphische Überlieferung hält [...] einen sicheren [...] Beleg26 für /ā/ < /au/ bereit: das Vorderglied von ska2no-modu [...] gehört zu urgerm. *skauni(ja)- Adj. ‘schön’ (Nedoma 2014, 346).

askew’)’, Old West Frisian skān- ‘schräg (‘slanted’)’ in skānwīs ‘in schräger Weise (‘in a slanting manner’)’, with reflexes of similar meanings in Early New West Frisian (schean) and New West Frisian (skean), cf. Faltings (2010, s. v. skauna- [467]), which is not in Heidermanns 1993. Old West Frisian skānwīs ‘in schräger Weise (‘in a slanting manner’)’ may survive as a loanword in English (Late Middle English) askance, which has no generally accepted etymology (but cf. OED: s. v.). According to Laker, names need not always be positive sounding. Indeed, at least in certain organisations, such as bands and gangs of various kinds, both adopted and bestowed names may be facetious or have a mean or threatening ring. Assuming this +skaun(a)- as the base for the first constituent of skanomodu would yield the same result for the corpus assignment of the inscription as does the traditional assumption of +skaun(i)-; it may even have the advantage of being less troubled by the connecting vowel -o-. However, semantic reasons disfavor this alternative if the interpretation of Skanomod as a female name assumed in this paper is correct. In any event, the Frisianness of skanomodu remains doubtful on account of the possible alternative derivation proposed below. 24 The only name beginning with the base skan- in Reichert’s Lexikon der altgermanischen Namen (names recorded no later than a. 600, regionally a. 700) is Scanomod (Reichert 1987, 591), with reference to Düwel (1983, 45). Förstemann 1900 lists a few names with +skaun- (s. v. scaunja), none in combination with +mōd-; and with +skain-, or rather +skīn- or +skin-, only Scinus, “der einzige zu got. skeinan, ahd. ags. skīnan splendere, lucere gehörige n[ame]” (‘the only name belonging to Gothic skeinan, Anglo-Saxon and Old High German skīnan’) (1307‒1308, between scilf and scipa). This is not much, but it shows at least that Förstemann considered names integrating the shine root onomastically warranted. Förstemann (1900, s. v. MODA) also offers a list of 137 male and female names with -mod- as second constituent. 25 Also cited approvingly (from the first edition) in Bremmer (1981, 55). 26 On page 348, Nedoma says, modalising, “das Vorderglied ist wohl zu urgerm. *skauni(ja)- Adj. ‘schön’ zu stellen” (‘the first constituent probably has to be associated with Proto-Germanic *skauni(ja)- adj. ‘beautiful’’). On page 357, he says “sicher bzw. sehr wahrscheinlich” (‘secure or very probable’).

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

207

‘The Frisian rune-epigraphic tradition holds [...]one secure [...] piece of evidence27 for /ā/ < /au/: the first constituent of ska2no-modu [...] belongs to Proto-Germanic *skauni(ja)- adj. ‘beautiful’.’

(3) Zu dem früh bezeugten m[ännlichen] R[uf]N[amen] skanomodu in einer wohl friesischen Runeninschrift des 6. Jahrhunderts, dessen Bestimmungswort nach allgemeiner Ansicht das Namenelement germ. *skaun- ‘schön’ enthält, vgl. Bammesberger 1990a [...], 457 ff. mit weiterführender Literatur (Faltings 2010, s. v. skauni-). ‘For the early-attested male personal name skanomodu in a probably Frisian runic inscription of the 6th century, whose specifying constituent is generally seen to contain the name element Gmc. *skaun- ‘beautiful’ [emphasis added], cf. Bammesberger 1990a [...], 457 ff. with further literature.’

Alfred Bammesberger likewise makes the classification of the inscription as Frisian dependent on this most common interpretation: Wenn man aber die derzeit wohl geläufigste Deutung der Inschrift als Name annimmt, dann ist eine Deutung des Namens nach friesische[m] Sprachmaterial durchaus wahrscheinlich. Der Name kann als Kompositum in skano-modu aufgeteilt werden. Im Zweitbestandteil wird das Substantiv für ‘Sinn’ vorliegen, im ersten Bestandteil kann man ein Adjektiv vermuten, das den Diphthong /au/ als Wurzelvokal hat. Für das Bahuvrihi-Kompositum skāno-mōdu ergibt sich dann eine mögliche Vorform *skauna-mōdu-. Die Entwicklung von /au/ > ā entspricht der friesischen Lautgeschichte. Bemerkenswert ist die Tatsache, dass für die Darstellung von ā die ‘geneuerte’ Rune ᚪ verwendet wurde, dagegen hat ᛟ den Lautwert ō (und nicht œ)28 (Bammesberger 2006, 186). ‘If however the presently most current interpretation of the inscription as a name is accepted, then the interpretation of the name with Frisian language material is indeed probable. As a compound the name can be divided as skano-modu. The second constituent will be the substantive for ‘mind’, the first constituent may be supposed to be an adjective with the diphthong /au/ as root vowel. This bahuvrihi compound skāno-mōdu then leads to an earlier form *skauna-mōdu-. The development /au/ > ā is in harmony with Frisian phonetic history. It is a remarkable fact that the ‘innovated’ rune ᚪ was used to represent ā; by contrast ᛟ has the sound value ō (and not œ).’29

27 Cf. the preceding note. 28 More precisely, runologists have assigned ᛟ the sound value ō rather than œ in order to make the desired interpretation with +(skān)-o- and +mōd- possible. The regular Anglo-Frisian rune for ō is ᚩ, the o-rune. If, however, the ᛟ rune were granted its ordinary Anglo-Frisian sound value, the inscription would have to be transliterated as skanœmœdu; the second constituent would then have to be reconstructed as a -j- stem, perhaps +mōd-j-ō, with loss of the -j-, and the connecting vowel rune as anticipatorily fronted – all very unwelcome consequences, quite apart from the question whether palatal umlaut was phonemic at the time the inscription was made. See also note 44 below. 29 Cf. the preceding note.

208

Theo Vennemann

4 An Alternative Etymology for skanAs the Frisianness of skanomodu primarily depends on the identification of skanwith Gmc. +skaun-, I asked at the conference whether the alternative Gmc. +skain-, ablauted base derived from the class I strong verb Gmc. +skīnan ‘to shine’, had been considered, because the monophthongisation of Gmc. +ai into ā occurred both in Anglo-Saxon and Frisian. Reconstructing +skain-o+mōd- ‘(having a) shining mind, (having) visible [or brilliant] courage’ on the tradional model +skaun-o-mōd- ‘(having a) beautiful mind’ would thus seem to make the exclusive assignment of the skanomodu inscription to the Frisian corpus questionable, +skain- > +skān- (skan-) in the given context being the regular development both in Anglo-Saxon and in Frisian. In the discussion Gaby Waxenberger drew the conclusion that this might indeed reopen the question of the corpus assignment of skanomodu. Back home, I therefore pursued the question whether reconstructing a base + skain- paralleling +skaun- is legitimate, with the result that it is, namely with the same degree of plausibility (or lack thereof) as in the traditional reconstruction. The adjectival base assumed for skan-o- in the tradional bahuvrihi reconstruction is that in +skauni-,30 i- or ja-stem adjective +skaun-i- (Goth. +skauns or +skauneis,31 OE scīene, OFris skēne, OS skōni, OHG scōni).32 Likewise an adjectival base +skain- is reconstructible from an i- or ja-stem adjective +skain-i- ‘shiny, visible’ in OHG ursceini ‘becoming visible, apparent, manifest’,33 attested in a gloss as ursceiniist translating appareat. Heidermanns (1993, s. v. -skaini*-) declares the word a “Verbaladj[ektiv] zu dem starken Verb *skeina- [i.e., +skīnan] ‘scheinen’” [a deverbal adjective of the strong verb *skeina- [i.e., +skīnan] ‘to shine’] and refers to the prefix verb OHG irscīnan ‘erscheinen [to appear]’.34 Meid (1967, § 70.2) writes about adjectives 30 Whether this adjectival stem is originally to be analysed as +skau-ni- (with a verbal base, the weak verb +skauwō(ja)-) or as +skaun-i- (as in the historical Germanic languages) is a matter of dispute; cf. Heidermanns (1993, s. v.); Faltings (2010, s. v.) It is irrelevant for the present purpose. 31 Attested as skaun-j-ai ‘beautiful (nom. plur. masc.)’, also in the dat. sing. neut. adjective form ibnaskaunjamma ‘equal in form’ (Lehmann 1986, s. v. *skauns). 32 Cf. Orel (2003, s. v.; Heidermanns 1993, s. v.). The adjective stem +skauni- (viz. +skaun-i-) is flanked by a number of further derivatives which in part likewise indicate the i-stem: a feminine noun + skaunīn (Goth. +guda-skaunei ‘form of god’, only attested as dat. sing. gudaskaunein, cf. Lehmann 1986, s. v. *skauns); a causative verb +skaunjan- (OHG scōnen ‘beautify, adorn’); an ō-stem verb (OFris. skēnia ‘to clean’, OHG giscōnnōn ‘to ornate, polish’); a feminine noun +skaunethō- (Old Low Franconian skōnitha ‘beauty’, OHG unscōnida ‘deformity’); and an ǣ-stem verb (OFris. skēnia ‘to become beautiful’), cf. Heidermanns (1993, s. v. skauni-). The base +skaun- of +skaun-i- is clearly visible also in the OHG adverb scōnō which is continued in Modern German as umlautless schon ‘already’ alongside umlauted schön ‘beautiful’, and in the OHG comparative scōnōro/scōnero (alongside scōniro, whence umlauted Modern German schöner ‘more beautiful’). Umlautless forms also occur in OFris. but are there considered borrowings from Low German (Faltings 2010, s. v. skauni-). 33 With prefix ur- < Gmc. +uz-. 34 Seebold (1970, s. v. *skei-na-), in his list of derivates of this class I verb +skīnan-, regards the OHG gloss word ursceini ‘sichtbar’ [‘visible’] as the only attestation for an adjective skain-i- (in

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

209

derived from root forms of strong verbs: “Bildungen von Wurzeln der 4. und 5. Ablautklasse haben meist Dehnstufe, solche von diphthongischen Wurzeln Schwundstufe.” (‘Forms derived from roots of ablaut classes IV and V usually have lengthened grade, forms derived from diphthongal roots, zero-grade.’)35 Meid’s examples are: OE bryc-e ‘breakable’ (brecan class IV ‘to break’); OE ge-myn-e ‘remembering, mindful of’ (munan pret.-pres. class IV ‘to be mindful of, remember’); OE ȳđ-fynd-e ‘easily findable’ (findan class IIIa ‘to find’); and, in the two diphthongal classes, Goth. un-nut-s ‘useless’ (niutan class II ‘attain’); OE swice ‘deceitful, fallacious, deceptive’ (swīcan class I ‘to fail, turn traitor, deceive’).36 Hence for the class I verb scīnan ‘to appear, shine, be resplendent’ an i-stem adjective would most likely be based on the zero-grade root form scin- (< +skin-) rather than on the full-grade (agrade, abtönung) form scān- (< +skain-). However, Mailhammer (2008) has shown that any ablaut grade can occur in all kinds of nominal stem formation, especially when based on strong verbs, so that an i-stem based on an a-grade base of a strong verb is absolutely natural; cf. also his list of “Ablaut grades and types of stem formation in the Proto-Germanic noun” (Mailhammer 2008, 286‒288). He writes with reference to my earlier concerns about the probability of +skain-i-: D. h. allerdings, dass beide Bildungen gleich wahrscheinlich sind, tut also Ihrem Argument nichts, insofern muss man sich m. E. um +skain- keine Sorgen machen, das ist eine relativ normale Bildung. Es kommen immer nicht-kanonische Ablautbildungen vor und die insbesondere von (starken) Verben, weil die durch Analogiebildungen37 von verschiedenen Stufen im Paradigma abgeleitet werden können. D. h. dass letztlich +skain- genauso (un)wahrscheinlich ist wie +skaun-. (Robert Mailhammer in an e-letter of 17 June 2014). ‘This means, to be sure, that both formations [+skin-i- and +skain-i-] are equally probable, so that your argument is not affected; in that regard there is in my view no need to worry about + skain-: it is a relatively normal formation. Non-canonical ablaut formations occur again and again, in particular those from strong verbs, because they can be derived from different grades in the paradigm by analogical formations.38 That means in the final analysis that +skain- is just as (im)probable as +skaun-.’

boldface and without asterisk), thereby suggesting that the Proto-Germanic adjective could exist without the prefix. 35 Cf. also Bammesberger (1990b, 260). 36 Goth. niutan and OE swīcan are both diphthongal verbs in a Germanic perspective: Goth. iu < Gmc. +eu; OE ī < Gmc. +ī < pre-Gmc. +ei. 37 Namely by a special type of analogy that does not proceed via formal similarities but via categorial identities: “This process is called systemic analogy here, because the shape of the derivational base is calculated with the help of the inflectional system, in this case the paradigm of the strong verb used as a derivational base. It works in a similar way to the more familiar proportional analogy, except for the fact that the categories of the bases and formations that are set in relation to each other match and not their phonological shapes” (Mailhammer 2008, 296 f.). Mailhammer (2008, 297, note 40) attributes the identification of this type of analogy to Becker (1990, 47) who includes, in his typological overview of analogical formations, a type “Sekundärer Bezug einer Ableitung auf andere Basen” (‘Secondary relation of a derivative to other bases’). 38 Cf. the preceding note.

210

Theo Vennemann

Bammesberger (1990b, 133), too, writes: “Neben der bei Verbalwurzeln in Klassen I‒ IV vorherrschenden Schwundstufe findet sich vor dem Suffix -i- auch nicht selten eine o-Stufe [die germanische a-Stufe, T. V.].” (‘Alongside the dominant zero-grade, verbal roots of classes I‒IV not infrequently also show o-grade [Germanic a-grade] before the -i- suffix.’) An example with class II ablaut is Gmc. +rauk-i- (OE rēc, OHG rouh etc.) ‘smoke’ from *reuk-an- ‘to (emit) smoke’, OE rēocan (ModE to reek), OHG riohhan (Gm. riechen), etc. But Bammesberger has no examples of a-grade i-stems from class I verbs (cf. Bammesberger 1990b, 133). Yet it would be a mistake to infer from this that this type of formation was less probable, because the lack of attestation does not really prove anything.39

5 Conclusion: Anglo-Saxon or Frisian I conclude that a derivation of skan-o- in skanomodu from an adjectival base + skain- of +skain-i- ‘shining’40 is formally and semantically no less plausible than the traditional derivation from an adjectival base +skaun- of +skaun-i- ‘beautiful’.41 39 Robert Mailhammer in an e-mail of 1 July 2014. 40 Mailhammer writes in his e-mail of 17 June 2014: “Eine Möglichkeit [die Angemessenheit von + skain-i- zu prüfen] wäre noch die Semantik, indem es nämlich sowohl +skain- als auch +skein- [i.e. + skīn-] als Adj[ektiv] gibt. Letzteres bedeutet Heidermanns zufolge ʻanscheinend, offenbarʼ und + skain- hat eher mit ʻerscheinenʼ zu tun, wenn ich Heidermanns richtig lese.” (‘One possibility [to test the suitability of +skain-i-] could be the semantics, considering that both +skain- and +skeinexist as adjectives. The latter means ‘apparent, evident’ according to Heidermanns [(1993, s. v. skeina-)] and +skain- is closer to ‘appear’, if I read Heidermanns [(1993, s. v. -skaini-)] correctly.’) I think that the somewhat narrower meaning assigned by Heidermanns to +skain-i- results from its more restricted attestation, namely solely in a compound ursceini whose meaning is determined by OHG urscīnan ‘to appear’ and the Latin verb apparēre ‘to appear, be visible’ which it translates (in its verbal construction ursceini ist). Seebold (1970, s. v. *skei-na-) glosses +skein-i- as ‘sichtbar [visible], offenbar [evident]’ and +skain-i- as ‘sichtbar’. From this I cannot infer any significant semantic difference between the two adjectives. The narrow definitions given for both adjectives are a consequence of their limited occurrence in texts of the old Germanic languages. The full range of meanings-in-context of deverbal +skain-i- may have resembled that of the basic strong verb +skīnan- and thus may have included ‘shiny’ alongside ‘being visible’, just as the meaning of the verb includes ‘to shine’ alonside ‘to appear’. That no i-stem noun +skain-i-z ‘shine’ is attested alongside the adjective +skain-i- whereas there is West-Gmc. +skein-a-z ‘light, shine [Gm. Licht, Glanz]’ (cf. Seebold 1970, s. v. +skei-na-) alongside the adjective +skein-a- may be nothing more than a consequence of our limited corpora. Since Gmc. +mōd-a-z also meant ‘courage’, even +skein-o-mōd- read as ‘(having) visible courage’ would transport an utterly positive sense and would thus be suited as a Germanic name, at least for a male. 41 Seebold (1970, s. v.) introduces the strong verb Goth. skeinan, ON skína, OS OHG skīnan, OE scīnan, OFris. skīna (in Old Frisian and dialectally in some of the other branches with a weak preterite) ‘to shine’ with the lemma heading “*SKEI-NA- (skein-a-) ‘scheinen’”. The division skei-na- is offered solely for etymological reasons. The parenthesis indicates that from a Germanic standpoint the division is skein-a-. Cf. notes 24, 34 above.

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

211

For +skaun-i- (or +skaun-ja-) we have the derivatives mentioned above. For +skain-iwe have the ursceini gloss and in addition the independently existing strong verb + skīnan with its ablauted root form +skain-. All the arguments that can be raised against the +skainomōd- reconstruction based on the +skain- of +skain-i-, e.g., the termination -o- of the first constituent of the compound 42, are no more damaging than the same arguments raised against the +skaunomōd- reconstruction based on the +skaun- of +skaun-i-. Even if the stem-final -i- is considered to have played a role in the first constituent (I do not think that such an influence need be assumed), this role would have been the same in both reconstructions; and since the skanomodu inscription is older than i-umlaut in both Anglo-Saxon and Frisian, Gmc. +ai would yield unfronted ā in both languages as a result of monophthongisation of +ai > ā, which was probably a shared Anglo-Frisian change.43 But these considerations are in my view beside the point: The connecting vowel -o- cannot possibly derive from -i- at this early date but most likely reflects -a-.44 Hence the skano- part of skano-

42 Robert Mailhammer, in his email of 17 June 2014, assures me that as a fugenelement [linking element] this -o- is nothing to worry about. “Das stammbildende Suffix war im Germ[anischen] schon längst Teil der Flexionsendung.” (‘The stem-forming suffix had in Germanic long since become part of the flexional ending.’) – Gaby Waxenberger adds: “For -o (< Gmc. *-a-) in skano(< *skauna-) Bammesberger (1990[a], 462[f.]) takes the process of ‘rounding medial -a- in labial surroundings’ into consideration (Waxenberger forthc., chapter 2 no. 71).” Indeed, Bammesberger writes: “OHG berumēs ‘we carry’ goes back to Gmc. *berame- and exhibits rounding of -a- > -obefore a nasal.” Even if this is so (Braune and Reiffenstein (2004, § 307a) write, “Die Formen mit -u- scheinen die ursprünglichen zu sein” (‘The forms with -u- seem to be the original ones’)), it may be doubted that the same rounding would have taken place in the putative *skauna+mōdu-, i.e., across a compound boundary. 43 This only applies in case the skān- in skāno- is the presumed i-stem obscured by the -o-. Before -o- the reflex of Gmc. +ai would be ā in both Anglo-Saxon and Old Frisian anyway (cf. de Vaan 2011 for Frisian). Nielsen argues that monophthongisation precedes i-umlaut: “The skanomodu legend thus gives indirect evidence of the chronological precedence of monophthongization over i-mutation” (Nielsen 1993, 84). – Gaby Waxenberger adds: “At that time i-umlaut must have been in its allophonic phase and therefore the rune āc /a:/ would also have denoted the allophone [æ] (cf. Waxenberger forthc., chapter 3).” 44 Elmar Seebold writes in his e-letter of 20 September 2014: “In der Zeit von Ende 5./Anfang 6. J[ahr]h[underts], mit der offensichtlich mit Recht gerechnet wird, ist es ausgeschlossen, dass ein stammbildendes -i- als -o- erscheint. Deshalb sind sowohl *skaini- wie auch *skauni- als Grundlage ausgeschlossen. ... Die Annahme eines mittleren Elements o ... ist von Ray Page höflich, aber bestimmt (und sicher im Ergebnis richtig) zurückgewiesen worden. Da ein u-Stamm kaum in Frage kommt, wird das Vorderglied wohl ein a-Stamm sein ([...] ich nenne das den Typ Dagobert).” (‘For about the end of the 5th/the beginning of the sixth century, which is the time obviously and with justification reckoned with, it is impossible for a stem-forming -i- to appear as -o-. Therefore both *skaini- and *skauni- are excluded as bases. ... The assumption of a middle o has been rejected by Ray Page politely but firmly (and certainly in the final instance correct in its result). Since a u-stem hardly comes into question, the first constituent is likely to be an a-stem ([...] I call this the Dagobert type).’)

212

Theo Vennemann

modu offers no criterion for deciding whether the inscription is Frisian or AngloSaxon.45 The other argument commonly proposed in favor of a Frisian origin of the inscription, namely that the termination -u derives from the masculine nominative singular a-stem termination +-az, has met with reservations in the literature, where an interpretation as a feminine dative singular ō-stem termination has been suggested instead. Even though tracing this termination in the historical grammar of English is not a straightforward matter,46 a feminine dative on a piece of jewellery seems to me a priori more probable for pragmatic reasons than a masculine nominative. In short, taking all aspects of the skanomodu solidus into account, I propose that its runic inscription reflects a personal ō-stem name in the dative, +skaina+mōd-ō ‘for Skainamōd (fem.)’, as an alternative to +skaun-a+mōd-ō ‘for Skauna-

45 Stephen Laker advised me in an e-mail of 17 June 2014 to add a footnote concerning the monophthongisation of Gmc. +ai in Old English: “There is the early raïhan ‘roe/Reh’ inscription on deer bone from Caistor[-by-Norwich] that has aï for [ai] < PGmc. *ai. Possibly somebody could argue that it should therefore be skainomodu as it is an early inscription. This seems a weak argument though, but possibly it would be worth mentioning in a footnote.” The raïhan incription on a roedeer astragalus found in a cremation urn is dated ca. 400 CE and thus antedates the skanomodu inscription by approximately a century (Hines 1990, 442; cf. Derolez 1990, 415). Hines (1990, 442) says the object is “regarded as a foreign intruder in Anglo-Saxon England”, and Derolez (1990, 415 f.) cites several authorities declaring it non-Anglo-Saxon, e.g. Odenstedt (1984, 124, note 19): “The Caistor-by-Norwich astragalus is too early [...] to be English – it must have been brought from the Continent by the Anglo-Saxon invaders.” It seems therefore that no argument can be derived from the raïhan inscription concerning the monophthongisation of Gmc. +ai in Anglo-Saxon. However, Bammesberger (2006, 172, 178) treats the inscription as Anglo-Frisian, without assigning it a date and hence without any reference to the time problem. Without discussion of this discrepancy Gaby Waxenberger sent me the following addition to this note: “The monophthongization of */ai/ > /a:/ had obviously not been completed at ca. 425‒475 as -ᚨᛇ- in the word rᚨᛇhᚨn shows. Bammesberger (2006, 178) suggests either [åi] or [åe] for the combination -ᚨᛇ-. Although probably Frisian, the skanomodu solidus (ca. 575‒610) sheds light on the terminus ante quem of the creation of the rune a ᚪ. Consequently, the rune a ᚪ must have been designed between ca. 425‒475 (= Caistor-byNorwich Astragalus where it is not yet attested) and ca. 575‒610 (skanomodu solidus). Since, however, there is reason to believe that the new rune a ᚪ was in all probability contemporaneous with the new rune o ᚩ, its emergence must have been around 500, the latest possible date for the Undley (450‒500) inscription, on which the new rune o ᚩ appears. (Waxenberger forthc., chapter 3).” 46 The dative singular termination of ō-stem feminine nouns in West Germanic and in Old Norse was -u: OS OHG gebu, ON giof (< +gebu) ‘gift (dat.)’ (Prokosch 1939, § 81.a). Old English giefe deviates. On the extension of -u cf. Campbell (1959, § 588.6‒7) and Brunner (1965, 255.3). ON, besides showing a reflex of the -u in the labial umlaut of giof, shows -u in the dative kerlingu of kerling ‘old woman’: “-u in the accusative and dative singular have, besides kerling, many feminine personal names, e.g. Áslaug, Bergljót, Katrín, Guðrún, Rannveig, Ingileif, and other similar” (Stefán Einarsson 1945, 38). A somewhat radical solution for the final -u is suggested by two formations similar to skano-mod-u, OE ofermettu and ofermēdu ‘pride’ (Campbell 1959, § 589.6‒7), ēaðmēttu ‘Demut [humility]’, ofermēttu ‘Übermut [high spirits, pride]’, cf. OHG *ôtmuotida, ubarmuotida (Brunner 1965, 254.3). Here the final -u is the reflex of the final -ō of the suffix +-iþō whose -i- was syncopated in Old English, with assimilation of the dentals and generalisation of the -u.

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

213

mōd (fem.)’47 (and also to reconstructions deriving the final -u from +-az); and I conclude that the likelihood of the inscription’s being Old English is at least as great as that of its being Frisian.

Postscriptum 1: The Location of the Coin, Again The assumption of the skanomodu inscription being Frisian would gain strength if there were some indication that the coin was brought to England from the Continent. Here is what the specialists say: Der Solidus war ursprünglich im Besitz König Georgs III. König Georg IV. schenkte ihn nach seiner Thronbesteigung 1820 dem Britischen Museum. (Schneider 1967, 55, with reference to Stephens vol. II, 879) ‘The solidus was originally in the possession of King George III. King George IV, after his accession to the throne in 1820, gave it to the British Museum as a gift.’

Karl Schneider mentions the possibility that the solidus was found on the Continent and acquired by the monarch (of the House of Hanover); but no evidence exists for this supposition. Ray Page writes: The skanomodu solidus was part of George III’s collection, given to the British Museum in 1825, but we do not know how it came into his possession, whether by inheritance or gift, from his English or continental territories or farther afield. The manuscript catalogue, compiled in 1771 and checked up to 1814, gives no details of it, nor has any information been traced in the Queen's Archives, Windsor Castle. Professor A. Aspinall tells me that he has found no mention of the solidus in George III’s correspondence (Page 1968, 12, n. 3 [cf. 1995, 145, n. 3], also cited in Bammesberger 1990a, 464).

Alfred Bammesberger’s own last words on the question in his 1990 article (1990a, 464) are the following:

47 The final nominative -a has reached only few names with -mod- as second constituent; most of them show no mark in the nominative. Förstemann in his list of -mod- names (cf. note 24 above) has the following names with final -a: Angelmoda, Adalmoda, Hathumoda, Hildimoda, Raguemoda, Redmoda, Wandelmoda, Wolfmota. It is not clear, however, which of these may be Latinisations of German suffixless nominatives. That names in -mod- may be both masculines and feminines is evident from Förstemann’s intoduction to his list: “Als grundwort gehört -môd eben so gern masculinen wie femininen an ...; einige Formen, die sich sowol männlich als weiblich nachweisen lassen, bezeichne ich mit mf. Mein Verzeichnis umfasst ... 137 namen.” (‘As a second (or head) constituent -môd belongs with equal ease to the masculines as well as the feminines. ... I mark a number of forms which can be shown to occur with either gender as mf. My catalog contains 137 names.’) Thus the gender of Skanomod- is not a priori clear, it may be feminine as well as masculine. I have opted for a feminine name, because that permits an interpretation of the -u termination as reflex of a dative case marker.

214

Theo Vennemann

With regard to the provenance of the skanomodu-solidus practically nothing seems ascertainable. Brown 1915: 69 reports the British Museum Catalogue according to which the solidus ‘is believed to have been found in this country’.

Thus, not only is the linguistic evidence for a Frisian origin of the skanomodu solidus inconclusive; also the fact that within its known history the coin has never been outside England does not make a Frisian origin appear more probable than an English one.

Postscriptum 2: An Earlier Class I Ablaut Etymology The Abtönung (a-grade) form +skain- occurs twice in the verbal paradigm of +skīnan, in the preterite indicative singular and in the causative.48 I am aware of one attempt to arrive at a scan- < +skain- analysis on this basis, viz. in Schneider 1967. The essential point is summarised by Page (1995, 153): Scholars have commonly interpreted the legend skanomodu as a personal name, but to this Schneider objects because of the ‘namenmorphologischen und auch sprachhistorisch-runologischen Schwierigkeiten’ it presents. He prefers to read three words, scan o modu, the first one 3rd sg. pret. ind. of scinan [i.e. scīnan], ‘shine, glitter’, the second o [to be understood as ō], an emphatic form related to OE a [i.e. ā] (Gothic aiw), ‘ever’, the third instr. sg. of OE mod [i.e. mōd], ‘courage, might’: ‘Er hat immer durch Mut (Macht) geglänzt’ [He always shone by courage (might).]. [...] This interpretation needs a good deal of justification. The form o, ‘ever’, Schneider compares with the Old Frisian emphatic negative no ‘never’, but it is not clear if he regards the skanomodu example as an Old English equivalent otherwise unrecorded, or as essentially Frisian – he calls it Urfries[isch] [Proto-Frisian]. [...] Yet there does seem a lot of special pleading required by Schneider’s interpretation of skanomodu.

Alfred Bammesberger (1990a, 459), too, finds Schneider’s proposal unacceptable: Linguistically the most difficult part is perhaps Schneider's identification of ‹o› with OE ā ‘always’,49 because at the early date which Schneider posits for the inscription50 the rounding of ā (from Gmc. ai) > ō would be totally unexpected. [...] More troublesome is the alleged verbal form scān (belonging to scīnan ‘shine’), since a preterite does not regularly have presential reference, i.e. scān ought to mean ‘he shone’. But saying ‘he always shone’ would seem to fit a funerary inscription rather than an amulet.51

48 Cf. Seebold (1970, s. v. +skei-na-). The causative +skain-eja- is only attested in OHG: sceinen, gisceinen ‘zeigen’. Among the Modern German translations provided for the 15 occurrences of gisceinen in Köbler (1994, s. v.) is ‘glänzen (‘to shine, gleam, sparkle’)’. 49 “OE ā ‘always’ goes back to Gmc. *aiw-, cf. Go. aiw (acc. sg. of aiws)” (Bammesberger 1990a, 459 n. 9). 50 “Schneider 1967: 59 dates the skanomodu-solidus ‘ca. 500‒525 n. Chr.’” (Bammesberger 1990a, 459 n. 9). 51 “Schneider insists on the solidus having functioned as an amulet. One could object generally that solidi do not seem to carry clauses of the type envisaged, but clearly this objection carries

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

215

I agree with Page and Bammesberger that Schneider’s reading of skanomodu is hard to accept as a coin inscription, even if the solidus was also worn as an amulet or ornament as suggested by the eyelet. The reading supposes the inscription to be sentential 52, which would be unusual in itself. And there is no possible explanation for the fact that with Schneider’s reading the subject of the sentence is unexpressed. Page rejects the idea that a divine name may have been omitted for taboo reasons, and adds: “A clue ought to be given by other magical texts within a similar context or date range, for, if Schneider’s interpretation is correct, one or other of these should bear a structure something like scan o modu” (Page 1995, 153). Nevertheless, Schneider’s proposal deserves mention as the only attempt so far, even if an unsuccessful one, to relate the a of skan- in skanomodu to the +ai of class I ablaut and by implication to relativise its corpus membership.

Acknowledgements I thank Stephen Laker (Kyushu University, Fukuoka), Robert Mailhammer (University of Western Sydney), and Roland Schuhmann (Jena and Leipzig) for helpful comments on a draft of this paper. I also would like to thank Gaby Waxenberger (University of Munich) for her friendly reaction to my proposal concerning skanomodu and for implying (in an e-mail of 12 June 2014) that I publish it. She also sent me, in an e-mail of 10 July 2014, the added notes that have appeared above. Tineke Looijenga (University of Groningen), in a friendly e-letter of 23 June 2014, encouragingly wrote: “Now there is still less need for the supposition of the existence of a separate Frisian runic corpus. The Frisian and Anglo-Saxon runic corpus is a common phenomenon which I warmly advocate.” Finally, I thank the following colleagues for their generous comments on earlier versions of this article in e-letters: Alfred Bammesberger (Catholic University of Eichstätt-Ingolstadt, 25 August 2014), Klaus Düwel (University of Göttingen, 19 September 2014), and Elmar Seebold (University of Munich, 20 September 2014). Prof. Bammesberger: Ich habe jetzt alles genau gelesen – keinerlei Einwände. *skain- ist sicherlich möglich [als Alternative zu *skaun- in skanomodu]. Bei Seebold findet sich genug Vergleichsmaterial, so dass rein theoretisch keinerlei Einwand besteht. Freilich: *skain- und *skaun- halten sich ungefähr die Waage, meine ich.53

little weight, since the number of comparable solidi is small, and their interpretation hardly certain anyway” (Bammesberger 1990a, 459, n. 13). 52 In this regard Schneider follows Stephens (I: LXVIIIf.) who offered two interpretations, namely one as a personal name which is now the dominant one, and one as a sentence scan o modu ‘Scan owns this mot (stamp, die, coin)’; cf. Schneider (1967, 52). 53 ‘I have now carefully read everything (i.e., an earlier version of the paper, T. V.) – no objections. *skain- is certainly possible [as an alternative to *skaun- in skanomodu]. Seebold offers sufficient

216

Theo Vennemann

Prof. Düwel: Ich denke, gegen Ihre Interpretation des stammhaften -a-, entstanden aus -ai- und Charakterisierung als sowohl Englisch als auch Friesisch, lässt sich kaum etwas einwenden. Allerdings ist damit eine friesische Provenienz des Stückes insgesamt noch nicht vom Tisch. Der merkwürdige Fugenvokal bleibt in beiden Fällen, aber was machen Sie mit dem schließenden -u? Soweit ich sehe, spielt das in Ihrer Argumentation keine Rolle. Trotz gelegentlicher Gegenstimmen hat sich, soweit ich sehe, die Erklärung als spezifisch friesischer Endungsvokal für germ. -az etabliert. Damit bliebe immerhin ein Kennzeichen für friesische Zugehörigkeit erhalten.54

I have taken up these hints above. Prof. Seebold has made no specific suggestions, but his questions and his pessimistic attitude toward the entire enterprise have had a sobering effect which has caused me to make small attributed additions, but also changes in formulation that cannot all be traced to individual expressions of discontent.

References Århammar, Nils. 1995. “Zur Vor- und Frühgeschichte der Nordfriesen und des Nordfriesischen”. Friesische Studien II: Beiträge des Föhrer Symposiums zur Friesischen Philologie vom 7.–8. April 1994. NOWELE Supplement Series 12. Eds. Volkert F. Faltings, Alastair G. H. Walker, and Ommo Wilts. Odense: Odense University Press/Amsterdam: Benjamins. 63‒96. Bammesberger, Alfred. 1990a. “skanomodu: Linguistic Issues”. Britain 400‒600: Language and History. Eds. Alfred Bammesberger and Alfred Wollmann. Heidelberg: Winter. 457‒466. Bammesberger, Alfred. 1990b. Die Morphologie des urgermanischen Nomens. Heidelberg: Winter. Bammesberger, Alfred, and Alfred Wollmann (eds.). 1990. Britain 400‒600: Language and History. Heidelberg: Winter. Bammesberger, Alfred. 2006. “Das Futhark und seine Weiterentwicklung in der anglo-friesischen Überlieferung”. Das fuþark und seine einzelsprachlichen Weiterentwicklungen: Akten der Tagung in Eichstätt vom 20. bis 24. Juli 2003. RGA-E 51. Eds. Alfred Bammesberger and Gaby Waxenberger. Berlin: De Gruyter. 171‒187. Becker, Thomas. 1990. Analogie und morphologische Theorie. München: Fink. Berghaus, Peter. 1967. “Die Runensolidi in numismatischer Sicht”. Anglofriesische Runensolidi im Lichte des Neufundes von Schweindorf (Ostfriesland) Eds. Peter Berghaus and Karl Schneider. Köln/Opladen: Westdeutscher Verlag. 9‒40.

comparative material, so that theoretically no objection arises. To be sure: *skain- and *skaunappear broadly balanced, in my opinion.’ 54 ‘I think there can be no objection against your interpretation of the stem vowel -a- developed from -ai-, and the characterisation as English as well as Frisian. However, that does not yet brush aside the question of a Frisian provenience of the piece once and for all. The peculiar connecting vowel remains an issue in both cases, but what will you do about the closing -u? That does not play a role in your argumentation, as far as I can see. Despite occasional statements to the contrary the explanation as a specifically Frisian suffixal vowel for Germanic -az has established itself, as far as I see. With this at least one mark of a Frisian connection would remain.’

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

217

Berghaus, Peter, and Karl Schneider (eds.). 1967. Anglofriesische Runensolidi im Lichte des Neufundes von Schweindorf (Ostfriesland). Köln/Opladen: Westdeutscher Verlag. Braune, Wilhelm. 2004. Althochdeutsche Grammatik I: Laut- und Formenlehre. Tübingen: Niemeyer. Bremmer Jr., Rolf H. 1981. “Frisians in Anglo-Saxon England: A Historical and Toponymical Investigation”. Fryske Nammen 3: 45‒94. Brown, G. Baldwin. 1915. The Arts in Early England, Vol. III: Saxon Art and Industry in the Pagan Period. London: John Murray. Brunner, Karl. 1965. Altenglische Grammatik: Nach der Angelsächsischen Grammatik von Eduard Sievers. 3rd reworked ed. Tübingen: Niemeyer. Campbell, Alistair. 1959. Old English Grammar. Oxford: Clarendon Press. Derolez, René. 1990. “Runic Literacy among the Anglo-Saxons”. Britain 400‒600: Language and History. Eds. Alfred Bammesberger and Alfred Wollmann. Heidelberg: Winter. 397‒436. Düwel, Klaus. 1983. Runenkunde. 2nd ed., with an additional appendix. Stuttgart: Metzler. Düwel, Klaus. 2008. Runenkunde. 4th updated and revised ed. Stuttgart: Metzler. Düwel, Klaus, and Wolf-Dieter Tempel (eds.). 1970. “Knochenkämme mit Runeninschriften aus Friesland: Mit einer Zusammenstellung aller bekannten Runenkämme und einem Beitrag zu den friesischen Runeninschriften”. Paläohistoria 14: 353‒391. Einarsson, Stefán. 1945. Icelandic: Grammar, Texts, Glossary. Baltimore, MD: Johns Hopkins Press. Faltings, Volkert F. 2010. Etymologisches Wörterbuch der friesischen Adjektiva. Berlin: De Gruyter. Förstemann, Ernst. 1900. Altdeutsches Namenbuch. Erster Band: Die deutschen Personennamen. 2nd ed. Bonn: P. Hanstein. (URL: ‹www.archive.org/stream/altdeutschesnam00seelgoog#page/ n1/mode/2up›, last accessed 3 July 2014) [1st ed.: Nordhausen: Ferdinand Förstemann, 1857]. Gysseling, Maurits. 1962. “Het oudste Fries”. It Beaken 24: 1–26. Heidermanns, Frank. 1993. Etymologisches Wörterbuch der germanischen Primäradjektive. Berlin: De Gruyter. Hines, John. 1990. “The Runic Inscriptions of Early Anglo-Saxon England”. Britain 400‒600: Language and History. Eds. Alfred Bammesberger and Alfred Wollmann. Heidelberg: Winter. 437‒455. Hofmann, Dietrich. 1976. “Eine friesische Runeninschrift in England”. Us Wurk 25: 73‒76. [Also in: Gesammelte Schriften II: Studien zur Friesischen und Niederdeutschen Philologie. Eds. Gert Kreutzer, Alastair Walker and Ommo Wilts. Hamburg: Buske, 1989. 376‒379]. Hofstra, T. 1989. “Ier-Aldfrysk neist Aldingelsk en Aldsaksysk”. Philologia Frisica anno 1988: 38‒49. Köbler, Gerhard. 1994. Taschenwörterbuch des althochdeutschen Sprachschatzes. Paderborn: Schöningh. Krause, Wolfgang. 1968. Handbuch des Gotischen. 3rd ed. München: C. H. Beck. Laker, Stephen. 2007. “Palatalization of Velars: A Major Link of Old English and Old Frisian”. Advances in Old Frisian Philology. Amsterdamer Beiträge zur älteren Germanistik 64. Eds. Rolf H. Bremmer Jr., Stephen Laker, and Oebele Vries. Amsterdam: Rodopi. 165–184. Lehmann, Winfred P. 1986. A Gothic Etymological Dictionary, Based on the Third Edition of Vergleichendes Wörterbuch der gotischen Sprache by Sigmund Feist. Leiden: Brill. Looijenga, Tineke. 2013. “Die goldenen Runensolidi aus Harlingen und Schweindorf / De gouden runensolidi van Harlingen en Schweindorf”. Land der Entdeckungen: Die Archäologie des friesischen Küstenraums/Land van Ontdekkingen: De archeologi van het Friese kustgebied. Ed. Jan F. Kegler. Aurich: Ostfriesische Landschaftliche Verlags- und Vertriebsgesellschaft. 430‒431. Mailhammer, Robert. 2008. “Ablaut Variation in the Proto-Germanic Noun: The Long Arm of the Strong Verbs”. Sprachwissenschaft 33: 279‒300. Meid, Wolfgang. 1967. Germanische Sprachwissenschaft, Vol. III: Wortbildungslehre. Sammlung Göschen 1218/1218a/1218b. Berlin: De Gruyter. Miedema, Henricus Theodorus Jacobus. 1974. “Dialect en runen van Britsum en de oudste Anglofriese runeninscripties”. Taal en Tongval 26: 101‒128.

218

Theo Vennemann

Miedema, Henricus Theodorus Jacobus. 1977. “Noordzeegermaans (Ingveoons), Fries en Nederlands”. De Nieuwe Taalgids 70: 465‒472. Nedoma, Robert. 2007. “Die voraltfriesischen Personennamen der Runeninschriften auf dem Webschwert von Westeremden, dem Schwertchen von Arum und anderen Denkmälern”. Advances in Old Frisian Philology. Amsterdamer Beiträge zur älteren Germanistik 64. Eds. Rolf H. Bremmer Jr., Stephen Laker, and Oebele Vries. Amsterdam: Rodopi. 299‒324. Nedoma, Robert. 2014. “Voraltfriesisch -u im Nominativ und Akkusativ Singular der maskulinen a-Stämme”. Directions for Old Frisian Philology. Amsterdamer Beiträge zur älteren Germanistik 73. Eds. Rolf H. Bremmer Jr., Stephen Laker, and Oebele Vries. Amsterdam: Rodopi. 343‒368. Nielsen, Hans Frede. 1993. “Runic Frisian skanomodu and aniwulufu and the relative chronology of monophthongization and i-mutation”. Twenty-Eight Papers Presented to Hans-Becker Nielsen on the Occasion of his Sixtieth Birthday 28 April 1993. NOWELE 21/22. Odense: Odense University Press. 181‒188. Nielsen, Hans Frede. 1996. “Development of Frisian Runology: A Discussion of Düwel & Tempel’s Corpus from 1970”. Frisian Runes and Neighbouring Traditions. Amsterdamer Beiträge zur älteren Germanistik 45. Eds. Tineke Looijenga and Arend Quak. Amsterdam: Rodopi. 123‒130. Nielsen, Hans Frede. 2000. The Early Runic Language of Scandinavia: Studies in Germanic Dialect Geography. Indogermanische Bibliothek, Erste Reihe. Heidelberg: Winter. Odenstedt, Bengt. 1984. “The Chessell Down Runic Inscription”. Papers in Northern Archaeology (Archaeology and Environment 2). Ed. Evert Baudou. Umeå: Dept. of Archaeology, University of Umeå. 113‒126. Orel, Vladimir. 2003. A Handbook of Germanic Etymology. Leiden: Brill. Page, Raymond Ian. 1968. “The Runic Solidus of Schweindorf, Ostfriesland, and Related Runic Solidi”. Medieval Archaeology 12: 12‒25. Page, Raymond Ian. 1973. An Introduction to English Runes. London: Methuen. Page, Raymond Ian. 1995. “The Runic Solidus of Schweindorf, Ostfriesland, and Related Runic Solidi”. Runes and Runic Inscriptions: Collected Essays on Anglo-Saxon and Viking Runes. Ed. Raymond Ian Page. Woodbridge: The Boydell Press. 145‒160. [Reprint of Page 1968, with added Postscript, 159‒160]. Page, Raymond Ian. 1999. An Introduction to English Runes. 2nd ed. Woodbridge: The Boydell Press. Prokosch, Eduard. 1939. A Comparative Germanic Grammar. Special Publications of the Linguistic Society of America, William Dwight Whitney Linguistic Series. Philadelphia, PA: University of Pennsylvania. Quak, Arend. 1990. “Runica Frisica”. Aspects of Old Frisian Philology. Amsterdamer Beiträge zur Älteren Germanistik 31‒32. Eds. Rolf H. Bremmer Jr., Geart van der Meer, and Oebele Vries. Amsterdam: Rodopi. 357‒370. Reichert, Hermann. 1987. Lexikon der altgermanischen Namen. Teil 1: Text. Thesaurus Palaeogermanicus 1. Wien: Österreichische Akademie der Wissenschaften. Schneider, Karl. 1967. “Die Runensolidi – runologische, sprach- und religionshistorische Betrachtung”. Anglofriesische Runensolidi im Lichte des Neufundes von Schweindorf (Ostfriesland). Veröffentlichungen der Arbeitsgemeinschaft für Forschung des Landes Nordrhein-Westfalen: Geisteswissenschaften 134. Eds. Peter Berghaus and Karl Schneider. Köln/Opladen: Westdeutscher Verlag. 41‒75. Seebold, Elmar. 1970. Vergleichendes und etymologisches Wörterbuch der germanischen starken Verben. Ianua Linguarum, Series Practica 85. The Hague: Mouton. Stephens, George. 1866‒1901. The Old-Northern Runic Monuments of Scandinavia and England. 4 vols. (I: 1866‒1867, II: 1867‒1868, III: 1884, IV: 1901). London: John Russell Smith. de Vaan, Michiel. 2011. “West Germanic *ai in Frisian”. Amsterdamer Beiträge zur älteren Germanistik 67: 301‒314.

The Runic Inscription skanomodu: Frisian or Anglo-Frisian?

219

Versloot, Arjen P. 2001. “76. Vergleichende Aspekte friesischer Lautgeschichte”. Handbuch des Friesischen/Handbook of Frisian Studies. Ed. Horst Haider Munske. Tübingen: Niemeyer. 767‒775. Waxenberger, Gaby. 2017. “How ‘English’ is the Early Frisian Runic Corpus? The evidence of sounds and forms”. Frisians and their North Sea Neighbours: From the Fifth Century to the Viking Age. Eds. John Hines and Nelleke IJssennagger. Woodbridge: The Boydell Press. 93‒124. Waxenberger, Gaby. Forthc. A Phonology of Old English Runic Inscriptions with a Concise Edition and Analysis of the Graphemes. RGA-E. Berlin/Boston, MA: De Gruyter.

Kerstin Kazzazi

Introductory Remarks on the Contributions by Leslie Webster and Gaby Waxenberger on the Franks Casket The aim of this volume, as outlined in the “Introduction”, is to show the value of bringing together the different disciplines involved in research on objects with runic inscriptions and to illustrate the different approaches used to classify and interpret them. The way in which such interdisciplinary collaboration may be fruitful in corroborating and refining results is exemplified by the following tripartite section on the Franks Casket:1 Both Leslie Webster (Part A) and Gaby Waxenberger (Part C) look at this object independently from their respective disciplines with regard to its dating and provenance, applying the specific methods of their fields, art history on the one hand, and historical linguistics on the other:2 Webster compares the casket with other objects from the art-historical perspective according to motifs, materials and cultural context, whereas Waxenberger attempts to find out more about date and place of origin by comparing the inscriptions on the casket with other Old English texts (written in runes or the Latin script) with regard to their phonology, their rune-forms and corresponding sound-values. The two analyses are complemented by Part B, containing the images of the casket referred to in Parts A and C as well as a commented master-transliteration and interpretation by Gaby Waxenberger. Webster’s aim is to corroborate or, if possible, refine the accepted opinion that the Franks Casket “was made somewhere in Northumbria in the first half of the eighth century” (p. 225); the method she employs for this objective is to contextualise the casket by comparison with other “decorated Northumbrian objects from the later seventh and first half of the eighth centuries which can be both reliably dated and localised”, in order to arrive at a “secure art-historical sequence” (p. 225). The objects she draws on are thus “six securely dated and provenanced artefacts” (p. 232): the animal decorations in the porch of St Peter’s Church in Wearmouth, St Cuthbert’s Coffin, the binding of the Cuthbert Gospel of St John, the Lindisfarne Gospels, the Codex Amiatinus and the St Petersburg Bede. However, as the Franks

1 Generally, runic inscriptions are named after their find-spots. However, the Franks Casket is an exception, because this name, which it is most commonly known under, goes back to Sir Augustus Wollaston Franks, who purchased parts of the casket (Lid, Front Panel, Left Panel, Back Panel, Bottom Panel and a small piece of the Right Panel: see below Figs. 8a and 8b) in Paris in 1857. Most of the right Panel is in the Museo Nazionale del Bargello, Florence (Carrand Collection no. 25): for more details see Webster (2012). In research, the Franks Casket is also labelled as, e.g., the Casket of Auzon, the Auzon/Franks Casket and the Clermont (Runic) Casket. 2 For the way in which these and other disciplines work together in the ‘interdiscipline’ runology, see the “Introduction”. https://doi.org/10.1515/9783110796834-011

222

Kerstin Kazzazi

Casket “is without parallel, both as an object, and in terms of its overall appearance” (p. 234), the comparison is extended to objects from outside its assumed date and provenance. Similarities with a fifth-century reliquary box from Brescia thus point to a possible Italian influence. One of the problems with contextualising such an object is, however, the scarcity of comparable objects that have survived, e.g., due to the loss of most wooden objects. Thus, the “densely crammed narrative of text and image” (p. 234) includes motifs from Germanic mythology, and “late Roman conventions”, and draws on Christian traditions in a way possibly inspired by Italian models, making the casket unique and to a certain extent incomparable, which poses a problem for placing it in context. The way Webster seeks to overcome these difficulties is a close analysis of the details of the decorative elements on the casket to uncover possible clues towards its date and provenance. To this end, she looks at the following categories of elements: Human figures; animals, in particular birds; plants; architectural detail; interlace; letter forms. For each category, the closest parallels come from objects datable to the early eighth century and linked to Lindisfarne or Wearmouth-Jarrow. In sum, Webster shows that it may not be sufficient to look at the larger context; rather, it is only the close analysis of details that yields a satisfying result. Waxenberger also proceeds from an earlier proposition on the date and provenance of the casket, namely a study by Napier, which he based on criteria of the language used and the runes it is written in. While his approach does involve a close analysis of details and leads to a conclusion similar to the one drawn by Webster, Waxenberger demonstrates that for the results to be reliable, the data and method used must stand the test of plausibility and relevance. As Napier, like Webster, believed that the casket dates from the “beginning of the eighth century”, and that its home was “the coast of Northumbria” (p. 267), Waxenberger examines each of his criteria before the backdrop of the “Nhb. sub-corpus of the Old English Runes Corpus (OERC)” as well as comparable non-runic sources, such as the Ms. L of Cædmon’s Hymn. Napier’s list of criteria for dating comprises the occurrence of full vowels in unstressed syllables, the spelling ‹eu› for the Class. OE diphthong ēo, the use of the rune f to denote the allophone [v] of the phoneme OE /f/, Nhb. loss of -n in an alleged form “sefu”, and final -u in the word flodu. In a close analysis of the sources, Waxenberger is able to show that only the first of these, the still extant full vowels, is in fact an indication of an early date, whereas the other criteria are not valid for different reasons, e.g., incorrect readings (sefu instead of correct se(fa) or inconclusive evidence (the spelling ‹eu›). The same holds for Napier’s dialect criteria for Anglian, specifically Northumbrian, comprising the vowel changes of Anglian smoothing, absence of diphthongisation after palatals, development of extra, socalled parasite vowels, second fronting and loss of final -n as an inflectional ending. While the last criterion is invalid due to its alleged evidence being the incorrect reading “sefu”, the occurrence of the first vowel change (smoothing) and the lack of the second (diphthongisation) on the casket is shown by Waxenberger indeed to

Introductory Remarks on the Contributions on the Franks Casket

223

corroborate an Anglian provenance of the casket, whereas the lack of second fronting only rules out the more southerly region of Mercia. Napier’s linguistic criteria thus having been proven insufficient in particular for narrowing the origin of the casket down to Northumbria, five further pieces of evidence are adduced, which in fact lead to an even closer delimitation, i.e., to the east coast of Northumbria: Nhb. retraction of æ > a and the reflex of Gmc. *a + nasal constitute “additional support for Nhb.” (p. 281). However, the casket differs clearly from the runic inscription on the Ruthwell Cross, in the north-western region of Northumbria, in its runic representation of the clusters [xt], [çt], the runic representation of the velar consonants in the palatalisation/assibilation processes, and the fact that it does not show the vowel change of back mutation, which relates it in turn to the non-runic Cædmon’s Hymn (Ms. L). As a final step, this evidence from the language system and its representation in spelling is then underscored by looking at the runic inventory representing the language, which Napier had not taken into account at all. This approach also proves fruitful in linking the casket to other runic objects and in particular some from the east of Northumbria, such as the Whitby Comb, e.g., by the use of a certain type of the rune y. In particular, it is the lack of certain innovations to the runic inventory found only in the western inscriptions such as the Ruthwell and Bewcastle Crosses (rune calc ᛣ [k] and rune no. 31 ᛤ [ḵ]) that make an eastern Northumbrian provenance of the Franks Casket highly probable. While Waxenberger thus comes to roughly the same conclusions as Napier before, her results are based on a methodologically stricter investigation of sounder and more varied evidence, making them a more reliable foundation for further research. Through these different methodological paths, the two authors each not only succeed in narrowing down the possible time and place of the casket. What is even more exciting is that their results are highly congruent, pinpointing the origin of the casket to an area on the northeast coast of England, and its date to sometime in the early eighth century. Results corroborated by such methodological diversity may claim a high degree of validity. It therefore seems promising to continue this path of interdisciplinary collaboration.

Leslie Webster

A. The Dating and Provenance of the Franks Casket: An Art-Historical Perspective In what follows,1 I have assumed that it is generally accepted that both the linguistic and runological evidence suggest that the Franks Casket 2 (see this volume Part B, Fig. 1 FC total) was made somewhere in Northumbria in the first half of the eighth century; the purpose of this paper is to explore whether the art-historical evidence supports this and allows this judgment to be refined in any way. This is not an easy task; the nature of the art-historical evidence from this period is severely constrained. Other than contemporary coins, which have almost no links to the decoration of the casket,3 there are only a very few decorated Northumbrian objects from the later seventh and first half of the eighth centuries which can be both reliably dated and localised, and from which one might construct a secure art-historical sequence. As most of these are frequently referred to in the following discussion, I list them below in chronological order: 1. The sculptural decoration of intertwined animals inside the original entrance porch of St Peter’s Church, Wearmouth, part of the original monastery founded by Benedict Biscop in 672/3 (see Figs. 1a and 1b; Speake 1980, Fig. 12e). 2. The wooden reliquary coffin of St Cuthbert, made for the saint’s translation at Lindisfarne in 698 (see Fig. 2; Kitzinger 1956).

1 This paper has its origins in an earlier paper on style and the Franks Casket (Webster 1982) but considerably extends the discussion, and the conclusions drawn. 2 See fn. 1 in Kazzazi, this volume, “Introductory Remarks on the Contributions by Leslie Webster and Gaby Waxenberger on the Franks Casket”. 3 A sole exception is discussed below. https://doi.org/10.1515/9783110796834-012

226

Leslie Webster

Fig. 1a: Decoration on the entrance porch of St Peter’s Church, Monkwearmouth; photo: J. Hines.

Fig. 1b: Decoration on the entrance porch of St Peter’s Church, Monkwearmouth; © G. Speake.

Fig. 2: St. Cuthbert’s coffin, symbols of Sts Matthew, and Mark (Battiscombe 1956, Plate VII).

The Franks Casket, Part A: An Art-Historical Perspective

227

3. The binding of the Cuthbert Gospel of St John, discovered in the saint’s coffin when it was opened on 29 August 1104 (see Fig. 3; Webster 2015). Palaeographical and textual evidence suggest that the gospel text was written at Bede’s twin monastery of Wearmouth-Jarrow sometime between 710 and 730; however, the condition of the outer leaves of the manuscript indicates that it was in use for some time before the decorative binding was added (Pickwoad 2015, 45‒46; Gameson 2015b, 130). The closest parallels for the decoration of the binding lie within the first half of the eighth century, most probably within the period ca. 700‒730. Comparative analysis of the decoration and pigments on the front and back boards of the binding suggests that, though they are significantly different in style, they were made at the same time and in the same workshop. On stylistic evidence, this could have been at Wearmouth-Jarrow or at Lindisfarne itself; but the fact that the text was written at Wearmouth-Jarrow may make an origin there more likely, given that monastery’s strong Mediterranean links, and that Mediterranean traditions are evident in both the binding technique and the decoration and layout of the front cover.4

Fig. 3: The Cuthbert Gospel of St John; © London, British Library, Add. MS 89000, Front Cover.

4. The Lindisfarne Gospels, dated by a later colophon to the time of Bishop Eadfrith of Lindisfarne (d. 722), and written and decorated by Eadfrith himself at the monastery on Lindisfarne (see Fig. 4).

4 On the text, see Gameson (2015a); on the binding, see Pickwoad (2015); on the decoration of the front cover, see Webster (2015).

228

Leslie Webster

Fig. 4: St John image, Lindisfarne Gospels, London, British Library, Cotton MS Nero D. IV fol. 209v; © The British Library Board.

The Franks Casket, Part A: An Art-Historical Perspective

229

Recent evaluation of the evidence points to the manuscript being written during the period ca. 690‒710 (Gameson 2013, 18‒19, 71). However, a suggestion that the Gospels, like the Cuthbert Gospel of St John, were made for presentation at the time of the translation of St Cuthbert’s remains in 698 is now regarded as unlikely (Gameson 2013, 18). Nevertheless, it seems probable that this supremely prestigious manuscript was made for the shrine of Northumbria’s major saint. 5. The Codex Amiatinus, one of three great Bibles commissioned by Abbot Ceolfrith of Wearmouth-Jarrow (see Figs. 5 and 6). This was copied from a sixth-century South Italian manuscript, the Codex Grandior, produced in Cassiodorus’ monastery at Squillace in Calabria, and which was probably brought back to Wearmouth-Jarrow by Benedict Biscop or Ceolfrith on one of several forays into Italy to gather books and other essential equipment for the new monastery. The Codex Amiatinus was written and decorated at Jarrow between 689 when Ceolfrith became abbot, and 716 when he left for Rome with the Codex, intending to present it to the Pope; a date towards the latter end of this spectrum is generally considered likely (Nees 2011).

Fig. 5: Depiction of the Tabernacle, Codex Amiatinus, Florence, Biblioteca Medicea Laurenziana, MS Amiatino 1, f. fols. IIv‒IIIr; reproduced with permission of MiBACT; further reproduction by any means is prohibited.

230

Leslie Webster

Fig. 6: Jerome’s Prologue, Codex Amiatinus, Florence, Biblioteca Medicea Laurenziana, MS Amiatino 1, f. fol. IVr; reproduced with permission of MiBACT; further reproduction by any means is prohibited.

6. The St Petersburg Bede, the second-oldest Bede manuscript, also made at Wearmouth-Jarrow (see Fig. 7; see also Figs. 15 below). This has commonly been dated to ca. 746 on internal evidence (Wright 1961; Alexander 1978, cat. no. 19; Brown 2007, 43); however, a slightly later date, in the third quarter of the century, is preferred by Gameson (2015a, 25, 27, 33), following Dumville (2007; see also Hines, this volume). This very short list makes plain the problem: These exceptional treasures have survived from only two major centres of Northumbrian monasticism. In the case of Lindisfarne, the association of these treasures with the shrine of St Cuthbert, and their consequent veneration, helped them survive the many vicissitudes of the years between the Viking sack of Lindisfarne and the re-establishment of the shrine at Durham in the late tenth century. In the case of Wearmouth-Jarrow, survival was probably linked to the exceptional output from this powerhouse of learning, and Bede’s reputation. Although both centres were clearly pre-eminent in their time, sheer luck will also have played a significant part in the survival of these, as indeed it seems to have done in the case of the Codex Amiatinus, whose identity was completely lost for hundreds of years (Alexander 1978, cat. no. 7). In addition, and despite their different religious and cultural origins, these two centres also shared very close ties, as their use of common textual exemplars has shown (Brown 2007, 10; Gameson 2013, 45‒47). From all the other important Northumbrian monastic centres, such as Hoddom, Abercorn, Melrose, Hexham, Tynemouth, Hartlepool, Lastingham, Whitby, Ripon and York (see Map 1 below), little survives other than sculpture (itself usually decontextualised) and minor items

The Franks Casket, Part A: An Art-Historical Perspective

231

Fig. 7: Decorated initial with the image of Gregory the Great, The St Petersburg Bede, MS Lat. Q. v. I. 18 fol. 26; © The National Library of Russia, St. Petersburg.

which have been found casually, or through archaeological excavations of varying scope and quality. The latter provide enough to tell us that at Hartlepool and Whitby, for example, fine metalworking and glass manufacture was being carried out, just as Rosemary Cramp’s extensive excavations have shown was the case at Wearmouth-Jarrow (Cramp 2006). Some of this industrial activity was for producing decorative mounts for book covers, presumably for the very manuscripts which were produced at these royal monastic centres. But not a single manuscript survives today which can be associated with these other Northumbrian foundations. At the same time, the remarkable sophistication of the Ruthwell and Bewcastle stone crosses, erected by communities in the north-western outposts of Northumbria, remind us that foundations of which we know virtually nothing were capable of producing or commissioning work of the highest intellectual and artistic quality in the middle of the eighth century (see Fig. 8).5 Such huge lacunae in the evidence thus make it impossible to construct anything more than a rather sketchy and inevitably unbalanced pedigree for the production of fine manuscripts and other high-status artworks in seventh- and eighthcentury Northumbria. Outside the magic circle of those six securely dated and prov-

5 There is no independent dating evidence for these two technically and artistically sophisticated, and closely related, stone crosses, though a date in the middle of the eighth century is generally accepted on stylistic and (in the case of the Ruthwell Cross) linguistic and runological grounds.

232

Leslie Webster

Map 1: Northumbrian monastic centres mentioned in the text; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München.

enanced artefacts listed above, other decorated items may of course be related to these known points by shared or similar features. These features might include: shared textual exemplars, scribes or scribal practices, or decorative signatures, along with any circumstantial evidence for dating or localisation which might exist. But many manuscripts and other portable decorated artefacts simply cannot be ascribed to any particular scriptorium or workshop, and/or are only loosely datable. Others may be assigned a provenance or close date through long sequences of association, but these are fragile constructs, easily disrupted by new discoveries or theories. To take one very recent example: a panel of animal decoration on the seventhcentury gold processional cross in the Staffordshire Hoard shares an identical physical model with panels on the rim mounts of the Sutton Hoo maplewood bottles,

The Franks Casket, Part A: An Art-Historical Perspective

Fig. 8: The Ruthwell Cross; © Historic Environment Scotland; photo: K. Majewski.

233

234

Leslie Webster

suggesting that these two objects are rather closer in date, if not actually contemporary, than the forty or so years’ chronological difference that their contexts would suggest (Høilund Nielsen 2010; below, Fig. 11b, d).6 By association, it may also invite reassessment of the date of the stylistically close Book of Durrow, which contains animal ornament stylistically and functionally very similar to that on the processional cross, but has been conventionally dated on palaeographical and textual grounds to c. 680, and located within a general Insular orbit that includes both Ireland and Northumbria (see Fig. 9; Alexander 1978, cat. no. 6; Brown 2007, 9‒10, 26). An earlier dating for the manuscript, and its association with an East Anglian stylistic repertoire, might even suggest a reattribution from a Columban monastery on Iona or Northumbria to one in East Anglia (Henderson/Henderson 2010; Webster 2016, 35‒36). However, the relationship of portable art in other media to high-end manuscripts is not always as close as in that case, leaving questions of provenance and date sometimes a matter of faith rather than of fact. In the case of the FC, there is an additional problem, in that it is without parallel, both as an object, and in terms of its overall appearance. The nearest formal parallel is not Anglo-Saxon, but a fifthcentury Italian ivory box reliquary now at Brescia (see Fig. 10) in northern Italy (Webster 1982 and 2012b, 30‒32), suggesting that an Italian import of similar type and appearance probably served as a model for the casket’s carver. The use of ivory or bone, the similar scale and method of construction, and the decorative scheme and organisation, cross-referencing from the main images to the borders which surround them, is identical on both objects. Nothing else like this survives from AngloSaxon England; and no other Northumbrian artefact – not the Lindisfarne Gospels, not the Ruthwell Cross – has the kind of densely crammed narrative of text and image which is such a distinctive signature of the casket. However, the casket was possibly less unique than it appears today. St Cuthbert’s coffin apart, no wood carving to speak of now survives from the seventh or eighth century, though it must have been one of the principal media for both large- and small-scale artistic output. The enduring portrayal of apparently mythological motifs in sixth- and seventh-century Anglo-Saxon metalwork also testifies to an established tradition of visualising native stories. It is then possible that the Germanic scenes depicted on the casket draw on an indigenous tradition of narrative carving, now largely lost to us; but they also employ a visual vocabulary and framework based largely on late Roman conventions, and copy their Italian model in the use of an early Christian form of condensed narrative (Webster 2012a,

6 The Sutton Hoo Mound 1 ship burial is generally thought to have been assembled around 625; the mounts on the maplewood bottles show little wear and are likely to have been among the most recently made objects in the burial. The Staffordshire Hoard gold processional cross, which also shows little sign of use, is one of the later objects in the hoard, which was probably buried in the third quarter of the seventh century.

The Franks Casket, Part A: An Art-Historical Perspective

235

Fig. 9: Carpet page, Book of Durrow, Trinity College Dublin Library MS 57, fol. 192v; © The Board of Trinity College Dublin.

236

Leslie Webster

Fig. 10: The Brescia Casket (lipsanotheca), Museo di Santa Giulia, Brescia; © Photographic Archive of the Brescia Museums, Fotostudio Rapuzzi.

84, 95‒96). This possibly experimental amalgam of traditions makes the casket a unique exemplum, with all the problems that such uniqueness brings for locating it in a secure art-historical context. Nevertheless, despite such obstacles, the FC contains a variety of decorative elements, which may help to date, and possibly, locate its place of origin a little more precisely (Webster 1982).

Human Figures The casket teems with human figures, many of which are engaged in movement, something less often seen in contemporary Insular manuscripts or stone sculpture. The overall treatment of the human figure is limited by the strong grain of the whale bone, which does not permit much fine detail, but within that constraint, the figures are carved with some subtlety, and the distinctive strong linear grooving of their garments gives a dynamic verve to the scenes. A marked linearity and stylisation of

The Franks Casket, Part A: An Art-Historical Perspective

237

the human figure is a feature of fourth- to seventh-century Italian art, and manuscripts and panel paintings exhibiting this are known to have reached England in the seventh century, and to have served as models for books and other ecclesiastical objects produced for the Anglo-Saxon Christian church.7 It has been suggested that the image of Mary displaying the infant Christ in a clipeus in the Adoration scene may be based on a Marian icon of the kind brought back to Wearmouth-Jarrow by Benedict Biscop and Ceolfrith (Bailey 2005, 9‒10).8 The depictions of the Sack of Jerusalem by Titus (Back Panel: cf. Part B Fig. 5), the finding of Romulus and Remus (Left Panel: cf. Part B Fig. 4), and the Adoration of the Magi (Front Panel: cf. Part B Fig. 3) on the casket must also have been adapted from imported early Christian sources, including cosmographical manuscripts (see also below). As noted above, the three scenes from Germanic legend draw in part on a different iconographical and narrative tradition, which was mainly expressed in perishable materials such as wood and textiles. However, it is noteworthy that notwithstanding a common cultural inheritance, Anglo-Saxon seventh-century portrayals of human figures from Germanic tradition are rather different in style to those on the casket; in those warrior and god images, dress is either absent or highly textured, and does not normally exhibit the strong linearity of the Germanic figures on the casket,9 which are stylistically indistinguishable from the figures in those scenes taken from early Christian sources. Working in a learned environment, the maker of the casket clearly chose to base his figure style on the Continental models which were available to him. The FC figural style is thus a version of the same late Roman linear figural style that has influenced a number of well-dated, mostly eighth-century Northumbrian artefacts: these include St Cuthbert’s coffin (Fig. 2, and Kitzinger 1956), the Lindisfarne Gospels (Fig. 4), the Codex Amiatinus (Fig. 5), the Durham Cassiodorus,10 and the St Petersburg Bede (Fig. 7). In all of these, the human form

7 For example, the St Augustine’s Gospels, Cambridge, Corpus Christi College MS 286. This sixthcentury gospel manuscript was already in England by the late seventh or early eighth century, when the text was annotated in an Insular hand. It was kept at St Augustine’s Abbey, Canterbury, from at least the eleventh century. It may well have been brought to England at an early stage of the conversion, and certainly represents the kind of manuscript that St Augustine and his fellow missionaries would have imported for the nascent church. Its simple linear depictions of the human form typify a prevailing current in early Christian art, and indeed the small-scale figures in the narrative scenes on fols. 125r and 129v are stylistically very reminiscent of the figures on the casket, indicating the kind of model that would have been available to the maker of the casket. 8 Bailey has pointed out that the famous seventh-century icon of Mary in Sta. Maria Antiqua at Rome displays just such an image of Christ in a clipeus; this is an object which Benedict Biscop and Ceolfrith could have seen for themselves. It is noteworthy that the icons brought back for the church at Wearmouth included depictions of the Virgin Mary. 9 For example, the warrior figures on the Sutton Hoo helmet, and on the Finglesham buckle, or the three-dimensional amulets of gods; see Webster (2012a). 10 Dated to the second quarter of the eighth century, and possibly made at York (Gameson 2010, 34‒37), though the traditional ascription to Wearmouth-Jarrow is not excluded.

238

Leslie Webster

is depicted in a two-dimensional format, defined by its flattened linear draperies and its simplified gestures and features, the latter usually fully frontal. All these figures are, as noted above, static rather than active images, and most are portraits of sacred figures, in contrast to the casket’s mostly secular cast of animated characters; as large-scale finely incised wood carving, or painted manuscript images, they also exhibit greater technical freedom, compared with the constraints of carving strongly grained whale bone. Nevertheless, these stylised presentations of the human figure, portrayed as generic types, not individuals, and swathed in ribbed draperies, are strikingly similar, even in such different media. With the possible exception of the Durham Cassiodorus,11 the origins of the closest parallels for this figure treatment focus on Lindisfarne or Wearmouth-Jarrow, and, with the slightly later exceptions of the David and Christ images in the Durham Cassiodorus figures and the portrait of Gregory the Great in the St Petersburg Bede (Fig. 7), they cluster in date around the first quarter of the eighth century.

Animals (see Webster 1982, Fig. 1) Animal motifs also figure prominently on the casket. These comprise the wolves on the Romulus scene on the left hand end (cf. Part B Left Panel: Fig. 4), the horse on the opposite end (cf. Part B Right Panel: Fig. 6a), a number of birds (cf. Part B Front Panel: Fig. 3; Right Panel: Fig. 6a), symbolic creatures (e.g., Part B Back Panel: Fig. 5), and the small crouching beasts in the corners of some of the panels (Webster 1982, Figs. 1 and 3; see also, for example, Part B Front Panel: Fig. 3; Left Panel: Fig. 4). With their dipped or backward-biting heads with pointed ears and long, lobed jaws, the crouching beasts in the corners of some of the panels are clearly linked to late → Style II processional beasts. Similar creatures occur in two manuscripts

Fig. 11: (a) Carpet page animal, Book of Durrow, Trinity College Dublin Library MS 57, fol.192v; (b) animal from initial, Corpus-London Gospels, Corpus Christi College Cambridge, MS 197B, fol. 2; (c) animal from initial, Durham Gospels, Durham Cathedral Library, MS A. II.17, fol. 2r; drawing: M. O. Miller.

11 Folios 81v and 172v of Durham Cathedral Library MS. B.II.30 Cassiodorus on the Psalms can be viewed online, provided by Durham Priory Library Project, a collaboration between Durham University and Durham Cathedral: ‹n2t.durham.ac.uk/ark:/32150/t2mrn3011371.html›, last accessed 6 December 2020.

The Franks Casket, Part A: An Art-Historical Perspective

239

which, on scribal evidence, were possibly made at Lindisfarne or one of its daughter houses: the Corpus-London Gospels, which date to the early eighth century, and the Durham Gospels, which date to c. 700 (see Figs. 11b and c). Somewhat earlier are the related processional creatures on the large cross from the Staffordshire Hoard (Webster 2016, Pls. II‒IV) and in the Book of Durrow (Figs. 9 and 11a,), the dating of which remains under discussion, but which falls within the period 650–700 (Webster 2016, 35‒36; see also above). Another piece of metalwork with a more distantly related small beast comes from excavations in the seventh-century royal fortress at Bamburgh, just opposite the tidal island of Lindisfarne (Webster/Backhouse 1991, cat. 45). A telling sculptural parallel to the pointed, laid-back ears and long jaws of some of the casket creatures occurs in the heads of the beasts in the church porch at Monkwearmouth (Fig. 1b). Almost all of these comparanda can be linked to specifically Northumbrian parallels, and, like the figures on the casket, find their closest counterparts in artefacts with Lindisfarne or Wearmouth-Jarrow connections belonging mostly to the first quarter of the eighth century. The larger animal figures have no direct parallels from Northumbria, though wolves and (unmounted) horses occur very occasionally on the Pictish symbol stones of eighth-century date; but their postures and appearance are somewhat different to those on the casket. The suggestion of musculature in the portrayal of the FC horse (Webster 1982, Fig. 1a; cf. this volume Part B Right Panel: Fig. 6), and its general posture, are more reminiscent of the calf which symbolises St Luke in the Echternach Gospels (see Fig. 12). This illuminated manuscript is linked paleographically to the Lindisfarne Gospels and was probably made there between 690 and 710, around the time of the foundation of the monastery at Echternach, now in Luxembourg, by the Northumbrian Willibrord in 698 (Alexander 1978, cat. no. 11; Gallica, View 236, fol. 115v); a very similar calf symbol appears in the related Trier Gospels, which were probably written at Echternach in the second quarter of the eighth century (Alexander 1978 cat. no. 26). The casket craftsman seems here to be adapting the Insular evangelist symbol convention to give shape to the horse in the Germanic scene, just as he uses a late Roman stylistic convention to portray the human figures in these native stories. Also of possible relevance to the date and origin of the casket are the distinctive triple knots which appear under the horse’s belly and forelegs (cf. Part B Right Panel: Fig. 6). The coins of King Aldfrith of Northumbria (685‒704), and of one of his successors, Eadberht (737–758), depict a prancing horse-like creature, perhaps some kind of regnal emblem; on the coins of Eadberht, this creature also has a triple knot between its legs (Webster 2012b, 52‒53, Fig. 44). Parallels on eighth-century stone monuments from Gotland showing an eightlegged horse with a triple knot between its legs suggest that this may reflect a common Germanic tradition, possibly indicating the god Odin’s horse, Sleipnir. However,

240

Leslie Webster

Fig. 12: Symbol of St Luke, Echternach Gospels, Bibliothèque nationale de France: MS Latin 9389, f. 115r; © Paris, Bibliothèque nationale de France, Département des Manuscrits.

The Franks Casket, Part A: An Art-Historical Perspective

241

Fig. 13: Die Stamp from Fen Drayton, Cambridgeshire; © Portable Antiquities Scheme: Record, ID: NLM-468D41 (CC BY-SA 4.0) ‹finds.org.uk/database/images/image/id/ 392444/recordtype/artefacts›, last accessed 12 November 2020.

since no other Anglo-Saxon examples of this distinctive image survive, these coin parallels might be significant in terms of locating and dating the image on the casket. The pair of winged beasts below the Ark on the back panel (Webster 1982, Fig. 1f; Part B Back Panel: Fig. 5), which seem to portray oxen, are probably meant to evoke, with the eagles above the Ark, the creatures of the Temple described in Ezekiel’s vision (Ezekiel, 1:5‒10; Webster 1999, 238); but they appear to have no early medieval parallel. The same is true for the strange hybrid creature on the right end panel (Webster 1982, Fig. 1e; Part B Right Panel: Fig. 6). Examples of creatures with human bodies and animal heads are known from Continental examples, particularly wolf-headed warriors, as on the late sixth or early seventh-century foils on sword scabbards from Gutenstein and Obrigheim, Germany, and on the Torslunda, Öland, dies (cf. Høilund Nielsen 2003, 187–218). A recent discovery of a late sixth- or early seventh-century Anglo-Saxon die for making foils with similar images, from Fen Drayton, Cambridgeshire (Fig. 13), confirms that such motifs were current in England. These, however, are different from the winged and horse-headed creature on the casket. A tradition of Germanic human/ animal hybrids undoubtedly lies behind this image, but it remains quite unique. However, two other symbolic creatures, the two sets of conjoined heads on the lid (cf. Part B Fragmentary Lid: Fig. 2), are recognisably from the Anglo-Saxon repertoire (Webster 1982, Fig. 3, e, f). The larger pair, below the female figure under an arch, have a close parallel in the mid-eighth-century Durham Cassiodorus manuscript. Here the composite David/Christ figure treads the beasts under his foot, in a reference to Psalm 91:13 ‘You shall tread upon the lion and adder: the young lion and the serpent shall you trample under feet’. The smaller pair of conjoined, long-

242

Leslie Webster

jawed beast-heads curve around the female’s head, in an image which may be intended to represent protection; a range of similar motifs occurs in sixth- and seventh-century Anglo-Saxon jewellery and other prestige metalwork, where they appear to be apotropaic in function (Speake 1980, Fig. 11; Webster 2003, 18, Fig. 6). A related motif dating to the late seventh century adorns the original porch at Wearmouth (Speake 1980, Fig. 12e; above, Fig. 1). But this is a widespread and evidently long-lived motif which is not susceptible to more precise attribution.

Birds (see Webster 1982, Fig. 3) Like the Lindisfarne artist, the carver of the FC had more than one kind of bird in his repertoire, and we can identify here ducks (cf. Part B Front Panel: Fig. 3), birds of prey, and maybe, on the right-hand end, a raven (cf. Part B Right Panel: Fig. 6). They share stylistic features, including scaly feathers, and interlacing heads, with birds in the Lindisfarne Gospels itself (Webster 1982, Fig. 4 a, d), and related manuscripts. More particularly, the scrolled junction of wing and body in the birds seized in the Weland scene, where the scroll distinctively sits at the back of the neck, has direct parallels in the Lindisfarne Gospels (e.g., fols. 139r, panel infill, and 210v), the Corpus/London Gospels, another possible Lindisfarne manuscript (fol. 2; Webster 1982, Fig. 4c) and with another early eighth-century Northumbrian manuscript, the Cologne Collectio Canonum (fol. 1r; Alexander 1978, cat. no. 13). A sculptural parallel is provided by the birds on a cross-shaft from Aberlady (Midlothian, Scotland), which share similar long wing feathers with the Weland scene birds (Fig. 14; Webster 1982, Pl. 29). This probably dates to the first third of the eighth century, on account of this and other decorative elements which it shares with the Lindisfarne Gospels; it was erected during the period when Northumbria still held control over this part of southern Scotland. The raptorial bird heads on the fine early eighth-century memorial of the priest Herebericht at Wearmouth recall the bird-heads above the Ark (Webster/Backhouse 1991, cat. 72; cf. Part B Back Panel: Fig. 5) and the duck-like bird at the Adoration scene has a close parallel in the small duck-head from a sculptural frieze at the Jarrow monastery (Cramp 2006, Fig. 28.2.9). These parallels centre on objects which are linked to Lindisfarne or Wearmouth-Jarrow, and are datable to the early eighth century.

The Franks Casket, Part A: An Art-Historical Perspective

Fig. 14: Cross-shaft from Aberlady, Midlothian; © Trustees National Museums Scotland; photo: A. Maldonado.

243

244

Leslie Webster

Fig. 15: Decorated initial, St Petersburg Bede, MS Lat. Q. v. I. 18. f. 3v; © The National Library of Russia, St. Petersburg.

The Franks Casket, Part A: An Art-Historical Perspective

245

Plants (Webster 1982, Fig. 5) The casket’s plant ornament, whether sprigs or forest trees (e.g., Part B Left Panel: Fig. 4; Right Panel: Fig. 6), appears as pliant, supple branches with round buds or berries, and oval or pointed leaves, often in triplets, evoking its late Antique origins. It derives directly from late Roman equivalents; fol. 129b of the St Augustine’s Gospels display characteristic examples of undulating plants with pointed leaves and berries. Less directly, it also relates to the classical and early Christian vine-scroll/ Tree of Life motif which plays such a prominent part in the decorative vocabulary of eighth-century (e.g., the Ruthwell Cross, see Fig. 8) and later Northumbrian sculpture. Related plant motifs appear in other artefacts of Northumbrian origin, such as the front cover of the early eighth-century Cuthbert Gospel of St John (Fig. 3). This shares distinctive round berries with the FC plants, as well as other features, discussed below. An initial in the St Petersburg Bede, another Wearmouth-Jarrow product, also has plants with drooping pointed leaves (Fig. 15); but perhaps the closest parallels are to be found in the Echternach Uncial Psalter, a grand Echternach manuscript dated to the middle of the eighth century, where very similar undulating plants with pointed leaves and round berries occur (Alexander 1978, cat. 28, Figs. 83 and 140‒142). Given the direct connexions of this plant ornament with Mediterranean exemplars, it is perhaps not surprising to observe parallels with Wearmouth-Jarrow products; but the close Echternach parallel reminds us that such exemplars were widely used. The date range suggested by these comparisons covers the first half of the eighth century, possibly into the third quarter.

Architectural Detail The casket’s lid and back display rich architectural detail, though close parallels are hard to come by. On the lid, the elaborate fortress defended by the Ægili figure has no contemporary parallel. It is shown in plan, while within it, the female figure stands beneath an arch seen in elevation. This same strategy is employed in the depiction of the Tabernacle in the Temple of Jerusalem in the Codex Amiatinus, in which the curtained perimeter is shown in plan, and the objects within it in elevation (Fig. 5); no other comparable Anglo-Saxon attempt at depicting an architectural complex survives from this period, and the casket’s depiction clearly draws on late Antique convention, as seen in the Codex Amiatinus image. The depiction of the Temple on the back of the casket is very different, concentrating on its symbolic aspects, again an image without contemporary Insular parallel. The form of the arch here is, however, related to the arches on the lid and on the front, with their stepped bases and capitals, and their ribbed or interlace-decorated

246

Leslie Webster

columns. There is a clear resemblance to the stepped arches of canon tables and other arched features in early Gospel manuscripts, with once again particularly close resemblances to elements in the Lindisfarne Gospels and the Codex Amiatinus; in the latter, Jerome’s prologue has arches with simple two-strand interlace very similar to that on the columns of the arch on the lid (Fig. 6). The other two arches on the casket, representing the Temple itself and framing the enthroned Virgin and Child, carry metaphorical weight as gateways to understanding (as indeed has also been suggested for canon table arches). In this case, they may allude to the Old and New Covenants symbolised in the two scenes. The more elaborate intermediate stepping which both share is not seen in the arcading of Anglo-Saxon canon tables of this period, but it does recall the midway horizontal moulding seen on the arch over the Good Shepherd on the Brescia reliquary; the prominent ribbing of the Franks Casket columns also suggests that the carver was copying fluted columns, like those on the Brescia casket, where also stepped bases can be seen (Fig. 10). This may indicate that the ribbing/fluting was copied from a plastic model, rather than from a manuscript exemplar. Finally, while not strictly architectural, the distinctive use of pellets as infill on the Front Panel (Part B: Fig. 3) and Lid (Part B: Fig. 2) of the casket finds a parallel in the Salaberga Psalter, fol. 2 (Fig. 16), which has been assigned to Northumbria and the first half of the eighth century. Overall, the comparative evidence is slight, but links to manuscripts which can be ascribed to Lindisfarne and Wearmouth-Jarrow in the first quarter of the eighth century are suggestive, as well as to less closely dated and located Northumbrian manuscripts.

Interlace The appearance of two-strand interlace on the columns of the arch on the lid (Part B: Fig. 2) directs us to another feature, present on every side of the casket. With the exception of the lid, where the original framing panels are lost, a twostrand interlace surrounds each panel. This simple form of interlace is actually quite hard to find in the grand Northumbrian manuscripts, where more elaborate versions are preferred. Three significant examples present themselves, however. The first appears in the opening initial of St Mark’s Gospel (fol. 2r), in a Durham Gospel fragment (Durham Cathedral Library, MS A.II.10), made in Northumbria in the middle of the seventh century (Alexander 1978, cat. 5, Fig. 9; see also Webster 2012a, Fig. 50). The second is on the leather binding of the Cuthbert Gospel of St John, where an identical interlace, coloured in yellow, surrounds the decoration on the front cover, performing the same framing function as that on the FC (Fig. 3; also Webster 2015, Fig. 3.ii). As noted earlier, this was made at Wearmouth-Jarrow, probably no later than ca.730. The third appears in the arches which frame St Jerome’s prologue

The Franks Casket, Part A: An Art-Historical Perspective

Fig. 16: Decorated initial, Salaberga Psalter: Staatsbibliothek, Berlin, MS Hamilton 553, fol. 2; © bpk, Staatsbibliothek zu Berlin, C. Seifert.

247

248

Leslie Webster

in the Codex Amiatinus (Fig. 6). The use of this simple motif does not seem to outlast the middle of the eighth century, and the closest parallels are with WearmouthJarrow.

Letter Forms No discussion on the art-historical dating of the casket can ignore its lettering, both runic and non-runic. Uniquely in this period, it is carved in relief, in itself exceptional, and an extraordinary accomplishment in this strongly grained medium. What relief carving does, however, is to enable the carver to indicate in some of the letters the ductus, the sequence of the letter strokes – in both runic and non-runic – as though he were writing with a pen; another indication, were it needed, of the close links between the casket and the scriptorium (Webster 2012b, Figs, 41 and 42). The form of the Roman letters takes us even closer. These are Phase II Insular halfuncial, except for the forms of ‹C›, ‹E›, ‹G› and ‹T›, which are capitals. The distinctive mixture of capitals and half-uncial is characteristic of the display script in the monogram pages of the Lindisfarne Gospels and related manuscripts such as the Echternach Gospels. The letter forms are also characteristic, including the ‘horned’ ‹a›, and the ductus of ‹u›, ‹a›, ‹n› and ‹m› is clearly visible. Here the principal link is to the Lindisfarne Gospels and its circle, suggesting a date in the early eighth century.

Contextual Evidence Though this paper has focussed on art-historical comparisons, in the spirit of ‘reading’ the casket as a whole, it is appropriate to look at two pieces of contextual evidence which deserve consideration, in attempts to narrow down the dating and provenance of the casket. First, the material itself: whale bone. As the riddling verse on the Front Panel implies, the use of this particular medium to construct the casket was significant in more than one way. The carcass of a beached whale was of course a very rare and much-prized commodity, indeed, from at least the reign of Edward II, it was the perquisite of the crown; so this was a very precious medium, a worthy substitute for the even rarer elephant ivory.12 The whale was also an animal of particular symbolic significance in terms of the overall message of the casket – which is a highly learned construct, probably made to hold an object fit for a king or prince (Webster 1999,

12 Walrus ivory does not seem to have been available to Anglo-Saxon carvers before the opening up of Viking trade in the later tenth century, after which its use becomes widespread.

The Franks Casket, Part A: An Art-Historical Perspective

249

2012b). So could the importance of the whale that was transformed into the casket receive some added enhancement from the inscription on the front, if that was a direct account of this particular whale’s stranding, rather than a mere trope? Napier, perhaps with tongue in cheek, suggested that maybe the description of the cliff or rocky shore on which the whale was stranded referred to the coast below the AngloSaxon monastery at Whitby (Napier 1901, 377). Of course that was pure speculation; but in fact a number of Northumbrian coastal monastic and royal sites such as Coldingham, Bamburgh, Tynemouth and Hartlepool, as well as Whitby – are located by cliffs and/or rocky shores, and so the possibility that the inscription on the casket refers to a significant local event may not be unreasonable. Of course the bone would be portable, once cut up, and could in theory have travelled a long way to the place where it was transformed into the casket. But as a valuable commodity, the whale’s carcass would have been swiftly controlled by local secular or ecclesiastical authorities, and the jaw-bone’s transformation into a learned object of probable royal significance could most readily have taken place at a favoured workshop close to the place of its discovery. And in this context, as we have repeatedly noted in dissecting the decoration of the casket, the monastic foundations with which it appears to have the greatest number of links are the two coastal ones of Lindisfarne and Wearmouth-Jarrow, each a wealthy community, distinguished by its influence, and by its output of prestigious manuscripts. The counterclaim to this is of course a strong one. The absence of surviving manuscripts that can be attributed to other, inland, foundations such as Hexham, Ripon and York, is particularly frustrating, given the tantalising references made by Alcuin and Stephen of Ripon to the lavish gifts that were bestowed on some of these wealthy foundations in the late seventh century and first decades of the eighth; these included gold altar vessels, imported silks and a Bible written in gold on purple vellum (a gift from Wilfrid to Ripon). York, in particular, with its noted tradition of scholarship must have produced books and other artefacts of comparable importance and sophistication to those made in the scriptoria at Lindisfarne and Wearmouth-Jarrow with its well-equipped library, and might have provided just the kind of learned context in which the casket could have been produced. The possibility that the imposing manuscript of the Durham Cassiodorus might have been produced at York hints at what else might have been made there. However, I am tempted to resist this alternative scenario, and not just because I suspect that the casket might most naturally have been made in a learned coastal centre not far from where ‘the sea tossed up the fish onto the rocky shore… [and] became sad where he swam aground on to the shingle’. One of the things we know about the casket is that its particular mixture of narratives suggests that it was made in a context where there was knowledge of a universal history – a cosmography, in which events or tales of the past, in all their diversity, formed part of that greater history that culminates in God’s message, embodied in Christ (Webster 2012b, 49‒ 50). The Sack of Jerusalem and the saving of Romulus and Remus are both named

250

Leslie Webster

as themes that were illustrated in these manuscripts. Moreover, Bede – who was certainly interested in such matters – tells us that the learned Northumbrian king, Aldfrith (685‒704), had given land to Wearmouth-Jarrow in exchange for just such a manuscript. If Bede knew about cosmographies and their contents, it is a fair bet that he had used one, or that there was, at the very least, oral knowledge of the nature of such a book at Wearmouth-Jarrow in his day. If we put this intellectual background together with the art-historical evidence which suggests that the casket was made in a centre where the influence of the Roman church was particularly strong, where Roman styles predominated, and where many of the casket’s elements are paralleled in manuscripts and other artefacts associated with that centre, then – in the absence of stronger competing evidence – the case for WearmouthJarrow as the place of origin for the casket appears very strong. That the casket also displays stylistic links with decorated manuscripts associated with Lindisfarne does not contradict this, for we know from Bede and other sources that intellectual contacts between the two foundations, including gifts of texts, flourished in his day (and see above). And indeed, there are suggestive traces of Hiberno-Saxon motifs among the sculpture at Wearmouth-Jarrow which hint that manuscripts decorated in a more Insular style could also have been produced there.13 A date in the opening decades of the eighth century would thus fit comfortably with both the art-historical evidence and the intellectual background – though sadly that would probably be too late to allow the cosmography given to King Aldfrith to have been housed in this splendid presentation box, appropriately decorated as it is with scenes which both encapsulate a universal history and deliver moral guidance for rulers.

References Alexander, Jonathan James Graham. 1978. Insular Manuscripts, Sixth to the Ninth Century: A Survey of Manuscripts Illuminated in the British Isles. Vol. 1. London: Miller. Battiscombe, Christopher Francis (ed.). 1956. The Relics of St Cuthbert: Studies by Various Authors Collected and Edited with an Historical Introduction. Oxford: Oxford University Press for the Dean and Chapter of Durham Cathedral. Bailey, Richard. 2005. Anglo-Saxon Sculptures at Deerhurst (Deerhurst Lecture 2002). Deerhurst: Friends of Deerhurst Church. Breay, Claire, and Brendan Meehan (eds). 2015. The St Cuthbert Gospel: Studies on the Insular Manuscript of the Gospel of John. London: British Library. Brown, Michelle. 2007. Manuscripts from the Anglo-Saxon Age. London: British Library. Cramp, Rosemary. 2006. Wearmouth and Jarrow Monastic Sites. Vol. 2. London: English Heritage. Dumville, David. 2007. “The Two Earliest Manuscripts of Bede’s Ecclesiastical History”. AngloSaxon 1: 55‒108.

13 For example, the Herebericht stone at Wearmouth, and the bird-head from Jarrow (see p. 243 above).

The Franks Casket, Part A: An Art-Historical Perspective

251

Gallica: Evangelia quattuor [Evangiles dits d’Echternach ou de Saint Willibrord] URL: ‹gallica.bnf.fr/ark:/12148›, last accessed 16 April 2012. Gameson, Richard. 2010. Manuscript Treasures of Durham Cathedral. London: Third Millenium. Gameson, Richard. 2013. From Holy Island to Durham: The Contexts and Meanings of The Lindisfarne Gospels. London: Third Millennium. Gameson, Richard. 2015a. “Materials, Text, Layout and Script”. The St Cuthbert Gospel: Studies on the Insular Manuscript of the Gospel of John. Eds. Claire Breay and Brendan Meehan. London: British Library. 13‒39. Gameson, Richard. 2015b. “History of the Manuscript to the Reformation”. The St Cuthbert Gospel: Studies on the Insular Manuscript of the Gospel of John. Eds. Claire Breay and Brendan Meehan. London: British Library. 129‒136. Høilund Nielsen, Karen. 2003. “Ulve, heste og drage: ikonografisk analyse af dyrene i stil II–III”. Hikuin 29: 187. Høilund Nielsen, Karen. 2010. “Style II and All That: The Potential of the Hoard for Statistical Study of Chronology and Geographical Distributions”. Papers from the Staffordshire Hoard Symposium. Ed. Helen Geake. URL: ‹finds.org.uk/staffs hoardsymposium›, last accessed 20 July 2018. Henderson, George and Isabel Henderson. 2010. “The implications of the Staffordshire Hoard for the understanding of the origins and development of the Insular art style as it appears in manuscripts and sculpture“. Papers from the Staffordshire Hoard Symposium. Ed. Helen Geake. URL: ‹finds.org.uk/staffshoardsymposium›, last accessed 20 July 2018. Kitzinger, Ernst. 1956. “The Coffin-Reliquary”. The Relics of St Cuthbert: Studies by Various Authors Collected and Edited with an Historical Introduction. Ed. Christopher Francis Battiscombe. Oxford: Oxford University Press for the Dean and Chapter of Durham Cathedral. 202–304. Napier, A. S. 1901. “The Franks Casket”. An English Miscellany Presented to Dr. Furnivall. Ed. William Paton Ker. Oxford: Oxford University Press. 362–381. Nees, Lawrence. 2011. “Recent Trends in Dating Insular Art”. Insular and Anglo-Saxon Art and Thought in the Early Medieval Period. Ed. Colum Hourihane. Princeton, NJ: Princeton University with Penn State University Press. 14‒30. Pickwoad, Nicholas. 2015. “Binding”. The St Cuthbert Gospel: Studies on the Insular Manuscript of the Gospel of John. Eds. Claire Breay and Brendan Meehan. London: British Library. 41‒64. Speake, George. 1980. Anglo-Saxon Animal Art and its Germanic Background. Oxford: Clarendon. Webster, Leslie. 1982. “Stylistic Aspects of the Franks Casket”. The Vikings. Ed. Robert T. Farrell. Chichester: Phillimore. 20–31. Webster, Leslie. 1999. “The Iconographic Programme of the Franks Casket”. Northumbria’s Golden Age. Eds. Jane Hawkes and Susan Mills. Stroud: Sutton Publishing. 227‒246. Webster, Leslie. 2003. “Encrypted Visions: Style and Sense in the Anglo-Saxon Minor Arts, A.D. 400‒900”. Anglo-Saxon Styles. Eds. Catherine E. Karkov and George Brown. Albany, NY: State University of New York. 11‒30. Webster, Leslie. 2012a. Anglo-Saxon Art: A New History. London: British Museum Press. Webster, Leslie. 2012b. The Franks Casket. London: British Museum Press. Webster, Leslie. 2015. “Decoration of the Binding”. The St Cuthbert Gospel: Studies on the Insular Manuscript of the Gospel of John. Eds. Claire Breay and Brendan Meehan. London: British Library. 65‒82. Webster, Leslie. 2016. “Imagining Identities; the Case of the Staffordshire Hoard”. Anglo-Saxon England and the Visual Imagination (Essays in Anglo-Saxon Studies, 6). Eds. John D. Niles, Stacy S. Klein, and Jonathan Wilcox. Tempe, Arizona: Arizona Center for Medieval and Renaissance Studies. 25‒48. Webster, Leslie and Janet Backhouse (eds). 1991. The Making of England: Anglo-Saxon Art and Culture AD 600‒900. London: British Museum Press. Wright, David. 1961. “Some Notes on English Uncial”. Traditio 17: 441‒456.

Gaby Waxenberger

B. The Franks Casket and its Inscriptions 1 Full View of the Franks Casket

Fig. 1: Full view of the Franks Casket; © The Trustees of the British Museum.

Note: See fn. 1 in Kazzazi, this volume, “Introductory Remarks on the Contributions by Leslie Webster and Gaby Waxenberger on the Franks Casket”. https://doi.org/10.1515/9783110796834-013

254

Gaby Waxenberger

2 The Fragmentary Lid of the Franks Casket

Fig. 2: The fragmentary Lid of the Franks Casket; © The Trustees of the British Museum.

ægili1 Ægili (masc. personal name)

1 For the graphic conventions used in this article see → Transliteration in “A Concise and Selected Guide to Terminologies” in this volume.

The Franks Casket, Part B: The Franks Casket and its Inscriptions

255

3 The Front Panel of the Franks Casket

Fig. 3: The Front Panel of the Franks Casket; © The Trustees of the British Museum.

fisc·flodu· ahofonferg enberig warþga:sricgrornþærheongreutgiswom← hronæsban The caption: mægi

(autopsies: 2001, 2005)

256

Gaby Waxenberger

3.1 The inscription of the Front Panel is fraught with three major difficulties:

3.1.1 The unexpected ending -u in flodu (flōd; DOE online 2018, s. v. flōd: orig. masc. ustem; later mainly masc. and neutr. a-stem). -u should have been lost after a long syllable at the beginning of the 7th cent. (Luick 1921, § 350).

3.1.2 The dots after fish and flodu are, in my opinion, not word dividers but have some other function.

3.1.3 The meaning of fergenberig is not quite clear. Before I discuss these problems, I would like to point out that, in my view, the texts of the top lines on the Front and the Back panels show abbreviated forms; this may have to do with the fact that there was less space on the Front Panel due to the originally extant lock and on the Back Panel because the inscription panel was divided by the image. I would also like to underline Leslie Webster’s observation (see Part A, this volume) and her comment that “no other Northumbrian artefact – not the Lindisfarne Gospels, not the Ruthwell Cross – has the kind of densely crammed narrative of text and image which is such a distinctive signature of the casket”. In order to fit the text into the panels, the runes are – with the exception of the Back Panel – normally relatively high and narrow in shape. According to Bammesberger (1991, 632) and also Napier (1901, 370), there is an abbreviated word on the Back Panel: giuþeasu (= Iudea su[m(æ)]). There is not enough space for an -m.

Fig. 4: Top line of Back Panel; © The Trustees of the British Museum.

On the Front Panel the sequence flodu· is also incomplete (cf. Waxenberger forthc., Chap. 2 no. 3a).

The Franks Casket, Part B: The Franks Casket and its Inscriptions

257

Fig. 5: Top line of Front Panel; © The Trustees of the British Museum.

Normally, dots are used as word-dividers but there seems to be a more specifically meaningful, but as yet enigmatic system of dots on the Franks Casket. In my opinion, the dots on the Front Panel may indicate runes that could not be inserted for lack of space. For a possibly different function see below 3.4, Possibility 3. Regarding points 3.1.1 and 3.1.2, two possibilities of parsing the sequence flodu· are imaginable.

3.2 Possibility 1 In my reading (Waxenberger 2006), I adapted Bammesberger’s (1991, 632) hypothesis for giuþeasu, in which he assumes that the rune carver’s original template may have been written in Latin characters and it may therefore have contained an abbreviation ‹v̄›; ‹ū› for ‹vm›; ‹um› (see SB 1965, § 4 note 4). In this case, the sequence giuþeasu would have been intended as giuþea su*[m] ‘someone of the Jews’ (with ‹u› representing the ending -um). I suggest the same approach for the sequence flodu, which would then have been intended as flodu*[m] (flōdum). So, I read: fisc flōdu (= flōdum) ahōf on fergenberig In this constellation, the noun fisc would be the subject and flōdum an instrumental in the form of a dat. pl. In this case, the interpretation of āhebban is problematic, as this verb is usually used transitively (see DOE online 2018, s. v. ahebban), whereas it would be intransitive here. However, there is in fact one example of intransitive use of hebban by Wulfstan, according to DOE online (2018, s. v. hebban 12. intransitive: up hebban ‘to rise, mount’).2 So, the Franks Casket may show an attestation of the intransitive use of āhebban. In this case,

2 “HomU 58 49: and ðær eall þæt folc on locode, he stah up to ðam stepele and of ðam stepele hof upp on lyfte, swylce he wolde wið þæs heofones weard” (Napier 1883, 98).

258

Gaby Waxenberger

fisc·flodu·ahof could be read as fisc flōdu (= flōdum) ahōf (…) and translated as follows: ‘The fish arose by means of the floods/high tide (…)’.

3.3 Possibility 2 As can be inferred from the DOEW, the verb āhebban can be accompanied by a reflexive pronoun.3 Taking this possibility into consideration, I assume that the dot after fisc· may indicate the reflexive use of the verb, as an expected pronoun, such as hinæ, could not be added due to lack of space. So, the sequence could be interpreted as: fisc·flodu·ahof fisc [hinæ] flōdu[m] ahōf on fergenberig ‘The fish arose (= raised himself) by means of the floods/high tide (…)’.

3.4 Possibility 3 However, it may be possible to parse flodu· differently: u may not stand for a caseending but it may be the first letter of the preposition up, in which case the missing p would be indicated by the dot. In this case, the parsing would be: fisc flod u·ahof on fergenberig fish flōd u[p] āhōf on fergenberig ‘The high-tide raised up the fish …’ In this case, the dot after fisc· may have been a ‘reading aid’ for the beholder, as neither fisc (masc. a-stem: DOE online, s. v. fisc) nor flod (if it is declined according to a masc. a-stem: see DOE online, s. v. flōd) are accompanied by, for example, a definite article to disambiguate the case (nom. sg. or acc. sg.). Regarding point 3.1.3 above, it can be surmised that the sequence fergenberig is a compound consisting of the two elements fergen- and -berig, with the latter showing a parasite vowel -i- and Anglian smoothing. The WS form would be beorg ‘mountain, hill’ (DOE online 2018, s. v. beorg). As the environment for breaking is the same in the first element, I also assume a smoothed Anglian form fergen- for WS *feorgen-, which could be linked to OE feorh ‘life’ and possibly ‘world’. Bammes-

3 E.g., Bede, HE: “Ond æfter þon he hine gereste medmicel fæc, ða ahof hine up & ongan aweg gan, gif he hwær ænigne freond metan meahte, þe his gymenne dyde & his wunda læcnian wolde” (DOEW [0449 (23.326.9)]; Miller 1890–1898, 326).

The Franks Casket, Part B: The Franks Casket and its Inscriptions

259

berger (1991, 630) suggests Gmc. *ferg(w)-an-a, “which could be ultimately connected with Goth. fairhwus ‘world’ and also *ferguni”. For Bammesberger, “some concept like ‘world, life’ would also be thinkable” (see also Holthausen 1974, 101).

3.5 The complete top line and the sequence of the right side of the front panel: The inscription fisc·flodu·ahofonfergenberig could be parsed and interpreted as

3.5.1 fisc flōdu[m] ahōf on fergenberig fisc flōdu[m] ahōf on fergenberig ‘The fish arose by means of the floods/high tide onto the world/life-mountain (= cliff; world).’

3.5.2 fisc [hinæ] flōdu[m] ahōf on fergenberig fisc [hinæ] flōdu[m] ahōf on fergenberig ‘The fish arose (= raised himself) by means of the floods/high tide onto the world/ life-mountain (= cliff; world).’

3.5.3 fisc·flod u·ahof on fergenberig fish flōd u[p] āhōf on fergenberig ‘The flood/high tide raised/lifted up the fish onto the world/life-mountain (= cliff)’

3.6 The bottom line of the front panel warþga:sricgrornþærheongreutgiswom← warþ gās(t)rīc grorn þær he on greut giswom ‘Gās(t)rīc was sad, as (= when) he swam on to the sand (= aground).’ [Waxenberger 2006, 306]

260

Gaby Waxenberger

I translate gās(t)rīc (= gāst ‘breath, soul, spirit, life’; rīce ‘strength, power’ ja-stem) as ‘the one who has the power of breath/life’. Since the compound must refer to he, I consider it to be a name for the whale. rīce appears as -rīc in second elements of compounds (Campbell 1959, § 348, n. 5). As the paralinguistic character : would be a very unusual word-divider since neither ga nor sric are words (see above Figure 3), I regard this sign as a hint for the reader. In my opinion, the double dot served two purposes: first, it identified t as the rune that had to be inserted and second it indicated the position of this rune (= immediately after the s) in the sequence in question. In the fuþorc row the rune t (no. 17) follows the s (no. 16; see also below Guide to Terminologies Figure 4). For this reason the carver had to put in one dot (= s*) for s + 1 = t. Another dot, however, became necessary because there was no space after the s as the runes of this line run from right to left, thus facing to the left. The complete lack of space in this line (see above Figure 3) not only forced the carver to leave out the character t but also forced him to make use of space where it was available, namely between the lower side-twig of the a and the right upper stave of s ᛲ .

3.7 The left line of the front panel hronæsban hronæs bān ‘Whale’s bone/s (bones = skeleton).’ The caption: mægi magi ‘the Three Wisemen’

4 The Left Panel of the Franks Casket romwalusandreumwalustwœgen gibroþær afœddæhiæwylifinromæcæstri oþlæunneg

(autopsies: 2001, 2005)

I generally follow Bammesberger (1998, 15) in his reading and translation (see below) but parse the sequence wylifinromæcæstri as wylif/wylifin in romæcæstri (see Waxenberger forthc., Chap. 6 INDIVIDUAL RUNES 6.1.3 rune i 8.3).

The Franks Casket, Part B: The Franks Casket and its Inscriptions

261

Romwalus and Reumwalus twœgen gibroþær afœddæ hiæ wylif/wylifin in Romæcæstri oþlæ unneg

‘Romulus and Remus (were) two (twin) brothers – a she-wolf raised them – in the city of Rome far from the native land’ (cf. Bammesberger 1998, 19).

Fig. 6: The Left Panel of the Franks Casket; © The Trustees of the British Museum.

262

Gaby Waxenberger

5 The Back Panel of the Franks Casket

Fig. 7: The Back Panel of the Franks Casket © The Trustees of the British Museum.

herfegtaþ titusendgiuþeasu4 HICFUGIANTHIERUSALIM5 afitatores Captions: dom

gisl

Hēr fegtaþ Titus end giūþea su[m] Hic fugiant (= Class. Lat. fugiunt) Hierusalim afitatores (Class. Lat. habitātōres)

(Bammesberger 1991, 632) (Page 1995, 311, 1999, 177)

4 For the interpretation of the sequence giuþeasu see above no. 3.2 Possibility 1. 5 It may be debated whether S is a rune s or an uncial. I read it as an uncial because all other s-runes are either ᛋ→ or ←ᛋ.

The Franks Casket, Part B: The Franks Casket and its Inscriptions

Captions: dōm

263

gīsl

For the sequence titusendgiuþeasu I follow Bammesberger’s translation: ‘Here Titus and a Jew (= someone of the Jews) fight’ or ‘here Titus and some Jews fight’. For the sequences herfegtaþ, HICFUGIANTHIERUSALIM, and afitatores I follow Page (1995, 311, 1999, 177): ‘Here the/its inhabitants flee from Jerusalem’. Captions: ‘judgement’

‘captive’

Fig. 8a: The Right Panel of the Franks Casket; © The Trustees of the British Museum.

264

Gaby Waxenberger

Fig. 8b: The original fragment of the Right Panel in the Museo Nazionale del Bargello, Florence, Italy; © Museo Nazionale del Bargello, Florence.

The Right Panel has cryptic runes for the vowels (cf. Waxenberger forthc., Chap. 10).

risci wudu

bita

I read and translate the cryptic panel as follows:6 Her hos sitaþ on harmberg a ægl[æ(ca] drigiþ swah iri ertæ egi sgræf sar den sorga ænd se(fa torna Hēr hōs sittaþ on harmberg ā. ǣglǣca drīgiþ. swāh (= swāc)

6 For my linguistic analysis see Waxenberger (forthc., Chap. 10).

The Franks Casket, Part B: The Franks Casket and its Inscriptions

265

irri. E/Ērtæ egi sgrǣf (= scrǣf) sār, den sorga ænd sefa torna. ‘Here7/listen, the company preside/s [3rd pers. sg. pres. or 3rd pers. pl. pres.] over the harmful burial place/mound [acc.] (always). The awesome opponent/ferocious fighter always performs/acts/endures. Anger has left. Ertæ assigned/decreed by means of the horse: distress, the grave of sorrows, and the sad mood’.

References Bammesberger, Alfred. 1991. “Franks Casket: Editor’s Notes”. Old English Runes and their Continental Background. Anglistische Forschungen 217. Ed. Alfred Bammesberger. Heidelberg: Winter. 629–632. Bammesberger, Alfred. 1998. “The ‘Romulus Plate Inscription’ on the Franks Casket”. Amsterdamer Beiträge zur Älteren Germanistik 50.1: 13–20. DOE online = Dictionary of Old English: A to I Online. Eds. Angus Cameron, Ashley Crandell Amos, Antonette diPaolo Healey et al. Toronto: Dictionary of Old English Project, 2018. URL: ‹www.doe.utoronto.ca/›, last accessed 20 June 2021. DOEW= The Dictionary of Old English Web Corpus. Compiled by Antonette diPaolo Healey with John Price Wilkin and Xin Xiang. Toronto: Dictionary of Old English Project, 2009. URL: ‹www.doe.utoronto.ca/›, last accessed 20 June 2021. Holthausen, Ferdinand. 1974. Altenglisches etymologisches Wörterbuch. 3rd. ed. Heidelberg: Winter. Luick, Karl. 1921. Historische Grammatik der englischen Sprache: Band I, 1. Abteilung. Leipzig: Tauchnitz. Miller, Thomas (ed.). 1890–1898. The Old English Version of Bede’s Ecclesiastical History of the English People. 4 vols. EETS 95, 96, 110, 111. London: OUP. Napier, Arthur S. 1883. Wulfstan. Sammlung englischer Denkmäler 4. Berlin. Repr. with an Appendix by K. Ostheeren, 1967. 98–101. Napier, Arthur S. 1901. “Contribution to Old English Literature”. An English Miscellany Presented to Dr. Furnivall in Honour of his Seventy-Fifth Birthday. Eds. William P. Ker and Arthur S. Napier. Oxford: Clarendon. 355–381. Page, Raymond Ian. 1995. “Runenkyndige Risteres Skriblerier: The English Evidence”. Runes and Runic Inscriptions, Collected Essays on Anglo-Saxon and Viking Runes. Ed. David Parsons, with a Bibliography by Carl T. Berkhout. Woodbridge: The Bodyell Press. 295–314. Page, Raymond Ian. 1999. An Introduction to English Runes. 2nd ed. Woodbridge: The Boydell Press. SB = Sievers, Eduard and Karl Brunner. 1965. Altenglische Grammatik. Nach der angelsächsischen Grammatik von Eduard Sievers. 3rd ed. Tübingen: Niemeyer.

7 The meaning ‘here’ was suggested by Mary Blockley in an unpublished paper, “Speech Acts and Inscriptions: The Syntax of the Right Side of the Franks Casket”, given at the 7th International Symposium on Runes and Runic Inscriptions in Oslo, 2010.

266

Gaby Waxenberger

Waxenberger, Gaby. 2006. “The Representation of Vowels in Unstressed Syllables in the Old English Runic Corpus”. Das ältere Fuþark und seine einzelsprachlichen Weiterentwicklungen. RGA-E 51. Eds. Alfred Bammesberger and Gaby Waxenberger. Berlin/New York: De Gruyter. 272‒314. Webster, Leslie. 2012. The Franks Casket. London: British Museum Press.

Gaby Waxenberger

C. The Dating and Provenance of the Franks Casket: A Linguistic and Runological Perspective 1 Introduction On a basis of only five linguistic criteria each for the date and the provenance of the Franks Casket, Napier (1901, 380 f.) stated in 1901 that the “most likely date” for the Franks Casket was the “beginning of the eighth century” and its home was “the coast of Northumbria”.1 Napier is basically right, but the evidence he uses does not support some of his claims. More importantly, he neglected other kinds of linguistic evidence that would in fact have supported his conclusions. I have discussed Napier’s (1901) arguments elsewhere (Waxenberger 2017), but in the present contribution some of my analyses and discussion points are set out in a more detailed way and have been extended to other objects (see below 5.5). I have also introduced tables of applicability/relevance so that the reader can see at one glance the reliability of Napier’s criteria. In order to obtain a more reliable and also more accurate result, it is necessary to examine the runic context, for which not only the language but the runes themselves must be taken into account. The Nhb. sub-corpus of the Old English Runes Corpus (OERC) is the direct runic context for the Franks Casket. It covers a period from the late 7th to the 10th/(11th) cent. (see Map 1 below).2 Where the runic data is insufficient for my analysis, I have consulted non-runic sources but limited my analysis to the Ms. L of Cædmon’s Hymn3 for two reasons: first, it has been dated to the period in question (= 8th cent.) and second, it is clearly Nhb. and has even been assigned to a certain place of origin: Wearmouth-Jarrow (Lowe 1958; Schapiro 1958; Parkes 1994). However, methodologically, two things must be kept in mind: firstly, Map 1 is merely a map of find-spots, which, especially for small objects, may not be the place 1 Elliott’s (1989, 138 f.) comment shows the tremendous influence of Napier’s dating on later studies of the casket: “The date and provenance of the Franks casket have been established beyond doubt by Napier’s [1901] linguistic analysis. The language is unmistakably Anglian and certain forms limit it further to Northumbria, and, in point of time, to the early eighth century.” 2 I did not plot the inscriptions found on the west coast (St Ninian’s Cave: […]ate; Mote of Mark: aþili; Whithorn Stone I: [...]ferþs[; Whithorn Stone II: [..]r[.]hwitu […]) because they are too fragmentary and – with the exception of the Mote of Mark Bone – too late. 3 The Moore Ms. (M = Cambridge University Library) is dated to probably 737 or possibly between 734–737 (Smith 1978, 2; 19 ff.), and the Leningrad Ms. (L = Leningrad Public Library) is dated to 746 (Smith 1978, 2; 19 ff.); see also SB (1965, § 2). See Webster, this volume, for Dumville (2007) and Gameson (2015, 25, 27, 33), who prefer a date in the 3rd quarter of the 8th cent. https://doi.org/10.1515/9783110796834-014

268

Gaby Waxenberger

Map 1: Runic inscriptions in Northumbria (late 7th–10th/11th cent.); map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München.

of origin. Secondly, very often inscriptions cannot be dated more narrowly than for a timespan of a couple of centuries. However, there are two objects still in situ: the Ruthwell and Bewcastle crosses. The language of the Ruthwell Cross (RC) is definitely Nhb., and Page (2006) has convincingly dated the runes to ca. AD 750 (see also Waxenberger 2006 for the unstressed vowels, which corroborate Page’s dating; see, however, Hines, this volume). Hence these two points provide a reliable basis for my further investigation regarding date and place of origin of the Franks Casket.

2 Napier’s Criteria for Dating the Casket In the following sections, Napier’s criteria for dating the casket will be discussed; but, as Table 1 shows, only two of his criteria are applicable/relevant and only one of them yields an unambiguous result; the other criteria are accordingly shaded in grey in the following table.

269

The Franks Casket, Part C: A Linguistic and Runological Perspective

Tab. 1: Napier’s criteria for dating the casket. Napier’s criteria for dating the casket

Applicability/Relevance

Comment

Full vowels in unstressed syllables (see below 2.1)

applicable/relevant

‹eu› (< Gmc. *eu) for the class. OE diphthong ēo (see below 2.2)

applicable/relevant

The spelling ‹eu› also appears in a few words of the Épinal/Erfurt Glossary. This criterion supports the assumption of an early date: see the discussion below 2.2.

Use of the rune ‹f› instead of ‹b› to denote the allophone [v] (see below 2.3)

not applicable to runic writing

In runic spelling ‹f› represents the fricatives OE [f] and [v] and ‹b› only represents the plosive OE [b].

Nhb. loss of -n in Napier’s form “sefu” (see below 2.4)

not applicable because Napier’s form is incorrect

The sequence in question reads se(fa.

Final -u in the word flodu (see below 2.5)

not applicable for phonological and chronological reasons

-u in final unaccented syllables after a long, accented syllable should have been lost by the beginning of the 7th cent. (Luick 1921, § 350).

In the following sections, Napier’s criteria listed in the table above will be discussed in more detail.

2.1 Full Vowels in Unstressed Syllables As can be inferred from Table 2, the Franks Casket is the only object in the OERC proper (ca. AD 650–11th cent.) whose inscriptions show consistent use of the runes -i and -æ (instead of -e), suggesting that the full vowels /ɪ/ and /æ/ were still in place in unstressed syllables. Tab. 2: The distribution of unstressed -æ and -i vs. -e (cf. Waxenberger forthc. for the complete list of OE inscriptions). Inflectional endings: early -æ vs. later -e

Inflectional endings: early -i vs. later -e

Particles, Prefixes: early -i vs. later -e; ni-, bi- vs. ne-, be-; gi-, ge-

Franks Casket

-æ: 4×

-i: 2×

gi-: 2×

RC: ca. 750

-æ: 10×

Great Urswick Stone: -æ: 5× ca. 750–9th cent.

-e: 1×

ni-: 1× bi-: 2× gi-: 2× ge-: 1×

270

Gaby Waxenberger

Tab. 2 (continued) Inflectional endings: early -æ vs. later -e Thornhill Stone I:

Inflectional endings: early -i vs. later -e

Particles, Prefixes: early -i vs. later -e; ni-, bi- vs. ne-, be-; gi-, ge-

-i: 1×

ca. 750–9th cent. Thornhill Stone II

-e: 2×

ca. 750–9th cent. Thornhill Stone III:

-e: 2×

-i: 1×

-e: 1×

ge-: 1×

ca. 750–9th cent.

2.2 ‹eu› (< Gmc. *eu) for the Classical Old English Diphthong ēo Franks Casket Front Panel: greut ‘grit, sand’ (DOE online s. v. grēot ‘grit, gravel, sand, shingle (of the seashore)’). Napier (1901, 380) uses the three ‹eu› spellings in the Épinal Glossary “as compared with about six times as many of the later” ‹eo›, ‹io› spellings to date the Franks Casket. Regarding the dating of the Épinal Glossary he follows Chadwick (1899, 240), who dates it to “about 730 at the latest”.4 For Hogg (1992, § 2.33) “a few early spellings with ‹u› as second element” suggest an earlier pronunciation /eu/. For long ēu (< Gmc. *eu) Hogg lists EpGl 940 “stēupfæder ‘stepfather’” (EpGl 1070 “uitricius steupfaedaer”: Pheifer 1974, 56) and EpGl, ErfGl 726 “trēulesnis ‘faithfulness’”.5 In the OERC, there is no other spelling with ‹eu› but instead only two more OE runic inscriptions showing the classical OE spelling with ‹eo› (< Gmc. *eu). Including the data of the Épinal Glossary (Épinal, Bibliothèque municipale 72 (2), fols. 94– 197), which is now dated to ca. AD 700 (Herren/Sauer 2016, 129; 136), i.e., a little earlier than the dating by Chadwick taken as a basis by Napier, it may be concluded that the spelling ‹eu› indicates an early date. It should, however, not go unmentioned that the ‹eu› spellings form the minority, ‹eo› representing the majority of the spellings in the Épinal/Erfurt Glossaries. This fact may suggest that the ‹eu› spellings go back to Épinal/Erfurt’s common archetype EE, which “would more likely have been made in the decade 680‒690 than in the 670’s” (Herren/Sauer 2016, 143 f.). If this is the case, it would explain the relatively small percentage of ‹eu› spellings as conservative spellings, whereas the larger part of the spellings was modernised by the scribes.

4 For an overview of the forms in Épinal/Erfurt see also Pheifer (1974, § 41). 5 “perfidia treulesnis” (Pheifer 1974, 39).

The Franks Casket, Part C: A Linguistic and Runological Perspective

271

Tab. 3: Attested forms of -eo- (WGmc. *eu) in the OERC. Findspot Mercia

Inscription Brandon Antler Piece (from a Middle Saxon context, ca. AD 700–850; Page 1999, 80): wildumde[.]r)an (dēoran = dēor an ‘on a wild animal’)

Italy

Monte Sant’ Angelo Inscription E (ca. AD 650–850; Waldispühl 2018–2019, 136): leofwini (masc. PN; Searle 1897, 334 ff.)

Considering ca. AD 700 for the dating of the Épinal manuscript and the fact that the majority of the attestations there already have the regular OE spelling ‹eo›, it may be argued that the Franks Casket inscription was made not much later than AD 700, although there may be a time lapse between the different dialects. For in contrast to the Franks Casket, Épinal is generally considered Mercian (SB 1965, § 2 note 6; Pheifer 1974, § 89; Sauer/Waxenberger 2013, 348; Herren/Sauer 2016, 144). In conclusion, Napier’s criterion is relevant for the label “early”, but the relatively wide datings of the other two attested runic forms (see Table 3 below) make determining the end of the ‹eu› period difficult. Additionally, it must be seen that there is no reference data of ‹eu›, such as a second and later form, in the OERC. Thus, keeping in mind that only an estimate ‘around AD 700’ but no end for ‹eu› spellings and therefore no cover period can be inferred from the data, this spelling nevertheless supports the general tendency towards the early 8th cent. Although the second element o of the diphthong ēo on the Brandon Antler Piece is hardly recognisable, some very faint visible remains make the assumption of a spelling ‹eo› (rather than ‹eu›) fairly probable.

2.3 Use of Rune f to Denote the Allophone [v] of the Phoneme OE /f/ In contrast to early Nhb. manuscript use, in which both the graphemes ‹f› (hefænrice ‘the kingdom of heaven’) and ‹b› (heben ‘heaven’) are used to denote the allophone [v], the Franks Casket shows only rune f (se(fa ‘mind’ and possibly wylifin ‘shewulf’ as haplography in the sequence wylifinromæcæstri parsed as wylifin in romæcæstri; cf. Bammesberger 1998a, 15; Waxenberger forthc., Chap. 2B no. 3c; see also Part B no. 4 in this volume) to denote the allophone [v]. As can be inferred from Table 4, the rune b is exclusively used for the plosive in the Pre-OERC and OERC.

272

Gaby Waxenberger

Tab. 4: Attested distribution of the runic graphemes ‹f› and ‹b› representing the allophones [f], [v], and [b]. ᚠf

ᛒb

Representing [f]

Representing [v]

Representing [b] exclusively

Franks Casket: fisc; flodu; ahof; fergenberig afœddæ; fegtaþ Collingham Stone: æft[..] Falstone Stone: æftær (2×) Monte Sant’ Angelo A: wigfus Lancaster Cross: foræ Monkwearmouth Stone I: tidfirþ Overchurch Stone: folcæ; fote (= fore) Rome Catacomb ad Duas Lauros: faᚼhild; æþelferþ; RC NE: fusæ; f(earran Thornhill Stone I: æfter Thornhill Stone II: æfte[?] Thornhill Stone III: æft[.] Whithorn Stone I: [...]ferþs[ (second element of a personal name)

Franks Casket: wylifin; se(fa RC SE: h(eafunæs; hlafard

Franks Casket: fergenberig; hronæsban; gibroþær; bita Bewcastle: kyniburuᚸ Caistor-by-Norwich Brooch: gibœtæ Crowle Stone: ][…]cbæcun[...][ Falstone Stone: gebidæd Great Urswick Stone: bekun; bæurnæ; gebidæs Lancaster Cross: gibidæþ; cynibalþ; cuþbere[.] London National Portrait Gallery Bone: tatberht London Thames scramasax: b(eagnoþ Loveden Hill Urn: sᛇþᚨbᚨd [sĩ:ð{a/æ}bæd] Maughold Stone I: blagc[?]mon Monte Sant’ Angelo Graffiti D: }her(eb(er(ehct Overchurch Stone: bec[; biddaþ Rome Cimitero di Commodilla Graffito: (ea(dbald RC SE: ba; blodæ; bistemi[.] Thornhill Stone I: [.]þelbe[..] Eþelberht Thornhill Stone III: berhtsuiþe; bekun; bergi; gebi)ddaþ Watchfield Purse Mount: hᚨribᛟ^i [hɛrıbœ:ḵı]

Conclusion The use of ‹b› for [v] in the Latin script cannot be taken as a criterion here, because epigraphical runic writing does not share this feature with the non-runic manuscript tradition (cf. Hogg 1992, § 7.55): the runic grapheme ‹f› is exclusively applied to the OE fricatives [f] and [v], and the runic grapheme ‹b› is only used for the OE voiced bilabial plosive [b].6

6 The rune ‹f› in afitatores (Back Panel) may also represent [v] (cf. habitatores ‘inhabitants’ in Class. Latin); see also Elliott (1959, 101 f.).

The Franks Casket, Part C: A Linguistic and Runological Perspective

273

2.4 Northumbrian Loss of -n in Napier’s Form “sefu” Napier’s (1901, 380) assumption that there is a loss of -n in “sefu” allows him to attribute the inscription to the “first half of the eighth” century. However, the sequence in question stands on the cryptic panel and reads

© The Trustees of the British Museum.

se(fa

sefa, a masc. n-stem ‘mind, spirit, heart; understanding’ (BT 1898, 856; CH 1960, 301). Therefore, this criterion of Napier’s is not relevant because it is based on a non-existent form. Nevertheless, in early Nhb. there are already attestations where -n had been dropped (Campbell 1959, § 472; SB 1965, §§ 188.2; 276 n. 6; Waxenberger 2006, 300 f.). On the basis of this evidence, se(fa may represent the nom. sg. but it may also represent the gen./dat./acc. sg. or the nom./acc. pl. without -n and is therefore also of no use as a criterion.

2.5 Final -u in the Word flodu Napier’s (1901, 380) most problematic argument is that the final -u in the word flodu is a plural form and therefore “the preservation of the u […] points to a date not later than the end of the seventh century”.7 In fact, -u in that position (= final unaccented syllables after a long, accented syllable) should have been lost by the beginning of the 7th cent. (Luick 1921, § 350). There has been much controversy over this form, but I consider it to be a dat. pl. with final -m not expressed in writing (flodu = ‹flodv̄› in the rune carver’s original text in Latin characters, which was intended as flōdum: Waxenberger 2006, 305 f.): for a more detailed explanation see Part B 3.2, Possibility 1. Napier’s (1901, 381) statement that the “most likely date” for the Franks Casket is “the beginning of the eighth century” is based on the criteria discussed above. His conclusion, although basically correct, as will be shown, rests on shaky

7 In his footnote 2, Napier (1901, 380) comments: “I attach great weight to the preservation of u in flodu. This form cannot have been copied from an older original, as the inscription on this side was evidently composed for the occasion, viz. the stranding of the whale. This shows that it cannot be much later than 700.”

274

Gaby Waxenberger

grounds, however. Now, I would like to turn to Napier’s criteria for determining the provenance of the Franks Casket.

3 Napier’s Dialect Criteria The overview in the following table shows that only two of Napier’s five criteria are applicable/relevant. Tab. 5: Napier’s (1901, 379 f.) dialect criteria. Napier’s dialect criteria

Applicability/Relevance Comment

Anglian smoothing (see below 3.1)

applicable/relevant

Absence of diphthongisation applicable/relevant after palatals: WS ea vs. Nhb. æ (see below 3.2)

The dialect can clearly be narrowed down to Anglia. Corroboration of an Anglian provenance, see also below 3.2.

Napier’s criteria specifically for Northumbria 1. The sequence reads se(fa rather than “sefu”. 2. It would be dangerous to use sefa (masc. n-stem) ‘mind’ as an example for the loss of -n as the noun could, at least theoretically, stand in the nom. sg. and thus regularly not contain final -n.

1. Loss of inflectional -n in “sefu” (see below 3.3.1)

not applicable

2. The parasite vowels in the second element of fergenberig and in wylif/wylifin8 ‘she-wolf’ (see below 3.3.2)

not applicable because The limitation of parasite vowels it is not entirely certain to exclusively early Nhb. can neither in its restriction to Nhb. be wholly justified by OE grammars [Campbell 1959, § 360; SB 1965, § 164; Hogg 1992, § 6.35] nor by the OERC itself. See below 3.3.2

3. Second fronting OE æ = Merc. e and OE a = Merc. æ in the Mercian Vespasian Psalter (see below 3.3.3)

not applicable/ not relevant

The lack of this sound change does not restrict the casket to Nhb., but merely excludes that part of Mercia where the Vespasian Psalter was written.

In the following sections, Napier’s criteria listed in the table above are discussed in detail: 8 For the reading wylifin see also above no. 2.3 and Part B no. 4.

The Franks Casket, Part C: A Linguistic and Runological Perspective

275

3.1 Anglian Smoothing The dialect can clearly be narrowed down to Anglia by the phenomenon of Anglian smoothing: the following three forms -berig, unneg, and fegtaþ are reliable pieces of evidence. In contrast, “fergen” (as the first element of fergenberig) is suspect evidence on the grounds of its uncertain etymology, “drigiþ” may never have had a diphthong,9 and berga (see Waxenberger forthc., Chap. 10), an ambiguous form on the cryptic panel, can be disregarded as the Front Panel has an unambiguous attestation of the word for ‘cliff, mountain’ (-berig in fergenberig). I concur that Angl. smoothing does limit the text to Anglia, as can be inferred from Table 6 below. Tab. 6: Anglian smoothing of eo/ēo.10

Anglian smoothing of eo > e

Anglian smoothing of ēo > ē (Campbell 1959, § 227) (= Nhb. ‹ea›)

Angl./(Nhb.) inscriptions

Franks Casket

Thornhill Stone III: bergi (Angl. berg- vs. beorg)

Front Panel: fergenberig (Angl. berh vs. beorg) Left Panel: fegtaþ (Angl. fehtaþ vs. feohtaþ) Left Panel: unneg (Angl. nēh < Angl. *nēoh)

3.2 Absence of Diphthongisation after Initial Palatals: Nhb. æ vs. WS ea The ea/ēa diphthongs due to influence of initial palatal consonants are a regular change in WS but unknown to all Kent. and Merc. texts (Campbell 1959, § 187). However, there are a few forms with ea from æ (e.g., ċeaster beside ċæster) in the late Merc. dialect of the interlinear glosses of the Rushworth Gospels1 (Campbell 1959, § 185; SB 1965, § 91; Hogg 1992, § 5.52).

9 The diphthong would have been brought about by i-umlaut. Normally, i-umlaut should have been carried through in the 3rd pers. sg. ind., but in Nhb. there is “practically never syncope, and the mutation of the root vowel is levelled away” (Campbell 1959, § 733(a); see also SB 1965, § 371, note 5 for Angl. texts and note 7 particularly for Nhb. texts). 10 See Waxenberger (forthc., Chap. 4 A5.2.1) for the dialect criteria of smoothing, according to which -berht in names is not smoothing proper. Thus, names with -berht as a second and as a first name element are not considered in the corpus of the other Nhb. inscriptions in Table 6. Note: Diphthongisation after initial palatals is caused not only by original palatal j, but by the new palatals which arose from Gmc. *k and *g in OE before all front vowels which still existed after the restoration of a before following back vowels (Campbell 1959, § 185).

276

Gaby Waxenberger

Tab. 7: Lack of diphthongisation after palatals in Nhb. inscriptions: æ vs. ea (Waxenberger forthc., Chap. 4.7A/B2.0). Franks Casket

Ruthwell Cross

Left Side: romæcæstri

RC Upper East: gisgæf[.] = gescæft (WS gesceaft ‘creature’: Campbell 1959, § 185)

Regarding Nhb., Campbell (1959, § 186) and SB (1965, § 92.e) confirm the diphthongisation only for late Nhb.,11 whereas Hogg (1992, § 5.51) differentiates between NorthNhb. (“including the e[arly] N[h]b. texts” (the Franks Casket) and also the late texts, such as the late Old English interlinear glosses of the Lindisfarne Gospels and the Durham Ritual), and South-Nhb. (the late interlinear glosses of the Rushworth2 Gospels), generally observing “a large number of spellings indicating diphthongization” only for the late North-Nhb. texts. The absence of the diphthong after an initial palatal in Franks Casket’s cæstri is further corroboration of an Anglian provenance, as Napier suggested. What is more, in the Nhb. runic sub-corpus, the absence of diphthongisation after initial palatals is also found on the RC as can be inferred from Table 7 above.

Conclusion Together with the other criteria for Nhb. in Section 4 below, in my opinion, this feature corroborates the restriction of the Franks Casket inscriptions to early Northumbria whereas Napier considered it merely valid for the wider Anglian dialect area.

3.3 Napier’s (1901, 379 f.) Criteria for Northumbrian 3.3.1 As I have already pointed out (see above 2.4), the loss of inflectional -n in “sefu” (cryptic panel) must be excluded because firstly, the sequence reads se(fa, and secondly, it would be dangerous to use se(fa (masc. n-stem) ‘mind’ as an example for the loss of -n as the noun could, at least theoretically, stand in the nom. sg. and would thus regularly not contain final -n.

11 The two examples (sċeal ‘he shall’; ġeatum ‘ornamentally’) in the Leiden Riddle may be left out of account since the text was written on the Continent.

The Franks Casket, Part C: A Linguistic and Runological Perspective

277

3.3.2 The parasite vowels in the second element of fergenberig and in wylif/wylifin12 ‘she-wolf’: The limitation of parasite vowels to Nhb. as proposed by Napier (1901, 379 f.) can neither be wholly justified by OE grammars nor by the OERC itself. The unprovenanced inscriptions or those of uncertain origin containing parasite vowels in Table 8 (Mortain Casket: Angl. possibly Merc.; Kirkheaton Stone: Non-Anglian; Monte Sant’ Angelo Inscription D: linguistically indecisive) do not show any clear dialectal criteria for Nhb. Tab. 8: Parasite vowels in the OERC. Object

Findspot

Provenance13

RC NE: ]geredæ (ondgeredæ ‘disrobed’; 3rd pers. sg. pret. indic.; wk. 1)

Nhb.

Nhb. (cf. Waxenberger 2006; Waxenberger forthc., esp. Chap. 6.1.9)

Bewcastle Cross: kyniburuᚸ (Cyneburh; fem. PN)

Nhb.

Nhb. (Page 1973, 28)

Lancaster Cross: cuþbere[ (Cūþberht; masc. PN)

Nhb.

Linguistically indecisive; -e- in -berht as second element in names is seen as development in unaccented syllables rather than as smoothing proper (SB 1965, §§ 43.2;120 note 5). Similarly, the -a- in cynibalþ in the second element of the compound seems to be more susceptible to vowel change due to loss of primary accent and is therefore not a totally reliable criterion for an Anglian provenance (Waxenberger forthc., Chap. 6.1.2).

Inscription RC NE: ]ᚷeredæ hinæ ᚸod almeᛇttiᚷ * þa he walde on ᚸalᚸu ᚷistiᚸa modiᚷ f[…] [….] men[ RC SE: [..]smæræ[.]u (bismærædu ‘disgraced’; 3rd pers. pl. pret.; orig. wk. 3) Inscription RC SE: ] iᛷ riiᛷnæ ᛤyniŋᛷ * h(eafunæs hlafard hælda iᛷ ni dorstæ [..]smæræ[.]u uŋᛤet men ba ætᚸad[..] iᛷ […] [.]iþ b[.]odæ bistemi[.] bi[

Inscription: gibidæþfo ræcynibal þcuþbere[

12 For the reading wylifin, see above no. 2.3 and Part B no. 4. 13 The linguistic evidence in the inscriptions in question is given in this column.

278

Gaby Waxenberger

Tab. 8 (continued) Object

Findspot

Provenance

Whitby Comb: helipæ ‘may help’ (3rd pers. sg. pres. subj.; str. III) aluwalu}da (WS eallwealda ‘all-ruling’)

Nhb.

Angl. Linguistic evidence: -walu}da (WS -wealda) Angl. retraction of a instead of breaking (æ + l + Co): see Waxenberger (forthc., Chap. 6.1.6 rune a 2.4.2)

Nhb.

Angl. Linguistic evidence: Angl. smoothing (Angl. ē < ēa) in bekun ‘monument’

Nhb.

Non-Angl. Linguistic evidence: eoh shows breaking.

France

Angl. (possibly Merc.) Linguistic evidence: gewarahtæ: Angl. (especially Merc.) a instead of OE o < Gmc. *o: see Waxenberger forthc., Chap. 6.1.6 rune a 2.5)

Italy

Linguistically indecisive -e- in -berht as second elements in names is seen as development in unaccented syllables rather than as smoothing proper (SB 1965, §§ 43.2;120 note 5).

Inscription: d*[æ]us}mæus godaluwalu )da}helipæcy[ Great Urswick Stone: toroᛇtredæ (Torhtred; masc. PN) Inscription: +tunwinisetæ æftertoroᛇ tredæbeku næfterhisb æurnægebidæsþe rs au læ lylþi sw[ Kirkheaton Stone: worohtæ ‘made’ (3rd pers. sg. pret. indic.; wk. 1) Inscription: eoh : woro htæ Mortain Casket: gewarahtæ ‘made’ (3rd pers. sg. pret. indic.; wk. 1) Inscription: +goodhelpe:æadan þiiosneciismeelgewar ahtæ Monte Sant’ Angelo Inscription D: )her(eb(er(ehct (Hereberht; masc. PN)

The Franks Casket, Part C: A Linguistic and Runological Perspective

279

SB (1965, § 164) confirm parasite vowels for the ‘oldest texts’ and list examples from Northumbria and runic attestations from (possibly) Mercia (Mortain Casket) but exclude Mercia from the later texts; parasite vowels only occur in later Nhb. texts. A similar argument is found in Campbell (1959, § 360). Hogg (1992, § 6.35) defines Northumbria “and to a lesser extent Merc[ia]” as the area of distribution. 3.3.3 Second fronting (OE æ = Merc. e and OE a = Merc. æ in the Mercian Vespasian Psalter14 ): Second fronting is not a general Merc. change (Campbell 1959, § 168; Hogg 1992, § 5.90). Napier (1901, 380 and fn. 1) himself, too, limits this feature to the dialect area of the Vespasian Psalter admitting, however, the possibility of a “North Mercian origin” because the Rushworth1 Gloss “Matthew generally has æ”; see also Hogg (1992, §§ 5.87, 5.90). This sound change thus does not restrict the casket to Nhb. but merely excludes that part of Mercia where the Vespasian Psalter was written. Therefore, second fronting is not a sufficient criterion. Summary After discussing all the sound changes listed above, Napier (1901, 380) concludes that the language of the Franks Casket is Northumbrian15 and he suggests that it may be pinpointed even more narrowly: “Can the whale have been stranded at the foot of the cliffs on the summit of which stood the abbey of Streoneshalh [= Whitby]?” However, he focuses only on a fraction of the linguistic evidence, and more importantly, on evidence that does not lead to the conclusion he draws from it (e.g., parasite vowels: 3.3.2; second fronting: 3.3.3). In order to find out if Napier’s hypothesis is nevertheless plausible, a more reliable attempt to locate and date the text must be made, comprising a more detailed analysis of both its language and the runeforms. I would like to start with the language.

14 According to Campbell (1959, § 168) “the change of æ to e” is also “fairly common” in EpGl and CorpGl and “very frequent” in the Royal Glosses and St Chad; see also Hogg (1992, § 5.90). 15 “We may, I think, safely assert that the home of the casket was the coast of Northumbria” (Napier 1901, 380).

280

Gaby Waxenberger

4 More Linguistic Criteria I will now present five pieces of linguistic evidence that Napier overlooked.16 This evidence helps to establish that the text is not only Angl. but distinctly Nhb. These features are: – Nhb. retraction of æ > a, – the development of Gmc *a + nasal, – the runic representation of the clusters [xt], [çt], – the runic representation of the velar consonants after the palatalisation/assibilation processes, and finally, – back mutation.

4.1 Nhb. Retraction to a vs. Breaking of æ + r + Co > ea (cf. Waxenberger forthc., Chap. 4.6 A1.1.1) Nhb. retraction of æ > a in stressed syllables17 is “especially common when a labial consonant (f, p, b, m, w) precedes the vowel or follows the r” (Campbell 1959, § 144). As warþ ‘became’ (Franks Casket Front Panel) contains such an environment, it represents a clear dialect marker for Nhb.

4.2 Gmc. *a + Nasal (= [å]) The OE successor of Gmc. *[å] was spelt both ‹o› and ‹a› in early Nhb. but ‹o› greatly predominates (Campbell 1959, § 130 n. 2). In addition to the evidence from the OERC (see Waxenberger forthc., Chap.s 6.1.2 rune o 2.2; 6.1.6 rune a 2.1), the non-runic Nhb. Ms. L of Cædmon’s Hymn also has ‹o› at least for the stressed syllable. Tab. 9: Gmc. *a + nasal > early Nhb. ‹a›; ‹o›. Franks Casket

Cædmon's Hymn (L)

Stressed syllables Franks Casket Front Panel: hronæsban (hron/hran ‘whale’) giswom (3rd pers. sg. pret. swam/swom ‘swam’; swimman ‘to swim’, str. III)

Stressed syllables moncynnæs ‘mankind’

16 In both texts, the RC and the Franks Casket, the i-umlaut of o/ō is at a stage represented by œ and on the RC the i-umlaut of a + nasal is at a stage represented by e. Therefore, i-umlaut qualifies as a marker neither for dialect nor for time. 17 Retraction in unstressed syllables is not necessarily a dialect marker.

The Franks Casket, Part C: A Linguistic and Runological Perspective

281

I regard these ‹o› forms on the Franks Casket as additional support for Nhb.18 One of the most promising and enlightening points in locating the casket is the rendering of the voiceless palatal [ç] and velar [x] fricatives.

4.3 The Voiceless Palatal [ç] and Velar [x] Fricatives (Particularly when Followed by /t/) The following table shows the attempts that were made to depict those fricatives. Tab. 10: The clusters [xt] and [çt] within the OERC. 1. [xt] = ‹ht›

2. [çt] = ‹ht›

3. [çt] = ‹hct›

4. [çt] = ‹ᛇtt›

5. [xt] = ‹ᛇt›

6. [çt] = ‹gt›

Kirkheaton Stone: worohtæ Mortain Casket: gewarahtæ

Thornhill Stone III: berhtsuiþe London National Portrait Gallery Bone: tatberht

Monte Sant’ Angelo Inscription D: }her(eb(er(ehct

RC NE: almeᛇttig

Great Urswick Stone: toroᛇtredæ

Franks Casket: fegtaþ 6.1 [ç] = ‹g› Franks Casket: unneg

For the manuscript tradition SB (1965, §§ 4.7; 221 note 1) list four of the six possibilities in Table 10 showing the runic tradition. Naturally, the ones lacking are the variants using the yew-rune ᛇ as this runic grapheme has no equivalent in the Latin script and is therefore absent from the non-runic manuscript tradition. Although the original sound value of the yew-rune is by no means clear, it could definitely denote /i(:)/ in OE (as can be inferred from Map 2). However, as /i(:)/ could also be expressed by the rune īs ᛁ, the yew-rune ᛇ could be ‘re-cycled’ and used for other sounds: in the RC inscription it was used for the voiceless palatal fricative [ç], whereas it denoted the voiceless velar fricative [x] on the Great Urswick Stone. The RC carver/designer applied the yew-rune ᛇ because it had, in principle, become obsolete (for more details see Waxenberger forthc., Chap. 6.2.2 rune ēoh/īh, esp. 5, and Waxenberger 2017a). Apparently, the Franks Casket carver/designer was also looking for other ways than h ᚻ to depict the voiceless palatal allophone [ç] in fegtaþ and unneg, but chose the rune g ᚷ. I am well aware of the fact that the following statement may lead us into the realm of speculation. But looking at the distribution maps of ht and ᛇt(t) (see below 18 However, prepositions are excluded because they occur in unstressed position and are therefore no dialect criterion (SB 1965, § 79 n. 3): see also Waxenberger (forthc., Chap. 6.1.2 rune o 2.2). The preposition on occurs on the Franks Casket Front (2×) and also on the RC (NE and SW) as well as on the Thornhill Stone III.

282

Gaby Waxenberger

Map 2: Inscriptions applying the yew-rune in the Pre-OE and OERC (fricative, vowel, rune no. 13 as a fuþorc-unit). Map base: © EuroGeographics, linguistic details: RuneS research unit EichstättMünchen.

Maps 2 and 3) it becomes clear that, even if we take into account the scarcity of the attested forms, it is notable that ht does not appear in the north-west of Nhb. Considering that the Thornhill Stone III may be a special case,19 ht is thin on the ground in all of Nhb.20

19 On the assumption that the Thornhill Stones were carved by the same ‘school’, the yew-rune was probably used (in the traditional way) for [ɪ] as the second element of the diphthong [eɪ] in the Thornhill Stone I (+(eadred sete æfter (eateᛇnne ‘+Eadred set up (this memorial) after Eadthegn’: Page 1973, 144; 1999, 141). As eᛇ was used for [ej] or possibly [eı] in (eateᛇnne, the sequence [xt] had to be rendered as ‹ht› in the Thornhill Stone III (+ᚼilsuiþ : arærde : æft[.] | berhtsuiþe . bekun | onbergigebi)ddaþ | þær : saule: Page 1973, 156; 1999, 141). 20 For the sake of completeness the personal name on the Kirkheaton Stone should also be mentioned here although the -h in eoh does not occur in the cluster ht. The non-smoothed form eoh

The Franks Casket, Part C: A Linguistic and Runological Perspective

283

Map 3: Rune h used in the cluster ‹ht› in the OERC; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München.

instead of *eh (cf. DOE online 2018, s. v. eoh 1. ‘horse, steed, stallion’; 2. Personal name) on the Kirkheaton Stone is non-Angl. and can therefore not be taken into account.

284

Gaby Waxenberger

But why did the Franks Casket carver/designer use the g-rune instead of the yewrune, like the RC carver/designer did? Does this mean that the Franks Casket is older than the RC, or does it mean that the Franks Casket carver was from a different part of the country, or does it mean both? A closer look at the runes ġ(i)efu g ᚷ, gār ᚸ and the yew rune ᛇ may shed more light on this matter. For there is another major difference between the Franks Casket and the RC with respect to the rune ġ(i)efu g ᚷ. The OE allophones ([ʝ], [ɤ], [ɉ], and [g]21) of Gmc. */ɤ/ were initially represented by the rune ġ(i)efu g ᚷ exclusively, in the inscriptions in Mercia and the more southerly dialect areas and also in the inscriptions in the east of Nhb. It was only some time later that rune gār ᚸ emerged in Nhb. to represent the velar allophones ([ɤ] and possibly [g]).22 The most significant difference between Tables 11 and 12 is the complete absence of gār before ca. AD 750. Table 11 and Map 4 clearly show that the earliest appearance of the new gār-rune is on the Ruthwell and Bewcastle Crosses. Thus, the RC and Franks Casket differ in two major usages of the rune ġ(i)efu g ᚷ: 4.3.1 While the Franks Casket uses the rune ġiefu g ᚷ to denote both the palatal (OE [ʝ]) and velar (OE [ɤ]) allophones of Gmc. */ɤ/, on the RC the rune ġiefu exclusively depicts the palatal allophone (OE [ʝ]), whereas the velar allophone (OE [ɤ]23) is represented by the new rune gār ᚸ.24 Tab. 11: Rune ġiefu vs. rune gār. Franks Casket

Ruthwell Cross

Rune ġiefu ᚷ palatal and velar allophones of Pre-OE */ɤ/: [ʝ] (initially, medially, and finally); [ɤ]/[g] (initially); [ɤ] (medially and finally)

Rune ġiefu ᚷ palatal allophones of Pre-OE */ɤ/: [ʝ] (initially, medially, and finally)

21 22 23 24

Rune gār ᚸ velar allophones of Pre-OE */ɤ/: [ɤ]/[g] (initially); [ɤ] (medially and finally)

[dʒ] and later /dʒ/ is not attested in the OERC. For the exact distribution see Waxenberger (forthc., Chap. 5). For the problematic word-initial sound value [ɤ]/[g] see Waxenberger (forthc., Chap. 5.1). A similar process can be observed for the rune c ᚳ (see below 5.4.4).

The Franks Casket, Part C: A Linguistic and Runological Perspective

285

Map 4: Attested distribution of the runes ġ(i)efu ᚷ and gār ᚸ to denote [ɤ] initially, medially and finally; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München.

Legend to Map 4 I. Unprovenanced Objects Derby(shire) Bone Plate (probably Anglian: Waxenberger forthc., Chap. 2B no. 23) Mortain Casket (probably Anglian; possibly Merc.: Waxenberger forthc., Chap. 2B no. 61) Lancashire Ring (Waxenberger forthc., Chap. 2B no. 37) Inscriptions that show the rune ġ(i)efu ᚷ, without complete certainty as to whether the allophone is [ɤ]. Kingmoor Amulet Ring: the inscription has not yet been satisfactorily interpreted: Waxenberger (forthc., Chap. 2B no. 35). Bramham Moor Ring: the inscription has not yet been satisfactorily interpreted: Waxenberger (forthc., Chap. 2B no. 8). II.

III. The Thornhill Stone III The potential allographic differentiation for the allophones of early OE /ɤ/ (> class. OE /g/) may also explain the seeming exception of the Thornhill Stone III. I consider it possible that the pronunciation of g in a smoothed form bergi became more palatal as the consonant cluster -rg- occurs between two front vowels. However, SB (1965, § 119 n. 6) state that there is a lack of evidence for a more palatal pronunciation; this may be due to a lack of graphemic representation of such a phenomenon in the Latin script. Possibly, however, such an indication is provided by the rune g ᚷ ġ(i)efu on the Thornhill Stone III. For more details on the statements of the OE grammars: see Waxenberger (forthc., Chap. 5.4.3.3.5).

286

Gaby Waxenberger

4.3.2 Similarly, the two texts also differ in assigning characters to the voiceless palatal fricative [ç]: it is the rune ġiefu ᚷ for the Franks Casket but the yew-rune for RC. Tab. 12: Rune ġiefu ᚷ vs. the yew-rune ᛇ. Franks Casket

Ruthwell Cross

[ç]: rune ġiefu ᚷ

[ç]: rune ē(o)h/īh (yew-rune) ᛇ

In view of these striking differences, the question arises as to whether more differences can be found between the RC and the Franks Casket? And, indeed, they do in fact differ in the presence/absence of back mutation.

4.4 Occurrence and Lack of Back Mutation (Waxenberger forthc., Chap. 4.8 A4.2.1), u-umlaut of e > eo before labials (SB 1965, § 110.1) As can be seen in the left column of the table below, the RC shows u-umlaut before the voiced labio-dental fricative [v] (with the Nhb. spelling ‹(ea› for ‹eo› here), whereas in the case of the Franks Casket o/a-umlaut of e > eo (Nhb. also ‹ea›) before the voiced labio-dental fricative [v] is absent.25 Tab. 13: Occurrence of u-umlaut of e > eo vs. lack of o/a-umlaut of e > eo. Findspot

u-umlaut of e > eo before [v]

Findspot

lack of o/a-umlaut of e > Angl. eo before [v]

Nhb.

RC: h(eafunæs (= heofunæs) ‘heaven’

Italy

Franks Casket Right Panel: s (fa (= se(fa) sefa ‘mind, spirit’

So why is there no o/a-umlaut in s (fa (se(fa)?26 Is it too early for this phenomenon, dated by Luick (1921, § 291) to the “end of the 7th/beginning of the 8th cent.”? It 25 Owing to their uncertain etymologies, Wadstein’s (1900, 24, 51) suggestion that -eu- in reumwalus (FC Left Panel) might have undergone mutation “Remu” > “Reumu-” (see Waxenberger forthc., Chap. 6.1.4 rune e 9.2.) is not considered here, especially since there are probably no other attested forms for back mutation in the FC. 26 Back-umlaut seems to be lacking unless we want to read the bind-rune (fa as }afa and therefore Nhb. *seafa for seofa ‘mind, spirit, heart’. Such a bind-rune, however, would be unparalleled in the OERC.

The Franks Casket, Part C: A Linguistic and Runological Perspective

287

may be interesting in this context that the combinative u-umlaut (WGmc. *i > OE u when preceded by w) was in fact carried through in Franks Casket’s wudu ‘wood’. The time for this particular umlaut is the 7th cent. according to Luick (1921, § 291). Unfortunately, the OERC does not have any other inscriptions with instances showing back-mutation of e or the lack of it. However, the non-runic Cædmon’s Hymn (Ms. L) may shed new light on the Franks Casket’s provenance and date since Cædmon’s Hymn also lacks u-umlaut of e > eo.

Tab. 14: Lack of u-umlaut. No u-umlaut of e > eo

Ms. L: hefenricæs (Smith 1978, 38.1) Ms. L: hefen (Smith 1978, 38.1) Ms. L: metudæs (Smith 1978, 38.2)27

Conclusion With all due caution, the data presented leads me to the following conclusion. As the RC and the FC differ in three major criteria (back-umlaut, yew-rune vs. ġiefu ᚷ for [xt]; gār ᚸ vs. ġiefu ᚷ for [ɤ]/[g]) on the one hand, and as the Franks Casket and Cædmon’s Hymn share at least one item of the three (lack of back-umlaut) and additionally have at least some of the ‹o› spellings for Gmc. *a + nasal in common, on the other hand, I suggest that the Franks Casket was probably not produced in the area around Ruthwell nor the larger north-west of Nhb. Now I would like to turn to the rune-forms.

5 The Runes: Allograph Types and Sound Values 5.1 The Rune s The allograph type ᛋ→; ᛋ→ of the rune s proves only that the Franks Casket is in congruence with the other Nhb. inscriptions of the 8th and 9th cent. because it is the type predominant at that time (see Waxenberger forthc., Chap. 3 for the distribution of the allograph types of the rune s in the OERC).

27 In contrast to u-umlaut of e > eo before labials not being a dialect marker, u-umlaut of e > eo before dentals is clearly Angl.

288

Gaby Waxenberger

5.2 The Rune d The rune d has two allographs in the OERC: the large d ᛞ and the small d . The large d ᛞ was used until ca. AD 700. After that time the small allograph type was introduced; both types seem to have been used side by side for the next three hundred years or so. No prevalence of one type can be seen, not even in the west of Nhb., where the small form is exclusively used on the RC (at least 18×) but the large variant is applied on the Great Urswick Stone (2×).

5.3 The Rune o There are four allograph types in the OERC, two of which are, by and large, prevalent in Nhb. and also in the OERC: the chevron type ᚩ (7th/8th–9th/(10th) cent.) and the pointed angle type (mid 7th–9th cent., late Anglo-Saxon period). Both types are common from the 7th–9th cent. and there is no obvious distributional pattern. The Franks Casket shows the chevron type 1B (= the pointed chevron type: ) and is therefore well integrated into the runic usage at the time in question (see Waxenberger forthc., Chap. 7.2). Summary As far as the runes s, d, and o are concerned, the Franks Casket is well-integrated into the runic context of the 8th cent. This means the rune-forms do not point to the later dating suggested in some approaches (e.g., Frank 1973). However, there are some other runes that help in dating the inscription even more precisely.

5.4 More Runes and Rune Forms In the following sections, the allographs of the runes y, )ea and c will be analysed.

5.4.1 The Rune y (see Waxenberger forthc., Chap. 7.7) Compared to the later and often Southern inscriptions (Bristol/Linstock Castle Agate Ring28 ( ); Morton Runic Strapend ( ); London Thames scramasax ( ), the Nhb. east coast inscriptions have a more uniform and also more transparent allograph type showing the individual elements of rune y, that is, the runes u ᚢ and i ᛁ. While

28 See Waxenberger (forthc., Chap. 2B no. 12) for the places Bristol and Linstock Castle.

The Franks Casket, Part C: A Linguistic and Runological Perspective

289

the Hartlepool Stones, the Whitby Comb, and the Franks Casket share almost the same allograph type (transparent type A ᚣ) of the rune y, the Bewcastle, Ruthwell, and Lancaster crosses have the broader allograph type (transparent type B ) of the rune y: see Waxenberger (forthc., Chap. 7.7 Table 1) for an overview of the attested allograph types in the OERC and for the transparent type A ᚣ see also below Table 17.

5.4.2 Rune No. 28 )ea ᛠ This rune is normally used for the diphthongs ea; ēa but in Nhb. it also denotes eo; ēo, e.g., in the RC inscription. Tab. 15: Rune no. 28 )ea ᛠ. ᛠ )ea used for eo on the RC

ᛠ )ea used for ēo on the RC

RC SE: h(eafunæs (gen. sg. of heofon ‘heaven’: back mutation)

RC NW: [...](ea[.]du[. ..] ([bih](ea[l]du[n] = WS behēoldon ‘watched’: 3rd pers. pl. pret.)

RC SW: f(earran (= feorran ‘from afar’: breaking)

However, the rune (ea ᛠ does not occur on the Franks Casket, because the diphthong ēo is spelt as e and u (‹eu› for later ēo which was also denoted by ‹ea› in Nhb.). Moreover, there is a sequence -ea- on the Franks Casket which consists of the individual runes e and a; it may be assumed, however, that this is not a diphthong as it has been parsed as giuþea su*[m] and interpreted as the gen. pl. ‘some(one) of the Jews’: see also Part B 3.2 Possibility 1. The forms in question on the Franks Casket: 1. greut = grēot ‘grit, sand’ 2. giuþeasu (giuþea su; iudeasu (= iudea su*[m] ‘some(one) of the Jews’ (CH 1960, 207: Iudeas ‘the Jews’ masc. pl.; BT 1898, 602: Iūdēas; gen. -a; pl. masc. ‘The Jews’)) The possible interpretation of the forms on the Franks Casket listed above is threefold: Firstly, the sequence was not considered to be the diphthong ea, especially since it occurs in the gen. pl. of a loan word (giuþeasu), and therefore the quality and the position of ‹ea› may have been felt to be different from the native pattern. Secondly, the bind-rune )ea did not exist at the time of the Franks Casket, or thirdly, the Franks Casket carver pronounced the sounds represented by ‹ea› and ‹eu› (slightly) differently from the RC carver/designer.

290

Gaby Waxenberger

5.4.3 The Franks Casket in Comparison to the Ruthwell Cross and Other Inscriptions Looking at the RC as the most prominent and sophisticated representative of NorthWest Nhb., it thus differs in many points from the Franks Casket: Tab. 16: The Franks Casket and the Ruthwell Cross. Franks Casket

Ruthwell Cross

Rune ġiefu ᚷ is used in the cluster [xt]

Yew-rune is used in the cluster [xt]

Rune ġiefu ᚷ is used for the OE velar and palatal allophones of Gmc. */ɤ/

Rune ġiefu ᚷ is used for the OE palatal allophone [ʝ] of Gmc. */ɤ/ Rune gār is used for the OE velar allophone [ɤ] of Gmc. */ɤ/

Lack of back-umlaut

Back-umlaut

No use of rune )ea ᛠ (giuþeasu; greut)

Rune )ea ᛠ is used (h(eafunæs; [...](ea[.]du[. ..] ([bih](ea[l]du[n] = WS behēoldon); f(earran)

On the other hand, the Franks Casket shares features with some of the inscriptions found on the east coast of Nhb., and it seems very close to one in particular, namely the one on the Whitby Comb. This inscription is also on bone, but a different technique was applied: the Franks Casket is cut in relief, whereas the Whitby Comb inscription is incised. Nevertheless, one might argue that the same material requires the same shapes of runes. This argument is, however, not totally conclusive, as can be demonstrated by the different shapes on the later Derby(shire) Bone Plate ((later) 8th–10th cent.).

Fig. 1: Derby(shire) Bone Plate; © The Trustees of the British Museum (8th–10th cent.; small allograph type d, different allograph type r, different (later) form of c ᛷ: see Waxenberger forthc., Chap. 5.5 Table 5).

The Franks Casket, Part C: A Linguistic and Runological Perspective

291

If it is thus not the material that is responsible for the similar rune forms, the place of origin and the time of making the inscription may in fact be a decisive factor. Apart from the allographs of the more common runes such as s, the Whitby Comb and the Franks Casket share the pointed, narrow u ᚢ, the chevron type of the rune o ᚩ, the early wynn-rune w with the small and pointed hook ᚹ; the early y-type ᚣ and, last but not least, the early allograph type of the rune c ᚳ. As a last item, the rune c ᚳ will be investigated in connection with the new rune ᛣ calc. Tab. 17: Runes o, u, y, w, and c on the Whitby Comb29 and the Franks Casket; © The Trustees of the British Museum; © Whitby Literary and Philosophical Society, Whitby Museum. o ᚩ

u

y

w ᚹ

c ᚳ

Whitby Comb

Franks Casket

It is noteworthy that the Franks Casket and the Whitby Comb also share linguistic features, such as the same parasite vowels -i- after a stressed vowel -e- (Franks Casket: fergenberig; Whitby Comb: h } elipæ), whereas (somewhat) later inscriptions show a parasite vowel -e- after a stressed vowel -e- (Lancaster Cross: cuþbere[.] Cūþberht; Monte Sant’ Angelo Inscription D: h ) er(eb(er(ehct Hereberht; RC NE: ]geredæ ]geredæ).30 If the quality of the parasite vowels (e.g., -i- vs. -e-) indicate dates of origin rather than provenance (see above section 3.3.2), then the Franks Casket and the Whitby Comb may be earlier than the RC (RC NE: ]geredæ ]geredæ). Criteria such as the rune-forms (especially the allograph, the transparent type A of y: see above section 5.4.1) and the verbal ending -æ (see also Waxenberger forthc., Chap. 6.3) place the Whitby Comb inscription in the late 7th/early 8th cent. This timespan does not contradict the dating of the comb: John Hines (private communication: 6th July 2009) comments that the form of the Whitby Comb “is essen-

29 Some allographs cannot be clearly seen in the photographs but they were unambiguously confirmed by an autopsy in 2010. 30 For more details on parasite vowels in the OERC see Waxenberger (forthc., Chap. 6.3).

292

Gaby Waxenberger

tially that which is typical of almost anywhere in England of the 7th and 8th centuries [...] quite likely a bit later too”. Before I can arrive at a more reliable conclusion, the rune c ᚳ and its distribution need to be be analysed in the following section.

5.4.4 The Rune ċēn c ᚳ Although the Franks Casket lacks attested forms for velar [k], there is sufficient proof in the OERC (cf. Waxenberger forthc., Chap. 5 Table 28) that the OE velar plosive [k] as well as the palatalised [ḵ] and assibilated [tʃ ] of Gmc */k/ were represented by ċēn c ᚳ in the earlier (east) Nhb. as well as in non-Nhb. inscriptions (see Waxenberger forthc., Chap. 5.7.1). However, similar to rune g, the situation changes for the rune ċēn c ᚳ and again, the starting point seems to be Ruthwell. As can be seen in the distribution patterns of ċēn c ᚳ and the new runes calc ᛣ and rune no. 31 no. ᛤ (see Waxenberger forthc., Chap. 5 esp. 5.9), the rune ċēn ᚳ was reserved for the assibilated sound [tʃ ], whereas velar [k] was denoted by the rune calc ᛣ and palatal [ḵ] by the rune no. 31. ᛤ on the RC (for more details see Waxenberger forthc., Chap. 5). Looking at Maps 4 and 6, it becomes obvious that the graphemically more elaborated inscriptions belong to the west of Nhb.31 whereas the others appear more in the eastern part of Nhb. as well as in the more southerly dialect areas. Inscriptions that have not been (satisfactorily) interpreted, are excluded: Blythburgh Writing Tablet, London National Portrait Gallery Bone, London Thames Mount, Scotterthorpe Lead Fragment (see Waxenberger forthc.: Chap. 2B nos. 6, 48, 50, 72). 31 Regarding the runes calc ᛣ and no. 31 ᛤ, the Lancaster Cross seems to deviate from the general picture, but this may merely be a question of date, as there are indications that the Lancaster inscription is earlier than the RC. There are two reasons for this assumption: First, the full vowels -i and -æ in unstressed positions in the Lancaster Cross text hint at an early date. Second, the Hartlepool Stones, the Whitby Comb, and the Franks Casket share almost the same allograph type (transparent Type A ᚣ) of the rune y, while the Bewcastle, Ruthwell, and Lancaster crosses have the broader allograph type (transparent Type B ) of the rune y (cf. Waxenberger forthc., Chap. 7.7 Table 1). With regard to the relation of the Lancaster Cross to the Franks Casket and the Whitby Comb: the latter two share the parasite vowel -i- after a stressed vowel -e- (Franks Casket: fergenberig; Whitby Comb: }helipæ), whereas the Lancaster Cross (cuþbere[.] Cūþberht) and others (Monte Sant’ Angelo; RC NE) have a parasite vowel -e- after a root vowel -e-. This seems to represent a weakened vowel and therefore a later development. If the quality of the parasite vowels (e.g., full vowel -i- vs. weakened vowel -e-) indicates dates of origin rather than provenance, then the Franks Casket and the Whitby Comb may be slightly earlier than the Lancaster Cross inscription, which in turn seems to be somewhat earlier than the RC. To sum up, from a linguistic and runological point of view the Lancaster inscription may have been written in the first half of the 8th cent. Therefore, the use of the rune no. 6 ċēn ᚳ for both the plosives [k] and [ḵ] is not surprising, because the new creations, rune no. 30 calc ᛣ and no. 31 ᛤ, were probably designed by the Ruthwell carver/designer in ca. AD 750, that is, slightly later than the Lancaster Cross.

The Franks Casket, Part C: A Linguistic and Runological Perspective

293

Unprovenanced objects: Derby(shire) Bone Plate, probably 8th cent.

Map 5: Attested distribution of rune no. 6 ċēn ᚳ; and its variants ᛷ, in inscriptions proper32 for the plosive allophones [k] and [ḵ]; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München. 32 The inscriptions showing fuþorc rows or parts of them are excluded.

294

Gaby Waxenberger

Map 6: Attested distribution of ċēn ᚳ [k]; [ḵ], calc ᛣ [k]; [ḵ], and rune no. 31 ᛤ [ḵ]; map base: © EuroGeographics, linguistic details: RuneS research unit Eichstätt-München.

The Franks Casket, Part C: A Linguistic and Runological Perspective

295

6 Summary and Conclusion What can we conclude about the Franks Casket’s language and its runes in the context of the OERC in general and the Nhb. sub-corpus in particular? Regarding the language of the casket, some dialectal features (Anglian smoothing, lacking diphthongisation of æ after palatals33) assigned it to Anglia, while others (retraction of a; ‹o› for Gmc. *a + nasal) made a narrower attribution to Nhb. possible. In comparing the RC and the other north-west Nhb. inscriptions, two points became increasingly obvious: first, the Franks Casket must be older than the RC, and second, it differs in its runes and also in some language features (e.g., no backumlaut) from the north-west, but shows clear affinities to other east coast inscriptions (e.g., rune y). Additionally, its runes have particularly close similarities with the Whitby Comb, which, according to Page (1973, 168; 1999, 165), “comes from a rubbish dump near the ruins of Whitby Abbey, and is traditionally associated with the Anglo-Saxon monastery of Streoneshalh, founded in 657 and destroyed in the second half of the ninth century”. Last but not least, the Ms L of Cædmon’s Hymn has dialect features that are also shared by the Franks Casket (lacking back-umlaut; ‹o› for Gmc. *a + nasal). Regarding the dating of the Franks Casket, some of the features, e.g., the use of the rune g ᚷ ġiefu exclusively for both the palatal [ʝ] and velar allophones [ɤ] of Gmc. */ɤ/, the application of the rune g ᚷ ġiefu instead of the yew-rune ᛇ for [ç]; [x], the presence of the earlier shape of rune c ᚳ ċēn instead of the apparently later c ᛷ, the spellings ‹-i› and ‹-u› for the unstressed vowels, absence of back-mutation but presence of combinative umlaut (“gesteigerter Velarumlaut”) as well as the comparison to the other inscriptions indicate that the Franks Casket must be slightly older than the RC; therefore, I suggest the earlier decades of the 8th century. Regarding the place of origin, it has been shown that the Franks Casket has close parallels with the Whitby Comb inscription and Cædmon’s Hymn. If these two are accepted as narrowly located,34 then the home of the casket could be the area

33 In the context of and together with the other features, the lack of diphthongisation also corroborates the restriction of the Franks Casket texts to early Northumbria. 34 From a purely linguistic point of view, the Whitby Comb inscription only allows for an analysis as Anglian. John Hines (private communication: 6 July 2009) comments on the archaeological evidence: “(…) So far as I know, the archaeological evidence from Whitby does not specifically include boneworking waste, but these excavations were so early, and relatively poorly recorded, that nothing could be built on such negative evidence.” In my opinion, the inscription on the Whitby Comb (see above 3.3 Table 8) is probably a simple pious maker formula (Waxenberger 2003, 170 f.) although the person’s name is incomplete (in contrast to the Mortain Casket: see above 3.3 Table 8); nevertheless, the Whitby Comb inscription (‘my God, God almighty may help Cy(n)-’ (Bammesberger 1991, 134) or possibly ‘My God: may God the Allruler help Cy-’ (Bammesberger 2017, 294)) indicates a monastic context especially with its Latin beginning d*[æ]us}mæus.

296

Gaby Waxenberger

between Whitby and Wearmouth-Jarrow.35 Leslie Webster’s data allows also including Lindisfarne as a possible place of origin. My data on Lindisfarne (= Lindisfarne stones I–VI: Waxenberger forthc., Chap. 2b nos. 41–46) is very scarce. It yields no comparative material apart from the rune y (Lindisfarne Stone I: Waxenberger forthc., Chap. 2b no. 41: osgyþ Ōsgȳ þ), which also belongs to the transparent type like the ones on the Whitby Comb and the Franks Casket (see above section 5.4.1). The rune y on its own would have been absolutely insufficient, but in synopsis with the art-historical evidence, it means that Lindisfarne also becomes a possible candidate for the provenance of the Franks Casket.

References Bammesberger, Alfred. 1991. “Three Old English Runic Inscriptions”. Old English Runes and their Continental Background. Anglistische Forschungen 217. Ed. Alfred Bammesberger. Heidelberg: Winter. 125–136. Bammesberger, Alfred. 1991a. “Franks Casket: Editor’s Notes”. Old English Runes and their Continental Background. Anglistische Forschungen 217. Ed. Alfred Bammesberger. Heidelberg: Winter. 629–632. Bammesberger, Alfred. 2017. “A Note on the Whitby Comb Runic Inscription”. Notes and Queries 64: 292‒295. Brunner, Karl. 1965. Altenglische Grammatik: Nach der angelsächsischen Grammatik von Eduard Sievers. 3rd ed. Tübingen: Niemeyer. BT = Bosworth, Joseph and T. N. Toller. 1882‒1898. An Anglo-Saxon Dictionary: Based on the Manuscript Collections of the Late Joseph Bosworth. Ed. and enlarged by T. N. Toller. Oxford: Oxford University Press. Campbell, Alistair. 1959. Old English Grammar. Oxford: Oxford University Press. CH = Clark Hall, John R. 1960. A Concise Anglo-Saxon Dictionary. 4th ed. with a supplement by Herbert D. Meritt. Cambridge: Cambridge University Press. Chadwick, Hector Munro. 1899. “Studies in Old English”. Transactions of the Cambridge Philological Society 4: 85 ‒265. Derolez, René and Ute Schwab. 1983. “The Runic Inscriptions of Monte S. Angelo (Gargano)”. Academiae Analecta 45: 95–130. DOE Online = Dictionary of Old English: A to I Online. Eds. Angus Cameron, Ashley Crandell Amos, Antonette diPaolo Healey et al. Toronto: Dictionary of Old English Project, 2018. URL: ‹www.doe.utoronto.ca/›, last accessed 20 June 2021. Dumville, David. 2007. “The Two Earliest Manuscripts of Bede’s Ecclesiastical History”. AngloSaxon 1: 55‒108. Elliott, Ralph W. V. 1959. Runes: An Introduction. Manchester: Manchester University Press. Elliott, Ralph W. V. 1989. Runes: An Introduction. 2nd ed. Manchester: Manchester University Press. Gameson, Richard. 2015. “Materials, Text, Layout and Script”. The St Cuthbert Gospel: Studies on the Insular Manuscript of the Gospel of John. Eds. Claire Breay and Bernard Meehan. London: British Library. 13‒39.

35 Franks Casket: þær ‘there, where, when’ is not an intrusion of a non-Anglian form. See Bammesberger (1991a, 631 f.).

The Franks Casket, Part C: A Linguistic and Runological Perspective

297

Herren, Michael W. and Hans Sauer. 2016. “Towards a New Edition of the Épinal-Erfurt Glossary: A Sample”. The Journal of Medieval Latin 26: 125‒198. Hogg, Richard. 1992. A Grammar of Old English. Vol. 1: Phonology. Oxford: Blackwell. Lowe, Elias A. 1958. “A Key to Bede’s Scriptorium: Some Observations on the Leningrad Manuscript of the Historia Ecclesiastica Gentis Anglorum”. Scriptorium 12: 182‒190. Luick, Karl. 1921. Historische Grammatik der englischen Sprache: Band I, 1. Abteilung. Leipzig: Tauchnitz Verlag. Napier, Arthur S. 1901. “Contribution to Old English Literature”. An English Miscellany Presented to Dr. Furnivall in Honour of his Seventy-Fifth Birthday. Eds. William P. Ker and Arthur S. Napier. Oxford: Clarendon. 355‒381. Page, Raymond Ian. 1973. An Introduction to English Runes. London: Methuen. Page Raymond Ian. 1999. An Introduction to English Runes. 2nd ed. Woodbridge: The Boydell Press. Page, Raymond Ian. 2006. “Bone Fragment with Runic Inscriptions: 2252”. The Mote of Mark: A Dark Age Hillfort in South-West Scotland. Eds. Lloyd Laing and David Longley. Oxford: Oxbow. 92‒93, and Pl. 17. Parkes, Malcolm B. 1994. “The Scriptorium of Wearmouth-Jarrow”. Bede and His World I: The Jarrow Lectures 1958‒1978. Aldershot: Variorum. 555‒586. Pheifer, Joseph D. (ed.). 1974. Old English Glosses in the Épinal-Erfurt Glossary. Oxford: Clarendon. Sauer, Hans and Gaby Waxenberger. 2012. “Old English: Dialects”. English Historical Linguistics: An International Handbook. Vol. 1. Eds. Alexander Bergs and Laurel J. Brinton. Frankfurt/New York: De Gruyter. 340–361. SB = Sievers/Brunner, see Brunner, Karl (1965) Schapiro, Meyer. 1958. “The Decoration of the Leningrad Manuscript of Bede”. Scriptorium 12: 191–207. Searle, William G. 1897. Onomasticon Anglo-Saxonicum: A List of Anglo-Saxon Proper Names from the Time of Beda to that of King John. Cambridge: Cambridge University Press. Sievers, Eduard. 1891. “Zu Cynewulf”. Anglia 13: 1–25. Smith, Albert H. (ed.). 1978. Three Northumbrian Poems: Cædmon’s Hymn, Bede’s Death Song and The Leiden Riddle. With a Bibliography Compiled by Michael J. Swanton. Exeter Medieval English Texts, rev. ed. Exeter: Exeter University Press. Waldispühl, Michelle. 2018–2019. “Roman and Runic in the Anglo-Saxon Inscriptions at Monte Sant’ Angelo: A Sociolinguistic Approach”. Futhark 9–10: 135–158. Waxenberger, Gaby. 2003. “The Intriguing Inscription of the Gandersheim Runic Casket Revisited”. Bookmarks from the Past. Eds. Lucia Kornexl and Ursula Lenker. Frankfurt: Lang. 143‒176. Waxenberger, Gaby. 2017. “Date and Provenance of the Auzon or Franks Casket”. Life on the Edge: Social, Political and Religious Frontiers in Early Medieval Europe. Eds. Sarah Semple, Celia Orisini, and Sian Mui. Braunschweig: Uwe Krebs. 121–133. Waxenberger, Gaby. 2017a. “The Development of the Old English fuþorc”. Von den Hieroglyphen zur Internetsprache: Das Verhältnis von Laut, Schrift und Sprache/ From Hieroglyphs to Internet Language: The Relation of Script, Sound and Language. LautSchriftSprache 2/ScriptandSound 2. Eds. Gaby Waxenberger, Hans Sauer, and Kerstin Kazzazi. Wiesbaden: Reichert. 209–247. Waxenberger, Gaby. Forthc. A Phonology of Old English Runic Inscriptions with a Concise Edition and Analysis of the Graphemes. RGA-E. Berlin/Boston, MA: De Gruyter.

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

A Concise and Selected Guide to Terminologies The first step to interdisciplinarity is to understand each other’s terminologies. This Guide is intended to be such a first step, proceeding from the terms occurring in this volume.

1 Linguistic Terms Gaby Waxenberger and Kerstin Kazzazi

Alliteration Alliteration is the relationship between words which begin with the same sound. In the early → Germanic languages (and indeed more widely) it is a frequently used discourse-stylistic feature for forming and unifying poetic (‘metrical’) units. Alliteration between consonants tends to be strictly phonemic (→ phoneme) while it is often found that any vowel is treated as alliterating with any other vowel irrespective of their actual → phonetic quality. This has been attributed to the glottal stop. In other words, alliteration is used as a poetic device, defining the rhythm of a verse-line. Thus, it has a structuring function. In an Old English alliterating verse, the stressed syllables usually have the same initial consonant; all vowels and the consonant clusters sc-, st-, and sp- alliterate with each other. An example of an Old English alliterative verse-line is Beowulf l. 18: Bēowulf wæs brēme, / blǣd wide sprang. An example of a line with alliterating vowels is Beowulf l. 3: hū þā æþelingas / ellen fremedon.

Allograph Allographs (e.g., ‹a›, ‹ɑ›, ‹A›) are sub-variants of a → grapheme. In most writing systems (such as the Latin script), graphemes can have allographs with the same function, such as upper-case (e.g., ‹A›) and lower-case (‹a›). For Old English runes, allographs for the phoneme /s/ include variants such as ‹ᛊ› and ‹ᛋ›. Allographs are given in pointed brackets ‹ ›. https://doi.org/10.1515/9783110796834-015

300

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Allophone (positional variant) Allophones are sub-variants (e.g., [l], [ɫ]) of a → phoneme (e.g., /l/). In Modern English, for example, the phoneme /l/ has two variants, depending on the context in which it occurs: one is termed clear l [l], which is always pronounced before a vowel (leave [li:v], hello [heləʊ]). The other variant is called dark l [ɫ], which occurs everywhere else (held [heɫd], fall [fɔ:ɫ]). Such allophones are also termed positional variants. Allophones are usually not differentiated in writing (e.g., [ɫ] in ‹held› – [l] in ‹leave› in British English). For example, in → Pre-Old English the allophones [a], [æ], [å], and [ε] were all represented by the rune ansuz a ᚨ. Allophones are given in square brackets [ ].

Anglo-Frisian Anglo-Frisian is a term for a) the traditionally assumed hypothetical, reconstructed common predecessor of the language stages Old English and Old Frisian, characterised by → major Anglo-Frisian sound-changes common to both. b) a rune-row: Compared to the Older → fuþark, the Anglo-Frisian → fuþorc has several new runes, e.g., āc ᚪ a, ōs ᚩ o, and æsċ ᚨ æ, which reflect similar changes in the two languages (cf. Waxenberger 2017). In this context, however, AngloFrisian is mainly used as a descriptive term, without implying a causal link between a purported common language stage (see the definition in a) above) and the development of the new rune row. Thus, the new rune āc ᚪ was used for the new long /a:/ developed both from Gmc. */aɪ/ in Pre-Old English and from Gmc. */au/ in Pre-OFrisian (→ monophthongisation). In any case, the modifications to the inherited rune-row found both in English and in Frisian inscriptions certainly suggest continuing close ties between the two language communities, regardless of their possible genetic relationship (cf. on this question see Hines 2017, 30 ff.). c) a special meaning in Vennemann, this volume: Anglo-Saxon or Frisian. The major Anglo-Frisian sound-changes are: 1. Anglo-Frisian Compensatory Lengthening: The loss of a nasal consonant (/m/, /n/) before /f/, /s/, /θ/ (= English th) caused lengthening and nasalisation (→ nasalisation of a vowel) of the preceding vowel; eventually the nasalisation was lost. Therefore, these vowels became OE /u:/ and /i:/; in the case of Gmc. /a/ there was also a change of vowel quality (see below). Modern German did not undergo this sound change, retaining the nasal consonant. 1.1. Gmc. *gans- developed to Pre-OE nasalised long open o /ɔ̃:/ (> /ɔ:/) and became /o:/ in OE (for the nasal consonant cf. English goose vs. German Gans)

A Concise and Selected Guide to Terminologies

301

1.2. Gmc. *munþa- developed to OE mūþ, which changed to ModE mouth (cf. ModG Mund). 1.3. Gmc. *fimf- developed to OE fīf, which became ModE five (cf. ModG fünf). 2. Fronting of Gmc. *a > OE æ (the sound /æ/ as in ModE cat) or OFris e: e.g., Gmc. *aski- developed into OE æsċ, which became ModE ash. OE WS ǣ (dǣd ‘deed’) 3. WGmc. *ā OE Angl. and OFris. ē (dēd ‘deed’) 4. WGmc. *ā + nasal > OE/OFris. ō (OE/OFris. mōna (‘moon’) These changes are now often placed within the larger perspective of a common North-Sea Germanic dialect continuum (→ Germanic languages) also partially including Old Saxon (cf. Nielsen 2001, 512 f.).

Anglo-Saxon (see below: 2. Archaeological and Art-Historical Terms)

Archetype As used by Bammesberger in this volume, an archetype is an original text of which versions or adaptations are made. If two or more texts have a set of specific features in common, they may be said to go back to the same archetype. For example, it is assumed that The Dream of the Rood (Vercelli Book), the short inscription on the Brussels Cross, and the runic poem on the Ruthwell Cross go back to the same archetype because they share some features, e.g., the OE phrase miþ blōde bestēmed (‘made wet with blood’): Ruthwell Cross (south-east face): The Dream of the Rood (l. 48b): Brussels Cross:

miþ b[l]odæ bistemi[d] Eall ic wæs mid blōde bestēmed blōde bestēmed

Asterisk A star (termed asterisk) * before a sound or word means that it has been reconstructed, i.e., the form is not actually documented (e.g., Gmc. *fehu ‘cattle’).

302

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Classification of vowels The → Pre-Old English vowels are depicted in a vowel chart, which is a simplified representation of the oral cavity (see below Fig. 1).

front

back high

low

Fig. 1: The Pre-Old English vowel chart.

Vowels can be classified in three ways: a) the height of the highest point of the tongue during the articulation (high vs. low), b) the position of the highest point of the tongue, i.e., whether the highest point of the tongue is more in the front of the mouth, or more in the back of the mouth (front vs. back or palatal vs. velar), c) the lips: they can be rounded or unrounded during articulation. High vs. low: If you pronounce the vowel /i:/ as in English bee and then the vowel /ɑ:/ as in English bah, and go back and forth several times, you will notice the different positions of the tongue and the directions in which it moves. While the tongue is humped up high in the mouth for /i:/, it is flattened out and relatively low in the mouth for /ɑ:/. For this reason, /i:/ is called a high vowel and /ɑ:/ a low vowel. The essential difference is the distance between the highest part of the tongue and the roof of the mouth. In a high vowel such as /i:/, the distance is very small as the highest part of the tongue is very close to the roof of the mouth (cf. Ladefoged 1975). Front vs. back (palatal vs. velar): If you pronounce /i:/ as in English bee, and then /u:/ as in English boo, and again go back and forth several times, you will notice that they are both high vowels. Relatively speaking, the hump for /i:/ is in the front of the mouth (the roof of the mouth, the so-called hard palate) and that for /u:/ is in the back of the mouth (the so-called velum). Therefore /i:/ is a front (palatal) vowel, whereas /u:/ is a back (velar) vowel (cf. Ladefoged 1975: 82).

A Concise and Selected Guide to Terminologies

303

According to points a) and b), /i:/ is a high front (palatal) vowel and /u:/ is a high back (velar) vowel. Regarding point c), /i:/ is an unrounded (spread) vowel as the lips are spread and not rounded. The English /u:/ is only slightly rounded.

Classification of consonants Consonants are classified in three ways (cf. Ladefoged 1975): a) Voice: A consonant can be voiced or voiceless. This means that the only difference between the phonemes /b/ and /p/ is that the former is voiced and the latter is voiceless. When the vocal cords are not open, as they are for normal breathing, but closed and vibrating, then this sound is called voiced. You can try it yourself by putting your hands on your throat or on your head: pronounce /p/ (not /pi:/), then you will not feel any vibration. If, however, you pronounce /b/ (not /bi:/), then you will feel the vibration of a voiced sound. b) Manner of articulation: When we articulate a consonant, the airstream is influenced by an obstruction, this means it cannot flow unhindered through the vocal tract (area from the glottis to the lips). Consonants are categorised by the manner in which the obstruction is overcome: Plosives (= stops): /p/, /b/, /t/, /d/, /k/, /g/. They usually consist of three phases: A blockage of the airstream is formed; the blockage is opened and the air escapes with an explosive sound. In the articulation of /p/ the lips are tightly closed, and the air pressure opens them. Fricatives: /f/, /v/, /θ/, /ð/, /s/, /z/, /ʃ/, /ʒ/. A narrowing of the air passage by the articulatory organs produces a sound when the airstream is forced through. In the articulation of /f/, for example, the narrowing is formed by the lower lip and the upper front teeth coming together. Affricates (a combination of plosive and fricative): /tʃ/, /dʒ/ Nasals (the air flows through the nose): /m/, /n/, /ŋ/. These are sounds created by blocking the outflowing air with the lips or at some point in the mouth while allowing the air to escape through the nasal cavity. In the articulation of /m/, for example, both lips are tightly closed. Laterals (the air flows along the sides of the tongue): /l/. Laterals are formed by placing the front or back of the tongue against the roof of the mouth (hard palate). However, the outflowing air is not blocked (cf. stops), rather, it escapes laterally around the sides of the tongue, which do not make contact with the roof of the mouth. Approximants (also called frictionless continuants): /r/, /w/, /j/. c) Place of articulation: This is the point of contact of the speech organs (e.g., teeth, lips) within the vocal tract where the obstruction occurs. Moving from front to back there are the following places of articulation: Bilabial (with both lips): /p/, /b/, /m/

304

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Labiodental (upper front teeth on lower lip): /f/, /v/ Interdental (the tip of the tongue touches both upper and lower front teeth): /θ/ (as in teeth), /ð/ (as in father) Alveolar (the tongue touches the alveolar ridge): /d/, /t/, /s/, /z/, /n/ Post-alveolar (the tip of the tongue touches behind the alveolar ridge): Received Pronunciation /r/ Palatal (the tongue touches the palate): [ɡ], [ḵ] (e.g., ModE give, kitchen) Velar (the tongue touches the velum): [ɡ], [ḵ] (e.g., ModE goose, cool).1

Fig. 2: The oral cavity (after Davis 1998, 29).

Compound A compound is a complex word made up of at least two lexical → morphemes, i.e., two elements that also occur as independent words, e.g., foot-stool from foot and stool. Compounding is a common type of word-formation in the → Germanic languages. An example in the OERC is hlafard hlāfard ‘lord’ (literally ‘bread-keeper’) developed from OE hlāf ‘bread’ (ModE loaf) and OE ward ‘keeper’.2 Compounding is an important strategy in Germanic name formation, e.g., as a possessive compound (bahuvrihi) (see Vennemann, this volume).

1 Note: /ɡ/ and /k/ can have palatal or velar allophones depending on the environment (→ allophone; also: → palatalisation/assibilation). 2 Note: The OE compound hlafard on the Ruthwell Cross was already an ‘opaque compound’ then, i.e., the individual elements (hlāf and ward) in the compound were no longer transparent for the user.

A Concise and Selected Guide to Terminologies

305

Editing runic inscriptions Runic inscriptions are edited in three steps: First, they are transliterated (→ transliteration) in bold and lower-case Roman letters: The Caistor-by-Norwich astragalus inscription, for example, has been transliterated as raïhan. For a different transliteration by Waxenberger (2020 and forthc.), retaining the yew-rune and the ansuz runes (rᚨᛇhᚨn), see → transliteration. Whenever possible and necessary, a phonetic → transcription is given if the sound level is clear; sometimes, only individual runes are transcribed. The transcription of the Caistor-by-Norwich inscription is given by Waxenberger as [råxån] (the circle above the vowels indicates nasalisation of the vowel). Finally, if possible, the → translation into Modern English is provided, which in this case is ‘of/from the roe-deer’.

Etymology Finding an etymology for a word involves not only searching for its origins of, but also studying how the form and meaning of a word change over time. This includes looking at the development of the sound-shape of a documented form as well as at its morphological shape and its meaning; for instance, the question of whether it is a simple word or possibly a derived or compound word (→ compound, morpheme). This is then compared to potentially related words in other languages, or other language stages of the same language. From this group of cognate words, either a → reconstruction of the predecessor form along with its meaning becomes possible, or a connection with a successor in a later language stage. All this together makes up what is called the etymology of a word. For example, the presumed etymology of the runic sequence transliterated (→ transliteration) as raïhan on the Caistor-by-Norwich astragalus is that it is the gen. sg. or dat. sg. of the precursor of OE rāha (in the Corpus Glossary), later rā, which then became Modern English roe (cf. ModG cognate Reh ‘roe-deer’). The runic form still seems to show a reflex of the original Gmc. diphthong */aɪ/; that was later monophthongised (→ monophthongisation), as can be seen in OE rā. Proposing an etymology for a word is an important step in the interpretation of a runic inscription, cf. Vennemann, this volume.

fuþark (Older fuþark) The term Older fuþark (named after the first six runes of the row) refers to the original rune-row that consists of 24 characters, used in Scandinavia until the 8th century AD and on the Continent until the early 7th century. In England, the Older fuþark developed to the → Pre-fuþorc and then to the → fuþorc (Old English).

Fig. 3: The Common Germanic fuþark and Pre-Old English innovations.

306 Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Fig. 4: The Old English fuþorc in ca. AD 650 and 750.

A Concise and Selected Guide to Terminologies

307

308

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

fuþorc (Old English) The term fuþorc refers to the rune-row that developed out of the → Pre-fuþorc after ca. AD 650 in England. It consists of 27 runic characters; in Northumbria in ca. AD 750, the rune-row had been extended to 31 runes.

Germanic languages The Germanic languages form a language family today spoken predominantly in north-western Europe. Germanic languages have a common ancestor, Proto-Germanic, which was spoken in the first millennium BC. Proto-Germanic was an IndoEuropean language which shared features with many other early languages in Europe, e.g., with the Greek, Italic, Celtic, Baltic, and Slavic languages. Eventually, Germanic languages differentiated themselves from other Indo-European languages by a consonant shift known as Grimm’s Law and some distinctive vowel-changes. Later, different dialects emerged and divided the language family further into East Germanic (e.g., Gothic), West Germanic (e.g., Old English, Old High German), and North Germanic (e.g., Old Norse). Regarding the further differentiation within the West Germanic group, the → Anglo-Frisian hypothesis is currently being re-evaluated in the light of recent suggestions by archaeologists (cf. Looijenga this volume; see also Kaiser 2021, 31– 33, for a detailed history of the Anglo-Frisian Hypothesis). The exact division of the Germanic languages has long been a matter of controversy. We have here retained the concept of North Sea Germanic (Old English, Old Frisian, partly Old Saxon), as it has come to play a more prominent role in the debate on Anglo-Frisian.3 The focus of English runology is on the North Sea Germanic continuum of West Germanic. For the Pre-Old English inscriptions there may be a connection to the North Germanic language stage Early Runic/Urnordisch. The periods given for the individual language stages are generally based on attested linguistic data; the apparent gaps in the continuity of some language stages of a certain language (see, e.g., Early Runic Frisian and Old Frisian) are due to a lack of data. The relationship between the different Germanic languages can be illustrated in a family tree (see Fig. 5), e.g., Early Runic is defined by Hans Frede Nielsen (2000, 32) as “the language of the 24-letter-futhark inscriptions of Scandinavia from ca. AD 200 to 500”; these texts “are distributed within a geographical area comprising

3 For a recent summary from the perspective of Old High German see Braune/Heidermanns (2018, 8‒9), with extensive references. For a general recent overview of the Germanic languages, see also Fulk (2018).

309

Fig. 5: The Germanic language family.

A Concise and Selected Guide to Terminologies

310

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Norway and Sweden as far north as respectively, Trøndelag and Uppland, the island of Gotland and Denmark (with Skåne and Slesvig), the medieval Danish territory exhibiting the largest concentration of Early Runic monuments”. Urnordisch is defined by Schulte (2018, 20) as ‘the language of the early runic inscriptions which covered wide areas of Scandinavia, that is today’s Norway, Sweden, Denmark to as far south as Northern Germany’. Urnordisch consists of two periods, the older (ca. AD 150–500) and the younger (ca. AD 500–700), according to Schulte.

Grapheme The term grapheme denotes the individual signs of a script (generally) representing a → phoneme (e.g., in the Latin script, ‹d›). A grapheme is given in pointed brackets ‹ ›. The sub-variants of a grapheme are called → allographs.

IPA (International Phonetic Alphabet) The International Phonetic Alphabet was developed by the International Phonetic Association. Its aim was to create a universal phonetic alphabet for all languages. In the following the most important phonemes of Present Day English (and two OE allophones) are listed in IPA.4 Short vowels: /ɪ/ as in hit /hɪt/ /e/ as in pet /pet/ /æ/ as in trap /træp/ /ʌ/ as in hut /hʌt/ /ʊ/ as in foot /fʊt/ /ɒ/ as in BE lot /lɒt/ /ǝ/ as in America /ǝˈmerɪkǝ/

Long vowels: /i:/ as in feet /fi:t/ /ɑ:/ as in father /fɑ:ðǝ/ (BE), /fɑ:ðɚ/ (AE) /ɔ:/ as in thought /θɔ:t/ /u:/ as in goose /gu:s/ /ɜ:/ as in BE bird /bɜ:d/ /ɝ:/ as in AE bird /bɝ:d/ /ɚ/ as in AE sister /sɪstɚ/

Diphthongs: /eɪ/ as in face /feɪs/ /aɪ/ as in nice /naɪs/ /ɔɪ/ as in choice /tʃɔɪs/ /aʊ/ as in house /haʊs/

BE /ɪǝ/ as in near /nɪǝ/ BE /eǝ/ as in care /keǝ/ BE /ʊǝ/ as in sure /ʃʊǝ/ /ǝʊ/ (BE); /oʊ/ (AE) as in low (BE /lǝʊ/, AE /loʊ/)

4 The presentation is based on: Jones et al. (2011).

A Concise and Selected Guide to Terminologies

Consonants: Voiceless plosives (stops): /p/ as in pea /pi:/ /t/ as in tea /ti:/ /k/ as in key /ki:/

311

Voiced plosives (stops): /b/ as in bee /bi:/ /d/ as in did /dɪd/ /ɡ/ as in give /gɪv/

Voiceless fricatives: Voiced fricatives: /f/ as in fat /fæt/ /v/ as in vat /væt/ /θ/ as in thin /θɪn/ /ð/ as in this /ðɪs/ /s/ as in sip /sɪp/ /z/ as in zip /zɪp/ /ʃ/ as in ship /ʃɪp/ /ʒ/ as in measure BE /ˈmeʒǝ/, AE /ˈmeʒɚ/ /h/ as in hat /hæt/ [x] as in German ach /ax/ and in OE worohtæ (Kirkheaton Stone) *[woroxtæ] [ç] as in German ich /ɪç/ and probably in OE berhtsuiþe (Thornhill Stone III) *[berçt]Voiceless affricate: /tʃ/ as in chin /tʃɪn/

Voiced affricate: /dʒ/ as in gin /dʒɪn/

Voiced nasals: /m/ as in mail /meɪl/ /n/ as in nail /neɪl/ /ŋ/ as in sing /sɪŋ/ Voiced lateral: /l/ as in lead /li:d/ Voiced approximants: /r/ as in red /red/ /w/ as in wet /wet/ /j/ as in yet /jet/ The prosodic symbol ˈ indicates that the primary stress is on a certain syllable, e.g., on the second syllable in America /ǝˈmerɪkǝ/.

i-umlaut (also known as i-mutation) i-umlaut or i-mutation is a → sound-change that occurred not only in Old English, but also in Old High German and Old Norse (→ Germanic languages). Velar (= dark/

312

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

back) → root vowels were palatalised (= fronted/brightened) when there was either an i or a j in the following syllable (→ classification of vowels). This process first yielded new → allophones, which then became → phonemes in the individual languages. As far as can be judged from the early runic material in England, i-umlaut was at work in the period from ca. AD 400–600, the new phonemes appearing in ca. AD 650. Example: Gmc. *fōt (singular) yielded Modern English foot, whereas the plural shows i-umlaut: The -i- in the original Gmc. plural form *fōt-iz ‘feet’ had an effect on the velar vowel ō, changing it first to a palatalised (fronted) vowel with lip-rounding œ/œ̄ (similar to the Modern German ö) and then to the unrounded vowel ē, yielding Old English fēt, which later becomes Modern English feet. The evidence of runic spelling clearly tracks the sequence of change in various cases.

Monophthongisation Monophthongisation is the process by which an original diphthong (a sequence of two vowels within a single syllable) becomes a monophthong (one vowel); e.g., in Frisian, the original Germanic diphthong */au/ became /a:/. See → relative chronology and → Anglo-Frisian, definition b).

Morpheme The morpheme is defined as the smallest meaning-bearing unit of a language. A word may be formed with one (e.g., foot) or several morphemes (e.g., the → compound foot-wear or the derivation footing).

Nasalisation of a vowel A vowel can have an additional quality, namely nasalisation. In American English vowels tend to be more nasalised than in British English. In Pre-OE vowels became nasalised through Anglo-Frisian Compensatory Lengthening (→ Anglo-Frisian, major Anglo-Frisian sound-changes).

North-Sea Germanic → Anglo-Frisian; → Germanic languages

A Concise and Selected Guide to Terminologies

313

Old English dialects Old English had several dialects. This entry only deals with the major ones. Generally, three main Old English dialects are distinguished, namely, West-Saxon (WS) in the South, Kentish (Kent.) in the South-East (Kent and neighboring areas), and Anglian (Angl.) in the Midlands and the North. Anglian can be subdivided into Mercian (basically the Midlands) and Northumbrian (generally, north of the Humber). These dialects are thought to be based on the speech of the different groups of Germanic settlers: West-Saxon is connected with a part of the Saxons (the SouthSaxons and East-Saxons hardly play a role in the OE dialects). Generally, Anglian is connected with the Angles, and Kentish with the Jutes, who settled in Kent and on the Isle of Wight. Most likely, Frisians also settled in England, but linguistically a specific link with one of the OE dialects is difficult to prove, although → Pre-Old English and Pre-Old Frisian are closely related as they have some sound-changes in common (→ Anglo-Frisian, major Anglo-Frisian sound-changes). The OE dialects are mainly distinguished by phonological differences (which can be deduced from spelling differences) and differences in vocabulary. Further differences, e.g., in inflectional morphology and word-formation, are not frequent. As far as Old English texts go, differences in syntax are difficult to prove (OE glosses usually follow the arrangement of the Latin matrix text and should only be used with great caution as evidence for Old English syntax). The map below shows political boundaries in ca. AD 825 based on SB (1965, no page number), who point out that it is not possible to determine dialect boundaries.

314

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Map 1: Old English dialect areas; map base: © EuroGeographics, linguistic details based on SB (1965): RuneS research unit Eichstätt-München.

A Concise and Selected Guide to Terminologies

315

Palatalisation/assibilation in Pre-Old English and Old English According to Campbell (1959, § 426), the consonants Gmc. */ɤ/ (> OE /g/) and */k/ “developed sensitivity” to their palatal (→ classification of vowels and consonants) environments, probably already on the Continent. Campbell (1959, § 427) points out that initially palatal k [ḵ] “probably differed from velar k hardly more than do the initial consonants” of ‹kit› [ḵɪt] and ‹cat› [kæt]. Based on the fact that palatalisation occurred in both OE and OFris., Hogg (1992, § 7.25) assumes that the Gmc. velar consonants (→ classification of vowels and consonants) had already acquired “a tendency to assimilate to a neighbouring front vowel at a very early stage”. Already in → Pre-Old English (Waxenberger 2021), the → allophones of the velar consonants */ɤ/ (> OE /g/) and */k/ became sensitive to an immediately adjacent palatal vowel or the approximant /j/, as a result developing palatal allophones. The palatal stops [ɡ:] (= long palatal g in gemination) and [ḵ] (= palatal k) eventually developed into voiced or voiceless affricates (→ classification of vowels and consonants): [ɡ] > /dʒ/ (= the sound depicted by ‹dg› in the word ‹bridge›), *[ḵ] > /tʃ/ (= the sound denoted by ‹ch› in the word ‹chase›) → IPA). This ‘sensitivity’ was to culminate in the long and complex processes of palatalisation (velar sounds becoming palatal) and assibilation (velar sounds becoming the affricates /tʃ/ and /dʒ/), enlarging the OE phoneme inventory by three new phonemes: /tʃ/, /dʒ/, and /ʃ/ (‹sh› as in the initial sound of ‹shoot›), and new palatal allophones (palatal g [ʝ] pronounced as the initial sound in the Modern English ‹yellow› (OE ġealu ‘yellow’) and palatal k [ḵ] as in the initial sound of ‹kitchen›).

Phoneme A phoneme is an abstract unit of a language system defined as the smallest meaning-differentiating unit. For example, in English, the voiced stop /b/ has a function different from the voiceless stop /p/. This means that words may differ in this one sound only, as can be seen in cap /kæp/ and cab /kæb/; cf. also hit /hɪt/ and heat /hi:t/, where there are two different vowel phonemes (short /ɪ/ and long /i:/). The sub-variants of phonemes are called → allophones. Notation: While phonemes are transcribed between slashes / / (→ transcription), allophones (→ phonetics) are placed in square brackets [ ]. For example, the phoneme /l/ in English has two allophones: clear [l] and dark [ƚ].

Phonetics Phonetics deals with the analysis and description of the sounds used in human speech, primarily in terms of how the various speech organs (particularly the

316

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

tongue, in combination with parts of the mouth, lips, throat and nose) are used to produce particular sounds (→ classification of vowels, classification of consonants; esp. Figs. 1 and 2). In relation to phonemics (→ phoneme), phonetic analysis is descriptively more detailed and language-independent, although phonetic variation may be grammatically and structurally insignificant. See also → transcription.

Pre-fuþorc The Pre-fuþorc represents an intermediate stage in the development from the Common Gmc. → fuþark to the OE → fuþorc. It covers the → Pre-Old English period from ca. AD 425 (Caistor-by-Norwich astragalus) until ca. AD 650 (Caistor-by-Norwich/Harford Farm brooch), which is the beginning of Old English proper (cf. Waxenberger 2019; forthc.). In this period, the two vowel runes of the Older fuþark, ansuz ᚨfuþark afuþark and ōþil ᛟfuþark ofuþark , probably each represented a set of allophones, some of which later developed to new→ phonemes and were therefore represented by new or modified runes in OE. The rune ansuz ᚨ afuþark probably denoted [a], [æ], [å] and i-umlauted [ɛ] of Gmc. */a/, and both [a:] and [æ:] of Gmc. */a:/. The allophones [æ] and [æ:] developed to phonemes and were then expressed by the rune-form ᚨ = æ in OE, whereas [a] and [a:] were rendered by the new rune āc ᚪ a in OE. Rune ōþil ᛟfuþark ofuþark denoted the allophones [o] and [o:] as well as the i-umlauted allophones [œ] and [œ:] in Pre-OE. The allophones [œ] and [œ:] developed to phonemes and were represented by the ‘recycled’ rune form ᛟ = œ in OE. The new sound /ɔ̃:/ (→ Anglo-Frisian major sound-changes 1.1) was denoted by the new rune ᚩ. From the 7th cent. onwards, the rune ᛟ was exclusively used for the new i-umlaut phoneme /œ(:)/ whereas long /ɔ̃:/ (> /ɔ:/) and Gmc. long /o:/ as well as short /ɔ/ and short /o/ had fallen together under /o:/ and /o/ in OE and were represented by the new ᚩ exclusively (Waxenberger forthc.). The end of these developments marks the transition from the intermediate stage of the Pre-fuþorc to the new rune-row of the OE fuþorc used in the inscriptions from the mid-7th century onwards.

Pre-Old English Pre-Old English, dated to ca. AD 425–650, is the immediate predecessor of Old English (OE), that is the dialectal continuum which was spoken in England after the → adventus Saxonum until 610/650 (Caistor-by-Norwich brooch: → Old English dialects). The Caistor-by-Norwich/Harford Farm brooch (ca. AD 610–650) marks the beginning of Old English proper because the new and modified runes, ᚪ a; ᛟ œ; ᚨ æ, with their new phonemic sound-values appear in this inscription for the first time. This demonstrates that the early sound changes (e.g., → monoph-

A Concise and Selected Guide to Terminologies

317

thongisation of Gmc. */aɪ/; → i-umlaut) had been phonemicised (→ phoneme) by AD 610/650.

Reconstruction Reconstruction is a method used in comparative historical linguistics. A reconstructed linguistic form represents the result of the comparison of a group of cognate words (i.e., words descended from the same precursor language, → etymology) and is regarded as a hypothetical, i.e., non-documented, linguistic precursor form, marked by an → asterisk. For example, the Germanic verb *kuss-ija- is the reconstructed precursor of English (to) kiss and German küssen, both showing the effect of → i-umlaut in the vowel. The basic theoretical assumption underlying this method is that of the regularity of → sound-change.

Root vowel The term root vowel refers to the stressed vowel of a word or of a word-stem, which can change due to ablaut (gradation). For example, Old English strong verbs changed their root vowels regularly: rīdan – rād – riden – (ge)riden (cf. Modern English ride – rode – ridden).

Sound-change Sound-change is defined as a change at the sound-level of a given language that is assumed to be regular, meaning that it will affect all words of a language containing the sound, or sequence of sounds, in question, regardless of the meaning or function of the word or → morpheme, and independent of the intention of the speaker. Such regular sound-changes characterise a specific idiom in a specific area during a specific period of time; as an example see → i-umlaut. Sound-changes form the basis for defining languages (English vs. Frisian) and language stages (→ Pre-Old English vs. Old English). Often, sound-changes affect not just one individual sound, but occur in conjunction with other sound-changes, e.g., the so-called Old High German consonant shift responsible for the difference in the underlined consonants of English two /t/ vs. German zwei /ts/, English pound /p/ vs. German Pfund /pf/, English make /k/ vs. German machen /x/.

318

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Transliteration In a transliteration, a word or text is transferred from one writing system (e.g., runes) into another, usually well-known writing system (e.g., the Latin script) so that a non-expert in the field can read the text.

Transliteration principles in Old English runology Runic inscriptions are normally transliterated. The letter chosen to represent the rune is usually one with a similar sound value, e.g., a for the ansuz-rune. However, as sound values of both runes and Latin letters are not static but vary in time and place (e.g., the letter in German and English orthography), this relation may be complex. In any case, a transliteration is not the same as a → transcription of sounds. Every edition of runic inscriptions uses transliterations (→ editing runic inscriptions). However, the different philologies follow different conventions, e.g., in Scandinavian Studies an uncertain rune is marked by a dot underneath it, e.g., ạ, while in English Studies the rune in question is italicised, e.g., a. Transliteration principles have been developed by various scholars over the years. The systems most commonly used are based on Dickins (1932) and Page (1973; 1984; 1999). The following principles, developed by Gaby Waxenberger for → Pre-Old English and Old English, are applied in the editions (forthc.) and explained below. Runes are transliterated in lower case and in bold print (e.g., fisc on the Franks Casket) whereas letters of the Latin script are represented by regular capital letters (e.g., HICFUGIANTHIERUSALIM on the Franks Casket). Mixed inscriptions are transliterated as follows: e.g., agrOf ‘engraved’ (Lancashire (Manchester) Ring). However, there are some exceptions: 1. The character eth ‹ð›, representing the → allophones [ð] and [θ], is not capitalised although it is not a rune, e.g., æðRED ‘Æthred’ (personal name) on the Lancashire (Manchester) Ring. It is given in lower case to avoid confusion with its upper-case counterpart ‹Ð› ([ð], [θ]) in the Latin script of the manuscripts. 2. In contrast to Dickins and Page, in Waxenberger’s (forthc.) transliterations, the runes are retained: a) when the rune itself has no equivalent in the Latin script of Modern English, e.g., yew-rune , ᛇ, and/or b) when its sound value is not clear and/or differs in different periods or dialect areas (e.g., the yew-rune , ᛇ, probably represents nasalised, long ī [ĩ:] in Pre-OE, but denotes the ach-sound [x] (the sound that occurs in the German word ach) in the OE of 8th century Northumbria (i.e., on the Ruthwell Cross).

A Concise and Selected Guide to Terminologies

319

In these cases, the rune itself is presented rather than a non-Latin script symbol (e.g., ï, i for the yew-rune). In the transliterations of the Pre-OE period (written in the → Pre-fuþorc: ca. AD 425–650), the runes ᚨ afuþark and ᛟ ofuþark are sometimes also retained in the transliterations. The subscripted index fuþark indicates that the rune in question is still the rune from the Common Germanic (= Older) fuþark and not yet the new rune āc ᚪ a or ōs ᚩ o of the OE fuþorc. This is done because both the allophones [a], [æ], and [å] of Gmc. short */a/ and the allophones [a:] and [æ:] of Gmc. long */a:/ were rendered by the old rune ansuz ᚨfuþark. The same is true for the old rune ōþil ᛟfuþark which must have denoted [o(:)] and also the → allophone [œ(:)] brought about by → i-umlaut. For the runes ^ and ᚴ denoting a (more) palatalised k (→ palatalisation/assibilation) in Pre-OE, ḵ is used in the transliteration and [ḵ] in the narrow → transcription. The same is true for the new rune no. 31 ᛤ ḵ on the OE Ruthwell Cross (ca. AD 750). For further deviation from Dickins and Page’s system see Waxenberger (forthc.: Chapter 1). Further transliteration symbols used by Waxenberger (forthc.) are:

Tab. 1: Transliteration symbols used by Waxenberger (forthc.). ]

] marks the break-off of an inscription on the left-hand side

[

[ marks the break-off of an inscription on the right-hand side

[.]

one character is missing or is illegible

[..]

two characters are missing or are illegible

[…]

three or more characters are missing or are illegible

[. ..]

This indicates that three characters are missing; however, they do not belong to the same word. In the following case, the first character belongs to the first word, whereas the other two missing characters belong to a second word. [...])ea[.]du[. ..] ([bih])ea[l]du[n] = WS behēoldon): RC NW

[?]

The question mark in square brackets expresses uncertainty as to whether the character is a rune, a miscut or something that cannot be defined, or even an empty space.

*

The asterisk * is used for reconstructed sounds, words, and runes.

)ha

Bind-runes have one stave in common. They are marked by an arc ) above the transliteration of the combined runes. ᚻᚪ = ᚻ and ᚪ The opaque bind-rune is presented as )ea by some scholars, but some use ea͜ with the arc below it.

It should be emphasised that the deviations from Dickins’s and Page’s transliteration system are triggered by the wish for more exact transliterations and transcriptions. Admittedly, this means that the transliterations based on these principles do

320

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

not always strictly follow the principle of transferring a runic character into a Latin character without taking the exact sound value into account but tend to give a more phonetically oriented transliteration. This principle is particularly applied to the Pre-OE inscriptions. As Waxenberger’s approach is a more phonetically oriented transliteration, also taking into account the allophonic level, she transliterates the inscription on the Chessell Down Scabbard Mouthpiece as follows: ᚨ

ᛸ

ᚩ

a ‘Ako

k

ɔ

: : :

ᛟ

ᚱ

ᛁ

ḵ œ chooser’

ᚴ

r

i

In an exclusively sign-oriented transliteration, the inscription on the Chessell Down Scabbard Mouthpiece, ᚨ

ᛸ

ᚩ

:

ᚴ

ᛟ

ᚱ

ᛁ

k3

o1

r

i

could be transliterated as: a1

k2

o2

:

a1 = the a ᚨ of the Older fuþark o1 = the o ᛟ of the Older fuþark o2 = the new o ᚩ of the Pre-fuþorc (from ca. AD 450/500 onwards) k2 ᛸ and k3 ᚴ are variants of k in the Pre-fuþorc A slightly different system for the English and Frisian inscriptions is employed by the RuneS database (‹www.runesdb.eu›: transliteration principles), which differs, for example, in the following items: a = the a ᚨ of the Older fuþark a = the new a ᚪ of the Pre-fuþorc (from ca. AD 575/610 onwards) o = the o ᛟ of the Older fuþark o = the new o ᚩ of the Pre-fuþorc (from ca. AD 450/500 onwards) k1 = ᚲ; k2 = ᛸ; k3 = ᚴ are variants of k in the various rune-rows

Transcription A phonetic transcription (→ phonetics) involves presenting the sounds in the standard phonetic alphabet → IPA, devised for the representation of any language so as

A Concise and Selected Guide to Terminologies

321

to make all languages comparable at the sound level. In a broad transcription only the → phonemes are transcribed. The broad or phonemic transcription of the second part of the Chessell Down Scabbard Mouthpiece is /kori/, whereas the allophonic or narrow transcription (in square brackets) transcribes the → allophones: [ḵœri].

Translation The aim of a translation is to express the meaning of an utterance in a given language in a different language or a different language stage. Runic texts are normally rendered in a threefold way: 1. → transliteration, parsing into individual words and marking vowel length (= reading/interpretation), 2. → translation, and 3. → transcription. A transcription is only given if possible and/or necessary. Cf. Bammesberger, Kopár, Vennemann, Waxenberger, this volume, and → editing runic inscriptions.

2 Archaeological and Art-Historical Terms John Hines

Adventus Saxonum This Latin term is used to describe the arrival and settlement of Germanic people (→ Germanic languages) from the Continent in Great Britain as the starting point of the establishment and development of → Anglo-Saxon culture in an area that came to be England. This was described and dated, after a fashion, as having taken place around the year AD 449 by Bede in his work that was completed around 731: Historia ecclesiastica gentis Anglorum (The Ecclesiastical History of the English People), I.xv and V.xxiv. It is not clear when this now common expression was first used: It is partly a scholarly adaptation of Bede’s statement (I.xv) that the Germanic incomers Advenerant … de tribus Germaniae populis fortioribus, id est Saxonibus, Anglis, Iutis (‘they came from three of the more powerful nations of Germania, that is the Saxons, the Angles, the Jutes’). This highly contested term therefore makes explicit reference only to Saxon colonists, but the newcomers apparently also included not only Angles and Jutes but Franks and other Scandinavian and Continental peoples too. Indeed, the formulation is an extraordinary misrepresentation of the fact that the closest phrasing Bede did use was adventus Anglorum (as ‘the coming of the English’) in his Chronica Maiora, s. a. 627 (a date there identified as being approximately 180 years after the coming of the English) and Historia ecclesiastica V.xxiii (where the year 731 is calculated as approximately 285 years after that event). Recently, the traditional view of this period has been discussed in a more differen-

322

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

tiated way from various disciplinary angles (Sims-Williams 1983), cf. Scull and Looijenga, this volume.

Anglo-Saxon a) Anglo-Saxon is a term now used mainly in History and Archaeology, principally to label the period of more than six centuries between the end of Roman Britannia and the Norman Conquest of England (AD 1066), and also as a general term (1) for the Germanic-speaking inhabitants of Great Britain in that period and (2) for their culture (→ material culture). The first known appearance of the term is as an exonym used by the Langobard Paulus Diaconus in the 8th century (Lentzsch 2018, 17), derived from the ethnonyms Anglii and Saxones (‘Angles’ and ‘Saxons’): two of the Germanic ethnic groups involved in the → adventus Saxonum (cf. Scull, this volume). In the 19th century the term was regularly used for the phase of the English language represented from the first known records to the earlier 12th century for which Historical Linguistics now prefers the more specific label Old English, focusing on the convergent development of the various Germanic dialects into a new variety of North-Sea Germanic, → Anglo-Frisian, Germanic languages, Old English dialects, Pre-Old English. b) For a special use of the term, see Looijenga, this volume.

Bayesian modelling Bayesian modelling is a means of processing statistics that allows one to refine measurements of probability in circumstances where there are multiple known probabilities that are or can be inter-related within a defined structure or model. The primary data of this kind are known as prior information while the outputs of modelling are known as posterior estimates. In respect of the current volume, Bayesian modelling is referred to as it has been used to enhance the information value of → radiocarbon dating. The model may involve secure factual evidence of → relative chronology (e.g., → stratigraphical sequence) and/or a hypothetical belief of what the relative chronological sequence is. In the latter case, the process of Bayesian modelling can also measure the fit between the hypothesis and the prior data, with a conventionally set threshold of agreement at which the model can be regarded as valid (see Hines/Bayliss 2013, 73–87).

Bracteate A bracteate is an artefact of thin precious or semi-precious metal foil (from Latin bractea ‘gold leaf’). In Archaeology and Numismatics the term may refer to single-

A Concise and Selected Guide to Terminologies

323

sided coins produced in Germany in the 12th and 13th centuries, but it is best known as the label for a series of disc pendants, mostly gold, produced in Scandinavia and some adjacent areas of the Continent and in England during the 5th and 6th centuries. The prototypes for the earliest bracteates appear to lie in Late Roman-period gold medallions; the bracteates have a pressed design in the central field of the disc on one face only and usually have attached rims and loops; they may also have an outer decorated zone on the face. The 5th- and 6th-century bracteates are divided into four main types (A–D) according to the principal elements of the motif in the central field. Later bracteates of the 7th century decorated in Style II zoomorphic art (→ Style: Salin’s Style I and II) are known from England, and a further series (Type E) from 7th- to 8th-century Gotland.

Calibration In the context of the present volume calibration is an essential process for converting the probabilistic measurements of the radiocarbon age of a sample into a radiocarbon date; → radiocarbon dating.

Central place The term central place has become widely used in Archaeology recently to label sites of special prominence in a particular district or region, usually because of the special nature of the finds made there and the activities that those finds appear to represent, and/or because of the distinct constructed character of the site. Evaluation of the relevance of the term, or attempts to explain the location of central place, will also usually consider the precise geographical and topographical features of the site in relation to the surrounding area. Central places are usually considered to be primarily sites of social pre-eminence in the regional population, but also to have special economic functions too, and may additionally serve as or be closely associated with regionally primary religious sites (cf. Pestell, this volume).

Correspondence analysis Correspondence analysis is a computerised technique for exploring the structure of data that can be presented in the form of a matrix or spreadsheet. The data-matrix will typically involve a series of objects to be compared (which may be individual artefacts, or assemblages of artefacts in the form of, for instance, individual graveassemblages or discrete hoards), the composition of which can be expressed in terms of the presence, absence, or quantity of a series of variables. Correspondence

324

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

analysis is commonly used to search for sequences of regular obsolescence and replacement in the artefact inventory over time in order to produce a → seriation of the objects, but it can also be used to examine different forms of clustering relevant to the topic under investigation (Hines/Bayliss 2013, 60–73).

Dendrochronology This term denotes the method of dating samples of timber by identifying sequences of variation in the size of the annual growth-rings that reflect the specific growing conditions, year by year. Normally 60 growth-rings are required for the sample to be datable. Dendrochronological master sequences, with which the sample can be compared in order to identify it with a particular period of growth, are species-specific: in northern Europe the main species used for dendrochronology are oak and pine. Master sequences also vary from region to region, so that the identification of a sample with a master sequence will also indicate what part of Europe that sample has come from (dendroprovenancing). The presence of sapwood (outer growth-rings) and, better still, bark allows for more precise identification of the felling year of the timber; in exceptional cases the presence of microfaunal remains at a specific stage of the annual life-cycle under the bark can narrow the date of burial of the sample to a span of just a few weeks within the year in question (cf. Baillie 1995).

Deposition This is the process whereby archaeological evidence is placed in the ground to be retrieved by excavation. Deposits may be deliberate, as in the case of the burial of a corpse along with grave goods, or represent casual accumulation over time. The latter are often referred to, especially in Scandinavian archaeology, as culture layers. A closed deposit is an archaeological context of this kind, the contents of which can be assumed to represent the discrete period of time in which the deposit was formed with no later disturbance either removing contents or intruding later items into the context.

Iconography Iconography is the general term used to label those features of visual art which have a readily identifiable and describable significance beyond and separate from the object or pattern portrayed. These may take the form of symbols (e.g., the Christian cross), emblems (e.g., ferocious animals, real or mythological) or figures (e.g., Mary Magdalene). In relation to the Early Medieval Period, iconography is used in

A Concise and Selected Guide to Terminologies

325

contexts where the art in question is understood to have religious or mythological meaning or connotations, so that we consistently refer to ‘Christian (or ‘pagan) iconography’ but ‘secular/social symbolism’.

Landscapes of power This term can refer to the social and political demonstration of power, e.g., through the possession of materials or objects or by certain skills that the more powerless lack. An example may be runic literacy in the Migration Period, which would demonstrate the power of the literate elite.

Material culture Material culture is usually regarded as the primary evidence and domain of interest of Archaeology. It refers to the material life of the past and to the ‘clues’ that provides for what has happened, in contrast to ‘historical’ evidence derived from textual/documentary sources. Rather than the reconstruction of the past alone, however, an equal concern of Archaeology is to understand what the dynamics, constraints and patterns in the relationship between the human population and its material circumstances as a whole may be. Following the science of Anthropology, culture is regarded as a way of life that is variable between groups rather than entirely predetermined by human biology; the regularities in behaviour and practice of particular populations, areas and periods may allow us to define or at least to suggest the identification of specific cultures (e.g., → Germanic, Celtic or Slavic Cultures; → Anglo-Saxon or → Viking Cultures), in a manner analogous to the identification of specific languages as opposed to human language as a universal phenomenon. The material aspects of material culture will normally include certain regular categories of manufactured products such as clothing; utensils, tools and weapons; vehicles. It will include the built environment and how the landscape is managed and exploited for production. It may even include forms of modification of the human body, such as distorting the shape of the skull, filing and decorating teeth, or tattooing.

Radiocarbon dating This is a method of measuring the age of a sample of organic material by comparing the levels of stable and unstable/decaying radioactive carbon isotopes within it. The ratio between these isotopes will start to diverge from a particular equilibrium in the living organism at the point of death and the cessation of the ingestion of current atmospheric carbon through the process of respiration.

326

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

The radiocarbon age thus calculated is expressed in terms of ‘n ± y’ years BP (= ‘Before Present’), where ‘Present’ has been set by convention at AD 1950, and the expression ‘± y’ represents 1 standard deviation (1σ in conventional mathematical notation) or circa 68 % of probability. The radiocarbon age, which in fact is in units of ‘radiocarbon years’ that are not identical with calendar years, then has to be calibrated (→ calibration), because we know that the levels of radioactive carbon in the atmosphere are not constant and have fluctuated in the past. This process produces distributions of probability that can be expressed in a simplified term either as ‘cal BC/AD DATE + x years’, or ‘cal BC/AD DATE–DATE (95 % probability)’, or more precisely represented purely in graphic form. It is also recognised that organisms living in different environments (e.g., seawater, freshwater, terrestrial) will be ingesting different reservoirs of atmospheric carbon, and processes have been developed to adjust for these effects (Hines/ Bayliss 2013, 35–73; see also Hines, this volume).

Relative chronology Relative chronology is a chronological sequence involving objects or entities placed in relationships of ‘earlier than’ or ‘later than’, and sometimes ‘contemporary with’. Contemporaneity may mean no more than that entities B and C are known both to be later than an entity A and earlier than an entity D, but no more specific relationship between them can be defined. Relative chronological dating plays a major role in Archaeology (→ stratigraphical sequence) and may be combined with → radiocarbon dating in → Bayesian modelling. Relative chronology is also used in Historical Linguistics, where it may help in relative dating of, among others, runic inscriptions; e.g., the inscription on the Caistor-by-Norwich astragalus raïhan [råxån] must be earlier than the → monophthongisation of Gmc. *ai to Pre-OE ā because the diphthong is still in place (→ sound-change).

Rehydroxylation dating This is a relatively new laboratory technique for determining the age of ceramic material, which will principally mean pottery, tile or brick produced by firing moulded clay. Such fired clay will have been ‘dehydroxylised’ (i.e., had all combined water, H2O, driven off) during manufacture, and the molecules of the fabric will then start to recombine with atmospheric moisture at a very slow but generally constant rate. The amount of water thus recombined can be captured and measured by controlled dehydroxylation of a sample of the material.

A Concise and Selected Guide to Terminologies

327

Sceatt (or sceatta, plural sceattas) This term designates a type of silver coin which had emerged by ca. AD 675 and came to have a wide distribution across England and Frisia and into a number of trading sites subject to Danish control. Except in Scandinavia, the coin-type was superseded by what are known as early silver pennies or deniers from the third quarter of the 8th century: coins which are usually slightly heavier and use a thinner and broader flan. In Northumbria, however, the sceattas evolved into the lowvalue bronze coinage known as stycas for most of the 9th century (see Hines/Bayliss 2013, 493–515; also: Hines, this volume).

Seriation (also known as Ordination) Seriation refers to the process of placing a set of archaeological features or complexes (e.g., grave-assemblages, artefacts of relatively complex composition, works of art, hoards) in a sequence or order of similarity so that those closest together in the series will be most similar in terms of the variables compared (→ correspondence analysis) and those at the opposite ends of the series should be much more different. Seriation is often used to try to identify sequences of gradual chronological change, but it can also reflect scales of variation up and down a scale of social rank and wealth, for instance, or over a geographical range.

Stratigraphical sequence Stratigraphical sequence between layers and/or features is a form of → relative chronology that is archaeologically fundamental and totally secure, whereby, for instance, a layer of earth that underlies another layer, or is cut by pits, post-holes or ditches, must be earlier than any of those overlying layers and features (Harris 1979). Even without direct ‘vertical’ stratigraphical contiguity, it can be shown that features in a ‘horizontal’ stratigraphical relationship to one another ‒ for instance, graves in a regular row, or a group of pits or a ditch which block the doorway of a building ‒ may also have determinable relative-chronological relationships (cf. Hines and Hills, this volume).

Style Style is in a manner analogous to the identification of specific cultures (→ material culture) and languages, particular styles can be defined where certain features of technique, form and/or composition recur amongst a group of products and are

328

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

sufficiently salient for the observer to recognise their similarity. Amongst art-styles there is a form of generic life-cycle that follows a more predictable route from phases of intense creativity through periods of greater stylisation and a simplification of the elements and motifs of the style, before a new burst of creativity introduces a new style. The recognition of this cycle has led to typological classification of artstyles and their assignation to a historical sequence as an instrument for dating artefacts in a non-utilitarian way in Archaeology and Art History (cf. Hines, Hills, Marth, and Webster, this volume). Salin’s Style I, II, etc.: This term refers to an art style coined by Bernhard Salin in Die altgermanische Thierornamentik (1914). The art styles were dominant between the 5th and the 7th centuries in northern Europe. Anthropomorphic designs typical of Roman animal ornament became abstracted in Germanic art. One distinguishes between Salin Style I (5th/6th cent.), which features animals (quadrupeds) in a close‐knit pattern. In Salin’s Style II (6th/7th cent.) the animals are larger, have ribbon and tendril designs, and their body parts (e.g., tails) are often interlaced. Salin’s Style III (late 7th/8th cent.) has a more naturalistic representation (cf. Webster, this volume).

Typology In Archaeology, typology is the primary taxonomy by which archaeological data are ordered and discussed. The term tends now to be used for a level of the taxonomic process that is a sub-division of a primary functional identification (e.g., ‘brooch’, ‘sword’, ‘cart’, etc.). It can, however, be pursued to many alternative levels of detailed sub-division, for which labels such as ‘class’, ‘type’ and ‘form’ may be encountered, not in any agreed order of precedence. It is, however, generally recognised that typological classification is largely heuristic, so that the ‘best’ typological sub-division depends entirely upon the analytical and interpretative objectives of the archaeological study in hand (Hines/Bayliss 2013, 20–25). In Linguistics, typology refers to one way of classifying languages, i.e., by their structural and functional features such as word order, morphological structure, structure of the vocabulary etc. Another way of classifying languages is by their genealogy, i.e., their descent from a common precursor language (stage), cf. → PreOld English as the precursor of Old English.

Viking Age and Viking Period Many histories of the Vikings indicate that we can identify the very day on which the Viking Age began: 8 June 793, with the famous sack of Lindisfarne. That view was in fact contested even in the 9th century, for the initial West Saxon version

A Concise and Selected Guide to Terminologies

329

of the Anglo-Saxon Chronicle insists that a violent encounter, perhaps at Portland, sometime between 787 and 793 was the first Viking raid – and pointedly omits to record the Lindisfarne raid (ASC MS A). The sack of Lindisfarne was of course ‘iconic’: this starting point thus serves to define the beginning of a historically conceived Viking Age, a period documented in textual sources and also widely re-conceived and represented in many different literary genres. Defining when the Viking Age ended is a more complex matter. Scandinavian military ambitions overseas continued at a major scale at least into the 13th century; in an English view, the failed invasions of Haraldr Harðráði in 1066 and Knútr Sveinsson in 1083 serve as appropriate end-markers (Loyn 1994, 64–69). It is, however, consistent with a cultural characterisation of the Viking Age as one of violent opposition between a pagan Scandinavia and Christian Europe that Scandinavian historians commonly describe a gradual end of the Viking Age in step with the progress and consolidation of Christianisation. A separate but equally important perspective on the definition of a Viking Period is the archaeological one: defined by aspects of → material culture. Phenomena that have been considered definitive in this respect are, for instance, the occurrence of Insular fine metalwork (i.e., from Britain and Ireland) as possible ‘loot’ in Norwegian grave finds, or the evolution of a series of urbanised port sites around Scandinavia reflecting new levels and types of long-distance maritime contact and exchange (Myhre 1993; Brink/Price 2008, 83–149). Archaeological and historical perceptions of the Viking Period cannot be completely independent, even if, as disciplinary frameworks, they are autonomous. Some attempts have been made to use the term Viking Period consistently for archaeological dating, and Viking Age when the perspective is historical and literary, and the regularisation of this distinction would be welcome.

References Baillie, Mike. 1995. A Slice through Time: Dendrochronology and Precision Dating. London: Routledge. Braune, Wilhelm and Frank Heidermanns. 2018. Althochdeutsche Grammatik 1: Laut- und Formenlehre. 16th ed. Berlin/Boston, MA: De Gruyter. Brink, Stefan and Neil Price (eds.). 2008. The Viking World. London: Routledge. Campbell, Alistair. 1959. Old English Grammar. Oxford: Oxford University Press. Davis, John F. 2015. Phonetics and Phonology. Stuttgart: Klett. [1st ed. 1998]. Dickins, Bruce. 1932. “A System of Transliteration for Old English Runic Inscriptions”. Leeds Studies in English 1: 15–19. Fulk, Robert D. 2018. A Comparative Grammar of the Early Germanic Languages. Amsterdam/ Philadelphia, PA: Benjamins. Harris, Edward C. 1979. Principles of Archaeological Stratigraphy. London: Academic Press. Hines, John and Alex Bayliss (eds.). 2013. Anglo-Saxon Graves and Grave Goods of the 6th and 7th Centuries AD: A Chronological Framework. Leeds: Society for Medieval Archaeology.

330

Gaby Waxenberger, Kerstin Kazzazi, and John Hines

Hines, John. 2017. “The Anglo-Frisian Question”. Frisians and their North Sea Neighbours: From the Fifth Century to the Viking Age. Eds. John Hines and Nelleke IJssennagger. Woodbridge: The Boydell Press. 25–42. Hogg, Richard. 1992. A Grammar of Old English. Vol. 1: Phonology. Oxford: Blackwell. Jones, Daniel, Peter Roach, Jane Setter, and John Esling. 2011. Cambridge English Pronouncing Dictionary. 18th ed. Cambridge: Cambridge University Press. Kaiser, Livia. 2021. Runes Across the North Sea from the Migration Period and Beyond: An Annotated Edition of the Old Frisian Runic Corpus. Runische Schriftlichkeit in den germanischen Sprachen 2. Berlin: De Gruyter. Ladefoged, Peter. 1975. A Course in Phonetics. New York: Harcourt Brace Jovanovic. Lentzsch, John. 2018. Beharrungsvermögen und Verdrängung. Polytheisten und Christen in den angelsächsischen Reichen des 7. Jahrhunderts. Berlin: De Gruyter. Loyn, Henry. 1994. The Vikings in Britain. 2nd ed. Oxford: Blackwell. Myhre, Bjørn. 1993. “The Beginning of the Viking Age – Some Current Archaeological Problems”. Viking Revaluations. Eds. Richard Perkins and Anthony Faulkes. London: Viking Society for Northern Research. 182–204. Nielsen, Hans Frede. 2000. The Early Runic Language of Scandinavia, Studies in Germanic Dialect Geography. Heidelberg: Winter. Nielsen, Hans Frede. 2001. “Frisian and the Grouping of the Older Germanic Languages”. Handbuch des Friesischen/Handbook of Frisian Studies. Eds. Horst Haider Munske, Nils Århammar, Volker F. Faltings, Jarich F. Hoekstra, Oebele Vries, Alastair G. H. Walker and Ommo Wilts. Berlin/Boston, MA: De Gruyter. 512–523. Page, Raymond Ian. 1973. An Introduction to English Runes. London: Methuen. Page, Raymond Ian. 1984. “On the Transliteration of English Runes”. Medieval Archaeology 28: 22‒45. Page, Raymond Ian. 1999. An Introduction to English Runes. 2nd ed. Woodbridge: The Boydell Press. RuneS database: URL: ‹www.runesdb.eu›, last accessed 7 October 2021. SB = Brunner, Karl. 1965. Altenglische Grammatik: Nach der angelsächsischen Grammatik von Eduard Sievers, 3rd ed. Tübingen: Max Niemeyer. Schulte, Michael. 2018. Urnordisch: Eine Einführung. Wien: Praesens. Sims-Williams, Patrick. 1983. “The Settlement of England in Bede and the Chronicle”. AngloSaxon England 12: 1–41. Waxenberger, Gaby. 2017. “The Development of the Old English fuþorc”. Von den Hieroglyphen zur Internetsprache: Das Verhältnis von Laut, Schrift und Sprache/ From Hieroglyphs to Internet Language: The Relation of Script, Sound and Language. LautSchriftSprache 2/ScriptandSound 2. Eds. Gaby Waxenberger, Hans Sauer, and Kerstin Kazzazi. Wiesbaden: Reichert. 209–247. Waxenberger, Gaby. 2019. “Absolute Chronology of Early Sound Changes Reflected in Pre-OE Runic Inscriptions”. NOWELE 72: 60–77. Waxenberger, Gaby. 2020. “Appearances are Deceiving: The Caistor-by-Norwich Astragalus (ca. AD 425–475)”. Ihr werdet die Wahrheit erkennen – Ye Shall Know the Truth. Eds. Hans Sauer and Rüdiger Pfeiffer-Rupp. Trier: WVT. 47‒56. Waxenberger, Gaby. 2021. “The Runes c ċēn ᚳ and g ġ(i)efu ᚷ and Their Velar Counterparts in the OE fuþorc and Pre-fuþorc”. Wege zur Konfiguration der Zeichen-Phonem-Beziehung. LautSchriftSprache 3/ScriptandSound 3. Eds. Alessia Bauer and Gaby Waxenberger. Wiesbaden: Reichert. 185‒204. Waxenberger, Gaby. Forthc. A Phonology of Old English Runic Inscriptions with a Concise Edition and Analysis of the Graphemes. RGA-E. Berlin/Boston, MA: De Gruyter.

Indices Index of Objects This index lists only objects bearing runic inscriptions (pure runic or mixed-script). Naming and numbering follow the runological conventions; this may lead to discrepancies in comparison to naming and numbering of the same object in other disciplines (for a detailed discussion of crossdisciplinarily different classification systems, see Kopár, this volume). Page numbers in bold indicate an image of the object, page numbers in italics mean that this page contains a transliteration of the runic inscription on the respective object. For more information on the objects, see the RuneS database (‹runesdb.eu›).

Aberlady cross 243 æniwulufu thrymsa 106, 116 Alnmouth stone 86, 87, 88, 90, 93 Amay comb 106, 120 Arlon bulla = Arlon capsule 110 Arum sword 118 Ash/Gilton sword-pommel 69, 118 Baconsthorpe tweezers 16, 153, 154, 170 Bakewell stone 86 Bardney Abbey pin 167 Barkston tweezers see Honington tweezers Belgorod comb 120 Bergakker scabbard mouthpiece 11, 103–122 Bernsterburen staff 118 Bewcastle Cross 70–75, 83, 86, 90, 91, 223, 231, 268, 272, 277, 284, 289, 292 Bingley font or cross-base 88, 89 Binham bracteates 68, 146, 147, 149, 172 Blythburgh tablet 16, 155, 159, 292 Boarley disc brooch 67, 68 Borgharen belt buckle 107, 110 Bramham Moor ring 285 Brandon antler handle 153, 166, 271 Brandon pin 153 Brandon tweezers 153, 170 Bristol/Linstock Castle Agate ring 288 Britsum rune stick 117, 119 Broughton Lodge brooch 67 Brunswick Casket see Gandersheim Casket Caistor-by-Norwich astragalus 52, 70, 106, 112, 118, 139, 140, 144, 212, 305, 316, 325, 326 Caistor-by-Norwich brooch (see also = Harford Farm brooch) 145, 272, 316 Charnay-lès-Chalon brooch 110 https://doi.org/10.1515/9783110796834-016

Chéhéry disc brooch 110 Chessell Down pail 66, 67 Chessell Down scabbard mouthpiece 69, 118, 119, 320, 321 Chester-le-Street stone 84, 88, 90, 92 Cleatham hanging bowl 67, 68 Collingham stone 86, 90, 272 Congham sword pommel 149 Coppergate York lead pendant 169 Crowle stone 86, 89, 272 Cumwhitton pin 167 Dagworth spindle whorl 167 Derby(shire) bone plate 285, 290, 293 Dover brooch 67, 68 Dover stone 81, 89, 92 Dunton lead plaque see Toftrees lead plaque Elisenhof comb 120 Engers, Mayen bow brooch 108 Eye fragment 150 Fallward (Wremen) footstool 108, 111 Falstone stone 81, 89, 93, 272 Ferwerd comb case 118 Folkestone/Glasgow thrymsa/tremissis 116 Franks Casket 2, 10, 16, 24, 41, 70, 115, 221– 223, 253–265, 267–292, 318 Fréthun sword pommel 110 Frienstedt comb 120 Gandersheim Casket 16, 98, 127–136 Great Urswick stone 11, 84, 93, 94, 96, 98, 269, 272, 278, 281, 288 Grenay sword pommel 110

332

Indices

Hackness stone 88 Hamwih bone see Southampton bone Hantum bone plate 118 Harford Farm brooch (see also = Caistor-byNorwich brooch) 9, 60, 67, 118, 145, 146 Harlingen solidus 117, 118 Hartlepool stones 88, 92, 289, 292 Heacham tweezers 152, 170 Hitsum bracteate 106 Ingham urn 140, 141 Ipswich belt buckle 165 Ipswich whale bone fragment 166 Kent gold coin 118 Kent sceattas 118 Keswick runic disc 166, 167 Kingmoor amulet ring 285 Kirby Misperton graffiti 81, 88 Kirkheaton stone 93, 277, 278, 281–283, 311 Lackford urn 140, 141 Lancashire (Manchester) ring 285, 318 Lancaster cross 93, 95, 272, 277, 289, 291, 292 Leeds stone 88, 90, 98 Lindisfarne stones 86, 88, 92, 95, 296 Lundenwic sheep’s vertebra (see also = London National Portrait Gallery bone) 166 London National Portrait Gallery bone (see also = Lundenwic sheep’s vertebra) 272, 281, 292 London Thames scramasax 287 Loveden Hill runic pot (see also = Loveden Hill urn) 51 Loveden Hill urn (see also = Loveden Hill runic pot) 60, 70, 112, 118, 141, 143, 272 Maughold stone 272 Midlum sceat 118 Monkwearmouth stones 88, 90, 92, 93, 272 Monte Sant’ Angelo graffiti 88, 271, 272, 277, 278, 281, 283, 291, 292 Mortain casket 277, 278, 279, 281, 283, 285 Morton strapend 288 Mote of Mark bone 267 Near Fakenham lead plaque see Toftrees lead plaque Near March lead plaque 16, 160, 161, 162 Norham fragments 88

Oostum comb 116, 118, 120, 204 Overchurch stone 93, 272 Quidenham lead pendant 169 Rasquert sword handle 118 Reedham spindle whorl 168 Rome catacomb ad Duas Lauros graffiti 272 Rome Cimitero di Commodilla graffito 272 Rome graffiti 88 Roughton disc 166, 167 Ruthwell Cross 8–10, 21–42 70, 71, 72, 73, 74, 75, 76, 77, 84, 88, 90, 91, 223, 231, 233, 234, 245, 256, 268, 269, 272, 276, 277, 281, 284, 286, 289–292, 295, 301, 304, 318 Saltfleetby spindle whorl 167 Sandwich/Richborough stone 81, 89 Sarre sword-pommel 69 Schweindorf solidus 106, 117, 118, 205 Scotterthorpe plaque 164, 292 Sedgeford spoon handle 156 Setre comb 120 Shropham lead plaque 16, 157, 158, 165 skanomodu solidus 6, 8–9, 106, 115, 117, 199– 216 Skelton-in-Cleveland sundial 89 Southampton bone = Hamwih bone 115, 118 Spong Hill runic pot (see also = Spong Hill urn) 45–56, 60 Spong Hill urn (see also = Spong Hill runic pot) 70, 92, 112, 141, 142 St Benet lead plaque 16, 162–165 St Cuthbert’s coffin 70, 221, 225, 226, 234, 237, 268 St Dizier sword pommel 110 St John’s College, Oxford, MS 17, fol. 5v 162 St Ninian’s Cave stone 93, 267 Stoke Quay buckle 16 Suffolk shilling 118 Surrey Docks pin-heads 167 Thames mount 292 Thames scramasax 92, 272, 288 Thames Valley gold coin 118 Thornhill stones 86, 93, 94, 270, 272, 275, 281, 282, 285 Tiluwald sceat 151, 152 Tiverton spindle whorl 167

Index of Persons

Toftrees lead plaque 160, 165 Toornwerd comb 118, 120, 204 Undley bracteate 68, 114, 118, 149, 212 Vimose comb 120 Wakerley brooch 67 Watchfield mount 69, 118, 272 Welbeck Hill bracteate 67

333

West Heslerton brooch 66, 67 Westeremden B rune stick 117, 118 Westeremden weaving-slay 204, 205 Whitby comb 10, 118, 120, 223, 278, 289–292, 295 Whithorn stones 86, 90, 267, 272 Wijnaldum pendant 117 Wijnaldum piece of antler 117 Wishaw spindle whorl 167 Wremen footstool see Fallward footstool

Index of Persons The following index contains only names of historically documented persons, e.g., “Bede, the Venerable”; no personal names recorded as names only, e.g., “Lyl” (Great Urswick). It also does not list the names of academic scholars working on runic objects. For these names, please consult the reference lists of the contributions in this volume. The information on the persons is not complete. It is only meant to place the person in the respective historical context. The non-bold cross-references are index-internal, the bold cross-references refer to the “Guide of Terminologies” (p. 299 ff.).

Acca, Bishop of → Hexham from 709–732 75 Ælfwold, Abbot of St. Benet’s Home Abbey at the time of → Harold Godwineson 164, 165 Æthelstan (ca. 894–939), first king of the → Anglo-Saxons 130, 131 Æthelwald, Bishop of → Lindisfarne from 721– 740 74 Alcuin (ca. 735–804), renowned scholar from Northumbria at the Frankish court of Charlemagne 249 Aldfrith, King of Northumbria (685–704) 239, 250 Aldhelm (ca. 639–709), Abbot of Malmesbury, Bishop of Sherborne 75 Alhfriþ, King of Deira (655–664) 71 Anna, King of East Anglia from ca. 636–654 156 Arbogast (d. 394), Romanised Frankish military leader and governor of Gaul, son of → Bauto 108 Augustine of Canterbury (d. 604/605), “Apostle of the English”, first bishop of Canterbury (from 597) 237, 245 Bauto (d. 388), Frankish military leader in Roman service, father of → Arbogast 108

Bede, the Venerable (b. 672/673 near → Wearmouth, d. 735 in the monastery of → Jarrow), renowned Northumbrian scholar, author of the Historia ecclesiastica gentis Anglorum (Ecclesiastical History of the English People) (→ Anglo-Saxon) 14, 73, 74, 75, 106, 114, 121, 138, 185, 227, 230, 250, 321 Bedwin, attested as Bishop of Elmham for the year 673 143 Benedict Biscop (ca. 628–690), founder of → Wearmouth-Jarrow Priory 225, 229, 237 Beonna, King of East Anglia from ca. 749 152 Cassiodorus (ca. 485–585), Roman statesman and scholar, in the service of the Ostrogothic king Theoderic in Ravenna, and founder of the monastery Vivarium 73, 229 Cassius Dio (ca. 155–ca. 235), Roman historian 121 Ceolfrith (d. 716), Abbot of → WearmouthJarrow 229, 237 Charles III (879–929), King of France 131 Cnut the Great (d. 1035), King of England from 1016–1035, also King of Denmark, and Norway 158, 164

334

Indices

Columba (d. 579), saint and founder of Iona Abbey 234 Constantius II., (317–361), Roman Emperor, son of Constantine the Great 112 Constantius Chlorus (ca. 250–306), Roman emperor, father of Constantine the Great 104 Cuthbert of Lindisfarne (ca. 634–687), Northumbrian monk, Bishop of Hexham (684–685) and Lindisfarne (685–687), revered as a saint; his relics were taken from Lindisfarne to Durham, where the coffin (→ St Cuthbert’s coffin) is now kept at the Cathedral 23, 70, 225, 226, 229, 230, 234, 237 Cyneburh, Queen Consort of King Oswald of Northumbria (637–642) 71 David, King of the United Kingdom of Israel and Judah (cf. Old Testament) 238, 241 Eadberht (d. 768), King of Northumbria 239 Eadfrith (d. 722), Bishop of → Lindisfarne 227 Eadgifu (d. ca. 951), daughter of Edward the Elder, King of Wessex; became queen of the West Franks, wife of Charles the Simple 131 Eadgyth (ca. 910–936), Queen Consort of → Otto I 130, 135 Eadmund, King of East Anglia from 855–869, after his death revered as St Edmund 138, 146, 158 Edward II (1284–1327), King of England 248 Edward the Confessor (1003–1066), King of England, last king of the House of Wessex 158, 168 Edwin (ca. 584–633), King of Northumbria 71 Emperius, Johann Ferdinand Friedrich (1759– 1822), professor and museum director in Brunswick, Germany (→ Gandersheim Casket) 129, 130, 133, 135 Felix of Burgundy (d. 647 or 648), first bishop of East Anglia 169 Franks, Augustus Wollaston (1826–1897), eponym of the → Franks Casket 221 Fraomar, King of the Alamannic Bucinobantes from 370–371 108 George III (1738–1820), King of (the United Kingdom of) Great Britain and Ireland 199, 201, 213

George IV (1762–1830), King of the United Kingdom of Great Britain and Ireland and King of Hanover 213 Gildas the Wise (ca. 500–570), British monk, author of De Excidio et Conquestu Britanniae 106, 185 Gregory I, also – the Great, (ca. 540–604), Pope and Bishop of Rome; sent out → Augustine to missionise the Anglo-Saxons 231, 238 Haraldr III Harðráði (b. 1015 in Norway; d. on 25 September 1066 in the Battle at Stamford Bridge), Norwegian king 329 Harold Godwineson (1022–1066), also – Harold II, last crowned Anglo-Saxon king, died in the Battle of Hastings against → William the Conqueror 164 Hassel, Johann Georg Heinrich (1770–1829), German geographer 129, 130 Herebericht, priest at → Monkwearmouth, died probably in the first quarter of 8th century 242, 250 Honorius = Flavius Honorius (384–423), Roman emperor 199 Hugh de Montfort (d. ca. 1088), Norman nobleman in the service of → William the Conqueror 168 Jerome (ca. 342–420), saint and church father, translated Bible into Latin 230, 246 Jurminus (also – Hiurmine of Blythburgh), 7th-century AngloSaxon prince, son of → Anna 156 Knútr Sveinsson, also Canute IV (1043–1086), King of Denmark 329 Luke, the Evangelist (cf. New Testament) 239, 240 Mallobaudes, 4th-century Frankish king who also served in the Roman army 108 Merobaudes (d. 383 or 388), Frankish officer in Roman military service 108 Offa, King of Mercia from 757–796 158 Otto I, also – the Great, (912–973), Holy Roman Emperor 130, 131, 135

Index of Place-Names

Paulus Diaconus (ca. 720–ca. 799), Langobard historian 122 Penda (ca. 605–655), King of Mercia 156 Pliny the Elder (d. 79 AD), Roman philosopher, historian, and naval commander; author of Naturalis Historia 121 Procopius (ca. 500–ca. 560), Greek historian 121 Ptolemy (ca. 100–ca. 160), Greek mathematician and astronomer from Alexandria 121 Sergius (ca. 650–701), Pope and Bishop of Rome 74 St Edmund see → Eadmund Stephen of Ripon (ca. 670–ca. 730), author of the hagiography Vita Sancti Wilfrithi (Life of St. Wilfrid) (→ Wilfrid) 249 Symeon of Durham (d. after 1129), monk of Durham Priory and English chronicler 74 Tacitus (ca. 58–ca. 120), Roman historian and politician, author of the historical work Germania 121

335

Theodore (602–690), 7th archbishop of Canterbury 143 Theudebert I (ca. 500–547 or 548), Frankish (Merovingian) king 114 Titus Caesar Vespasianus (39–81), Roman military leader and later Emperor 237 Venantius Fortunatus (ca. 540–600/610), Bishop of Poitiers 110, 114 Wilfrid (ca. 634–709/710), Bishop of York 249 William de Warenne (ca. 1030–1088), Norman nobleman in the service of → William I the Conqueror 152 William I the Conqueror (ca. 1028–1087), Duke of Normandy, from 1066 King of England 152 Willibrord (ca. 658–749), Bishop of Utrecht 239 Wulfstan (d. 1023), Bishop of London, Archbishop of York, and Bishop of Worcester 257

Index of Place-Names Due to the focus of this volume being on the Old English Runic Corpus, the majority of places mentioned are located in the United Kingdom and are listed without a country code. For placenames outside of the UK, the country of location is indicated by the following international country code used on car number plates: B D DK F I

Belgium Germany Denmark France Italy

Abercorn 230 Aberlady 241, 242 Abingdon 184 Amay, B 109, 120 Arlon, F 109 Arum, NL 109 Auzon, F 221 Baconsthorpe 16, 153 Bamburgh 239, 249 Bardney 167

IE IL LU N NL

Ireland Israel Luxembourg Norway Netherlands

RUS S SYR VA

Russia Sweden Syria Vatikan

Barkston 170 Bawsey 171 Belgorod, RUS 120 Bergakker, NL 11, 103, 107–109 Bernsterburen, NL 109 Bewcastle 72, 223 Billingford 45 Binham 15, 50, 68, 146–149, 172 Blakeney 147 Blekinge, S 171 Blythburgh 16, 155, 156, 159

336

Indices

Borgharen, B 109 Brandon 16, 153, 155–156, 159, 166, 170–172 Brescia, I 222, 234, 236 Brighthampton 184, 189 Brinton 50, 147 Britsum, NL 109 Broughton Lodge 68 Brussels, B 91, 301 Buckenham 159 Buckland 68 Burton in Kendal 95 Bury St Edmunds 138, 146, 156, 169 Caister-on-Sea 164 Caistor-by-Norwich 52 Caistor St Edmund 139, 144, 146, 166, 172 Charnay-les-Chalon, F 109 Chéhéry, F 109 Cleatham 52 Clermont(-Ferrand), F 221 Coldingham 249 Congham 150 Coppergate 169 Croft 51 Crondall 148 Crowland 162 Cumwhitton 167 Dagworth 167, 168, 170 Damascus, SYR 74 Deerhurst 87 Donderberg, NL 110, 113 Dublin, IE 72, 88 Durham 73, 74, 230, 238, 246, 268 East Anglia, kingdom 137–172 East Walton 152 Echternach, LU 239, 245, 248 Edgefield 148 Ely 138, 162 Eriswell see Lakenheath 150 Eye 150 Fakenham 160 Fallward, D 103, 106, 111 Feddersen Wierde, D 108 Fen Drayton 241 Ferwerd, NL 109 Field Dalling 147 Finglesham 237

Flixborough on Humberside 171, 172 Florence, I 221, 264 Fréthun, F 109 Gandersheim, D 127–136 Grenay, F 109 Grimston 51 Gudme, DK 50, 148 Gutenstein, D 241 Halton 87, 95 Hantum, NL 109 Harford Farm 146 Harlingen, NL 109 Hartlepool 230, 231, 249 Hastings 152, 164, 168 Haughley 168 Heacham 152, 153, 156, 170 Hexham 75, 230, 249 Hitsum = Hitzum, NL 109 Hoddom 230 Hogebeintum (Dutch) = Hegebeintum (Frisian), NL 109 Hope 95 Hough-on-the-Hill 143 Ingham 140 Iona 234 Ipswich 16, 138, 146, 157, 165, 166 Isle of Wight 69 Jarrow (see also → Wearmouth-Jarrow) 87, 222, 227, 229, 230, 231, 237, 238, 239, 242, 245, 246, 249, 250, 296 Jerusalem, IL 74, 237, 245, 249 Kantens, NL 109 Keswick/Eaton 166 Kirby Misperton 88 Kirkby Wharfe 95 Kirkdale 87 Lackford 140, 141 Lakenheath 150 Lancaster St. Mary 95 Lastingham 230 Leeds 98 Lindisfarne 23, 74, 93, 95, 221, 222, 225, 227, 230, 238, 239, 242, 246, 248, 249, 250, 268, 296, 327, 328

Index of Place-Names

Little Carlton 171 London 167 Loveden Hill 51, 172 Lund, S 119 Lundenwic 166 Maastricht, NL 107 Maeshowe 88 March 162 Melrose 230 Monkwearmouth 239 Monte Gargano, I 81, 88 Mucking 70, 189 North Elmham 45, 141, 143, 172 Norwich 138, 139, 144, 152, 160, 166, 168, 170 Obrigheim, D 241 Oostum, NL 109 Oxford 162 Patching 148 Pensthorpe 51 Peterborough 162 Petersfinger 184 Portland 329 Quidenham 169 Ramsey 162 Rasquert, NL 109 Raynham 160 Reedham 168, 169 Rhenen, NL 110, 113 Ringlemere 70 Ripon 230, 249 Rome, I 81, 88, 229, 237, 272 Rothbury 75 Roughton 166 Ruthwell 21–42, 72, 84, 90, 223 Ryther 171 Saltfleetby 167 Samson, B 110 Schweindorf, D 109 Sedgeford 156, 157 Shropham 157, 159, 165 Snettisham 157 Solway Firth 71, 74 Southmere 157

337

Spong Hill 12, 45–56 60, 70, 141, 143 Squillace, I 229 St Benet, Horning 162, 164, 165 St Petersburg, RUS 73, 221 Staunch Meadow 153 St-Dizier, F 109 Stoke Quay 165 Streoneshalh 292, 295 Sutton Hoo 139, 148, 234, 237 Thetford 138, 139, 157, 160 Thorney 162 Tiel, NL 107, 108, 110, 111 Tiverton 167 Toftrees 160, 165 Toornwerd, NL 109 Torslunda, S 241 Trier, D 239 Tynemouth 230, 249 Undley 15, 149, 150 Uppåkra, S 148 Urswick 84, 94 Vatican, VA 74 Venta Icenorum 144 Vercelli, I 9, 29, 91, 301 Vron, F 110 Wasperton 190 Watchfield 69 Wearmouth 221, 222, 225, 227, 229, 230, 231, 237, 238, 239, 242, 245, 246, 249, 250, 296 Wearmouth-Jarrow 248, 267, 296 Weasenham 170 Welbeck Hill 68 Wells-next-the-Sea 148 West Stow 150 Westeremden, NL 109 Westergo, NL 104, 121 Whitby 171, 223, 230, 231, 249, 279, 295, 296 Whithorn 74, 75, 90 Wijnaldum, NL 109 Willoughby-on-the-Wolds 68 Winchester 170 Wishaw 167 York 169, 230, 237, 249