eBooks

Words, Words, Words: Philology and Beyond

Festschrift for Andreas Fischer on the Occasion of his 65th Birthday

0215
2012
978-3-7720-5435-8
978-3-7720-8435-5
A. Francke Verlag 
Sarah Chevalier
Thomas Honegger

This Festschrift comprises a series of papers written in honour of the philologist Andreas Fischer, on the occasion of his sixty-fifth birthday. As in Andreas Fischer's own research, the main focus of the volume is on words: words in modern varieties, such as emergent conjunctions in Australian, American and British English, words in their cultural and historical context, such as English keywords in Old Norse literature, and words in a diachronic perspective, such as Romance suffixation in the history of English. Many contributions are anchored in the philological tradition that has informed much of Andreas Fischer's own scholarship, such as the study of verbal duelling in the late thirteenth-century romance Kyng Alisaunder. Others examine the construction ofdiscourses, such as those surrounding the Black Death. The volume, with its innovative studies,offers fascinating insights into words, discourses,and their contexts, both past and present.

<?page no="0"?> Sarah Chevalier / Thomas Honegger (eds.) Words, Words, Words: Philology and Beyond Festschrift for Andreas Fischer on the Occasion of his 65 th Birthday <?page no="1"?> Sarah Chevalier and Thomas Honegger (eds.) Words, Words, Words: Philology and Beyond Festschrift for Andreas Fischer on the Occasion of his 65 th Birthday <?page no="2"?> Andreas Fischer <?page no="3"?> Sarah Chevalier / Thomas Honegger (eds.) Words, Words, Words: Philology and Beyond Festschrift for Andreas Fischer on the Occasion of his 65 th Birthday <?page no="4"?> Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http: / / dnb.d-nb.de abrufbar. Cover illustration: Beowulf by Anke Eißmann Picture of Andreas Fischer by David Werner We would like to thank the following institutions and foundations for their support: Dr. Wilhelm Jerg-Legat Hochschulstiftung der Universität Zürich Zürcher Universitätsverein © 2012 Narr Francke Attempto Verlag GmbH + Co. KG Dischingerweg 5 · D-72070 Tübingen Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Gedruckt auf chlorfrei gebleichtem und säurefreiem Werkdruckpapier. Internet: www.francke.de E-Mail: info@francke.de Druck und Bindung: Hubert & Co, Göttingen Printed in Germany ISBN 978-3-7720-8435-5 <?page no="5"?> Table of Contents Sarah Chevalier & Thomas Honegger Introduction vii-viii List of Publications by Andreas Fischer ix-xviii Dieter Bitterli Two Old English Prose Riddles of the Eleventh Century 1-11 Lukas Bleichenbacher Discourse about Diglossia Swiss German and Standard German in the Newspapers 13-22 Margaret Bridges Verbal Duelling on the Beach: Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 23-40 Hans-Jürgen Diller Discourse, Matrix and Genre in Historical Semantics: ofermōd and emotion 41-58 Udo Fries Melancholy Accidents in Early English Newspapers 59-76 Andreas H. Jucker “What’s in a name? ” Names and Terms of Address in Shakespeare’s Romeo and Juliet 77-97 Dieter Kastovsky Recursiveness in Word-Formation Grammatical, Conceptual and Typological Factors 99-108 Hans-Peter Naumann English Keywords in Old Norse Literature 109-125 Pam Peters Emergent Conjunctions 127-144 <?page no="6"?> Ursula Schaefer Romance Suffixation in the History of English New Ways of Solving an Old Problem 145-164 Daniel Schreier Words and the New Englishes Borrowing as Evidence of Contact and Ancestral Effects 165-180 Annina Seiler The Function of the Sword-Hilt Inscription in Beowulf 181-197 Gunnel Tottie On the History of try with Verbal Complements 199-214 Richard J. Watts The Black Death and the Development of English A Tale of Two Archives 215-235 <?page no="7"?> Sarah Chevalier & Thomas Honegger Introduction This volume has been prepared to honour our colleague and friend, Andreas Fischer. As his former students, we are delighted to be able to pay tribute to him with this Festschrift, on the occasion of his sixty-fifth birthday . Andreas Fischer was born in Basel on 25 January 1947. He attended the Humanistiches Gymnasium there, and went on to study English, German and Art History at the University of Basel; one year of his undergraduate studies was also spent at the University of Durham. He completed his studies in 1975 with a doctorate, summa cum laude, on lexical aspects of dialects in the south-west of England. For the following six years Andreas Fischer worked as a teaching and research assistant at the English Department of the University of Basel. He completed his Habilitation, on engagement, wedding and marriage in Old English, in 1981 and subsequently became Privatdozent at the University of Basel. In 1984, he left Switzerland again for a year, this time not to study but to teach, namely as Visiting Associate Professor at the University of Michigan, Ann Arbor; and this time not alone but accompanied by his wife Melinda. Andreas Fischer’s professorial appointment came soon after the couple’s return to Switzerland. In 1985, he was appointed Professor of English Philology at the University of Zurich. The couple moved to Zurich, where their daughter Emma was born. Andreas Fischer has had a long and distinguished career at the University of Zurich. For a quarter of a century he has taught and inspired hundreds of students, published significant research in the fields of lexicology and English historical linguistics, and made a vital contribution to the running of the university. His teaching has covered a wide range of topics and we mention a few courses here which have been particularly memorable for us: “Animals in Medieval Literature”, which led to Thomas Honegger’s interest in this topic and his dissertation on that very theme; “Scandinavian-English Language Contact” co-taught with Hans-Peter Naumann, which included a field trip to the Orkney and Shetland Islands; “English in Australia”; “The Bible in English”; and “Case Studies in Bilingualism”, which laid the groundwork for Sarah Chevalier’s Habilitation on trilingual language acquisition. His inspirational teaching is further attested by the number of stu- <?page no="8"?> Sarah Chevalier & Thomas Honegger viii dents who have chosen to write their theses with him: Andreas Fischer has supervised 182 Lizenziat theses and, to date, 27 doctoral dissertations, with several more still being completed. Andreas Fischer’s research has centred on lexis, especially from a historical perspective, as is evidenced by his many publications in this field (e.g. “The Vocabulary of Very Late Old English”, 1996); this interest in lexis includes issues of lexicography (e.g. “Global English as a Challenge for Lexicography”, 2002). Further areas of interest are historical linguistics in general (e.g. “The Hatton MS. of the West Saxon Gospels: The Preservation and Transmission of Old English”, 1997), as well as sociolinguistics (e.g. “Language and Politics in Switzerland”, 2001). Perusal of his publication list also reveals a focus on the analysis of linguistic data and concepts through classification (e.g. “Graphological Iconicity in Print Advertising: A Typology”, 1999 and “The Notional Structure of Thesauruses”, 2004). This ability to make sense of phenomena via various ways of ordering and structuring them has no doubt played a role in Andreas Fischer’s illustrious administrative career, which began with the position of Head of the English Department. Later, he was appointed Vice-Dean, then Dean of the Faculty of Arts, and following, Vice-President of the University. Andreas Fischer’s administrative career has culminated in his current position as President (Rektor) of the University of Zurich. But it has not only been a gift for analysis that has led him to succeed in and enjoy his work; as Andreas Fischer has told us, he is basically “interested in people”. This warmth and interest has been appreciated by colleagues and students alike. Turning to the present volume, it is both colleagues from around the world as well as former students who have contributed to this Festschrift. As in Andreas Fischer’s own research, the focus is on words: descriptive adjectives in newspapers, emergent conjunctions, the inscription on the sword-hilt in Beowulf - to mention just a few areas of analysis. The studies appear in alphabetical order according to the name of the author; an abstract precedes each paper and a biographical sketch of the author follows. We hope that Andreas Fischer, as well as other linguists, will take pleasure in the findings, discussions, and analyses in Words, Words, Words: Philology and Beyond. <?page no="9"?> Sarah Chevalier & Thomas Honegger List of Publications by Andreas Fischer 1976 1. Dialects in the South-West of England: A Lexical Investigation. (The Cooper Monographs 25). Berne: Francke. 2. “Wörterbuch mit Anspruch auf Universalität: Zum Abschluss der Neubearbeitung des ‘Muret-Sanders’”. Neue Zürcher Zeitung 164, 16 July 1976, p. 37. 3. “Harold Pinter”. Basler Nachrichten 130, 8 June 1976, p. 25. 1978 4. “Hamlet als Historie? Überlegungen zu Brechts Deutung von Shakespeares Hamlet”. Deutsche Shakespeare-Gesellschaft West, Jahrbuch 1978/ 79. Heidelberg: Quelle & Meyer, pp. 40-57. 5. Charles Dickens. Meistererzählungen, aus dem Englischen übersetzt von Trude Fein. Zürich: Manesse. Auswahl und “Nachwort”, pp. 551-568. [see also 42.] 6. Review of Karl-Gustav Ek. The Development of OE æ (i-mutated a) before Nasals and OE æ in South-Eastern Middle English. (Acta Universitatis Lundensis, Sectio I, 22; Lund: CWK Gleerup, 1975). English Studies 59, p. 369. 1979 7. “Poetry and Drama: Pinter’s Play The Birthday Party in the Light of his Poem ‘A View of the Party’”. English Studies 60, pp. 484-497. 8. Review of Steven H. Gale. Butter’s Going Up: A Critical Analysis of Harold Pinter’s Work (Durham, North Carolina: Duke University Press, 1977). English Studies 60, pp. 529-533 (531-533). 9. Review of Rüdiger Imhof. Harold Pinters Dramentechnik: Gestalterische Mittel im Kontext des Gesamtwerks (Gesamthochschule Wuppertal, Schriftenreihe Literaturwissenschaft, Band 2; Bonn: Bouvier Verlag Herbert Grundmann, 1976). English Studies 60, pp. 529-533 (529-531). 10. “Ein ungewöhnlicher Atlas”. Basler Zeitung 136, 14 June 1979, p. 27. <?page no="10"?> Sarah Chevalier & Thomas Honegger x 1980 11. Review of S. E. Gontarski. Beckett’s Happy Days: A Manuscript Study (Columbus, Ohio: Publications Committee, The Ohio State University Libraries, 1977). English Studies 61, pp. 571-572. 12. Review of Horst Oppel (ed.). Das englische Drama der Gegenwart: Interpretationen (Berlin: Erich Schmidt, 1976). English Studies 61, pp. 84-86. 13. Review of Clas Zilliacus. Beckett and Broadcasting: A Study of the Works of Samuel Beckett for and in Radio and Television (Acta Academiae Aboensis, Ser. A, Vol. 51, Nr. 2; Åbo: Åbo Akademi, 1976). English Studies 61, pp. 379-382. 1981 14. Arthur Conan Doyle. Sherlock Holmes-Geschichten/ Der Hund von Baskerville, aus dem Englischen übersetzt von Trude Fein. Zürich: Manesse. Auswahl und “Nachwort”, pp. 413-428/ 31. 1983 15. Review of Gertrud Walter. Kompendium Didaktik Englisch (München: Franz Ehrenwirt, 1981). Buchbesprechungen, Beilage zur Schweiz. Lehrerzeitung und zum Basler Schulblatt 3, p. 4. 1984 16. “Linguistic Aspects of Interior Monologue in James Joyce’s Ulysses”. Anglistentag 1982, Zürich: Vorträge. Eds. Udo Fries und Jörg Hasler. Giessen: Hoffmann, pp. 239-251. 17. John Galsworthy. Meistererzählungen, aus dem Englischen übersetzt von Irma Wehrli. Zürich: Manesse. Auswahl und “Nachwort”, pp. 459-471. 1986 18. Engagement, Wedding and Marriage in Old English. (Anglistische Forschungen 176). Heidelberg: Carl Winter. 19. “A Room with a View”. LSA [Ann Arbor, Michigan] 9/ 2, pp. 11-14. 1987 20. Rudyard Kipling. Meistererzählungen, aus dem Englischen übertragen von Sylvia Botheroyd, Monika Kind, Sabine Kipp, Ilse Leisi und Irma Wehrli. Zürich: Manesse. “Nachwort”, pp. 419- 441. <?page no="11"?> List of Publications by Andreas Fischer xi 21. Review of John M. Kirk, Stewart Sanderson and J. D. A. Widdowson (eds.). Studies in Linguistic Geography: The Dialects of English in Britain and Ireland (London: Croom Helm, 1985). Anglia 105, pp. 447-454. 22. Review of Martyn F. Wakelin. The Southwest of England (Varieties of English Around the World, Text Series 5; Amsterdam and Philadelphia: John Benjamins, 1986). English World-Wide 8, pp. 301-304. 1988 23. “Context-Free and Context-Sensitive Literature: Sherwood Anderson’s Winesburg, Ohio and James Joyce’s Dubliners”. In Neil Forsyth (ed.). Reading Contexts. (SPELL, Swiss Papers in English Language and Literature 4). Tübingen: Gunter Narr, pp. 13-31. 24. “How to Create a World: Beginnings in Fiction”. Bulletin CILA 48, pp. 29-44. 25. Review of Peter M. Anderson. A Structural Atlas of the English Dialects (London, New York, Sydney: Croom Helm, 1987). English World-Wide 9, pp. 313-318 (315-318). 26. Review of Christine Fell, Cecily Clark and Elizabeth Williams. Women in Anglo-Saxon England and the Impact of 1066 (A Colonnade Book; London: British Museum Publications, 1984). Anglia 106, pp. 488-492. 27. Review of Clive Upton, Stewart Sanderson and John Widdowson. Word Maps: A Dialect Atlas of England (Cartography by David Brophy; London, New York, Sydney: Croom Helm, 1987). English World-Wide 9, pp. 313-318 (313-315). 1989 28. (ed.). The History and the Dialects of English: Festschrift for Eduard Kolb. (Anglistische Forschungen 203). Heidelberg: Carl Winter. 29. “Aspects of Historical Lexicology”. In Udo Fries and Martin Heusser (eds.). Meaning and Beyond: Ernst Leisi zum 70. Geburtstag. Tübingen: Gunter Narr, pp. 71-91. 30. “Lexical Change in Late Old English: From æ to lagu”. In Andreas Fischer 1989 (ed.). The History and the Dialects of English: Festschrift for Eduard Kolb. (Anglistische Forschungen 203). Heidelberg: Carl Winter, pp. 103-114. 31. Review of Wilhelm Busse. Altenglische Literatur und ihre Geschichte: Zur Kritik des gegenwärtigen Deutungssystems (Studia humaniora, Band 7; Düsseldorf: Droste, 1987). Archiv für das Studium der neueren Sprachen und Literaturen 226, pp. 139-141. <?page no="12"?> Sarah Chevalier & Thomas Honegger xii 1990 32. “Story and Discourse in Sir Gawain and The Franklin’s Tale”. In Rüdiger Ahrens (ed.). Anglistentag 1989 Würzburg: Proceedings. Tübingen: Max Niemeyer, pp. 310-319. 33. “Strange Words, Strange Music: The Verbal Music of the ‘Sirens’ Episode in Joyce’s Ulysses”. In Margaret Bridges (ed.). On Strangeness. (SPELL, Swiss Papers in English Language and Literature 5). Tübingen: Gunter Narr, pp. 39-55. [see also 59.] 1991 34. (with Daniel Ammann). An Index to Dialect Maps of Great Britain. (Varieties of English Around the World, General Series 10). Amsterdam and Philadelphia: John Benjamins. 1992 35. “Laughing and Smiling in the History of English”. In Wilhelm G. Busse (ed.). Anglistentag 1991 Düsseldorf: Proceedings. Tübingen: Max Niemeyer, pp. 51-62. 36. Review of Wolfgang Viereck (dialectological editor) in collaboration with Heinrich Ramisch. The Computer Developed Linguistic Atlas of England 1 (Computational production: Harald Händler, Petra Hofmann, Wolfgang Putschke; Tübingen: Max Niemeyer, 1991). English World-Wide 13, pp. 302-306. 1993 37. “Bedbur, an Old English Ghost Word? ” In Klaus R. Grinda and Claus-Dieter Wetzel (eds.). Anglo-Saxonica: Beiträge zur Vor- und Frühgeschichte der englischen Sprache und zur altenglischen Literatur. Festschrift für Hans Schabram. München: Wilhelm Fink, pp. 327- 333. 38. “Medieval English Studies in Switzerland”. Medieval English Studies Newsletter (Tokyo) 29, pp. 1-6. 1994 39. (ed.). Repetition. (SPELL, Swiss Papers in English Language and Literature 7). Tübingen: Gunter Narr. 40. (with Roland Lüthi). “Der deutsche Chaucer: Eine bibliographische Übersicht mit Kommentar”. Archiv für das Studium der neueren Sprachen und Literaturen 231, pp. 44-58. 41. “‘Sumer is icumen in’: The Seasons of the Year in Middle English and Early Modern English”. In Dieter Kastovsky (ed.). Studies in Early Modern English. (Topics in English Linguistics 13). Berlin, New York: Mouton de Gruyter, pp. 79-95. <?page no="13"?> List of Publications by Andreas Fischer xiii 1995 42. Charles Dickens. Weihnachtsgeschichten, aus dem Englischen übersetzt von Trude Fein. (manesse im dtv). München: Deutscher Taschenbuch Verlag / Zürich: Manesse Verlag. “Nachwort”, pp. 253-262. [a shortened reprint of 5.] 43. “Wie buchstabiert man SPELL? ” unizürich, Journal der Universität Zürich 1/ 95, p. 19. 1996 44. “Dieth, Eugen”. In Harro Stammerjohann (ed.). Lexicon Grammaticorum: Who’s Who in the History of World Linguistics. Tübingen: Max Niemeyer, p. 240. 45. “Dream Theory and Dream Lexis in the Middle Ages”. In Jürgen Klein and Dirk Vanderbeke (eds.). Anglistentag 1995 Greifswald: Proceedings. Tübingen: Max Niemeyer, pp. 245-257. 46. “The Vocabulary of Very Late Old English”. In M. J. Toswell and E. M. Tyler (eds.). Studies in English Language and Literature. ‘Doubt Wisely’: Papers in Honour of E. G. Stanley. London and New York: Routledge, pp. 29-41. 1997 47. (ed. with Martin Heusser and Thomas Hermann). Aspects of Modernism: Studies in Honour of Max Nänny. With two drawings by Wilhelm Föckersperger. Tübingen: Gunter Narr. 48. “The Hatton MS. of the West Saxon Gospels: The Preservation and Transmission of Old English”. In Paul E. Szarmach and Joel Rosenthal (eds.). The Preservation and Transmission of Anglo-Saxon Culture: Selected Papers from the 1991 Meeting of the International Society of Anglo-Saxonists. (Studies in Medieval Culture 40). Kalamazoo: Western Michigan University, Medieval Institute Publications, pp. 353-367. 49. “Hemingway in the Oxford English Dictionary”. In Andreas Fischer, Martin Heusser and Thomas Hermann (eds.). Aspects of Modernism: Studies in Honour of Max Nänny. Tübingen: Gunter Narr, pp. 151- 167. 50. “Lexical and Semantic Change: An Evaluation of Structural, Pragmatic and Cognitive Explanations”. In Uwe Böker and Hans Sauer (eds.). Anglistentag 1996 Dresden: Proceedings. Trier: WVT Wissenschaftlicher Verlag Trier, pp. 61-69. 51. “Max Nänny, Zürich, zur Emeritierung und zum 65. Geburtstag”. Anglistik 8/ 2, pp. 243-245. <?page no="14"?> Sarah Chevalier & Thomas Honegger xiv 52. “The Oxford English Dictionary on CD-ROM as a Historical Corpus: To Wed and to Marry Revisited”. In Udo Fries, Viviane Müller and Peter Schneider (eds.). From Ælfric to The New York Times: Studies in English Corpus Linguistics. (Language and Computers: Studies in Practical Linguistics 19). Amsterdam, Atlanta, GA: Rodopi, pp. 161-172. 53. “‘With This Ring I Thee Wed’: The Verbs to Wed and to Marry in the History of English”. In Raymond Hickey and Stanislaw Puppel (eds.). Language History and Linguistic Modelling: A Festschrift for Jacek Fisiak on his 60th Birthday. (Trends in Linguistics, Studies and Monographs 101). Berlin and New York: Mouton de Gruyter, pp. 467-481. 54. Review of Nadine Van den Eynden. Syntactic Variation and Unconscious Linguistic Change. A Study of Adjectival Relative Clauses in the Dialect of Dorset (Bamberger Beiträge zur Englischen Sprachwissenschaft 33; Frankfurt am Main etc.: Peter Lang, 1993). English World-Wide 18/ 1, pp. 160-162. 1998 55. “Marry: From Religious Invocation to Discourse Marker”. In Raimund Borgmaier, Herbert Grabes and Andreas H. Jucker (eds.). Anglistentag 1997 Giessen: Proceedings. Trier: WVT Wissenschaftlicher Verlag Trier, pp. 35-46. 56. “‘Milly Bloom, fairhaired, greenvested, slimsandalled’: Joyce’s Compound Adjectives and the Oxford English Dictionary”. In Ruth Frehner and Ursula Zeller (eds.). A Collideorscape of Joyce: Festschrift for Fritz Senn. Dublin: The Lilliput Press, pp. 171-179. 57. “Rejoinder: The Role of Language Studies in Editing Shakespeare”. The European English Messenger 7/ 2, pp. 72-75. 1999 58. “Graphological Iconicity in Print Advertising: A Typology”. In Max Nänny and Olga Fischer (eds.). Form Miming Meaning: Iconicity in Language and Literature. Amsterdam and Philadelphia: John Benjamins, pp. 251-283. 59. “Strange Words, Strange Music: The Verbal Music of ‘Sirens’”. In Sebastian D. G. Knowles (ed.). Bronze by Gold: The Music of Joyce. (Border Crossings 3/ Garland Reference Library of the Humanities 2061). New York and London: Garland Publishing, pp. 245-262. [a reprint of 33.] <?page no="15"?> List of Publications by Andreas Fischer xv 60. “What, if Anything, is Phonological Iconicity? ” In Max Nänny and Olga Fischer (eds.). Form Miming Meaning: Iconicity in Language and Literature. Amsterdam and Philadelphia: John Benjamins, pp. 123-134. 2000 61. “The Greening of Greening”. In Christiane Dalton-Puffer and Nikolaus Ritt (eds.). Words: Structure, Meaning, Function. A Festschrift for Dieter Kastovsky. (Trends in Linguistics, Studies and Monographs 130). Berlin and New York: Mouton de Gruyter, pp. 75-86. 62. “Lexical Gaps, Cognition and Linguistic Change”. In Julie Coleman and Christian J. Kay (eds). Lexicology, Semantics and Lexicography: Selected Papers from the Fourth G. L. Brook Symposium, Manchester, August 1998. (Amsterdam Studies in the Theory and History of Linguistic Science, Series IV: Current Issues in Linguistic Theory 194). Amsterdam and Philadelphia: John Benjamins, pp. 1-18. 63. Review of Hugh Magennis. Anglo-Saxon Appetites: Food and Drink and their Consumption in Old English and Related Literature (Dublin: Four Courts Press, 1999). Anglia 118, pp. 595-597. 2001 64. “Einleitung”. In Thomas Honegger (ed.). Authors, Heroes and Lovers: Essays on Medieval English Literature and Language. Selected Papers from the Studientage zum englischen Mittelalter SEM I & II (Potsdam 1999 & 2000). / Liebhaber, Helden und Autoren: Studien zur alt- und mittelenglischen Literatur und Sprache. Ausgewählte Beiträge der Studientage zum englischen Mittelalter SEM I & II (Potsdam 1999 & 2000). (Sammlung/ Collection Variations 2). Berne: Peter Lang, pp. xv-xx. 65. “Language and Politics in Switzerland”. In John M. Kirk and Dónall P. Ó Baoill (eds.). Linguistic Politics: Language Policies for Northern Ireland, the Republic of Ireland, and Scotland. (Belfast Studies in Language, Culture and Politics 3). Belfast: Cló Ollscoil na Banríona, pp. 105-122. 66. “Lexical Borrowing and the History of English: A Typology of Typologies”. In Dieter Kastovsky and Arthur Mettinger (eds.). Language Contact in the History of English. (Studies in English Medieval Language and Literature 1). Frankfurt am Main: Peter Lang, pp. 97-115. <?page no="16"?> Sarah Chevalier & Thomas Honegger xvi 67. Review of Lynne Magnusson. Shakespeare and Social Dialogue: Dramatic Language and Elizabethan Letters (Cambridge: Cambridge University Press, 1999). Journal of Historical Pragmatics 2/ 2, pp. 317-321. 68. Review of Andrea Simmelbauer. The Dialect of Northumberland: A Lexical Investigation. (Anglistische Forschungen 275; Heidelberg: C. Winter, 2000). English World-Wide 22/ 1, pp. 145-147. 2002 69. (ed. with Gunnel Tottie and Hans Martin Lehmann with the assistance of Therese Lutz and Peter Schneider). Text Types and Corpora: Studies in Honour of Udo Fries. Tübingen: Gunter Narr. 70. (with Peter Schneider). “The Dramatick Disappearance of the <-ick> Spelling, Researched with Authentick Material from the Zurich English Newspaper Corpus”. In Andreas Fischer, Gunnel Tottie and Hans Martin Lehmann 2002 (eds.), pp. 139-150. 71. “Global English as a Challenge for Lexicography”. In Frances Ilmberger and Alan Robinson (eds.). Globalisation. (SPELL, Swiss Papers in English Language and Literature 15). Tübingen: Gunter Narr, pp. 127-150. 72. “Notes on Kinship Terminology in the History of English”. In Katja Lenz and Ruth Möhlig (eds.). Of Dyuersitie & Chaunge of Langage: Essays Presented to Manfred Görlach on the Occasion of his 65th Birthday. (Anglistische Forschungen 308). Heidelberg: C. Winter, pp. 115-128. 73. “Ernst Leisi (29 June 1918 - 30 December 2001)”. English Studies at Swiss Universities 2001/ 2002, pp. 6-7. 74. “Prof. Dr. Ernst Leisi”. Nekrologe 2001: Zum Gedenken an unsere verstorbenen Professorinnen und Professoren. Zürich: Universität Zürich, pp. 21-22. 2003 75. “Ernst Leisi †”. Anglia 121/ 1, pp. 1-3. 2004 76. “‘Non Olet’: Euphemisms We Live By”. In Christian Kay, Carole Hough and Irené Wotherspoon (eds.). New Perspectives on English Historical Linguistics: Selected Papers from 12 ICEHL, Glasgow, 21-26 August 2002, Glasgow. Vol. 2: Lexis and Transmission. (Amsterdam Studies in the Theory and History of Linguistic Science, Series IV: Current Issues in Linguistic Theory 252). Amsterdam and Philadelphia: John Benjamins, pp. 91-107. <?page no="17"?> List of Publications by Andreas Fischer xvii 77. “The Notional Structure of Thesauruses”. In Christian Kay and Jeremy Smith (eds.). Categorization in the History of English. (Amsterdam Studies in the Theory and History of Linguistic Science, Series IV: Current Issues in Linguistic Theory 261). Amsterdam and Philadelphia: John Benjamins, pp. 41-58. 78. “Sceaf”. Reallexikon der Germanischen Altertumskunde. Band 26. Berlin and New York: Walter de Gruyter, pp. 557-558. 79. “[Vorwort: ] Zauberhafte Erfolgsgeschichte”. In Thomas Hermann, Daniel Ammann and Heinz Moser. Harry war hier: Lesen, Magie und Projekte im Klassenzimmer. Materialien zu Harry Potter und der Stein der Weisen. Mit einem Vorwort von Andreas Fischer und einem englischen Übungsteil von Michael Prusse. Winterthur: ZKM, Verlag der Zürcher Kantonalen Mittelstufenkonferenz/ Zürich: Verlag Pestalozzianum an der Pädagogischen Hochschule Zürich, pp. 6-7. 2005 80. “Die Joyce Stiftung und die Universität Zürich”. In Ursula Zeller and Ruth Frehner (eds.). 20 Jahre Zürcher James Joyce Stiftung. Zürich: Zürcher James Joyce Stiftung, pp. 24-25. 2006 81. “Of Fæderan and Eamas: Avuncularity in Old English”. In Graham D. Caie, Carole Hough and Irené Wotherspoon (eds.). The Power of Words: Essays in Lexicography, Lexicology and Semantics. In Honour of Christian J. Kay. (Costerus New Series 163). Amsterdam and New York, NY: Rodopi, pp. 67-77. 82. Review of N. F. Blake. Shakespeare’s Non-Standard English: A Dictionary of His Informal Language. (Athlone Shakespeare Dictionary Series; London and New York: Thoemmes Continuum, 2004). Archiv für das Studium der neueren Sprachen und Literaturen 243, pp. 158-159. 2007 83. “Prof. Dr. Max Nänny”. Nekrologe 2006: Zum Gedenken an unsere verstorbenen Professorinnen und Professoren, Privatdozentinnen und Privatdozenten. Zürich: Universität Zürich, pp. 21-22. 2009 84. Weltsprache(n). Rede des Rektors gehalten am Dies academicus 2009. Zürcher Universitätsschriften 11. <?page no="18"?> Sarah Chevalier & Thomas Honegger xviii 2010 85. Baum des Lebens, Baum der Sprache. Rede des Rektors gehalten am Dies academicus 2010. Zürcher Universitätsschriften 12. [see also 86.] 2011 86. “Baum des Lebens, Baum der Sprache”. In Heinz-Ulrich Reyer and Paul Schmid-Hempel (eds.). Darwins langer Arm - Evolutionstheorie heute. (Reihe Zürcher Hochschulforum 47). Zürich: vdf Hochschulverlag an der ETH Zürich, pp. 169-188. [a slightly shortened reprint of 85.] <?page no="19"?> Dieter Bitterli Two Old English Prose Riddles of the Eleventh Century Abstract The two short Old English riddles preserved in a mid-eleventh-century manuscript from Winchester are the only surviving examples of their kind written in prose. Yet despite their uniqueness, the two prose riddles show close similarities not only to the celebrated verse riddles of the Exeter Book, but also to the earlier Latin enigmata that circulated in Anglo-Saxon England. Their exact subject matter is unknown (there are no solutions in the manuscript), but they both deal with paradoxical family constellations that not only are similar to those of the traditional relationship riddles about biblical characters but also occur in mathematical problems known from the period. This paper explores some of these intriguing intertextualities as it attempts to situate the two Winchester prose riddles in the context of the rich Anglo-Saxon riddle tradition. Its primary concern is not to propose a new or “correct” solution to either of the two riddles; rather, it seeks to identify the literary framework in which they are rooted (and perhaps can be solved), and to provide further evidence for the manyvoiced dialogue between Anglo-Latin and Old English enigmatography. Riddles, puzzles, and conundrums are a perennially popular genre, yet in the literate circles of Anglo-Saxon England, the literary riddle - in both verse and prose, and in Latin as well as in the vernacular - appears to have been held in particularly high esteem, constituting a major literary genre from as early as the seventh century. It was Aldhelm of Malmesbury (d. 709/ 10) who initiated the tradition of riddle-making in early England. His prototypical collection of one hundred Latin enigmata in hexametrical verse was studied in the Anglo-Saxon classroom and often copied in manuscripts together with the verse riddles of the late Roman poet Symphosius (4th/ 5th c. AD) and those of the Anglo-Saxon churchmen Tatwine (d. 734), Eusebius (8th c.), and Boniface (d. 754). Their elegant collections of poetic riddles, in turn, were known to Alcuin of York (d. 804), whose literary oeuvre includes both occasional verse enigmata, a riddle dialogue in prose, and a series of riddle-like mathemati- <?page no="20"?> Dieter Bitterli 2 cal problems. 1 The distinctively English tradition of early medieval enigmatography is further evidenced by two Anglo-Saxon texts once attributed to Bede: the Collectanea Pseudo-Bedae, a compilation of monastic wisdom questions and riddles in prose and verse, and the so-called Joco-seria, a group of grammatical puzzles that survive in the famous “Cambridge Songs” manuscript (Cambridge, University Library, Gg. 5.35) alongside the complete riddle cycles of Symphosius, Aldhelm, Tatwine, and Eusebius. Symphosius’ pithy tristichs, in particular, also circulated outside these collections: ten of them were incorporated into the Late Latin Apollonius romance (which was translated into Old English in the tenth century), and Aldhelm quotes several of them in his metrical treatises, fashioning his own century of enigmata after that of his Roman model. Given such a rich and long-standing practice of the riddle genre in Anglo-Saxon England, it comes as no surprise that riddles in vernacular Old English were composed not only in alliterative verse - as exemplified by the ninety-odd Riddles of the 10th-century Exeter Book - but also in prose. Only two genuine Old English prose riddles, however, are known to have come down to us. They uniquely survive in a mid-eleventh-century psalter manuscript from Winchester (British Library, Cotton MS Vitellius E.xviii) as part of a short entry about secret writing. The two riddles, which were printed for the first time in Wanley’s catalogue (1705), are virtually unknown even to the specialists in the field, although they were discussed in a series of notes by Max Förster (1905, 1906a, 1906b, 1916) and have occasionally been referred to in studies of the Exeter Book riddles. One aspect that connects the two Winchester prose riddles with both the Latin enigmata and the earlier Exeter Book collection is the combination of riddling and cryptography. The riddles occur in the section of the manuscript (fols. 2r-17v) which precedes the glossed psalter and also contains computistical tables and notes, charms, prognostications, and prayers. 2 As they stand in the manuscript, the riddles form part of a one-page treatise on secret codes that occupies about two thirds of fol. 16v and elucidates a writing technique known in the Middle Ages as notae sancti Bonifatii. 3 In this cryptographic method, the vowels a e i o u are either substituted by the consonants which follow them in the roman alphabet 1 The Disputatio Pippini cum Albino and the Propositiones ad acuendos iuvenes; see Bitterli (2010). 2 See Ker (1957: 298-301, no. 224); Pulsiano & Treharne (1998), with a facsimile of fol. 16v (plate 12); Gneuss (2001: 73, no. 407). 3 The best edition is Howlett (2002: 109-12). A brief section of the treatise, without the riddles, occurs also in the earlier manuscript Oxford, Bodleian Library, MS Bodl. 572, fol. 40r; see Ker (1957: 376-77, no. 313); Howlett (2002: 107). For the notae sancti Bonifatii, see Levison (1946: 290-94); Bischoff (1992: 177 and 255); Howlett (2002: 105-12). <?page no="21"?> Two Old English Prose Riddles of the Eleventh Century 3 (b f k p x), or expressed by dots (one dot for a, two for e, three for i, etc.). Copyists of medieval codices, especially, would occasionally use this device to encode their names in signatures and subscriptions. In the Winchester text, both codes are explained and illustrated with some examples in Latin and Old English, including an encrypted scribal colophon that associates the short treatise with Ælfwine of Winchester, who was abbot of the New Minster from 1031 to 1057. 4 Strictly speaking, the two riddles precede the cryptographic treatise, but they are introduced by a formula which employs the system of vowel substitution explained later in the text: Nys þks frfgfn syllkc þknc to rædfnnf. [Nys þis fregen syllic þinc to rædenne.] Þu þe færst on þone weg gret ðu minne broðor, minre modor ceor[l], þone acende min agen wif. And ic wæs mines broðor dohtor and ic eom mines fæder modor geworden. And mine bearn syndon geworden mines fæder modor. a e i o u. a b f e k i p o x u - (This question is not a strange thing to guess: You who journey on the way, greet my brother, my mother’s husband, whom my own wife gave birth to. And I was my brother’s daughter, and I became my father’s mother. And my children became my father’s mother. a e i o u. a b f e k i p o x u -) 5 Then, on a new line, the explanations of the vowel substitution code follow: “Đis is quinque vocales. mid þysum fif stafum man mæg writan swa hwæt swa he wile” (This is [the system of] five vowels; with these five letters one can write whatever one wishes), and so on. It is interesting to note that riddles and secret writing also occur together elsewhere in the early medieval riddle corpus. In some manuscripts of the Latin enigmata, for instance, the Bonifatian method of vowel substitution was used to encrypt riddle solutions or add scribal comments, and in the Exeter Book Riddle 36, there is an interpolated line with encrypted vowels in Old English and Latin providing a hint to the solution. 6 However, no such connection between the riddles and the explanations about secret writing which follow them on the page seems to exist in the case of the two Winchester riddles. It would be tempting to read the riddles as an oblique instruction of how to use the secret writing code shown in the opening formula and expounded in the treatise. If this was the case, the 4 The conclusion reads, transliterated: “Ælfwine me wrat ræd nu ðu ðe cunne” (Ælfwine wrote me, now you guess who might be able). This is not Tolkien’s Ælfwine (cf. Honegger 2006), but the Winchester monk and owner of a tiny Prayerbook (British Library, Cotton Titus D.xxvi + xxvii) of about 1023/ 31, which also contains an encrypted scribal note using vowel substitution; see Pulsiano & Treharne (1998: 99); Lapidge (1999). 5 Howlett (2002: 109), except w for Anglo-Saxon “wynn”, and and for “7”; my punctuation and translation. 6 See Orchard (2005: 285); Bitterli (2009: 71-75) and Bitterli (2010: 19). <?page no="22"?> Dieter Bitterli 4 siblings, parents, and children mentioned in the text might stand for the vowels and consonants (and the dots) following and replacing one another in the code - comparable to Aldhelm’s riddle about the alphabet (Enigma 30), in which the letters appear as ‘voiceless’ yet speaking sisters 7 - yet here, the family relations are too complex and paradoxical to support such a reading. Nor do the cryptographic explanations encode the solutions to the riddles or help in any way to solve them. Rather, we should take the two riddles as constituting a separate sub-entry, independent, as it were, of the cryptic code used and explained on the same manuscript page, yet as part of a scribal note whose purpose seems to have been to collect a few brainteasers from medieval writing practice and monastic sport. 8 Obviously, we are dealing with two separate riddles: in the first of them, beginning with “Þu þe færst on þone weg”, there seems to be a male speaker and the subject to be guessed must be a man (or a subject metaphorically disguised as a man), while the phrase “ic wæs mines broðor dohtor” changes the perspective to introduce a female speaker who is also the subject of the second riddle. This was already noticed by Dietrich (1859: 489-90), who suggested that the solution to the first riddle is “day”, while the subject of the second (and perhaps even of the first) is the biblical Eve, a solution supported by Grein (1865: 309), Förster (1906a), and Tupper (1910: 175-76). 9 “Eve” indeed appears to be the solution to the second item: Eve can be said to be her “brother’s daughter” (“mines broðor dohtor”) because she was born from Adam’s rib (Gen 2.21); she became her “father’s mother” (“mines fæder modor”) when she died and turned to earth, from which Adam, her “father”, was made (Gen 2.7); and her “children became [her] father’s mother” (“mine bearn syndon geworden mines fæder modor”) because she is the typological figure of Mary, the Mother of God, whom she preceded in the biblical genealogy. As such, the riddle is indeed “full of popular motives common in riddle and dialogue literature”, as Tupper (1910: 176) has noted. Witty puzzles and wisdom questions about biblical characters, especially Adam and Eve, Cain and Abel, and the Old Testament patriarchs, circulated widely in the early medieval Ioca monachorum and related monastic dialogues, both in Latin and the vernacular languages. The paradoxical nature of Adam and Eve was among their frequent subjects, as in this 8th-century Latin version of the Ioca monachorum from St. Gallen: 7 See Bitterli (2009: 114-15), with further examples. 8 The discreteness of the entries is also suggested by the use of paragraphs and capital letters in each of the first words, Nys, Þu and Đis. 9 Dietrich must have thought of Aldhelm’s Enigma 48 “De die et nocte” (Glorie 1968: 258), in which the day and the night are referred to as unlike “sisters”. “Day and Night” also seems to be the solution to Collectanea Pseudo-Bedae no. 122 (Bayless and Lapidge 1998: 136). <?page no="23"?> Two Old English Prose Riddles of the Eleventh Century 5 Qui est mortuos et non est natus? - Adam. Qui dedit et non accepit? - Eva lacte. (Who died and wasn’t born? - Adam. Who gave and didn’t receive? - Eve the milk.) 10 In the anonymous Collectanea of Pseudo-Bede, a compilation of wisdom questions and riddles from mostly Hiberno-Latin and Anglo-Saxon sources of the eighth century, there is a similar item about Adam: Dic mihi quis homo, qui non natus est, et mortuus est, atque in utero matris suae post mortem baptizatus? Est Adam. (Tell me, what man was not born, and died, and was baptized in the womb of his mother after his death? It is Adam.) 11 According to medieval tradition, Adam was buried at Golgotha and baptized after his death by Christ’s blood flowing from the Cross which had been erected over his tomb (“the womb of his mother”, i.e. the earth). The legend is also referred to in the late Old English prose dialogue of Adrian and Ritheus: Saga me, hwilc man wære dead and nære acenned and æfter þam deaðe wære eft bebyried in his moder innoðe. Ic þe secge, þæt wæs Adam se æresta man, for þam eorðe wæs his moder and he wæs bibiriged eft in þære eorðan. (Tell me which man died and was not born, and after death was later buried in his mother’s womb. I tell you, that was Adam the first man, because the earth was his mother and he was buried afterwards in the earth.) 12 Adam may indeed also be the subject of the first of the two Winchester prose riddles. If we assume that here, too, Eve (and not a man) is the speaker, the elements could be explained as follows: “gret ðu minne broðor”: Adam is Eve’s “brother”, because they both were created by God; “minre modor ceor[l]”: Adam is also the “husband” of Eve’s “mother”, the earth; “þone acende min agen wif”: the earth, of which Adam was made is both the mother and the metaphorical “wife” of the first parents who both were brought forth by it and subdued it (Gen 1.28). 13 The solution “Adam” is perhaps not as satisfactory as the one suggested for the second riddle 10 Suchier (1955: 108); my translation. 11 Text and translation from Bayless and Lapidge (1998: 136-137, no. 123), whose note (p. 228) lists numerous medieval analogues. 12 Adrian and Ritheus 28, ed. and trans. Cross and Hill (1982: 38 and 76); see their commentary pp. 75-79. 13 Alternatively, “ceorl” might take the sense of “peasant” so that the “mother’s peasant” would denote Adam as the first husbandman. This reading was favoured by Grein (1865: 309). <?page no="24"?> Dieter Bitterli 6 (“Eve”) - it also leaves unexplained why the speaker addresses traveling on their way - but the point here is the seemingly contradictory combination of both brother, husband, and son, which the reader must unravel in order to arrive at a solution. Similar paradoxical family constellations are described in a number of so-called relationship or kinship riddles found elsewhere in the Anglo-Saxon riddle corpus. In the Exeter Book Riddle no. 46, for instance, the incestuous consanguinity of Lot and his offspring (another familial group of Old Testament figures) is turned into a grotesque orgy of wives, sons, daughters, sisters, sons, uncles, and nephews, who all join their father in his wine: Wer sæt æt wine mid his wifum twam ond his twegen suno ond his twa dohtor, swase gesweostor, ond hyra suno twegen, freolico frumbearn; fæder wæs þær inne þara æþelinga, æghwæðres mid eam ond nefa. Ealra wæron fife eorla ond idesa insittendra. (A man sat at wine with his two wives and his two sons and his two daughters, dear sisters, and their two sons, noble first-born children. In there was the father of these noble youths, (who were) at the same time each other’s uncle and nephew. In all there were five men and women sitting within.) 14 Here, the solution is obscured by the fact that Lot’s children stood in more than one relationship with their father: the two daughters were also Lot’s “wives”, and their sons were at the same time each other’s uncle and nephew. Such perplexing puzzles about marital relations already occur in the Carolingian Propositiones ad acuendos iuvenes, a collection of mathematical and combinatorial “Problems to Sharpen the Young” now generally attributed to Alcuin of York (d. 804), as well as in the anonymous Enigmata risibilia or “Amusing Riddles” preserved in a late 10th-century manuscript from the Abbey of Reichenau. They all deal with strange family constellations resulting from remarriage and intermarriage. Alcuin’s “Problem about two men marrying each other’s mother” (Propositio 11a) is a typical example. It is one of three related combinatorial puzzles that follow each other in some of the many surviving manuscripts of the Propositiones: Si duo homines alter alterius matrem similiter in coniugium sumpserit, quali cognatione filii eorum sibi coniungantur. (If two men each take each other’s mother in marriage, what would be the relationship between their sons? ) 15 14 Krapp and Dobbie (1936: 205); my punctuation and translation. 15 Alcuin, Propositio 11a, ed. Folkerts (1978: 50), trans. Hadley and Singmaster (1992: 109). <?page no="25"?> Two Old English Prose Riddles of the Eleventh Century 7 The answer is that the two sons - like Lot’s sons - are both each other’s uncle and nephew; or as a ninth-century manuscript of the Propositiones explains: “Filius igitur meus et filius matris mee avunculi et nepotes sunt” (Hence my son and my mother’s son are uncles and nephews). 16 In the Propositiones, the same unusual setting is duplicated by the analogous problem about a father and a son marrying a widow and her daughter (Propositio 11b), and there is a similar problem about two men who marry each other’s sister (Propositio 11). 17 Obviously, matters are complicated when the people involved are widowed and remarry, so that not only natural parents and children, but widows, stepfathers and stepsons come into play, as in this particularly knotty item from the tenth-century Reichenau Riddles, the last of a group of only six short riddles in Latin: Porto filium filii mei, mariti mei fratrem, alterum unicum filium meum. (I carry the son of my son, my husband’s brother, my second only son.) 18 There is no solution to this riddle in the Reichenau manuscript, but as Schmidt (1936: 199) has demonstrated, its speaker appears to be a woman who is pregnant by her own stepson and second husband - a pattern not entirely unusual in the Middle Ages: the woman must have been married before to a man who had a son from an earlier marriage; after her husband’s death she married her stepson whose baby she is now carrying. Her baby is, therefore, the son of her stepson (“filium filii mei”) and at the same time her husband’s stepbrother (“mariti mei fratrem”) since the baby will be born from one and the same mother and be her second yet only natural son (“alterum unicum filium”). This solution, as far-fetched as it may seem, is somewhat corroborated by the even shorter relationship riddle which precedes it in the manuscript. Like the remaining items of the group, this one is accompanied by its solution, which was added by the scribe in secret writing, employing the very code of the notae sancti Bonifatii explained in the Winchester treatise. It reads: 16 Folkerts (1978: 50). 17 Folkerts (1978: 50-51); see Folkerts (1993: 308-310); Bitterli (2009: 58). 18 Reichenau Ridldle 6, ed. Müllenhoff and Scherer (1873: 14); repr. in Schupp (1972: 31); my translation. <?page no="26"?> Dieter Bitterli 8 Equitavit homo cum femina: mater eius matris meae socrus fuit. xktrkcxs. [vitricus.] (A man rode on a horse with his wife: his mother was my mother’s mother-inlaw. Stepfather.) 19 The solution, “vitricus” (stepfather), is given in the encoded line, but the point of view is that of the stepdaughter: the couple she describes as riding on a horse are her mother (“femina”) and her stepfather (“homo”) who is also her uncle. Her natural father, who is dead, and her uncle had the same mother (“mater”) who was also her mother’s mother-in-law (“matris meae socrus”). 20 With its use of an observational perspective and direct speech this riddle employs the rhetorical strategy to the so-called ‘‘I saw” type of riddles, in which there is a third-person description rather than a personified subject as in the more frequent “I am” riddles. Both patterns occur in the Old English riddles of the Exeter Book, and they also seem to govern the two Winchester prose riddles as there is an external point of view in the first (“gret ðu minne broðor …”) and a speaking subject in the second of them (“ic wæs mines broðor dohtor …”). The use of such formulas is, as we have seen, not the only link between the two Old English prose riddles and their better-known analogues in the early medieval, and especially Anglo-Saxon, riddle corpus. In the two Winchester puzzles, the words “modor” (used tree times), “fæder” (twice), “broðor” (twice), “ceorl”, “wif”, “dohtor”, and “bearn” are repeated and combined to bewildering effect, creating an intricate knot of seemingly paradoxical family relations and ultimately defying a single solution. In terms of their diction and (obscure) contents, the two riddles echo both the traditional conundrums about Old Testament figures known from monastic literature and the secular relationship puzzles preserved in early medieval collections of mathematical and combinatorial problems. Like the popular Adam and Eve questions and the jocose riddles about unlikely marriages, the two Old English prose riddles from Winchester play with our notions of what is a legitimate marriage and what would be considered incest. As such, they are also reminiscent of the famous riddle posed by the incestuous Antiochus in the romance of Apollonius of Tyre. Its Old English version reads: 19 Reichenau Riddle 5, ed. Müllenhoff and Scherer (1873: 13); my translation. 20 This solution differs from the one proposed by Schmidt (1936: 199) and adopted in Schupp (1972: 315). <?page no="27"?> Two Old English Prose Riddles of the Eleventh Century 9 “Ic sece minne fæder, mynre modor wer, mines wifes dohtor and ic ne finde.” (I seek my father, my mother’s husband, my wife’s daughter, but I do not find them.) 21 The first two clues in this (apparently corrupt) version seem to refer to Antiochus’s daughter, who is speaking and unaware of the fact that her own father has taken her in marriage, while the latter might be the speaker of the phrase “mines wifes dohtor”, who is untraceable because his daughter and wife are one and the same person. In the two Winchester riddles it is equally unclear who is speaking, as the perspective seems to shift from a male to a female speaker, suggesting that there must be a double answer. “Adam and Eve” is perhaps this double answer, but what is more important is the striking fact that the two pieces reverberate with motives and elements which can be found in a number of both Latin and vernacular riddles and related texts from the period. That they, moreover, occur in the context of secret writing, provides yet another intriguing connection to the genre and the scribal practices associated with it. All this places these two elusive pieces squarely in the Anglo-Saxon riddle tradition, whose many voices can be heard to echo in them. Biographical sketch Dieter Bitterli teaches Old and Middle English language and literature in the English Department of the University of Zurich. He has written about Counter-Reformation sermons, Early Modern emblematics and Old English poetry. His recent book, Say What I Am Called: The Old English Riddles of the Exeter Book and the Anglo-Saxon Riddle Tradition (2009) was originally inspired by a seminar course taught by Andy Fischer, who also midwived his Habilitation at the University of Zurich. 21 OE Apollonius, 4, ed. Goolden (1958: 6); my translation. For a possible solution, see Goolden (1955: 247), who discusses the Old English version in connection with those in Gower’s Confessio amantis (8.423-27) and Shakespeare’s Pericles (I.i.64-71). <?page no="28"?> Dieter Bitterli 10 References Bayless, Martha, and Michael Lapidge (eds. and trans.). 1998. Collectanea Pseudo- Bedae. Dublin: School of Celtic Studies, Dublin Institute of Advanced Studies. Bischoff, Bernhard. 1992. Latin Palaeography: Antiquity and the Middle Ages, trans. D. ó Cróinín and D. Ganz. Cambridge: Cambridge University Press. Bitterli, Dieter. 2009. Say What I Am Called: The Old English Riddles of the Exeter Book and the Anglo-Latin Riddle Tradition. Toronto: University of Toronto Press. Bitterli, Dieter. 2010. “Alkuin von York und die angelsächsische Rätseldichtung”. Anglia 128, pp. 4-20. Cross, James E., and Thomas D. Hill (eds.). 1982. The ‘Prose Salomon and Saturn’ and ‘Adrian and Ritheus’. McMaster Old English Studies and Texts 1. Toronto, Buffalo and London: University of Toronto Press. Dietrich, Franz. 1859. “Die Räthsel des Exeterbuchs: Würdigung, Lösung und Herstellung”. Zeitschrift für deutsches Altertum 11, pp. 448-490. Folkerts, Menso. 1978. Die älteste mathematische Aufgabensammlung in lateinischer Sprache: Die Alkuin zugeschriebenen ‘Propositiones ad acuendos iuvenes’: Überlieferung, Inhalt, Kritische Edition. Österreichische Akademie der Wissenschaften, Mathematisch-naturwissenschaftliche Klasse, Denkschriften 116.6. Wien: Springer. Folkerts, Menso, and Helmuth Gericke. 1993. “Die Alkuin zugeschriebenen Propositiones ad acuendos iuvenes”. In Paul Leo Butzer and Dietrich Lohrmann (eds.). Science in Western and Eastern Civilization in Carolingian Times. Basel: Birkhäuser, pp. 283-362. Förster, Max. 1905. “Ein altenglisches Prosa-Rätsel”. Archiv für das Studium der neueren Sprachen und Literaturen 115, pp. 392-393. Förster, Max. 1906a. “Die Lösung des ae. Prosarätsels”. Archiv für das Studium der neueren Sprachen und Literaturen 116, pp. 367-371. Förster, Max. 1906b. “Ae. fregen ‘die Frage’”. Englische Studien 36, pp. 325-328. Förster, Max. 1916. “Nochmals ae. fregen ‘Frage’”. Archiv für das Studium der neueren Sprachen und Literaturen 135, pp. 399-401. Glorie, F. (ed.). 1968. Collectiones Aenigmatum Merovingicae Aetatis. 2 vols., Corpus Christianorum Series Larina 133 and 133A. Turnhout: Brepols. Gneuss, Helmut. 2001. Handlist of Anglo-Saxon Manuscripts: a List of Manuscripts and Manuscript Fragments Written or Owned in England up to 1100. Medieval and Renaissance Texts and Studies 241. Tempe: Arizona Center for Medieval and Renaissance Studies. Goolden, Peter. 1955. “Antiochus’s Riddle in Gower and Shakespeare”. Review of English Studies n.s. 6, pp. 245-251. Goolden, Peter (ed.). 1958. The Old English ‘Apollonius of Tyre’. London: Oxford University Press. Grein, Christian W.M. 1865. “Kleine Mitteilungen”. Germania. Vierteljahrsschrift für deutsche Alterthumskunde 10, pp. 305-310. <?page no="29"?> Two Old English Prose Riddles of the Eleventh Century 11 Hadley, John, and David Singmaster. 1992. “Problems to Sharpen the Young: an Annotated Translation of Propositiones ad acuendos iuvenes, the Oldest Mathematical Problem Collection in Latin, Attributed to Alcuin of York”. The Mathematical Gazette 76, pp. 102-126. Honegger, Thomas. 2006. “Aelfwine”. In Michael D.C. Drout (ed.). J.R.R. Tolkien Encylopedia: Scholarship and Critical Assessment. New York: Routledge, pp. 4-5. Howlett, David. 2002. “‘Tres Linguae Sacrae’ and Threefold Play in Insular Latin”. Peritia 16, pp. 94-115. Ker, Nigel R. 1957. Catalogue of Manuscripts Containing Anglo-Saxon. Cambridge: Clarendon. Krapp, George Philip, and Elliott Van Kirk Dobbie (eds.). 1936. The Exeter Book. Anglo-Saxon Poetic Records 3. New York: Columbia University Press. Lapidge, Michael. 1999. “Ælfwine, Prayerbook of”. In Michael Lapidge, John Blair et al. (eds.). The Blackwell Encyclopaedia of Anglo-Saxon England. Oxford: Blackwell, pp. 9-10. Levison, Wilhelm. 1946. England and the Continent in the Eighth Century. Oxford: Clarendon. Müllenhoff, Karl, and Wilhelm Scherer (eds.). 1873. Denkmäler deutscher Poesie und Prosa aus dem VIII-XII Jahrhundert. 2nd ed. Berlin: Weidmann. Orchard, Andy. 2005. “Enigma Variations: The Anglo-Saxon Riddle Tradition”. In Katherine O’Brien O’Keeffe and Andy Orchard (eds.). Latin Learning and English Lore: Studies in Anglo-Saxon Literature for Michael Lapidge. Vol. 1. Toronto: University of Toronto Press, pp. 284-304. Pulsiano, Phillip, and Elaine M. Treharne (eds.). 1998. Anglo-Saxon Manuscripts and their Heritage. Aldershot: Ashgate. Schmidt, Arno. 1936. “Die Reichenauer Rätsel”. Zeitschrift für deutsches Altertum und Literatur 73, pp. 197-200. Schupp, Volker (ed.). 1972. Deutsches Rätselbuch. Stuttgart: Reclam. Singmaster, David. 1998. “The History of Some of Alcuin’s Propositions”. In Paul Leo Butzer et al. (eds.). Charlemagne and his Heritage: 1200 Years of Civilization and Science in Europe: Karl der Grosse und sein Nachwirken: 1200 Jahre Kultur und Wissenschaft in Europa. Vol. 2: Mathematical Arts. Turnhout: Brepols, pp. 11-29. Suchier, Walther (ed.). 1955. Das mittellateinische Gespräch Adrian und Epictitus nebst verwandten Texten (Joca Monachorum). Tübingen: Max Niemeyer. Tupper, Frederick. 1910. The Riddles of the Exeter Book. Boston: Ginn. Wanley, Humfrey. 1705. Antiquae Literaturae Septentrionalis Liber Alter, seu Librorum Vett. Septentrionalium, qui in Angliae Bibliothecis extant, nec non multorum Vett. Codd. Septentrionalium alibi extantium Catalogus Historico-Criticus, cum totius Thesauri Linguarum Septentrionalium sex Indicibus. Oxford: Sheldonian Theatre. <?page no="31"?> Lukas Bleichenbacher Discourse about Diglossia Swiss German and Standard German in the Newspapers Abstract While sociolinguists disagree about the extent to which the term diglossia captures the reality of present-day German-speaking Switzerland, there is a widespread consensus on the unusually high prestige of the Swiss German dialects, paired with less favourable attitudes towards Standard German. The aim of my contribution is to trace this discourse in a small corpus of opinion articles and letters to editors in two Zurich newspapers. The analysis shows that both negative statements about Standard German, and favourable ones about Swiss German, are rarer than one could expect. The diglossic situation can be considered as stable as ever, and in future research, much can be gained from further enquiry into pro-standard ideologies in different kinds of metalinguistic discourse. Introduction The notion of diglossia has occupied linguists for more than half a century, and interest shows no sign of abating. Diglossia continues to inform much work in fields such as the sociology of language and sociolinguistics and has been used, from the outset, to describe standard-dialect allocations in settings throughout history (for instance in Anglo-Norman Britain) and around the globe. Recently, the term has gained renewed relevance in selected fields of English linguistics, in areas such as Jamaica or Singapore. Moreover, concerns about domain loss among established national languages in Europe, in contexts such as academia or commerce, have resulted in a discourse about diglossia where it is feared that languages other than English might be eventually relegated to the status of mere dialects. These factors have certainly contributed to the fact that a number of Swiss researchers whose primary focus is English linguistics have crossed the linguistic boundary to comment on the case of diglossia in German-speaking Switzerland. It is, after all, also a part of their daily lives, and a rich area from which to draw inspiration for teaching. The diglossic use of Standard German in more formal contexts, and Swiss German dialects in the more relaxed ones, is generally unproblematic <?page no="32"?> Lukas Bleichenbacher 14 for, and taken for granted by their speakers, though typically not so by foreign visitors to the area, or immigrants facing the pressure of linguistic integration. However, vigorous debates periodically arise in the case of even minor changes to the established practises. In my contribution, I analyze a small corpus of newspaper articles and letters to editors in order to take stock of current metalinguistic discourse about diglossia. Specifically, I address an idea that is prevalent in much previous work on the issue, namely to what extent the Swiss German dialects are invested with a prestige that is unusually high for a diglossic situation, whereas the opposite is true for the standard variety. My aim is to demonstrate that the discourse about diglossia is more complex: While the sociolinguistic situation remains overall stable despite minor shifts, much could be gained by a closer inquiry into discourses that are critical of dialect use, and favourable towards Standard German. Diglossia or not diglossia? The academic debate German-speaking Switzerland is considered by many to be a clear case of so-called diglossia, although a number of eminent sociolinguists have disagreed with this term, which is also not generally known or used in the general public. In his 1959 article, Ferguson defined the phenomenon as “one particular kind of standardization where two varieties of a language exist side by side throughout the community, with each having a definite role to play“ (Ferguson 2003: 345). The varieties are in complementary distribution, without the continuum between standard and dialects that is typical of most sociolinguistic settings, and with little code-switching between the varieties, except of the situational kind. The main distinction is between the high or H variety, which is fully standardized, used in more formal and prestigious situations, taught at school and to foreigners, and which generally enjoys a high (overt) prestige. The low or L varieties display incomplete or no standardization; crucially, they are the only varieties admissible in everyday conversation, irrespective of the speakers’ social background. According to Ferguson, other typical contexts for the use of L include radio soap operas and folk literature, whereas H is allocated to most written and/ or formal situations of language use, including religious sermons, political and parliamentary speeches, news broadcasts and newspaper articles, and personal letters (Ferguson 2003: 347). Since 1959, however, some important changes to this distribution have been noticed with regard to Swiss German. Swiss German is now also used in some regional parliaments, in an increasing number of radio and television programmes and, perhaps most strikingly, in personal written communication (text messages, emails, but also postcards and normal letters; see, for instance, Fischer 2001: 112). Not surprisingly, this development had led to a debate <?page no="33"?> Discourse about Diglossia 15 whether the concept of diglossia is still accurate for German-speaking Switzerland. One established view is that these changes, important and interesting though they may be, do not call the overall applicability of diglossia for German-speaking Switzerland into question (see, for instance, Haas 2002; 2004). In other contributions, though, scholars have endeavoured to improve on the original model by suggesting a variety of new terms, aiming to encapsulate the Swiss situation more precisely. Most of these definitions highlight concerns about how much prestige the H variety, Standard German, actually still enjoys among the Germanspeaking Swiss, how successfully they acquire it, and whether they use it much at all. For instance, Kolde and Näf (1996: 393) use the terms “medial diglossia” (suggesting that, in general, Standard German is only written, Swiss German only spoken) and even “productive-receptive diglossia”, which implies that the use of Standard German occurs only passively, i.e. through reading and listening. Similarly, the notions of “bilingual within one (the German) language”, “bilingualism” (Ris 1990) or “asymmetric bilingualism” (Werlen 1998) point to the question of how linguistically close the H and L varieties are perceived by the speakers, which can be seen as a reaction to the frequent claim that Standard German is actually a foreign language, rather than just a different variety of one’s first language. Also, these commentators highlight how the speakers’ proficiency varies according to whether they are reading or writing, listening or speaking; specifically, it is the active use of the standard in oral communication that is perceived to be a challenge for many people, especially those with a rural and / or less educated background. More recently, Berthele (2004) has made a case for accepting the label of a (kind of) foreign language for Standard German, while Hägi and Scharloth (2005) propose the term “secondary language”. The importance of the latter contribution lies in the authors’ demonstration of how the idea of Standard German as a foreign language is essentially a folklinguistic heterostereotype. In their questionnaire study, Hägi and Scharloth found that the majority of people interviewed did not consider Standard German as a foreign language themselves. However, when asked whether they think that the Swiss in general think so, an even larger majority agreed. Analysis of newspaper articles The data compiled for this study consists of 20 opinion articles and 83 letters to the editor, published between 2001 and 2007 in three Zurichbased publications, the two daily newspapers Neue Zürcher Zeitung and Tages-Anzeiger, and Das Magazin, a weekly magazine accompanying the latter. All texts selected discuss an aspect of diglossia, in particular the use of Standard as opposed to Swiss German, as their central topic. Apart from <?page no="34"?> Lukas Bleichenbacher 16 content analysis (whether there is a clear preference or support for any variety in the argument), I analyzed specific patterns of argumentation, language choice and depiction of different varieties, and language names. In both the opinion articles and the letters, the two major issues are the spread of Swiss German (real or perceived) on the one hand, and the virtue of educational measures aimed at higher standard competence (e.g. by using the standard already in kindergarten) on the other. In many cases, there is a topical incident or development which has triggered the debate. These incidents can appear to be minor, such as the decision of the national German-speaking television channel to change the language of the main weather forecast from Standard German to Swiss German. In the opinion articles, such instances of a spread of Swiss German are most often considered problematic, especially from the point of view of intercultural communication. Authors criticize what they understand to be an increasing lack of accommodation towards those people who understand Standard German, but not the dialects, such as Frenchor Italian-speaking Swiss, and most foreigners. The use of Swiss German with these interlocutors is labelled as instances of impoliteness, narrow-minded provinciality, laziness and incompetence, or even a symptom of linguistic and cultural disintegration. The blame is put alternatively on the speakers themselves, the school teachers, and the audiovisual media. In only two out of 20 articles is there a clear pro-Swiss German argumentation. There, the dialects are considered in danger and worthy of protection; their speakers are encouraged to use them more assertively, and even with foreigners, in order to ease their integration into Swiss society. In turn, an increased use of the standard at school and elsewhere is judged as a futile endeavour because of its artificiality and complexity, which underlies its rejection by many speakers. None of the articles, however, contain any extreme statements, neither for a complete abandonment of the dialects, and even less so for an abandonment of the standard. Regarding style, there is a tendency for writers to use hyperbolic and humorous expressions. One commentator describes the Swiss German speaker’s unwillingness to use the standard as “einen so tief sitzenden Knacks” (‘deeply-seated psychological issues’; von Rohr 2006: 14). To illustrate the opposing viewpoint, another author compares the use of Standard German by school children to walking crabwise or on stilts (Vogel 2003). Linguists themselves also contribute to the debate, sometimes highlighting the benefits of the diglossic situation itself, presenting a viewpoint in support of the standard in its traditional domains. For instance, the linguist Karin Landert describes, in an interview, that young children can indeed develop positive attitudes towards Standard German if they have been exposed to stories in that variety from an early age (Arnet 2006). In sum, pro-standard arguments seem to enjoy higher popularity than previous research would suggest - unless <?page no="35"?> Discourse about Diglossia 17 one would want to claim that the newspapers systematically express unpopular views, with the risk of losing readers in already difficult times for print journalism. Analysis of letters to editors In the letters written to the editors of three publications, there is a broader variety of viewpoints. Still, the most frequently expressed arguments are those in support of Standard German; they appear in 37 out of 83 letters (45%). 28 letters (34%) are pro-dialect, while the remaining 18 do not take a clear stance. These percentages cannot be understood, though, as a direct reflection of public opinion and discourse in German-speaking Switzerland in general. There are changing patterns of adherence to and defense of either variety, depending on the context and content of the letter. For instance, the letter writers react more critically towards dialect spread in the media than in other contexts. Also, the readers of the Tages-Anzeiger, which is slightly less of a high-brow publication than the Neue Zürcher Zeitung, are overall more favourable towards the dialect in general. In turn, the Tages-Anzeiger readers are more supportive of an intensified use of Standard German in kindergarten - which shows that the picture is far from straightforward. The letters are generally written in Standard German, much like the newspapers themselves, with a few interesting exceptions. These include some letters containing parodies of different non-standard forms of German, as well as six letters written partly and entirely in Swiss German itself. In three cases, the letters match the category of pro-dialect arguments, where the choice of variety reflects the message. One such letter contains the following statement: “schliesslich sind die meischte stolz uf euse Dialekt, und das isch scho immer so gsi und wird sich au nie ändere! ” (‘after all, most people are proud of our dialect, and this is the way it has always been, nor will it ever change’; J. 2006). While this excerpt is obviously in Swiss German, the orthographic choices are fairly moderate. For instance, the writer avoids the popular use of <ä> for schwa, writing “ändere” (Standard ändern, ‘to change’) rather than ändärä, and uses the Standard German digraph <eu> in euse (‘our’) rather than <öi>, which represents more closely the actual pronunciation of the diphthong. The orthography is as close to Standard German as can be, making the text easily comprehensible even for non-speakers of Swiss German, which supports the author’s claim that an increased use of this variety would be unproblematic even in formal contexts. The opposite argument appears in a letter criticizing the abovementioned shift to Swiss German in the weather forecast that immediately follows the major evening news (which is still exclusively in Standard <?page no="36"?> Lukas Bleichenbacher 18 German), on the Swiss TV channel SF 1. The letter writer parodies a Swiss German weather forecast as follows: “Deet, wotzunä tuät fürügüxlä, ischäs mäischtsunnig. Ond übrau viumiudr, wenns mitem näbu de nit chewtr wiird.“ (‘Wherever the sun manages to peak through, it is mostly sunny. And everywhere [it gets] a lot milder, unless it gets colder again due to fog’; Eichenberger-Reichmuth 2006). Apart from the obviously trivial content, the parody becomes apparent in the orthographical choices which make the text almost incomprehensible, even for first language speakers of Swiss German. These include lack of observance of word boundaries, as in wotzunä, where the relative pronoun wo ‘where’ is joined to the following subject noun phrase tzunä ‘the sun’, which, following the orthography in the previous example, could be rendered as <wo d’Sunne> (Standard German wo die Sonne). Likewise, viumiudr (‘a lot milder’, Standard viel milder) lacks the boundary between adjective and adverb. Moreover, while the writer can be seen to imitate a Bernese variety of Swiss German with its characteristic / l/ -vocalization (spelt with <u> in viumiudr and <w> in chewter ‘colder’, respectively), there is a very intensive use of umlaut letters, most strikingly in the diminutive verb fürügüxlä (‘peak through’), rendering the excerpt even less accessible. Overall, this parody fits nicely into a European tradition of dialect stereotyping, where the non-standard indexes lack of intelligence, if not (at least from a Zurich point of view) quaint rural backwardness. We can observe here an attempt to demonstrate, through marked orthographic choices, the unsuitability of a dialect for a specific public function. The large span of attitudes towards the varieties also becomes evident in their ways they are named. For Swiss German, the neutral term is Schweizerdeutsch, and positive attitudes are accompanied by the names Mundart (‘vernacular’, lit. ‘way of the mouth’), regional specifications such as Züritüütsch (‘Zurich German’), or the use of the possessive pronoun in expressions such as ‘our language’, ‘our own German language’, and ‘our (colourful) dialect(s)’. In contrast, opponents of more dialect use employ terms such as Kauderwelsch, Bauerndeutsch, Dialektgeplapper and Mundart-Slang, which can be translated as ‘double Dutch’, ‘farmers’ German’, ‘dialect babbling’ and ‘vernacular slang’, respectively. These terms display the common associations of non-standard varieties with the idea of limitations (in terms of speakers and contexts: only farmers, or only in informal situations), if not their rejection as modes of communication altogether, as in the first and third term. The terminology for the Standard German is equally complex. The three most general terms are Standardsprache, Standarddeutsch and Hochdeutsch (‘standard language’, ‘Standard German’ and ‘High German’). Writers favourable towards this variety also use Schriftsprache (‘written language’ or ‘language of writing’) and Hochsprache (‘high language’), <?page no="37"?> Discourse about Diglossia 19 whereby the term high points to the notion of formality rather than, as in Hochdeutsch, a geographical region. These terms also collocate repeatedly with attributes such as korrekt ‘correct’ and gepflegt ‘cultivated’. The more negative labels include humorous creations suggesting that the kind of Standard German spoken in Switzerland is either ‘too Swiss’, e.g. used with exaggerated admixture of features from Swiss German, or indeed ‘too German’. The first idea is illustrated by the terms Parlamentarierdeutsch ‘MPs’ German’, as well as Allemand-fédéral and Deutsch-fédéral (both meaning ‘federal German’, modelled on the ironic term français fédéral ‘federal French’, for a variety of French used as a foreign language in communication across the Swiss language borders, especially in formal and political settings). These terms refer to one aspect of the stereotype of Standard German as a foreign language, namely that it is especially badly spoken by politicians (who can be witnessed using Standard German frequently in news reports). In contrast, varieties devoid of any Swiss features are labelled as Tüütschlandig (‘Germany-ish’) or Hauchtüütsch, a variation of Hochtüütsch/ Hochdeutsch, jokingly based on the imitation of a hypercorrect pronunciation. A strong sense of humor is also evident in a further feature of some letters, namely a line of argumentation where one recent shift in the allocation of a variety is criticized by speculation on what is likely to follow, which results in a kind of hyperbolic chain of events. In one letter, the shift to Swiss German in the weather forecast could lead, it is suggested, to the exclusive use of Swiss German place names, which in turn would make Standard German obsolete for newspapers as well, and eventually, all documents would appear in Swiss German only. In another letter, an author criticizes the introduction of Standard German into kindergarten with the underlying aim for the children to acquire the standard as early and easily as possibly. The letter writer ironically suggests that German women should be brought into the country in order to become mothers, and then transmit the standard language to children, even before kindergarten. Both examples are quite obviously light-hearted, which testifies to an absence of any serious kind of language strife, but at the same time, they are firmly based on the notion that patterns of language choice should best be kept exactly the way they are - an ideology which underlies many of the texts analyzed. Conclusions The results of the content and discourse analysis of metalinguistic discourse on the use of Standard and Swiss German in German-speaking Switzerland contradict the picture of almost unanimous pro-dialect attitudes or even a general ideology of the dialect as it has been proposed in most previous <?page no="38"?> Lukas Bleichenbacher 20 research. There are contradictory opinions which depend on the different contexts and also the different forms of Standard German and Swiss German dialects, and there is no clear tendency in either direction. Instead, there appears to be a strong sense of stability, which is argued to be a defining feature of diglossic communities itself: “Any attempt to alter these habits might be considered threatening and radical; it would certainly take a considerable effort on the part of the speech community […] to change the status quo” (McColl Millar 2005: 2). If the use of Swiss German does spread, in the near future, into further contexts (e.g. in electronic communication), it is unlikely not to be noticed or to lead to further intensive debates, and to more official or unofficial support for Standard German in turn. The discourses surrounding the latter are especially worthy of scholarly scrutiny because they have been somewhat neglected so far, which may have stemmed from an impetus to present the Swiss case as something even more particular and special than it already is. In sum, diglossia in Germanspeaking Switzerland has been, and is likely to remain (to exploit another Swiss stereotype) courant normal - and as long as public discourse remains moderately humorous, not a major source of problems. Biographical sketch Lukas Bleichenbacher (lukas.bleichenbacher@phsg.ch) has been inspired, in many ways, by Andreas Fischer ever since his first week as a student of English at the University of Zurich, in October 1995. After graduating in 2002, he had the privilege to work as an assistant for Andreas Fischer until 2008. He is now Lecturer of English Studies at the University of Teacher Education St.Gallen (PHSG). His research interests include discourseanalytic and sociolinguistic approaches to multilingualism, multiculturalism, language policies and language ideologies. He has published several articles and book chapters in these fields, as well as, based on his doctoral thesis, a book entitled Multilingualism in the Movies (Francke, 2008). <?page no="39"?> Discourse about Diglossia 21 Works cited Primary sources Arnet, Helene. 2006. “Standardsprache als Beziehungssprache akzeptiert”. Tages- Anzeiger, 22 September 2006, p. 25. Eichenberger-Reichmuth, Helen. 2006. Letter to the Tages-Anzeiger, 12 January 2006, p. 18. J., D. 2006. Letter to Das Magazin 08/ 2006, p. 5. Vogel, Alfred. 2003. “Hochdeutsch-Zwang ist keine Lösung”. Tages-Anzeiger, 27 July 2003, p. 15. von Rohr, Mathieu. 2006. “Fremdsprache Deutsch”. Das Magazin 06/ 2006, pp. 14-15. Secondary Sources Auer, Peter. 2005. “Europe’s Sociolinguistic Unity, or: A Typology of European Dialect/ Standard Constellations”. In Nicole Delbecque et al. (eds.). Perspectives on Variation: Sociolinguistic, Historical, Comparative. Berlin: Mouton de Gruyter, pp. 1-42. Barbour, Stephen and Patrick Stevenson. 1990. Variation in German: A Critical Approach to German Sociolinguistics. Cambridge: Cambridge University Press. Berthele, Raphael. 2004. “Vor lauter Linguisten die Sprache nicht mehr sehen - Diglossie und Ideologie in der deutschsprachigen Schweiz”. In Helen Christen (ed.), pp. 111-136. Christen, Helen (ed.). 2004. Dialekt, Regiolekt und Standardsprache im sozialen und zeitlichen Raum. Wien: Edition Praesens. Ferguson, Charles. 2003. “Diglossia”. In Christina Bratt Paulston and Richard Tucker (eds.). Sociolinguistics: The Essential Readings. Malden, MA: Blackwell, pp. 345-358. (First published 1959). Fischer, Andreas. 2001. “Language and Politics in Switzerland”. In John M. Kirk and Dónall P. Ó Baoill (eds.). Linguistic Politics: Language Policies for Northern Ireland, the Republic of Ireland, and Scotland. Belfast: Cló Ollscoil na Banríona, pp. 105-122. Haas, Walter. 2002. “Comment” [on Alan Hudson's focus article “Outline of a Theory of Diglossia”]. Focus on Diglossia. International Journal of the Sociology of Language 157, pp. 109-115. Haas, Walter. 2004. “Die Sprachsituation in der deutschen Schweiz und das Konzept der Diglossie”. In Helen Christen (ed.), pp. 81-110. Hägi, Sarah and Joachim Scharloth. 2005. “Ist Standarddeutsch für Deutschschweizer eine Fremdsprache? Untersuchungen zu einem Topos des sprachreflexiven Diskurses“. Linguistik Online 24/ 3. Kolde, Gottfried and Anton Näf. 1996. “Die Westschweiz”. In Robert Hinderling and Ludwig M. Eichinger (eds.). Handbuch der mitteleuropäischen Sprachminderheiten. Tübingen: Narr, pp. 385-412. McColl Millar, Robert. 2005. Language, Nation and Power: An Introduction. Houndmills, Basingstoke: Palgrave Macmillan. <?page no="40"?> Lukas Bleichenbacher 22 Ris, Roland. 1990. “Diglossie und Bilinguisme in der deutschen Schweiz: Verirrung oder Chance? ” In Jean-Pierre Vouga and Max Ernst Hodel (eds.). La Suisse face à ses langues = Die Schweiz im Spiegel ihrer Sprachen = La Svizzera e le sue lingue. Aarau etc.: Sauerländer, pp. 40-49. Sieber, Peter and Horst Sitta. 1986. Mundart und Standardsprache als Problem der Schule. Aarau etc.: Sauerländer. Watts, Richard J. 1999. “The Ideology of Dialect in Switzerland”. In Jan Blommaert (ed.). 1999. Language Ideological Debates. Berlin: Walter de Gruyter, pp. 67-103. Werlen, Iwar. 1998. “Mediale Diglossie oder asymmetrische Zweisprachigkeit? Mundart und Hochsprache in der deutschen Schweiz”. Babylonia 1/ 98, pp. 22-35. <?page no="41"?> Margaret Bridges Verbal Duelling on the Beach: Cultural Translation and the Agonistic Vocabulary of Kyng A lisaunder Abstract In the romance tradition, Alexander the Great has no patrilineal claim to succeed Philip on the throne of Macedon. One of the narrative’s legitimizing moves consists in representing the adolescent hero passing a test that takes the form of both a verbal and a non-verbal combat. This study, which is indebted to historical pragmatics, looks at the lexis, form, function and setting of Alexander’s interaction with his first antagonist in Kyng Alisaunder, a late thirteenth-century romance strongly marked by linguistic and literary contact with Anglo-Norman French. Tracing the trajectory of the episode back through several ascertainable stages of rewriting allows us to reflect upon the continuities and discontinuities marking the cultural translation of this testing - and insulting - of Alexander in his Greek, medieval Latin, Anglo-Norman and English literary contexts. Introduction In the course of his study of Old English terminology for engagement, wedding and marriage, our honoree and distinguished lexicographer also included linguistic evidence of terms used to gloss or translate Latin words for marriage-breaking or adultery (Fischer 1986: 67, 73 n. 136, 91). It may not be inapposite, therefore, to take as a starting point for this article one of the first attested intradiegetic uses in English literature of the expression ‘son of an adulteress’. However extensive their usage outside literature may have been, horeson and fiz a putains do not make their (bilingual) appearance in Middle English writing as terms of insult until the late thirteenth century, when they are levelled at the romance hero Alexander the Great. Although the Thesaurus of Old English suggests that our early lexis had no dearth of terms for designating extramarital sex or men and women engaging in promiscuous sexual practices (lucrative or not), there is very little written evidence of a ‘consensual’ term for the illegitimate offspring of those practices, the bastard. In addition to the Anglo-Norman loanword bastard itself, which comes into English as an epithet of William in the <?page no="42"?> Margaret Bridges 24 1066 entry of the D version of the Anglo-Saxon Chronicle 1 and continues to be collocated with William’s name in Middle English chronicles, a variety of Old English terms (cifesboren, dooc, docincel, hornungsunu) are found as glosses, which occur either once or twice only, and are never embedded in a context of ‘flyting’ or taunting. 2 In the case of the term horeson, Anglo- Norman French seems to have paved the way for this Middle English calque of filz a putain, whose degree of ‘Frenchness’, or foreignness, is difficult to gauge in this early English romance, the bilingual author of which was inspired by an Anglo-Norman roman to write for what may well have been a diglossic audience. 3 As a lexical item, filz a putain has been identified by the historical pragmatist Dominique Lagorgette as part of a subgroup of insults stigmatizing its addressee as unworthy of belonging to the group (1994: 324-326). But like other historical linguists, Lagorgette has pointed out that the lexical item alone yields little information about the speech act of which it is part. Moreover, it does not in and of itself function as an insult. In what follows I am therefore not so much concerned with the linguistic history of this lexical item as with its function in its early literary context, in an episode pitting the youthful Alexander the Great against his first opponent. In so doing, I gratefully acknowledge the pioneering studies of verbal violence in the context of medieval French literature (Lagorgette 1994), of The Canterbury Tales (Jucker 2000), of medieval Dutch romance (Bax 1981), and of Homeric and Old English poetry (Parks 1986 and 1990). Both within and outside literature, the utterance ‘son of a whore’ is likely to be a non truth-conditional insult that may be ritualistic in nature 1 The entry reads Wyllelm Bastard, once only (Cubbin 1996: 79). 2 Although it is tempting to speculate that the late development in English of ‘son of an adulteress/ a whore’ as a term of insult may be related to the lack of stigma attached to bastardy in the Middle Ages, as argued by Given-Wilson and Curteis (1984: 48-51), early Anglo-Norman usage of filz a putain and its variants (see next footnote) does not support such speculation. 3 Kyng Alisaunder, a free adaptation of the twelfth-century Anglo-Norman Roman de toute chevalerie (in which the expression does not, however, seem to occur) has been variously dated between the last quarter of the thirteenth century and the first quarter of the fourteenth. Most recently, critics have opted for a date before 1300 - a dating which locates this poem well before the perceived ‘loss of French’, even according to older language histories of Britain; the current trend is to query that loss for virtually the whole of the English Middle Ages (see e.g. Butterfield 2009: 51). Fiz a putaines also appears in an early manuscript of Cursor Mundi and houres sone is attested in Bevis of Hampton, dated 1325 by its latest editors Ronald B. Herzman, Graham Drake, and Eve Salisbury (1999). All the ‘English’ literary uses of fiz a putaines and its variants postdate the term’s appearance in Anglo-Norman literature, as evidenced, for instance, by Marie de France’s Eliduc (843), or, more interestingly for our purposes, by the late twelfthcentury Roman d’Alexandre, where Alexander is led to enclose the people of Gog and Magog after they have called him a fil a putain (III, 2162). <?page no="43"?> Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 25 (uttered in play) or personal (uttered in earnest and intended to hurt). 4 In the relevant episode from the early Middle English romance, however, the truth conditionality of this insult turns out to be of unsettling pertinence. Moreover, the accusation of bastardy levelled at Alexander is part of a dialogic exchange that functions as a contract for a fight - a fight that is also a vital part of the process of the hero’s legitimization. When studied against the backdrop of the cross-cultural, literary transmission of the Alexander legend, it will become obvious that the flyting as well as the fighting are marked by both continuity and change. In order to trace this double trajectory, I will be drawing attention to the prehistory of the episode in which Alexander the Great becomes the first English romance hero to have the dubious honour of having the insult of bastardy flung at him in these words that are still part of our agonistic vocabulary. The transformations undergone by the lexis, form and function of this agonistic testing of the hero reflect the longevity and cultural spread of the Alexander legend, whose numerous translations and rewritings across the centuries inevitably involved suppressing or acculturating obsolescent or culture-specific forms and contexts of verbal violence. The setting is a coastline, where water meets land. After disembarking with his companions, a fifteen-year old warrior-knight, as yet untried on the battlefield, encounters another young warrior, king of the land, whom he has sought out because of a family history of feuding. The visitor meets with verbal hostility. The two engage in an increasingly violent exchange of verbal and non-verbal insults until one of them, the lord of the land, has a lucky escape to the safety of his home camp. After a restless night, the two armies engage in battle: further threatening words (wordes) precede the encounter of spears (ordes), leading to general carnage and the death of many a noble knight on both sides. The visiting knight fulfils his oath to requite the young king’s insults by beheading him, despoiling and destroying his illustrious city and embarking on his return journey with his newly acquired crown. This is a succinct summary of lines 839-992 of Kyng Alisaunder (henceforth KA), a poem of over 8000 four-stressed lines arranged in rhymed couplets. 5 The visiting warrior-knight is the adolescent Alexander, freshly dubbed and proclaimed as heir to Philip of Macedon, in spite of a genealogy as heroically uncertain as that of the legendary King Arthur or, say, the Greek hero Hercules, for in the romance tradition Alexander’s natural 4 This distinction is derived from Jucker (2000: 375). Parks’s comparable distinction is between referential modes that are “ludic” or “serious” (1986: 446-47). 5 Unless specified otherwise, I will be quoting from and referring to MS Bodleian Laud Misc. 622 (mid fourteenth century), as edited by G. V. Smithers (vol. I, 1952; vol. 2, 1957). <?page no="44"?> Margaret Bridges 26 father is the oriental Egyptian pharaoh, astrologer and magician Nectanebus, who has tricked Olympias into committing adultery in Philip’s absence. These lines follow the episode in which Alexander’s chivalric mettle has been acknowledged by his mastery of the wild, man-eating horse Bucephalus; together these two episodes can be read as attempts to remove the stain of illegitimacy which implicitly functions as an impediment to Alexander’s accession to the throne of Macedon. After all, at more or less the same time as the London poet was composing his version of Thomas of Kent’s Roman de toute chevalerie (henceforth RTC), Peter of Langtoft wrote that la lay ne chaunte/ Ke bastard enporte regioun saunz haunte/ Ou saunz coup d’espeye gaynez sur tyraunt (in Robert Mannyng’s words: a bastard no kyngdom suld hald/ Bot if þat he it wan with suerd or with lance/ Of tirant or of sara 3 in). 6 In other words, conquest (“with sword or lance”) cancels out the stigma of bastardy and from this medieval political perspective on right rule it is vital that the young Alexander should acquire a crown through successful warfare against an opponent who rules unjustly (as tyrant). 7 The fact that the spectre of illegitimacy will continue to haunt the hero of this poem lends current relevance to the insults that are not unique to this passage. 8 Alexander’s opponent in verbal and military combat is Nicholas of Carthage, and the episode bears many of the hallmarks of the ritual challenge between medieval knights. But before I proceed to a more systematic analysis of the striking verbal and non-verbal forms of communication marking the hostilities of Alexander and Nicholas in Kyng Alisaunder, I would like to summarize another version of the same encounter, written at least one thousand years earlier. 6 Langtoft (1866 I, p. 372; no line numbers). Robert Mannyng of Brunne’s Chronicle, lines 1224-1226 (1996: 519). 7 Unjust rule was one form of tyranny recognized by the scholastic Middle Ages, the other form being the usurper’s tyranny of title. 8 Immediately following this episode, Alexander returns to Macedon only to discover that his mother has been repudiated on account of rumours that she hore had ybene (998) and that Alisaunder was false ayre (1000). During the siege of Thebes, an unnamed knight storms towards Alexander, flooding him with insults: Whare artou, hores son, whare? / An hore þee to man bare./ Þou auetrolle! Þou vile wrecche! (2687-2689) The word auetrolle (from French avoultre, “adulterer” - a term used by the King of Thessaly to insult Alexander in a different episode of Kyng Alisaunder’s Anglo-Norman source) possibly relates to the addressee’s own sexual promiscuity rather than to the fact that Alexander is the offspring of promiscuity, as glossed by Smithers (1957: 163). The same combination of insults (for the offspring and the practitioner of promiscuity) recurs during the hostilities with Darius’s Persians, when Alisaunder himself addresses a “foreign knight” as fitz a puteyne and lecchoure (3912), without any obvious current relevance as he doesn’t know the knight in question. Although both terms are clearly general terms of insult without truth conditionality, they surely raise the spectre of illegitimacy in a work written at a time when the price being paid for cultural colonization - in the widest sense of the term - was anxiety over genealogy and legitimate rule. <?page no="45"?> Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 27 The promising but as yet unproven Alexander begs his father to allow him to compete in the four-horse chariot races of the Olympic Games at Pisa, on the Peloponnese peninsula. Alexander - who insists on going with his own horses rather than with those proffered to him by Philip - receives his father’s permission and, after travelling by sea, Alexander, his friend Hephaestion and the slaves in charge of the chariots and horses disembark at Pisa, where they encounter Nicolaus, son of the king of Arcanania, also entered for the races. The greetings exchanged between the two rivals are tense. There follows a vivid description of the chariot race, initially involving nine competitors. Alexander reveals his sense of strategy in allowing Nicolaus, driven by the hatred of a man whose father was killed by Philip, to overtake him and to believe that he is the victor; but Alexander, controlling the situation from behind, catches the axle of Nicolaus’s chariot, which overturns, killing its occupant. In the temple of Zeus a prophecy announces that “as [he] has conquered Nicolaus, so [he] will conquer many others in war.” The crown with which he returns to Macedon is the victor’s garland. 9 This summary of the agonistic encounter between two rivals in a competition of both political and religious significance comes from the Greek Alexander Romance (henceforth GAR), also known as Pseudo- Callisthenes, the oldest (lost) version of which is dated some time between the first century BC and the third century AD; the oldest extant translations diverge greatly, as do the various Greek manuscripts which they translate and rewrite. 10 The centuries separating the Greek from the Middle English episodes just summarized trace a history of textual transmission and translation that has many gaps. What we do know, however, is that some of the major stages through which this episode transited in the course of its transformation from a Greco-Byzantine chariot race to a medieval ritual challenge between knights on territorial borderland included a ninthcentury summary (Epitome) of an early Latin version of the Greek Romance 11 as well as an Anglo-Norman romance, probably written in the last 9 For the purposes of this paper, I have consulted the recent English translation by Richard Stoneman (1991). My summary of the Olympic-Games episode is based on Book I, chapters 18 and 19 of the Greek romance (pages 49-52). All references to the Greek Alexander Romance (GAR) are to this (composite) edition and translation, by chapter and page number. 10 The most helpful overview of the tangled relationship of the various manuscripts and manuscript-groups is to be found in Jouanno (2009: 11-28). I am further greatly indebted to Corinne Jouanno for generously responding to my queries about GAR. 11 The (anonymous) Epitome of Julius Valerius’s Res gestae Alexandri Macedonis is generally known by the name of its first editor, Julius Zacher (1867), whose edition I refer to throughout, by chapter and page number. The dating of Julius Valerius has proven controversial; most recently, Callu (2010) has firmly embedded the Res gestae in between the mid first-century BC and the year 16 AD. <?page no="46"?> Margaret Bridges 28 quarter of the twelfth century by a self-inscribed Thomas of Kent, who claims to be writing in a language that is not his own. 12 Kyng Alisaunder (henceforth KA), written anonymously in or near London, in a ‘minstrel’s voice’, was inspired by Thomas’s work, which in turn goes back to the Epitome. These are four discernable stages, then, in one of the many strands of Western European rewriting of the Alexander legend. Long-lived as its appeal may have been (as suggested by its manuscript history), 13 KA seems from our vantage point to terminate this strand of transmission, since there is no evidence of this Middle English romance having affected any of the other extant English writings about Alexander. 14 Although all four stages may here serve to foreground some characteristic features of cultural translation, my analysis will perforce be privileging the wording of the Middle English romance. The following reading of KA 839-992 and its corresponding passages in earlier stages of the testing-of-Alexander episode will take into account its embedding in its larger narrative frame, setting, contestants, the verbal and non-verbal contest, and resolution. Embedding As we shall be seeing, it is not very helpful to classify the context for Alexander’s eristic flyting as an inter-societal one, 15 even though in all texts under consideration, the testing of adolescent Alexander takes place away from ‘home’, in an overseas locus agendi that receives special emphasis through reference to the building of Alexander’s ship (in GAR), its loading and the (dis)embarkation processes prominent in the Anglo-Norman and English versions. I have already suggested that the implicit rationale for the 12 The Roman d’Alexandre ou Le Roman de toute chevalerie (RTC) was edited by Brian Foster and Ian Short, whose text was in turn integrated in the bilingual edition by Catherine Gaullier-Bougassas and Laurence Harf-Lancner (2003). References throughout are to line numbers in this edition. I am grateful to Catherine Gaullier-Bougassas for pointing out to me some major differences between RTC and the roughly contemporary continental Roman d’Alexandre, in which the insult of bastardy is frequently levelled at the hero by a variety of antagonists, even though the narrator himself firmly denies that Alexander was born out of wedlock. 13 The romance was still being read after the introduction of printing, as witnessed by the early sixteenth-century fragments bound up in The Bagford Ballads. 14 The same can be said of all the writings about Alexander the Great in English; while KA and the two extant Scottish Alexander romances appear to reflect French source materials, the alliterative Alexander fragments and the Thornton prose Alexander, like the prose Letter of Alexander to Aristotle, seem to be translations from Latin. There is no evidence of any one English Alexander text being in any way indebted to any other English Alexander text - perhaps a reflection of the bookish nature of the ‘matter of antiquity’ and of the authority conferred on a text by its status as translation. 15 Ward Parks (1986: 452) has distinguished inter-societal eristic flyting from intra-societal ludic flyting. See further my reflections below on the contestants. <?page no="47"?> Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 29 need to conquer abroad lies in the fact that the romance hero’s genealogical claim to rule at home is open to dispute. Where the versions differ from one another relates to the explicit reason proffered for Alexander's embarkation as well as to the sequence of episodes within which the voyage is embedded. Whereas the Greek and Latin versions have a sequence moving from Alexander’s mastery of Bucephalus to his mastery of a four-horse chariot - the most prestigious of the Olympic disciplines - the Anglo- Norman and Middle English texts develop an episode in between these two testing-of-the-hero ‘narremes’ 16 : the dubbing episode (KA 801-838; RTC 533-563) not only affords an obvious instance of medieval acculturation of classical materials that were no longer self-explanatory, but also provides a new rationale for a contest episode that did not stand in a relationship of causality to the episode preceding it. In the English KA, as in RTC, Alexander has just been dubbed by a loving Philip, together with a hundred other young knights, and generously equipped with horses carrying rich loads of gold (KA 827, 837). In both works, Philip’s lavish gift-giving creates a bond binding the recipient to the donor. The English poet alone, in a further move of legitimization, assimilates Philip’s gift of arms and dubbing to a “coronation” in Corinthe toune (word-play in 805-806 makes up for geographical fuzziness), a coronation which in turn is glozed by the English poet as Philip’s designation of Alexander as his heir. 17 Adding further to these legitimizing precautions, the English poet alone grounds Alexander’s first adventure in an ethos of kinship revenge: before the untried warrior leads his horses, elephants and camels onto his ships, he has sworn an oath to wreke his fader of his foon (844). Alexander heads for a specific destination sumdel fer þenne, where there was a king þat had greued mychel his kynne (839-840). The shining armour of the knights as they eagerly disembark on the shore of the land they have sought out furthermore affords a glimpse, unparalleled in other versions of this episode, of an epic motif generally referred to as ‘the-heroon-the-beach’. 18 The embedding of Alexander’s first adventure abroad in the English romance, then, suggests that the courtly narrative stands in a palimpsestic relationship to earlier narrative conventions, which it overwrites while continuing to occupy the same space. They remain, I suggest, 16 I use the term loosely in the sense of a narrative unit standing in a functional (usually causal) relationship to other narrative units in the general narremic substructure of a work. The term ultimately derives from Dorfman (1969). 17 For he wolde in Corinthe toune/ Alisaundre his sone coroune/ Þat is, forto sigge ari3th/ 3iue hym armes and maken hym kni3th/ And maken couþ to alle gyng/ After hym he shulde be kyng (805-810). There is no precedent for this reference to Corinth and the other English MS containing this episode, Lincoln’s Inn 150, omits any reference to location. 18 Although criticism seeking for oral formulaic ‘hero-on-the-beach’ type scenes has itself been beached in recent years, there is no denying that the motif has been shown to permeate much early epic narrative. <?page no="48"?> Margaret Bridges 30 in play. To be sure, the animosity between the contestants is also related to a family feud in GAR. There it is, however, in the middle of the excitement of the chariot race that we learn that Alexander’s opponent was “less keen to win than to destroy Alexander. Nicolaus’s father had been killed in battle by Philip. Alexander was intelligent enough to realize this [….]” (GAR, 19, 51). Through thought-representation, twice removed, the narrator reports how Alexander realizes that Nicolaus’s thoughts are intent upon revenge and that he uses this realization to his strategic advantage in the contest. The hero’s own motivation is not in any way represented as being related to this killing, and nothing more than an ambitious boy’s fancy seems to account for his leaving home: “One day, when Alexander found his father relaxing, he kissed him and said, ‘Father, I beg you to allow me to go to Pisa to the Olympic Games; I want to take part [….] in the chariot races’” (GAR, 18, 49). Without any attempt to link chapter seventeen (in which the fourteen-year old Alexander fulfils a prophecy that the identity of Philip’s successor - and future world conqueror - will become apparent only when he rides Bucephalus through the streets of Pella) to chapter eighteen (fifteen-year old Alexander heads for Pisa), the ninthcentury Epitome reports Alexander’s request and Philip’s response in two short sentences (Epitome, 18, 21). Pisa is mentioned as the place where the chariot races will take place; there is no reference to Olympic Games, and what a ninth-century West-European cleric imagined chariot races (quadrigis certando) to have been like can only be a matter of conjecture. The epitomizer certainly wastes no time in describing them. How then has cultural translation affected the embedding of the contest that will allow the hero to establish his worthiness to rule? To sum up, Alexander’s chariot race in GAR and Epitome directly followed another equestrian test. In the West, this kind of race had ceased to be a social reality well before the ninth century, when the Latin Epitome was written. Yet the epitomizer, for whom the races well may have been recognizable in name only, reports in considerable detail (as we shall be seeing) the verbal contest that precedes the dies certaminis; details concerning the race itself are simply omitted in favour of the result. Together, the two episodes serve to underline the young Alexander’s equestrian skills and courage; only the verbal contest preceding the chariot race, however, allows him to demonstrate his philosophical skills, while the race itself reveals his tactical brilliance. Accretions in the Anglo-Norman and Middle English romances, in the form of a dubbing episode, provide a rationale for the hero’s contest, in which horsemanship is no longer a major issue and which is carried out in the name of a family feud (KA). By linking the dubbing with the contest abroad, the medieval poets additionally construe Alexander’s first conquest as Philip’s just reward for his generosity (KA and RTC), performed in the name of chivalric fealty. <?page no="49"?> Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 31 The Setting The setting of the encounter between Alexander and Nicholas both resembles and does not resemble the settings that have generally been associated with verbal contests, since we are dealing neither with a battlefield (as in The Iliad) nor with a stretch of water across which contestants taunt one another (as in The Battle of Maldon, or in Harbard’s Song). In all four texts under consideration the young men meet somewhere near the coast, but in the earliest (Greek) version of the episode it is probably merely fortuitous, since the contestants are on their way to the Olympic Games, which take place in the North-Western Peloponnese. 19 More to the point is the fact that the Epitome imagines the encounter at Elim 20 resulting from a diversion caused by a storm at sea: in the ninth century, this change of setting marks the end of boasting and taunting within a space set aside for game and religious ritual. The historically determined loss of such a place of licensed competition leads later medieval writers to locate the confrontation on a threshold to territory, sovereignty over which is threatened or contested. When the visitor reaches the stronde (KA 865), or the coastline beneath the cité marage (RTC 567), the outcome of his verbal and nonverbal contest with the lord of the land will determine which of the two has power over that space (centred around the city of Elim in RTC, and around the city of Carthage in KA). In the course of its cultural translation, the competition between adversaries in a circular hippodrome whose starting point and finish are determined ‘in game’ has been transformed into a deadly contest between sons of kings for the right to step over and beyond the line into the other’s space, which in the worlds of RTC and, to a lesser degree, of KA can only be imagined as feudal space in which power is measured by tribute and where there is only room for one at the top. The Contestants In KA, Alisaunder’s opponent is called Nicholas of Carthage. There is little to indicate just what the associations of the thirteenth-century poet were with that epithet, but surely the name of an illustrious coastal city with a no-less illustrious literary past would be a good place for the young hero’s first conquest, while the other names associated with Nicholas (Elim is his kingdom in RTC, even though he is said to have been born near Carthage; 19 The ancient town of Pisa has not yet been located with any certainty, but it was probably very close to Olympia, some 20 kilometres inland. 20 Elim is a variant of Elis, a city 60 kilometres North of Olympia, which supplanted Pisa as the place where athletes trained as of the sixth century BC (Golden 2003: 59). <?page no="50"?> Margaret Bridges 32 Acarnania in GAR and the Epitome) 21 were clearly insignificant enough in their new cultural environment to warrant substitution. More important than this free-for-all regarding the kingdom over which Alexander’s opponent rules is the type of relationship that he is represented as having with the hero. This too varies considerably as the episode is rewritten over time and cultural space. In GAR, Nicolaus is one of nine competitors entered for the prestigious chariot races, each one representing a different socio-political group. 22 He and Alexander meet when the latter, freshly disembarked, goes for a walk with his friend Hephaestion: intent as he is only on the game, the apparent aimlessness of Alexander’s movements at this point underscores the Macedonian’s lack of hostile intention. When they meet and exchange greetings - of which more below - the two competitors do not seem to have met before: the animosity arising from the ‘father’ of the one having killed the father of the other is not reported until well into chapter 19, when the race itself is already half run. The verbal exchange preceding the competition by several days essentially reflects the prestige attached to participating in, and winning, the chariot race. The ninth-century Epitome reproduces the chance encounter between the two participants in the races, but we find an accretion that problematizes the status of the contestants as representatives of different social groups, and therefore undermines the inter-societal nature of the competition. For the Epitome characterizes Nicolaus as a former ‘schoolmate’ of Alexander’s: Nicolaus rex Acernanum quondam conscolasticus eius (18, 22). It is a detail of which much will be made in Thomas of Kent’s RTC. Nicholas is there said to have known Alexander and many members of his family well, for they learned their lettrure et langage (literature and rhetoric) together. 23 For all the difference between kin and kingdom, the two young men were part of the same intellectual élite by virtue of their shared education, perhaps assimilated by Thomas to that of an Anglo-Norman clerc. Nicholas’s animosity towards Alexander therefore resembles a betrayal from within, a breach of propriety. At any rate, the distinction between inter-societal and intra-societal conflict has become blurred and the conflict, which is 21 In the popular eleventh-century Historia de Preliis, which the poet of KA may also have been familiar with, Nicolaus is styled both King of the province of the Peloponnese as well as King of Aridea. The former was, as we have seen, associated with the Games, whereas the latter lay to the north of Pella, the principal residence in Macedon of both Philip and Alexander (Kratz 1991: 9). 22 Four of these competitors are said to be the sons of kings, the five others being sons of generals and satraps (GAR 19, 50). 23 Bien conust Alisandre e mult de son lignage,/ Apris orent ensemble lettrure et langage (572- 573). <?page no="51"?> Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 33 inscribed neither in game nor in an ethos of revenge, seems to arise out of the meeting on the shores of Elim. The English Alisaunder is unique in singling out Nicholas (who has not gone to school with Alexander) as his opponent for his first confrontation, as I have already indicated. Rooted in revenge, and in the ‘proper’ desire to recompense Philip for his generosity, Alexander’s confrontation with Nicholas has been overtly and covertly embedded in a contest that ennobles its victor as the representative of chivalric and philosophical values. His ill-tempered and foul-mouthed opponent, as we shall be seeing, makes a number of verbal (and non-verbal) moves that challenge those values, so that the opposition becomes one between a future king who passes the test of knighthood and a tyrant who is unworthy of his kingship: the embedding of this episode (growing out of a double ceremony of dubbing and crowning) has already made it abundantly clear that only he who best embodies the values of chivalry has the right to rule, bastard or not. Yet running parallel to this discourse of difference we shall be noting a tendency in KA to reduce that difference: as feelings begin to run high on both sides, the narrator empathizes with both Nicholas and Alisaunder as ‘angry young men’, mutually provoked to engage in combat. Verbal and Non-verbal Communication Flyting, in the sense of an exchange of insults and boasts either preceding or substituting fighting, has been studied in a variety of literary and cultural contexts. While many of these studies reveal patterns of verbal interaction that overlap with what we find in the testing of Alexander, these patterns are only minimally useful in tracing the long trajectory of cultural translation undergone by this episode. The most striking changes affecting the verbal encounter can be summarized as follows. The opening conversational move is one of qualified politeness in the Greek and Latin versions of the text. To Nicolaus’s “Greetings, young man! ” Alexander responds with “Greetings to you, too, whoever you may be and wherever you come from” (GAR 18, 49). The potential for insult contained in the qualifications “young man” (the belittling puer of the Epitome is said to be part of a salutatio non sine contumelia, 18, 22) and “whoever you may be” (this real or feigned non-recognition may be compared with the Epitome Alexander’s decision not to return Nicolaus’s greeting) 24 will be emphasized in its medieval rewriting by Thomas of Kent, whose Alexander is addressed as enfes 24 Nec ille resalutat (Epitome, 18, 22). Since the author of this summary had characterized Nicolaus as someone who had gone to school with the young Alexander, he could hardly claim not to know the person greeting him. <?page no="52"?> Margaret Bridges 34 de mult jovene aage (RTC 575), i.e. “child, very young in years”, which is followed two lines later by the degrading address valet. KA omits the greetings altogether in favour of a request for information as opening conversational move. Following his salutation by Alexander as “whoever you may be and wherever you come from”, Nicolaus in GAR reveals his identity: “I am Nicolaus, the king of Acarnania.” This prompts Alexander to accuse him of pride and boasting, which fate may well put an end to at any moment: “Do not pride yourself so, King Nicolaus, and glory in the assumption that your life will last to the morrow; for fate is not accustomed to stay in one place, but a turn of the balance makes mock of the boastful man” (18, 49). In its essence this philosophical reflection - whose perlocutionary thrust is evidenced by it being simultaneously an insult (‘you are an ignoramus whose pride is in the wrong place’), a boast (‘I know better’) and a threat (‘that fate will be yours if you lose the chariot race’) - survives with minimal variations across place and time. 25 The twelfth-century Anglo-Norman poet increases the verbal violence and emphasizes the boasting inherent in Nicholas’s affirmation of his identity and feudal power. His insolent opening conversational move, spoken in mult grant culvertage, had been to greet the trespassing knight as a very young child. The perceived insult of Alexander’s astonished silence subsequently leads Nicholas to challenge the trespasser, degraded to the status of a feudal inferior, by demanding tribute of him: […] Valet, tu n’es pas sage. Tu ne sies qe jeo sui, ne quel est mon parage. Roy sui de cest regné e des autres ay truage, De l’orgoil qe tu fez ore me doing ton gage. (RTC 577-580) 26 This elicits less of a philosophical response from Aristotle’s student than a consideration on the need to complement the advantages of birth with the chivalric virtues of pruesce (“prowess”, 584) and bon corage (“resolution”, ibid.). Alexander’s medieval moralizing on the worthlessness of noble birth without noble character traits is preceded by an outright insult: to Nicholas’s earlier accusation that he lacks sense (tu n’es pas sage, 577) Alexander replies that Nicholas speaks but folly and madness (folur e rage, 581). Even if (covert or overt) verbal violence already characterizes these opening speech actions, as far as their sequencing is concerned, the verbal exchange here summarized in three of the four works can roughly be subsumed 25 Quid […] prodest tibi ista vana imperii iactatio de secundis crastinis fluctuanti (Epitome 18, 22) is a compact enough summary of GAR’s insult/ boast/ threat. 26 “Boy, you are ignorant. You don’t realize who I am and what my lineage is. I rule over these lands and receive tribute from others. For your affront, you owe me compensation” (my translation). The term truage (“tribute”), used by both interlocutors, is a key word in this laisse. <?page no="53"?> Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 35 under the one heading “salutation/ establishment of identity” of the interlocutors. In Bax’s characterization of the medieval Request-for-Information sequence (1981: 426-430), these introductory conversational moves are an optional feature of the verbal interaction that leads medieval knights to fight. Salutations do not always precede the request for information, with which some knights initiate the exchange. The English KA is a case in point, since Nicholas’s opening conversational move when he meets Alisaunder wandering on the shore seems from its interrogative mode to be a request for information, not unlike Nicolaus of Acarnania’s query “Why are you here? As a spectator or as a competitor? ” in GAR (18, 50). In fact, in the Middle English romance Nicholas’s query (“Who authorized you to be here? ”) 27 is at once an affirmation (‘You have no right to be here’) and a threat (‘Get out of here, or else…! ’): Who 3af þee leue here to gon? Now quyk do þee hennes sone, For þou ne hast nou3th here to done. (KA, 870-872) Whereas in the earlier versions of this episode, Alexander’s reply is to announce that he has come to compete with Nicholas in the chariot-race, provoking some incredulous scoffing, 28 the young philosopher-knight of KA contents himself with a provocative non-verbal gesture of indifference: Alisaundre loked a-skof/ As he ne had y 3 oue þere-of (873-874). 29 This in turn causes an enraged Nicholas to fling a number of verbal insults at Alisaunder, who is called a misbi 3 ete gome (877: “misbegotten chap” is possibly a reference to his adulterous conception) and a vyle ateynt hores sone (879: “vile, tainted son of a whore” is unmistakably a slur on his bastard status). 30 The flurry of verbal violence - which raises the issue of legitimacy only in the Middle English romance - culminates in a request for compensation, or tribute (wed, 881), inciting young Alisaunder to speak up in resolute defiance. In what follows, it is tempting to recognize Bax’s move of “refusal 27 The question is of course also a rhetorical one, implying the affirmation “I, who alone decides who walks here, didn’t give you my authorization.” 28 Nicolaus tells Alexander that he should participate in the less prestigious games of wrestling or boxing (GAR, 18, p. 50). For a demonstration that horse-racing was the preserve of older competitors and for a sociological analysis of differences of age and class involved in these competitions, see Golden (1998: 104-175). 29 “Alisaunder glanced mockingly, as if he couldn’t care less.” Translations from KA are my own. 30 Strangely, ateynt is not glossed by Smithers, although this use of the participial adjective of French origin may well precede its first attested use mentioned by the Oxford English Dictionary, which cites Robert Mannyng’s Handlyng Synne. It is not cited in connection with a verbal insult before Sir Launfal, written roughly a century after KA. In fact, there is no feature of the verbal violence in this episode that incites any commentary from this distinguished editor of KA. <?page no="54"?> Margaret Bridges 36 and challenge” with which the medieval Dutch knights of his corpus respond to the request for information accompanied by a threat (1981: 427). Nicholas’s initial challenge is met by a counter-challenge (‘no, it is I who will receive tribute from you, and gain possession of the land you claim to rule over’, KA 882-886) and a counter threat (887-888: ‘it would be best for you to flee’). This in turn is followed by more verbal violence from Nicholas, who threatens to have Alisaunder hanged, drawn, flayed and burned (889-895), conjuring up the staging of a public execution that is unique to the Middle English text and that is just one of the many instances of the poet’s construction of a new ‘popular’ context for the ‘courtly’ text he is translating. Contrasting with this creation of a new context for an old text and the English poet’s concomitant capacity for lexical innovation is the stability of the non-verbal gesture of spitting in which this exchange of insults culminates throughout the roughly thousand years of its transmission. This cannot be the place to develop a cultural, linguistic or anthropological history of spitting. 31 Suffice it to note that, for all their differences, GAR, the Epitome, RTC and KA represent Nicholas flinging not only insults but also spittle in the hero’s face. Whereas they concur in representing the spittle not only as highly insulting but also as tantamount to a curse or an illicit punch, they differ slightly in their reporting of the hero’s response. The Greek and Latin texts seem to be emphasizing an opposition between the uncontrolled emotions of the spitter (who, overcome with anger, curses his rival) and the precocious self-control of the philosopher-hero (who smilingly wipes the insulting spittle from his face); both texts moreover explicitly comment on Alexander’s ability to contain his feelings - whether this is said to come naturally (as in GAR 18, 50) 32 or whether it is seen as the result of his training (as in Epitome 18, 22). These comments are subsequently lost in translation, even though RTC reports Alexander’s silent smile (Alisandre s’en test si rist de cel hontage, 592) 33 before going on to report his resolution - involving the swearing of an oath - to kill Nicholas in combat. Alisaunder’s response to being spat upon is not directly reported by the KA narrator, who assures his audience that Nicholas would have paid dearly for his spitting, had he not fled from the scene (896-898). In so doing, he is in fact suggesting that Alisaunder is not repressing but 31 Joachim Knuf (1992: 466-482) has studied the communicative use of spitting as part of a more general semiotic investigation of the relation between linguistic and nonlinguistic sign systems, emphasizing the greater potential of non-linguistic signs for ritualizing the communicative situation. 32 Stoneman translates the gamma version of GAR as saying that it came naturally to Alexander to control his feelings (18, 50), whereas Jouanno (2009: 58) translates the same passage as saying that Alexander had learned to dominate his nature. 33 “Alexander is silent and laughingly shrugs off this insult.” <?page no="55"?> Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 37 expressing his licit anger. 34 What was once an opposition between selfdefeating irascibility and empowering self-control has become in the hands of the Middle English poet a quarrel between two equally angry young men, who are coupled through expressions like eiþer to oþere and boþe, and whose verbal violence has functioned as a ritual prelude to a battle that both are eager to engage in: Alisaunder was sore awhaped that he [Nicholas] was so sone ascaped, And suore he shulde sore abugge, And his hede for that gilt legge [….] Her eiþer to oþere ost is went, Ful of yre and mautalent. Þe ni3th þai resteþ litel, forsoþe, Bot as men that ben wroþe. Wel warded thai weren boþe that ni3th, Al forto spronge þe dayes li3th. (899-910) 35 Resolution/ Conclusion The verbal and non-verbal insults exchanged by the young ‘heroes on the beach’ concluded with their temporary separation and with Alexander’s resolve to defeat his opponent, whether in the competition (GAR and Epitome) or in battle (RTC and KA). Unlike the chariot race, in which Alexander and Nicolaus would have been competitors even without their preliminary verbal skirmishes, the battle opposing the armies of Alexander and Nicholas was induced by their exchange of insults: both the Anglo- Norman and the Middle English poems represent Alexander addressing Nicholas’s severed head as having received its due reward for the insults it proffered the day before. 36 Only the later medieval texts, therefore, establish a causal relationship between the flyting and the fighting, themselves embedded in the need to legitimize the rule of the freshly dubbed bastard 34 Moreover, after he has won the battle, Alisaunder specifically links his resolution to decapitate Nicholas with the insult of spitting (976-977). 35 “Alisaunder was immensely troubled that he [N] had made such a rapid escape, and swore that he would pay sorely for his offence, with his head. [….] Each of them went back to his own army, full of anger and resentment. At night they had little rest, to be sure, for they were but angry men, and they both spent that night on their guard, waiting for day to break.” 36 If Alexander’s victory in this episode was assimilated by its Anglo-Norman and English translators to a victory in a medieval judicial duel, then it would have made more sense to foreground the truth-conditional insults relating to bastardy than to foreground the gesture of spitting. <?page no="56"?> Margaret Bridges 38 heir to the throne of Macedon. Given the stated objectives of this study, I am here less concerned with the form taken by the resolution of the conflict (Nicholas’s defeat and death; Alexander’s crown of the charioteer’s laurels or of chivalric conquest) than with the verbal testing of Alexander that has led to that conflict. It does seem worth noting, however, that some spectacular prose is devoted to the race itself in GAR, and that both RTC (35 lines) and KA (72 lines) seem to relish the opportunity for describing the hue and cry of battle and its aftermath. The laconic affirmation of Epitome to the effect that Alexander returned home triumphant having defeated Nicolaus and the others 37 is surely an indication of the fact that the specific historical community for which this ninth-century text was produced was no longer able to appreciate the nature of the contest opposing Alexander and Nicolaus, the tethrippon or four-horse chariot race. It is hardly surprising that a gradual process of acculturation should have led later medieval writers to imagine the contest developing out of the kinds of verbal interaction that marked the agonistic encounters of their epic and romance heroes. Or is it? This is a question that we must address by way of conclusion. On the one hand we have been able to observe in this episode (the testing of Alexander) a number of features that undergo modification as a result of the cultural changes marking its long process of transmission. Yet we have also observed that the conversational moves and their sequence remain surprisingly stable across contexts as diverse as those of post-Hellenistic Greece, ninth-century Latinitas, the Anglo-Norman court of Henry II 38 and urban London around 1300. This stability seems to defy the historical pragmatist’s conviction that we can only describe specific speech conventions in specific historical communicative communities. In the case of the texts in question, not much is known about the specific historical communicative communities that produced and received them, beyond the (not incontrovertible) ways they are constructed in the works themselves. This is especially true of the Middle English Kyng Alisaunder, which deserves to occupy a pole position in cultural translation studies by virtue of its pervasive linguistic and literary contact with Anglo- Norman French, which it both reflects and distances itself from, as the analysis of a single episode relating to the hero’s testing has intimated, if not conclusively demonstrated. The scholarly debate over the universality or relativity of agonistic behaviour and language in literature is likely to remain a debate without resolution. 37 Non multo post, ubi dies certaminis advenit, Alexander primum Nicolao superato et reliquis qui certamini adfuerant, corona redimitus repatriat Macedoniam (19, 23). 38 This has been postulated as the milieu for which Thomas - the only author of our corpus to have inscribed himself in his text - wrote his Roman (Gaullier-Bougassas and Harf-Lancner 2003: xx). <?page no="57"?> Cultural Translation and the Agonistic Vocabulary of Kyng Alisaunder 39 Biographical sketch Margaret Bridges recently retired as professor of Medieval English at the University of Berne, where she still co-directs a graduate programme on “Gender Scripts and Transcripts”, supported by the Swiss National Science Foundation. She is currently responsible for the field of English literature in a five-year interdisciplinary research project, based in Lille, on mythmaking processes as evidenced by medieval European representations of Alexander the Great between the eleventh and the sixteenth centuries. She also is one of the general editors of a new series on Alexander redivivus being launched by Brepols. Recent publications have been in the fields of Gender Theory (Gender Scripts, 2009), Chaucerian reception of the classical topos of the abandoned woman addressing her bed (in Mythes à la cour, Mythes pour la cour, 2010) and Early Modern English trading vocabularies (in Fiction and Economy, 2007). Publication projects, in addition to work on aspects of cultural translation of the Alexander myth, include figurations of the ‘Embedded Author’ as well as a study of ‘Suicide’ in Medieval Literature. References Primary Texts Anon. 1996. The Anglo-Saxon Chronicle: A Collaborative Edition, Volume 6. MS D. (Edited by G. P. Cubbin). Cambridge: Brewer. Anon. Bevis of Hampton. 1999. (Edited by Ronald B. Herzman, Graham Drake, and Eve Salisbury). Kalamazoo, Michigan: Medieval Institute Publications. Anon. 1867. Julii Valerii Epitome. (Edited by Julius Zacher). Halle: Buchhandlung des Waisenhauses. Anon. The Greek Alexander Romance. 1991. (Translated from Greek by Richard Stoneman). Harmondsworth: Penguin. Anon. Historia de Preliis J1. 1991. In The Romances of Alexander. (Translated from Latin by Dennis M. Kratz). New York: Garland, pp. 1-82. Anon. 1952-1957. Kyng Alisaunder. (Edited by G. V. Smithers). Early English Text Society O.S. 227 and 237. London: Oxford University Press, 1952 (volume I) and 1957 (volume II). Iulius Valerius ‘Alexander Polemius’. 2010. Roman d’Alexandre. (Translated from Latin and edited by J.-P. Callu). Turnhout: Brepols. Langtoft, Pierre. 1866. The Chronicle of Pierre de Langtoft. In French Verse from the Earliest Period to the Death of King Edward I. (Edited and translated from Anglo-Norman by Peter Wright). London: Longmans. Mannyng, Robert. 1996. Robert Mannyng of Brunne: The Chronicle. (Edited by Idelle Sullens). Medieval & Renaissance Texts & Studies 153. Binghamton, N.Y.: MRTS. <?page no="58"?> Margaret Bridges 40 Pseudo-Callisthenes. L’Histoire merveilleuse du roi Alexandre maître du monde. 2009. (Translated from Greek by Corinne Jouanno). Toulouse: Anacharsis. Thomas of Kent. 2003. Le Roman d’Alexandre ou le Roman de toute chevalerie. (Translated from Anglo-Norman and edited by Catherine Gaullier-Bougassas and Laurence Harf-Lancner). Paris: Champion. Criticism Bax, Marcel. 1981. “Rules for Ritual Challenges: A Speech Convention Among Medieval Knights”. Journal of Pragmatics 5, pp. 423-444. Butterfield, Ardis. 2009. “Chaucerian Vernaculars”. Studies in the Age of Chaucer 31, pp. 25-51. Dorfman, Eugene. 1969. The Narreme in the Medieval Romance Epic: An Introduction to Narrative Structures. Toronto: University of Toronto Press. Fischer, Andreas. 1986. Engagement, Wedding and Marriage in Old English. Heidelberg: Carl Winter Universitätsverlag. Given-Wilson, Chris and Alice Curteis. 1984. The Royal Bastards of Medieval England. London and New York: Routledge. Golden, Mark. 1998. Sport and Society in Ancient Greece. Cambridge: Cambridge University Press. Golden, Mark. 2003. Sport in the Ancient World from A to Z. London: Routledge. Jucker, Andreas H. 2000. “Slanders, Slurs and Insults on the Road to Canterbury. Forms of Verbal Aggression in Chaucer’s ‘Canterbury Tales’”. In Irma Taavitsainen, Terttu Nevalainen, Päivi Pahta and Matti Rissanen (eds.). Placing Middle English in Context. Berlin: Mouton de Gruyter, pp. 369-389. Knuf, Joachim. 1992. “‘Spit First and Then Say What You Want! ’ Concerning the Use of Language and Ancillary Codes in Ritualized Communication”. Quarterly Journal of Speech 78/ 4, pp. 466-482. Lagorgette, Dominique. 1994. “Termes d’adresse, acte perlocutoire et insultes : la violence verbale dans quelques textes des XIV e , XV e et XVI e siècles”. La violence dans le monde médiéval. Senefiance 36, pp. 317-332. Parks, Ward. 1986. “Flyting, Sounding, Debate: Three Verbal Contest Genres”. Poetics Today 7/ 3, pp. 439-458. Parks, Ward. 1990. Verbal Dueling in Heroic Narrative: The Homeric and Old English Traditions. Princeton: Princeton University Press. Roberts, Jane, Christian Kay and Lynne Grundy (eds.). 1995. A Thesaurus of Old English. London: King’s College. <?page no="59"?> Hans-Jürgen Diller Discourse, matrix and genre in Historical Semantics: ofermōd and emotion Abstract The usefulness of two categories, discourse and genre, for historical semantics is explored. After a plea for a rather narrow definition of discourse (Foucault 1972: 48 [French original 1969: 67]: “discourses create the objects of which they speak”) and a rather loose definition of genre (Biber 1989: 5-6: “text categories readily distinguished by mature speakers”), it is claimed that discourse, supplemented by Hacking’s (2000) concept of matrix, is useful diagnostically, whereas the usefulness of genres is primarily heuristic. The claim is substantiated in two exemplary analyses: the first brings the concepts of discourse and matrix to bear on the much-debated meaning of ofermōd in the Old English Battle of Maldon, while the second demonstrates, against Dixon (2003), a nonor pre-discursive use of emotion. 1. Introduction, mainly on discourse and matrix It is a commonplace that meaning depends to a large extent on context, both linguistic and extra-linguistic. Among contexts, the kinds of text in which linguistic units tend to occur form a special group, and it is on these that this paper will concentrate (‘linguistic units’ includes both lexemes and constructions). The two most prominent ‘kinds of kind of’ text are discourse and genre. Discourse is more fashionable but has become “wideranging and slippery” (Taylor 2001: 8), and “it is hard to see how we could avoid the proliferation of discourses until there were as many as words in the dictionary” (Burr 2003: 175). And there is no sign that proliferation will stop there! We find, for instance, “[d]iscourses [plural! ] of ‘lone’ motherhood”, including a “counter-discourse” (Carabine 2001: 271), Burr (2003: 65) refers to “fox-hunting discourses”, which she sub-divides into a “pro-” and an “anti-fox-hunting discourse”. Such distinctions may be overly delicate, but they highlight an important point which some German scholars tend to neglect, perhaps owing to the influence of Habermas. In the French and Anglo-Saxon traditions discours and discourse are very different from discussion. A discourse is not just a publicly conducted debate with opposing opinions, but a set of statements or assertions which support each other. Habermas (German original <?page no="60"?> Hans-Jürgen Diller 42 1981) conceives of Diskurs as a negotiation of concept and their definitions, where the best argument will (or at any rate should) win and where power (ideally) plays no part. In this respect, Foucault’s discours is the direct opposite. It exercises power, and the individuals participating in it (who empirically do make it into a debate) are of little or no account (Hacking 2004). In this paper I shall argue that discourse is potentially a powerful diagnostic category in historical semantics while genre is, above all, heuristically useful. At the same time, I will plead for a more rigorous definition of discourse and a highly flexible definition of genre, as proposed by Biber (1989: 5-6): genres are “text categories readily distinguished by mature speakers.” I will illustrate my argument with two vastly different and historically distant examples: Byhrtnoth’s ofermōd in the Old English Battle of Maldon, and the establishment of emotion (as opposed to passion) in the post-medieval period. A moment’s reflection will tell us that Foucault’s notion of discours is highly intractable material for historical semantics. The primary material used by historical semanticists will be words and word-groups in texts, as they are now available in large computer-readable corpora, a fair number of them in the public domain. If you are interested in the denotations or connotations of words like fox, vixen, hunt, etc., you will be equally interested in the pro-fox-hunting discourse and in the anti-fox-hunting discourse. You may even look at texts or utterances which take no position pro or anti. Above all, you can take your evidence from a corpus which has been compiled with no view to either proor anti-fox-hunting. This is not how the discourse analyst will, or even can, proceed (cf. Busse and Teubert 1994: 14-18; Maset 2002: 195f.). The discourse analyst has to start with a “pre-understanding” (Gadamer, German original 1960) of where the meanings which he or she wants to study are most likely to be created. Discourse analysts will also work with corpora, but they will usually compile their corpora ad hoc, from textual universes which they expect to be rewarding. A highly popular source is media texts and interviews. From what has been said so far discourse analysis does not look like a promising method for historical semantics. But there is one aspect of discourse which the historical semanticist cannot ignore without serious risk. That aspect is part of Foucault’s definition of discours and has nothing to do with Habermas. It is also relatively unimportant in the Critical Discourse Analysis which is associated with names like Norman Fairclough (1989), Teun van Dijk (1998), and Siegfried Jäger (1999). <?page no="61"?> Discourse, matrix and genre in Historical Semantics: ofermōd and emotion 43 According to Foucault, one of the important effects of discourses is to “create the objects of which they speak” (1972: 48, French original 1969). 1 A good example of the ‘object-creating’ power of discourse is La loi sur les aliénés of 1838, 2 which describes not so much the criteria by which an insane person is defined as the procedure which turns an individual into a member of the class or category of insane persons. Together with the procedures which it describes and prescribes, la loi creates a social space where acts become possible which we may call acts of ‘performative categorization’: if, at the end of the appropriate procedure, a person is declared insane, his or her insanity is a social fact. A disturbing consequence of a purely procedural-performative definition is that it does not provide for the possibility of psychiatry being abused for illegitimate purposes: when a person is declared insane the question whether he or she is ‘really’ insane cannot even be asked under such a definition. Although not referring directly to la loi, Foucault (1965: 269) deplores the “apotheosis of the medical personage” which is implied in the psychiatric thinking of its time. Performative categorization is closely associated with what is often called ‘social construction’. In my favourite field, the emotions, there is an important school called Social Constructionism (SC) which claims that emotions are not “invariable physiological occurrences” but “socially constructed” “discursive practices” (Stedman 2002: 7; see above all Harré 1986, Lutz 1988, Abu-Lughod and Lutz 1990). Like discourse, SC has become an indiscriminately used passe-partout term which is losing precision by over-use. The most lucid analysis of the (limited) validity of SC is Hacking (2000). Hacking, a philosopher at the University of Toronto, stresses a few facts which all too often are side-lined by other practitioners of discourse analysis. (1) The objects created by discourses are not individual objects or events existing in space and time, but categories or, as Hacking prefers to say, kinds. (2) Turning individual people or acts into members of kinds requires not only the linguistic practices which we primarily associate with discourses but also authorized institutions and procedures which give a discourse its proper form. Hacking calls this ‘matrix’ rather than ‘discourse’. To illustrate, I will concentrate on one of the many examples of SC given by Hacking: the woman refugee. A woman refugee is first of all a woman who has fled her country because she found conditions unbearable. But to be recognized as a woman refugee in Canada she has to convince the relevant authorities in a prescribed procedure that those conditions are indeed unbearable by Canada’s standards. In Hacking’s own words: 1 For a convenient bibliography of Foucault (both French originals and English translations) see http: / / en.wikipedia.org/ wiki/ Bibliography_of_Michel_Foucault (last visited 30 November 2010). 2 See http: / / www.ch-charcot56.fr/ textes/ l1838-7443.htm (last visited 30 November 2010). <?page no="62"?> Hans-Jürgen Diller 44 The matrix in which the woman refugee is formed is a complex of institutions, advocates, newspaper articles, lawyers, court decisions, immigration proceedings. Not to mention the material infrastructure, barriers, passports, uniforms, counters at airports, detention centers, courthouses, holiday camps for refugee children. You may want to call these social because their meanings are what matter to us, but they are material, and in their materiality make substantial differences to people. (2000: 10) The ‘material infrastructure’ makes it clear that the matrix isn’t just words and linguistic practices. It emphatically includes the extra-linguistic context. I suppose the list is not meant to be exhaustive. Still, I find it a little strange that it should not contain something like an ‘authoritative text’, a law or statute regulating the status of the woman refugee and the procedures by which she can be recognized as one. It will be noticed immediately that there is an important and, I believe, necessary ambiguity in the meaning of the term: it designates two different categories. To understand the word, we have to know whether its meaning depends on a discourse (as Foucault would say) or on a matrix (as Hacking would say) - or whether it occurs in ordinary, unregimented language. Discourse and matrix, we may say, ‘officialize’ a category. This, I submit, is a frequent problem in textual interpretation, and it makes a distinction necessary which I have not yet seen in the semantic literature: that between the ‘discursive’ and the ‘intuitive’ use of a word. Intuitively the woman will be a woman refugee to everybody who has heard and believes her story. Discursively, she will be a woman refugee only after she has successfully undergone the appropriate procedure. 3 English and German soccer fans will immediately think of a much handier example: Lampard’s goal against Germany during the World Cup in South Africa in 2010 was a goal intuitively, but not discursively. The woman refugee and Lampard’s goal have one important and uncomfortable thing in common: the distinction between discursive and intuitive is not so much one between words as between our use of words. The decision whether a word is used to designate a ‘discursive’ or an ‘intuitive’ category is not at all self-evident. 2. Ofermōd and its matrix Bythnoth’s ofermōd in The Battle of Maldon, line 89, one of the most discussed cruxes in historical semantics, is an eloquent witness to this claim. 3 I borrow the distinction from Murdoch (1961: 1). The meaning of intuitive is related to the philosophical use of intuition (OED, sense 5) which int. al. translates Kant’s Anschauung. <?page no="63"?> Discourse, matrix and genre in Historical Semantics: ofermōd and emotion 45 As will be remembered, it is ofermōd which makes the English leader yield “too much land” to the Vikings who have landed on a small islet off the coast near Maldon. Gneuss (1976) gives a very full survey both of the translations of the word by modern scholars and of its use in Old English texts. His conclusion was widely accepted for a long time: in every other case ofermōd clearly means ‘pride’ or ‘proud’, to assume a different meaning in the case of Maldon 89 would be literally special pleading. Gneuss admits that the other occurrences come from religious writings, whereas Maldon, while certainly a Christian poem, is not a religious one. In Gneuss’s view, however, this is not a compelling counter-argument. In its time Gneuss’s interpretation found general assent. 4 Halbrooks (2003: 236), however, finds this agreement “extraordinary considering Gneuss’s own guardedness concerning his conclusions.” Gneuss’s view might be called a ‘weak genre’ view of historical semantics: it accepts that the genre context can make a difference to the meaning of a word (cf. Diller 1986; Diller 2009 for some Old English examples), but when a word is frequent in some genres and rare in others, it would be unwise to postulate a separate sense or meaning for the genre(s) with few occurrences - even if some “semantic discomfort” (Lewis 1967: 1) remains. The weak genre approach may be heuristically sound, but it is insufficient to settle a doubtful point of interpretation. Against this approach I would ask, in the spirit of the ‘discourse’ or ‘matrix’ approach, not: how should we translate ofermōd, but: what kind of thing is it, and what makes it the kind of thing it is? The answer is not as clear as in the case of the recognized woman refugee or the given goal. Recognizing a woman refugee and giving a goal are performative utterances or acts which are possible by virtue of the matrix within which they take place. A person becomes a woman refugee and a goal becomes a goal because a person in authority has declared them such in an appropriate situation, or more technically, because the respective acts are ‘felicitous’ (Austin 1962). They are felicitous because the corresponding discourse has created the categories which they apply. As the juridical discourse and the discourse of soccer create the categories of refugee woman and goal, so the discourse of Christian ethics has created the categories of vice and sin, enabling persons in authority to subsume individual acts or attitudes under these categories and thus to create social facts. Readers of The Battle of Maldon are faced with the question: is ofermōd in line 89 used in a sense that is germane to that discourse? Ofermōd is a frequent translation of Latin superbia, a core concept in the discourse of traditional Christian ethics. That discourse also has kindmaking powers. More accurately: the ethical discourse of the Roman 4 See Halbrooks (2003) for a survey of scholarship after Gneuss. <?page no="64"?> Hans-Jürgen Diller 46 Catholic church is embedded in a matrix which enables performative acts very much like giving a goal or recognizing a woman refugee. Important components of the matrix are the office of the priest, the physical structure and location of the confessional, the ritual and penances which the penitents have to accept and which categorize them as sinners - within a subtly graded hierarchy of sins and sinners. Evidence of this matrix is to be found, above all, in the penitential literature. 5 The penances and the authority to impose them are anchored in an ensemble of texts which form the discourse of traditional Christian ethics in a narrower sense and in which concepts like vice and sin are defined. Prominent in this discourse are writers like ‘Saint’ John Cassian (ca. 360-435) with his De institutis coenobiorum et de octo principalium vitiorum remediis (420x429) 6 and Pope Gregory I, the Great, with his Moralia in Iob (591 at the latest; cf. Greschat 2005: 26). The debate about the sins has been treated very fully by a literary historian (Bloomfield 1952). Bloomfield wrote of course long before our notions of discourse or matrix were developed, his subtitle refers to a ‘religious concept’, and the book describes various traditions articulating that concept. It now has to be shown that the concepts of discourse and matrix can contribute something to the understanding of Maldon 89 that Bloomfield’s ‘religious concept’ cannot contribute. First of all, there is the distinction between ‘vice’ and ‘sin’. That distinction is not applied consistently in the medieval discourse on sin but, simplifying somewhat, we may say that a vice is a mental condition, while a sin is an individual act prompted by that condition in a given situation. The practice reflected in the penitentials is primarily concerned with sins, but the penalties which they impose require of course a grounding in the vices as defined in the theoretical discourse. Theologically, a vice is an ailment of the soul. The remedia against these ailments occur already in the title of John Cassian’s work (see above), and ch. xxvii of Book XII is headed “Expositio vitiorum, quae per morbum superbiæ generantur.” [“A description of the faults which spring from the sickness of pride.”] 7 John Cassian treats superbia in the last, twelfth book of De institutis but he insists that the spirit of pride is the beginning of all sin (“origine tamen et tempore primus est” [PL 45, coll. 421f.]). In the same 5 See, int. al., Wasserschleben (1851, repr. 1958), McNeill and Gamer (1938), Bieler (1963), and Frantzen (1983). 6 “Although never formally canonized, St. Gregory the Great regarded him as a saint.” (Catholic Encyclopedia, loc. cit. http: / / www.newadvent.org/ cathen/ 03404a.htm, last visited 21 November 2010). The x between the figures means that the work was composed some time between 420 and 429; a hyphen, by contrast, would mean that its composition was begun in 420 and ended in 429. 7 De institutis and its English translation are best accessed through http: / / en.wikipedia.org/ wiki/ John_Cassian (last visited 20 November 2010). <?page no="65"?> Discourse, matrix and genre in Historical Semantics: ofermōd and emotion 47 vein Gregory, quoting Scripture, defines superbia as the root of all evil: “Radix quippe cuncti mali superbia est, de qua, scriptura attestante, dicitur: Initium omnis peccati superbia (Eccl[esiasticus] 10, 15).” [“The root of all evil is pride, of which it is said by witness of Scripture: For pride is the beginning of sin.” (Sirach 10, 13)] 8 In the Vulgate this verse is preceded by verse 14: “Initium superbiae hominis apostatare a deo.” [“The beginning of pride is when one departeth from God.” (Sirach 10, 12)], which is a constant theme with countless variations in the Moralia (Baasten 1986, esp. ch. 1). It will also be important for the understanding of Maldon 89. The archetype of superbia is Lucifer, the fourth chapter of Cassian’s twelfth book is headed: “Quod ob superbiam Lucifer ille de archangelo factus est diabolus.” (PL p. 187, coll. 425ff. [How by reason of pride Lucifer was turned from an archangel into a devil.]) Lucifer’s rebellion as described in the Old English poem Genesis shares many features with John Cassian’s account. It is clearly a ‘departing from God’: God realizes that Lucifer is ofermōd (adj.) in l. 262, and in ll. 271f. he is described either as the “angel of ofermōd” or as uttering “words of ofermōd”. 9 In those words he claims to have more power and strength than God (ll. 269-70). He also plans to build himself a throne which will be higher than God’s (ll. 273-4). The Old English Genesis is not discourse in the sense that a category is ‘created’. But the narrative illustrates the essential features by which the ethical discourse of the medieval Church defines that category. It is an important point in what Bloomfield calls a tradition and, we may add in passing, it is a classic example of that ‘status violation’ which the new science of the sociology of the emotions regards as a human universal and as the chief antecedent of wrath (Kemper 2006): by raising himself above God he would violate God’s ‘status’. 10 Human superbia will appear on a smaller scale than Lucifer’s, but both share the essential feature as identified by Gregory: “[…] pride is a dangerous elation […] in which the proud fail to recognize their true relationship with God or other men. In regard to God, they consider themselves to be on the same level with God, instead of ‘under him’” (Baasten 1986: 19). Turning now to The Battle of Maldon, it is clear that Byrhtnoth is not guilty of this “dangerous elation”. The speech which is prompted by his 8 The transmission history of Ecclesiasticus is extremely complex. Because the medieval church knew only of a Greek not a Hebrew version, it is regarded as ‘apocryphal’ by the churches of the Reformation; to the Roman Catholic church it is ‘deuterocanonical’. Since 1896 more and more portions of the Hebrew original have come to light (Deissler and Vögtle 1985: 947). The Latin quotations are from Colunga and Turrado (1959), English translations from the Polyglot Bible. Titles like “Sirach”, or “Jesus (Ben) Sirach” are due to an ascription in the Epilogue. 9 “feala worda gespræc / se engel ofermodes.” (Doane 1991: 209 and commentary on p. 260). All line references are to this edition. 10 For a more extensive discussion of status violation see Diller (forthcoming). <?page no="66"?> Hans-Jürgen Diller 48 ofermōd could even be read as an intended contrast to Lucifer’s illusion of self-sufficiency: ‘Nu eow is gerymed: gað ricene to us guman to guþe. God ana wat hwa þære wælstowe wealdan mote.’ (Scragg 1981, lines 93-95) ‘Now a path is opened for you: come quickly against us, men at war. God alone knows who will control the battlefield.’ (Scragg 2006: 143) We should also note that status violation, which is the defining feature of superbia in the medieval discourse on sin, plays no part in The Battle of Maldon. The question simply does not arise between military commanders one of whom has landed in the territory which the other has to defend. Even more important is the fact that not only the penitential matrix is absent from the poem but that its absence is not even treated as a problem. Dying in combat, Byrhtnoth has no chance of confession and penitence, and in the entire fragment there is no hint that this might damage his prospects in life hereafter. If the poet had meant to say that the orders given by Byrhtnoth for his ofermōde were motivated by superbia in the discursive sense, the dying prayer would probably have been very different. Instead of thanking God for “ealra þæra wynna þe ic on worulde gebad” (l. 174; [all the joys that I have experienced in this world], Scragg 2006: 145) and trustfully asking for protection against helsceaðan (=thieves from hell) Byrhtnoth could be expected to ask forgiveness for that specific act. Byrhtnoth’s ofermōd and the ofermōd ‘created’ in the penitential matrix are two very different things and we should think at least twice before we apply the ‘discursive’ meaning to the word as it is used in Maldon. To borrow a phrase from German juridical discourse: with a probability bordering on certainty, it is not applicable. The action prompted by Byrhtnoth’s ofermōd is “unwise” (Scragg 2006: 153) and “posthumously criticized by our poet” (Gneuss 1976: 133), but it is not sinful. More colloquially, ofermōd in Maldon is too much of a good thing, not a bad thing in itself. In this (important) respect it differs from apparently all other instances of ofermōd. 3. Emotion and its genres The previous section was an attempt to use the categories of discourse and matrix in the elucidation of a semantic crux, i.e. the interpretation of a word occurring in a specific context. The existence of the relevant discourse <?page no="67"?> Discourse, matrix and genre in Historical Semantics: ofermōd and emotion 49 with its matrix was taken as given, the question was: does the word occurrence have a place in that matrix? This section will present data from the pre-history of a discourse which has been insightfully described by a theologian. Dixon (2003) has published a book with a telling title: From Passions to Emotions. The Creation of a Secular Psychological Category. Right at the beginning he claims that “the emotions did not exist until just under two hundred years ago” (Dixon 2003: 1). The evidence for his claim he finds in ‘psychological discourse’, which he defines as the “tradition of systematic thought about mental life” (Dixon 2003: 12). The ‘kind-making’ power of discourse is assumed rather than demonstrated in Dixon’s account; since performative categorization plays no part with him, we may say that he is using the concept of discourse in a weak sense. I will return very briefly to this problem in Section 4. Since he does not look for evidence outside that discourse it seems natural that it should have ‘created’ the emotions. That claim feeds what has been called the “fallacy of autonomous discourse” (Dreyfus and Rabinow 1982). The fallacy is at best a “professional oversimplification” of social and intellectual historians (Richter 1996: 11). It is even pernicious because it mistakes the conventional for the arbitrary, a mistake which was castigated by Karl Popper as early as 1945 (2005: 70) and, following him, by F. A. Hayek (1970). The category ‘emotion’ was not created; it emerged, and it did not emerge in psychological discourse as defined by Dixon, or at least not there alone. 11 To counter Dixon’s claim we have to look at a massive and messy body of data which can be given some little structure only by a highly fuzzy category like ‘genre’. Biber’s definition, quoted above, underlies the divisions of virtually all language corpora; it clearly does not refer only to literary genres. Emotion does not appear in the 17 th century part of ARCHER, 12 and Chadwyck-Healey’s Full-Text English Poetry Database returns only one 17 th century example, which uses the word in its earlier, socio-political sense. 13 A certainly unexhaustive and perhaps unrepresentative search of texts in Gutenberg, Wikisource and the Online Books Page yielded only isolated examples in Samuel Pepys (1), John Evelyn (1), Aphra Behn (4), and John Locke (2). 14 A slightly richer harvest is to be reaped from the drama section 11 Thomas Dixon has kindly drawn my attention to a more recent publication (Dixon 2006), in which he stresses the influence of medicine in establishing the category ‘emotion’. 12 For a description of ARCHER see Biber, Finegan and Atkinson (1994); Biber (2001). 13 In a poem by William Herbert dated 1604. 14 S. Pepys, Diary, 17 August 1660 (sense of vomiting) http: / / www.gutenberg.org/ files/ 4200/ 4200.txt; J. Evelyn, Diary, 31 May 1672, http: / / www.archive.org/ details/ diary- <?page no="68"?> Hans-Jürgen Diller 50 of LION, 15 which has eleven 17 th century plays using the word, giving a total of 13 occurrences, mainly but not exclusively in the psychological sense. Compared with the 132 instances of passion in the same plays, that is still modest. But even these few examples show two things: in spite of its comparative rarity emotion is well established in non-specialized language; and emotion and passion are by no means synonymous, most of the time the substitution of passion for emotion would have produced extremely awkward results. Emotion can often be replaced by words like apprehension, fear, or sorrow, not by passion. So far, I can offer only slight modifications to Dixon’s claim. The word emotion is rare everywhere, far rarer than passion. A very different picture emerges when we look at 17 th and 18 th century newspapers, which are accessible in the 17 th -18 th Century Burney Collection of Newspapers. Table 1 shows that, taking the period as a whole, there is an almost perfect balance between emotion and passion. But until 1786 there are far more uses of emotion than of passion: emotion passion before 1701 2 1 1701-10 11 0 1711-20 35 0 1721-30 60 18 1731-40 77 96 1741-50 49 23 1751-60 165 28 1761-70 228 19 1771-80 396 64 1781-90 703 566 1791-1800 1251 2201 2977 3016 Table 1: 17 th -18 th Century Burney Collection of Newspapers johnevelyn04dobsgoog; A. Behn, Sir Patient Fancy I, i (Works IV, 14), Agnes de Castro (Works V, 238, 248), The Dumb Virgin (Works V, 437); John Locke, Some Thoughts Concerning Education (1693), sect. 8, 115 (= Works IX, pp. 12, 109); http: / / socserv2. mcmaster.ca/ ~econ/ ugcm/ 3ll3/ locke/ index.html). 15 LIterature ONline http: / / lion.chadwyck.co.uk/ (last visited 26 November 2010). <?page no="69"?> Discourse, matrix and genre in Historical Semantics: ofermōd and emotion 51 The swing seems due to the launching of a single new newspaper in 1787, The World, 16 which in its opening issue promises a lighter, more entertaining tone than its older rivals. The number of occurrences is far too large for a detailed analysis, especially as the optical character recognition is not always flawless. Spurious and multiple attestations would have to be eliminated before the data can be interpreted. This was done only for the years before 1711. Interestingly, the only record of passion in these years is from a murder trial report; the murder in question was committed “in the midst of” a passion. 17 Emotion is commonly used in the context of verbal communication, which is well illustrated by the very first example: “I love to write without any Emotion, and as we say in cold blood.” 18 Violent action is not associated with the emotions. Against Dixon’s view of emotion as a secular category another quotation is also of interest: “Religion is of that admirable frame and temper, it enflames us with a true Emotion to our great maker […].” 19 This picture is re-enforced by another collection, the Eighteenth-Century Collections Online (ECCO), which has its own genre division and shows the following uses of emotion: Subject areas N size of area (% of files) 20 1. History and Geography 52 7.5 2. Fine Arts 7 3. Social Sciences 30 25.0 4. Medicine, Science and Technology 30 7.8 5. Literature and Language 140 23.8 6. Religion and Philosophy 132 26.4 7. Law 3 7.0 8. General Reference 18 2.5 Total: 412 100.0 Table 2: Books containing emotion published 1700-10 16 The World and Fashionable Advertiser until November 26, 1787. 17 Tryal of Rowland Waters Etc. (for Murthering of Sir Charles Pym, Bart) (London, England), Friday, June 1, 1688, p. 4. 18 Present State of Europe or the Historical and Political Mercury (London, England), Monday, August 1, 1692; Issue 7, p. 313 (“A Letter written by an Officer about the Combat fought near Enguien”). 19 Post Man and the Historical Account, Thursday, July 20, 1699; Issue 633, front page (letter from New York, dated March 21). 20 Percentages are taken from ECCO’s “FAQ” section (last visited 15 March 2011). <?page no="70"?> Hans-Jürgen Diller 52 As regards emotion, the section “Religion and Philosophy” runs a close second to the only slightly smaller “Literature and Language”. A look at the first dozen texts revealed only religious and no philosophical texts. As a test of Dixon’s claim it is thus quite adequate. Even these few instances provide powerful evidence against the view that meanings are the exclusive effect of discourse. While the psychological sense of emotion is clearly in the majority there are, equally clearly, nonpsychological uses as well: (1) [Meteors] cause an emotion, attrition, or fermentation, (like water on unslaked lime, or as Spirits of Nitre and Tartar do) which may so encrease as to become flame. (Athenian Gazette or Casuistical Mercury, August 17, 1695; Issue 10, front page, in answer to a question concerning the nature of meteors) (2) The disgrace of the Count d’Arcos has made some Malecontents, who excited a Popular Emotion […]. (Post Boy, Tuesday, September 16, 1701; Issue 989, front p., r.col.) (3) assoon [sic] as he [the King of France] had appeas’d that Emotion among his people …,” (Daily Courant, January 30, 1710; Issue 2580) In (1), emotion describes a physical phenomenon, in (2) and (3) the word is used in a socio-political sense, attested since 1579 and now obsolete (OED, s.v. ‘emotion’, sense 3). These different senses are readily understood without recourse to the concept of discourse, the syntactic microcontext provides a sufficient clue and makes historical knowledge, whether linguistic or conceptual, unnecessary for the immediate understanding. The last sentence is true of 12 of the 13 occurrences of emotion which were found in the Burney Collection before 1711. It is somewhat less true of the 13 th : “Sins of infirmity” are defined as: (4) such as proceed from a sudden Emotion of the sensitive Appetite, which by it’s violence hurries a Man to act before he has made a due Deliberation of the Matter. (British Apollo, March 16, 1710, Issue 114) This clearly invokes a discourse in the strict, Foucauldian sense, which ‘makes’ the objects of which it speaks. ‘Appetites’, rational, natural, or sensitive, are categories postulated by St. Thomas Aquinas and established in what we may usefully call the discourse of scholastic ethics and psychology. But in our context it is important to note that it is not the ‘emotion’ that is formed by the discourse. Rather the reverse: by attributing ‘emotion’ to the sensitive appetite, the writer makes an implicit, metaphorical statement about the nature of appetites: like the populace, the ocean, the earth or the human person, they can be metaphorized as non-solid or non- <?page no="71"?> Discourse, matrix and genre in Historical Semantics: ofermōd and emotion 53 homogeneous masses. These few remarks must suffice to show that the category ‘emotion’ is not created ex nihilo. 21 4. Conclusion This paper has dealt with two rather different tasks of Historical Semantics: placing an individual word-use in the network of contemporary practice, and drawing a (rather small-scale) map of the increasing ‘intuitive’ use of a lexical item which enabled that item to become the label of a central category in a discourse. At the beginning of Section 3 I said that Dixon uses the category of discourse in a weak sense. In the rest of the section the ‘kind-making’ power of discourse was rather neglected, and the notion of matrix was conspicuously absent. Comparing Sections 2 and 3, readers may well ask whether the two concepts of discourse have enough in common to be discussed under the same label. In partial defence, I would admit that the psychological discourse is indeed unlikely to contain performative acts like declaring a person insane, giving a goal, recognizing a woman refugee, or determining the gravity of a sinful act. This admission, though, does not deny Dixon’s claim that the substitution of emotion for passion is more than a mere change of labels. To replace one category by the other, the new category had indeed to be ‘created’ as a scientific, nontranscendental category. In the medieval discourse of Christian ethics, gluttony, lechery, avarice, wrath, sadness, sloth, vanity and pride were categorized as ‘vices’: 22 they were moral as much as psychological categories. To the extent that they were adopted into modern psychological discourse, they were stripped of their moral evaluation and newly ‘created’ as new kinds. In that sense, the ‘kind-making’ power of discourse can be observed here, too. Discourse in a weak sense still seems an appropriate name for Dixon’s use of the term. 21 See Diller (2010) for a fuller discussion. 22 See Cassian, V, i: gula, fornicatio, avaritia, ira, tristitia, acedia/ anxietas/ taedium cordis, cenodoxia/ vana gloria, superbia (PL XLIX, col. 203, website page 76). <?page no="72"?> Hans-Jürgen Diller 54 Biographical sketch Hans-Jürgen Diller, Professor of English Language and Literature of the Middle Ages (Giessen 1969-73, Bochum since 1973, Emeritus since 1999), Visiting Professor. (‘Herder-Professur’) Jagiellonian University, Cracow, 2001. Studied English, Romance Languages and Philosophy at Universities of Kiel, Munich, Besançon and at Wesleyan University (Middletown, Conn.). PhD Munich 1963, Habilitation Göttingen 1968. Main publications: Redeformen des englischen Misterienspiels (1973, Engl. transl. The English Mystery Play. A Study in Dramatic Speech and Form Cambridge 1992), Linguistische Probleme der Übersetzung. Deutsch-Englisch (with J. Kornelius, 1978), Metrik und Verslehre (1978), English Romantic Prose (co-ed. with G. Ahrends, 1990), Unconventional Conventions in Theatre Texts (co-ed. with G. Ahrends, 1990), Chapters from the History of Stage Cruelty (co-ed. with G. Ahrends, 1994), Gewalt im Drama und auf der Bühne (co-ed. with U.K. Ketelsen and H.U. Seeber, 1997), Theatre and Religion (co-ed. with G. Ahrends, 1998), Towards a History of English as a History of Genres (co-ed. with M. Görlach, 2001). Articles chiefly on medieval drama, English metre, historical semantics, stylistics. Co-founder and co-editor of anglistik & englischunterricht (1977-99). References 17th-18th Century Burney Collection Newspapers: http: / / infotrac.galegroup.com/ itweb/ bochum? db=BBCN (last visited 16 August 2010). Abu-Lughod, Lila and Catherine A. Lutz. 1990. “Introduction: Emotion, Discourse, and the Politics of Everyday Life”. In Catherine A. Lutz and Lila Abu- Lughod (eds.). Language and the Politics of Emotion. Cambridge: Cambridge University Press, pp. 1-23. Austin, John Langshaw. 1962 (First edition). How to Do Things with Words. Cambridge MA: Harvard University Press. Baasten, Matthew. 1986. Pride according to Gregory the Great [=Studies in the Bible and Early Christianity, 7]. Lewiston NY: Edwin Mellon Press. Biber, Douglas. 1989. “A Typology of English Texts”. Linguistics 27, pp. 3-43. Biber, Douglas. 2001. “Dimensions of variation among 18 th -century registers”. In Hans-Jürgen Diller and Manfred Görlach (eds.). Towards a History of English as a History of Genres. Heidelberg: Winter, pp. 89-109. Biber, Douglas, Edward Finegan and Dwight Atkinson. 1994. “ARCHER and its Challenges; Compiling and Exploring a Representative Corpus of Historical English Registers”. In Udo Fries, Gunnel Tottie and Peter Schneider (eds.). Creating and Using English Language Corpora. Papers from the Fourteenth Inter- <?page no="73"?> Discourse, matrix and genre in Historical Semantics: ofermōd and emotion 55 national Conference on English Language Research on Computerized Corpora Zürich 1993. Amsterdam and Atlanta GA: Rodopi, pp. 1-13. Bieler, Ludwig (ed.). 1963. The Irish Penitentials [=Scriptores Latini Hiberniae, 5]. Dublin: The Dublin Institute for Advanced Studies. Bloomfield, Morton. 1952. The Seven Deadly Sins. An Introduction to a Religious Concept, with Special Reference to Medieval English Literature. Ann Arbor MI: Michigan State University Press. Burr, Vivien. 2003. Social Constructionism. (Second edition, first edition An Introduction to Social Constructionism, 1995). London: Routledge. Busse, Dietrich and Wolfgang Teubert. 1994. “Ist Diskurs ein sprachwissenschaftliches Objekt? ”. In Dietrich Busse, Fritz Hermanns and Wolfgang Teubert (eds.). Begriffsgeschichte und Diskursgeschichte. Opladen: Westdeutscher Verlag, pp. 10-28. Carabine, Joan. 2001. “Unmarried Motherhood 1830-1990: A Genealogical Analysis”. In Wetherell, Taylor and Yates, pp. 267-310. Cassianus, Johannes. 1874 [420x429]. De institutis coenobiorum et de octo principalium vitiorum remediis. In Ioannis Cassiani Opera Omnia, ed. J.-P. Migne. Tomus I [=Patrologia Latina, 49]. Paris. Chadwyck-Healey. 2000. English Poetry Full-Text Database on CD-ROM-DVD. Second Edition. Colunga, Alberto, and Lorenzo Turrado (eds.). 1959. Biblia Sacra iuxta Vulgatam Clementinam Nova Editio. Madrid: Biblioteca de Autores Cristianos. Deissler, Alfons and Anton Vögtle (eds.). 1985. Neue Jerusalemer Bibel. Einheitsübersetzung mit dem Kommentar der Jerusalemer Bibel. (Revised and enlarged edition, first edition Jerusalemer Bibel 1968.) Freiburg: Herder. Diller, Hans-Jürgen. 1986. “Wortbedeutung und literarische Gattung. Ein Versuch am Beispiel von ae. wlanc”. In Raimund Borgmeier (ed.). Gattungsprobleme in der anglo-amerikanischen Literatur. Beiträge für Ulrich Suerbaum zu seinem 60. Geburtstag. Tübingen: Niemeyer, pp. 1-11. Diller, Hans-Jürgen. 2009. “Old English mōdig: Cognitive Semantics and Cultural Contact”. In R. W. McConchie, Alpo Honkapohja, and Jukka Tyrkkö (eds.). Selected Proceedings of the 2008 Symposium on New Approaches in English Historical Lexis (HEL-LEX 2). Somerville MA: Cascadilla Proceedings Project, pp. 41-48. http: / / www.helsinki.fi/ varieng/ HELLEX/ Diller, Hans-Jürgen. 2010. “‘Emotion’ vs. ‘Passion’: The History of Word-Use and the Emergence of an A-moral Category”, Archiv für Begriffsgeschichte 52 (2010), pp. 127-151. Diller, Hans-Jürgen. Forthcoming. “Soong on Ifaluk, ANGER and WRATH in Middle English: Historical Semantics as bridge-builder”. In Laura Wright and Richard Dance (eds.). Forthcoming. Fourteen Essays on Middle English. Frankfurt: Peter Lang. [=Proceedings of Sixth International Conference on Middle English]. Dixon, Thomas. 2003. From Passions to Emotions: the Creation of a Secular Psychological Category. Cambridge: Cambridge University Press. Dixon, Thomas. 2006. “Patients and Passions: Languages of Medicine and Emotion, 1789-1850”. In Fay Bound Alberti (ed.). Medicine, Emotion and Disease, 1700-1950. Basingstoke & New York: Palgrave Macmillan, pp. 22-52. <?page no="74"?> Hans-Jürgen Diller 56 Doane, Alger N. (ed.). 1991. The Saxon Genesis. An Edition of the West Saxon Genesis B and the Old Saxon Vatican Genesis. Madison WI: University of Wisconsin Press. Dreyfus, Hubert L. and Paul Rabinow. 1982. Michel Foucault: Beyond Structuralism and Hermeneutics. With an Afterword by Michel Foucault. Chicago: The University of Chicago Press. Eighteenth-Century Collections Online: http: / / infotrac.galegroup.com/ itweb? db=ECCO (last visited 16 August 2010). Fairclough, Norman. 1989. Language and Power. London: Longman. Foucault, Michel. 1965. Madness and Civilization: A History of Insanity in the Age of Reason. (Translated from French by Richard Howard. Original title: Histoire de la folie à l’âge classique - Folie et déraison. 1961). New York NY: Pantheon Books. Foucault, Michel. 1972. The Archaeology of Knowledge, and, The Discourse on Language. (Translated from French by A.M. Sheridan Smith. Original title: L’Archéologie du Savoir. 1969). New York NY: Pantheon Books. Frantzen, Allen J. 1983. The Literature of Penance in Anglo-Saxon England. Brunswick NJ: Rutgers University Press. Gadamer, Hans-Georg. 1989. Truth and Method. (Second, revised edition of translation from German by J. Weinsheimer and D.G. Marshall. Original title: Wahrheit und Methode. Grundzüge einer philosophischen Hermeneutik. 1960). New York: Crossroad. Gneuss, Helmut. 1976. “The Battle of Maldon 89: Byrhtnoth’s ofermod Once Again”, Studies in Philology 73, pp. 117-137. Gregory. Moralia in Iob: see Sancti Gregorii Magni Moralia in Iob. Greschat, Katharina. 2005. Die Moralia in Job Gregors des Großen. Ein christologisch-ekklesiologischer Kommentar. [=Studien und Texte zu Antike und Christentum, 31]. Tübingen: Mohr Siebeck. Gutenberg: see Project Gutenberg. Habermas, Jürgen. 1984, 1987. The Theory of Communicative Action. (Translated from German by Thomas McCarthy. Original title: Theorie des kommunikativen Handelns. 1981). 2 vols., vol.1: London: Heinemann; vol.2: Cambridge: Polity. Hacking, Ian. 2000. The Social Construction of What? (First Harvard University Press paperback edition, first edition 1999). Cambridge MA: Harvard University Press. Hacking, Ian. 2004. “Between Michel Foucault and Erving Goffman: Between Discourse in the Abstract and Face-to-face Interaction”. Economy and Society 33 (3), pp. 277-302. Halbrooks, John. 2003. “Byrhtnoth’s Great-hearted Mirth, or Praise and Blame in The Battle of Maldon”. Philological Quarterly 82, pp. 239-255. Harré, Rom. 1986. “The Social Constructionist Viewpoint”. In Rom Harré (ed.). The Social Construction of Emotions. Oxford: Blackwell, pp. 2-14. Hayek, F[riedrich] A[ugust]. 1970. Die Irrtümer des Konstruktivismus und die Grundlagen legitimer Kritik gesellschaftlicher Gebilde. Munich and Salzburg: Fink. <?page no="75"?> Discourse, matrix and genre in Historical Semantics: ofermōd and emotion 57 Jäger, Siegfried. 1999. Kritische Diskursanalyse, eine Einführung. (Second edition, revised and extended, first edition Text- und Diskursanalyse 1994). Duisburg: Duisburger Institut für Sprach- und Sozialforschung. Kemper, Theodore D. 2006. “Power and Status and the Power-Status Theory of Emotion”. In Jan E. Stets and Jonathan H. Turner (eds.). Handbook of the Sociology of Emotions. New York NY: Springer Science+Business Media, pp. 87-113. Lewis, Clive Staples. 1967. Studies in Words. (Second edition, first edition 1960). Cambridge: Cambridge University Press. LION = Literature Online: http: / / lion.chadwyck.co.uk/ (last visited 26 November 2010). Loi sur les aliénés n° 7443 du 30 juin 1838: http: / / www.ch-charcot56.fr/ textes/ l1838-7443.htm (last visited 30 November 2010). Lutz, Catherine A. 1988. Unnatural Emotions. Everyday Sentiments on a Micronesian Atoll and Their Challenge to Western Theory. Chicago: University of Chicago Press. Maset, Michael. 2002. Diskurs, Macht und Geschichte. Foucaults Analysetechniken und die historische Forschung. Frankfurt: Campus. McNeill, John Thomas and Helena M. Gamer (eds.). 1938 (repr. 1965). Medieval Handbooks of Penance: a Translation of the Principal ‘Libri poenitentiales’ and Selections from Related Documents. New York NY: Columbia University Press. Murdoch, Iris. 1961. Sartre. Romantic Rationalist. [=Studies in Modern European Literature and Thought, 9]. (Third edition, first edition 1953). New Haven CT: Yale University Press. Online Books Page: http: / / onlinebooks.library.upenn.edu/ . Polyglot Bible: http: / / www.sacred-texts.com/ bib/ poly/ sir.htm. Popper, Karl R. 2005. The Open Society and its Enemies. Vol. I: The Spell of Plato. (Fourth revised edition, transferred to digital printing 2005. First edition 1945.) London and New York: Routledge. Project Gutenberg: http: / / www.gutenberg.org. Richter, Melvin. 1996. “Appreciating a Contemporary Classic: The Geschichtliche Grundbegriffe and Future Scholarship”. In Hartmut Lehmann and Melvin Richter (eds.). The Meaning of Historical Terms and Concepts. New Studies on Begriffsgeschichte. Washington DC: German Historical Institute. (Occasional Paper no. 15), pp. 7-19. http: / / webdoc.sub.gwdg.de/ ebook/ serien/ p/ ghi-dc/ op15.pdf Sancti Gregorii Magni Moralia in Iob Libri XXIII-XXXV, ed. by Marc Adriaen. [=Corpus Christianorum Series Latina, 143B]. Turnhout: Brepols 1979. Scragg, Donald G. (ed.). 1981. The Battle of Maldon. Manchester: Manchester University Press. Scragg, Donald G. 2006. The Return of the Vikings. The Battle of Maldon 991. Stroud, Gloucestershire: Tempus Publishing. Stedman, Gesa. 2002. Stemming the Torrent. Expression and Control in the Victorian Discourses on Emotions, 1830-1872. Burlington VT: Ashgate. Taylor, Stephanie. 2001. “Locating and Conducting Discourse Analytic Research”. In Wetherell, Taylor and Yates, pp. 5-48. <?page no="76"?> Hans-Jürgen Diller 58 Van Dijk, Teun A. 1998. Ideology: A Multidisciplinary Approach. London: Sage. Wasserschleben, F.W.H. (ed.). 1851 (repr. 1958). Die Bußordnungen der abendländischen Kirche. Halle: Graeger (repr. Graz: Akademische Druck- und Verlagsanstalt). Wetherell, Margaret, Stephanie Taylor and Simeon J. Yates (eds.). 2001. Discourse as Data. A Guide for Analysis. London: Sage. Wikisource: http: / / en.wikisource.org/ wiki/ Main_Page. <?page no="77"?> Udo Fries Melancholy Accidents in Early English Newspapers Abstract Newspaper reports of accidents increase dramatically in the 18 th century. This study looks at 259 accident reports from the ZEN Corpus: accidents caused by thunderstorms and lightning, drowning, falls from a horse, road accidents, and injuries or deaths from fires. The reports vary in length: single sentence reports are the standard, but longer reports become more common as the 18 th century develops and the need for human interest stories (about the upper classes) increases. Sensational details and comments, the inclusion of descriptive adjectives (melancholy accidents) are introduced to give an emotional touch: a distinction between hard and soft news can be observed for the first time. On the other hand, sentences become shorter (from 65 to 30 words), they follow similar patterns, and use stereotype formulations. 1 Introduction In the late 17 th and early 18 th centuries, the major concern of English newspapers was the transmission of foreign news. Among the home news, reports of accidents of various kinds were gradually included, and became more numerous in the course of the 18 th century. In this study I want to analyse the reports of accidents in these early newspapers by using the material collected in the ZEN Corpus (ZEN 2004), which includes newspapers from 1671 to 1791. The type of accident we are concerned with here refers - following the OED (sub accident) - to an “unfortunate event, a disaster, a mishap”. Many of these accidents involve serious injuries or the death of some person. Causes may be thunderstorms and lightning or events that came about through carelessness such as falls of many types: falls into the water with subsequent drowning, falls from a building (including the collapse of a building), or falls from a horse. Many accidents happened on the road, others include injuries or deaths due to fires, accidental gunshots or even rabid dogs. One may distinguish between accidents which affect a large number of people and others that concern only a few people or only one or two persons. To the former group tend to belong thunderstorms and fires, both of which were mentioned in the column REMARKABLE EVENTS, printed towards the end of the 18 th century in The Morning Chronicle. Here is an <?page no="78"?> Udo Fries 60 excerpt from January 1, 1791, in a survey of the news of the preceding twelve months: (1) MARCH. 2. Mr. Fox’s Motion to repeal the Test Act negatived by 294 to 105. 9. Dreadful fire at Hanway Yard, Oxford-street, by which several persons lost their lives. 14. Violent Thunder-storm at Thurso in Scotland. — Captain Bligh arrived in London, and brought intelligence of a mutiny on board the Bounty armed ship in the South See, on the 28th April, 1789; when the Captain and 17 others were put into an open boat, and after travelling the ocean for 46 days, the extent of 4000 miles, arrived at Timor on the 12th of June. (1791mcr06730: s: 278) Here, the fire that occurred on March 9 th and the thunderstorm of March 14 th were important enough to be mentioned on the same level with foreign and other home news. On the other hand, fires affecting a single house, accidents on the road involving one or two persons only, or someone drowning belong to a more personal sphere and would not appear in a summary of the sort shown above. 2 The Corpus The ZEN Corpus provides some help for the researcher looking for examples of accident reports, as it marks a separate text class ACCIDENT , which is provided to facilitate the study of these texts. There are 218 passages marked ACCIDENT , which form the basis of this study. A small number of these had to be eliminated as they did not fit the above definition of an accident. On the other hand, additional examples of accident reports, which were not marked by the compilers of the ZEN Corpus in the text class ACCIDENT , but in other text classes, mainly in HOME NEWS , CRIME , SHIP NEWS , or among FOREIGN NEWS or even DEATHS , were included. In order to retrieve them, a number of key words typical for accidents, such as fire, fall, light(e)ning, drowned or drown’d were used. The columns Casualties and Weekly Bills of Mortality were disregarded. On the other hand, all occurrences of the word accident were checked. In 49 instances this word occurs as part of a report of an accident, and thus gives us an indication of what contemporaries of the time considered an accident. The overall sum of accident reports discussed here is 259. We will divide the period investigated into three sub-periods, (1) from 1671 to 1691 representing the years dominated by The London Gazette, (2) from 1701 to 1741 the period of expansion, and (3) from 1751 to 1791 the period of much larger newspapers. The accident reports of these three periods are presented in section 2. <?page no="79"?> Melancholy Accidents in Early English Newspapers 61 3 Periods of Early English Newspapers 3.1 The Late 17 th Century: 1671-1691 The only English newspaper of the 17 th century was the The London Gazette (originally The Oxford Gazette; cf. Handover 1965). For the years 1671 and 1681, a number of newsbooks, among them The Domestic Intelligence (din) and The Impartial Prostestant Mercury (ipm) are also included in the ZEN Corpus, and so are a few issues of the short-lived Current Intelligence of 1666 (included in the ZEN Corpus under the year 1671). The entire number of accidents reported in this period is very small (24 reports). With two exceptions (see (4) and (5) below), all of them refer to fires (14 reports) and thunderstorms (8 reports). These reports are part of home news, ship news, or foreign news. Even The London Gazette wrote about storms and fires, but only if they had occurred abroad. Here is the beginning of a paragraph in which a report of a thunderstorm in Spain is embedded in the middle. (2) Madrid, Jan. 7. The last week died here Don Antonio de Vidania of the Consejo Real, by whose death and those others formerly deceased, there are at present Six places vacant in that Council, which are daily expected to be filled up, the beginning of the year being the usual time for bestowing such Graces in this Court. On the 5th instant at night, happened a fire in the street of the Preciados, which at the beginning was very violent, but by the care and timely assistance of the people that flocked thither, was happily quenched without any further damage, then the consuming of one house onely. The same evening happened a Quarrel between some of the Portugal Ambassadors servants, and others of this Town, where after sharp words being come to blows, a Postillion of the Ambassadors was killed, and Two of his Footmen desperately wounded. (1671lgz00540: s: 17.1-19.1) The oldest report of a thunderstorm in the Corpus is from The Current Intelligence (1666) from Ludlow in Shropshire, which stands on its own: (3) Ludlow, July 9. Some few daies since have happened very violent storms of Hail and Thunder, the effects of which these parts have been too sensible of; Mr. John Davies of Lanvihangel was kill’d himself, horse and dog; about five miles from the same place, a Man and his Wife scorcht up; a Woman found dead; with one Ear struck off; several Cattle and Sheep kill’d, all judg’d to be the consequences of this sad accident. (1671cui00013: s: 28.1) <?page no="80"?> Udo Fries 62 Special Cases Apart from the reports of fires and thunderstorms, there is one report of an accident of a person falling down, printed in The Domestick Intelligence of 1681. (4) On Wednesday last a Gentleman walking in the Fields near Hogsdon, fell down for Dead, and it was long e’re he was perceived to breath, notwithstanding his being carried into a house, and put into bed, he came not to himself again under 16 hours, though no means were omited for his Recovery, to the admiration of many. (1681din00001: s: 20.1) The following report stands out because of its length. The London Gazette opens its issue of January 5, 1671 with a report from Dublin, which consists of two items, a long report of an accident of the collapse of part of a playhouse followed in the same paragraph by a very short notice of the death of a bishop. (5) Dublin, December 27. Yesterday happened here a very unfortunate accident, most of the Nobility and Gentry being at a Play, at the publick Play-house, the upper Galeries on a sudden fell all down, beating down the Second, which, together with all the people that were in them, fell into the Pit and lower Boxes. His Excellency the Lord Lieutenant with his Lady, happened to be there, but thanks be to God escaped the danger without any harm; part of the Box where they were, remaining firm, and so resisting the fall of what was above; onely his two Sons were found quite buried under the Timber; the younger had received but little hurt, but the eldest was taken up dead to all appearance, but having presently been let blood, and other remedies being timely applyed to him, he is at present past all danger. There were many dangerously hurt, and Seven or Eight killed outright. Here is lately dead the Bishop of Dromore and its thought that Doctor Essex Digby may succed him in that See. (1671lgz00537: s: 2.1ff) 3.2 The Beginning of the 18 th Century: 1701-1741 New developments set in at the end of the 17 th century, when, in 1695, the existing licensing act was not renewed and publishers immediately tested their new freedom. Among the new newspapers we find The English Post (ept), The Post Boy (pby), The Post Man (pmn), The Flying Post (fpt), and The London Post (lpt), all of which can be found in the ZEN Corpus for 1701, although only The London Post and The English Post carry a few reports of accidents in that year. This first batch of papers was followed in the 18 th century by The Daily Courant (dct), The Daily Post (dpt), The London Daily Post (ldp), and The Daily Advertiser (dat). In 1711 there are hardly any reports of accidents in the ZEN Corpus. Only from 1721 does their number increase. <?page no="81"?> Melancholy Accidents in Early English Newspapers 63 For this period, the ZEN Corpus also includes a number of weeklies, Applebee’s Original Weekly Journal (apb) and The Weekly Journal: OR British Gazetteer (wjb) of 1721, Reed’s Weekly Journal (rwj), and The Country Journal, or, the Craftsman (cjl) of 1731 and 1741. These weeklies carry almost half of all the accident reports. Altogether we find 66 reports of accidents in this period. The largest number in this period once again refers to fires that broke out (23 reports). Reports about thunderstorms (6 reports), however, are clearly less frequent both in absolute figures and in relationship to all other accident reports. While fire reports are normally brief, there are some exceptions, the most extensive one of which is the report of a fire of the seat of the Earl of Rochester at Petersam, which runs up to 731 words (1721wjb02045s: 93. 1-6), giving all the details of how the fire was detected, of how the servants jumped out of the window, either miraculously surviving or falling to their deaths, of what was saved and what was lost, and where the family were going to stay after the destruction of their house. The new types of reports are about falls into the water that result in drowning (16 reports), about other falls (6 reports), and about accidents that happened on the road (13 reports). The most common reason for drowning is falling over-board: (6) Last Tuesday a young Lad being in great danger of drowning in the Long Reach, by falling out of a Boat, his Father attempted to save him, and thereby they perished both together; the Man’s Wife and his own Mother were Spectators of this Tragical Scene From the side of a Ship. (1721wjb02045: s: 78.1) Falls include, e.g., falls from a window, but also buildings that collapse: (7) On Sunday last a Child fell out at a Window in Peter’s-Street near Golden-Square, and dash’d its Brains out with the Fall. (1721apb02095: s: 67.1) (8) Clonmell, Sept. 23. On Monday last was acted in this Town the Recruiting Officer, with the Devil in the Wine-Cellar, and as the Actors began the Farce the Lofts of the House not only fell down, but also the Stage and above 200 Persons of those that sat in the front Seats and on the Stage, several of whom had their Arms and Legs broke, and others so much bruis’d that their Lives are despair’d off. (1741dpt06888: s: 22.1) Frequently people, often children, are run over by a coach. (9) Yesterday in the Afternoon a Child about 5 Years old, Daughter of Mr. Hyde, a very eminent Cabinet-maker in Maiden-Lane, Covent Garden, was unfortunately run over by a fore-Wheel of a Hackney-Coach, as she was playing before her Father’s Door, and bruised in so terrible a manner that it’s thought that she cannot live. (1731dat00088: s: 32.1) <?page no="82"?> Udo Fries 64 Special Cases There are two reports that do not fall into any of the categories mentioned above. Both mention that the report is about an accident, a fatal one in (10) and a sad one in (11). The former appeared in The Weekly Journal and looks very much like a mock report of an accident. (10) Some Days ago, a fatal Accident happened near St. James’s; a large Cat playing with a Lady’s little lap Bitch, scratch’d her miserably; poor Liddy howl’d and bled for half an Hour, whilst her Mistress wash’d her Wounds with her Tears. The good Lady can scarce Eat, Drink, or Sleep since, and ’tis thought will not be able to receive a Visit, not even on her Bed, till Liddy’s Wounds are closed, and the poor Creature quite out of Pain. (1721wjb01835: s: 108.1) The second is about an accidental shooting and its dire consequences. (11) On Tuesday the following sad Accident happen’d at Star-Cross: one John Jeffrys, a Baker, going into the Room of one Mary Wills, and seeing several Fowling Pieces behind the Door, he unfortunately took one up of them, and twirling it round his Hand, it went off, shot the Woman in the Head, and blow’d her Brains about the Room. (1741ldp02074: s: 98.1) 3.3 The Second Half of the 18 th Century: 1751-1791 The number of newspapers, including weeklies, rises steadily during this period, which is mirrored by a wide variety of selections in the ZEN Corpus. The London Chronicle (lcr) is represented for the years 1761, 1781 and 1791, The Morning Chronicle (mcr) and The Public Advertiser (pad) for 1781 and 1791, all of them carry many reports of accidents, so does Lloyd’s Evening Post and Chronicle (lep). The Times, which was founded as the Daily Universal Register in 1785, is not included in the ZEN Corpus, as it is easily available elsewhere. Of those newspapers available for only one decade, The General Advertiser (gat) and The London Daily Advertiser (lda) of 1751 carry many reports of accidents. The London Morning Penny Post (lmp) of 1751 published under the headings C OUNTRY N EWS , L ONDON and S COTLAND short news on its front page, many of which are reports of accidents. In 1771 The Craftsman; or SAY’s Weekly Journal (csw) published many reports of accidents. All the other papers have only an occasional report. The London Gazette is no longer a newspaper but has turned into a mere journal of public announcements and advertising: There is just one report of an accident among its foreign news section. For 1781, although all the major newspapers are included, there are hardly any reports of accidents, a fact that may be coincidental, but needs further attention. <?page no="83"?> Melancholy Accidents in Early English Newspapers 65 This is the period with the largest number of accident reports: altogether there are 168 reports, more than twice as many as in the period before. As usual, fires (43 reports) are the most frequent reports, followed by reports of accidents on the road (37) and drowning (35). Falls (27) are clearly more frequent than in the period before, and so are reports about thunderstorms and lightening (20). Among the reports of fires, there are a few which occur in the foreign news sections, as this report from Cologne about a fire in Linz, Austria, from The London Gazette: (12) Cologne, Oct. 7. We learn by our last Letters from Upper Austria, that a dreadful Fire had lately broke out at Ottensheim, a small Town about a Mile from Lintz, which burnt with great Violence, and reduced to Ashes one hundred and ten Houses besides the Parish Church. (1751lgz09098: s: 30.1) In 1791 both The London Chronicle and The Public Advertiser published exactly the same report of “a truly melancholy accident” of 14 people drowning. (13) Extract of a Letter from Minehead, Somerset-Shire, Feb. 3. “A truly melancholy accident happened off this Port last Wednesday; a party of thirteen ladies and gentlemen went in a pleasure yacht to dine at Wachet, a small town seven miles distant from this; the morning was extremely fine, the water placid, and every thing promised a agreeable day; but, alas! How soon was the scene changed! They had got about a league off the shore, and were going on briskly before the breeze, when, in the sight of many people standing on the Pier, the boat suddenly went down, and every soul on board was drowned. It is generally supposed that one of the planks sprung. Thus were they all, without the least previous warning, plunged into eternity. The unfortunate sufferers were Mr. and Mrs. Carpenter, Mr. John Deake, and Miss Rawle, Mr. Adams and his daughter, Mr. and Mrs. P. Ball, Miss H. and C. Bastome, Miss Bouchier, Miss Broadmead, and William Wickland, Esq. None of their bodies have yet been taken up. It is supposed that a strong Easterly wind which sprung up the next day blew them down the Channel to the Westward. This town is now truly a scene of mourning and distress. They have all left either large families or friends to deplore their untimely fate.” (1791lcr05378: s: 51ff and 1791pad17660: s: 228ff) This report is really a letter, it is much longer than the usual matter-of-fact reports, and consists of almost poetic phrases: every soul on board, plunged into eternity, deplore their untimely fate. An even longer report “of thunder, lightning, hail, rain, and trembling of the earth, such as was never seen or heard of in this part of Britain before” (1781mcr03783: s: 241ff), also as an extract from a letter, appeared in The Morning Chronicle of 1781. <?page no="84"?> Udo Fries 66 Special Cases Five reports do not fit the established categories, one about a father hitting his son (14), two of rabid dogs (one in 15), another gunshot by accident (cf. 11 above), and - among the foreign news - of a whole powdermagazine blowing up (16). (14) Extract of a letter from a Gentleman near Colchester, to his Friend in town. “A Gentleman’s son, a boy about six years of age, and of whom he was vastly fond, happened to do some trifling thing which displeased his father, he gave him a box on the side of the head, which immediately caused him to bleed at the nose and ear, and of which he died in three hours; the father was so much shocked at the accident and the fatal consequence, that he went raving mad, and died so in three days.” (1771wlp01388: s: 175.1f) (15) From Aberdeen we are inform’d of a melancholy accident that happened lately near that place. As a young girl was asleep, a dog, not suspected of madness, crept softly to her bed, and licked her face for some time, which, upon waking, she found all over in a lather of froth, and complained of no ailment that day. But soon after she was violently seized with all the symptoms of that dreadful infection, and died raving mad. (1761rwj08901: s: 72.1ff) (16) Stralsund, Dec. 5. Wednesday last the powder-magazine in this city blew up at the time the workmen were all there; by which terrible accident upwards of an hundred lives were lost, and more than a thousand persons were dangerously wounded. Seventy houses were entirely thrown down, and the windows of all the other houses in the city were broken. (1771gep05813: s: 290.1) 4 Analyses 4.1 Paragraphs and Sentences In the first period, reports of accidents normally consist of one sentence only, which is embedded at the beginning, in the middle, or at the end of a longer paragraph. In 1681, The Domestick Intelligence and The Impartial Protestant Mercury, however, have a number of reports, which show the pattern that was to be the common one in the 18 th century. The reports consist of one sentence each, printed as a separate paragraph, preceded and followed by other reports of local affairs. (17) On Tuesday last a Fire happened in the Kings-Bench Prison in Southwark, occasioned by a certain Prisoner setting a Pan of Coals under his Bed to <?page no="85"?> Melancholy Accidents in Early English Newspapers 67 kill the Buggs as he pretended, the which, had it not been timely discovered by the smoak and smother, might have proved of fatal consequence, and in short space have reduced that Antient Pile to Ashes. (1681din00002: s.16.1) (18) On Sunday the 5th instant, a Fire began at an Hotpressers House, within two doors of Skinners-Hall, where the Right Honourable Sir Patience Ward keeps his Mayoralty; all the Family was abroad, but being happily discover’d by Passengers, they broke open the door, master’d it, without damage to the contiguous Houses. (1681ipm00013: s: 21.1) In the second period, only five reports - about thunderstorms and fires - are embedded in longer reports dealing with other matters as well. They are all reports from abroad (i.e. foreign news). 72 percent are of the type of one paragraph consisting of one sentence only. The remaining 28 percent consist of 2 to 4 sentences. Among them are two reports about the same fire, which appeared in The Daily Post and in The Country Journal. In both reports there is a suggestion that the fire was set on purpose. (19) The same Day, about Two o’Clock in the Morning, a Fire was discover’d at the House of Mr. Crookshanks, a Virginia Merchant, in Poor-Jury- Lane near Aldgate, which before it could be conquer’d consume’d five Dwelling Homes and several Warehouses. ’Tis uncertain how it began, but many People say it was set a-fire on purpose by Villains, but that does not appear plain enough to depend on the Truth thereof. (1741dpt06658: s: 58.1) (20) Wednesday Morning early a Fire broke out in the Compting-House of Mr. Crookshanks, an eminent Virginia Merchant, in Poor-Jewry-Lane by Aldgate, which consumed the said House, and another adjoining, with several Buildings backwards. Mr. Crookshanks and his Family narrowly escaped the Flames, and the Furniture, &c. are all destroy’d. ’Tis said that the House was set on Fire by some Villains. (1741cjl00758: s: 74.1) The following report about the same fire in The London Daily Post is much more detailed and suggests it not be classified as an instance of the text class ACCIDENT , but rather of the text class CRIME . (21) Yesterday Morning about One o’Clock Capt. Robert Crookshanks, a very eminent Tobacconist in Poor-Jury-Lane near Aldgate, was awaken’d by a Noise in his House, as of People forcing open a Door, on which he got out of Bed, to alarm his Servants, but was stopp’d on the Stairs by two Fellows, one of whom snapp’d a Pistol at him, but it flash’d in the Pan, and he saw no more of them, soon after he discover’d his Comptinghouse on fire, which spread with such Violence, that himself, his Wife, and Servants had scarce Time to save their Lives; in a few Hours it intirely consum’d the said House, and three or four Warehouses backwards fill’d with Tobacco, and Goods belonging to several Shop keepers <?page no="86"?> Udo Fries 68 in Leadenhall-street, and damag’d some other Houses. Mr. Crookshanks had the Day before received 1500 l. in Bank-Notes, which ’tis probable was known by some Person acquainted with his Affairs; and were, very likely, concern’d in this Villany. (1741ldp01938: s: 101.1 and 102.1.) In the third period, the number of reports of accidents consisting of a paragraph of one sentence decreases slightly from 72 percent to 67 percent and the number of instances with 2 to 5 sentences per paragraph increases to 33 percent. Among accidents on the road or fires, a second sentence typically gives information of what happened to the victim. (22) Wednesday morning several persons had got upon the top of a hackney coach at the end of Hatton Garden, to see the prisoners pass to execution, and some words happening between two men, a struggle ensued, and one of them fell from the carriage and broke his leg. He was immediately carried to St. Bartholemew’s Hospital. (1771wlp01388: s: 141.1f) (23) By the fire which broke out yesterday morning, at Mr. Cope’s sugarhouse, in Thames-street, opposite College-hill, a great quantity of whalebone was consumed in Fishermens-hall; and a high party-wall belonging to the sugar-bakers falling on Mr. Pyke’s Meeting-house, the greatest part of it was beat in. Several persons who came to assist, were buried in the ruins, but were all dug out; one, however, had his leg and thigh broke, and another was dangerously bruised, and carried to St. Thomas’s Hospital. (1761lcr00675: s: 74f) But other topics occur as well in the second setence, as in this early case of hit-and-run: (24) On Tuesday a Man in a Chair drove over a Child in Brewer-street, who with the bruises died in less than half an Hour. The Surprize this Accident occasioned, gave time for the Man to make his Escape unknown. (1751lmp03102: s: 79.1f) 4.2 Sentence Length In the first period, more than 50 percent of the sentences are longer than 50 words. The average sentence length is 65 words. 1 In the second period the figure comes down to 42 words. Among the sentences about fires there are three reports that stand out: they have three sentences with 132, 210 and 263 words respectively. If we disregard these three sentences, the average sentence length is reduced to 37 words. Reports on falls, fire and thunderstorms have shorter sentences, those on accidents on the road, 1 As in previous studies, sentence length is measured in s-units: a sentence finishes with a full stop, a question mark or an exclamation mark (cf. Fries 2010). <?page no="87"?> Melancholy Accidents in Early English Newspapers 69 drowning, and the small group of miscellaneous accidents have longer sentences. In the third period, the average sentence length drops to 30 words. There is less difference between the individual groups of accident reports, those on the road and of fires tend to have longer sentences, whereas those on drowning have shorter ones. All the others cluster around 30 words. The pattern observed here corresponds to the findings of the development of sentence length in newspapers of the 18 th century, beginning with very long sentences around 1700 and gradually decreasing in subsequent decades (cf. Schneider 2002). 4.3 Time Reference The prototypical report of an accident begins with a time reference. This is particularly the case with reports of accidents on the road and of falls, but also reports of drowning and of fires show a very high percentage. There is a development towards this type of report, which is shown by the figures for the third period, which are almost always higher than those for all three periods together. total time adverbial percentage 3 rd period only in initial position on the road 53 52 98.11% 97.43% falls 32 31 96.8% 100% drowning 53 44 83.01% 83.33% fires 79 61 77.2% 88.37% thunder 34 19 55.8% 60% Table 1: Time reference in first slot of a report The most frequent time reference is the mentioning of a weekday, with or without the preceding preposition on, or some specification of the time of the day (morning, afternoon, evening). (25) On Monday Afternoon a Daughter of Mr. Fitz-Symonds, a Linenweaver in Marrowbone-Lane, was run over, and killed by two Boys who were riding on a Cart. (1731rwj00328: s: 31.1) Last is frequently added, before or after the weekday (see (7) above and (26) below). Yesterday occurs in all periods, it is particularly frequent in reports on fires, with the addition of the time of the day, which is an important part of these accident reports (see (27)). <?page no="88"?> Udo Fries 70 (26) Last Monday the Wife of Mr. Peacock, who keeps Livery-Stables in Blue Ball Yard, St. James’s Street, had the Misfortune to fall from a Hayloft in the said Stables, and fractured her Skull in such a Manner that there are no Hopes of her Recovery. (1751lmp01416: s: 46.1) (27) Yesterday morning, between seven and eight o’clock, a fire by some accident broke out on board a vessel lying off Blackwall, which burnt down the same to the water’s edge. (1791lcr05362: s: 47.1) Unspecified time references (some days ago or some few days since) occur very occasionally in the first and second periods, last week crops up in all periods, but is generally infrequent. 4.4 The Victims: Persons and Places Whereas in reports about thunderstorms and, to a lesser degree about fires, individual people are rarely mentioned, accidents on the road, falls and drowning usually concern one or two persons only and they are most frequently referred to with a general term, like a man, a child, a young woman, an elderly lady, or a labourer. This is the case in 60% of all instances, the figures vary very little between the types of accident or over the three periods. The attribute poor is found together with man and woman, the former, however, only once. It is rare in the entire ZEN Corpus: there is only one other instance about the victim of a murder. A poor woman, on the other hand, is much more frequent, there are 19 instances in the Corpus altogether, 13 of them in accident reports. (28) Tuesday last a poor woman in Pye-street fell down stairs, and broke both bones of her leg. She was immediately carried to the Westminster Hospital. (1761lep00577: s: 194.1-2) In 40% of instances the name of the victim of an accident is given in one way or another. There is, very rarely, just a name without any hint of who this person may be. (29) On Wednesday last George Manley was brought to the London Hospital, having been run over by a Post Chaise, four Miles beyond Marlborough, as he was sitting on the Side of a Bank to rest himself, on his Journey to London. (1751gat05209: s: 55.1) An alternative would be a phrase beginning with one, followed by a reference to the bearer’s profession or status, but that is very rare, there are just three instances, all from the beginning of the 18 th century. (30) Wednesday Morning one Mr. Barton, a Dealer in Coals, had the Misfortune to fall from the Side of a Lighter near Billingsgate-Stairs, into the Thames at Highwater, and was drowned. (1731rwj00308: s: 52.1) <?page no="89"?> Melancholy Accidents in Early English Newspapers 71 One and the same accident is reported in The Weekly Journal and The Post Boy of the beginning of October 1721. Both reports consist of one sentence, but the report in The Post Boy is twice as long and much more detailed and colourful. (31) Last Week Mr. James Eaton, a young Student riding through Barlow, on the Road to Cambridge, his Horse threw him, and tumbling over his Body, bruis’d him so, that he died in a few Hours after. (1721wjb02045: s: 41.1) (32) On Friday last there happen’d an unfortunate Accident to one Mr. James Eaton, a young Gentleman who studied Physick: As he was riding through Barlow, his Horse being upon a full Gallop, made a false Step, and pitching upon his Head, threw him forward in such a manner, that he tumbled over with the whole Weight of his Body upon him, which bruis’d his Stomach so violently, that he vomited Blood, and died in a very few Hours. (1721pby05024; s: 31.1) In the majority of cases, the victim’s profession, status or home is added to his name. (33) Last Week Robert Cooke, Esq; Counsellor at Law, had the Misfortune, by a fall from the Gallery in the Four Court Marshalsea, to break his Left Leg, besides receiving a violent Contusion on his Head; but we hear he is in a fair Way of Recovery. (1751gat0525: s: 83.1) There are a few instances, where the name of a woman is given. (34) On Sunday last as Mrs. Thatcher, who lives near St. Martin’s Church, was coming to Town from Kilburne, with some other Company, she had the Misfortune, as she was getting over a Gate to fall, by which Accident she fractured her Skull in such a manner that her Life is despaired of. (1751lmp00406: s: 70.1) Otherwise, for children, wives, 2 and servants the name of their father, husband, or employer is provided, as in (35) and (26 above): (35) Last Monday the Wife of Mr. Peacock, who keeps Livery-Stables in Blue Ball Yard, St. James’s Street, had the Misfortune to fall from a Hayloft in the said Stables, and fractured her Skull in such a Manner that there are no Hopes of her Recovery. (1751lmp01416: s: 46.1) With reports of thunderstorms and fires, it is in most cases a building that is destroyed or on fire, and persons are not mentioned unless they lose their lives in these accidents. Of the 79 fires reported in the Corpus, the owner of the burnt house is referred to in 22 instances, which amounts to 27% only. In the first period the usual phrase is that a fire happened, and only once is a person mentioned. 2 For a similar situation in death notices the 19 th century cf. Fries (1990). <?page no="90"?> Udo Fries 72 (36) On Sunday Night last, between the hours of eleven and twelve happened a Fire at one Mr. Wrights house in Crutched Fryers, which by the speedy Assistance of the Neighbourhood was quenched with the Dammage of about fifty pound onely. By what means it happened is yet unknown, but is suspected to be done by Treachery, the Master of the house being last up, and leaving all things secure when he went to bed. (1681pdi00094: s: 23.1-2) From the second period onwards, the most common phrase is that a fire broke out, while a fire happened disappears almost completely. (37) On Sunday last a fire broke out in the Countess of Brandon’s house, on Lazor’s Hill, which was burnt to the ground, and all the rich furniture consumed. (1771csw00657: s: 142.1) When a house of the gentry or the aristocracy is on fire, it is clear that their names are mentioned. The destructive force of accidents, in particular fires, may be mitigated by the timely assistance of people, water or fire engines. The phrase occurs throughout the Corpus. (38) Saturday evening a fire broke out at one Mrs. Wood’s, a green-grocer, in Honey-lane-market, which did considerable damage to the furniture, &c. but by the timely assistance of the engines was prevented from spreading any further. (1771csw00657: s: 117.1) 4.5 Vocabulary: Loss of Life Persons losing their lives are either described as killed (20 instances), died (on the spot) (19 instances), dead (5 instances), perished (in flames or water, 7 instances), expired (5 instances), or, given the circumstances, drowned (38 instances). A person killed by lightening is struck dead. (39) On Wednesday last there was a great deal of thunder and lightening in this neighbourhood, by which a woman at Kebb Coat, near Heptonstall, was struck dead. (1771lev06777: s: 274.1) With 5 instances, expired is relatively seldom used, although the word is quite frequent in reports of illnesses, suicides and crimes (altogether 17 instances). An unusual expression is destroyed, found only once in a report about a thunderstorm in The London Post of 1701. (40) Several Stacks of Chimneys blown down, Persons destroy’d thereby, Houses Until’d, their Roofs carried away, and Trees blown up by the Roots. (1701lpt00256: s: 22.2) Other descriptions are rare, consumed, which is the regular word for fires consuming a building (41 instances in the ZEN Corpus), is extended to human beings in three instances, one of which is <?page no="91"?> Melancholy Accidents in Early English Newspapers 73 (41) Early yesterday morning a fire broke out at the house of correction at Hulsted, in this county, which consumed the prison part of the building, and four prisoners confined therein. (1781lep03711: s: 49.1) The same extension applies to animals. (42) And Thursday Night, between Eleven and Twelve, a dreadful Fire broke out in Tooley-street, Southwark, which has consumed above thirty Houses, an Ox, and some Sheep. (1741cjl00758: s: 76.1) There are hardly any other expressions used for dying, but In Reed’s Weekly Journal of 1731, dying by the kick of a horse is referred to as kick out a Boy’s brain, and in Applebee’s Original Weekly Journal of 1721, the phrase dash’d its Brains out is used (cf. (7) above). (43) The same Night a Mare belonging to Mr. William Smith, of Golden Bridge, kick’d out a Boy’s Brains in Cork-street. (1731rwj00328: s: 32.1) In some cases, it is not clear whether a person involved in an accident will survive or not, but there are indications of which of the two will happen. In both cases the descriptions of the accidents are very detailed. (44) Some Days ago, a Servant of the Rev. Dr. Carter, Rector of Worplestone, and one of the Fellows of Eaton, coming to Guildford Market with a Load of Wheat, Part of the Way being narrow, he went to leap upon the Sharps but miss’d his Hold, and fell backward. Both the Wheels of the Waggon went over his Left Thigh and Right Leg: His Thigh was broke to pieces, so that the Bone came through the Skin: A Man coming by at the time, he was put into the Waggon upon the Sacks of Corn, and carried a Mile to Guildford. He lost a great deal of Blood, till it was set by Mr. John Howard, Surgeon at that Town; and he is in a fair way of doing well. (1731rwj00308: s: 72.1-3) All these instances occur from the middle of the 18 th century onwards. The typical formulation that there is no hope, is that a person’s life is despaired of. It is found in accidents happening on the road, in falls, and - outside the scope of this investigation - in reports of crimes or illnesses. These victims are so much, so terribly, so dangerously bruised, or hurt in such a manner that his or her life is despaired of. (45) One Day this Week, Mr. Benjamin Ludgater, an eminent Master- Builder, being on a Scaffold at Iron-gate, which gave way, he fell down, and broke two of his Ribs, and was otherways so terribly bruised that his Life is despaired of. (1751lda00156: s: 48.1) <?page no="92"?> Udo Fries 74 4.6 Emotional Language: Melancholy Accidents Two ways of adding an emotional touch to the bare facts are analysed here: the use of descriptive adjectives or adverbs and the explicit use of the word accident. Among the adjectives, dreadful or terrible is added to fire. A dreadful fire occurs through all the decades, from 1681 to 1791(cf. (12) and (42) above), but with a total of 16 instances it is far less common than the simple formulation a fire broke out. A terrible fire occurs only twice, once at the beginning of the period, in 1681 and once in 1711. The more common phrases with terrible are in so terrible a manner (cf. (9) above), in a (most) terrible manner, in such a terrible manner, or in a very terrible manner. The phrase is found from 1731 onwards. In 14 instances unfortunately is added to modify the contents of a verb. The most common is was unfortunately drowned. (46) On Monday evening a boat was overset near Cuckold’s Point, by which two men, a woman and a child, with the waterman, were all unfortunately drowned. (1771wlp01388: s: 155.1) Other verbs modified by unfortunately are run over, fall, killed, or perished. In some 50 instances, the word accident itself is used in an accident report. The word refers to all the types of accident discussed in this paper. It appears either as part of an introductory phrase that draws the attention of the reader to what follows or it occurs at the end of the report introducing the result of the accident. (47) On Sunday last happened a very melancholly Accident, Mr. William Compton, Surgeon, and Apothecary at Lamport, returning Home from Martock, was drowned in a Brook in Stapleton Parish, to the great Loss of the Publick, he being a Man very skilful in his Profession. (1751lmp00810: s: 17.1) (48) Friday a Woman cleaning the Windows of a House at Islington, unfortunately fell, by which Accident both her Legs were broke; and she was immediately carried to St. Bartholomew’s Hospital. (1751gat05209: s: 51.1) As an introductory phrase, the word accident is always specified by an adjective as a lamentable, a sad, an unhappy, a terrible, an unforeseen, a remarkable, and most frequently a melancholy accident (cf. (13), (15) and (47) above). References to a melancholy accident appear only in the third period, in 1751 (5 instances), in 1761 (1 instance), and in 1791 (5 instances). <?page no="93"?> Melancholy Accidents in Early English Newspapers 75 5 Conclusion As a category or special text class in newspapers ACCIDENTS cannot always be separated from FOREIGN or HOME NEWS , but rather form a subcategory of each. It was interesting to observe that The London Gazette right from the beginning refrained from publishing any accident reports apart from those that dealt with accidents that had happened abroad. In the second half of the 18 th century accident reports became an important part of news coverage. A large number of accident reports together with crime reports and other home news of a similar shape appear in The London Morning Post, The London Chronicle, Lloyd’s Evening Post, The Public Ledger, The Craftsman, The Westminster Journal, The London Chronicle, and The Morning Chronicle. The text class ACCIDENT takes an intermediate position between hard news and soft news. Although the majority of these texts in early English newspapers belong to hard news, there is also an important minority of reports that could be classified as soft news. The “hard news reports” have become an almost stereotype text class, which describe the various types of accidents in very straight-forward language in just one, or a very few sentences. In these reports there are hardly any sensational details or specific human-interest factors. The “soft news” reports, on the other hand, are more detailed; we find sensational details about members of the upper class, and comments about the accidents reported. The introductory phrases discussed in section 4.6 above can be seen as one type of comment: from a mere accident to a melancholy accident. The study has also shown that the ZEN Corpus can provide a reasonably good start for a text linguistic investigation of a particular text class of early English newspapers. There are, obviously, limitations that can only be overcome by a systematic extension of the material. To see the individual house styles of newspapers, more papers of the same dates would be necessary. In the material looked at here, we could see that occasionally the same report is printed in more than one paper without any changes whatsoever, whereas in other instances there are clear differences between the reports, which would deserve closer inspection. <?page no="94"?> Udo Fries 76 Biographical sketch Udo Fries is Professor Emeritus of English Linguistics at the University of Zurich. His main areas of interest include text linguistics, stylistics, corpus linguistics and the history of the English language. He wrote on Old English syntax and is the author of an introduction to Chaucer’s language. In the 1990s he compiled the ZEN-Corpus of early English newspapers, and has since published many papers on text and corpus linguistics. References Fries, Udo. 1990. “Two Hundred Years of English Death Notices”. In Margaret Bridges (ed.) On Strangeness. (SPELL, Swiss Papers in English Language and Literature, 5). Tübingen: Gunter Narr, pp. 57-71. Fries, Udo. 2010. “Identifying Diversity in English Newspapers in 1701”. In Nicholas Brownlees, Gabriella Del Lungo and John Denton (eds.). The Language of Public and Private Communication in a Historical Perspective. Newcastle upon Tyne: Cambridge Scholars Publishing, pp. 213-227. Handover, Ph. M. 1965. A History of The London Gazette 1665-1965. London: Her Majesty’s Stationery Office. Schneider, Kristina. 2002. The Development of Popular Journalism in England from 1700 to the Present. Corpus Compilation and Selective Stylistic Analysis. PhD Dissertation, Universität Rostock. ZEN: The Zurich Corpus of Early English Newspapers. 2004. Contact: Hans Martin Lehmann, at: hmlehmann@es.uzh.ch. <?page no="95"?> Andreas H. Jucker “What’s in a name? ” Names and terms of address in Shakespeare’s Romeo and Juliet 1 Juliet: ’Tis but thy name that is my enemy; Thou art thyself, though not a Montague. What’s Montague? It is nor hand nor foot, Nor arm nor face, nor any other part Belonging to a man. O be some other name! What’s in a name? That which we call a rose By any other word would smell as sweet; (2.2.38-44) 2 Abstract Linguistically, names are arbitrary designations of specific entities. They do not have any lexical meaning. But if they are used as terms of address (vocatives), they can become powerful designations that express a vast range of social or attitudinal meanings, and as such they interact with the pronominal terms of address thou and you. In this paper, I will explore this interaction between names used as vocatives and the pronouns thou and you in Shakespeare’s Romeo and Juliet. Introduction “What’s in a name? ” asks Juliet in Shakespeare’s Romeo and Juliet. “That which we call a rose / / By any other word would smell as sweet” (2.2.43- 44). According to the traditional linguistic approach signs are arbitrary. The word “rose” does not smell like a rose, neither does it look or sound 1 My thanks go to Elisabeth Bronfen, Beatrix Busse, Daniela Landert and Irma Taavitsainen for very helpful comments on a draft version of this paper. The usual disclaimers apply. 2 Quotations follow the New Cambridge Shakespeare edition (Evans 2003). Unless otherwise indicated they are from Romeo and Juliet. However, the choice of a particular edition is not trivial for pragmatic analyses of Shakespeare’s plays. See, for instance, Blake (2002) or U. Busse and B. Busse (2010: 249). <?page no="96"?> Andreas H. Jucker 78 like one. But words have conventional meanings. They have denotative meanings, which specify the range of concepts that the word can refer to, and they have associative meanings, which concern language users’ attitudes towards the concepts referred to by the word. As opposed to common nouns, proper nouns or names are generally said to have no sense but only reference. In Shakespeare’s play Romeo and Juliet names are particularly important and, therefore, worthy of investigation. Obviously, the dialogues in a play do not record “natural conversations”. They are fictional, and as such do not give us any direct evidence of the spoken language in the late sixteenth century. But linguistically they are sufficiently interesting in themselves. In recent years, the relatively new field of historical pragmatics has developed a new attitude towards historical language data and has rediscovered the language of fiction and drama as legitimate objects of linguistic investigation as long as no unwarranted conclusions to other types of data are drawn (see, for instance, U. Busse and B. Busse 2010; Culpeper and Kytö 2010; Kytö 2010). Romeo very explicitly comments on the significance of names. As if that name, Shot from the deadly level of a gun, Did murder her, as that name’s cursed hand Murdered her kinsman. O tell me, Friar, tell me, In what vile part of this anatomy Doth my name lodge? Tell me, that I may sack The hateful mansion. (3.3.102-108) In this extract from the third scene of act three, Romeo speaks to Friar Laurence after just having killed Juliet’s cousin Tybalt. Both Romeo and Juliet are concerned with the potent meaning of their family names. They are deeply in love with each other but their love is impossible because of their family names and the hostility between the two families which these two names represent. Names of people are used both to refer to them when talking about them and as vocatives, or nominal terms of address, when talking to a person. In this paper, I shall focus on their use as vocatives and nominal terms of address, and in this context it is important to relate them to the use of the pronominal terms of address ye/ you and thou/ thee because they interact in interesting ways with the nominal terms of address. Linguistically, a distinction is made between common nouns, proper nouns and names. Proper nouns generally refer to people, places, months, days, festivals, and so on (Quirk et al 1985: 288), and their formal characteristics differ from common nouns. Common nouns take articles (rose, the rose, a rose), while proper nouns generally do not (Stratford, *the Stratford, *a Stratford). In a given universe of discourse, proper nouns have unique <?page no="97"?> Names and terms of address in Shakespeare’s Romeo and Juliet 79 denotation, that is to say they unambiguously refer to one specific entity only and from a synchronic point of view they do not have sense. Quirk et al (1985: 288) define proper nouns as single words, while a name may consist of more than one word. Thus Shakespeare and William Shakespeare are both names consisting of one or two proper nouns. Proper nouns, and by implication also names, generally refer to one specific entity and are devoid of lexical meaning. In the words of Biber et al. (1999: 245) “The most typical proper nouns (e.g. Tom, Hobart) are arbitrary designations which have no lexical meaning. There are no defining characteristics for those who carry the name Tom, except that they are male.” And in the context of a play, names have an important function for the audience who needs to keep track of who is who in the play. For Romeo and Juliet clearly this is different. Their family names should be arbitrary labels (“What’s in name? ”), but they are full of significance standing in the way of their love. A name can even murder (“As if that name […] did murder her”). Shakespeare’s use of vocatives Shakespeare often presents his characters as being very much aware of the significance and power of how they address each other. When used as vocatives, names are often embedded in standard phrases and used as more or less empty formulas, but on occasions they can regain their literal force as in the following extract from King Henry VI, Part 3 (quoted from B. Busse 2006: 210): Richard Good day, my lord. What, at your book so hard? King Henry Ay, my good lord - my lord, I should say rather. ’Tis sin to flatter; ‘good’ was little better: ‘Good Gloucester’ and ‘good devil’ were alike, And both preposterous; therefore not ‘good lord.’ (3H6 5.6.1-5) The exchange takes place in the tower where King Henry has been captured by Richard and his followers. It is one of the final interchanges between the two rivals before King Henry’s abdication and his death. Henry reflects on a variety of forms of address - including the use of the name Gloucester modified by the epithet good - and the reasons why the epithet does not seem to fit the formulaic form of address in this case. The extract is from a different play but it shows how sensitive terms of address were or could be in Shakespeare’s plays. <?page no="98"?> Andreas H. Jucker 80 Vocatives are used by speakers to refer directly to the person or thing to whom or to which they are speaking. Quirk et al (1985: 773) define the term “vocative” as follows (see also Biber et al 1999: 1108-13): A vocative is an optional element, usually a noun phrase, denoting the one or more persons to whom the sentence is addressed. (Quirk et al 1985: 773) The most detailed study of vocatives in the plays by William Shakespeare has been carried out by B. Busse (2006). She defines vocatives as follows: [V]ocatives are direct attitudinal adjunct-like forms of address. Realised as a nominal group or head alone, vocatives are optional in form, they may be introduced in Shakespeare by the morphological marker O, and their position may be either initial, middle or final in the clause. (B. Busse 2006: 29) B. Busse (2006: 129) argues that next to their interpersonal and textual potential “Shakespeare’s vocatives, as so-called ‘experiential markers,’ semiotically re-construe and construe experience into meaning and are part of the experiential metafunction of language.” She distinguishes between a large range of different vocatives to account for all occurrences in Shakespeare’s work. She distinguishes between name-based vocatives that are based on modified or unmodified proper nouns and non-name-based vocatives (see also U. Busse and B. Busse 2010: 259). The non-name-based vocatives can be further subdivided into the three categories of those that are related to human qualities, those that are not related to human qualities and a third category which she labels EPITHETS in capital letters to distinguish it from epithets that are used as premodifiers of nouns or names. Figure 1, based on B. Busse (2006: 132), gives an overview of her categorization. <?page no="99"?> Names and terms of address in Shakespeare’s Romeo and Juliet 81 Figure 1: Vocatives in Shakespeare, based on B. Busse (2006: 132). The category of personal names comprises unmodified and modified personal names, such as Benvolio, Balthasar, good Juliet (5.3.159), gentle Romeo (1.4.13). Among the non-name-based names, the category of conventional terms describes those terms that are related to Early Modern English hierarchical society and its social structures. Relevant examples would be sir (1.1.37, 38, 39), my lord (1.2.6), your ladyship (3.5.106). Terms of family relationship are illustrated by examples such as gentle coz (1.5.64) or daughter (3.5.64). Generic terms are those that overtly refer to the male or female genus of the addressee and often to his or her age. Relevant examples are boy, girl, good fellow or man (5.3.1, 1.3.106, 5.3.42, 1.2.44). The category of emotion/ mind, thought refers to expressions that designate an emotion or thought, such as dear love (2.2.136). There are two more categories under the non-name-based vocatives. There are those that refer to specialised fields, such as the legal, the medical, or the metaphysical field, and there are those that refer to natural phenomena, such as sense and perception, parts of the body or food. Some of these terms are used to address humans, e.g. when Capulet addresses his daughter with you green-sickness carrion and you baggage (3.5.156). And some are used to address nonhuman entities as in the following example. <?page no="100"?> Andreas H. Jucker 82 Here’s much to do with hate, but more with love: Why, then, O brawling love, O loving hate, O any thing of nothing first create! O heavy lightness, serious vanity, Misshapen chaos of well-seeming forms, Feather of lead, bright smoke, cold fire, sick health, Still-waking sleep, that is not what it is! This love feel I, that feel no love in this. Dost thou not laugh? (1.1.166-174) In his first substantial speech Romeo addresses the feelings love and hate, and feelings evoked by them. At this point, Romeo has not yet met Juliet and he is characterized as “the lover in love with love (hence largely with himself)” (Evans 2003: 12). EPITHETS, finally, refer to “vocatives that describe a kind of quality already inherent in the semantics of the thing” (B. Busse 2006: 133, 137). As such they denote a quality that is more important than its generic quality. B. Busse gives the example of the vocative whore, which denotes the female sex of the addressee and as such has generic qualities, but the quality of being a prostitute is the more important aspect of the term. Sirrah (4.2.2) or good my friend (5.3.124) are relevant examples in Romeo and Juliet. Table 1 gives B. Busse’s statistics of the frequencies of these categories in Romeo and Juliet. It appears that the conventional terms are the most frequent category. EPITHETS, personal names and specialised fields are next in frequency. Generic terms are the least frequent category. occurrences frequency per 100 words conventional terms 152 0.64 emotion / mind, thought 24 0.10 EPITHET 91 0.38 generic terms 27 0.11 natural phenomena 74 0.31 personal names 91 0.38 specialised fields 81 0.34 terms of family relationships 54 0.23 total 594 2.48 Table 1: Vocative constructions in Shakespeare’s Romeo and Juliet (extracted from B. Busse 2006: 148, Table 3) <?page no="101"?> Names and terms of address in Shakespeare’s Romeo and Juliet 83 In the context of all the other plays by Shakespeare, Romeo and Juliet does not stand out in any way in its use of vocatives. The relative frequency of vocatives in the entire corpus investigated by B. Busse (2006: 146) is 2.31 per 100 words. In Romeo and Juliet it is slightly higher with 2.48 but considerably lower than the 3.38 vocatives per 100 words in Titus Andronicus, the play with the highest frequency of vocatives. Nominal and pronominal terms of address For B. Busse, vocatives are not so much a formal syntactic category of case but a functional category used by speakers to address human or nonhuman entities. Vocatives addressed to humans have been analyzed more widely under the label of (nominal) terms of address, particularly in the context of historical pragmatics (see Mazzon 2010 for an overview). In many contexts they have been shown to reflect complex social and interactional relationships between the speaker and the addressee, and Shakespeare’s work has been found to be particularly rich for such analyses. Shakespeare presents his characters as using an extremely broad and creative range of different nominal terms of address reflecting attitudinal and social relationships between the characters. This usage interacts with his use of the pronominal terms of address, thou and you, for characters who address a single interlocutor. In the thirteenth century, authors writing in English had started to adopt the French usage of using a plural pronoun in certain situations to address a single addressee. Thus the originally plural pronoun ye/ you (henceforth Y 3 ) became ambiguous between its usage for several addressees and the new usage of addressing only a single addressee, and the use of the singular pronoun thou/ thee (henceforth T) became situationally reduced. In certain situations it was no longer appropriate (see, e.g. Burnley 2003 for early evidence of the new usage). Chaucer’s use of the pronominal terms of address has been particularly well researched (see, for instance, Mazzon 2000; Honegger 2003; Jucker 2006; Knappe and Schümann 2006; and for an overview Pakkala-Weckström 2010). Various analyses have been proposed for Chaucer’s usage but generally scholars try to account micropragmatically for the individual choices that Chaucer’s fictional characters have to make when they use either Y or T for a single addressee. Shakespeare wrote his plays some two hundred years later, at a time when the use of T was already the clearly marked form (see Walker 2003, 2007). Shakespeare seems to be using a slightly archaic system which shows clear signs of erosion. This is very clear on the level of case marking. 3 The abbreviations Y and T include the different case forms, i.e. ye, you, your, yours and thou, thee, thy and thine. <?page no="102"?> Andreas H. Jucker 84 Shakespeare no longer consistently distinguishes between the subject and the non-subject forms of the pronouns of address (see U. Busse 1998). In the following speech, which the Nurse addresses to Romeo, she uses the pronouns ye and you indiscriminately both in subject and in non-subject position. In line 138, you is used in subject position while in lines 134 and 135, it is used three times in non-subject position. And in line 136, ye is used both as a subject and as a direct object. Pray you, sir, a word: and as I told you, my young lady bid me inquire you out; what she bid me say, I will keep to myself. But first let me tell ye, if ye should lead her into a fool’s paradise, as they say, it were a very gross kind of behaviour, as they say; for the gentlewoman is young; and therefore, if you should deal double with her, truly it were an ill thing to be offered to any gentlewoman, and very weak dealing. (2.4.134-140) As a result of such morphological inconsistency, it is plausible to assume that Shakespeare also used the pragmatics of the pronoun choice less consistently than Chaucer. Micro-pragmatic, turn-by-turn explanations are more difficult. The system clearly had changed in the two hundred years since Chaucer, but any attempt to ascertain the consistency of either Chaucer’s or Shakespeare’s system presupposes, of course, that we know the details of the underlying system from which the two authors might have deviated. Hope (2003: 73) describes the basic choice between Y and T as follows: The basic factor determining choice of thor ypronoun in Early Modern English is social relationship: th-forms are used down the social hierarchy; y-forms up it. This means that those in authority - kings, lords, husbands, fathers, masters - can use th-forms to those in subordinate roles: subjects, vassals, wives, children, servants. Subordinates use y-forms in return. Social equals usually exchange mutual y-forms in the Early Modern period. (italics original) However, things are often more complex as a brief analysis of the opening scenes of Romeo and Juliet clearly demonstrates. In the first scene Sampson and Gregory, two members of the house of Capulet, address each other with T, but when they encounter Abram and Balthasar, two servants of the house of Montague, Gregory switches to Y to address Abram and receives Y in return. When Benvolio, a Montague, and Tybalt, a Capulet, enter the scene, however, they immediately use T to each other. The tone is hostile and they do not waste any efforts on unnecessary decorum. It is no surprise that Montague, the head of the Veronese family, uses T to Benvolio, one of the members of his household, and receives Y in return (1.1.147-150), but the interactions between Romeo, Montague’s son, and Benvolio are again more complex. They start out with mutual T (1.1.174-180), but when Romeo threatens to leave, Benvolio switches to Y (1.1.187). In the <?page no="103"?> Names and terms of address in Shakespeare’s Romeo and Juliet 85 next scene, however, Benvolio uses T to Romeo and receives Y in return (1.2.44-51, 82-87), but soon afterwards - without discernable reason - Benvolio switches back to Y, too. The interaction between Lady Capulet, the Nurse and Juliet in scene 3 further follows the basic system as described by Hope. Lady Capulet uses T both to the Nurse and to Juliet, and she receives Y from both of them. The Nurse and Juliet, however, mostly use T to each other in spite of their status difference, but on certain occasions the Nurse switches to Y to address Juliet, e.g. when she announces her to her mother (1.3.6) or in a later scene when she calls Juliet away from her interaction with Romeo at the ball (1.5.110). In this case, the switch can easily be explained with the increased formality of the situation. Romeo and Juliet almost exclusively use T for each other except for Juliet, who uses Y to Romeo during their first encounter before they even know each other’s identities (1.5.92-109). Even this brief and cursory analysis shows that in many cases it is difficult to account for pronoun choices on a turn-by-turn basis. Many scholars have, therefore, tried to supplement or even replace the micro-pragmatic account with a more sociolinguistically informed approach focussing on relative frequencies of pronouns in different plays and in particular on the frequencies of the pronouns in specific dyads of characters. Thus, the choice of pronoun is accounted for on the basis of the constellation of a particular pair of interlocutors but not for each individual occurrence of a pronoun of address. The occurrences are accounted for on a sociolinguistic basis on their relative frequencies over longer stretches and often over entire plays (see U. Busse and B. Busse 2010 on historical pragmatic analyses of Shakespeare’s plays and Mazzon 2010 on address terms in general). Stein (2003), for instance, uses a markedness approach to study the pronominal terms of address in Shakespeare’s As You Like It and King Lear, that is to say he combines a approach that establishes general usage patterns with micro-pragmatic analyses that account for deviations from these patterns. He identifies socially defined dyads, such as fathers and daughters, servants and aristocrats, or lovers, and establishes the unmarked use of pronouns for each of them. Against the backdrop of this usage, he then focuses on specific cases of pronoun switches and studies them in detail within their dramatic context and the emotional involvement of the communicating characters. Positive or negative emotions may lead to a switch from an expected T form to a Y form or vice versa. He concludes that the pronoun usage in these two plays is in accordance with the general development that turned the Y pronouns into the default at the beginning of the seventeenth century. Mazzon (2003), in a similar fashion, investigates the terms of address in King Lear, Othello and Hamlet. She also establishes usage patterns for specific types of relationships between characters, such as husband and wife, <?page no="104"?> Andreas H. Jucker 86 father and daughter or the relationship between equals. However, she looks not only at pronominal terms of address, but also considers their interaction with nominal terms of address. She, too, focuses on contexts in which the characters deviate from the expected patterns and use address terms for specific dramatic effects. She concludes that [t]he system of nominal address responds to very subtle socio-pragmatic considerations, and is used as an indicator of relationships and of affective factors, in the same way as the sub-system of second-person pronouns is used. (Mazzon 2003: 240) U. Busse has developed a particularly clear way of visualizing the correlation of pronominal and nominal terms of address. For each individual nominal term of address he establishes how often it co-occurs with a T or a Y pronoun, i.e. its so-called thouor you-fulness. In order to visualize the result, he uses the logarithm of the division of the frequencies of Y and T. If the frequencies are identical, the division equals 1 and the logarithm 0. If the frequency of Y is higher than that of T, the division is larger than one and the logarithm positive, if Y is smaller, the division is smaller than one and the logarithm is negative. A logarithm of +1 indicates that Y is ten times as frequent as T, and a logarithm of -1 that T is ten times as frequent as Y. In order to enhance the visibility of the result, the logarithm is multiplied by 1000 (see U. Busse 2003: 217, fn 11). The results are bar charts with negative and positive bars representing what U. Busse calls youful and thouful terms of address, i.e. terms of address that tend to co-occur more often with you or more often with thou. Figure 2 represents the six categories of terms of address that U. Busse distinguishes and the extent of their thouor you-fulness. The three bars on the left of the diagram indicate negative values. These terms of address cooccur more often with the pronoun thou. Terms of endearment, such as bully, chuck, heart, joy, love or wag (U. Busse 2003: 196) show the highest predominance of thou over you. Terms of abuse, such as devil, dog, fool, knave, rascal, rogue; and generic terms of address, such as boy, friend, gentleman, gentlewoman or lad also co-occur more often with thou than with you but not to the same extent as terms of endearment. The remaining three categories of terms of address co-occur more often with you than with thou in Shakespeare’s work. These are terms indicating family relationships, such as brother, cousin, coz, daughter, father or husband; terms of address indicating occupation, such as captain, doctor, esquire, justice, knight or nurse; and titles of courtesy, such as Your Grace, Your (royal) Highness, Your Honour, Your Ladyship, Goodman, goodwife, lady, lord or sir. <?page no="105"?> Names and terms of address in Shakespeare’s Romeo and Juliet 87 Figure 2: Co-occurrence of pronominal and nominal forms of address in Shakespeare’s plays. Values: 1000 * log(Y/ T) (source: U. Busse 2003: 214) U. Busse observes that the order of these categories reflects Brown and Levinson’s (1987) politeness scale. The titles of honour and courtesy represent the extreme end of negative politeness or deference while the terms of endearment represent the extreme end of positive politeness. The scale in figure 2 suggests that the T pronouns are used together with terms of endearment to express intimacy and together with terms of abuse they express contempt and lack of respect. The predominant cooccurrence with terms of courtesy and occupation, on the other hand, indicates that Y pronouns are used to indicate deference and respect. The fact that terms of family relationship also occur more often with Y pronouns than with T pronouns indicates that solidarity in the sense of Brown and Gilman (1960, 1989) does not automatically call for T pronouns. In figure 3 I have applied U. Busse’s methodology to throw a more focussed light on the six most frequent terms of address in Romeo and Juliet. The Nurse, Romeo and Juliet are more often addressed with T than with Y, and the term villain, which is used to address different characters also co-occurs more often with T than with Y. The terms Madam and Sir, on the other hand, co-occur more often with Y. They are youful terms in U. Busse’s terminology. <?page no="106"?> Andreas H. Jucker 88 Figure 3: Co-occurrence of pronominal forms of address with selected nominal forms of address in Romeo and Juliet. Values: 1000 * log(Y/ T) It is noteworthy that the extreme values in figure 3, both positive and negative, are considerably higher (or lower) than in figure 2. It is plausible to assume that this is the result of the much smaller basis on which these figures have been calculated. In contrast to figure 2, the different bars in figure 3 do not comprise the data of large categories of terms of address, but only the data of single characters (the Nurse, Romeo, Juliet) or one single term of address used to a range of different characters (villain, Madam, Sir). It is also significant that Romeo and Juliet receive most address terms from each other (almost exclusively T) and from the Friar, who also almost exclusively uses T to both of them. The Nurse, on the other hand, mostly addresses Juliet with Y (8 T/ 38 Y) but Romeo’s and Juliet’s thou-fulness depends mostly on their own address terms and those of the Friar. For Romeo it is clearly T which is the unmarked pronoun of address. He uses 118 T pronouns and only seven Y. The Y pronouns are addressed to Juliet (twice), the Nurse (twice), Benvolio (twice) and Mercutio (once). But all four of them receive more T pronouns from Romeo (Juliet 32, the Nurse 13, Benvolio 6 and Mercutio 3). This pattern shows Romeo to be a character who enjoys a sufficiently high social position and who only rarely cares for the social decorum of the Y pronoun. Table 2 gives an overview of the pronoun choices between Romeo, Benvolio, Mercutio and Tybalt. This table once again reveals the multifunctionality of the T pronouns. Benvolio, Montague’s nephew, and Mercutio, a kinsman of the Prince, are both friends of Romeo. Tybalt, <?page no="107"?> Names and terms of address in Shakespeare’s Romeo and Juliet 89 Capulet’s nephew, is one of Romeo’s foes. Romeo addresses all of them mostly with T and only rarely with Y. He never uses the pronoun of respect for Tybalt. His friends also address him mostly with T but they use Y pronouns more often than Romeo. Tybalt only uses T to Benvolio. Benvolio and Mercutio only use T to each other. Among the foes, it is only Mercutio and Tybalt who switch between T and Y to address each other. speaker addressee Romeo Benvolio Mercutio Tybalt Romeo 6 / 2 3 / 1 12 / 0 Benvolio 14 / 8 2 / 0 1 / 0 Mercutio 25 / 6 13 / 0 3 / 4 Tybalt 3 / 0 5 / 0 2 / 5 Table 2: Pronominal terms of address (T/ Y) between friends and foes in Romeo and Juliet Table 3 tabulates the nominal terms of address used by the same four friends and foes of table 2. The results of this table are perhaps somewhat less clear-cut. They indicate that the friends tend to use each other’s names or family relationships (coz, cousin), and they tend to premodify these terms with good, fair or gentle, while the foes either use mostly unadorned names or insults. speaker/ addressee Romeo Benvolio Mercutio Tybalt Romeo my coz, cousin, good fellow Mercutio, gentle Mercutio, man Tybalt, good Capulet Benvolio cousin, coz, fair coz, man, Romeo good Mercutio Mercutio gentle Romeo, Sir, madman, lover, Signor Romeo Good Benvolio, Benvolio Tybalt, you rat-catcher, good king of cats, sir Tybalt Romeo, boy, wretched boy Benvolio Sir Table 3: Nominal terms of address between friends and foes in Romeo and Juliet <?page no="108"?> Andreas H. Jucker 90 It is interesting to compare these usages with the more neutral stance of the Prince. The Prince rarely uses address terms that are modified by adjectives. He addresses his interlocutors mostly with their plain names: Benvolio (3.1.142), Capulet (1.1.90) Montague (1.1.81, 91). In one case, he uses the term old Capulet (1.1.81). For the Prince, names are neutral designations. The adjective old in old Capulet is used to distinguish him from Romeo, the young Capulet. For the other characters, however, the names are not neutral, and appropriate adjectives are used to connote with the emotional value of the names themselves. For Romeo and Juliet, names, terms of address and referring expressions are perhaps even more important. They use a variety of vocatives and nominal designations to address each other, and as referring expressions when talking about each other to other characters. Table 4 lists all the vocatives and all the referring expressions that the two lovers use for each other. Speaker Terms of address Referring expressions Romeo Dear saint Juliet bright angel fair saint love my love, my wife The fair daughter of rich Capulet thy lady and mistress thy lady sweet Juliet Juliet Good pilgrim Romeo gentle Romeo fair Montague, gentleman dear love sweet Montague my Romeo Poor my lord Love, lord, ay, husband, friend Churl My love Romeo My dearer lord Beautiful tyrant, Fiend angelical Dove-feathered raven, wolvish-ravening lamb Despised substance of divinest show A damned saint, an honourable villain My husband That husband My sweet love Table 4: Nominal terms of address and referring expressions used by Romeo and Juliet for each other <?page no="109"?> Names and terms of address in Shakespeare’s Romeo and Juliet 91 In fact, the distinction between terms of address and referring expressions is not always as straight-forward as it might seem. Romeo and Juliet address each other not only when they actually talk to each other but also in their soliloquies, for instance in act 5 when Romeo encounters what he believes to be Juliet’s dead body (5.3.74-120) and when Juliet wakes up only to find her husband dead (5.3.160-170). But Juliet also seems to talk to her husband when she learns from the Nurse that he has killed Tybalt (see below). In such passages it is not entirely clear whether the nominal designations are used as address terms or as referring expressions. Discussion and conclusion In the light of the above considerations, I shall discuss two extracts from Romeo and Juliet in some more detail. In the first scene of act 3, Benvolio and Mercutio encounter Tybalt. After a verbal fight, Romeo enters and tries to defuse the situation. Because of his secret love to Juliet he no longer pursues the struggle between the houses of Montague and Capulet. He very clearly hints at his changed feelings to the house of Capulet but his allusions are not understood in the intended sense. Enter ROMEO TYBALT Well, peace be with you, sir, here comes my man. MERCUTIO But I’ll be hanged, sir, if he wear your livery. Marry, go before to field, he’ll be your follower; Your worship in that sense may call him man. TYBALT Romeo, the love I bear thee can afford No better term than this: thou art a villain. ROMEO Tybalt, the reason that I have to love thee Doth much excuse the appertaining rage To such a greeting. Villain am I none; Therefore farewell, I see thou knowest me not. TYBALT Boy, this shall not excuse the injuries That thou hast done me, therefore turn and draw. ROMEO I do protest I never injured thee, But love thee better than thou canst devise, Till thou shalt know the reason of my love; And so, good Capulet, which name I tender As dearly as my own, be satisfied. <?page no="110"?> Andreas H. Jucker 92 MERCUTIO O calm, dishonourable, vile submission! ‘Alla stoccata’ carries it away. [Draws.] Tybalt, you rat-catcher, will you walk? TYBALT What wouldst thou have with me? MERCUTIO Good King of Cats, nothing but one of your nine lives that I mean to make bold withal, and as you shall use me hereafter, dry-beat the rest of the eight. Will you pluck your sword out of his pilcher by the ears? Make haste, lest mine be about your ears ere it be out. TYBALT I am for you. [Drawing.] (3.1.49-75) At the beginning of this extract, Mercutio and Tybalt address each other with Y pronouns and with the honorific sir, which is Shakespeare’s most frequent vocative and used here as a neutral and socially safe term of address. The social decorum is still intact, but it immediately breaks down when Tybalt addresses Romeo. He switches to T and calls him a villain, which is a very serious insult “carrying not only the sense of ‘depraved scoundrel’ but undertones of ‘low-born fellow’ (=villein)” (Evans 2003: 137, footnote to line 3.1.54). The word love in line 53 is highly ironic, and some editors prefer the reading hate from the first Quarto because such irony seems untypical of Tybalt (Evans 2003: 137, footnote to line 3.1.53). Romeo in his attempt to defuse the situation ignores the insult and tries to end the precarious interaction, but he also uses T to address Tybalt, possibly because he feels closer to Tybalt than Tybalt can understand at this point, but Tybalt is either not placated by Romeo’s address or he takes the T pronoun as picking up the challenge he presented with his own T and the term villain. Tybalt sticks to T and addresses Romeo as boy, a further insult (see Crystal and Crystal 2002: 52). In his response, Romeo uses five T pronouns and addresses Tybalt as good Capulet. These signs are now clear enough for Mercutio, who is incensed by what he perceives as Romeo’s submission to Tybalt’s insults. So he verbally attacks Tybalt. He calls Tybalt alla stoccata, which literally means ‘at the thrust’ and refers to an Italian type of fencing despised by Mercutio (Evans 2003: 138, footnote to line 3.1.67). He addresses him as a cat with the terms rat-catcher and good King of Cats (according to Evans 2003: 119, 138 footnotes to lines 2.4.18 and 3.1.68, an allusion to a fictional cat called Tybert or Tibault) but, perhaps surprisingly, he sticks to the formal Y pronoun. In return Tybalt, who accepts the challenge for a fight, also addresses him with Y. <?page no="111"?> Names and terms of address in Shakespeare’s Romeo and Juliet 93 From this interaction it becomes clear why it is problematic to account for pronoun choices on a strict turn-by-turn basis. We might have expected Romeo to use a Y pronoun to show his new found respect and “love” for the house of Capulet and, therefore, for Tybalt. But as I have shown above, T is the unmarked choice for Romeo from which he does not deviate at this point. At the same time, we might have expected Mercutio and Tybalt to use T as an expression of their contempt for each other. The full significance of T and Y pronouns in this play can, therefore, not be gleaned easily from such examples. The broader perspectives provided above in which pronominal terms of address are correlated with nominal terms of address and tabulated for significant pairs of interlocutors or for specific categories of nominal terms of address provide a more comprehensive picture. The second scene of act two, following the scene from which the previous extract has been taken, is a turning point of the play. Juliet has been married to Romeo for only a few hours, and now she is to learn that her husband has killed Tybalt, one of her cousins. At the beginning of the scene, Juliet, who is still unaware of the events, awaits the Nurse’s report on Romeo. However, on her return the Nurse is so overwhelmed with the calamity of the news she has to convey that she presents a confusing and ambiguous account of the events so that it takes Juliet several turns to find out what has happened. When Juliet finally grasps the impact of the news, she is torn in her feelings. NURSE Tybalt is gone and Romeo banishèd. Romeo that killed him, he is banishèd. JULIET O God, did Romeo’s hand shed Tybalt’s blood? NURSE It did, it did, alas the day, it did! JULIET O serpent heart, hid with a flow’ring face! Did ever dragon keep so fair a cave? Beautiful tyrant, fiend angelical! Dove-feathered raven, wolvish-ravening lamb! Despisèd substance of divinest show! Just opposite to what thou justly seem’st, A damnèd saint, an honourable villain! O nature, what hadst thou to do in hell When thou didst bower the spirit of a fiend In mortal paradise of such sweet flesh? Was ever book containing such vile matter So fairly bound? O that deceit should dwell In such a gorgeous palace! (3.2.71-84) <?page no="112"?> Andreas H. Jucker 94 Juliet uses a string of pairs of incompatible referring expressions that are incompatible in themselves. Romeo is both a beautiful tyrant and an angelical fiend; he is both a raven and a lamb, but a dove-feathered raven and wolvish-raving lamb, and even a damned saint and honourable villain. Thus, Juliet’s conflicting feelings find a very direct expression in the incongruity and incompatibility of the terms that she uses to refer to the man she married only a few hours ago. Romeo is not present in this scene, but Juliet’s words seem to be addressed to him. Later in the scene she already regrets her words (“O what a beast was I to chide at him! ” 3.2.94) but the wording suggests that the words were addressed to her husband. It is, therefore, not quite clear whether these designations are terms of address or referring expressions. But they indicate once again the power of names (in the wider sense of the word). The two passages, furthermore, corroborate U. Busse’s claim about Shakespeare’s use of nominal and pronominal terms of address. The paper has shown how Shakespeare uses both nominal and pronominal address in his plays and that a correlation between these categories in terms of variable rules can be established. However, it has also proved that pronoun use is not fully predictable, because on the micro level of analysis apart from intersocial relationships other factors have to be taken into account. Obviously, Shakespeare must have been well aware of the social conventions of the day, and he surely exploited them skilfully for dramatic purposes. Nonetheless, on the basis of this investigation we can only construct a “social grammar” of Shakespeare, but we should not conclude that the language of drama with its carefully constructed speeches bore any close resemblance to real people talking, because it is not always possible to take such renditions at their face value. (U. Busse 2003: 216) It is certainly important to repeat the point made at the beginning of this paper that Shakespeare’s use of terms of address do not necessarily reflect the usage in everyday conversations at the turn from the sixteenth to the seventeenth century. The basis of our analysis is a fictional world, albeit a highly complex and intricate one. Even if we agree that Shakespeare exploited these resources “skilfully for dramatic purposes”, it is not always straightforward to interpret these purposes. The linguist’s striving for generalizations of usage patterns may be seen to clash with the reader’s or the literary scholar’s wish to interpret each passage individually taking into account the unique constellation of emotional involvement of the characters at any particular point in a play. <?page no="113"?> Names and terms of address in Shakespeare’s Romeo and Juliet 95 Biographical sketch Andreas H. Jucker is Professor of English Linguistics at the University of Zurich. His current research interests include historical pragmatics, speech act theory, politeness theory and the grammar and history of English. Recent publications include Diachronic Perspectives on Address Term Systems (Benjamins, 2003), Speech Acts in the History of English (Benjamins, 2008), both co-edited with Irma Taavitsainen, Early Modern English News Discourse (Benjamins, 2009) and Corpora: Pragmatics and Discourse (Rodopi, 2009) co-edited with Daniel Schreier and Marianne Hundt, and most recently the Handbook of Historical Pragmatics (Mouton, 2010) co-edited with Irma Taavitsainen and Communicating Early English Manuscript (Cambridge University Press, 2011) co-edited with Päivi Pahta. He is the editor of the Journal of Historical Pragmatics (with Irma Taavitsainen) and he is the Associate Editor of the Pragmatics & Beyond New Series (Benjamins). References Primary Text Shakespeare, William. 2003. Romeo and Juliet. Updated edition. (Edited by G. Blakemore Evans). Cambridge: Cambridge University Press. Secondary Literature Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan. 1999. Longman Grammar of Spoken and Written English. London: Longman. Blake, Norman. 2002. “Forms of Address in Hamlet”. In Katja Lenz and Ruth Möhlig (eds.). Of dyuersitie & chaunge of language. Essays Presented to Manfred Görlach on the Occasion of his 65th Birthday. Heidelberg: C. Winter, pp. 305- 318. Brown, Roger, and Albert Gilman. 1960. “The Pronouns of Power and Solidarity”. In Thomas A. Sebeok (ed.). Style in Language. Cambridge, Mass.: MIT Press, pp. 253-276. Brown, Roger, and Albert Gilman. 1989. “Politeness Theory and Shakespeare’s Four Major Tragedies”. Language in Society 18.2, pp. 159-212. Brown, Penelope, and Stephen C. Levinson. 1987. Politeness. Some Universals in Language Usage. (Studies in Interactional Sociolinguistics 4). Cambridge: Cambridge University Press. <?page no="114"?> Andreas H. Jucker 96 Burnley, David. 2003. “The T/ V Pronouns in Later Middle English Literature”. In Irma Taavitsainen and Andreas H. Jucker (eds.). Diachronic Perspectives on Address Term Systems. (Pragmatics & Beyond New Series 107). Amsterdam and Philadelphia: John Benjamins, pp. 27-45. Busse, Beatrix. 2006. Vocative Constructions in the Language of Shakespeare. (Pragmatics & Beyond New Series 150). Amsterdam and Philadelphia: John Benjamins. Busse, Ulrich. 1998. “‘Stand, sir, and throw us that you have about ye’. Zur Grammatik und Pragmatik des Anredepronomens ye in Shakespeares Dramen”. In Eberhard Klein and Stefan J. Schierholz (eds.). Betrachtungen zum Wort. Lexik im Spannungsfeld von Syntax, Semantik und Pragmatik. Tübingen: Stauffenburg Verlag, pp. 85-115. Busse, Ulrich. 2003. “The Co-occurrence of Nominal and Pronominal Address Forms in the Shakespeare Corpus: Who Says Thou or You to Whom? ” In Irma Taavitsainen and Andreas H. Jucker (eds.). Diachronic Perspectives on Address Term Systems. (Pragmatics & Beyond New Series 107). Amsterdam and Philadelphia: John Benjamins, pp. 193-221. Busse, Ulrich, and Beatrix Busse. 2010. “Shakespeare”. In Andreas H. Jucker and Irma Taavitsainen (eds.). Historical Pragmatics. (Handbooks of Pragmatics 8). Berlin and New York: De Gruyter Mouton, pp. 247-281. Crystal, David, and Ben Crystal. 2002. Shakespeare’s Words. A Glossary and Language Companion. London: Penguin. Culpeper, Jonathan, and Merja Kytö. 2010. Early Modern English Dialogues. Spoken Interaction as Writing. (Studies in English Language). Cambridge: Cambridge University Press. Evans, G. Blakemore. 2003. “Introduction”. In G. Blakemore Evans (ed.). Romeo and Juliet. Updated edition. Cambridge: Cambridge University Press, pp. 2- 62. Honegger, Thomas. 2003. “‘And if ye wol nat so, my lady sweete, thanne preye I thee, [...].’: Forms of Address in Chaucer’s Knight’s Tale”. In Irma Taavitsainen and Andreas H. Jucker (eds.). Diachronic Perspectives on Address Term Systems. (Pragmatics & Beyond New Series 107). Amsterdam and Philadelphia: John Benjamins, pp. 61-84. Hope, Jonathan. 2003. Shakespeare’s Grammar. London: The Arden Shakespeare. Jucker, Andreas H. 2006. “‘Thou art so loothly and so oold also’: The Use of Ye and Thou in Chaucer’s Canterbury Tales”. Anglistik 17.2, pp. 57-72. Knappe, Gabriele, and Michael Schümann. 2006. “Thou and Ye: A Collocationalphraseological Approach to Pronoun Change in Chaucer’s Canterbury Tales”. Studia Anglica Posnaniensia 42, pp. 213-238. Kytö, Merja. 2010. “Data in Historical Pragmatics”. In Andreas H. Jucker and Irma Taavitsainen (eds.). Historical Pragmatics. (Handbooks of Pragmatics 8). Berlin and New York: De Gruyter Mouton, pp. 33-67. Mazzon, Gabriella. 2000. “Social Relations and Form of Address in the Canterbury Tales”. In Dieter Kastovsky and Arthur Mettinger (eds.). The History of English in a Social Context. A Contribution to Historical Sociolinguistics. Berlin: Mouton de Gruyter, pp. 135-168. <?page no="115"?> Names and terms of address in Shakespeare’s Romeo and Juliet 97 Mazzon, Gabriella. 2003. “Pronouns and Nominal Address in Shakespearean English: A Socio-affective Marking System in Transition”. In Irma Taavitsainen and Andreas H. Jucker (eds.). Diachronic Perspectives on Address Term Systems. (Pragmatics & Beyond New Series 107). Amsterdam and Philadelphia: John Benjamins, pp. 223-249. Mazzon, Gabriella. 2010. “Terms of Address”. In Andreas H. Jucker and Irma Taavitsainen (eds.). Historical Pragmatics. (Handbooks of Pragmatics 8). Berlin and New York: De Gruyter Mouton, pp. 351-376. Pakkala-Weckström, Mari. 2010. “Chaucer”. In Andreas H. Jucker and Irma Taavitsainen (eds.). Historical Pragmatics. (Handbooks of Pragmatics 8). Berlin and New York: De Gruyter Mouton, pp. 219-245. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman. Stein, Dieter. 2003. “Pronominal Usage in Shakespeare: Between Sociolinguistics and Conversational Analysis”. In Irma Taavitsainen and Andreas H. Jucker (eds.). Diachronic Perspectives on Address Term Systems. (Pragmatics & Beyond New Series 107). Amsterdam and Philadelphia: John Benjamins, pp. 251-307. Walker, Terry. 2003. “You and Thou in Early Modern English Dialogues: Patterns of Usage”. In Irma Taavitsainen and Andreas H. Jucker (eds.). Diachronic Perspectives on Address Term Systems. (Pragmatics & Beyond New Series 107). Amsterdam and Philadelphia: John Benjamins, pp. 309-342. Walker, Terry. 2007. Thou and You in Early Modern English Dialogues. Trials, Depositions, and Drama Comedy. (Pragmatics & Beyond New Series 158). Amsterdam and Philadelphia: John Benjamins. <?page no="117"?> Dieter Kastovsky Recursiveness in Word-Formation Grammatical, Conceptual and Typological Factors 1 Abstract The notion of recursiveness originated in mathematics and was introduced into linguistics by Chomsky in order to account for the fact that on the basis of a finite inventory of elements and rules an infinite number of sentences can be generated by using part of the output of a rule as input of the same type of rule. In morphology (inflectional and derivational) a more precise definition is required. One has to distinguish between the repetition of the same morphological process such as compounding, prefixation, suffixation, but with different morphological material, and the repetition of the same morphological material. The second type of recursiveness, the repetition of the same morphological material within the same morphological process, is the limiting case of the former. It will therefore be subject to the same general constraints as the more general type of recursiveness, but in addition it will also be subject to more specific constraints that may involve features not relevant for recursiveness in the more general sense. 1 Recursiveness The notion of recursiveness - the reapplication of an operation to its output - originated in mathematics. It was introduced into linguistics by Chomsky (1957, 1965) in order to account for the fact that on the basis of a finite inventory of elements and rules an infinite number of sentences can be generated and that there was no longest sentence in a language. This was guaranteed by the fact that the initial symbol S (“Sentence”), which by phrase structure rules was expanded into sentence constituents, could occur both as input and output of these phrase structure rules, as the following very much simplified examples illustrate: 1 An earlier version of this paper was presented at the workshop “Recursitivity in wordformation” organised by Pavol Štekauer at the Meeting of the Societas Linguistica Europaea, Lisbon, September 2009. Since Andreas Fischer has always also been interested in morphological-lexicological problems, I hope this is a small tribute to this part of his many linguistic interests. <?page no="118"?> Dieter Kastovsky 100 (1) S NP + VP a. NP NP + S b. NP Det + N + S Thus, in (1 a), NP is recursively expanded into a sequence NP + S resulting in a relative clause, and in (1 b) NP is recursively expanded into Det + N + S resulting in a complement clause. Both rules were unrestricted from a grammatical point of view, although it was of course recognised that there were performance restrictions. But from a purely grammatical point of view there was no limit to the embedding of relative clauses and complement clauses. Since in this framework other recursive structures, e.g. multiple adjectival modifications such as (2) this little decrepit red old house or compounds like (3) singer-composer-director were transformationally derived from structures like (1 a), no other type of recursiveness was deemed necessary in this early generative-transformtional framework: syntax was recursive, and morphological recursive structures were derived from recursive syntactic structures. This transformational hypothesis was given up in connection with the emergence of X-bar theory (cf. Chomsky 1970) and the subsequent developments, which separated syntax from the lexicon, introducing wordformation processes into the lexicon (the “lexicalist hypotheses”). As a consequence, recursiveness had to be introduced also at the major category level by postulating rules such as (4) a. N A + N b. N N + N in order to account for adjective stacking (cf. 2), compounding (cf. 3), and other similar processes. But recursiveness was usually still regarded as part of syntax, whether at the sentence or the phrase level syntax, whereas its role in morphology - both inflectional and derivational - had not really been systematically investigated. This is why its application in this domain needs a more precise definition. In a relatively general sense, recursiveness might just refer to the repetition of the same morphological process such as compounding, prefixation, suffixation, but with different morphological material, as, e.g. in (5) for compounding, (6) for prefixation, and (7) for both preand suffixation. <?page no="119"?> Recursiveness in Word-Formation 101 (5) a. The clear dome of the Intergalactic Mining Company employee recreation hall shone black around and above them, silvered at its crossbars by the pale glow of Earth’s single moon […] (L. Ron Hubbard. 1982. Battlefield Earth. A Saga of the Year 3,000. Redhill: New Ear Publications, p. 1) b. As they approached the site of the Sling, Siobhan stood up in the tractor’s bubble-dome pressure compartment to see better. (Arthur C. Clarke and Stephen Baxter. 2005. Sunstorm. New York: Del Rey, p. 143) (6) pre-post-modernism period (7) anti-dis-establish-ment-arian-ism In fact, with a somewhat generous interpretation, one might even interpret the sequential stacking of inflectional suffixes in agglutinative languages such as Finnish or Hungarian as a kind of recursiveness in this sense, cf. also the following English and German examples: (8) a. ox-en (Pl.) -s (Gen.) b. Kind-er (Pl.) -n (Dat.) In a more restricted interpretation of recursiveness, only cases of the repetition of the same morphological material would be interpreted as recursive, as, e.g., in (9) re-re-…write, post-post-…post-modernism, great-great-...grandfather, G Urur-ur...großvater It could be argued that the same phenomenon is also responsible for identical constituent compounds - see Hohenhaus (2004) and also the following examples: (10) a. The Imperial government was determined that only one personnel on any planet would be able to build a teleportation console for use in times of emergency or to repair one. So they specially trained these groups of students. We used to call them ‘brain-brains’ (L. Ron Hubbard. 1982. Battlefield Earth. A Saga of the Year 3,000. Redhill: New Ear Publications, p. 1046) b. Almost a year earlier, during alcoholic talk-talk in a hotel room, Mike had promised us one chance in seven… (Robert A. Heinlein. 1966. The Moon is a Harsh Mistress, New York: Tom Doherty Associations, p. 167) <?page no="120"?> Dieter Kastovsky 102 c. Many comrades turned out to be talk-talk soldiers (Robert A. Heinlein. 1966. The Moon is a Harsh Mistress, New York: Tom Doherty Associations, p. 198) I would like to suggest that the second type of recursiveness, the repetition of the same morphological material within the same morphological process, is the limiting case of the former. It will therefore be subject to the same general constraints as the more general type of recursiveness, but in addition it will also be subject to more specific constraints that may involve features not relevant for recursiveness in the more general sense. 2 Discussion Let me begin the more detailed discussions of these two notions of recursion by looking at the more general notion of recursiveness, i.e. the repetition of the same morphological process, but with different morphological material. I will restrict myself to data from English and, to a certain extent, German, for practical reasons. The most conspicuous example of recursiveness of this type seems to be N + N compounding. German in this respect is well-known, cf. the proverbial (11) Donau-dampf-schiff-fahrts-gesellschafts-kapitäns-witwen-pensionskürzungs-verfahren ‘proceedings concerning the cuts of the pension of the widow of the captain of the company running travels of ships powered by steam on the Danube’ Thus, there does not seem to be any limit to the number of constituents, i.e. theoretically there is not any longest compound in German, and N N + N seems to be truly recursive. As for English, however, Marchand (1960) argues that there are no compounds containing more than three lexemes. This is based on the assumption that in English stress distinguishes between compounds (one main stress on the first element), and syntactic groups (level stress or end stress), cf. the contrast between íce crèam and góld wátch. This assumption, however, is very controversial and is rejected by many specialists in the field, since there is no hard and fast line between these formations and often both options exist - as with ice cream (cf., e.g. Bauer 1998, 2001; Giegerich 2004, 2006, 2009; Plag 2006, 2010; Plag et al. 2008). Thus, if we disregard this distinction and treat level stress combinations also as compounds, formations such as those in (5) would also count as compounds, and English thus behaves like German in this respect, although maybe there is a difference of frequency: such multimember compounds seem to be more current in German than in English, but obviously not impossible in the latter. <?page no="121"?> Recursiveness in Word-Formation 103 If we look at the grammatical-semantic-cognitive analogues of nominal compounds, this should not be surprising. In early generative-transformational grammar, such compounds were treated as reduced relative clauses, and stacking of relative clauses was permitted in syntax (cf. above). In a less formal analysis, these compounds can also be analysed (paraphrased) as parallel to a stacking of prepositional phrases, e.g. (12) Mining Company employee recreation hall = hall for the recreation of employees of the company for mining which again is a permissible grammatical structure. And cognitively, it is the progressive narrowing of the concept ‘hall’ to a very specific subcategory, where each constituent narrows the result of the previous specification - again, from a cognitive point of view, without a limit, except for practical purposes, viz. how specific do you want to be: thus here the constraint would basically only be a pragmatic one. But now take a compound of the type V - N like (13) rattlesnake, whetstone, bakehouse, washday, drawbridge Here, the first constituent is a verb, and the second constituent can be regarded as some kind of semantic argument of the verb (agent, patient, instrument, place, etc., or, depending on the framework used, a syntactic argument such as subject, object, adverbial); the term is used here in a rather loose sense. 2 Obviously, the verb by virtue of its grammaticalsemantic properties only allows a certain number of such arguments, and also only one of each kind, unless the arguments are conjoined. Here no recursiveness is possible for grammatical reasons. The same applies to the type writing table, baking powder, dancing girl, which is a morphological variant of the previous, the verb here having the -ing-form, which may be interpreted as participial or gerundial. The situation with Adj + N combinations is peculiar. The syntactic pattern seems to be freely recursive (cf. 2), but neither in English nor in German do we observe stacked adjectival modifiers in Adj + N compounds. Only one adjective occurs, cf. (14) blackbird, wildfire, mainland, freshman, hothouse, etc. This is somewhat puzzling in view of the fact that Adj + N compounds seem to have a relative clause analogue like the N + N compounds, cf. the earlier generative-transformational handling of these structures, deriving them from relative clauses. I do not have an explanation for this different 2 Whether argument structure or thematic structure or both should be used in the description of word-formation processes is a matter of controversy, which, however, would go beyond the present topic and will therefore not be further pursued. <?page no="122"?> Dieter Kastovsky 104 behaviour of free syntactic phrases, which allow adjective stacking, and compounds which do not. But it might have something to do with the cognitive aspect of word-formation, which creates names, and designations containing more than one property modifier - assuming that adjectives denote properties - may not be acceptable, whereas the function of N as a modifier is much wider and therefore allows stacking. No recursiveness seems to occur with adjectival compounds such as colourblind, grass-green, icy-cold, heart-breaking, man-made as well. This, again, may have to be related to the grammatical make-up of these formations: the first constituent can be regarded as filling an “argument” slot of the adjective, and only one slot is available. Thus no recursiveness of the first constituent is possible. And recursiveness of the second is conceptually not conceivable, cf. *grass-green-blue, *heart-breaking-rending. Let me now turn to prefixes. From a grammatical-functional point of view, prefixes are equivalent to adjectives, adverbs and prepositions, cf. (15) a. adjective: ex-husband ‘former (= ex-) husband’, co-author ‘joint (= co-) author’ b. adverb: pre-cook ‘cook before (pre-)’, hyper-accurate ‘overly (= hyper-) accurate’, re-write ‘write again (re-)’, un-fair ‘not (= un-) fair’ c. preposition: anti-American ‘against (against-) America’, international ‘between (= inter-) nations’ With (15 a), recursiveness seems to be possible, although it might require some additional pragmatic context. Thus, if a bigamist woman gets divorced, she might speak of her ex-co-husbands, and if a writing-team breaks up, the writers might refer to themselves as ex-co-writers. I have not actually encountered such examples, but they seem possible. Here, I think, the restrictions are rather conceptual-referential, but not grammatical. It is interesting, however, that in contradistinction to the restriction with real adjective compounds, which do not seem to allow recursiveness, the adjective-related prefixes do so. With (15 b), i.e. prefixes in adverbial function, we have the following situation. These prefixes modify adjectives and verbs. And again there are instances where recursiveness is possible and instances where it is not. With examples such as re-re-re-write, denoting repetition, recursiveness is possible, whereas with degree-denoting prefixes such as hyper-active, overanxious it seems to be excluded. This looks like a semantic-conceptual rather than a strictly grammatical constraint. With (15 c), i.e. prefixes having prepositional force, it would seem that recursiveness is not possible, since a preposition only allows one comple- <?page no="123"?> Recursiveness in Word-Formation 105 ment, and the complement can only be linked to one preposition. Moreover, such formations might be semantically contradictory: for formations such as post-pre-war period, pre-post-war-period it would be rather difficult to find a plausible semantic interpretation. Stacking of prefixes with different functions is possible, however, cf. (16) anti- (Prep) de- (Adv) militarisation campaign With suffixes, recursiveness in this general sense is common, cf. formations such as (17). But there are suffix-specific restrictions, i.e. some suffixes allow further suffixation, e.g. -al, -ise: nation-al-is-ation, -less: hair-less-ness, whereas others do not, e.g. -ness (*idle-ness-less), -ity (*impeccability-less). (17) -ise: nation-al-is-ation, -less: hair-less-ness, -ness: *idle-ness-less, -ity: *impeccabil-ity-less Whether these are purely morphological restrictions, or whether at least some of them are also semantically-pragmatically based, requires further investigation. The examples in (17) point to a purely morphological constraint on -less, since the semantically roughly equivalent un-idle-ness ‘state of not being idle’, un-impeccabil-ity ‘state of not being impeccable’ exist. One area where recursiveness seems to be fairly widespread is diminutives, usually in combination with a hypocoristic shade. Thus, in German we find Bub ‘boy’> Bub-i ‘dear boy’ > Bub-i-lein ‘dear little boy’. No such stacking seems to be possible in English, but it is very common in the Romance and the Slavic languages, cf. Sp. chica ‘girl’: chiqu-ita ‘little girl’: chiqu-it-ita ‘very little girl’, Pol. dziewcznyna ‘girl’: dziewcznyn-k-a ‘little girl’: dziewcznyn-escz-k-a ‘very little girl’; this process even applies to adjectives, cf. mał-y ‘small’: mal-utk-i ‘very small’: mal-ut-enk-i ‘very, very small’ (cf. Dressler and Merlini-Barbaresi 1994; Kryk-Kastovsky 2000). 3 Conclusion Let me finally look at recursiveness in the narrow sense, i.e. the repetition of identical elements. As we have already seen, such formations occur, but they seem to be tied to specific semantic domains such as repetition, emphasis or diminutives, i.e. they are controlled by cognitive-semantic factors. Thus, the examples in (10) are emphatic in the sense that they specify that the denotatum of the second element is a real, most prototypical representative of the category in question - see also the recent German coinage Film-film, referring to a real movie, and not a television movie. Similarly, the example re-re-re-write not only denotes multiple repetition of the same activity, but also conveys emphasis, and in fact seems to be infinitely recursive, since conceptually speaking the action allows indefinite repetition. <?page no="124"?> Dieter Kastovsky 106 A similar situation obtains with preposed elements (either lexemes or prefixes) denoting a genealogical relationship, such as great-great-great- …grand-father, G Ur-ur-ur…grossvater, etc., which also are, at least theoretically, infinitely recursive, because there is no conceptual limit to the number of possible ancestors. Again, it is the conceptual-pragmatic situation that allows indefinite recursion in such cases. This brief survey has, I hope, shown that recursiveness in wordformation needs a more specific definition and an attempt to distinguish different types of recursiveness. Moreover, constraints on recursiveness involve grammatical factors, e.g. the properties of syntactically equivalent structures, but also morphological properties of the constituents, typological features of the languages in question, as well as semantic-conceptual, pragmatic and cognitive factors. Biographical sketch Dieter Kastovsky was born on December 26 th , 1940 in Freudenthal (Bruntal), Germany (now Czech Republic). He studied English, Romance, German and General Linguistics in Tübingen, Berlin (Free University) and Besançon (France). He received his Dr. Phil. in 1967 from the University of Tübingen. 1967-1973 he was research assistant at the English Department of the University of Tübingen. 1973-1981 he was Full Professor of English Linguistics at the University of Wuppertal. From 1981-2009 was Full Professor of English Linguistics at the University of Vienna; 1986- 1990, 1994-2004 he was also Head of Department of the Department of Interpreting and Translation at the University of Vienna, 1990-1992 Head of the English Department, 2004-2006 Director of the Centre for Translation Studies, University of Vienna. From 1991 till September 2006 he was Secretary and Treasurer of Societas Linguistica Europaea. He is a member of the ‘Sudetendeutsche Akademie der Wissenschaften und Künste’, Munich and a Fellow of the English Association, London. Since November 2009 he is Professor Emeritus at The University of Vienna and teaches as Professor at the Academy of Management, Warsaw Branch, School of English, Warsaw. His main research interests are English morphology and wordformation (synchronic and diachronic), semantics, history of linguistics, language typology. <?page no="125"?> Recursiveness in Word-Formation 107 References Alonso, Roberto Torre. 2010. “Morphological Process Feeding in the Formation of Old English Nouns”. SKASE Journal of Theoretical Linguistics 7.1, pp. 57- 69. Arista, Javier Martín. 2010. “Old English Strong Verbs Derived from Strong Verbs: Affix Variation, Grammaticalisation and Recursivity”. SKASE Journal of Theoretical Linguistics 7.1, pp. 36-56. Bauer, Laurie. 1998. “When is a Sequence of Two Nouns a Compound in English? ” English Language and Linguistics 2, pp. 65-86. Bauer, Laurie. 2001. “Compounding”. In Martin Haspelmath, Ekkehard König, Wulf Oesterreicher and Wolfgang Raible (eds.). Language Typology and Language Universals / Sprachtypologie und sprachliche Universalien / La typologie des langues et les universaux linguistiques. An International Handbook / Ein internationales Handbuch / Manuel international. Volume / Halbband / Tome 1. Berlin and New York: Walter de Gruyter, pp. 695-707. Bisetto, Antonietta. 2010. “Recursiveness and Italian Compounds”. SKASE Journal of Theoretical Linguistics 7.1, pp.14-35. Chomsky, Noam. 1957. Syntactic Structures. Mouton: The Hague. Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge, Massachussetts: M.I.T. Press. Chomsky, Noam. 1970. “Remarks on Nominalisation”. In Roderick A. Jacobs and Peter S. Rosenbaum (eds.). Readings in English Transformational Grammar. Waltham, Massachussetts: Ginn, pp. 184-221. Dressler, Wolfgang U. and Lavinia Merlini Barbaresi. 1994. Morphopragmatics. Diminutives and Intensifiers in Italian, German and Other Languages. Berlin: Mouton de Gruyter. Giegerich, Heinz. J. 2004. “Compound or Phrase? English Noun-plus-noun Constructions and the Stress Criterion”. Journal of English Language and Linguistics 8, pp. 1-24. Giegerich, Heinz. J. 2006. “Attribution in English and the Distinction between Phrases and Compounds”. In Petr Rösel (ed.). Englisch in Zeit und Raum - English in Time and Space. Forschungsbericht zu Ehren von Klaus Faiß. Trier: Wissenschaftlicher Verlag Trier, pp. 10-27. Giegerich, Heinz J. 2009. “The English Compound Stress Myth”. Word Structure 2, pp. 1-17. Hohenhaus, Peter. 2004. “Identical Constituent Compounding - a Corpus-based Study”. Folia Linguistica 38, pp. 297-331. Kryk-Kastovsky, Barbara. 2000. “Diminutives: An Interface of Word-formation, Semantics and Pragmatics”. In Christiane Dalton-Puffer and Nikolaus Ritt (eds.). Words: Structure, Meaning, Function. A Festschrift for Dieter Kastovsky. (Trends in Linguistics 130). Berlin: Mouton de Gruyter, pp. 165-174. Marchand, Hans. 1960. “Die Länge englischer Komposita und die entsprechenden Verhältnisse im Deutschen”. Anglia 78, pp. 411-416. Plag, Ingo. 2006. “The Variability of Compound Stress in English: Structural, Semantic, and Analogical Factors”. English Language and Linguistics 10, pp. 143-172. <?page no="126"?> Dieter Kastovsky 108 Plag, Ingo. 2010. “Compound Stress Assignment by Analogy: The Constituent Family Bias”. Zeitschrift für Sprachwissenschaft 29, pp. 243-282. Plag, Ingo, Gero Kunter, Sabine Lappe and Maria Braun. 2008. “The Role of Semantics, Argument Structure, and Lexicalization in Compound Stress Assignment in English”. Language 84, pp. 760-794. <?page no="127"?> Hans-Peter Naumann English Keywords in Old Norse Literature Abstract This contribution deals with cultural historical semantics. By first discussing the English influence on Old Norse in general, we argue that neologisms of a certain period (i.e. loanwords per se) reveal information about a speech comunity’s attitudes and cultural values. The subsequent discourse analysis summarises a set of criteria with which to determine culturally embedded ‘keywords’ from different sources, mainly saga literature, but also ecclesiastical scripture as well as eddic and scaldic poetry. As non-literary sources, runic inscriptions have been included. Selected keywords, both with highly positive and negative connotations are explored in greater depth in their respective linguistic contexts. 1. Introduction In his introduction to the Altnordische etymologische Wörterbuch, Jan de Vries (1977) presents word lists providing a broad outline of the origin and frequency of loanwords in Old Norse. An overview not unexpectedly shows a significant dominance of Middle Low German (52%), but already in second place one finds English (21%), followed by Latin (11%) and French (7%). Further languages like Irish/ Scottish (3.5%), Slavic, Dutch (both less than 2%) and others give interesting information about language contact at that time, but only peripherally contribute to the extension of the Old Nordic lexicon. Even though the lists in de Vries (pp. XIII-XLI) contain some irregularities, they do represent the proportions of the various source languages appropriately: <?page no="128"?> Hans-Peter Naumann 110 In this article, the term Old Nordic is used for the totality of medieval North Germanic. This refers to the speech community regarded as an entity by the contemporary speakers themselves, and by Icelanders even as late as the 14th century. It was named dönsk tunga ‘lingua danica’ after the country with the largest population. The term Old Norse, on the other hand, will have its traditional meaning ‘Old West Nordic’ and will particularly denote the written language used in medieval Norway and Iceland. It is generally accepted that Old Nordic and Old English were not “Abstandlanguages”, but rather mutually intelligible (Townend 2002: 9-11, 53; Palm 2004: 330; Askedal 2008: 193-196). In various sagas of the Icelanders set in England (e.g. Egils saga, Njáls saga) the intelligibility between Old Nordic and Old English is taken for granted. In the Saga of Gunnlaug Serpent-tongue (written late 13th century), however, this special linguistic relationship is explicitly mentioned: 7. King Ethelred, the son of Edgar, was ruling England at that time. He was a good ruler. [...] In those days, the language in England was the same as that spoken in Norway and Denmark, but there was a change of language when William the Bastard conquered England. Since William was of French descent, the French language was used in England from then on. (Gunnlaugs saga ormstungu, ed. Hreinsson, vol. I, 1997: 315) The description could refer to Danish-Norwegian settlers in England, but from the saga author’s perspective it is more probable that the Icelanders or Norwegians who visited England during Ethelred’s reign (reg. 978-1013, 1014-1016) are meant. It is questionable, however, whether individual bilingualism can be assumed for that period. The 21% of the borrowed lexicon shown above comprise approximately 250 Old English loanwords in Old Nordic and may be seen as a rather high proportion. At the same time, this vocabulary represents a quite clearly defined time period, spanning the Danish expansion into England after the invasion of Yorkshire in 876 and subsequent Danish-Norwegian <?page no="129"?> English Keywords in Old Norse Literature 111 immigration, resulting in successive expansion of settlement until the 11th century. The distribution of Scandinavian place names indicates a concentration of Danish-Norwegian settlement mainly in Yorkshire and Lincolnshire, but also in Cumberland and Westmoreland (Fellows-Jensen 1972, 1978; Cameron 1985; Crawford 1987; Andersen 2006). Another factor is the Anglo-Saxon mission in Scandinavia and the encounter of Scandinavians with the English church in general. These short-lived but close relations are attested from the 10th century until the mid-12th century. They have also been extensively studied from an etymological perspective. Given the complexity of the mutual contact, well documented in recent studies (Townend 2002; Sandred 2005), only a few political, cultural and ecclesiastical historical points will be emphasised in the following. Concerning the dialectal classification of the Old English/ Anglo-Saxon borrowings in Old Nordic, Anglian in particular, but also West-Saxon and Kentish come into consideration (Hofmann 1955: 19). This differentiation is hardly relevant to the resulting words, however. Therefore, the general term Old English will be used in this paper. Middle English loanwords are very rare because the area of language contact shifted to the North following events at Hastings in 1066 and their political consequences. And as it is known, Scottish English borrowed incomparably more from Old Nordic than the other way round (Crawford 1987; Kries 2003). After the assignment of the Norwegian territories Orkney, Shetland and Caithness to Scotland, English influence cannot be attested in Nordic records anymore. Only after 1500 can a few English traces be detected in the lexicon of the modern Scandinavian languages. Contrary to widespread assumption, the proportion of English loanwords in the Scandinavian lexicon remains, however, remarkably small until the 20th century. The great historical Swedish Academy Dictionary (SAOB) lists only six English loanwords for the 16th century and 28 for the 17th century. Many of these words disappear later for culture-bound reasons. In the first half of the 19th century, words of English origin account for only 1% of the Swedish lexicon (compared to 30.3% German, 8.9% Latin and Greek, 6.3% French). English influence on the Swedish (as well as Danish and Norwegian) lexicon has increased significantly in recent times, but from a historical perspective there still is no talk of “the English disease of the Swedish language” (Edlund and Hene 1996: 59, 62-65). In this paper, neither the historical development of the lexicon nor single etymological problems will be discussed. Rather, I take the assumption as a starting point that neologisms of a certain period reveal information about a speech community’s disposition, attitudes and values. Additionally, I consider that there are identifiable types of loanwords that had a discursive relevance and function within the Old Nordic speech community beyond their mere lexical meaning. The corpus consists of different <?page no="130"?> Hans-Peter Naumann 112 sources, mainly from saga literature, but also from ecclesiastical scripture, as well as eddic and scaldic poetry. As non-literary sources, runic inscriptions have been included. 2. Research on keywords In the following, the linguistic discourse analysis as it is applied in current Historical Semantics is taken as a basis. A “discourse” in this context means a text corpus defined by common semantic properties, a consistent space, time and communication relation (text type, genre) and intertextual references. This kind of discourse analysis focuses on: a) single words in their contextual usage; b) their intertextual links; and c) semantic presuppositions and implications of statements - in other words, strategic aspects of an expression (Busse/ Teubert 1994: 22f.). Following a first dissertation in semantics by the English linguist John Rupert Firth, current research in discourse analysis has not without reason turned its attention mainly to politically or historically embedded keywords. Firth (1957), however, is not the one who coins the term “keywords”, but launches the terms “focal words” and “pivotal words”. The former build the centre of a social discourse focusing on communicative matters. The latter form an essential hub of the communicative process. Both terms refer to the user’s perspective, while the term “keywords”, which was used by Stephen Ullmann (1964), rather emphasises the observer’s perspective. Both concepts stress at the same time the affective components of the lexical meaning and request a contextual investigation of relevant subjects. In the study of affective contents, in German research the term “Hochwertwort” was coined for keywords of highly positive value. As an antonym, “Unwertwort” or “Stigmawort” was established. The former refers to central values in a society, which under certain circumstances can be crucial to creating that society’s identity. The “stigmaword” on the other hand stigmatises “Personen, Gegenstände, Sachverhalte” (Hermanns 1994: 20). As a conclusion to the discussion on keywords I do not wish to make a further attempt at a definition, but prefer to summarise a bundle of criteria relevant to discourse analysis of historical languages. I consider the following points to be particularly relevant (for a summary see Naumann 2005: 60-63): • affective semantic components (positive/ negative: stigmawords) • semantic relation to context • semantic/ lexical competition (co-occurrence/ oppositions) • reference to abstract values of social and clerical life (high value words) <?page no="131"?> English Keywords in Old Norse Literature 113 • lexical dynamics: neologisms (loanwords per se), ad-hoc compounds, derivations • relative (temporary) frequency of use. 3. Categories of Old English loanwords Before questions concerning cultural history can be answered it is important to know how the lexicon that derived from Old English is actually composed. In the following, I rely on the standard survey by Fischer (1909), which presents most of the relevant material, classified according to source languages and genres of literary sources. Fischer’s etymologies have been discussed in other studies, and some were rejected (Höfler 1931, 1932; de Vries 1977). However, his survey, last supplemented by Simensen (2002), still provides considerable information about specific texts and literary genres in which certain loanwords occur. Etymologically obscure cases have not been considered in this study. A rough classification shows the following specifics of the borrowed lexicon (selected evidence for categories 1-3): 3.1 Ecclesiastical Culture bjalla ‘bell’: OE bella; bleza ‘to bless’: OE bletsían; guðsifjar fem. pl. ‘spiritual relationship, sponsorship’: OE godsibb; guðspjall ‘gospel’: OE gōdspell; hringja ‘to ring bells’, hringing fem.: OE hringan; húsl ‘housel’: OE hūsel; kofi masc. ‘convent cell’: OE cofa; læra ‘to teach’, part. lærðr ‘learned’: OE læran, gelæred; reykelsi ‘incense’: OE rēcels; róða ‘rood, crucifix’: OE rōd; sál, sála ‘soul’: OE sāwl, sāwol ‘soul, life’; sálask ‘to die’: OE sāwlian. E-Lat.: e.g. áboti, biskup, djákn ‘deacon’; engill, erki- ‘arch-’; funtr, fontr ‘font’; kalikr ‘cup, chalice’; kantiki ‘song’; kanóki ‘canon’; kirkja, klaustr, klerkr, messa, munkr, nunna, pápa, prestr, salmr, saltari ‘psalter’; skript ‘confession’. Via OE or OSax.: bókstafr ‘character, letter’; offra ‘sacrifice’; prédika ‘to preach’. 3.2 Courtly and chivalric culture, state and warfare byrla ‘to wait upon’: OE byrelian; byrlari, byrli ‘waiter’: OE byrele; hirð fem. ‘a king’s or earl’s bodyguard, the king’s men’: OE hīrd, hīrēd masc. < *hīwræd (ON originally drótt, verðung), cf. hirðmaðr etc.; láðmaðr ‘guide’: OE lādman; lafði ‘lady’: ME leafdi < OE hlæfdige; lávarðr ‘lord, master’: ME lāuerd < OE hlāfweard; frakka fem., frakki masc. ‘spear, lance’: OE franca; harri ‘lord, king’: OE hearra; stívarðr ‘steward’: OE stīweard; targa ‘shield’: OE targe; þingmannalíð, þingmenn ‘the king’s housecarles in England’: OE þēningmanna; and possibly tákn ‘sign’; vimpill ‘wimple, a kind of hood, veil’. E-Lat: bóla ‘buckler’: ME boule; kastali; kempa ‘champion’: OE cempa; leó, león, ljón (herald.); múrr <?page no="132"?> Hans-Peter Naumann 114 ‘wall’; stallari masc. ‘a king’s marshal’: OE steallere; tafl ‘a board game, tables’; tollr ‘toll’. 3.3 Material culture (food and drink, new objects) barlak ‘barley’: OE bærlic; bjórr ‘beer’: OE bēor; glófi ‘glove’: OE glōf; klæði ‘cloth, stuff’: OE clād, clæð-; krukka ‘pot’; skrúð ‘ornament; furniture; a kind of cloth’. E-Lat.: belti ‘belt’; bytta ‘small tub, bucket’; diskr ‘plate’; fiðla ‘fiddle’; gim, gimsteinn ‘gem-stone’; kál ‘cabbage, kale’: OE cāwl; kalkr ‘chalice’; kanna ‘can’; kápa ‘cloak made with a cowl or hood’; kista ‘chest, box’ (ult. Greece); klefi ‘closet’: OE cleofa; koparr ‘copper’; kyndill ‘candle’: OE candel; kyrtill ‘gown, tunic’; næpa ‘turnip’: OE næp (possible Egyptian origin); ostra ‘oyster’; pell ‘kind of costly stuff’: OE pæll; pera ‘pear’: OE peru; pipari, piparr ‘pepper’ (ult. from the Orient); port ‘gate’; ME skarlat ‘scarlet’ (or MLG, ult. Persian); skutill ‘plate, table’; skyrta ‘shirt, a kind of kirtle’; sokkr ‘stocking’; sútari ‘tanner, shoemaker’; url ‘veil’: OE orl; and possibly mynt ‘coin’; plóma ‘plum’; tigl ‘tile, brick’. 3.4 Trade and shipping bátr masc. ‘boat’: OE bāt (vs. ON beit neut.). E-Lat.: akkeri ‘anchor’; blakt ‘white cloth’: ME blancet; búza, bússa ‘vessel’: OE būtse; dreki ‘ship of war; baring a dragon’s head as an ornament at the prow’: OE draca (or Ir. drac); forkr ‘pole, stick on a boat’: OE forclas pl.; mang ‘barter, peddling’: OE mangīan; pund ‘pound’; sekkr ‘package, truss; in a merchant ship’. The ecclesiastical loanwords form the category with the highest proportion (35%). There is no doubt that the Scandinavians owe their central clerical lexicon primarily to Old English. Only a small part of this lexicon derives from a Germanic root, the far bigger part of the ecclesiastical terminology is Latin in its origin. Also Middle Low German has contributed to the ecclesiastical vocabulary, but in a far smaller range. Much of Christian terminology was adapted through loanshifts (Simensen 2002: 960f.). The prime example for a semantic loan within the ecclesiastical field is fjándi, a highly affectively connotated keyword with numerous compounds and derivations: OE fēond > ON fjándi ‘enemy, foe > fiend, devil’, fjándinn ‘the evil foe, the devil’, fjánda-kraptr ‘fiendish power’, fjánda-limir ‘devil’s limbs, devil’s servants’, fjánda-villa ‘fiendish heresy’ etc. Examples of loan translations are: OE fæstan vb., fæsten neut. > ON fasta fem., fasta vb.; rōdfæstian > ON krossfesta ‘crucify’. ON helvíti is a morpheme-tomorpheme substitution of either OE hellewīte or OSax. helliwīti. The Old Saxon word means ‘hell-torment’ or ‘hell-punishment’. The Old English word also has both these meanings, but can additionally mean ‘hell’, in the <?page no="133"?> English Keywords in Old Norse Literature 115 sense of Old Nordic helvíti. Old English was probably the source language (Halldórsson 1970: 376). The profound influence of Christianity has been extensively studied by Kahle (1890), Taranger (1890), Thors (1957) and Walter (1976). The linguistic evidence merely reflects the history of conversion in the North. Christianity was introduced in Norway and Iceland directly from England, mainly by Anglo-Saxon - and also Irish and Scottish - missions. The important function of the vernacular languages in place of Roman Latin in ecclesiastical life was mainly owed to the Rome-distant English church. Moreover, some Norwegian settlers emigrated to Iceland, either from the Northern British Isles or passing by them as an intermediate stop. From there, they brought Christian wives and slaves (Loyn 1994: 58, Forte et al. 2005: 311). In Denmark, Sven Haraldsson’s conquest of England (recognition as King of England in 1013) and the Empire’s weakness at that time also had an influence on the strong English impact in the ecclesiastical field. Sven employed the English bishop Gotebald as a teacher in the region of Skåne. His successor Knud (Cnut) the Great brought more English bishops to his home country during the years before 1035. Hundreds of churches were established. In the new reign after 1035, and even more intensively after 1066, the Danish-Norwegian nobility from Sven’s and Knud’s retinue (or at least parts of it) were to some extent forced to repatriate. According to the Scandinavian names in the Liber Vitae of Thorney Abbey (ed. Gerchow 1988) quite a number of them might have been newly converted (Whitelock 1940). In England, they had proprietary churches and their own burial places and probably did not part with their recently acquired cults after their return. Already at the time of the early raids the Vikings seem to have shown a certain respect towards ecclesiastical institutions. They robbed monasteries but did not always destroy them. This explains why the relics of St. Columba were still present in 825, even though Iona was raided in 795, 802 and 806 (Kries 2003: 22). Interesting in this context is the fact that within the Anglo-Saxon Chronicle there is a “terminus technicus” for the early raids of the Vikings, gān up (wendan up) with the specific meaning ‘invade the country’. In Old Norse sources the term is attested as ganga up (Hofmann 1955: 62, 200). Whichever the perspective, whether from the view of the Vikings or the Anglo-Saxons, the term is highly connotated. In its function as a new mental and cultural spirit, the courtly and chivalric culture triggers a number of neologisms (see 3.2) which become keywords of the time in saga literature. A proportion of them definitely have the features of high value words. In the sagas of the Icelanders their distribution is very unequal, but some texts stand out in showing high frequencies. Especially rich in Old English loanwords from this category (31 records) is the Saga of the People of Laxardal (written mid-13th century). <?page no="134"?> Hans-Peter Naumann 116 Here, chivalric neologisms serve to emphasise the status of the persons involved. Or in other words: the author endows them to some extent with the attributes of English knights, suggesting that they - however anachronistically - bring with them some courtly glamour to the altogether rural life in Iceland. Chapter 77 describes in a colourful way how the main character Bolli returns from abroad. It gives a striking example of this narrative form: He had become such a fine dresser by the time he returned from his journey abroad that he wore only clothes of scarlet or silk brocade (han vildi engi klæði bera nema skarlatsklæði ok pellsklæði) and all his weapons were decorated with gold. He became known as Bolli the Elegant (Hann var kallaðr Bolli inn prúði). He took eleven men with him […] They were all comely men, but Bolli was in a class by himself (allir váru þeir listuligir menn, en þó bar Bolli af). He wore a suit of silk brocade given to him by the emperor of Byzantium, with a cloak of red scarlet outermost (Han var í pellsklæðum, er Garðakonungr hafði gefit honum; hann hafði yzta skarlatskápu rauða). […] he held a red shield at his side with the figure of a knight drawn on it in gold (han hafði rauðan skjöld ... ok á dreginn riddari). He had a lance in his hand, as is common in foreign parts (han hafði glaðel í hendi, sem títt er í útlöndum). […] On such style did Bolli ride westward … (Með slíkri kurteisi ríðr Bolli vestr í sveitir). (Laxdœla saga, ed. Hreinsson, vol. V, 1997: 118) In this extract, there are three Old English loanwords (klæði, prúðr, pell); two words derive from Middle English (glaðel and skarlat); while listuligr ‘polite, elegant’, riddari ‘knight’ and kurteisi ‘courtesy, chivalry’ are originally Middle Low German. Many Old English loanwords with Latin roots are terms for new edibles, trade goods, implements, cloths, tools and often important as migrant and cultural words (see 3.3). Among the words in category 3.4 three shipping terms are eye-catching: bátr for older Nordic beit, buza from OE búza, bússa and dreki, which denotes a fast battleship ornamented with a dragon head. These loans are surprising because the Vikings were much more proficient in ship technology than the Anglo-Saxons. In fact, this explains the numerous Scandinavian loanwords from this semantic field in Old English (Peters 1981: 172ff.). For dreki, the onomasiological background seems fairly clear: it is a high value word for an extravagant prestigious object (in the owner’s eyes, of course), the Rolls Royce among the different ship types, so to say. ON bátr, on the other hand, is an everyday word. It displaced the Nordic form beit and denotes a longboat or ferry boat corresponding to its use in Anglo-Saxon. But it is also related to the poetic word bāt which, for instance, is used for true ships in Beowulf: flota was und yðum, bat under beorge (Beow. 210ff.). The process of borrowing is obviously complicated, but we can suppose that the integration of the new <?page no="135"?> English Keywords in Old Norse Literature 117 word form has led to a differentiation of already existing terms (Hofmann 1955: 119f.; Peters 1981: 184). However, the lively exchange of words is strong evidence for particularly intensive contact in the maritime sector. So much for the English borrowings according to traditional categories. A significant part of the loanwords, however, has to be assessed in a different way: 3.5 Lexicon of cognition, emotion, intention bílifi ‘luxury’: OE bileofa, bigleofa ‘support, food, nourishment’; fox neut. ‘fraud’ (only in a metaph. sense): OE/ ME fox ‘cheater/ cheating’; loddari ‘juggler, jester’: OE loddere ‘beggar’; prettr masc. ‘trick, deceit, fraud’ (pretta vb., prettótr adj.): OE prætt masc. (prættig adj.); púki ‘a devil, imp’: OE pūca, ME pouke, ON púki > Ir. puca; strákr ‘landlouper, vagabond’: OE strāc; ægis-hjálmr ‘a helmet of terror’ (metaph. and mythol.): OE eges-, (-grīma). E-Lat.: portkona ‘gate-woman’ after OE portcwēn; próf ‘proof, evidence’: OE prōfían; prúðr ‘fine, magnificent, stately’: OE prūt, ME prūd, via OFr. prud; pryða vb. ‘to adorn, ornament’, prúðiligr adj., pryði fem. ‘an ornament’: OE pryto, pryde; pynda or pynta ‘to extort, compel by brute force’: OE pyndan ‘arrest’, ON pynding fem. ‘extorsion, tyranny’. At first glance, these words seem to be, at least in parts, just ordinary everyday or vogue words. Also words for human behaviour, deficiency or even vice are part of the sample. But on closer inspection, the words can be grouped in semantic fields titled “cognition”, “emotion”, “intention”. They hardly have less diagnostic value for cultural historical questions than more specifically cultural words. Among some positive terms, words for deceit and cunning catch attention in this category, but also words for the lowest levels of society. With púki we have another word for the devil (see below). 4. Criteria for keywords As one criterion to determine keywords I have mentioned relative or temporary frequency of use. This can possibly mean that a specific word or semantic field is a keyword only in a single text, as has been shown for the courtly and chivalric lexicon in Laxdæla saga. Normally, however, keywords possess a relatively high frequency in texts. These frequencies can be investigated more or less reliably by current methods in Corpus Linguistics. Relatively high numbers in text usage (i.e. occurrence in 20-40 texts) apply for the following words: bátr, hirð, prettr; from the ecclesiastical field: fasta, lærðr (lærðir ok ólærðir menn) and sál, sála respectively. Bátr has already been mentioned above. In context it is only used referentially and is therefore not interesting as a keyword. Why prettr was borrowed remains enigmatic. We can assume that it constrained the meaning <?page no="136"?> Hans-Peter Naumann 118 of the older words vél or list ‘deceit, cunning’ or even displaced them partially. The borrowing appears in a 10th century skaldic poem and becomes very popular later on. A semantically highly accentuated field is hirð. Its focal context is quite obvious. ON hirðmaðr ‘king’s man’ and hirð ‘a king’s bodyguard, the king’s men’ both derive from OE hīrēdman and hīrēd. In Anglo-Saxon its usage is wider, however. Here, hīrēd not only determines the king’s or noblemen’s household, but also those of monasteries or clerical convents. The origin is the meaning ‘family’. In Scandinavia, however, the word was only used to determine the king’s men (for the history of the borrowing process see Hofmann 1955: 57f.). From a historical perspective, there is certainly a connection with the introduction of courtly behaviour influenced by Anglo-Saxon customs. In Norway, these customs are first recorded at the time of king Ólafr Haraldsson (reg. 1015-1030). They are of central interest in numerous sagas of the Icelanders, regardless of the fact that they take place in the 10th century. For this special perspective, both a historical-economical and a narratological explanation could be possible. It is important to keep in mind that in the 11th and 12th centuries, the number of inhabitants of Iceland was about a quarter of the number in Norway (ca. 70-80,000 versus ca. 320,000). Iceland was not only a potent trading partner, but also a political force. Young Icelanders from good families sallied forth to the Norwegian court during their first journey abroad, striving to be affiliated with the king’s entourage. The kings for their part had great interest in engaging the Icelanders in their court. The journey abroad, on the other hand, is identical with the narrative patterns of a “Journey out” as we know them from many kinds of Old Norse prose with affiliation to folktales. These have to end happily if the story shall continue at all, and lead to the narrative function “Hero’s glorification”. The arrival at the court and the reception into the hirð (or also rejection in favour of a rival) is a favourite plot. The discourse circles around these topics and thereby deduces the society’s attitudes. Njáls saga (“The Saga of Njáll”), the most widely acclaimed work of saga prose, serves as an example. The young, noble and handsome Icelander Hrútr comes to Norway. His arrival in Bergen is instantly reported to the king’s mother Gunnhildr, one of the most remarkable women in Nordic Antiquity. Literary sources speak of her great erotic appeal, her cruelty and magical powers. As a historical figure, she followed her husband, King Eiríkr blóðøx (“blood-axe”) Haraldsson, on his escape to York, returned to Norway in 961 and died in exile on the Orkneys in 970. Hrútr naturally wants to become part of the hirð but first has to deal with Gunnhildr. She commands him to her chamber every evening for half a month before finally, upon his passing the exam, giving him a privileged <?page no="137"?> English Keywords in Old Norse Literature 119 position at the king’s table and in the hirð. All this is discussed extensively in the prologue to the saga. Ecclesiastical discourses mostly involve the following three terms: for the first, it is the loan translation fasta vb. and fasta f. ‘to fast; fast, fasting’, a phenomenon that apparently was discussed in detail during the period of conversion. For the second, it is the word group læra ‘to teach’, part. lærðr ‘learned, clerically educated’ or lærðir menn ‘clerics’ in contrast to the illiterate laity, which is also embedded in the context of the time. In the formula lærðir ok ólærðir menn ‘clerics and laymenn’, the new term also serves to emphasise the Christian community. The North gains a central lexeme of the upcoming written culture: the vocalism / æ/ instead of expected / a/ clearly indicates English origin (Kristensen 1906: 24). Reference to a central value in Christian life gives sál or sála. The root of this Old West Nordic form is Old English sāwl or sāwol, whereas Swedish själ and Danish sjæl derive from Old Saxon siala. Nordic had and still has a special word for ‘spirit’ or ‘life’, namely the feminine önd, but no word for ‘soul’ in the Christian meaning. Both Nordic forms appear early, around the year 1000, as authentic records sial or sal, saul, sol on Danish and Swedish rune stones. These monuments were raised in memory of already christened persons. On the stone of Husby, Västergötland, it says: Assur ok Sven ok ÞoriR, þeir lagđu sten þenna yfiR mođur sina Oluf. Guđ hialpi sial hennaR ok Guđs mođiR ok allir Guđs ænglaR. ‘Özurr and Sveinn and Thórir, they laid this stone over their mother Ólöf. May God and God’s mother and all of God’s angels help her soul.’ (Vg 50) The inscription also bears the memory of a dead mother in a special intercession formula: ‘May God and God’s mother and all of God’s angels help her soul’. It contains another English loan word: ænglaR. The memorial stone of Husby and further inscriptions provide the first records of ON engill. The word, however, must have a linguistic prehistory because its ending indicates an integration into the Nordic lexicon. How serious the converts were about the new religion is shown in Swedish inscriptions from the same period. On the stone from Eggeby, Uppland, for example, it is mentioned that a bridge was dedicated to the deceased. Naturally this bridge was a physical memorial, but metaphorically the “soulbridge” into heaven was meant: RagnælfR let gærva bro þessi æftiR Anund sun sinn gođan. Guđ hialpi hans and ok salu bætr þæn hann gerđi til [...] ‘Ragnelfr had this bridge made in memory of Önundr, her good son. May God help his spirit and soul better than he deserved.’ (U 69) This kind of inscription also reveals that the borrowed word sál, sála did not displace the Nordic term önd. Rather, this word is kept as a relative term reflecting runic and ok salu. <?page no="138"?> Hans-Peter Naumann 120 What is also striking about this type of inscription is that similar types of memorial stones in England might have served as a model. There are records of several runic inscriptions on stone crosses from Western England that anticipate the soul formula, two of which are metrical (Thornhill III, Yorkshire; Great Urswick, Cumbria): Gilswiþ arærde / æfte Berhtswiþe / becun on bergi; / gebiddaþ þær saule. ‘Gilswith raised a memorial on a mound for Berhtswith. Pray for her soul.’ (Thornhill III, ed. Page 1973: 144f.) Tunwini settæ / æfter Torhtredæ / becun æfter his bæurnæ; / gebiddæs þer saulæ. ‘Tunwini set up a monument after Torhtred his son. Pray for his soul.’ (Great Urswick, ed. Page 1973: 144f.) In fact, it cannot be ruled out that these memorial formulae had an impact on Scandinavian inscriptions, even though Christian memorial practice in runic script within Scandinavia only emerges much later (from the 10th century in Denmark). Fragments from stones of the same type (Thornhill I and II, cf. Page 1973: 145f.) show, however, that this type of memorial was spread more widely in the Danelaw than the few remnants would suggest. Negative ideals in the medieval Nordic narrative community trigger a number of neologisms with high frequency of use. Partly there are terms for persons with stigmatised characteristics, partly words for unacceptable or at least dubious social practices. The words loddari, portkona, strákr belong to the former group; fox, prettr and pynda to the latter. The rather peculiar term ægishjálmr ‘helmet of terror’, probably coined after OE ægisgríma originates from the mythological field and is properly used of serpents in Eddic Poetry. In saga prose, the word has adopted the character of an idiom (bera ægishjálm yfir e-m ‘to bear the ægis over or before another’) and is still used metaphorically in Modern Icelandic (Claesby-Vigfusson s.v.). The counterworld to the kingdom of God and the trust in an immortal soul evokes the denizens of hell. In Nordic Antiquity faith is infiltrated with fear of the devil and demons. Although the pagan Scandinavians also had demons - trolls, giants, Loki and the Midgard serpent - there was no word for the devil. This semantic gap was filled in various ways: via Middle Low German/ Old Saxon came djöfull < diabol < diabolus (Swedish djävul, Danish djævel), via English the loan translation fjándi and the curious devil figure púki. Around the stigma words par excellence, djöfull and fjándi, a dense framework emerges, which spreads to nearly all types of sources. The word púki on the other hand only occurs in more recent records in contexts that are particularly accentuated in a literary way. Neither is there a particular word for ‘hell’. The pagan kingdom of the dead Hel cannot cover the concept. Instead, the morpheme-to-morpheme substitution helvíti fills the gap. The loan translation fjándi is still in use at present. In <?page no="139"?> English Keywords in Old Norse Literature 121 the individual modern languages the corresponding lexemes are not used as appellatives anymore, however, but play a significant role in the semantics of swearing. ON púki is associated with a complicated word history. Most commonly, the word is seen as an English loan (< OE pūca or ME pouke). The fact that the word only appears late and mainly in literary texts influenced by Christian culture (Fischer 1909: 23, 173) speaks in favour of this thesis. As a loan from Scandinavian (via Shetlandish) respectively, Irish puca ‘sprite, hobgoblin’ has been discussed (de Vries s.v.). The semantic field is widely ramified in Scandinavian and serves among others in the form pocker/ pokker ‘devil, deuce’ in Modern Swedish/ Danish as an amplifier lexeme in expletives (Falk and Torp s.v. pokker, puge). According to Claesby and Vigfusson (s.v.) the meaning of púki in Icelandic texts is ‘a devil, but with the notion of a wee devil, an imp’. In this guise the figure arises in Þorsteins þáttr skelks, an entertaining anecdote which portrays the supernatural and is somewhat farcical in character, written about 1300 and transmitted in the “Great saga of Olaf Tryggvason” in Flateyjarbók (ed. Nordal 1944-45: 462-64). Even for the genre þættir the text is quite short, but captures within this limited span a message influenced by Christian thought and therewith a significant group of “focal words”, many of which are English in origin. The word púki occurs no less than ten times and also alternates with semantically related terms like draugr “ghost, spirit” and skelmir “rogue, devil” (< MLG schelmer, schelm). On a banquet one summer night in Norway the Icelander Þorsteinn leaves the house and wends his way alone to the privy, even though this has been expressly forbidden by king Ólafr Tryggvason (reg. 995-1000). There, he meets a demon (púki) who appears directly from hell (hellvíti) and starts a conversation about those who dwell in the underworld. In this dialogue, the agonies (pínsl, pín) of the best and worst among pre-Christian heros, Sigurðr Fáfnisbani and Starkaðr inn gamli, are portrayed. Sigurðr bears his torments in the fire best, Starkaðr worst: his screams of pain are too vigorous even for the devils (pl. fjándr) in hell. Þorsteinn calls on the púki to imitate Starkaðr’s screams. The púki screams louder with every attempt and Þorsteinn nearly faints. But when he screams for the third time, the church bells start to ring (hringja) and the demon sinks down into the ground. Þorsteinn owes his rescue to the king himself, who suspected danger and had the church bells ring. Impressed by the Icelander’s deceit the king forgives him and calls him by the nickname skelkr ‘shudder’. He becomes one of the king’s retainers by oath (hirđmađr) and dies at his side on Ormrinn langi (“The Long Dragon”) in the battle of Svoldr. <?page no="140"?> Hans-Peter Naumann 122 In recent research, Þorsteins þáttr skelks has been discussed from an ethnographic perspective as “the embodiment […] of a legend of the supernatural” (Lindow 1986: 269). However, as early as the beginning of last century the Danish philologist Axel Olrik (1910: 217-19) called attention to the fact that the king’s anecdote seems to use a specific source. This might be the Visio Tnugdalis (or Tundali), a Latin text written around 1149 in Regensburg - probably by an Irish monk - and likely to be the best known conception of the afterlife from the Middle Ages. The text describes the transmigration of the soul of the Irish knight Tnugdal through hell, purgatory and paradise. Fergus and Conall, the two bravest warriors from the epic cycle of Cuchulin, languish in a fire, one standing straight up and the other headfirst in the fire. The intertextual references between Visio and the þáttr go into detail and are convincing, but still the methods of transmission are unresolved. In any case, the Old Norse translation of the work, Duggals leiðla, is definitely not a possible reference. It might be a coincidence that the target text contains an Anglo-Irish signal word, but it should not be totally excluded that language awareness and reflexion are behind it. Addressing language, i.e. discourse about words and discussion of terminology, comprises present methodic principles of keyword analysis in public language use. Considerations about cultural historical semantics can in only a few cases be based on such systems of reference. In cases of explicit reference in Old Norse literature, the semantic fields of “glory and honour”, “law and order” or central concepts of Nordic memorial culture are most likely referred to. We are therefore concerned with a system of values and norms that has yet to be influenced by lexical borrowing, and in this context the extension of the term “keyword” has to be put into perspective. In any event, linguistic discourse analysis certainly provides new insights that extend beyond research into loanwords. Translation by Michelle Waldispühl. <?page no="141"?> English Keywords in Old Norse Literature 123 Biographical sketch Hans-Peter Naumann is Professor emeritus in Nordic philology at the University of Zurich. He has published widely on such subjects as the history of German-Swedish translations, the laws of the Icelandic Commonwealth, Old Icelandic and modern Swedish literature, medieval personal names, language contact in Scandinavia, Old Norse syntax and Nordic language history, and runic inscriptions. He is expert advisor for the Reallexikon der Germanischen Altertumskunde, and co-editor of The Nordic Languages. An International Handbook of the History of the North Germanic Langagues (2002-2005) and Beiträge zur nordischen Philologie. Work in progress: Die metrischen Runeninschriften Skandinaviens (2011). References Andersen, Per Sveaas. 2006. Det siste norske landnåmet i Vesterled: Cumbria - Nordvest-England. Oslo: Unipub. Askedal, John Ole. 2008. “Språklig kulturutveksling over havet i norrøn tid”. In John Ole Askedal and Klaus Johan Myrvoll (eds.). Høvdingen. Om Snorre Sturlasons liv og virke. Oslo: Vidarforlaget, pp. 191-228. Busse, Dietrich and Wolfgang Teubert. 1994. “Ist Diskurs ein sprachwissenschaftliches Objekt? Zur Methodenfrage der historischen Semantik”. In Dietrich Busse et al. (eds.). Begriffsgeschichte und Diskursgeschichte. Opladen: Westdeutscher Verlag, pp. 10-28. Cameron, Kenneth. 1985. “Viking Settlement in the East-Midlands. The Place- Name Evidence”. In Rudolf Schützeichel (ed.). Giessener Flurnamen- Kolloquium 1. bis 4. Oktober 1984. Heidelberg: Winter, pp. 129-153. Claesby, Richard and Gudbrand Vigfusson. 1957. An Icelandic-English Dictionary. (Second edition, suppl. William A. Craigie, first edition 1874). Oxford: Clarendon Press. Crawford, Barbara E. 1987. Scandinavian Scotland. Leicester: Leicester University Press (Scotland in the Early Middle Ages 2). Edlund, Lars Erik and Birgitta Hene. 1996. Lånord i svenskan. Om språkförändringar i tid och rum. (Second edition, first edition 1992). Stockholm: Norstedts. Falk, Hjalmar and Alf Torp. 1910-11. Norwegisch-dänisches etymologisches Wörterbuch. Heidelberg: Winter. Fellows-Jensen, Gillian. 1972. Scandinavian Settlement Names in Yorkshire. Copenhagen: Akademisk forlag (Navnestudier 11). Fellows-Jensen, Gillian. 1978. Scandinavian Settlement Names in the East Midlands. (Navnestudier 16). Copenhagen: Akademisk forlag. Firth, John Rupert. 1957. Papers in Linguistics 1934-1951. London: Oxford University Press. <?page no="142"?> Hans-Peter Naumann 124 Fischer, Frank. 1909. Die Lehnwörter des Altwestnordischen. (Palaestra 85). Berlin: Maier & Müller. Forte, Angelo, Richard Oram and Frederik Pedersen (eds.). 2005. Viking Empires. Cambridge: Cambridge University Press. Gerchow, Jan. 1988. Die Gedenküberlieferung der Angelsachsen. Mit einem Katalog der libri vitae und Necrologien. (Arbeiten zur Frühmittelalterforschung 20). Berlin and New York: Walter de Gruyter. Halldórsson, Halldór. 1970. “Determining the Landing Language”. In Hreinn Benediktsson (ed.). The Nordic Languages and Modern Linguistics. Reykjavík: Vísindafélag, pp. 365-378. Hermanns, Fritz. 1994. “Linguistische Anthropologie. Skizze eines Gegenstandsbereichs linguistischer Mentalitätsgeschichte”. In Dietrich Busse et al. (eds.). Begriffsgeschichte und Diskursgeschichte. Opladen: Westdeutscher Verlag, pp. 29-59. Höfler, Otto. 1931, 1932. “Altnordische Lehnwortstudien I-III”. In Arkiv för nordisk filologi 47, pp. 248-297; 48, pp. 1-30, 213-241. Hofmann, Dietrich. 1955. Nordisch-englische Lehnbeziehungen der Wikingerzeit. (Bibliotheca Arnamagnæana XIV). Copenhagen: Ejnar Munksgaard. Hreinsson, Viðar (ed.). 1997. The Complete Sagas of Icelanders. I-V. Reykjavík: Leifur Eiríksson. Kahle, Bernhard. 1890. Die altnordische Sprache im Dienste des Christentums. I. Teil: Die Prosa. (Acta Germanica 51, 4). Berlin: Maier & Müller. Kries, Susanne. 2003. Skandinavisch-schottische Sprachbeziehungen im Mittelalter: Der altnordische Lehneinfluss. (NOWELE. Supplement 20). Odense: University Press of Southern Denmark. Kristensen, Marius. 1906. Fremmedordene i det ældste danske skriftsprog (før omtr. 1300). Copenhagen: Schubothe. Lindow, John. 1986. “Þorsteins þáttr skelks and the Verisimilitude of Supernatural Experience in Saga Literature”. In John Lindow et al. (eds.) Structure and Meaning in Old Norse Literature: New Approches to Textual Analysis and Literary Criticism. (The Viking collection 3). Odense: University Press, pp. 264- 280. Loyn, Henry R. 1994. The Vikings in Britain. Oxford, England and Cambridge, MA: Blackwell. Naumann, Hans-Peter. 2005. “Niederdeutsche Schlüsselwörter im Altwestnordischen. Überlegungen zu einer kulturellen Semantik”. In Lennart Elmevik et al. (eds.). Niederdeutsch in Skandinavien V und VI. (Osloer Beiträge zur Germanistik 37). Frankfurt a. M.: Peter Lang, pp. 59-76. Nordal, Sigurður (ed.). 1944-45. Flateyjarbók. 4 vol. Akranes: Prentverk Akraness. Olrik, Axel. 1910. Danmarks heltedigtning. Vol. 2. Starkad den Gamle og den yngre Skjoldungrække. Copenhagen: Gad. Page, Raymond Ian. 1973. An Introduction to English Runes. London: Methuen. Palm, Rune. 2004. Vikingarnas språk 750-1100. Stockholm: Norstedts. Peters, Hans. 1981. “Zum skandinavischen Lehngut im Altenglischen”. Sprachwissenschaft 6, pp. 85-124, pp. 169-185. <?page no="143"?> English Keywords in Old Norse Literature 125 Sandred, Karl Inge. 2005. “Language Contact outside Scandinavia III: with England and Scotland”. In Oskar Bandle et al. (eds.). The Nordic Languages. An International Handbook of the History of the North Germanic Languages. Vol. 2. Berlin and New York: Walter de Gruyter, pp. 2062-2073. Simensen, Erik. 2002. “The Old Nordic Lexicon”. In Oskar Bandle et al. (eds.). The Nordic Languages. An International Handbook of the History of the North Germanic Languages. Vol. 1. Berlin and New York: Walter de Gruyter, pp. 951-963. Taranger, Absalon. 1890. Den angelsaksike kirkes indflydelse paa den norske. Kristiania: Grøndahl. Thors, Carl-Eric. 1957. Den kristna terminologien i fornsvenskan. Helsinki and Copenhagen: Svenska litteratursällskapet i Finland. Townend, Matthew. 2002. Language and History in Viking Age England. Linguistic Relations between Speakers of Old Norse and Old English. (Studies in the Early Middle Ages 6). Turnhout: Brepols. Ullmann, Stephen. 1962. Semantics. An Introduction to the Science of Meaning. Oxford: Blackwell. de Vries, Jan. 1977. Altnordisches etymologisches Wörterbuch. (Third revised edition, first edition 1956). Leiden: E. J. Brill. Walter, Ernst. 1976. Lexikalisches Lehngut im Altwestnordischen. Untersuchungen zum Lehngut im ethisch-moralischen Wortschatz der frühen lateinischaltwestnordischen Übersetzungsliteratur. (Abh. d. Sächs. Akad. d. Wiss. Leipzig: Philol.-hist. Kl. 66, 2). Berlin: Akademie-Verlag. Whitelock, Dorothy. 1940. “Scandinavian Personal Names in the Liber Vitae of Thorney Abbey”. Saga Book of the Viking Society XII, pp. 127-153. <?page no="145"?> Pam Peters Emergent Conjunctions Abstract Conjunctions are usually discussed as one of the closed classes of words, yet their ranks are quietly expanding with emergent subordinators and semicoordinators, grammaticalizing out of preexisting adverbials. This paper focuses on three such items: however, once, the way. Despite differences in form and semantics, their established role as adverbial adjuncts allows them to act as the syntactic link to a following clause, and thus to take on the additional role of subordinator or coordinator. Corpus data illustrating their use as conjunctions come from comparative corpora (ACE, Frown, FLOB), and spoken and written discourse (the ICE corpora and ART), as resources for evaluating their status in three major varieties of English. The emergence of new subordinators out of several different kinds of adverbial argues against the idea that subordinators are a residual word class (Huddleston and Pullum 2002). This research suggests rather that the extending of an adverbial’s grammatical role to subordinating (or coordinating) conjunction is a natural process; and that the so-called “closed classes” can be replenished by the grammaticalization of amenable elements from elsewhere in the lexicon. 1 Introduction The modeling of language as a set of discrete levels - as phonology, lexicon, grammar - can seem to imply separation between them which in practice is often bridged. It happens between phonology and the lexicon in the process of lexicalization with phonetically motivated words, e.g. whoosh; and in the process of grammaticalization, by which new grammatical forms are generated from the open classes of the lexicon, e.g. the use of lots (of) as a complex determiner. The interplay between grammar and lexicon is continually evident in the interchange of items such as the modal verb may in it may be and the modal adverb possibly. This interplay between grammar and the lexicon may be seen as “grammatical metaphor” (Halliday and Matthiessen 2004) and as evidence of the fact that grammar and lexicon are vitally interconnected in the lexicogrammar of English. The interplay between lexicon and grammar is especially evident in new and emerging varieties of English. It is conspicuous in pidgins and creoles, e.g. in the use of longa (< ‘belonging to’) as an all-purpose preposition meaning ‘relating to’ in Melanesian Pidgin. In “settler” varieties of English <?page no="146"?> Pam Peters 128 too such as Australian and New Zealand English, the lexicogrammar is the locus of distinctive regional lexicogrammar (Schneider 2007). For example the application of the zero complementizer rule in verb and adjectiveheaded subordinate clauses (e.g. He decided [] the color was too dark; She was afraid [] the cat would get lost) was found (Kearns 2009: 244-9) to be noticeably higher in Australian and New Zealand English than in British and American English. It seems that this syntactic change may be generated from the antipodes (p. 260). Emergent elements of grammar can no doubt consolidate sooner in new varieties than in the established northern hemisphere varieties (Peters 2011 forthcoming). 2 Conjunctions and other word classes 2.1 Conjunctions and adverbial adjuncts Grammarians have long commented on the affinity between adverbials and certain other word classes. The parity of adverbial adjuncts with conjunctions is noted by Quirk et al. (1985: 442-43), especially those for which there are positional restraints on their use. This is clear with the conjunctive adverbs so and yet, which behave like conjunctions in occurring at the start of a clause, e.g. We brought a computer, (and) so there was no problem. Under the tree we were protected from the hail, (and) yet we got rather wet. The bracketed and shows that in each case they allow asyndetic connection, and are in this respect more like coordinators than subordinators - hence the name “semi-coordinators” given to them by Quirk et al. (1985: 927-28). But the authors also acknowledge a coordination - subordination gradient, and that there are conjunctive properties (e.g. inability to link subordinate clauses) which they share with subordinators. Large reference dictionaries such as the Oxford English Dictionary online and Merriam-Webster online both accord the role of conjunction to so and yet, alongside their adverbial roles. A variety of other adverbials are also capable of taking on conjunctive roles, especially those that express time (when), place (where) and reason (why). For example: They put it where we expected. There the adverbial where and following clause function as the obligatory adjunct of the verb put in the main clause, while where itself also plays the role of subordinator. The potential for other types of adverbial adjunct to take on the subordinator role can be seen with examples such as directly, in its temporal rather than spatial sense; and immediately in the sense of ‘as soon as’. <?page no="147"?> Emergent Conjunctions 129 2.2 Conjunctions and prepositions A kind of gradience also exists between certain conjunctions and prepositions, demonstrable by their ability to perform both roles. See for example the use of after in the following: After he spoke, he continued to smile at the audience. After speaking, he continued to smile at the audience. He continued to smile at the audience after his speech. The first two examples contrast the use of after as a conjunction with a finite clause and with a nonfinite clause (where it may be regarded as either a conjunction or preposition: Quirk at al. 1985: 660). Other conjunctions/ prepositions to which the same indeterminacy applies are before and since (Biber et al. 1999: 76-77). The third example has after used unequivocally as a preposition heading a noun phrase. The ambiguous status of after in the second example and others like it is resolved in the Cambridge Grammar of the English Language by a radical reanalysis of the role of prepositions, allowing that they may not only head noun phrases and nonfinite clauses, but also finite clauses (Huddleston and Pullum 2002: 1011-16). By this analysis, the after in all three examples above is a preposition. The effect of this position is to reduce the set of subordinators to just two: that, which can only take finite clauses; and whether/ if, as used in reported speech or thought (she wondered/ didn’t say whether…). Whatever the merits of this analysis, the implication is that subordinators are a residual word class with declining membership, losing their distinctive role. This goes against findings of corpus-based research demonstrating the emergence of new subordinators and semi-coordinators heading finite clauses which could never be construed as prepositions with nonfinite clauses or noun phrases. They include complex subordinators such as the way, the time, appearing without any following relative pronoun (Biber et al. 1999: 62); and the movement of conjunctive adverbs such as however, therefore, thus into the role of semi-conjunctions (Peterson 2009). The movement of adverbs into subordinator roles can in fact be traced back to Old English, as found by Kjellmer (1997) for both now and once, and it continues in contemporary British English with examples like directly. Previous research on emergent subordinators like these has focused on northern hemisphere data (Kjellmer 1997; Biber et al. 1999), and/ or on written rather than spoken English (Peterson 2009). There is therefore room for extending and updating those studies with material from Australia and Australian speakers, to see whether such conjunctions are also <?page no="148"?> Pam Peters 130 beginning to emerge in the new southern hemisphere variety, and whether perhaps the spoken medium fosters their appearance. Corpus data from the northern hemisphere will be used for comparative purposes. 3. Research plan and data to be used This research into emergent conjunctions focuses on three quite different adverbials, formally and semantically: however as an adversative semicoordinator; once in the sense of ‘as soon as’; and the way in the sense of ‘in the manner in which’, ‘as’. Data will be sought in Australian, British and American corpora dating from 1986 on, to elucidate their relative frequencies in different regional varieties, and in spoken and written discourse. The assumption is that evidence for their use as conjunctions is an aspect of the emergent grammar of English, i.e. that which becomes visible in actual discourse, rather than being related to underlying rules, constraints, or the logic of the language (Hopper 1988). In order to examine their use in Australian English, data will be extracted from: ACE a sample corpus of 1 m. words of published Australian writing from 1986 ICE-AUS a sample corpus of 1m. words of Australian English (60% spoken / 40% written), collected in the early 1990s ART a corpus of 257,000 words of Australian radio talkback, from 2004-06 As contrastive evidence from the northern hemisphere, data will be extracted from: FLOB a sample corpus of 1m. words of published British writing from 1991-95 FROWN a sample corpus of 1m. words of published American writing from 1991-95 These corpora are all mounted on the corpus website of Macquarie University’s Linguistics Department. The corpus data retrieved for all three case studies of emergent conjunctions involves searches with an immediately following subject pronoun as a means of identifying the presence of a following clause. This was done to ensure the comparability of searches for all three items across all corpora used in this research. It does however exclude cases where the following clause began with an NP subject, on which further comment will be found below in the analysis of individual cases. <?page no="149"?> Emergent Conjunctions 131 4. Three case studies: however, once, the way 4.1. HOWEVER as an adversative conjunction The use of however as an indefinite conjunction is recognized in all major reference dictionaries (Oxford English Dictionary online; Merriam-Webster online). Yet neither dictionary documents the use of however as an adversative conjunction alongside its use as an adversative adverb meaning ‘nevertheless’, and equivalent to ‘but’ at the start of a sentence. Its conjunctive use in “run-on” sentences following a comma is nevertheless commented on in a number of late C20 usage books, including Burchfield (1996) speaking of British English, and Garner (1998) for American English. Earlier mention of it can be found in Murray-Smith (1989: 195) in Australian English, where he makes no bones about it being “illiterate”. This ambivalence is also reflected in more recently compiled Oxford dictionaries, e.g. Australian Oxford Dictionary (2 nd ed. 2004), which registers the use of adversative however but adds a cautionary usage note, to the effect that “many people object” to this usage. This objection has probably prompted editors for some decades to “edit out” this conjunctive use of however in mid-sentence - which is simply done by changing the preceding punctuation mark, i.e. replacing the comma with a semicolon, or else using a full stop to make however the first word of a fresh sentence. This second practice would have contributed to more widespread use of however in sentence-initial position, which is the chief mid-C20 concern about it (Strunk and White 1959; Gowers 1965). Scattered uses of however in run-on sentences can nevertheless be found in the ACE corpus in published Australian writing from 1986, although Peterson (2009) found most instances in the unedited writing from the early 1990s contained in the ICE-AUS corpus. Because of his interest in the coincidence of adversative however and a preceding comma, he did not look for data in the spoken sections of the ICE-AUS corpus, though they too might show conjunctive use of adversative however - to the extent that they were not edited out by the transcribers. It is therefore of some interest to see whether examples can be found in spoken material in the Australian corpus. Table 1 below presents the findings from both written and spoken (Australian) sources, to test the question as to whether however used as an adversative conjunction might be more evident in speech, where the speaker’s intonation (a lowered pitch) would have signaled its presence at the junction of two otherwise independent clauses. In this and all subsequent tables, the unbracketed numbers are the raw frequencies, set alongside the normalized ones (per 10 000 words) in brackets. <?page no="150"?> Pam Peters 132 ACE 1m. w FLOB 1m. w FROWN 1m. w ICE-AUS written 400,000 w ICE-AUS spoken 600,000 w ART 257,000 2* (0.02) 0* 0* 6* (0.15) 0 0 * Additional instances of however as adversative conjunction with NP subjects following were found by Peterson (2009) Table 1: Raw and normalized frequencies of however as adversative conjunction with subject pronoun following Table 1 shows the paucity of examples of adversative however as an adversative conjunction in written Australian English, and its total absence from the two spoken corpora on the right-hand side (ICE-AUS and ART). The largest number of instances were found in the ICE-AUS written data, in run-on sentences used in unedited writing or letters (none in published writing). Yet they do not seem to be generated in ordinary spoken discourse. There were no straightforward examples of conjunctive however with adversative meaning in conversational texts, although it may underlie the transcription in lengthy monologic discourse from the courtroom or classroom, where the transcriber treated the speaker’s utterance as consisting of two sentences divided by an initial however, in accordance with editorial practice. s2b-028 (A): 244 The flying fox has got an opposible thumb but it just looks like a long hook on the front of its wing. s2b-028 (A): 246 However it certainly uses those thumbs to hang onto things. In an example from talkback radio, however seems to signal a change of focus, though the syntax is too convoluted for analysis in this fragment from a discussion about growing roses: ComE1: [16] Want them to be y’know cover you from the street a little bit however they get to the point where they’re not <inaudible>. Both these examples point to discoursal use of however as a change-of-topic marker in public discussions, but we still lack evidence of its use in private conversations. The data in Table 1 show only that however as an adversative conjunction is primarily associated with informal prose - likely to be edited out in the publication process. The data also suggest that conjunctive usage of adversative however is more visible in the southern hemisphere, which aligns with the fact that it was an Australian commentator (Murray-Smith) who first noted it in 1989. Yet the grammatical role of however is an artifact of the punctuation. The preceding comma makes it <?page no="151"?> Emergent Conjunctions 133 the contentious conjunction, whereas a heavier stop (noncontentiously) makes it an adverb (Peters 2004). Earlier evidence of conjunctive use of adversative however could probably be found going back centuries, when the punctuation system itself was more fluid than it is now. The Oxford English Dictionary records from C17 include citations with adversative however appearing in mid-sentence taking the conjunctive role, one of which certainly prefigures the C20 run-on sentence: Locke 1690 The Idea of Black is no less positive in his Mind, than that of White, however the cause of that Colour in the external Object, may be only a Privation. According to the C19 editors of the Dictionary, this usage of however is obsolete or archaic. Yet in the following century it comes under the scrutiny of prescriptive writers, who detailed a variety of prescriptive rules about the placement of however (whether it should or should not be used at the start of sentence (Webster’s Dictionary of English Usage, 1989). That this was symptomatic of its evolving status as a semi-coordinator does not seem to have been commented on. 4.2 ONCE as a temporal subordinator The earliest use of once as a temporal conjunction is an isolated example recorded in the Oxford Dictionary online from 1475. More regular use of once in this role only takes off in the mid-C18, in the fiction writing of Richardson (1747) and Sheridan (1761). Once operates as adverb, adjective and noun apart from its conjunctive role, but its versatility has never attracted critical comment from English usage commentators. Its use as a subordinator was endorsed by Fowler (1926) as “sound English enough”, echoed by Gowers (1965). While its reception in British English has been neutral, it seems to have passed under the radar in American English, and is not currently recognized at all in Merriam-Webster online. Actual usage of once as a subordinator was on a par in the standard corpora of American and British English of the 1960s (LOB and Brown), as reported in Kjellmer’s research (1997: 173). The use of once as a subordinator emerged slowly from earlier adverbial use through several stages. They are carefully documented by Kjellmer (1997) to show both the semantic extensions from its core meaning ‘on one occasion’ to ‘as soon as’; and the extension of its grammatical role from heading an adverbial phrase to acting as subordinator to nonfinite and ultimately finite clauses. Most of these intermediate roles are still current, as Kjellmer demonstrated, and the use of once as subordinator is only a minor percentage of the total (c. 16%). His data was based purely on written source material from the northern hemisphere, so it is of no small <?page no="152"?> Pam Peters 134 interest to add in later corpora from both hemispheres, as well as spoken discourse. The comparative frequencies, raw and normalized figures (in brackets) are presented below in Table 2. ACE 1m. w FLOB 1m. w FROWN 1m. w ICE-AUS written 400,000 w ICE-AUS spoken 600,000 w ART 257,000 33 (0.33) 42 (0.42) 39 (0.39) 9 (0.23) 51 (0.85) 28 (1.09) Table 2: Raw and normalized frequencies (per 10 000 words) for once as subordinating conjunction with subject pronoun following The data on the left-hand side of Table 2 support Kjellmer’s earlier findings (from the LOB and BROWN corpora) on the very similar frequencies of once as a subordinator in British and American English of the 1960s. The differences between the FLOB and Frown results here are again only minimal. The level of usage is relatively lower in both ACE and ICE-AUS written texts - but much higher in ICE-AUS spoken data. The level in ART is even higher, no doubt because once is often used in personal narrative to contextualize one’s experience and its longer-term effects, and common in free-ranging discussions on talkback radio. For example: COMe2 [C5] …You get used to it once you’ve lived there for a while. ABCe3 [C6] Once I’d started I found it a very interesting concept and a gripping story. Inspection of the corpus data showed that it was quite often used at start of sentence, as in the second example, to introduce background information. In less well-constructed sentences, a subordinate clause introduced with once was sometimes left hanging, without any following main clause. This routine use of once to head a sentence-initial clause shows it to be an incipient discourse formula, marking what is to be received as “given” information before proceeding to the new (Halliday and Matthiessen 2004: 93). Its increasingly frequent use in Australian speech at turn of millennium reflects its new-found value as a discourse structuring device. 4.3 THE WAY as subordinating conjunction The use of the way as a subordinating conjunction in the sense of ‘as, how’ stems from nonliteral use of way to mean ‘manner’, ‘mode of doing something’, which can be traced back to Old and Middle English. As a subordinator, the way is centuries old, with citations dating back to Shakespeare, Addison, Hardy, among others. Its use in “He doesn’t do it the way I do it” was counted as “literary English” in Marckwardt and Walcott’s study <?page no="153"?> Emergent Conjunctions 135 (1938: 28) of variable points of usage in American English. Elsewhere (1938: 109) they described it as “clearly good colloquial English” although “roundabout”. In British English its stylistic status was less secure, according to the findings of Mittins et al. (1970). The suggestion of using their sample sentence “Young girls do not dress the way their mothers did” in formal writing or speech elicited strong negative reactions. It was however tolerated in informal discourse. In the standard American and British corpora of the 1960s (Brown corpus, LOB corpus), use of the way as a conjunctive device was more common in fiction writing than nonfiction, although it appeared relatively more often in nonfiction in American English (Peters 1993). Grammatical commentary on this conjunctive use of the way is limited. The absence of relativizer after the way before a relative clause is illustrated by Quirk et al. (1985: 849, 1249-56) with some recognition of its adverbial role; but Huddleston and Pullum (2002: 1053-56) offer only examples of its nonadverbial use when commenting on zero relativization. In Biber et al. (1999: 624-29) the discussion of the way notes that it can cover an “adverbial gap” as well as the “relative gap”, and is strongly associated with a manner adverbial clause. However the fact that this is the context in which the way becomes a complex subordinator is not articulated. Biber et al. do nevertheless show that use of the way with zero relativizer and adverbial gap is very common in all registers of the Longman corpus, partly because there is no alternative wh-word (like where, when for adverbials of place and time). This strong association of the way with manner adverbial clauses is masked to some extent by the fact it can appear with zero relative and a complementary clause at several points of the main clause. The way can be prominently attached to the subject of the clause, as a kind of topicalizer for introducing a characteristic behavior or style, as in: FLOB K12: 24 The way he used my first name confirmed my suspicions. Or as a complement of clausal subject (dummy subject), in: FROWN L05: 131 It’s just the way they proceed. This use of the way as subject (predicative) complement following the verb BE has been noted as common in conversation (Biber et al. 1999: 448). The way also appears in other “object territory” roles, e.g. as object of verb, as in: ACE D13: 2575 He likes the way we care about the battler and the underdog. Here again, the way is an obligatory constituent of the main clause, and the subordinate clause attached to it by way of a zero relativizer remains a <?page no="154"?> Pam Peters 136 relative clause. Likewise when the way serves as complement of a preposition, it remains an obligatory element of the prepositional phrase or adjunct: ICE-AUS W2E-006 It is clear changes in the way we live and work and ultimately in the way we think about the world are necessary. ICE-AUS S1B-002(A) This is perhaps why she acts the way she does at the end of the play. In all these constituent roles, the way is an obligatory element of the main clause, and followed by a relative clause. But beyond them we find numerous instances in the corpus data where the way appears as a nonobligatory adjunct at the end of the main clause, making it the pivot between main clause and the following subordinate clause. In this context it seems to bear a dual role, as noted by Quirk et al. (1985: 442-43) with other adverbial constituents which serve both as adjunct in the main clause and subordinator to the following clause. This duality can be seen in relation to the way in examples such as: ICE-AUS S1A-031(B) It was quite incredible the way we got the case of beers over the fence. ICE-AUS S2A-033(A) They may not necessarily work the way you would expect them to. In cases like these, the way stands ambiguously on the boundary between the two clauses, and becomes head of the subordinate clause. This is particularly clear when the way is actually separated from the main clause by means of a comma, making it an appositional adjunct to main clause, though still followed by a subordinate clause. For example: FROWN K26 119 Allan’s dad stopped to talk to the cop, the way he always did. The punctuation puts the way and the subordinate clause outside the boundary of main clause, and the way takes on an independent grammatical role as a subordinator. There is no duality about it. The table below presents the relative frequencies for the various constituent roles of the way, as obligatory and nonobligatory clausal roles. They show the matrix from which its role as subordinator role has emerged, amid a plethora of instances where it simply connects without a relativizer to a following relative clause. <?page no="155"?> Emergent Conjunctions 137 Constituent roles of the way ACE 1m. w FLOB 1m. w FROWN 1m. w ICE-AUS written 400,000 w ICE-AUS spoken 600,000 w ART 257,000 1. As subject with rel. cl 4 (0.04) 7 (0.07) 10 (1.0) 3 (0.075) 21 (0.35) 3 (0.11) 2. As compl. to subject 11 (0.11) 5 (0.05) 12 (0.12) 1 (0.025) 25 (0.42) 5 (0.19) 3. As object of verb 17 (0.17) 5 (0.05) 6 (0.06) 2 (0.05) 10 (0.17) 6 (0.23) 4. As compl. to prep. 19 (0.19) 14 (0.14) 9 (0.09) 10 (0.25) 23 (0.38) 5 (0.19) 5. As obligatory adjunct 3 (0.03) 2 (0.02) 2 (0.02) 0 0 1 (0.04) 6. As nonoblig. adjunct of clause 15 (0.15) 8 (0.08) 11 (0.11) 1 (0.025) 12 (0.2) 5 (0.19) 7. Nonoblig. adjunct, following punctuation 3 (0.03) 1 (0.01) 5 (0.05) 0 0 1 (0.04) 8. Total nonoblig. adjunct (ll. 6+7) 18 (0.18) 9 (0.09) 16 (0.16) 1 (0.025) 12 (0.2) 6 (0.23) TOTAL (ll. 1-7) the way with following sub. clause 72 (0.72) 42 (0.42) 55 (0.55) 17 (0.43) 91 (1.52) 26 (1.01) Table 3: Raw and normalized frequencies (per 10 000 words) for the various roles of the way as obligatory clause constituent and optional adjunct (with following subordinate clause introduced by personal pronoun) Table 3 shows that the uses of the way as nonobligatory adjunct (line 8) are a relatively small total in comparison with the sum total of all uses of the way with a following subordinate clause. Though clearly related by virtue of the zero relativizer, the more independent grammatical status of the nonobligatory uses of the way affects its relationship to the following clause, and its inherent role. Its ambiguous grammatical status is evident even when it is an obligatory adjunct (e.g. with verbs like behave and treat as found in the examples of line 5 of the table); and this is also of some significance for the emergence of the subordinator. Grammarians now recognize the importance of ambiguous contexts of use for the reanalysis of <?page no="156"?> Pam Peters 138 language and language change (Diewald 2002: 103). The grammaticalization of the NP the way as a subordinato takes place through the existence of such contexts, and the persistence of its meaning through expanding contexts over a formative period (Traugott 2010). The data on left-hand side (first three columns) of Table 3 suggest that the subordinator role of the way in heading adverbial clauses is found relatively more often in written Australian and American English than in British English. This is in line with greater criticism of it found in Mittins et al.’s usage research (1970), and fact that Australian English is more tolerant of informal styles (Peters 2009: 397). Its popularity in Australian speech is clear in the normalized data from the ICE-AUS and ART data on the right-hand side, where use of the way as adverbial adjunct with a following subordinate clause is higher than in any of the other corpora. Also noteworthy are the levels of usage in ICE-AUS of the way in all its other constituent roles, making it a high frequency unit of Australian speech, public and private. One further question to explore in relation to the way is how common the elliptical construction is in comparison with its nonelliptical counterparts. Since the way is fostered in spoken discourse, we might expect its less elliptical paraphrase the way that, and the full form the way in which to be more common in writing. Comparative data on the three alternatives are set out in Table 4 below, to show their distribution over spoken and written discourse, as well as their appearance as obligatory and nonobligatory constituents of the main clause. <?page no="157"?> Emergent Conjunctions 139 ACE FLOB FROWN ICE-AUS written 400,000 w ICE-AUS spoken 600,000 w ART 257,000 1. Nonoblig. adjunct the way (= Table 3, row 8) 18 (0.18) 9 (0.09) 16 (0.16) 1 (0.025) 12 (0.2) 6 (0.23) 2. Oblig. constituent the way that 2 (0.02) 0 0 0 I7 (0.28) 0 3. Nonobligatory constituent the way that 0 0 0 0 1 (0.017) 0 4. Oblig constituent the way in which 2 (0.02) 4 (0.04) 4 (0.04) 1 (0.025) 8 (0.13) 0 5. Nonobligatory constituent the way in which 0 0 4 (0.04) 0 0 0 Table 4: Raw and normalized frequencies (per 10 000 words) of elliptical and nonelliptical forms of the way, as obligatory and nonobligatory constituent with a following subordinate clause The data in Table 4 is confined to examples with pronoun subjects, for the methodological reasons indicated in section 3 above. Additional data for the use of the way that with NP subjects as obligatory constituents can be found in ACE, FROWN, and ICE-AUS; and for the way in which with NP subjects, in line with Biber et al.’s (1999: 621) observation. Yet without them, Table 4 still shows that the less elliptical form the way that and the nonelliptical the way in which are associated with obligatory rather than nonobligatory constituents of the clause (compare lines 2/ 3 and 4/ 5 of the table). This relationship holds across the board in all the corpora, except for the cluster of 4 examples of the way in which found in FROWN. These actually occur in a single sentence, as a quartet of appositional clauses after a colon: FROWN G15: 191-201 This complicity between social text and its viewer takes many, familiar forms: the way in which we are complicit in the concepts of modes of address and ideological problematic used by David Morley and John Fiske, the way in which we think the relationship of viewer inter- <?page no="158"?> Pam Peters 140 discourse and television’s heterogeneity of discourse (Ien Ang), the way in which we think about ideologies contending with one another for our consent (in, among other places, Bill Nichols’ essay ‘Ideological and Marxist Criticism’), or the way in which television works to construct a complicity with the viewer around the construction of the ideal family (in John Ellis and Jane Feuer). This repeated use of the way in which in an annotated-bibliography-of-asentence (more than 100 words long) is clearly a rhetorical device by the writer concerned, and unique in comparison with all the other data sets in same line. That apart, the data suggest that fleshed-out forms of the way are rarely found as nonobligatory constituents. Somewhat surprisingly, the use of nonelliptical forms of the way is most noticeable in the spoken data of ICE-AUS, and holds for both constructions (the way that, the way in which). The latter seems to contrast with Biber et al.’s (1999: 628) finding that it was most common in academic prose in the Longman corpus, and hardly registered in fiction, news or conversation. However closer inspection of the ICE-AUS data shows that these nonelliptical constructions all occur in public speech (academic, legal, corporate, religious), where the discourse style is closer to institutional and formal writing than to conversation. The comparative data in Table 4 confirm the earlier finding, that the fully elliptical use of the way correlates strongly with optional adjuncts where adverbial gapping occurs. At that point the way with zero relativizer takes on the role of subordinator, heading an adverbial clause of manner. Elsewhere, when the way is an obligatory constituent of the main clause, the following clause remains a relative clause, whether it connects with the following clause as by a zero relativizer, or by the presence of a relativizer (that or (in) which). This indicates a kind of complementary distribution between the elliptical and nonelliptical constructions with the way, rather than the evolution of the elliptical form. 5. Findings The three case studies present some common patterns for all three items to emerge as conjunctions in spoken and written English. The subordinator role of once is well established in writing (Kjellmer 1997), though still expanding its reach into spoken discourse by the data reported above. For adversative however, the extension from conjunctive adverb to semicoordinator also seems to be taking place first in writing (at least in Australian English), and is not yet visible in speech. The way as subordinator for adverbial clauses following nonobligatory adjunct is well established in speech and writing, by the data presented in Tables 3 and 4. The highest <?page no="159"?> Emergent Conjunctions 141 levels found in this research are in the ACE (written) corpus and the ART (spoken) corpus. The fact that the elliptical construction is well established in academic writing in the Longman corpus was the more surprising finding for Biber et al. (1999: 629), but it suggests that use of the way as subordinator is now stylistically neutral. No longer is it felt to be colloquial - as in the findings of earlier surveys in the US (1930s) by Marckwardt and Walcott (1938), and in the UK (1960s) by Mittins et al. (1970). The research reported in this paper confirms its acceptability in standard published texts in the northern and southern hemispheres in the last decade of C20. In each case, the conjunctive role is an extension to the underlying adverbial identity of the item, and underpinned by fresh applications in discourse which come to light in corpus data. We noted that however serves to signal a change of focus, and once to package up given information in relation to the new. The way, used in relation to the subject of a clause is often a kind of topicalizer. These various discourse functions align with Hopper’s (1988) hypothesis, that emergent grammar is a product of discourse. 6. Conclusions on emergent conjunctions The three cases of emergent conjunctions discussed above are not isolated examples. Each one has its counterparts in other adverbs or adverbial phrases, all capable of taking on the role of conjunction. For example, alongside • however: the parallel cases of conjunctive adverbs thus, therefore, both also used as semi-coordinators though with different logical senses (Peterson 2009) • once: the similar cases of temporal adverbs directly, immediately, also used occasionally as subordinators (Kjellmer 1997) • the way: similar cases of the place, the reason, the time with zero relativizer (Biber et al. 1999: 627-28), also usable as nonobligatory adverbial adjuncts/ subordinators The three types of emergent conjunctions just listed are more and less obviously marked as adverbials, which might be thought to affect their extension into conjunctive roles. However and therefore are both marked as adverbs by their second elements (by indefinite suffix -ever, -fore a suffixal form of for). Though the adverbial marking of once is masked by medieval respelling (Oxford English Dictionary online), that of directly and immediately is patent. Clearly their adverbial identity has not prevented them from <?page no="160"?> Pam Peters 142 assuming conjunctive roles. The way is not obviously adverbial in its elliptical form, but marked as a noun phrase and often constrained by its other constituent roles. It has nevertheless established a subordinator role for itself where there is an adverbial gap, and shows how other NP adverbials such as the place, the time can extend into subordinator roles. The emergence of two of these conjunctive items (however, the way) has taken place in the face of prescriptive criticism of the linguistic extensions that they entail. That they have established or are establishing themselves in English written and spoken discourse is a reflection of the strength of emergent grammar. At the same time, the emergence of new subordinators such as once and the way to expand the word class argues against it being residual. Biographical sketch Pam Peters is an Emeritus Professor in the Linguistics Department at Macquarie University, NSW Australia, where she continues to research and write on aspects of English grammar and usage. She is the author of the Cambridge Guide to English Usage (2004) and the Cambridge Guide to Australian English Usage (2007). She led the collection of sample corpora in Australia (ACE, ICE-AUS), and is now President of the Australian National Corpus. She was a member of the Editorial Committee of the Macquarie Dictionary for twenty years, and is now engaged in developing online termbanks of key terminology in the sciences and social sciences, for use in early university study and by the general public. <?page no="161"?> Emergent Conjunctions 143 References Australian Oxford Dictionary. 2004. (Second edition.) Melbourne: Oxford University Press. Biber, Douglas, Geoffrey Leech, Stig Johansson, Susan Conrad, and Edward Finegan. 1999. Longman Grammar of Spoken and Written English. London: Longman. Burchfield, Robert. 1996. Fowler’s Modern English Usage. Oxford: Clarendon Press. Diewald, Gabriele. 2002. “A Model for Relevant Types of Contexts in Grammaticalization”. In Ilse Wischer and Gabriele Diewald (eds.). New Reflections on Grammaticalization. Amsterdam: John Benjamins, pp. 103-120. Fowler, Henry. 1926. A Dictionary of Modern English Usage. Oxford: Clarendon Press. Garner, Brian. 1998. A Dictionary of Modern American Usage. New York: Oxford University Press. Gowers, Ernest. 1965. Fowler’s Modern English Usage. Oxford: Clarendon Press. Halliday, Michael, and Christian Matthiessen. 2004. An Introduction to Functional Grammar. (Third edition). London: Hodder Education. Hopper, Paul. 1988. “Emergent Grammar and the a priori Grammar Postulate”. In Deborah Tannen (ed.). Linguistics in Context: Connecting Observation and Understanding. New Jersey: Ablex, pp. 103-120. Huddleston, Rodney, and Geoffrey Pullum. 2002. Cambridge Grammar of the English Language. Cambridge: Cambridge University Press. Kearns, Kate. 2009. “Zero Complementizer, Syntactic Context and Regional Variety”. In Pam Peters et al. (eds.). Comparative Studies, pp. 243-262. Kjellmer, Goran. 1997. “The Conjunction once”. In Udo Fries et al. (eds.). From Aelfric to the New York Times. Amsterdam: Rodopi, pp. 173-181. Marckwardt, Albert, and Fred Walcott. 1938. Facts about Current English. New York: Appleton, Century-Crofts. Merriam-Webster Dictionary: www.merriam-webster.com. Accessed 4/ iii/ 2011. Mittins, W.H., Mary Salu, Mary Edminson, and Sheila Coyne. 1970. Attitudes to English Usage. Oxford: Oxford University Press. Murray-Smith, Stephen. 1989. Right Words. (Second edition.) Melbourne: Viking, Penguin Australia. Oxford English Dictionary: www.oed.com. Accessed 4/ iii/ 2011. Peters, Pam. 1993. “Corpus Evidence on Some Points of Usage”. In Jan Aarts et al. (eds.) English Language Corpora: Design, Analysis and Exploitation. Amsterdam: Rodopi, pp. 247-255. Peters, Pam. 2004. Cambridge Guide to English Usage. Cambridge: Cambridge University Press. Peters, Pam. 2009. “Epilogue: Collective Findings and Conclusions”. In Pam Peters et al. (eds.). Comparative Studies, pp. 387-399. Peters, Pam. 2011 forthcoming. “Usage Guides and Australian English, Prescription and Description”. In Ulrich Busse et al. (eds.). Codification, Canons and Curricula. Bielefeld English and American Studies. Bielefeld: Bielefeld University. <?page no="162"?> Pam Peters 144 Peters, Pam, Peter Collins, and Adam Smith (eds.). 2009. Comparative Studies of Australian and New Zealand English: Grammar and Beyond. Amsterdam: John Benjamins. Peterson, Peter. 2009. “Commas and Connective Adverbs”. In Pam Peters et al. (eds.). Comparative Studies of Australian and New Zealand English: Grammar and Beyond. Amsterdam: John Benjamins, pp. 277-292. Quirk, Randolf, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman. Schneider, Edgar. 2007. Postcolonial Englishes. Cambridge: Cambridge University Press. Strunk, William and E.B. White. 1959. The Elements of Style. New York: Macmillan. Traugott, Elizabeth. 2010. “The Persistence of Linguistic Contexts over Time: Implications for Corpus Research”. Plenary paper presented at the 31st ICA- ME Conference, Giessen Germany. Webster’s Dictionary of English Usage. 1989. Springfield, Mass.: Merriam-Webster. <?page no="163"?> Ursula Schaefer Romance Suffixation in the History of English New Ways of Solving an Old Problem Abstract In the late Middle English period borrowings from French are legion. Simultaneously we see that Romance affixation on native lexical material becomes productive. My contribution looks at this phenomenon from the perspective that the borrowing from French (and Latin) is influenced by the norms of discourse traditions. This takes up a problem raised by Andreas Fischer (2001) with regard to different approaches to lexical borrowing. While he dismissed the sociolinguistic approach for this field as having to work on uncertain assumptions, I argue that discussing borrowing in terms of discourse-traditional norms within the domain of writing may serve a similar function as the sociolinguistic approach, yet work on quite safe grounds. I will substantiate my argument with the analysis of -able derivatives from the late 15 th century. 1 Introduction Some ten years ago Andreas Fischer published an article with the title “Lexical Borrowing and the History of English: a Typology of Typologies”, where he took stock of three “attempts to classify lexical borrowings” (2001: 97): a first set that runs along the lines of morphology, a second that applies semantic criteria and, finally, the sociolinguistic approach. In conclusion he states: Of the three types of typologies discussed here, the third, although the most attractive because of its sociolinguistic orientation, is possibly the least useful for a study of lexical borrowing […]. (Fischer 2001: 110) Fischer’s immediately convincing reasons for this dismissal run as follows: […] lexis is an unreliable indicator of the precise nature of a contact situation, and the socio-historical evidence necessary to reconstruct the latter is often not available. (Fischer 2001: 110) Although this is certainly true if we concentrate on the contact among speakers of the languages involved, Fischer himself assumes that “borrowing from written sources” (2001: 109) should also be taken into consideration when discussing processes of lexical borrowing from French (and Latin) in the later English Middle Ages. <?page no="164"?> Ursula Schaefer 146 In my contribution I will therefore suggest that if we heuristically concentrate on the domains of writing as conceivable - yet heavily understudied - places of linguistic contact the prospects of accounting for that “massive influx of French loanwords between 1250 and 1400” (Fischer 2001: 109) are, for a number of reasons, much more favorable. Recent studies in Contact Linguistics are, however, − at least at first sight − of little help because they have largely bracketed off contact in writing, as this does not seem to fit into the parameters of substrate/ superstrate/ adstrate contact situations. 1 As Fischer rightly states: These labels are appropriate in a general way, but they grossly simplify the individual contact situations. This is specifically true of the contact of English with French (and Latin). (Fischer 2001: 107, fn. 15) A widespread way of conceptualizing the conditions that triggered the borrowing from French (and Latin) is to discuss them in terms of “prestige” and “high” vs. “low” varieties. This suggests a link to Labov’s classification of ‘borrowings’ that trigger “changes from below” and “changes from above” (1994: 78). First it should be noted that Labov is speaking about phonological transfer. Moreover he does not use the notion of borrowing in the way Contact Linguistics does because his borrowing is that from one variety of a language into another. 2 Although scholars in Contact Linguistics explicitly only exclude geographical varieties, i.e. dialects, from their considerations in terms of “language contact” (cf., e.g., Thomason 2001: 2), they do not include social varieties of “one” lanuage in contact either. However, as we will see in a moment, register borrowing may indeed be a helpful concept when conceiving the contact situation in late medieval England as this kind of borrowing must have happened - more or less - with “full awareness” (Labov 1994: 78). Another critical point in Labov’s concept of changes from above is the fact that the “dominant class” that introduces such changes is not identical with the “speech communities that have higher prestige” (Labov 1994: 78). Therefore, if we attempt to apply Labov’s notion of “change from above” to account for the heavy lexical borrowings from French (and Latin) into late medieval (written) English in terms of prestige we have to wonder who, after all, actually were the agents of this change. In order to answer this question we first of all need to avoid the circularity against which Fischer warns when we interpret the data and apply socio-linguistic parameters such as that of adstrate or superstrate situations (2001: 110). This danger becomes apparent when we look at the following 1 For a recent survey of the relevant literature cf. Weber (2010: 76-83). 2 Further on in his 1994 monograph Labov uses the term in a stricter sense when he comments, e.g., on the fact that in today’s English minimal pairs of the kind fail vs. veil the v-initial forms are always French loans (1994: 141). <?page no="165"?> Romance Suffixation in the History of English 147 remarks by Dalton-Puffer. First she states that, with the exception of “the very upper echelons of the feudal aristocray”, within the span of “at most three generations after the Conquest, English was the mother tongue of all social classes” (1996: 6). Nevertheless, she goes on, “[…] French (and Latin) continued to play an important role in the intellectual life of that society, giving these two languages great social prestige” (1996: 6). Now, did those languages have prestige because of the content they transported, or was the fact that these languages had a specific prestige the reason why such content was expressed in no other language? In conclusion Dalton- Puffer observes: That French has had a great impact upon the English language is undeniable. […] The linguistic situation in England is probably best described in terms of diglossia. It was a diglossic situation where H kept losing ground to L. (Dalton-Puffer 1996: 6) This suggests, among other things, that the multilingual space of the period under consideration is organized in terms of a constant equilibrium: if the use of one language increases, that of the other(s) automatically decrease(s). This ignores two facts: first and foremost, English did not simply substitute French (or Latin). Moreover, the use of writing generally increased in this period, and it is in this increase that English was very much involved. A conspicuous example from the 15 th century is that of private letter writing (cf., e.g., Nevalainen 2004; Nurmi and Pahta 2004), as it can be observed, for instance, with the Pastons. This was a new social practice and it did not establish itself at the expense of French or Latin, but rather with their help (cf., e.g., Schaefer 2005), as we can see from the fact that the authors of these letters closely follow the available Latin and French models (cf., e.g., Davis 1954; Camargo 1995). Therefore the heuristic framework in which we conceptualize the borrowing processes in the late Middle English period has to be reconsidered. In doing so, we first need to take note of the way in which the trilingual communicative space of the period was actually structured with regard to the use of the available languages in writing. 3 This is necessary because this specific space sets the frame for linguistic contact and conditions the individual borrowing processes. In other words: the driving forces that have motivated the heavy influx of French (and Latin) linguistic material into English may, I suggest, indeed have been dominance and prestige. Yet, this dominance is not simply exerted by social class membership, and prestige is not simply a class-oriented asset. We should rather think of the borrowings as the results of complying with the established stylistic norms. Materializing in cultural practices, norms in general are, by definition, social, and the 3 I adopt the notion of “communicative space” (“Kommunikationsraum”) from Oesterreicher 2001, here: p. 1564. <?page no="166"?> Ursula Schaefer 148 various forms in which linguistic norms are realized are by no means random (Koch 1988). We speak of them in terms of “registers”, “styles”, or “discourse traditions” which shows that their realization runs along the lines of an internally organized linguistic space. Therefore, I hope to show that the concept of “diglossia” is by far too simple to account for the influx of borrowings in late medieval English. After having established this heuristic framework in the first part of the next section, I will turn to the specific field of word formation with borrowed affixes. As shall be seen, this field has been discussed in the literature in a very heterogeneous way. Suffice it here to point to the problem of whether affixation with Romance items in English should be considered as “structural borrowing” (cf. Winford 2005: 386-88). A related issue which shall be discussed at the end of the theoretical section is the received wisdom that in the history of late Middle English first came the affixation of borrowed words with indigenous elements and only later that of native words with borrowed affixes. Although this chronology is partly underpinned with the data, the theoretical “explanation” - as much as this may superficially stand to reason - seems to be much more the result of the (structuralist) approach than an account of what “really” happened in the borrowing processes. All of this should well prepare us to look at late Middle English affixation. To profile my argument I shall concentrate on the very productive suffixation with -able. My aim is to show that taking into account the socio-cultural frame of trilingual England in the 14 th and 15 th centuries may prevent us from rash conclusions as to the contemporary mechanisms of borrowing. 2 A Theoretical Framework for Late Middle English Borrowing Processes 2.1 The Trilingual Communicative Space Coming back to the question of who, after all, actually were the agents of the numerous borrowings from French (and Latin) in the 14 th and 15 th centuries, we first note with Weinreich (1966: 1) that it takes bilinguals in order for successful borrowing from one language into another to occur. Now, while it is certainly correct to follow Rolf Berndt’s observation that - when it comes to speakers of English and French - “bilingualism was not widespread” in post-Conquest England (Fischer 2001: 109), ever since the 13 th century multilingual English writers were the rule. However, using Latin, French or English was not a matter of free choice. A very simple example is the so-called Statute of Pleading issued in 1362. In the syntactically circumstantial so-called “courtly style” (cf. Burnley 1986) it says in <?page no="167"?> Romance Suffixation in the History of English 149 French that henceforth the language of pleading in court be English and that the records of the court procedures should be written in Latin. From our modern stance we may ask: if Edward III really wanted to promote the use of English, why did he not issue the Statute in English? The answer is simple: in the 14 th century − and well into the 15 th − the language for statutes was prevalently French, as much as the language of record was Latin (cf. Clanchy 1993: 197-223). This division of labor among the languages suggests that the “different” languages are, in fact, stylistic or register options. As Burnley states: Although Latin literacy was somewhat distinct from that in the vernaculars, it is quite apparent that the division between English and foreign languages was not comparable with that existing today. Indeed, Anglo-French was hardly a foreign language; rather a special style used for certain technical purposes. (Burnley 1989: 41) We may also surmise that once English is increasingly used in writing, the authors who are proficient in such a “foreign language” style take recourse to linguistic material from that language if they stay in the same register (cf., e.g., Rothwell 2000). In view of the “prestige” argument Irma Taavitsainen rightly observes that this may well apply to changes which “originate in casual and everyday usage rather than in monitored and self-conscious styles” (2001: 188). However, when “creating a functional variety like scientific English […] prestige is not a function of socio-economic power but education” (2001: 188). The register or style models in the other language may therefore be considered as providing “a standard which can be seen as the norm toward which vernacular writing tended, and influenced both the macrostructures of discourse and the microlevel of linguistics features” (Taavitsainen 2001: 185). Over twenty years ago Peter Koch called such norms […] Diskurstraditionen, die offensichtlich als Diskursnormen intersubjektiv gültig sind und den jeweiligen Sinn eines Diskurses mitkonstituieren: Textsorten, Gattungen, Stile usw. Es handelt sich dabei um Komplexe von Diskursregeln, die […] nicht universal sondern historisch und konventionell sind und […] nicht (oder allenfalls zufällig) an Sprachgemeinschaften gebunden sind. (Koch 1988: 342; italics in the original) 4 As vernaculars generally enter the realm of literacy within the frame of specific discourse traditions (cf. Oesterreicher 1993: 277), English also 4 ‘[…] discourse traditions which, as discours norms, are obviously intersubjectively valid and add to constituting the meaning of a discourse: text types, genres, styles etc. They are complexes of discourse rules which [are] not universal but historical and conventional and […] which are not or only coincidentally tied to a linguistic community’; cf. also Koch 1997. <?page no="168"?> Ursula Schaefer 150 seems to have re-entered the written tradition along such normative lines which had already been set by various non-English models. The result of one such late medieval English norm-compliance is the so-called Chancery English which stylistically follows what Burnley has dubbed the “curial style” (1989, 2001). This variety, which Laura Wright has acutely characterized as “one functional variety of written English, with contemporaneously a very limited readership” (1994: 113) is most visibly embedded in that literary scenario where Latin, French and English are used - literally - side by side (cf. Weber 2010). In such an environment - and in view of the stylistic norms at hand - “dominance” seems a suitable concept after all. However, as has been stated before, this would not be any kind of social class dominance but rather a linguistic one. To this and other considerations as to which forces are at work when languages - or, more pertinently here: styles - come into contact we shall turn in the next subsection. 2.2 The Contact Phenomenon of Borrowing Within the multilingual cultural scenario I have just tried to sketch, recent studies in Contact Linguistics at first sight seem of limited help as they usually consider speakers - in the strict sense of the notion - as agents (or patients) of the various processes triggered by linguistic contact. Here borrowing is a process in which the (native) speaker of a language L1 - the recipient language (RL) - transfers linguistic material from a foreign language L2 - the source language (SL) - into this RL. As van Coetsem (1988) has illustrated, borrowing and imposition are the two (main) results of language contact. Whether the one or the other is at work depends on which language is the dominant one for the speakers. Linguistic - as opposed to social - dominance is characterized by van Coetsem as follows: A bilingual speaker […] is linguistically dominant in the language in which he is most proficient and most fluent (which is not necessarily his first or native tongue). (van Coetsem 1995: 70) In commenting on van Coetsem’s approach Winford specifies that borrowing takes place from a non-dominant into the dominant (primary) language, while imposition takes place from the dominant into the nondominant language (2005: 376-77). The first type “typically involves mostly vocabulary, though some degree of structural borrowing is possible”, and the latter “tends to involve mainly phonology and grammatical features, though imposition of vocabulary can occur as well” (Winford 2005: 377). Albeit the distinction very much stands to reason, Winford also concedes that there may occur “interaction between the two transfer types” (2005: 383). And here he refers explicitly to the influence of French on Middle English: <?page no="169"?> Romance Suffixation in the History of English 151 […] it seems reasonable to assume that lexical incorporations from French into Middle English were introduced by both English-dominant speakers (under RL agentivity) and by French-dominant speakers (under SL agentivity). (Winford 2005: 383) By the 14 th century it is unquestionable that writers, say, in the Royal Chancery had English as their mother tongue. Yet they were proficient in the “curial style” in its Latin and French form, and we may infer that one of these two - or even both - must have been the dominant language(s) providing the discourse-traditional models when it came to writing in English. Thus, the import of lexical material into English by these writers could - in van Coetsem’s terminology - be conceived of as imposition. Unfortunately this would largely contradict the observation that imposition mainly takes place on the phonological and grammatical level. But again I would like to point out that we are dealing here not with speakers but with writers: the authors who imposed the discourse norms on their English may not have been French-dominant speakers, yet they certainly were French- (and Latin-)dominant writers. Looking at the heavy influx of loans (a) in terms of the writers’ multiligualism and (b) as instances of imposition of discourse norms thus solves the problem we saw with the Labovian “change from above” in which the socially dominant classes borrow from “speech communities that have higher prestige” (1994: 89). If we agree that here it is the discourse norm cast in the foreign language which has “higher prestige”, we can, for one thing, solve that infelicitous category of class-oriented ‘social dominance’. Discourse norms are, indeed, social norms, yet not the kind Labov was thinking of. Moreover, this conceptualization of the borrowing scenario as one that is oriented along the lines of discourse traditions allows us to conjoin social and linguistic dominance which, as Winford insists, otherwise should be strictly kept apart (Winford 2005: 377). This heuristic conjunction is supported by Matras. Though once again developed with regard to the effect of language contact on spoken language, he also accounts for borrowing by way of dominance. He does, however, shift the emphasis to pragmatic dominance and characterizes the pragmatically dominant language as […] the language which, in a given moment of discourse interaction, is granted maximum mental effort by speakers. This may be the speaker’s first language, or one that is dominant for a particular domain of linguistic interaction, or one that exerts pressure due to its overall role as the majority language that is culturally prestigious or economically powerful. (Matras 2000: 577) I do not think we are stretching this concept too far if, for the present purpose, we extend “dominant for a particular domain of linguistic interaction” to ‘dominant for a particular discourse tradition’. Thus I would argue that late medieval English scribes or authors who prevalently moved in the <?page no="170"?> Ursula Schaefer 152 realm of Latin and French literacy when dealing, say, with administrative matters in writing “granted” this “maximum mental effort” to the established non-English discourse tradition. This becomes easily apparent on the general level of lexematic borrowing and, more specifically, say, on the lexico-syntactic level with creations such as the calque connective notwithstanding (from Latin/ French nonobstant(e); cf. Weber 2010: 135-163), which has kept the smack of its original ‘legalese’ discourse tradition. Things may, however, prove more complicated when it comes to accounting for affixation with French (and Latin) bound morphemes. 2.3 The Status of Loan Affixation First and foremost we have to clarify whether word formations with French (and Latin) affixes really are instances of “structural borrowing”, as has been claimed by Sarah Thomason (2001: 79). Structural borrowing ranks relatively low on her ‘borrowing scale’ (cf. Thomason 2001: 70-71), which means that structural elements are much less likely candidates for borrowing than, say, non-basic nouns. To approach this problem adequately, it is necessary to ask whether prefixes or suffixes may be subsumed under “structural features”. In her discussion of the various degrees of contact intensity Thomason remarks: You need not be at all fluent in a language in order to borrow a few of its words; but since you cannot borrow what you do not know, control of the source language’s structure is certainly needed before structural features can be borrowed. This is the reason for the split between lexical and structural borrowing […]. (Thomason 2001: 69) Winford, in his turn, addresses the borrowing scenario that we are interested in here as follows: The important point, for our purposes, is that [… the] morphological innovations were introduced indirectly through lexical borrowing. Middle English speakers clearly did not isolate morphemes like -able in the relevant French words and import them independently of the stems to which they were attached. (Winford 2005: 387) 5 Both observations are certainly true in their own right. What needs to be discussed, however, is the conflation of both “structrual features” (Thomason) and “morphological innovations” (Winford) with the question of 5 In his 2003 monograph Introduction to Contact Linguistics Winford adds to the same course of reasoning the general observation that the otherwise heavy constraints on structural borrowing “may be weaker in the case of derivational morphology (which is more lexical in nature anyhow) […]” (Winford 2003: 62). The bracketed remark translates, I think, into a general doubt whether derivational morphology is a ‘structural’ phenomenon to begin with. <?page no="171"?> Romance Suffixation in the History of English 153 productivity of foreign-item affixation. For one thing, prefixation and suffixation have been established word formation processes in English well before French could have any influence on English. Hence individual affixations with Romance morphemes may be “morphological innovations” in substance but not in function: on the structural level etymologically foreign affixes largely perform the same structural word formation duties as the indigenous affixes do or did. In the following remark Marchand argues half way between Thomason and Winford. In contrast to the native affixation of loans, he states, when […] foreign affixes are added to native words […] the assimilation of a structural pattern is involved, not merely the adoption of a lexical unit. Before the foreign affix can be used, a foreign syntagma must have become familiar with speakers so that the pattern of analysis may be imitated and the dependent morpheme be used with native words. This is much more complicated. (Marchand 1969: 210-11) Marchand’s use of the notion ‘structural’ is ambiguous here, as it is with Thomason. This ambiguity is, however, easily neutralized if we make a clear distinction between structures relating to word-formation and any other structural phenomena such as, e.g., inflection. Besides this indispensible distinction to which Marchand would fully agree (cf. 1969: 4 and 210), there is another flaw in his argument, namely that of the historical order of the linguistic events. Before we turn to this question I have to take up again the argument that borrowing has much to do with complying with the demands of register. In this respect the following general remark by D. Gary Miller is worth considering: Chaucer and Wyclif use popular hybrid formations but avoid Romance suffixes on native bases for abstract nouns and other terms of higher register. […] That discourse canon [in the higher register] would block the application of prestigious Latin/ French suffixes to bases of lower register. (Miller 1997: 245) Miller does, in a way, follow a heuristic idea similar to mine, although I wonder how a suffix may be more or less ‘prestigious’. And yet another category is somewhat problematic: what does he mean by “popular hybrid formations”? What makes, say, knowable, which Chaucer uses in the prefixed negative form unknowable in his Boethius translation for ignorabilis (cf. OED ³2010, November, s.v. knowable), more popular than, say, seeable? 6 How may we identify “bases of lower register”? Is know ‘lower’ than see? These are questions to which I do not have an answer. I think that breaking register demands down to affixes as such is problematic. What we 6 All quotes from and other references to the OED have been double-checked on April 24, 2011. If the reference is made to an updated entry of the third edition, I note this as “OED ³+ YEAR . month”. <?page no="172"?> Ursula Schaefer 154 can do is look into such creations in their immediate context and - wherever this is possible - see them before the backdrop of their models. Hence this will be our next focus. 3 The Example of the Deverbal Suffix -able 3.1 General Considerations The OED ( ³2009, September) gives an extensive etymological entry for the deverbative adjectival suffix -able. After commenting on the occurence of the suffix in Latin and (Anglo-Norman) French it says: […] the suffix soon extended to the derivation of deverbal adjectives within English, as shown clearly by hybrid formations on bases ultimately of Germanic origin (e.g. MISTROWABLE adj., UNSPEAKABLE adj., n., and adv., SPEAK- ABLE adj., DOABLE adj., TAKEABLE adj., etc.), and probably also by formations on bases of Romance origin which are not paralleled in French (e.g. PRAISABLE adj., PRAYABLE adj.). (OED ³2009, September; s.v. -able suffix; Etymology) Here the OED argues that the productivity of -able on indigenous bases presupposes the influx of a large number of French loans thus suffixed. The sheer number of such loans would make the pattern transparent and enable the isolation of the suffix which then could create new forms in English with either native bases or previously not suffixed loans. 7 Although this reasoning may sound structural(istical)ly cogent, I strongly doubt that the quantitative argument really holds. Take the following first evidence for mistrowable in the OED: a1382 Bible (Wycliffite, E.V.) (Bodl. 959) Baruch i. 19 - We weren mystrowable [a1425 L.V. vnbileueful; L. incredibiles] to þe lord oure god. The first translator certainly did not need a host of French deverbal -able adjectives to get from Latin incred-i-bilis to English mistrow-able. 8 He saw 7 For an extensive discussion of the ME -able formations s. also Dalton-Puffer (1996: 182-184); Miller (1997: 245-246); Miller (2006: 225-232) also discusses the etymological background. 8 The L.V. solution vnbileueful for Latin incredibiles is very interesting. I have recently looked into Chaucer’s attempts to translate the Boethian notions of beatitudo and felicitas. His solutions blisfulnesse and welefulnesse show that adjectival derivations of the type ‘noun + -ful’ are good stepping-stones to get to the abstract noun which in Latin is also deadjectival (Schaefer in print). However, the E.V. also tries to render a Latin -bilis adjective with an indigenous derivate in -ful; cf. OED (²1989, s.v. herien - derivate heryful): “1382 Bible (Wycliffite, E.V.) Dan. iii. 25 - Blessid art thou, Lord God of our fadris, and heryful [a1425 L.V. worthi to be heried] or worthi to be preyside”. Here heryful translates laudabilis, and the redundant periphrasis worthi to be preyside shows the attempt to express the passive meaning of the Latin word. Moreover, in the entry for the suffix -ful the OED (²1989) states: “[…] in modern English adjs. in -ful are sometimes <?page no="173"?> Romance Suffixation in the History of English 155 and knew the Latin derivational pattern as much as he was familar with the equivalent French -able pattern. Moreover I would hold that the simultaneity of the -able suffixation on indigenous bases and on loans shows that there was no such thing as a dormant word formation pattern that had to be recovered by being confronted with a mass of loans demonstrating this pattern. And, thirdly, the following observation by Marchand itself speaks against this putative course of events: English derivatives [with -able] occur from the 14 th century on, almost simultaneously with the afflux of loans. From the very beginning, the suffix has been tacked on to French as well as native roots. (Marchand 1969: 230; italics added). In sum I would thus argue that it was the familiarity with the pattern as such - rather than with a number of examples following this pattern - that accounts for its increasing productivity on any suitable basis. 9 Although composed almost a century later than the initial period of productivity of -able, it is worthwhile observing the ongoing productivity of -able within a specific corpus, here the Cathlicon Anglicum, to which we will turn next. As this is a dictionary we must expect that its compiler has a morphological awareness that exceeds that of the ‘normal’ speaker/ writer. This would raise the question of how representative the evidence in such a work is. On the other hand one could also argue that in the context of the dictionary we may observe productivity at work under lab conditions, as it were. 3.2 Adjectives in -able in the Catholicon Anglicum As one of the first English-Latin dictionaries (Stein 1985: 107) the Cathlicon Anglicum (Cath. Angl.) is specifically situated in the trilingual communicative space we envisaged earlier. The Cath. Angl. exists in two manuscripts: the oldest is BL/ BM MS Addit. 15562, dated “? c1475” by the editors of the Linguistic Atlas of Late Middle English (LALME), and the younger one is Lord Monson MS 168, dated 1483. As to the latter MS we have to rely on the EETS edition from 1881 by Herrtage and Wheatley (now easily accessible in an electronic reproduction) as the original MS is filed as “wherebouts unknown” (LALME). Only one other English-Latin dictionary, the Promptorium Parvulum (s. Stein 1985: 91-106), precedes the Cath. Angl., which, however, the formed directly on verb-stems, the sense of the suffix being ‘apt to’, ‘able or accustomed to’, as in assistful, distractful, crossful, mournful […].” While in vnbileueful the base bileue may be interpreted as a noun, the base of heryful can only be verbal. 9 In his Grammar (from c1000) Ælfric renders Latin deverbal -bilis adjectives as ‘present participle of vb. + lic.’ Thus he has lufiendlic for Latin amabilis (Zupitza 1880: 55) and asecgendlic for Latin affabilis (Zupitza 1880: 135). This is paralleled by his Old English pattern ‘sb. + lic’ for Latin adjectives in -lis; cf. cildlic for puerilis (Zupitza 1880: 54) or feondlic for hostilis (Zupitza 1880: 55). <?page no="174"?> Ursula Schaefer 156 compiler seems not to have known. We therefore readily see that its compiler is inspired by Latin-English dictionaries (Stein 1985: 111) whose authors, in turn, had to find suitable solutions in the target language English. This becomes particularly visible with its -able adjectives because a considerable number of these are given as the first written evidence of their existence in the OED. The comfortable access to the electronic online version of the OED with its various recent updates allows us to generate lists that show the extent to which a specific text is the source for “first entries”. As we can see with the help of this tool of the OED, 26 of the overall 313 first entries from the Cath. Angl. are deverbal adjectives with -able, six of them additionally prefixed with un-. 10 Five of the 26 have a Romance base (set below in italics), two of which have the prefix un-. When cross-checking with the OED we see that 17 entries, that is, more than two thirds of these ‘first entries’, are hapax legomena, at least with the sense in which they are given in the Cath. Angl., which is mainly that of the active meaning ‘able to do’ (cf. Marchand 1969: 231). 11 Here is a list of these -able adjectives (with the lemma in the Modern English of the respective OED entry; the dagger indicates a hapax legomenon for the sense given in the Cath. Angl.): 1. bitable † 10. sendable † 19. unacceptable 2. bowable ubi pliabylle 11. shippable † 20. UNCLEANSABLE † 3. buyable † 12. speakable 21. unhearable † 4. CLEANSABLE † 13. spendible 22. unpraisable † 5. eatable 14. strainable 23. UNTEACHABLE 6. fillable † 15. suppable [1483 Suppabylle] † 24. unthirlable † 7. joinable † 16. TEACHABLE 25. weavable † 8. playable 17. tellable -vbi spekabylle † 26. yokeable † 9. seekable † 18. turnable [overcomable 12 ] At first sight this list is puzzling because we may wonder, e.g., what kind of person would be in need of finding the Latin equivalent jnpen[e]trabilis of the English adjective unthirable (Cath. Angl, p. 383). However, if we look 10 Stein (1985: 117) notes that “the listing together of morphological antonyms” is a characteristic of the Cath. Angl. 11 I consider those forms in the Cath. Angl. hapax legomena that are either not continued, only ‘reappear’ in an early 17 th century French-English dictionary or have the next entry after a gap of more than 250 years. 12 This is an additional form not mentioned in the OED. In this case the Cath. Angl. (p. 263) lists the verb to Ouer com (Lat. confundere, fundere […]) under ‘O’. The form is mentioned in Miller (1997: 246); cf. also Miller (2006: 227). <?page no="175"?> Romance Suffixation in the History of English 157 into the Cath. Angl. itself, it soon becomes clear that most of these -able adjectives are part of “word families” (Stein 2004: 112) clustering the verbal base and its derived forms. 13 Thus the entry Biteabylle (‘bitable’, Lat. morsalis ; Cath. Angl. p. 33) is preceded by the verb entry to Bite (Lat. mo[r]dere 14 […]) and followed by Bytynge (Lat. mordens, mordax; ibid.). Similarly Byabylle (‘buyable’, Lat. empticius; Cath. Angl. p. 81) is preceded by the verb to By (Lat. emo), here, however, the compiler next gives the collocation to By and selle (Lat. auccionari, mercari […]) (ibid.). And to name a third formation with a Germanic base: before Playabylle (‘playable’, Lat. ludibundus, ludicris […]; Cath. Angl. p. 282) the noun A Play (Lat. locus, loculus, ludus […]) and the verb to Play (Lat. iocari, ioculari, ludere […]) are listed, and both the deverbal noun a Player (Lat. iocista, lusor) and the compound a Playnge place (Lat. diludium) follow. Looking into the lemmatic context also helps to assess the un-prefixed forms with a Germanic basis which first appear in the Cath. Angl. Thus the form vn-Hereabylle (Lat. in Audibilis; Cath. Angl. p. 184) follows the positive form Hereabylle (Lat. Audibilis), which is separated from the preceding verb entry to Here (Lat. Audire, Accipere, Attendere […]) by the locative adverb Here (Lat. jstic, hie). The entries Clennessabylle (Lat. expiabililis […]) and vn Clenceabylle (Lat. jnexpiabilis […]) are, in turn, couched in the following family environment (Cath. Angl. p. 66): Clene (Lat. jntemieratus [! ]) vn Clene (Lat. jnexpiabilis […]) Clene rynynge (Lat. eliquus […]) a Clennes (Lat. honestas […]) vn Clennes (Lat. jmmundicia […]) Clennessabylle (Lat. expiabilis […]) vn Clenceabylle (Lat. jnexpiabilis […]) to Clense (Lat. acerare […]) A Clensynge (Lat. colacio […]) Clensynge (Lat. colans […]) The fact that Clennessabylle immediately follows the (derived) nouns a Clennes (‘cleanness’; Lat. honestas […]) and vn Clennes (Lat. jmmundicia […]) suggests that the compiler seems to relate the positive -able adjective to the noun. However, the negative vn Clenceabylle (Lat. jnexpiabilis […]) 13 It has to be emphasized that there are many more - prefixed and unprefixed - -able derivates in the Cath. Angl. which are not dealt with here because they are not ‘first evidences’. 14 My emendation. <?page no="176"?> Ursula Schaefer 158 sets this right again as the spelling links it to the next verbal entry to Clense (Lat. acerare […]). Two other instances where the adjective precedes the verb from which it is derived are found with Sendabylle (Lat. missilis) and to Sende (Lat. mandare, commendare […]; Cath. Angl. p. 329) and with Techeabylle (Lat. docibilis, qui faciliter docet alios, docilis, qui faciliter docetur), vn Techeabylle (Lat. Jndocibilis A.) and to Teche; catezizare (caterizare A.), docere, inbuere, jnformare […]; Cath. Angl. p. 378). 15 These examples should sufficiently illustrate that the compiler was acutely aware of all kinds of available word-formation patterns. Those that particularly interest us here may be illustrated systematically with the example (un-)hearable: verbal base adjectival suffixation prefixed negation of deverbal adjective Engl. here here-abylle vn-hereabylle Lat. aud-iaudi-bilis in-audibilis He could also apply these patterns to etymologically French material: This should, again, make us hesitate to follow Marchand’s idea that, when it comes to word formation, there is a clear distinction between borrowed and indigenous material. The last two examples enforce this, and so does the case of bowable which is also first attested in Cath. Angl. As I have marked in the list, the Cath. Angl.’s entry Bowabylle has a cross-reference to pliabylle. Again, the derived adjective on the Germanic base bow follows the verb entry to Bowe (Lat. flectere, de-, plectere, humiliare […]; Cath. Angl. p. 38). The adjective entry Pliabylle (Lat. flexuosus, flexibilis […]; Cath. Angl. p. 284), in turn, is preceded by the verb entry to Plye (Lat. flectere, & 15 “A.” indicates that the form is solely found in the earlier MS. Addit. 15562. 16 The first Latin entry for praise in the Cath. Angl. is commendare; laudare is only the eleventh interpretamentum, for praisabylle the translation laudabilis is second and commendabilis first (Cath. Angl. p. 290). As illaudabilis is the only translation for vnpraisabylle I chose the base laudto illustrate the pattern. verbal base adjectival suffixation prefixed negation of deverbal adjective < Fr. iunge iune-abylle --- Lat. adiung-ijungi-bilis --- < Fr. praise prais-abylle vn-praisabylle Lat. 16 laud-alauda-bilis vn-laudabilis <?page no="177"?> Romance Suffixation in the History of English 159 cetera) with the cross-reference “vbi to bowe”. The OED entry for pliable (³2006, September) gives the first entry from a text from 1392 in a manuscript dated c1475. From the following entries we must infer that this adjective was rare before the 17 th century and had a technical ring to it. For the verb ply the OED gives quotes from Gower’s Confessio Amantis and Chaucer’s Merchant’s Tale, but after this the verb, too, seems rare. In any event, for the Cath. Angl. compiler bow and ply were on par and the -able suffixation could work ‘simultaneously’, as it were, with both the indigenous and the French synonyms. In closing this section a remark is in order about a cross-refered entry which is quite telling for the whole problem addressed here: the adjective speakable. As the OED says, it is first noted in the 1483 version of the Cath. Angl. in the sense “1. That may or can be spoken; fit or possible to be expressed in speech” (Spekabylle (Lat. effabilis […]; Cath. Angl. p. 358)). Its negative form, which is also given in the 1483 speak-‘family’ after the verb (vn Spekabylle (Lat. Jneffabilis […]; Cath. Angl. p. 358)), is attested in the OED sense “1.a. Incapable of being expressed in words; inexpressible, indescribable, ineffable” already in 1400. As a translation of Latin ineffabilis it is one of those cases where the negative form appears in (written) English earlier than the positive. The second thing to remark about speakable in the Cath. Angl. are, again, two cross-references. For one we have in the 1474 version (and only there) Tellabylle (Lat. vbi spekabylle) and vn Tellabylle (Lat. inenarrabilis, Ineffabilis; Cath. Angl. p. 379) preceding the verb by five and four positions respectively. The second cross reference can be found in the Cath. Angl. with the construction Abylle to speke, which runs “ubi spekeable” (Cath. Angl. p. 2). The form Abylle is given above (before an Abydynge and Abylite) with Latin “hie hec Abilis & hoc le, […]” as translation (Cath. Angl. p. 2). This strongly supports Marchand’s speculation that the immediate prolificness of French -able on native bases might have been fostered by the association of that suffix “with the (unrelated) word able”, and he continues that “its early use with native stems is otherwise not easily accounted for” (1969: 230). Unfortunately Marchand does not disclose why the recourse to the lexeme able would make it easier to explain the fast nativization of the suffix, as the earliest evidence for able in the OED shows it in predicative use with the sense “1.a Capable of, having sufficient strength or power for; ready for, fit to cope with”. If I understand Marchand correctly he surmises that this is attributable to the semantic congruence of the lexeme and the suffix in its active meaning (cf. OED ³2009, September): “Forming adjectives denoting the capacity for […] performing the action <?page no="178"?> Ursula Schaefer 160 denoted or implied by the first element of the compound”; 17 s.v. -able). This proximity is, on the one hand, also indicated by abylle to speke and the cross reference to spekeable, but also by the next entry Abylle to yoke with the reference “vbi to yoke”. The yoke-‘family’ does indeed have 3 okabylle with the Latin equivalent iugalis (Cath. Angl. p. 427). Latin iugalis, in its turn, is a denominal adjective. Be this as it may (which is to say: much more detailed research would have to be done to answer this question): the Cath. Angl.’s compiler strongly suggests the putative relation between predicative able and the suffix -able. 18 Although as historical linguists we are trained not to pay too much attention to spelling, it is, however, remarkable, that in the 1475 and the 1483 version of the Cath. Angl. the suffix is spelt quite regularly as abyll(e), which would comply with the respective spellings in the constructions abylle to speke and abylle to yoke. Yet it is also striking that the crossreference for abylle to speke spells the adjective as spekeable. Again it would take a much closer look at the Cath. Angl. to decide whether the -abylle/ -able spelling mirrors a semantic difference which could, for instance, run along the active/ passive line. 19 4 Conclusions The initial concept for my contribution in honor of my dear friend and colleague Andreas Fischer had been to wrap up my theoretical answers to a number of questions which he raised in his 2001 article on “Lexical Borrowing and the History of English”, put a virtual bow around them to make this a birthday present. I will try a brief summary in order to see whether I have arrived at offering anything substantial. I have taken up the heuristic doubts and the resulting methodological postulates which Andreas Fischer phrased in his 2001 article. In doing so I tried to show why a socio-linguistic approach in the wake of (pseudo-? ) Labovian categories does not really take us to where we want to go when it comes to the question of lexical borrowing in late medieval England. Instead I have suggested looking into - implicit - linguistic norms along the lines of discourse traditions which, for the period under consideration, compose the multilingual scenario. Any lexical - and to my mind this also 17 The notion compound is misleading here, as in word formation this is restricted to composits of two free lexemes. 18 Cf. also, e.g., the following OED evidence for prayable: “a1382 Bible (Wycliffite, E.V.) (Bodl. 959) Psalms lxxxix. 13 - Preyable be þou [a1425 L.V. be thou able to be preied; L. deprecabilis esto] vp on þi seruauntis.” ( 3 2007, March). 19 The Cath. Angl. compiler is aware of this line as he renders, e.g., Engl. techeable as both “docibilis, qui faciliter docet alios” and “docilis, qui faciliter docetur” (Cath. Angl. p. 378). <?page no="179"?> Romance Suffixation in the History of English 161 includes affixational - borrowing needs to be considered as being geared by the French (and Latin) text models available. These were the models/ styles in which the authors and scribes were initially trained and which must have dominated their attempts to express themselves in English. Subsequently I have restricted my concrete data analysis to wordformation that involves the suffixation with the element -able, derived from its Latin (and French) word-formation pattern. In doing so I looked into a specific witness of the fourth quarter of the 15 th century that has ‘first evidence’ for -able suffixations with both borrowed and native bases. It should - by now - come as no surprise that once we look into the ‘real’ linguistic data, things turn out to look much less orderly than our initial heuristic approach would want them to be. Nevertheless I hope to have illustrated with my data analysis that Marchand’s structuralist theory as to the productivity of borrowed affixes has to be imbedded in the Latino-French context. To put it briefly: there were no ‘speakers’ of English who needed a heavy influx of borrowed adjectives in -able in order subsequenly to abstract the ‘structural’ word formation pattern. By way of explicit induction or implicit deduction they already formally knew this pattern from Latin and, in all likelihood, via Latin, probably for French. In sum I have tried to substantiate Andreas Fischer’s postulate that we should pay more attention to “borrowing from written sources” (2001: 109), and I hope to have provided a few heuristic ideas how this could be put into analytical practice. While my heuristic suggestions are, hopefully, well-founded, the results provided by the selective data analyses clearly indicate that much more research needs to be done to solve the questions addressed here. However, it is the unmatchable privilege of our philological profession that we may return to seeking solutions for problems even after we have retired from active teaching - or, in some cases, from preoccupying ourselves with administrative academic affair. In this sense, dear Andy: many happy returns! <?page no="180"?> Ursula Schaefer 162 Biographical sketch Ursula Schaefer is currently the Vice-Rector for Academic and International Affairs at the Technische Universität Dresden and therefore on leave from her position as Professor for English Linguistics. However her research interests are still focused on medieval English literature, culture and language, on orality and literacy and on the standardization of the English language in late medieval and early modern times. She is author of the monograph Vokalität: Altenglische Dichtung zwischen Mündlichkeit und Schriftlichkeit (1992) and has published a number of articles within the wide field of her resarch interests. Moreover she has edited in the recent past a volume on late Middle English standardization and on the history of the philologies. References Burnley, John David. 1986. “Curial Prose in England”. Speculum 61, pp. 593- 614. Burnley, John David. 1989. “Late Medieval English Translation: Types and Reflections”. In Roger Ellis et al. (eds.). The Medieval Translator. The Theory and Practice of Translation in the Middle Ages. Papers Read at a Conference Held 20 - 23 August 1987 at the University of Wales Conference Centre, Gregynog Hall. Cambridge: D. S. Brewer 1989, pp. 37-53. Burnley, John David. 2001. “French and Frenches in Fourteenth-Century London.” In Kastovsky and Mettinger (eds.), pp. 17-34. Camargo, Martin (ed.). 1995. Medieval Rhetorics of Prose Composition. Five English Artes dictandi and their Tradition. Binghampton, NY: MRTS. Catholicon Anglicum (Cath. Angl.) = Catholicon Anglicum: An English-Latin Wordbook, Dated 1483 (edited by Sidney J. H. Herrtage and H. B. Wheatley). EETS 75. London: N. Trübner 1881 [electronic version accessible under http: / / www.archive.org/ details/ catholiconanno7500herruoft]. Clanchy, Michael T. 1993. From Memory to Written Record: England 1066-1307. (Second edition, first edition 1979). Oxford: Blackwell Publishers. Dalton-Puffer, Christine. 1996. The French Influence on Middle English Morphology. A Corpus-based Study of Derivation. Berlin and New York: Mouton de Gruyter. Davis, Norman. 1954. “The Language of the Pastons”. Proceedings of the British Academy 40, pp. 119-144. Fischer, Andreas. 2001. “Lexical Borrowing and the History of English. A Typology of Typologies”. In Kastovsky and Mettinger (eds.), pp. 97-115. Kastovsky, Dieter and Arthur Mettinger (eds.). 2001. Language Contact in the History of English. Frankfurt am Main: Peter Lang. <?page no="181"?> Romance Suffixation in the History of English 163 Koch, Peter. 1988. “Norm und Sprache”. In Jörn Albrecht et al. (eds.). Energeia und Ergon. Sprachliche Variations − Sprachgeschichte − Sprachtypologie: Studia in honorem Eugenio Coseriu. 3 vols.; vol. II: Harald Thun (ed.). Das sprachtheoretische Denken Eugenio Coserius in der Diskussion (1). Tübingen: Gunter Narr, pp. 327-354. Koch, Peter. 1997. “Diskurstraditionen: zu ihrem sprachtheoretischen Status und ihrer Dynamik”. In Barbara Frank, Thomas Haye and Doris Tophinke (eds.). Gattungen mittelalterlicher Schriftlichkeit. Tübingen: Gunter Narr, pp. 43-79. Labov, William. 1994. Principles of Linguistic Change. Vol. I: Internal Factors. Oxford: Blackwell. LALME = A Linguistic Atlas of Late Mediaeval English. 1986. Edited by Angus McIntosh, Michael L. Samuels and Michael Benskin. 4 vols. Aberdeen: Aberdeen University Press. [References to LALME for the Catholicon Anglicum have been retrieved here electronically from the Middle English Compendium Hyper- Bibliography at http: / / quod.lib.umich.edu/ cgi/ m/ mec/ hyp-idx? type=byte&size =First+100&byte=373596&title=Catholicon%20Anglicum]. Marchand, Hans. 1969. Categories and Types of Present-Day English Wordformation. A Synchronic-Diachronic Approach. (Second, completely revised and enlarged edition; first edition 1960). München: C. H. Beck’sche Verlagsbuchhandlung. Matras, Yaron. 2000. “How Predictable is Contact-induced Change in Grammar? ” In Colin Renfrew, April McMahon and Larry Trask (eds.). Time Depth in Historical Linguistics. Vol. 2. Oxford: MacDonald Institute for Archaeological Research, pp. 563-83. Miller, D. Gary. 1997. “The Morphological Legacy of French Borrowed Suffixes on Native Bases in Middle English”. Diachronica 14, pp. 233-264. Miller, D. Gary. 2006. Latin Suffixal Derivatives in English and Their Indo- European Ancestry. Oxford: Oxford University Press. Nevalainen, Terttu. 2004. “Letter Writing. Introduction”. Journal of Historical Pragmatics 5, pp. 181-191. Nurmi, Arja and Päivi Pahta. 2004. “Social Stratification and Patterns of Code- Switching in Early English Letters”. Multilingua 23, pp. 417-456. Oesterreicher, Wulf. 1993. “Verschriftung und Verschriftlichung im Kontext medialer und konzeptioneller Schriftichkeit”. In Ursula Schaefer (ed.). Schriftlichkeit im frühen Mittelalter. Tübingen: Gunter Narr, pp. 267-292. Oesterreicher, Wulf. 2001. “Historizität - Sprachvariation, Sprachverschiedenheit, Sprachwandel”. In Martin Haspelmath et al. (eds.). Language Typology and Language Universals / Sprachtypologie und sprachliche Universalien / La typologie des langues et les universaux linguistiques - An International Handbook / Ein internationales Handbuch / Manuel international, vol. 2. Berlin and New York: Mouton de Gruyter, pp. 1554-1595. Rothwell, William. 2000. “Aspects of Lexical and Morphosyntactical Mixing in the Languages of Medieval England”. In David A. Trotter (ed.) Multilingualism in Later Medieval Britain. Cambridge: D.S. Brewer, pp. 213-232. <?page no="182"?> Ursula Schaefer 164 Schaefer, Ursula. 2005. “Emergente Macht. Die spätmittelalterliche Re- Institutionalisierung der Volkssprache in der Prosa”. In Gert Melville (ed.). Das Sichtbare und das Unsichtbare der Macht. Institutionelle Prozesse in Antike, Mittelalter und Neuzeit. Köln: Böhlau, pp. 335-354. Schaefer, Ursula. in print. “Travelling the Paths of Discourse Traditions: A Sample Analysis of the Lexical Innovation blisfulnesse in Chaucer’s Boece”. In Viktor Becher et al. (eds.). Multilingual Discourse Production. Amsterdam / Philadelphia: John Benjamins. Stein, Gabriele. 1985. The English Dictionary before Cawdrey. Tübingen: Niemeyer. Stein, Gabriele. 2004. “The Catholicon Anglicum (1583): A Reconsideration”. Nordic Journal of English Studies 3, pp. 109-124. Taavitsainen, Irma. 2001. “Language History and the Scientific Register”. In Hans-Jürgen Diller and Manfred Görlach (eds.). Towards a History of English as a History of Genres. Heidelberg: Universitätsverlag C. Winter, pp. 185-202. Thomason, Sarah G. 2001. Language Contact. An Introduction. Washington D.C.: Georgetown University Press. van Coetsem, Frans. 1988. Loan Phonology and the Two Transfer Types in Language Contact. Dordrecht: Foris. Weber, Beatrix. 2010. Sprachlicher Ausbau. Konzeptionelle Studien zur spätmittelalterlichen Schriftsprache. Frankfurt am Main: Peter Lang. Weinreich, Uriel. 1966. Languages in Contact. Findings and Problems. (Fourth printing with a preface by André Martinet, first printing 1953). Den Haag: Mouton & Co. Winford, Donald. 2003. An Introduction to Contact Linguistics. Maldon, CN/ Oxford: Blackwell Publishing. Winford, Donald. 2005. “Contact-induced Changes: Classification and Processes”. Diachronica 22, pp. 373-427. Wright, Laura. 1994. “On the Writing of the History of Standard English”. In Francisco Fernández, Miguel Fuster and Juan José Calvo (eds.). English Historical Linguistics 1992. Papers from the 7 th International Conference on English Historical Linguistics. Valencia, 22-26 September 1992. Amsterdam and Philadelphia: John Benjamins, pp. 105-115. Zupitza, Julius (ed.). 1880. Ælfrics Grammatik und Glossar. Berlin: Weidmannsche Buchhandlung. <?page no="183"?> Daniel Schreier Words and the New Englishes Borrowing as Evidence of Contact and Ancestral Effects Abstract In this paper, I argue that the analysis of lexical borrowing processes contributes substantially to our understanding how varieties of English form around the world. In particular, such mechanisms are essential to determine what varieties come into contact, (and also in what proportions of speaker groups) and which of them was most influential in a stratified community. Moreover, borrowing yields important information on the degree of interaction and quite possibly also of social stratification between ethnic groups in contact and sociocultural pressures of assimilation. Based on examples from New Zealand English, Fiji English, African American English and Tristan da Cunha English, the paper describes some dynamic processes of contact-induced borrowing and outlines their relevance for the disciplines of sociolinguistics and dialectology, arguing that the study of words can contribute to the investigation of ancestral effects and contact-induced developments in post-colonial English. 1 Introduction Not long ago, the study of lexis took a prominent role in historical linguistic research. In order to understand how languages change (very often as a result of direct contact with other languages), philologists and historical linguists were interested in lexical borrowing processes or the regional usage of words at large. For a long time lexical investigations, along with an array of related changes (blending, calquing, semantic shift), were at the forefront in the study of dialect obsolescence, spatial variation, contactinduced change, innovation-diffusion and language change in general. Until roughly the 1960s, many, if not most, researchers gave special attention to words, though phonology, morphosyntax and pragmatics were not neglected. The history of English illustrates the lexical foundations of contactinduced change in an exemplary fashion. Starting with the arrival of the Anglo-Saxons and the evolutionary inception phase of Old English, the language underwent several high-contact phases, as a result of which words were adopted from multiple sources. The initial contact scenario, for <?page no="184"?> Daniel Schreier 166 instance, saw an embryonic stage involving coexistence and competition of Saxon, Franconian, Jutish, Frankish, Hessian, Frisian and other varieties, speakers of which were in regular contact. Old English was in fact a hybrid of various Germanic inputs (Crystal 2003); in the words of Baugh and Cable (1993: 50), “the English language has resulted from the fusion of the dialects of the Germanic tribes who came to England”, and lexis, as stated above, received much interest from historical linguists (Strang 1970, Stockwell & Minkova 2001). Borrowings from the Celtic substrate (mostly toponyms) remain subject to debate to the present day (Vennemann forthcoming). Then, from the 7 th century onwards, extensive contact with Old Norse saw a different (and rather untypical) contact scenario; this contact setting gave rise to borrowing of a smaller number of items in domains of everyday usage. Frequent words such as sky, skirt or egg, even pronouns such as them and their, were borrowed from the Scandinavians (Stockwell & Minkova 2001). The rather unusual borrowing process (small number of lexical items in intimate domains) has been analysed in detail and become central in typological approaches to lexical change (Fischer 2001, Thomason & Kaufman 1988). Contact with Norman French from the 11 th century onwards provides an altogether different scenario. It is estimated that up to 10,000 words entered the word stock of English in the three centuries following the 1066 Norman invasion (Jespersen 1938, Dekeyser 1986, Stockwell & Minkova 2001) and the timing of borrowing provides important insights into the emergence and disappearance of bilingualism and its sociolinguistic effects on the host community. Jespersen (1938: 87), for instance, looked into the absolute frequencies of French loanwords (based on the Oxford English Dictionary) and found that they peaked in the second half of the 14 th century, when there were hardly any French monolinguals left and when the number of French-English bilingual speakers was on the decline. This, and other studies, led Thomason & Kaufman (1988) to formulate some detailed principles as to how the emergence and decline of societal bilingualism affects the rate of borrowings. Similarly, the domains for borrowing provided important insights too since most words come from superstrate domains such as the court, military, government, etc., all of which attest to the strength of social differences at the time. Finally, and this is the topic of the present paper, the diversification of English and the development of many so-called new varieties outside the British Isles saw a veritable explosion of its lexicon. Tens of thousands of words were adopted from languages all over the world, a process so extensive that English is now considered to have the largest lexicon of all languages (with an estimated 500,000-600,000 entries; Crystal 2003). This paper consequently discusses the ‘value’ of lexis for historical and contact linguistics in general, leaving aside why it has lost some of its im- <?page no="185"?> Words and the New Englishes 167 portance in recent years. Based on examples from varieties such as New Zealand, Fiji, African American and Tristan da Cunha English, this contribution describes some dynamic processes of contact-induced borrowing and outlines their relevance for the disciplines of sociolinguistics and dialectology. Above all, it aims at showing how a study of words can contribute to the investigation of ancestral effects and contact-induced developments in post-colonial English. 2 The sociolinguistic value of borrowing At the risk of simplification, one can generally distinguish between quantity and quality of borrowing. It is certainly relevant to distinguish in what domains words are adopted (home or learned/ erudite ones, for instance) as opposed to, quantitatively speaking, how many items enter the language via other sources. Notwithstanding, it is of central interest to be as accurate as possible and investigate precisely what features are borrowed (in what domain or from which category, free or bound morphemes, e.g. bound suffixes, etc.). While content words and free morphemes are most readily borrowed, other elements may enter other languages as well (such as closed-class function words: determiners, intensifiers, even bound morphemes, see below), a process known as structural borrowing. As is well known, any feature can be borrowed, depending on the circumstances (Comrie 1981 lists Russian verb inflections in Copper Island Aleut; Altaic case suffixes in Xinjiang Chinese; or the borrowing of personal pronouns from Old Norse into Old English, e.g. they, them, their, as rare borrowings), and contact intensity is considered to play an important role; following Thomason & Kaufman (1988), lexical borrowing typically occurs when there is low intensity contact between speech communities; structural borrowing, on the other hand, requires heavy contact and is particularly common in communities with sustained bior multilingualism leading to structural-systemic effects of contact-induced language change. Often, the source of borrowings is a reflection of the social organization and/ or stratification of a community; in pidgin and creole linguistics, for instance, the lexifier spoken by the socially influential (often colonial) groups provides most of the vocabulary, especially content words. 1 Following Thomason & Kaufman, these processes are to a large part socioculturally determined: “With a minimum of cultural pressure we 1 One should of course bear in mind that social pressures are only one facet and that there exist other reasons for borrowing: lexical gaps (borrowing of American Indian words by the first settlers in America: tepee, tomahawk, raccoon, etc.; this process is arguably found in every colonial scenario; see Schneider 2007), superstratal influence (up to 10,000 words were adopted via Norman and later Parisian French; Dekeyser 1986), or simple prestige and Zeitgeist (English terms in Swiss German). <?page no="186"?> Daniel Schreier 168 expect only lexical borrowing, and then only nonbasic vocabulary” (1988: 77). They propose a typological scale of borrowing, distinguishing between (1) casual contact; (2) slightly more intensive contact; (3) more intense cotact; (4) strong cultural pressure; (5) very strong cultural pressure, all of which vary in intensity and influence the rate and direction of borrowing. Languages such as English borrow from many different languages due to casual contact, which gives rise to lexical borrowing only (content words) and is motivated by cultural and functional reasons (gaps) and involves non-basic vocabulary. As contact intensifies, we find increased lexical and some slight structural borrowing: Thomason & Kaufman list function and content words (especially conjunctions, adverbial particles) and structural elements (minor phonological, syntactic, lexical semantic features, new phonemes in borrowed words, new syntactic functions). A scenario of more intense contact gives rise to extended structural borrowing of function words (e.g. pre-, postand adpositions), derivational affixes adopted from loan words (-ity, -ation), inflectional affixes adopted via loan words, personal/ demonstrative pronouns and low numerals (basic vocabulary). Phonological borrowing may involve prosodic features, stress rules, phonemicisation, even in native vocabulary, etc. Strong cultural pressures involve typological changes (new distinctive features, new syllabic constraints and allophonic rules, fairly extensive word order change (often to SVO) and borrowed inflectional affixes/ categories). It is only in very strong cultural pressure that one encounters major structural change and typological disruption (phonetic changes, loss of phonemic contrasts or morphophonemic rules, changes in word structure rules, categorical, ordering changes in morphosyntax, etc.). This view is now widely held and goes back to Weinreich (1953), who was one of the first to argue that sociolinguistic and cultural values are likely to promote or discourage borrowing/ interference. The focus on structural borrowing, generally assumed to have higher systemic significance, certainly helps to explain why examining lexical borrowing has fallen out of favour, and I would argue that the development of pidgin and creole studies as a field in its own right certainly played a role here. Similarly, contact linguistics has emancipated from historical linguistics and philology over the last 30 years and researchers tend to focus almost exclusively on grammar and phonology. However, even if there is reason to assume that structural changes entail systemic modification and long-term language change, it would be mistaken to downgrade the importance of lexical borrowing in general, as argued below. The English language in particular, with its global distribution and a myriad of local contact scenarios, goes some way toward furthering the understanding how lexical aspects vary (both in quality and quantity) in different contact settings, as a result of which lexis should not be underestimated in typological assess- <?page no="187"?> Words and the New Englishes 169 ments of contact histories and ancestral effects of varieties. I will use examples from four varieties to illustrate this. 3 Four different borrowing scenarios The history of African American English (AAE) has been amply discussed in the literature, so extensively in fact that it has become by far the mostwidely researched of all varieties of American English (Schneider 1996). The origins and ancestry of AAE have received much attention, and several theories have been put forward to account for its evolution. The so-called Anglicist position maintains that earlier, post-colonial African American speech was quite similar to the early British dialects brought to North America: By and large the Southern Negro speaks the language of the white man of his locality or area and of his education […] as far as the speech of uneducated Negroes is concerned, it differs little from that of the illiterate white: that is, it exhibits the same regional and local variations as that of the simple white folk. (Kurath 1949: 6) This would mean 1) that there was near-accommodation to targets (lexical and structural) in earlier American English, and 2) that AAE was more or less directly derived from British dialects and that present-day differences developed in the 19 th and 20 th centuries only. Others adopted a creolist position, maintaining that the Negro slaves […], even many who were born in the New World spoke a variety of English which was in fact a true creole language - differing markedly in grammatical structure from those English dialects which were brought directly from Great Britain, as well as from New World modifications of these in the mouths of descendants of the original white colonists. (Stewart 1967: 3) As a consequence, there would have been massive social, cultural and sociolinguistic differentiation between the two ethnic groups historically, the African descendants forming proto-AAE via processes of language contact and creolisation. What evidence was offered to substantiate these claims? Lexis was prominent in early research. George Philip Krapp, for instance, in his The English Language in America, took the position that American words [that] were brought into the language through the Negroes have been insignificant in number […] A few words like juba, a kind of dance, banjo, hoodoo, voodoo, pickaninny, exhaust the list of words of non-English origin […] very little of it, perhaps none, is derived from sources other than English. (Krapp 1925: 161) <?page no="188"?> Daniel Schreier 170 Krapp went so far as to insist that jazz was an old English dialect word and claimed that moke, formerly common to denote a black person, derived from Icelandic möckvi ‘darkness’ (and that it may have been a borrowing from Old Norse). Lorenzo Dow Turner, on the other hand, one of the pioneers of the historical study of AAE, argued in his Africanisms in the Gullah Dialect (1949) that some syntactic features of Gullah were ipso facto traceable to West African languages and that they appear in parts of the ‘New World’ that predate the arrival of British colonialists (mentioning Brazil and Haiti in particular). His long-term study of Gullah led Turner to suggest that as many as 6,000 Gullah words were African in origin, among others: hep-cat / hip ‘a person who is attuned to his/ her environment, sharp and cool’ (from Wolof hipikat ‘has his eyes open’), chigger ‘small species of flea’, gumbo, banjo, jitter, cola, yam, zombie, juke, goober ‘peanut’, tote ‘carry’, okra, boogie-woogie, bogus, funky, phoney, perhaps even jazz. These words were either shared with or else carried over into AAE. Looking at these borrowings in detail, one cannot help but notice that, though these borrowings cover a range of domains (topography, household, etc.), the vast majority of them are lexical and not structural. Hence, the significance of the quantity of borrowed items, though immense (if Turner’s estimate is correct, that is), would have to be reassessed somewhat by the observation that they can be mostly classified as due to casual or slightly more intensive contact by Thomason & Kaufman (1988). In any case, only few instances of structural borrowing are attested. Lexis has received attention in contemporary AAE as well, particularly in the case of the convergence-divergence hypothesis. Teresa Labov (1992) studied lexical variation in American adolescents, with particular focus on whether (and if so, to what extent) they understood and used (via borrowing) a specified set of terms. Quite strikingly, race was the strongest of all the social variables attested; Labov found that African American adolescents were much more familiar with AAE terms such as bougie (‘an uppityacting African American’), busting out (‘looking good’), fresh (‘cool’), and kitchen (‘hair at the nape of the neck’). Whites, on the other hand, were more familiar with schlep (‘drag along’) and bombed or smashed (’drunk’). Many words regarded as typical of AAE were not slang but commonly used by all age and most social groups. In other words, Labov found there was a stock of words that, no matter whether they derive from African languages or not, was restricted in usage to African Americans and practically unknown to European Americans (and vice versa). This is countered by a very strong trend for originally black terms to be transferred over into mainstream American English (as a result of borrowing via the media). Lee (1999) reported the following terms, all found in her local newspaper: chill out, threads, all that, boom-shaka-laka, main squeeze, you go, girl, high-five, homeboy, soulmate, got game, or the exclamation right on! (discussion in <?page no="189"?> Words and the New Englishes 171 Green 2002). Other candidates include you guys, awesome, bro, etc. Consequently, though borrowing occurred, the usage of lexis continues to correlate along ethnic boundaries in the US, so that a number of words are found in one particular ethnic group only: “One of the many fascinating features of black vocabulary is how sharply it can divide between blacks and whites, and how solidly it can connect blacks from different social classes” (Rickford & Rickford 2000: 93). Borrowings may thus give insights into intimate contacts between speech communities (e.g. the Old English/ Old Norse scenario) but they may also further our understanding how they remain apart and do not assimilate or integrate sociolinguistically. New Zealand English (NZE) provides an equally insightful case of the value of borrowing for contact linguistics. New Zealand English is one of the youngest varieties of post-colonial English (Gordon et al. 2004). Data analysed for the Origins of New Zealand (ONZE) project, the most extensive database of New Zealanders born in the 19 th century available, has made it possible to reconstruct the history and evolution of New Zealand English and to formulate/ test various theories of how it might have developed: split from Australian English, mixing theory, single vs. multiple origins, etc. Gordon and her associates focussed mainly on phonology and treated morphosyntax and lexis peripherally, in contrast to Bauer (1999, 2000), who argued that lexis could indeed contribute toward answering these questions. As for methodology, Bauer analysed all words labelled “dialectal” or “of British dialect origin” in Orsman’s Oxford New Zealand English Dictionary (1997). Upon identifying and classifying these words, he researched their areas of origins and distribution in dialect dictionaries and atlases of British English, listing them for each region. Bauer (1999: 295) found that their places of origins (and thus the donor dialects through which they were transported to New Zealand) were evenly distributed, with no marked regional preference and little localisation of donors. This even distribution of British dialect words was seen as evidence of mixing and as an important clue in pinpointing the varieties transported to New Zealand (and which ones eventually won out when New Zealand English formed). It is noteworthy (and also somewhat troublesome) that the historical analysis of NZE calls for separate conclusions, depending on whether one focuses on phonology or lexis. Whereas an impressionistic phonetic analysis would strongly point to the English Southeast as a donor area, an analysis of the lexicon suggests that there was a much larger pool of dialects that contributed to the emerging NZE. It is a great pity that the differing results were not seen as complementary and that the validity of both findings has been debated intensively; this is not the place to support either interpretation. Instead, it should simply be pointed out that structural and <?page no="190"?> Daniel Schreier 172 lexical borrowing may co-occur yet differ in their motivation, hence the old saying that “words travel lightly”. Furthermore, any theory offered to explain why new dialects are the way they are should attempt to integrate both, particularly since lexical borrowing, structural borrowing and phonological selection do not necessarily exclude one another. Fiji English provides an example of far-reaching, not to say intimate and creative, consequences of lexical borrowing with socio-cultural implications. “The Republic of the Fiji Islands is one of the most cosmopolitan and linguistically diverse Pacific island nations” (Tent 2001: 209) and three main varieties, Fijian, Fiji Hindi and English, are spoken side by side in the community. Fijian is predominant in the indigenous Fijian population (roughly 50% of the population), Fiji Hindi (a local, Hindi-derived koiné; Siegel 1986) is the language of the Indo-Fijians (44% of the population), and this includes both shudh (standard Hindi) and low-prestige Fiji Hindi (in a quasi-diglossic scenario). Moreover, a variety of languages all have their place in the community also: Pidgin Fijian, Pidgin Fiji Hindi, Cantonese, Mandarin, Rotuman, Samoan, Tongan, Kiribati (Gilbertese), Tuvaluan, Gujarati, Urdu, Punjabi, Tamil, Telugu and Malayalam. English, on the other hand, is a distinct nativised L2 variety spoken in differing competence by most of the population (but only 3% speak it natively). It was first introduced in the early 19 th century, the first missionaries arriving only in 1835 (Schneider 2007). The early formation phase of a local vernacular was characterised by restricted access to the English target. Fijian was the medium of communication in local administration, and English was not taught as a precautionary measure (in an Education Commission report from 1909, one reads that knowledge of English “would only tend to make them [the locals] despise manual labour”). The situation only changed in the 1930s when New Zealand education authorities took over, when newly appointed teachers were not keen on other languages and started to actively promote English as a lingua franca. From the administrative point of view, the perhaps somewhat romantic goal was that English should serve as a “neutral” medium and thus serve as a pathway to harmonious relationships between Fijians and Indo-Fijians. Today, however, Fijian and Hindi are still extensively used (Siegel 1973, 1989) and both carry high local prestige and function as identity-carriers in the communities, so the intention of a largescale adoption of English as a local lingua franca was not successful. Moreover, although English is the language of government, education, business, the judiciary and the media, it is not widely taught as a Second Language. If so, then Fijians learn it predominantly for instrumental (job opportunities, economic advancement, etc.) and not for integrative reasons. The local Fiji English variety, spoken along a lectal continuum (Siegel 1989), “is recognised as a unique and distinctive regional variety of Eng- <?page no="191"?> Words and the New Englishes 173 lish” (Tent 2001: 221) and lexis features prominently in its analysis. Tent (2001) looked at the vocabulary of Fiji English to demonstrate its ancestry and to also show contact and borrowing effects. With this aim, he conducted a thorough lexical analysis of Fiji English and found that despite the fact that there were loanwords from Cantonese, Tongan, Indian English, etc., a large number of borrowings was drawn from Fijian and Hindi (38% and 16%). Most of them were what he called “necessary loans” for native concepts, objects and customs, but there are numerous “unnecessary” ones as well. These can be exemplified as follows. Fijian loans (rather extensive) Fijian English has borrowings that are unnecessary from a functional perspective, since none of them fills lexical gaps: kana ‘food’, 2 ‘to eat’, kasou ‘drunk’, koro ‘village’, leqa ‘problem’, laloma ‘love’ etc. It is noteworthy that most of them are from the rather intimate domain of household. Hindi loans (less frequent) There are some cultural loans (vegetables and spices, etc.), and very few “unnecessary” loans (paidar ‘by/ on foot’, paisa ‘money’), which can be explained by the fact that Fijian has higher status and is of more importance in the community. It is possible, though not attested, that there exists regional variation and that there are more loans in sugar-cane growing areas, where Hindi speakers have lived traditionally. Of particular interest is the fact that Tent (2001) reports numerous hybrids, that is to say, compound items that derive from more than one language. free morphemes • bula smile ‘welcoming smile’ (Fijian + English) • nice bola! exclamation to congratulate on good looks, smart dress etc.; Fijian bola ‘bag, case, box’ • talanoa session ‘chat, informal talk, story-telling’ (Fijian talanoa ‘to chat’) • gang sirdar ‘cane-gang foreman’ (English + Hindi) • no ghar ‘homeless’ (Hindi ghar ‘house’) 2 Note, however, that khana (with aspirated / k/ ) is also ‘food, eat’ in Hindi (thanks to Claudia Rathore for pointing this out to me). <?page no="192"?> Daniel Schreier 174 bound morphemes: • bilibiliathon ‘annual race on traditional bamboo rafts’; Fijian bilibili + English -athon, marathon) • malosa ‘hopeless person with no prospects’ (Fijian ma, particle compounded with many words forming nouns of state or condition + English loser) There are even cases of what Tent (2001) calls re-borrowings. Words are borrowed from one language, nativised and adapted phonologically only to be re-borrowed by the original language. • Fijian English boso ‘chief, foreman’ < Fijian boso < English boss • Fijian English karasi ‘marijuana’ < Fijian karasi < English grass Multilingual and -cultural contexts with frequent face-to-face interaction, as on Fiji, entail and foster complex borrowing and, in contexts of proximity, familiarity and close intimacy, may also give rise to processes such as blending, calques and even re-borrowings. The fact that the sociolinguistic landscape of Fiji displays these mechanisms is interpreted as an important piece of evidence for language contact and the emergence of endemic features on Fiji. The final setting considered here is Tristan da Cunha, the world’s most isolated inhabited island (Schreier 2003). Tristan da Cunha English (TdCE) is one of the youngest post-colonial varieties of English and was formed in the 1820s. The island was settled in 1816 (by a South African garrison assisted by “Hottentots”) yet abandoned the year after. A Scottish corporal (with his South African wife and two children) as well as two stonemasons from Plymouth stayed behind, thus founding the present-day community. They were followed by sailors from various parts of England (Hull, London, Hastings), two Danish seamen, a group of women from St Helena, American whalers and a Dutchman. The social history is comparatively well-researched and has been used to retrace the development of the local variety (Zettersten 1969, Schreier 2003). Their research has shown that morphosyntactically, TdCE bears resemblances to Atlantic creoles, which is most convincingly explained by extensive contact with St Helenian English (StHE), as spoken by the women who arrived in 1827. Reduction of inflectional morphology as well as the characteristics of the TdCE tense and aspect systems offer strong evidence. The lexicon in particular indicates the amount of contact and mixing that must have occurred in the 1820s and 1830s, leading to borrowing from multiple sources. Several of them can be pinpointed regionally, namely (see Schreier and Lavarello-Schreier 2011 for more examples): <?page no="193"?> Words and the New Englishes 175 British dialectal dicelen ‘thistle’ (EDD lists it in Devon and Cornwall) gansey ‘knitted pullover’ (guernsey ‘a thick, knitted, closely fitting vest or shirt […] worn by seamen’, EDD) qualmish word applied to various stomach disorders scudda buddy ya ‘drunk, tipsy’ (Scots scud ‘to drink copious draughts’, EDD) snislens dessert made of boiled potatoes (cf. Scots snisle ‘to singe, burn partially, harden with heat’, EDD) wee ‘small’ (this is current usage in Scottish English) South African (Afrikaans and South African English) banki ‘stool, bench’ bakki ‘pick-up truck’ (Afrikaans bakki) braai ‘barbecue’ (Afrikaans braai) bredie ‘fried meat’ (usually beef) cooked with added vegetables (e.g. cabbage bredie, pumpkin bredie, etc.) kappi ‘bonnet’ (Afrikaans kappi) kraal ‘sheep pen’ koibichi ‘take a little nap, sleep for a little while’ (Afrikaans kooi ‘bed’ + bietjie ‘a little bit’) lekker ‘delicious’ (also ‘attractive’) (Afrikaans and/ or Dutch lekker) oukabaatjie person of racially mixed descent (Afrikaans oude kabaatjie ‘bogeyman’) tackies ‘running shoes’ American origins buddy kinship term used for addressing males gulch ‘steep, narrow ravine’ taters ‘potatoes’ St Helenian English catfish ‘octopus’ (this unusual meaning is attested in the OED as “2. The cuttle-fish or other cephalopod”, historically attested as follows: “1678 Phillips, Catfish, a sort of Fish in some parts of the West Indies, so called from the Round-head, and large glaring Eyes, by which they are discovered in the Concavities of the Rocks. 1758 Baker in Phil. Trans. L. 785 Sea Polypi are <?page no="194"?> Daniel Schreier 176 frequent in the Meditteranean. A different species … came from the West Indies, where it is called a Catfish.” This points to sociohistorical contacts between St Helena and the Caribbean, which might also explain structural parallels. hatchabali exclamation to frighten children (often reduced to hatcha; origin unknown) piccaninny ‘child’ (ultimately Portuguese, yet attested in practically all English-based pidgins and creoles) Moreover, we also find some independent creations and community-based innovations in TdCE: canteen local supermarket crawford ‘curious, inquisitive person’ (“you’s an old Crawford”), refers to Allan Crawford, a British cartographer who visited the island in the 1930s and 1940s station fella ‘foreigner’ (< expatriate staying on the island in the station building) Some words, finally, have origins that still need to be ascertained: fardi ‘godfather’ muddish ‘godmother’ heitamassie exclamation of astonishment jadda boys ‘penguin eggs’ kee-kee ‘ear’ (widely believed by Tristanians to be from Afrikaans or Dutch, but not attested) Of course, it is not always a straightforward task to delimit the origins of words, even if these are known. For instance, some borrowings may have multiple origins. The TdCE word tiddy ‘sister, female friend, term of endearment towards a woman’ (e.g. ‘How you is, tiddy? ’) is attested in Scottish English and in West African Creoles (and also in StHE), or the word padre ‘priest’ is originally Portuguese, yet also used in Indian English, British Army usage, and English-based Creoles. All in all, the existence of these words correlates with the social history and supports claims that TdCE as such is in fact a hybrid variety that drew features from several, perhaps all the inputs brought to the island. It is firm evidence (just as in Bauer’s analysis of NZE) that settlers arrived from various locations and that the first generations of native speakers selected features from their parents’ speech, forming a hitherto new and unattested contact variety by doing so. However, settings such as Tristan da Cunha <?page no="195"?> Words and the New Englishes 177 also show that speakers have the possibility to coin and form new words, which are consequently endemic and extremely diagnostic. 4 Conclusion To conclude, lexical borrowing is one of the earliest and most immediately obvious mechanisms of contact-induced language change. Words may travel lightly, even in the absence of face-to-face contact between speakers (hence the origins of English borrowings in countries with no historical ties to Britain or the US) but it is certainly misguided to disregard (let alone ignore) lexical evidence in present-day research. Aspects relating to the quantity or quality of borrowings offer important clues as to how speech communities interact, whether there are sociocultural pressures of accommodation and who is in a socially dominant position (super-, ador substrate) so as to serve as a source of borrowings. The correlation of borrowings with cultural pressure and social organisation is a most promising advance in contact linguistics and the lexicon yields important information in order to determine (1) exactly what varieties come into contact, perhaps also in what proportions, and which of them was most influential, and (2) the degree of interaction and quite possibly also of social stratification between ethnic groups in contact. Moreover, lexis complements findings from phonology and syntax; if there are differing results, then these may be complementary rather than contradictory and both can be integrated into one coherent explanation. In the NZE example, discussed above, lexical distribution of words from the donors is an indication how many inputs were brought to New Zealand; at the same time, they are indicative of the degree of strength of sociolinguistic levelling effects (Trudgill 1986), as a result of which phonetic features were not adopted. To put it differently, words may remain where phonemes do not, and why they differ should receive much more attention in the literature than has hitherto been the case. The lexicon therefore continues to be a fruitful area for investigation in historical linguistics, complementing, at times even openly challenging, results from the analysis of phonetics/ phonology, morphology and syntax. <?page no="196"?> Daniel Schreier 178 Biographical sketch Daniel Schreier studied English, French and General Linguistics at the Universities of Basel and Neuchâtel. He researched and taught at North Carolina State University, the University of Canterbury at Christchurch (New Zealand), the University of Regensburg and the University of Bern. Since 2006, he is Professor of English Linguistics at the University of Zurich. His area of specialisation is in sociolingustics and contact linguistics; among others, he has carried out research on the history and origins of English in the South Atlantic Ocean (Tristan da Cunha and St Helena), earlier New Zealand English as well as contact-induced phonotactic changes in English around the world. He has written several books and is on the editorial board of English Worldwide and Multilingua. References Bauer, Laurie. 1999. “On the Origins of the New Zealand Accent”. English World- Wide 20.2, pp. 287 - 307. Bauer, Laurie. 2000. “The Dialectal Origins of New Zealand English”. In Allan Bell and Koenraad Kuiper (eds.). New Zealand English. Amsterdam and Philadelphia: John Benjamins, pp. 40 - 52. Baugh, Albert C. and Thomas Cable. 1993. A History of the English Language. (Fourth edition, first edition 1951). London and New York: Routledge. Comrie, Bernard. 1981. Language Universals and Linguistic Typology. Chicago: University of Chicago Press. Crystal, David. 2003. The Stories of English. London: Penguin. Dekeyser, Xavier. 1986. “Romance Loans in Middle English: A Re-assessment”. In Dieter Kastovsky and Aleksander Szwedek (eds.). Linguistics across Historical and Geographical Boundaries: In Honour of Jacek Fisiak on the Occasion of His Fiftieth Birthday. Berlin: Mouton de Gruyter, pp. 253 - 265. The English Dialect Dictionary. 1898-1905. Edited by Joseph Wright. 6 volumes. London: Frowde, Corner. Fischer, Andreas. 2001. “Lexical Borrowing and the History of English: A Typology of Typologies”. In Dieter Kastovsky and Arthur Mettinger (eds.). Language Contact in the History of English. Frankfurt am Main: Peter Lang, pp. 97-115. Gordon, Elizabeth, Lyle Campbell, Margaret Maclagan, Andrea Sudbury and Peter Trudgill. 2004. New Zealand English: Its Origins and Evolution. Cambridge: Cambridge University Press. Green, Lisa J. 2002. African American English: A Linguistic Introduction. Cambridge: Cambridge University Press. Jespersen, Otto. 1938. Growth and Structure of the English Language. (Ninth edition, first edition 1905). Oxford: Basil Blackwell. <?page no="197"?> Words and the New Englishes 179 Krapp, George Philip. 1925. The English Language in America. New York: The Century Co. Kurath, Hans. 1949. A Word Geography of the Eastern United States. Ann Arbor: University of Michigan Press. Labov, Teresa. 1992. “Social and Language Boundaries among Adolescents”. American Speech 67.4, pp. 339 - 366. Lee, Margaret G. 1999. “Out of the Hood and into the News: Borrowed Black Verbal Expressions in a Mainstream Newspaper”. American Speech 74.4, pp. 369 - 388. Orton, Harold and Eugen Dieth (eds.). 1962-1971. Survey of English Dialects. 4 volumes. Leeds: Arnold. The Oxford English Dictionary. 1989. Prepared by John Simpson and Edmund Weiner. (Second edition, first edition 1928). Oxford: Clarendon Press. The Oxford New Zealand Dictionary. 1997. Edited by Harry W. Orsman. Auckland: Oxford University Press. Rickford, John R. and Russell J. Rickford. 2000. Spoken Soul: The Story of Black English. New York: John Wiley. Schneider, Edgar W. 1996. “Introduction: Research Trends in the Study of American English”. In Edgar W. Schneider (ed.). Focus on the USA. Amsterdam and Philadelphia: John Benjamins, pp. 1 - 12. Schneider, Edgar W. 2007. Postcolonial English: Varieties around the World. Cambridge: Cambridge University Press. Schreier, Daniel. 2003. Isolation and Language Change: Sociohistorical and Contemporary Evidence from Tristan da Cunha English. Houndmills/ Basingstoke and New York: Palgrave Macmillan. Schreier, Daniel and Karen Lavarello-Schreier. 2011. Tristan da Cunha and the Tristanians. London: Battlebridge. Siegel, Jeff. 1973. “A Survey of Language Use in the Indian Speech Community in Fiji”. Unpublished typescript. Field Study Project for the Culture Learning Institute, East West Centre, Honolulu. Siegel, Jeff. 1986. “Pidgin English in Fiji: A Sociolinguistic History”. Pacific Studies 9.3, pp. 53-106. Siegel, Jeff. 1989. “English in Fiji”. World Englishes 8.1, pp. 47-58. Stewart, William A. 1967. “Sociolinguistic Factors in the History of American Negro Dialects”. Florida FL Reporter 5.2, pp. 3-8. Stockwell, Robert & Donka Minkova. 2001. English Words: History and Structure. Cambridge: Cambridge University Press. Strang, Barbara. 1970. A History of English. London: Methuen. Tent, Jan. 2001. “A Profile of the Fiji English Lexis”. English World-Wide 22.2, pp. 209 -2 45. Thomason, Sarah G. & Terrence Kaufman. 1988. Language Contact, Creolization, and Genetic Linguistics. Los Angeles: University of California Press. Trudgill, Peter. 1986. Dialects in Contact. Oxford: Blackwell. Turner, Lorenzo Dow. 1949. Africanisms in the Gullah Dialect. Chicago: University of Chicago Press. <?page no="198"?> Daniel Schreier 180 Vennemann, Theo. forthcoming. “English as a Contact Language: Typological Considerations”. To appear in Daniel Schreier & Marianne Hundt (eds.). English as a Contact Language. Cambridge: Cambridge University Press. Weinreich, Uriel. 1953. Languages in Contact: Findings and Problems. New York: Linguistic Circle of New York, No. 1. Zettersten, Arne. 1969. The English of Tristan da Cunha. (Lund Studies in English 37). Lund: Gleerup. <?page no="199"?> Annina Seiler The Function of the Sword-Hilt Inscription in Beowulf Abstract 1 The article presents an analysis of an inscription on the sword hilt described in Beowulf 1687-98a. This inscription represents the only reference to writing in the entire poem and has consequently received much scholarly attention. In this paper, Seiler claims that the sword-hilt inscription should be examined as a piece of writing used in a situation of “vocality”, a term that characterizes a society which makes use of writing as a whole, but where access to literacy is mediated only by a small minority of experts. In this view, Hrothgar can be identified as the reader of the inscription. A comparison with two other runic texts serves to highlight the function of the sword-hilt inscription: unlike the rune-stick occurring in a parallel part of Grettis saga, the Beowulf inscription is not a fully functioning means of communication. Rather, it serves as a support for Beowulf’s oral account of his adventure and, like the South Germanic inscription on the Pforzen buckle, it establishes a connection to a legendary past, which serves as a kind of prophecy fulfilled by the present. Lines 1687-98a of Beowulf give an account of an inscription on a sword hilt, which Beowulf brings back from his fight with Grendel’s mother. The weapon to which the hilt belongs is first introduced into the story in line 1557ff., where we learn that it was originally made by giants. Beowulf seizes this sword when Hrunting, the weapon lent to him for the occasion by Unferth, fails in the fight against Grendel’s mother (1522-28, 1666- 68). The hero manages to kill the she-monster with this giant-made sword, but its blade melts due to the blood (1605-11, 1615-17). He takes the hilt back to the surface of the lake, together with Grendel’s head and Unferth’s sword. Back in Heorot, Beowulf retells his story, and King Hrothgar receives the hilt. In lines 1687-98a the poem gives a detailed description of it: 1 I had the opportunity to discuss a first idea for this paper with friends and colleagues in the “Mediävistische Mittelbaukolloquium” in Zurich and I would like to thank all those who provided valuable feedback, especially Michelle Waldispühl, Martin Graf, Gerald Schwedler and Luzius Thöny. <?page no="200"?> Annina Seiler 182 Hrōðgār maðelode - hylt scēawode, ealde lāfe, on ð m wæs ōr writen fyrngewinnes, syðþan flōd ofslōh, gifen gēotende gīganta cyn; frēcne gefērdon; þæt wæs fremde þēod ēcean Dryhtne; him þæs endelēan þurh wæteres wylm Waldend sealde. Swā wæs on ð m scennum scīran goldes þurh rūnstafas rihte gemearcod, geseted ond ges d, hwām þæt sweord geworht, īrena cyst rest w re wreoþenhilt ond wyrmfāh. 2 Hrothgar spoke, he looked at the hilt, the old relic, on which was carved the beginning of an old war, after the flood, the rushing torrent, had destroyed the race of giants. They fared terribly - it was a people estranged to the eternal lord. For this the ruler gave them a final reward through the surging of water. In the same way, there was on the scenns (plates? ) of shining gold with runes properly marked, set and said, for whom the sword had first been made, the best of blades, with a twisted hilt and serpentine ornamentation. This passage has received much critical attention; discussion has focussed, first of all, on the exact parts of the sword on which the inscription is located, centring on the interpretation of the hapax legomenon scenn (1694). 3 A further matter of debate is the question of whether the inscription consists only of runes, or also of a graphic illustration. 4 The point is controversial because of the ambiguous verb writan (1688), which very often refers to writing, but primarily means ‘scratch, carve’ and, consequently, can also be applied to the production of a drawing. Both lexical problems play a role in the discussion of the amount of space available on the sword hilt. Concerning the content of the first part of the inscription or drawing (1688b-93), scholars have looked to both the biblical account of the Flood 5 and to Germanic mythology 6 for analogues. The second part 2 All quotations from Beowulf come from Klaeber (1936). Translations, unless otherwise stated, are my own. 3 The form scennum is a dat. pl. of a fem. jōor masc. ja-stem (< *skanjō/ a); it is probably related to ON skán f. ‘bark, crust’ and to ON skinn n. ‘skin’ < *skenþan (which was also borrowed into English; Mod. Engl. skin); cf. Falk (1914: 30) quoted in Cramp (1957: 66), Vries (1962: 482), Ellis Davidson (1962: 138). Plausible translations are ‘plates’ or ‘covering’. 4 Cf. Klaeber (1936: 189f.); only runes: Chickering (1977: 343), Lerer (1991: 170f.), Anderson (2010: 148f.); runes and pictures: Cramp (1957: 66), Osborn (1978: 977), Gwara (2008: 183f.), Grünzweig (2009: 194-198). 5 Giants are mentioned in Genesis 6.4 just before the account of the Flood; cf. Osborn (1978), Irving (1997: 188). 6 A possible Germanic analogue occurs in the Snorra Edda (Gylfaginning, ch. 7): Synir Bors drápu Ymi i tun, en er hann fell, þá hlióp svá mikit blóð ór sárum hans at með þvi <?page no="201"?> The Function of the Sword-Hilt Inscription in Beowulf 183 of the inscription (1694-98a) is less disputed - everybody seems to agree that it contains a personal name written in runes (1695 rūnstafas), most probably of the first owner of the sword (1696bf. hw ā m þæt sweord geworht […] rest w re), which is paralleled by extant sword hilts from the Anglo- Saxon period. From a literary point of view, the function of this passage within the entire poem has particularly been discussed with regard to the connection between the sword-hilt inscription and Hrothgar’s subsequent discourse, which is commonly termed Hrothgar’s ‘sermon’. This speech is one of the longest speeches in the entire poem, but there is no clear reference to the inscription. What makes the sword-hilt passage especially interesting is the fact that it contains the only reference to writing in Beowulf (cf. Lerer 1997: 337-39). Some of the issues just mentioned will be touched upon below, but what I want to focus on in the main is the swordhilt inscription with respect to the functions and the use of writing in the context of early medieval (il)literacy. A typical aspect of literacy in Anglo-Saxon England is that society as a whole knows and makes use of writing, but that there is a large majority of people who can neither read nor write, and only a small minority of experts who actually do dispose of the skills of literacy (cf. Schaefer 1992: 21). As far as we know, this holds true for both alphabetic manuscript writing and for runic writing in Anglo-Saxon England, but there is very little evidence concerning the carvers of runic inscriptions aside from the inscriptions themselves. 7 This specific state of literacy has been described by Ursula Schaefer (1992: 15-42) as a situation of “vocality” (German “Vokalität”). It refers to the fact that access to writing is mediated by the literate minority through vocalisation (“Verlautlichung”) and its reverse-process scripting (“Verschriftung”). Scripting and vocalisation designate only the transfer from the spoken to the written medium and vice versa and are distinguished from the conceptual changes, oralisation (German “Vermündlidrekðu þeir allri ætt hrím[þu]rsa, nema einn komz undan með sínu hýski; þann kalla i tnar Bergelmi - ‘The sons of Bor killed the giant Ymir, and when he fell, so much blood flowed from his wounds that with it they drowned all frost-giants; only one escaped with his household; the giants call him Bergelmir’ (Lorenz 1984: 152). Scholars have argued that Snorri’s account depends on the biblical account of the Flood. However, according to Lorenz (1984: 153-34), an independent tradition seems more likely, given the differences between the versions. Apart from that, Slade (2007) has demonstrated that creation myths including a Flood are of common Indo-European stock. Of course, it is possible that the Beowulf poet is referring to both Genesis and to Germanic mythology, as suggested by Lane (2000): “The poet would have found in the Norse material a precious echo of an ‘original’ lore whose pristine form (in his view) had been preserved in his own handbook, the Bible”. See also Anderson (2010: 144) and Ellis Davidson (1962: 141). 7 On runic literacy relating to coin legends see Blackburn (1991); on “[w]riting in an oral society” (in runes and the Latin alphabet) see Green (1994: 35ff.). <?page no="202"?> Annina Seiler 184 chung”) and textualisation (“Verschriftlichung”), which concern stylistic adaptations resulting from the change in medium (cf. Koch and Oesterreicher 1985, Oesterreicher 1993, Schaefer [forthcoming]). Vocalisation and scripting alone occur, for example, when a letter is read out or when it is written down by the process of dictation. In many cases, however, processes of scripting and vocalisation were also accompanied by the corresponding changes on the conceptual level. For instance, the existence of charter drafts (often written on the back side of the final charter) suggests that, during negotiations, only some important words (such as names and bounds) were actually written down. Later, in the scriptorium, the text was elaborated and drafted according to the (literate) conventions of this text type, which is an example of textualisation. Given its ephemeral nature, evidence for processes of vocalisation is more difficult to obtain, but communicating the content of written texts to non-literate people in many cases included more than merely establishing character-sound correspondences (cf. Schaefer 1992: 17). 8 In general, communication was primarily effected as oral communication, and writing was used to support, but not to replace, oral communication (cf. Clanchy 1993: 278). The inscription described in the sword-hilt passage from Beowulf fits into the pattern of vocality: Hrothgar, the reader, is not depicted as reading out the inscription word by word (or character by character); instead the poem provides us with a general idea of the inscription and its content. Furthermore, the hilt inscription does not function like normal written communication; rather it plays a role in an essentially oral setting (cf. below). Most scholars have interpreted the insertion between the maðelode formula and the actual beginning of direct speech in line 1700 as a pause in which Hrothgar is silent. 9 Michael Near (1993: 324), however, argues that Hrothgar starts to speak while he is looking at the inscription: “The narration suggests, in fact, that the sermon and the investigation of the hilt proceed simultaneously, for Hrothgar begins his address (‘Hroðgar maðelode’ [1687a]) at the same time that he begins his examination of the sword (‘hylt sceawode’ [1687b])”. Furthermore, Anderson (2010: 148) has 8 Spurkland (2000) discusses Scandinavian examples of letters that are retold instead of read out, which would be an example of oralisation instead of vocalisation. I am grateful to Kevin Müller for pointing this out to me. 9 For example, Lerer (1991: 161) talks of “a jarring break in the narrative continuity of the story. It begins with the announcement of Hrothgar’s speechmaking, only to break off that speech before it begins. The verbs of its first line shift abruptly from speaking to looking, from the assertion of a verbal performance to the mute apprehension of a thing”. Köberl (1987: 121) remarks that the sword hilt “keep[s Hrothgar] speechless for twelve and a half lines” and comments on “Hrothgar’s reticence on this point (especially when compared with his usual lack of this quality)”. Similarly, Osborne states that “while Hrothgar gazes, considering what to say, the poet tells us what is written on the hilt” (1978: 977; original emphasis). <?page no="203"?> The Function of the Sword-Hilt Inscription in Beowulf 185 pointed out that “[w]hen maðelode is paired with another verb, or followed by an adverb or independent clause, the syntax suggests simultaneity”. I find this interpretation convincing, but - unlike Near - I would go even further and claim that what Hrothgar is talking about while he is analysing the sword hilt is not his admonition to Beowulf, but rather the inscription itself: that the passage in fact provides a summary of Hrothgar’s comments on the inscription. 10 This hypothesis is supported by an analysis of the end of the sword-hilt passage, which concludes with a kind of ‘second introduction’ to the sermon in lines 1698b-99: ([…] wreoþenhilt ond wyrmfāh.) Ðā se wīsa spræc sunu Healfdenes - swīgedon ealle -: ‘Þæt, lā, mæg secgan sē þe sōð ond riht […]’ In accordance with the punctuation in Klaeber’s (1936: 63) and most other editions, which put a period before ða and a colon at the end of the sentence, it is usually translated as ‘Then, the wise one, the son of Healfdene, spoke - all were silent -: “This indeed may say he who [has furthered] truth and right […]”’. Such a translation is in line with the general opinion that Hrothgar first examines the hilt and only starts to speak afterwards. Yet it is also possible to interpret the sentence as much more closely linked to the previous passage, i.e. […] wreoþenhilt ond wyrmfāh; ðā se wīsa spræc, sunu Healfdenes, swīgedon ealle. This would be translated as follows: ‘While the wise one, the son of Healfdene, was speaking, all were silent’. In the manuscript, there is no dividing period or line break after wyrmfah, and, from a syntactic point of view, it is actually more convincing to translate ða as the conjunction ‘when, while’ and not as the adverb ‘then’ since the word order þā - S (se wīsa) - V (spræc) commonly occurs in subordinate clauses. 11 This means that the sentence could refer not only to Hrothgar’s direct speech, beginning in line 1700, but also to the previous lines that describe the sword hilt and which are introduced by Hroðgar maðelode. This would indicate that Hrothgar 10 Anderson (2010: 145-151) discusses the question of whether “the Flood story [is …] told by the poet, or by Hroðgar in indirect discourse? ” in detail. He finds the textual arguments inconclusive, but from the point of view of comparative mythology considers it to be more likely that Hrothgar does read the runes. Cf. also Gwara (2008: 184-186). 11 Cf. Bosworth and Toller (1898: 1030): “When the word [þā] stands at the beginning of a clause and may be translated by then, the verb generally precedes its subject; if it is to be translated by when the subject generally precedes the verb”. In poetry, word order is “a less certain guide than it is in the prose” (Mitchell 1985: 251), but it is still consistent with prose word order in the majority of cases (cf. §2444, 2536, 2547). <?page no="204"?> Annina Seiler 186 indeed starts to use his voice simultaneously with his eyes, but that the poem does not render his speech in detail. 12 The question of who is speaking is closely linked to that concerning the audience of the story on the sword hilt. If we assume that Hrothgar is the reader it is possible that he communicates his reading to Beowulf, Wealhtheow and the Geatish and Danish warriors. On the other hand, if we take the narrator of Beowulf as the source of information, the retainers necessarily remain in the dark. Near (1993: 324) attributes the passage to the poet, who “makes dramatic use of his privileged knowledge” throughout the poem. In a similar way, Osborn (1978: 978) takes the poet as the source of information and not Hrothgar: “He gazes upon the hilt, and the information with which the poet provides us during this pause gives a scriptural context for the wisdom that Hrothgar subsequently reveals about the recurrent feud with mankind’s enemy within the human breast” (original emphasis). Both assume that Hrothgar does not actually read or understand the inscription and that the pun is only apparent to the audience of the poem (Anglo-Saxon or modern). However, Köberl (1987: 121) dismisses the idea that Hrothgar cannot read the inscription: “While possibly true of real-life kings of that period, this explanation does not sound particularly convincing within the literary framework of our poem: we are left with an apparently blind motif”. I agree and, furthermore, I believe that Hrothgar would not be referred to as se wisa ‘the wise one’ right after a scene in which he is depicted as not being able to read runes. Lerer (1991: 172f.) also sees Hrothgar as an understanding reader (which can be taken as a model for the audience of Beowulf) and analyses the sermon as “a ‘reading’ of the hilt or an oration ‘inspired’ by its text”, connecting the inscription and the sermon as rūn and r d - two words that are paired in a line from the Old English Maxims ræd sceal man secgan, rune writan ‘one shall give advice, write runes’ (quoted in Lerer 1991: 172). He points to the etymological connection between r d ‘advice’ and Mod. Engl. read ‘to read’ < OE r dan ‘to advise, counsel’ and concludes that giving advice is a public activity, which may include “interpretations of the written texts of secret or of learned instruction” (Lerer 1991: 172). In Lerer’s (1991: 174) opinion “the scene posits a view of discourse as written and private”, and is “not the presentation of a public performance but the controlled evocation of private reading”. Whereas I am fully convinced that the analogy of rūn and r d works very well as an interpretive pattern for the sword-hilt passage with the inscription (rūn) as a trigger for the sermon (r d), I feel that Lerer’s analysis runs into problems where it characterizes reading as an essentially private activity. First of all, Hroth- 12 On maþolian ‘speak’ in Beowulf and Old English literature see Rissanen (1998); cf. also Anderson (2010: 146, 148). <?page no="205"?> The Function of the Sword-Hilt Inscription in Beowulf 187 gar’s analysis of the sword-hilt inscription is not at all private - it takes place in the hall in front of Beowulf, Queen Wealhtheow and both the Danish and the Geatish warriors. Furthermore, the interpretation of reading as private is based on a modern assumption of reading, which - according to Saenger (1997: 1) - “is a silent, solitary, and rapid activity” and contrasts with “[a]ncient reading [which] was usually oral, either aloud, in groups, or individually, in a muffled voice”. So if we think of Hrothgar, bent over the sword hilt and slowly deciphering the inscription, we have to imagine as well that he is articulating the sounds of the runes he is looking at - similar to a school child spelling out a difficult word. Modern silent reading depends on the fact that every word has its “unique visual/ spatial representation in the [mental] lexicon” (Katz and Feldmann 1996: 1094). Such representations were not available to the medieval reader since word separation was not fully established and consistent word spelling was rare. 13 With respect to the reading of inscriptions, I feel that it is even less likely that silent reading was possible. 14 Apart from that, the oral aspect of writing is mentioned in the sword-hilt passage itself: the owner’s name is not only rihte gemearcod ‘correctly marked’ and geseted ‘set down, placed’ (1695f.), but also gesæd ‘spoken’ through the runes. In order to analyse the function of the runic inscription as a written text, I will compare the sword-hilt passage with two other examples of runic literacy: the inscriptions on the Pforzen buckle and on the rune-stick occurring in chapter 66 of Grettis saga. The Pforzen buckle displays functions of writing that are similar to those of the sword-hilt inscription in Beowulf, whereas the parallel passage from Grettis saga serves to demonstrate what the sword-hilt inscription is not, namely an example of fully functioning written communication. The Pforzen buckle is a rare example of a South Germanic runic inscription of some length. As I will show, there are several features of this inscription that can be compared to the one described in the Beowulf sword-hilt passage and they give us an idea of what the sword inscription may have looked like. Before I discuss the similarities in detail, I will address one relevant problem relating to the Beowulf inscription, namely that lines 1688-93 possibly describe a graphic illustration, rather than an inscription. Since the expression on ðæm wæs … writen 13 On the development of silent reading cf. Balogh (1927), Raible (1991), Green (1994), Saenger (1997). 14 A different view with respect to runic writing is advanced by Terje Spurkland (2001) who holds that in contrast to ‘oral’ Latin reading, runes were read silently. However, the material discussed in the quoted paper are rune-sticks from 13 th / 14 th c. Norway, where runic literacy (“runacy”) was used as an everyday means of communication. I believe that the same conditions do not apply to Anglo-Saxon or late migration-period runic inscriptions. <?page no="206"?> Annina Seiler 188 ‘on it was carved’ is ambiguous, either interpretation is possible. 15 Based on comparisons of the text passage with extant Anglo-Saxon swords, it has been argued that a sword hilt does not provide enough space for an inscription of such complexity and that, consequently, it had to be a drawing. 16 Cramp (1957: 66), for example, states “[h]ere writen must, I think, be taken in the sense of engraved, since any sort of inscription or poem lengthy enough to give this information […] would be highly improbable on such a small area as a sword-handle”. Webster (1998: 191) holds that “while no surviving sword carries anything resembling the story of the downfall of the giants (ll. 1687-93), there are a number of elaborately decorated hilts dating from the sixth to the eighth century in both England and Scandinavia where the complex motifs may carry special meaning”. Based on the same assumption, Grünzweig (2009: 194-199) discusses similarities between the possible symbolic meaning of animal style II ornamentation (represented, for example, by the Sutton Hoo purse lid) and the primordial fight mentioned in the sword-hilt passage (1689 fyrngewin ‘ancient strife’). Animal style II motifs have been interpreted as representing a fight between good and evil, light and darkness, or between chaos and order, which could correspond to the fight in the Beowulf sword-hilt passage (Grünzweig 2009: 197). However, it is possible to argue for an inscription entirely in runes on the basis of the text passage. If we take 1688f. on ðæm wæs ōr writen fyrngewinnes literally as ‘on it was recorded the beginning of an old war’, and not ‘the course’ or ‘all about the ancient strife’ (cf. for example Heyne and Schücking 1961: 102f.), I find it difficult to see how this could be represented in a pictorial way. On the other hand, it is very easy to represent the beginning of a story linguistically. As Chickering points out, a literal interpretation of ōr also saves the problem of space: “[t]here is enough room on surviving sword hilts for an inscription such as Hrothgar reads. It need only have been the first words of a verse from Genesis” (Chickering 1977: 343) - or, of course, from a Germanic legend. Furthermore, Lerer has suggested that the swā of line 1694 does not necessarily mean that the first part is not in runes, but that it can be interpreted as “also written in runes the name of the dedicatee” (Lerer 1991: 171; his 15 A different opinion is voiced by Anderson (2010: 146f.), who argues that in West and North Germanic, OE writan / ON rísta “refers specifically to runes”. This is true in Old English for most contexts; however, some examples quoted by Bosworth and Toller (1898: 1275) clearly refer to drawings, e.g. Writ ðisne circul mid ðines cnifes orde on anum […] stane - ‘Carve this circle with the point of your knife on a stone’ (Leechdoms I, p. 395). 16 Obviously, an inscription in a poem is not obliged to adhere to realism. However, in view of the fact that other details of the sword and of weapons mentioned elsewhere in Beowulf seem to be consistent with archaeological swords (cf. Ellis Davidson 1962: 211, Chadwick Hawkes and Page 1967) I think that such an interpretation should only be made as a last resort. <?page no="207"?> The Function of the Sword-Hilt Inscription in Beowulf 189 emphasis). Even though the exact significance of scenn is unclear, the scennum sciran goldes may indicate a different part of the sword hilt, i.e. that one part of the inscription is, for example, on the pommel, and the other one on the gold covering of the handle, but both are in runes. Whereas “no surviving sword carries anything resembling the story of the downfall of the giants” (Webster 1998: 191), the Pforzen buckle inscription does, and I believe that it is in many respects comparable to the sword-hilt inscription described in Beowulf. The silver belt buckle containing the inscription was found in 1991/ 92 in a grave in the Alemannic cemetery of Pforzen and dates to the 6 th century (cf. Looijenga 2003: 253- 255, Bammesberger and Waxenberger 1999). The grave was most probably a warrior grave since it also contained weapons and a shield. The inscription consists of two lines - probably representing an alliterative long line - which are transliterated by Nedoma (2004: 344) as follows: Aigil andi Ailrūn / (I)ltahu gasōkun - ‘Aigil and Ailrun / fought at the Ilzach (a river)’. Whereas the first part of the inscription is read by most scholars as two personal names Aigil and Ailrun - presumably a man and a woman -, the second part (especially the first word) seems to be less clear. The verb gasokun probably represents the third person plural, past tense, indicative of the strong verb gasakan, which has cognates in Old High German gasahhan ‘to condemn, censure’ and Gothic gasakan ‘to rebuke, convict, silence’ (Nedoma 2004: 360). In the context of the Pforzen inscription, it is translated as ‘cursed, threatened, rejected, scolded, disputed, fought’. 17 Interpretations of the first word(s) of the second line are much more diverse, such as, a ‘stag masquerade’, the ‘eel water’, a personal name, a river name, or just ‘all’ (my translations from German to English; for more suggestions and references see Nedoma 2004: 343-348, Marold 2004: 217- 219). The consensus seems to be that the inscription relates to some sort of fight involving the two persons Aigil and Ailrun. Whatever the exact meaning of the second part of the inscription, it seems likely that Aigil and Ailrun correspond to the archer Egill and his wife lrún (to use the Old Icelandic names) and that the inscription represents the earliest attestation of a legendary story attached to the Wayland myth; the story may even describe the same scene as that depicted on the lid of the Franks Casket. There, a man labelled Ægili is presented with a bow and arrow, defending a building with a woman inside against a group of attackers (Marold 2004: 219-223, 227-235, Nedoma 2004: 354-361, Looijenga 2003: 255). Now the parallels to the Beowulf inscription do not relate to the content of the story; rather, they are of a typological nature: first of all, the Pforzen buckle demonstrates that lack of space is not a valid argument against the assumption that the inscription in Beowulf is entirely in runes. The Pforzen 17 For a different opinion see Bammesberger (1999: 118). <?page no="208"?> Annina Seiler 190 inscription consists of about 26 characters (presumably five words), which are placed in an area of two lines, which are both about 5 centimetres long and 0.8 wide (measured on the 1: 1 drawing of the Pforzen buckle in Babucke 1999: 17). As much space is certainly available on a sword hilt; for example, the inscription on the Chessel Down scabbard mouth piece consists of seven characters and the one on one side of the Gilton/ Ash swordpommel of about 12 to 20 runes (Grünzweig 2004: 115-123, 2009: 203- 208; Chadwick Hawkes and Page 1967, Ellis Davidson 1962: 77-82, 96- 103). This means that an allusion to a complex story could easily be fitted on a sword hilt. If the Pforzen buckle inscription is a part or a topic sentence of an unknown Germanic legend, this would correspond to Köberl’s (1987) interpretation of the Beowulf sword-hilt inscription; Köberl argues that the name in the inscription is Heremod’s, and that the first part of the inscription includes details from an unknown Danish legend. Both inscriptions contain personal names, which may be true of many inscriptions, but the characters mentioned seem to be mythical characters rather than reallife people. This is certainly true for Beowulf, but Nedoma (2004: 355) and Marold (2004: 227) assume that it is also the case in the Pforzen inscription. An important point in this connection is that the inscription is placed on the front side of the buckle, where it was visible when the buckle was worn, rather than on the backside. In contrast, runic inscriptions of a private nature seem to be placed on portions of buckles or brooches, which are hidden from sight (Nedoma 2004: 354f). There are also some similarities relating to the content as both inscriptions seem to represent some sort of fighting (Pforzen: gasokan / Beowulf sword hilt: fyrngewin), and if Nedoma’s interpretation of (I)ltahu as a river name is correct, then both even have to do with water (but cf. Looijenga 2003: 255). Whatever the exact details of the story, a reference to a Germanic legend could have served as a motto or maxim for the bearer of the buckle (Marold 2004: 233f., Nedoma 2004: 363). This means that the writing may have had some sort of magic function: [D]ie erfolgreiche Bewältigung einer kritischen Situation aus sagenhafter Vorzeit wird evoziert, damit sich - nach dem Similitätsprinzip (“Wie einst, so jetzt”) - hic et nunc gleicher Erfolg wie im Präzedenzfall einstellen möge. (Nedoma 2004: 363) (The successful handling of a critical situation from a legendary past is evoked, so that - according to the principle of similitude (“As in former times, so now”) - the same success as in the precedent may be achieved in the here and now.) In the context of the Beowulf inscription, such an interpretation is very attractive, since the ‘principle of similitude’ applies to the Flood, which kills the giants, and to Beowulf, who kills the descendants of giants. We <?page no="209"?> The Function of the Sword-Hilt Inscription in Beowulf 191 could even argue that it is the magic of the runic inscription which makes it possible for Beowulf to kill Grendel’s mother, since his other sword cannot harm her. In addition to such a magic function of writing, both inscriptions certainly also fulfil a visual function (cf. Nöth 1990: 160f.). The Pforzen inscription is placed on an oblong piece of silver, which is divided into different sections by sets of parallel grooves. The writing space is almost completely used up by the runes and what is left is filled up with ornaments (cf. Babucke 1999: 17). The way in which the Beowulf sword hilt is described also gives the impression of a very precious piece of work; it is first of all decorated with gold, and the expressions wreoþenhilt ond wyrmfāh (1698) seem to refer to specific kinds of decoration, probably interlacing or serpentine patterns (Ellis Davidson 1962: 136f.). My second example, the Icelandic saga about Grettir the Strong, has long been recognized as an important analogue to Beowulf; sections 65 and 66 include many similarities to Beowulf’s fight with Grendel and Grendel’s mother. The historical events depicted in Grettis saga Ásmundarsonar date from the late 9 th to the mid-11 th century; the oldest extant manuscripts come from the late 15 th century (Seelow 1998: 243-260). In the following, I will briefly list the most important parallels: 18 Grettir, like Beowulf, fights against a male and a female monster, a troll-wife and her giant husband (the order is reversed in Grettis saga). The first fight takes place in a hall and ends with the monster’s arm being cut off by the hero; the second is situated in an underwater cave, which is the home of the monsters - under a mere in Beowulf and behind a waterfall in Grettis saga. A sword hanging on the wall of the cave plays a role in both stories: in Beowulf it is the sword with the hilt inscription, which Beowulf uses to kill Grendel’s mother and whose blade subsequently melts. In Grettis saga, the roles are inversed: the giant tries to grab a sword after Grettir cuts off the blade of the giant’s short-sword (heptisax), and Grettir uses this moment to kill the giant. Hrothgar in Beowulf and the priest waiting for Grettir in Grettis saga both think that the hero is dead when they see the blood of the dying monster in the water, and both consequently go home. After the fight, both heroes bring some proof of their adventure back from the cave: Beowulf takes Grendel’s head and the sword hilt, Grettir the bones of two men slain by the monster. In both texts, the adventure is told (at least) twice by the text: the first time by the narrator, the second time by the hero - orally in the case of Beowulf and on a rune staff by Grettir. A runic inscription (consisting of two parts) occurs at the end of both episodes - on the sword hilt and on the rune-stick respectively. On the rune-stick Grettir 18 For more details see, for example, Seelow (1998: 247f.), Garmonsway and Simpson (1968: 316), Chambers (1932: 451-485). There is general agreement about a genetic relationship between Beowulf and Grettis saga (cf. Swinford 2002); for a different opinion see Fjalldal (1998). <?page no="210"?> Annina Seiler 192 writes down what happened to him in the cave, deposits the stick in the church porch and leaves. The passage runs as follows (given in English translation except for some important expressions): Then he went back to Eyjardalsár, and brought the bag in which the bones were to the church porch, together with a wooden rod inscribed with runes, on which these verses were extremely well carved [rúnakefli því er vísur þessar voru forkunnlega vel á ristnar]: Into a dark gulf I went; / The rolling rock-filled torrent / Gaped its wet and icy mouth / To swallow up this swordsman. / The flowing current smote my breast / In the ogress’ dwelling; / The whirlpool’s hostile fury / Beat about my shoulders. And also these: The she-troll’s ugly husband / Left his cave to meet me; / Long he fought against me, / And hard, if truth be told. / The hard-edged hafted short-sword [heptisax] / I hewed from off its shaft; / The giant’s breast and belly / I clove with my bright blade. (Garmonsway and Simpson 1968: 316) There it also said that Grettir had taken the bones from the cave. And when the priest came to the church in the morning, he found the rod and what was with it and read the runes [fann hann keflið og það sem fylgdi og las rúnirnar]. But Grettir had gone home to Sandhaugar. (My translation, cf. Seelow 1998: 171) The sword-hilt inscription from Beowulf is similar to the message on the rune-stick in so far as it represents some sort of rendering of the preceding adventure. However, whereas the rune-stick in Grettis saga is used as a fully functioning means of communication with a clearly identifiable sender and receiver of the message (namely Grettir and the priest Steinn), the swordhilt inscription is much farther removed from the fight that just took place: it is about a mythical or biblical fight and about a flood in which a race of giants is drowned, and it represents Beowulf’s underwater fight against the descendants of giants only as a kind of mise en abîme. Though there is a name in the inscription, which may refer to the person who commissioned the writing, it is not recounted in the narrative. On the other hand, Grettir’s name is missing in the inscription, but it is clearly implied by the context and by the fact that the message is written in the first person. On the receiver side, Hrothgar turns into a reader only by accident. In Grettis saga, what has happened to Grettir in the cave is known exclusively because it is written down on the stick. Grettir leaves the stick in the church porch on purpose, so that it will be read by the priest. Concerning the medium of the message, the sword hilt functions first of all as (a part of) a weapon and as a precious object. Grettir’s rune-stick, on the other hand, serves no other purpose than that of carrying the message. In Grettis saga then, writing is used to extend communication over space and time. In Beowulf, however, <?page no="211"?> The Function of the Sword-Hilt Inscription in Beowulf 193 it is the verbal account of the hero himself that communicates his adventure to Hrothgar and the Danes. The inscription presents Beowulf’s feat as part of a larger frame of events, but it does not replace oral communication. To some extent, it functions on the same level as Grendel’s head, namely as proof of Beowulf’s story. In Grettis saga the bones act as proof, and the rune-stick provides the story. Ellis Davidson (1962: 139) surmises that Grettir’s rune-stick could actually be the shaft of the giant’s heptisax, which Grettir cut off during the fight. Yet, if this is the case, it is not mentioned - the rune-stick has no other function except as a means of written communication. A comparison of the sword-hilt passage from Beowulf with the Pforzen buckle and the rune-stick from Grettis saga makes it possible to assess the function of the inscription within the narrative. It becomes apparent that the written text on the hilt does not work as a fully functioning means of communication, as it might be used in a literate society. Whereas the runestick in the much younger Grettis saga exclusively serves the purpose of conveying a message in written form, the writing on the Beowulf sword hilt plays only a subsidiary role. The sword hilt functions not primarily as a means of communication; it is, first of all, a precious object and a part of a powerful weapon. It serves a visual function as a decoration on the hilt, with the incised runes enhancing the value of that object. It may also have a magic function, with the inscription supporting the weapon during the fight. But as a communication medium alone, it is rather inadequate. It cannot tell Beowulf’s adventure; its message only becomes meaningful in connection with the oral account of the hero. Furthermore, the sword has to fulfil its purpose as a weapon first before the runes are noticed and deciphered by Hrothgar. The function of the inscription is similar to that of the 6 th century Pforzen buckle inscription, which evokes a connection to a legendary past and can only be linked to the wearer of the buckle in a vague way as an implication of the fame he hopes to gain. Both inscriptions are typical examples of writing in a situation of “vocality”, where only few people can actually read or write, and where writing serves as a support for oral communication and not as a replacement of it. In Beowulf, Hrothgar represents the small minority of literates and acts as mediator of the sword-hilt inscription. The process of reading focuses on oralisation, a free rendering of the content, and not on vocalisation, an exact word for word account. His audience are both the characters in Beowulf and the readers of the poem. For the audience within the text, the sword-hilt passage is the second telling of a fight against giants; it appears as a prophecy which Beowulf has fulfilled and which elevates Beowulf to the league of legendary heroes. For the audience of the poem, it is already the third account, after the events themselves are described by the poet and Beowulf’s personal rendering. (A fourth version is finally given when Beowulf returns home to <?page no="212"?> Annina Seiler 194 the Geatish court). In an oral performance of the poem, the sword-hilt passage may even have played a performative role, with a scop holding up a sword (or a sword hilt 19 ) and establishing associations between the fyrngewin described in the inscription in the poem, Beowulf’s adventures, and the hopes and fears of the scop’s present audience in the hall - thereby extending the function of the sword-hilt inscription beyond the text. Biographical sketch Annina Seiler works as Wissenschaftliche Mitarbeiterin at the English Department of the University of Zurich. Her general research interests include English language history, Old English, historical graphemics and phonology, the West Germanic languages. She studied German, English Linguistics, and Comparative Germanic Philology at the University of Zurich. From 200--2009 she was a member of the National Center of Competence in Research (NCCR) Mediality, where she wrote her doctoral dissertation on the scripting of the Germanic languages, supervised by Andreas Fischer and Elvira Glaser. Annina Seiler teaches Old and Middle English at the English Department of the University of Zurich, and English linguistics for the Zurich University of Teacher Education. 19 Chadwick Hawkes and Page (1967: 3) point out that in the Sarre cemetery two detached sword pommels were buried (one of them containing an inscription). It seems likely that precious sword hilts were kept as valuable objects even if they could no longer be used as weapons. <?page no="213"?> The Function of the Sword-Hilt Inscription in Beowulf 195 References Anderson, Earl R. 2010. Understanding Beowulf as an Indo-European Epic. A Study in Comparative Mythology. Lewiston N.Y.: Edwin Mellen Press. Babucke, Volker. 1999. “Die Runenschnalle von Pforzen (Allgäu) - Aspekte der Deutung. 1. Zur Herkunft und Datierung: Archäologischer Befund”. In Alfred Bammesberger and Gaby Waxenberger (eds.). Pforzen und Bergakker. Neue Untersuchungen zu Runeninschriften. Göttingen: Vandenhoeck & Ruprecht, pp. 15-24. Balogh, Josef. 1927. “‘Voces paginarum’. Beiträge zur Geschichte des lauten Lesens und Schreibens”. Philologus 82, pp. 84-109, 202-240. Bammesberger, Alfred. 1999. “Pforzen: Eine Anmerkung zu gasōkun”. In Alfred Bammesberger and Gaby Waxenberger (eds.). Pforzen und Bergakker. Neue Untersuchungen zu Runeninschriften. Göttingen: Vandenhoeck & Ruprecht, p. 118. Bammesberger, Alfred and Gaby Waxenberger (eds.). 1999. Pforzen und Bergakker. Neue Untersuchungen zu Runeninschriften. Göttingen: Vandenhoeck & Ruprecht. Blackburn, Mark. 1991. “A Survey of Anglo-Saxon and Frisian Coins with Runic Inscriptions”. In Alfred Bammesberger (ed.). Old English Runes and their Continental Background. Heidelberg: Universitätsverlag Winter, pp. 137-189. Bosworth, Joseph and Thomas Northcote Toller. 1898. An Anglo-Saxon Dictionary. Based on the Manuscript Collections of the Late Joseph Bosworth. Oxford: Clarendon Press. Chadwick Hawkes, Sonia and Raymond I. Page. 1967. “Swords and Runes in South-East England”. The Antiquaries Journal 47, pp. 1-26. Chambers, Raymond W. (ed.). 1932. Beowulf. (Second edition, first edition 1921). Cambridge: Cambridge University Press. Chickering, Howell D., Jr. (ed.). 1977. Beowulf. New York: Anchor Books. Clanchy, Michael T. 1993. From Memory to Written Record: England 1066-1307. (Second edition, first edition 1979). London: Arnold. Cramp, Rosemary J. 1957. “Beowulf and Archaeology”. Medieval Archaeology 1, pp. 57-77. Ellis Davidson, Hilda 1962. The Sword in Anglo-Saxon England. Its Archaeology and Literature. Oxford: Clarendon Press. Falk, Hjalmar. 1914. Altnordische Waffenkunde. Skrifter utgivit av Videnskabsselskapet i Kristiania 2, Historisk-filosofisk klasse 6. Kristiania: Dybwad. Fjalldal, Magnus. 1998. The Long Arm of Coincidence: The Frustrated Connection between Beowulf and Grettis Saga. Toronto: University of Toronto Press. Garmonsway, George N. and Jacqueline Simpson. 1968. Beowulf and its Analogues. London: Dent. Green, Dennis Howard. 1994. Medieval Listening and Reading the Primary Reception of German Literature 800-1300. Cambridge: Cambridge University Press. Grünzweig, Friedrich E. 2004. Runeninschriften auf Waffen. Inschriften vom 2. Jh. nach Chr. bis ins Hochmittelalter. Wien: Edition Praesens. <?page no="214"?> Annina Seiler 196 Grünzweig, Friedrich E. 2009. Das Schwert bei den “Germanen”. Kulturgeschichtliche Studien zu seinem “Wesen” vom Altertum bis ins Hochmittelalter. Wien: Fassbaender. Gwara, Scott. 2008. Heroic Identity in the World of Beowulf. Leiden, Boston: Brill. Irving, Eward B. Jr. 1997. “Christian and Pagan Elements”. In Robert E. Bjork and John D. Niles (eds.). A Beowulf Handbook. Exeter: University of Exeter Press, pp. 175-192. Katz, Leonhard and Laurie B. Feldman. 1996. “The Influence of an Alphabetic Writing System on the Reading Process”. In Hartmut Günther and Otto Ludwig (eds.). Schrift und Schriftlichkeit. Writing and its Use. Ein interdisziplinäres Handbuch internationaler Forschung. Berlin, New York, pp. 1094- 1101. Klaeber, Friedrich (ed.). 1936. Beowulf and the Fight at Finnsburg. (Third edition, first edition 1922). Boston: Heath. Köberl, Johann. 1987. “The Magic Sword in Beowulf”. Neophilologus 71, pp. 120- 128. Koch, Peter and Wulf Oesterreicher. 1985. “Sprache der Nähe - Sprache der Distanz. Mündlichkeit und Schriftlichkeit im Spannungsfeld von Sprachtheorie und Sprachgebrauch”. Romanistisches Jahrbuch 36, pp. 15-43. Lane, Michael. 2000. “Remembrance of the Past in Beowulf”. In Geardagum 21, pp. 41-59. (Online version: www.alor.org/ Triumph%20of%20The%20Past/ Remembrance%20of%20the%20Past%20in%20Beowulf.htm [accessed on 14 February 2011]). Lerer, Seth. 1991. Literacy and Power in Anglo-Saxon Literature. Lincoln, London: University of Nebraska Press. Lerer, Seth. 1997. “Beowulf and Contemporary Critical Theory”. In Robert E. Bjork and John D. Niles (eds.). A Beowulf Handbook. Exeter: University of Exeter Press, pp. 325-339. Looijenga, Tineke. 2003. Texts and Contexts of the Oldest Runic Inscriptions. Leiden: Brill. Lorenz, Gottfried (ed.). 1984. Snorri Sturluson. Gylfaginning. Darmstadt: Wissenschaftliche Buchgesellschaft. Marold, Edith. 2004. “Die Schnalle von Pforzen und die altnordische Heldensage”. In Oskar Bandle, Jürg Glauser and Stefanie Würtz (eds.). 2004. Verschränkung der Kulturen. Der Sprach- und Literaturaustausch zwischen Skandinavien und den deutschsprachigen Ländern. Zum 65. Geburtstag von Hans-Peter Naumann. Tübingen: Francke, pp. 217-238. Mitchell, Bruce. 1985. Old English Syntax. Oxford: Clarendon Press. Near, Michael R. 1993. “Anticipating Alienation: Beowulf and the Intrusion of Literacy”. PMLA 108, pp. 320-332. Nedoma, Robert. 2004. “Noch einmal zur Runeninschrift auf der Gürtelschnalle von Pforzen”. In Hans-Peter Naumann (ed.). Alemannien und der Norden. Berlin & New York: de Gruyter, pp. 340-370. Nöth, Winfried. 1990. Handbook of Semiotics. Bloomington: Indiana University Press. <?page no="215"?> The Function of the Sword-Hilt Inscription in Beowulf 197 Oesterreicher, Wulf. 1993. “Verschriftung und Verschriftlichung im Kontext medialer und konzeptioneller Schriftlichkeit”. In Ursula Schaefer (ed.). Schriftlichkeit im frühen Mittelalter. Tübingen: Gunter Narr, pp. 267-292. Osborn, Marijane. 1978. “The Great Feud: Scriptural History and Strife in Beowulf”. PMLA 93, pp. 973-981. Raible, Wolfgang. 1991. Zur Entwicklung von Alphabetschrift-Systemen: Is fecit cui prodest. Heidelberg: Carl Winter Universitätsverlag. Rissanen, Matti. 1998. “Maþelian in Old English Poetry”. In Peter S. Baker and Nicholas Howe (eds.). Words and Works: Studies in Medieval English Language and Literature in Honour of Fred C. Robinson. Toronto: University of Toronto Press, pp. 159-172. Saenger, Paul. 1997. Space Between Words. Stanford, California: Stanford University Press. Schaefer, Ursula. 1992. Vokalität. Altenglische Dichtung zwischen Mündlichkeit und Schriftlichkeit. Tübingen: Gunter Narr Verlag. Schaefer, Ursula. (forthcoming). “Spoken and Written English - Orality and Literacy”. In Alex Bergs and Laurel Brinton (eds.). [forthcoming]. History of English. An International Handbook. Handbücher zur Sprach- und Kommunikationswissenschaft/ Handbooks of Linguistics and Communication Science. Berlin: de Gruyter. Seelow, Hubert (ed.). 1998. Die Grettis Saga. Die Saga von Grettir dem Starken. Darmstadt: Wissenschaftliche Buchgesellschaft. Slade, Benjamin. 2007. “Untydras ealle: Grendel, Cain, and V tra. Indo-European ruti and Christian sm ti in Beowulf”. In Geardagum 27, pp. 1-32. Spurkland, Terje. 2000. “‘Þeir báru fram bréf ok segja ørendi þau sem fylgðu.’ Om brevveksling i middelalderen”. In Jón Viðar Sigurðsson and Preben Meulengracht Sørensen (eds.). Den nordiske renessansen i høymiddelalderen. Oslo: Historisk institutt, Universitetet i Oslo, pp. 47-64. Spurkland, Terje. 2001. “Scandinavian Medieval Runic Inscriptions - an Interface Between Literacy and Orality? ” In John Higgitt, Katherine Forsyth and David N. Parsons (eds.). Roman, Runes and Ogham. Medieval Inscriptions in the Insular World and on the Continent. Donington: Shaun Tyas, pp. 121-128. Swinford, Dean. 2002. “Form and Representation in Beowulf and Grettis Saga”. Neophilologus 86, pp. 613-620. Vries, Jan de. 1962. Altnordisches etymologisches Wörterbuch. Leiden: Brill. Webster, Leslie. 1998. “Archaeology and Beowulf”. In Bruce Mitchell and Fred C. Robinson (eds.). Beowulf. An Edition with Relevant Shorter Texts. Oxford: Blackwell, pp. 183-194. <?page no="217"?> Gunnel Tottie On the History of try with Verbal Complements Abstract 1 The verb try is unique in English in that it can be followed by an infinitive complement with either to or and, as in I must try to finish/ I must try and finish. It has been suggested that the use of try and-constructions is a consequence of the so-called horror aequi principle, i.e. avoidance of repetition as in I want to try to finish. However, although horror aequi certainly operates as a stylistic device - especially in written English - it cannot be the impetus for the appearance of try and-complementation in the history of English. Research on historical corpora and dictionaries shows that try and + VERB is the older construction, and that the emergence of try to + VERB is a concomitant of the semantic change of try from ‘sift, select’ to ‘attempt’. 1 Introduction There are a number of verbs in English that can be used as conative verbs, or verbs of trying, corresponding to German versuchen or French essayer. The following list, containing 13 different verbs, is taken from the much longer diachronic inventory in Visser (1969: 1336): assay, attempt, endeavor, essay, make, offer, seek, strive, struggle, study, sweat, try, undertake Of these, only a few are at all common today: attempt, endeavor, seek, strive and try, with try at the top of the frequency list, by far surpassing the others. 2 1 I am deeply indebted to Nils-Lennart Johannesson and Sebastian Hoffmann for helping me find and interpret Middle English and Early Modern English examples, and for reading and commenting on an earlier version of this paper. Sebastian Hoffmann also provided the necessary skills for extracting and handling data from Early English Books Online (see further Tottie and Hoffmann, in preparation), and we thank Andrew Hardie for determining approximate word-counts for individual time periods in the data. I have also benefited greatly from ideas and suggestions from members of the lively English linguistics seminar at Stockholm University. The responsibility for any remaining inadvertencies is mine alone. 2 A search in the British National Corpus (BNC) yields 88.47 instances per million words of the base form try to, 39.65 instances pmw of try and, compared with probably less than 10 instances of the verb attempt to. (This search is only an approximation because attempt in attempt to is often a noun.) <?page no="218"?> Gunnel Tottie 200 Try is an unusual verb in English because it can take two different types of infinitival complements, either with to, as in (1) or with and, as in (2) - cf. e.g. Biber et al. (1999: 738): (1) I try to speak Swahili but it’s really hard (2) I try and speak Swahili but it’s really hard Many writers and scholars have claimed that there are meaning differences between try and + infinitive and try to + infinitive, e.g. that try and expresses “greater urgency or determination” (Wood 1965: 241). But there is strong evidence that such differences are non-existent or at least negligible - see the discussion in Hommerberg & Tottie (2007). Whether you use the to-infinitive or the construction with and (called pseudo-coordination by Quirk et al. 1985: 978), the meaning remains the same. 3 Notice that variation between try to and try and can only occur with the base form try: thus (3) and (5) are grammatical but (4) and (6) with inflected forms of try are not: (3) I tried to speak Swahili but it was really hard (4) *I tried and speak Swahili but it was really hard (5) I was trying to speak Swahili but it was really hard (6) *I was trying and speak Swahili but it was really hard Prescriptive grammarians, especially in the United States (e.g. Crews et al. 1989), have often condemned the try and variant, but in most recent British-oriented usage handbooks, (e.g. Peters 2004) it is regarded as fully acceptable (see further Hommerberg & Tottie 2007). Present-day use of the two constructions is described by Hommerberg & Tottie (2007) on the basis of 25 million words from two corpora, CobuildDirect and the Longman Spoken American Corpus English (LSAC). We show that surprisingly, the overall use of the verb try is greater in British English than in American English - cf. the findings of Vosberg (2006: 225f.), who shows that try was also much more frequent in early British fiction than in contemporary American fiction. We provide further evidence supporting Biber et al.’s finding (1999: 738) that British speakers 3 A third construction also exists, with a present participle as complement, as in (i): (i) I try speaking Swahili but he doesn’t understand (i) does not mean the same thing as the other two - in (1) and (2) you don’t actually succeed in speaking Swahili, but in (i) you do - but it doesn’t work, because the interlocutor does not understand. Try also sometimes appears without an infinitive marker, as in (ii): (ii) …try touch the tips of upper lashes only with the wand. (Kjellmer 2000: 118) This construction is rare and will not be treated here. See Kjellmer (2000). <?page no="219"?> On the History of try with Verbal Complements 201 are more likely to use try and while Americans tend to use try to, and we also show that there are differences between spoken and written language, as appears from Fig. 1. Figure 1. Try and and try to in speech and writing in British and American English. Base forms only. From Hommerberg & Tottie (2007). But why do we have this unique variation in complementation with either to or and plus infinitive after try, and how did it arise? Rohdenburg (2003) suggests that try and results from a horror aequi effect, i.e. that speakers and writers wish to avoid sequences that would contain a repetition of to. This effect would lead to a rejection of (7), and a preference for the version in (8): (7) I want to try to help him (8) I want to try and help him Rohdenburg applies the principle to a very limited dataset consisting only of cases where try is followed by indirect interrogative clauses, but Vosberg (2006), in a study based on 18 th and 19 th fiction, makes the claim that horror aequi is the overall historical impetus of try and-constructions: 4 4 Vosberg also claims that formal identity accounts for the spread of try and from infiniteve to imperative or subjunctive. “Während formale (lautliche) Identität (wie bereits gesagt) eine Voraussetzung für diese durch and erzeugte Pseudokoordinierung darstellt…” (2006: 230.) <?page no="220"?> Gunnel Tottie 202 Dabei stellt das Syntagma to try den Ursprung und damit die für die Ausbreitung der Struktur and + Verbstamm günstigste Umgebung dar (und and try die ungünstigste). (2006: 230f.) ‘The [preceding] syntagm to try is the original locus and thus the most favorable environment for the spread of the structure and + base form of verb (and and try the least favorable one).’ However, Hommerberg & Tottie (2007) show that although horror aequi does have an effect in avoidance of to…to sequences, it operates chiefly in the written language, as appears from Fig. 2. Figure 2. The distribution of try and+complement after to and in other environments in spoken and written British and American English. After Hommerberg & Tottie (2007). In written British English there is a difference of 29 percentage points between instances that follow to and those that don’t; try and appears in 47 percent of all cases following to but only in 18 percent in other contexts. In spoken British English, where try and is already dominant, it only increases by seven percentage points from 77 percent to 84 percent. In American English, there is an increase from 24 percent to 34 percent in speech and from four percent to 13 percent in writing. A multivariate analysis of the same data carried out by Gries (2010) also shows a strong effect in written American English. Considering the fact that the horror aequi-effect operates chiefly in writing, it seems unlikely that horror aequi would be the major factor behind the appearance of and-complements in the history of try, as changes originating in the written language - changes from above - are usually less frequent than changes originating in everyday speech. <?page no="221"?> On the History of try with Verbal Complements 203 Based on this observation, I set out to test the hypothesis that try and is the older construction. In this paper I will discuss the variation between try to and try and-constructions from a diachronic perspective, focusing on the early history and semantic development of the verb as well as its complementation. I shall argue that Vosberg (2006) puts the cart before the horse, and I will show that try and+infinitive is the older construction, appearing first in contexts that facilitate semantic change of try to mean ‘attempt.’ I will also argue that this is the most likely origin of present-day try andconstructions, and that they provided a handy resource for the language after try to-constructions had become current, activating the horror aequieffect in suitable contexts. To do this it will first be necessary to survey the history of try as a verb in English. I will begin by examining the semantic field of ‘trying,’ what we may call the conative field, from a historical perspective, looking at the different lexical items that have served conative functions in the history of the language, in section 2. Against this backdrop I will then trace the history of the use of the verb try and the competition between toand andcomplements in section 3 by considering dictionary evidence and corpus data. Section 4 will be devoted to a summary and discussion. 2 The conative semantic field before try The conative semantic field is one that has been in constant flux throughout the history of English. Lexical items with conative meanings have come and gone over the centuries, as appears from the list of 62 different items in Visser (1969: 1336). As he puts it, “[h]ere a conspicuous feature is the extremely large number of verbs belonging in this group, as well as the large number of them that are no longer patterned with an infinitiveobject.” As a French loanword, try is a relative latecomer in the English language - Old English made use of a number of other verbs, such as cunnian ‘attempt’, fundian ‘strive’, higian ‘strive’, underfon ‘attempt’, earnian ‘earn, strive’, tilian ‘strive’, and hyhtan ‘trust, hope’. These could all be used with to-complements as well as clausal complements. According to Los (2005: 298) the origin of the to-infinitive is what she calls “the purposive to-infinitive”, which expresses “a goal […] in time, i.e. a future, or at least non-actuated event.” This then became the prevailing construction in Middle English and still is the most common type of non-finite verbal complement in English (Mair 1990: 101f.). A couple of illustrative examples quoted from Los (2005: 48) and Visser (1969: 1338), respectively, are given in (9) and (10). All italics here and in other examples are added for expository purposes. <?page no="222"?> Gunnel Tottie 204 (9) Se ðonne se ðe fundige wislice to sprecanne, ondraede he him suiðlice…. ‘He then who strives to talk wisely should be wary…’ (CP 15.93) (10) Pharaoh …sohte Moises to ofsleanne ‘Pharaoh …tried to murder Moses’ (Ælfric, Exodus 2,15) Some of the OE conative verbs, like e.g. cunnian survived into Middle English, as can be seen from example (11) from Ormulum, written some time between 1150 and 1180. Orm also used fandenn, as in (12): 5 (11) Þe deofell lët te laferrd seon. Þiss middellærdess riche. Forr þatt he wollde cunnenn swa. To brinngenn inn hiss herrte. E‹o›rþlike þingess lufe & lusst. (12135-39; Holt 1878, vol. II: 67) ‘The devil let the Lord see the realms of this world, because he wanted to try thus to bring into His heart the love and desire for earthly things.’ (12) He fandeþþ þa to lacchenn þe. Þurrh trapp off modi3nesse. (12300-01; Holt 1878, vol. II: 73) ‘He then tries to catch you with the trap of pride.’ Many other verbs were used to mean ‘attempt’ in ME, including French loanwords such as enforce, study, aspie, procure and prove. There was clearly a need for words that could be used to express conativity, but most of these lexical items were short-lived in their sense ‘attempt’. The winner was another French loanword, try. 3 Try as a conative verb in the history of English Try comes from Old French trier, ‘to sift or pick out’. (The Middle English Dictionary - MED - also lists the forms OF triier and AF triher, treier.) Trier is still used in French with the same meaning, and can still only be 5 The verb that Orm uses is fandenn < OE fandian ‘to tempt,’ which is not a development of OE fundian. The verb usually means ‘to tempt sb.’ in Ormulum too, but occasionally it is used with a to-infinitive to mean ‘try’ as in (12). <?page no="223"?> On the History of try with Verbal Complements 205 used with nominal complements. Cf. (13), overheard in a French village store, where the shopkeeper observed a customer inspecting a basket of rather doubtful-looking potatoes: 6 (13) Il faut trier et trouver, Madame. ‘You need to pick (out) and find, Ma’am.’ Try is recorded in English with similar meanings from the 14 th century. (14) is the earliest example quoted in OED, with the gloss “To separate (one thing) from another or others; to set apart; to distinguish. Often with out. Obs. or arch.” (14) c1330 R. Mannyng Chron. Wace (Rolls) 13260 Þey turnde ageyn, And tryde þe Bretons fro ilk Romeyn. ‘They turned back and separated the Britons from the Romans’ [literally ‘each Roman’]. OED then lists a number of meanings developed from this first one, notably • ‘separate (metal) from the ore […] by melting; to refine, purify by fire’ (sense 3) • ‘examine and determine (a cause or question) judicially’ (sense 6) • ‘test the strength, goodness, value, truth, or other quality of; to put to the proof, test, prove’ (sense 7) The sense ‘make an effort, endeavour, attempt’ is only listed under 16.a, with an example from Dryden’s translation of Virgil’s Georgica from 1697, and then one from Gray’s translation of Propertius dated only as a1771, given here as (15) and (16). 7 Notice that the infinitival complement is indeed a to-infinitive, but also that it is preposed in both examples: (15) 1697 Dryden tr. Virgil Georgics iii, in tr. Virgil Wks. 107 To repair his Strength he tries: Hardning his Limbs with painful Exercise. (16) a1771 T. Gray Imit. Propertius in Wks. (1884) I. 154 While to retain the envious Lawn she tries. Try with and-complementation is only listed after try to, under 16.b, as follows: “Followed by and and a co-ordinated verb (instead of to with inf.) expressing the action attempted. colloq.” Notice that try and is labeled 6 The noun triage retains this meaning in English and is nowadays used to mean ‘sorting according to quality’, and especially according to degrees of urgency of wounds or illness in hospitals and on battlefields. Cf. OED and Merriam-Webster’s Online Dictionary. 7 Another example dates from 1638, but it is a nominalized use: “[implied in: T. Herbert Some Yeares Trav. (rev. ed.) 72 Ecbar is poysoned; ‥after foureteene dayes violent torment and trialls to expell the poyson, yeelds up his ghost. [at trial n.1 8]” <?page no="224"?> Gunnel Tottie 206 colloquial, and that it predates the Dryden example in (16) by a few years, as appears from (17): (17) 1686 J. S. Hist. Monastical Convent. 9 They try and express their love to God by their thankfulness to him. The OED examples square well with Visser’s statement (1969: 1336) that try ‘attempt’ did not “come into general use before the second half of the seventeenth century.” However, MED antedates the use of try to mean ‘attempt’ by several centuries, quoting (18) - (20), listed out of chronological order here for the purpose of the discussion - as early examples of try ‘attempt’: (18) c1350 (a1333) Shoreham Poems (Add 17376) 46/ 1290: Nou ich habbe of þe ferste yteld, Þat oþer wyl ich trye. (19) c1475 Wisd. (Folg V.a.354) 861: Ther ys no craft but we may trye yt. But notice that both (18) and (19) have nominal objects, not verbal complements, and that ‘put to the proof, test’ would probably be better glosses than ‘attempt.’ (20) is a much better candidate for both the meaning ‘attempt’ and complementation with to + infinitive, at least at first glance, and it is tempting to gloss it as I have done under (20): (20) a1400(a1325) Cursor (Vsp A.3) 15019: Þe riche men […] tempird tresun for to tri [Göt: trei] To tak iesum wit wogh. ‘The rich men plotted treason to try To take Jesus wrongfully.’ The quotation from Cursor Mundi is taken from the Palm Sunday episode where Jesus rides into Jerusalem and children and poor people strew palms and clothing before him, while the powers that be withdraw and conspire to kill him. The Latin text and a fuller quotation from Cursor Mundi are given in (21) and (22): (21) tunc congregati sunt principes sacerdotum et seniores populi in atrium principis sacerdotum qui dicebatur Caiaphas 4 et consilium fecerunt ut Iesum dolo tenerent et occiderent (Mt 26: 3-4) (22) þe lauerdinges and þe riche men To-quils o-bak þam drogh, And temprid tresun for to tri To tak iesum wit wogh. (15017-20; Morris 1874-93: 860) According to MED, temprid means ‘plotted’ here (s.v. tempren 10 c), and the quotation would then mean ‘The lords and the rich men withdrew […] and plotted to try treason [in order] to take Jesus unrightfully.’ In this interpretation, trien would not have a verbal complement, but a nominal <?page no="225"?> On the History of try with Verbal Complements 207 one like (18) and (19), and the meaning would then be that of try ‘test, put to the proof’ rather than ‘attempt to’. Another potential early example of try to can be found in the Chadwyck- Healey English Poetry Collection (CHP), given as (23) below. 8 It could be even earlier than (20) above, but is listed only as poem by an anonymous author between 1100 and 1400: (23) XXXII. Hou a man schal lyue parfytly. Seuen medicines for þe seuen dedly synnes. 683 Afftur þis he wol vs teche, 684 God þat is vr goode leche, 685 Þis seuen medicines, trie to vs, 686 To helen vs of seuen Maledius, Again, it is tempting to gloss this as follows: ‘After this, he wants to teach us, God who is our good doctor/ savior To try to use these seven remedies To heal us of seven maladies’ But the spelling vs in line 685 gives us pause. Is it the verb use that is intended, or the personal pronoun us, as in lines 683 and 686? OED does not include vs among the 33 spellings that it lists under use. And what about the rhyme in line 686 - would a presumably long vowel in vs rhyme with Maledius? We cannot be absolutely sure that the gloss should not be something like ‘try these seven medicines on us’ but admittedly, the meaning of the preposition to is a problem here. So far, we have not found a single absolutely certain instance where try to or try and + VERB means ‘attempt’ in OED, MED, or CHP. But by using the entire OED as a database as described in e.g. Hoffmann (2004) we can find more promising examples such as those in (24) - (25) and (27) - (34). (24) is actually listed under the headword try, but under sense 12, where the sense is listed as “[t]o endeavour, to ascertain by experiment or effort; to attempt to find out” occurring “a. with simple obj. (usually fortune, luck, or the like).” However, the meaning in (24) comes close to ‘attempt to,’ and I would argue that this is precisely the kind of polysemous context that invites a reanalysis of the original co-ordination as a case of complementation, with and grammaticalizing to an infinitive marker. (24) 1573 J. Baret Aluearie F 955 I will aduenture, or trie and seeke my fortune 8 An easy way to access it is also “Poems from the Vernon MS”, the full text of which can be found at: http: / / xtf.lib.virginia.edu/ xtf/ view? docId=chadwyck_ep/ uvaGenText/ tei/ chep_1.2223.xml <?page no="226"?> Gunnel Tottie 208 The example in (25) is even more convincing, as trie cannot be construed as having a nominal object: (25) 1589 A. Fleming tr. Virgil’s Georgiks i. 10 Thrise did they trie and giue assay vpon mount Pelius… The meaning ‘attempt’ becomes clear if we look at the Latin original shown in (26). The evil giants attempt to scale heaven by placing one high mountain upon another, but Jupiter thwarts them with lightning. The original provides proof: sunt conati definitely means ‘attempted’: (26) Ter sunt conati imponere Pelio Ossam Scilicet, atque Ossae frondosum involvere Olympum Ter Pater extructos disjecit fulmine montes. (VIRG. Geor. i., ver. 281.) The earliest instance of try to is (27). Again, we need context to interpret the example, and it is given in (28) from CHP, with some punctuation added: (27) 1573 T. Tusser Fiue Hundreth Points Good Husbandry (new ed.) f. 24v Dank linge forgot will quickly rot.‥ Here learne & trie to turne it and drye. (28) Dank ling, forgot, will quickly rot. Here learne, and trie, to turne it, and drie. According to OED the meaning of trie here is probably “to make experiment; †in quot. 1573 ? to practise” (sense 11c), so that the best gloss would be ‘Learn here, and practice to turn it and dry it.’ Arguing that this is a context where to could have grammaticalized, with semantic bleaching of to from ‘in order to’ to a mere infinitive marker would still require a prior shift in meaning of try to mean ‘attempt’. From the middle of the 17 th century we are on firmer ground with examples (29) and (30), which both antedate the ones under the headword try quoted above as (15) and (16): 9 (29) 1651 W. Cartwright Ordinary i. v. in Hazl. Dodsley XII. 231 Let us try To win that old eremite thing. 9 There is an earlier example in OED, from 1633, in Ford, Love’s Sacrifice III. ii, (i) My lord you were best to try to set a[t] maw. [a cardgame] But the reading is uncertain here; Chadwyck-Healey has an amended line: (ii) Bian[ca]. My Lord, you were best to try a set at Maw; I and your friend, to passe away the time, Will vndertake your Highnesse and your sister Duke. The game’s too tedious. <?page no="227"?> On the History of try with Verbal Complements 209 (30) 1664 Butler Hud. ii. ii. 483 But since no reason can confute ye, I’ll try to force you to your Duty. Inflected forms of try to are definite proof of its use to mean ‘attempt to’ - cf. (31) - (34): (31) 1665 Pepys Diary 15 Oct. Tried to compose a duo of counter point. (32) 1684 Lond. Gaz. No. 1898/ 4 Some Deal men have tried to go off to her in one of our Yaules. (33) 1655 W. Gurnall Chr. in Arm. i. 111 Joseph’s mistresse first tries to draw him to gratifie her lust… (34) 1658 Cleveland Clev. Vind. (1677) 28 In whom Dame Nature tries To throw less than Aums Ace [ambsace, i.e. two aces] upon two Dice. Notice that the examples with and-complements in (24) and (25) predate those with to-complements. This supports the hypothesis that the use of try to mean ‘attempt’ originated in co-ordinated and-constructions, and that and-complements were the precursors of to-complements. Quantitative corpus support for this can be found by making a search of the data in Early English Books Online, EEBO. 10 This collection consists of 25,277 texts totaling over 593 million tokens, covering the period from the beginnings of printing in England, c. 1475, until 1700. The results of a search for the collocations try to and try and appear in Table 1, which shows the frequency per million words of try toand try and-constructions over 50-year periods from 1500, and which is summarized in Figure 3. 11 Try and starts from 1.33 instances per million words (pmw) in the early 16 th century, peaks at 1.98 at the end of the 16 th century and then drops to 1.14 around 1700, whereas try to only appears in the second half of the 16 th century with a frequency of 0.21 pmw, then rises in the second half of the 17 th century to 1.95 a century later, and 7.86 around 1700. 10 According to their website, “this incomparable collection now contains about 100,000 of over 125,000 titles listed in Pollard & Redgrave’s Short-Title Catalogue (1475-1640) and Wing’s Short-Title Catalogue (1641-1700) and their revised editions, as well as the Thomason Tracts (1640-1661) collection and the Early English Books Tract Supplement.” About a quarter of all texts are accessible via a full-text search. 11 Examples of try and and try to were extracted by an automated search, and then irrelevant instances were removed manually. Where samples were large, the percentage of irrelevant examples was calculated based on random samples of 100 instances and that proportion was then deducted from the totals. Numbers arrived at in this way are marked with asterisks. <?page no="228"?> Gunnel Tottie 210 Decade N words (approx.) try and pmw try to pmw 1500 - 1549 18,800,000 25 1.33 0 0 1550 - 1599 87,500,000 173 1.98 18 0.21 1600 - 1649 169,100,000 330* 1.95 94 0.56 1650 - 1699 311,000,000 506* 1.63 608* 1.95 1700+ 7,000,000 8 1.14 55 7.86 Table 1. Collocations with try and and try to in EEBO listed by half-centuries Fig 3. Frequency per million words per half-centuries of try and and try to in EEBO. It must be immediately pointed out that Table 1 and Figure 3 are preliminary approximations of the parallel development of the two alternative tryconstructions. Especially the scant data from 1700 and the graph generated on the basis of them must be taken with a large grain of salt. The raw data also need to be “tried and fined” - thus for instance, text types and contexts need to be taken into account. Furthermore, as pointed out by Gries and Hilpert (2008), this type of organization of data is based on preconceived notions and should be replaced by corpus-driven periodization. Nevertheless, the figures shown here strongly suggest that in the period between 1500 and 1700, there was a rise and fall in try and-constructions and a rapid rise of try to-constructions co-occurring with the fall of try and, <?page no="229"?> On the History of try with Verbal Complements 211 with a break-even point in the latter half of the 17 th century. An in-depth analysis of EEBO data is beyond the scope of the present paper but is undertaken in Tottie & Hoffmann (in preparation). 4 Summary and discussion I have shown above that it is possible to predate the OED examples of try ‘attempt’ listed under the headword try (s.v. 16) by almost a century for try and - cf. (17) and (25) - and by nearly a half-century for try to - cf. (15) and (29) - using examples that can be demonstrated to be true conatives. The examples are repeated here for convenience: (17) 1686 J. S. Hist. Monastical Convent. 9 They try and express their love to God by their thankfulness to him. (25) 1589 A. Fleming tr. Virgil’s Georgiks i. 10 Thrise did they trie and giue assay (OED as corpus) (15) 1697 Dryden tr. Virgil Georgics iii, in tr. Virgil Wks. 107 To repair his Strength he tries: Hardning his Limbs with painful Exercise. (OED headword) (29) 1651 W. Cartwright Ordinary i. v. in Hazl. Dodsley XII. 231 Let us try To win that old eremite thing. I have also demonstrated that the meaning ‘attempt’ is likely to have arisen in potentially polysemous contexts, like (24), also repeated here for convenience. (24) 1573 J. Baret Aluearie F 955 I will aduenture, or trie and seeke my fortune ‘try my fortune and seek it’ > ‘attempt to seek my fortune’ The proportions of try andand try to-constructions in the early part of EEBO also support the assumption that the meaning ‘attempt’ in try tocollocations only appeared after it had already evolved in try and-contexts. Thus most of the 18 earliest recorded try to-constructions do not appear until after 1580, whereas there are almost 200 instances of try and in the 16 th century, as appears from Table 1 above. Several factors must have contributed to the rise of try to + infinitive constructions: • the use of try ‘attempt’ before to ‘in order to’ - the same process that took place in OE and ME when to turned into an infinitive marker <?page no="230"?> Gunnel Tottie 212 • analogy with the many other verbs taking to+infinitive complements • the fact that this construction was much more versatile and could be used with inflected forms of try as in He tries/ tried to speak, He is trying to speak Try and survives as the preferred form in spoken British English and as an alternative to try to in written British English. Try and has a more precarious existence in American English, where it either never really caught on or was killed by the denunciations of prescriptivists. The history of try in American English remains to be written, but try and still appears, and it is favored by horror aequi contexts. The bottom line is that try and did not originate as a consequence of horror aequi concerns. As I have shown, try and predates try to, and it was a handy pre-existing resource rather than a new invention. What has been called pseudo-coordination originated as true co-ordination. Biographical sketch After completing her Ph. D. in English linguistics at Stockholm University, Gunnel Tottie taught there and at the universities of Lund and Uppsala before being appointed to a professorship at the University of Zurich in 1991. She is a committed corpus linguist and has published works dealing especially with negation, relativization and American English. After retiring from Zurich in 2002, she continues to work on issues in syntax and pragmatics such as the use of tag questions and planning signals (a.k.a. hesitation markers). The present paper takes historical syntax as a point of departure for a foray into lexical semantics. <?page no="231"?> On the History of try with Verbal Complements 213 References Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan. 1999. Longman Grammar of Spoken and Written English. Harlow: Longman. The Chadwyck-Healey Literature Collections: English Poetry: http: / / collections.chadwyck.co.uk/ home/ home_ep.jsp. Accessed Oct. 2010. Crews, Frederick C., Sandra Schor, and Michael Hennessy. 1989. The Borzoi Handbook for Writers. New York: McGraw Hill. Early English Books Online: http: / / eebo.chadwyck.com/ home. Accessed Oct. 2010. Gries, Stefan Th. 2010. “Methodological Skills in Corpus Linguistics: a Polemic and Some Pointers towards Quantitative Methods”. In Tony Harris and María Moreno Jaén (eds.). Corpus Linguistics in Language Teaching. Frankfurt a.M.: Peter Lang. 121-146. Gries, Stefan Th. and Martin Hilpert. 2008. “The Identification of Stages in Diachronic Data: Variability-based Neighbor Clustering”. Corpora 3, pp. 59-81. Hoffmann, Sebastian. 2004. “Using the OED Quotations Database as a Corpus - a Linguistic Appraisal”. ICAME Journal 28, pp. 17-30. Holt, Robert (ed.). 1878. The Ormulum. With the Notes and Glossary of Dr. R. M. White. 2 vols. Oxford: The Clarendon Press. Hommerberg, Charlotte and Gunnel Tottie. 2007. “Try to or try and? ” ICAME Journal 31, pp. 43-62. Kjellmer, Göran. 2000. “Auxiliary Marginalities: The Case of try”. In John M. Kirk (ed.). Corpora Galore: Analyses and Techniques in Describing English: Papers from the Nineteenth International Conference on English Language Research on Computerised Corpora (ICAME 1998). Amsterdam: Rodopi, pp. 115- 124. Los, Bettelou. 2005. The Rise of the To-infinitive. Oxford: Oxford University Press. Mair, Christian. 1990. Infinitival Complement Clauses in English. Cambridge: Cambridge University Press. Merriam-Wester Online: http: / / www.merriam-webster.com/ dictionary/ triage, accessed Oct. 2010. The Middle English Dictionary: http: / / quod.lib.umich.edu/ m/ med/ , accessed Oct. 2010. Morris, Richard (ed.). 1874-93. Cursor Mundi. The Cursur of the World. A Northumbrian Poem of the XIVth Century: In Four Versions. Available online at http: / / xtf.lib.virginia.edu/ xtf/ view? docId=chadwyck_ep/ uvaGenText/ tei/ chep_ 1.1404.xml; chunk.id=d3; toc.id=d4; brand=default. The Oxford English Dictionary: http: / / www.oed.com/ , accessed Dec. 2010. Peters, Pam. 2004. The Cambridge Guide to English Usage. Cambridge: Cambridge University Press. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman. Rohdenburg, Günter. 2003. “Cognitive Complexity and horror aequi as Factors Determining the Use of Interrogative Clause Linkers in English”. In Günter Rohdenburg and Britta Mondorf (eds). Determinants of Grammatical Variation. Berlin: Mouton de Gruyter, pp. 205-249. <?page no="232"?> Gunnel Tottie 214 Tottie, Gunnel and Sebastian Hoffmann. In preparation. “Try to or try and? A Chicken-and-egg Story”. Visser, Fredericus Theodorus. 1969. An Historical Syntax of the English Language. Vol. 3: 1. Leiden: E.J. Brill. Vosberg, Uwe. 2006. Die grosse Komplementverschiebung. Aussersemantische Einflüsse auf die Entwicklung satzwertiger Ergänzungen im Neuenglischen. Tübingen: Gunter Narr. Wood, Frederick T. 1965 (1962). Current English Usage. A Concise Dictionary. London: Macmillan. <?page no="233"?> Richard J. Watts The Black Death and the Development of English A Tale of Two Archives Abstract One contentious point concerning the history of any language is the degree to which language change is conditioned by the socio-cultural, socio-political and socioeconomic contexts within which varieties of that language are used through time. In the histories of most major European languages one cataclysmic event is conveniently relegated to the realm of communal amnesia: the Black Death from 1347-1351. I argue that one of the reasons for this relegation is the convenient “explanation” for that devastating pandemic of plague derived from a medical, epidemiological discourse archive in which the plague is taken to be bubonic plague. This assumption has been discursively challenged in recent years, leading to a new set of perspectives on language change in the 14 th and 15 th centuries which considers forms of language contact in a world of population decline and internal migration to merit increased attention on the part of historical sociolinguists. Such a change of perspective involves more intensive interdisciplinary research with economic, cultural and urban historians and also, and in particular, with demographers. The contribution is thus of an exploratory nature. 1 Introduction In honouring Andreas Fischer on the occasion of his 65th birthday, we honour not only the man but also the academic. Although it is not so obvious in looking through the long list of his publications, one of his interests has always been “historical sociolinguistics”. Unlike its two close relatives variationist sociolinguistics and interactional sociolinguistics, historical sociolinguistics is constrained to forge research alliances with cultural history, historical demography, and economic history (among other historical disciplines that focus on the activities of social groups). As a sociolinguist myself, I have a distinct leaning towards the historicity of the disciplines of historical linguistics and sociolinguistics, and I have found it expedient to resort to the methods of discourse analysis in opening up what is commonly taken to be “commonsense” beliefs within those discourses. I understand the term “discourse” in the sense that Michel Foucault has defined it, and it is this understanding that will guide the structure of the present contribution to the Festschrift. I intend to focus on one strand each <?page no="234"?> Richard J. Watts 216 from two different kinds of discourse, the first of these being a medical/ epidemiological discourse, the second a discourse on language change and language standardisation. My aim will be to suggest that these discursive strands form part of what Foucault (1972) calls an “archive” (see also Blommaert 2005 and Watts 2011) that needs to be challenged in order to open up new research paths. I shall start in section 2 with a (necessarily brief) outline of how Foucault understands the two terms “discourse” and “archive”. In sections 3 and 4 I shall go on to outline the two discursive strands that I wish to challenge, the first of these being the interpretation of the plague that caused the Black Death from 1347 to 1351 as a pandemic of bubonic plague, the second being the common tendency to equate the history of English with the history of the standard language. In section 5 I argue that the crucial two centuries in the relatively rapid change from Middle English to Early Modern English from 1300 to 1500, might be approached more fruitfully by carrying out interdisciplinary research in conjunction with disciplines such as cultural history, historical demography and economic history in an effort to discover and interpret new kinds of language data which might help us to overcome the “funnel view” of the history of English (cf. section 4 and chapter 12 in Watts 2011). 2 Foucault’s understanding of discourse Foucault understands a discourse to be a body of statements (i.e. a subset of statements) that belong to a single system in the overall formation of statements, and he takes statements to be historically situated events. As such, the body of statements (the discourse) is markedly distinct from other systems of statements. Human interaction always takes place within discursive formations, the result being that the statements in the discourse are ultimately taken to represent true states of affairs, “true” not in the sense of logically true in a coherent logical system, but rather in the sense of a system of shared “beliefs”. The process of “coming to accept” a system of statements can be thought of as part of the larger cognitive process of socialising individuals into socio-cultural groups, and it is a process which frequently takes place below the level of consciousness. Foucault stresses that the system of statements constituting a discourse appears to the socio-cultural group to be coherent and to be characterised by its continuity over time. But he counters this apparent coherence and continuity by arguing that the fundamental characteristic of discursive statements is their discontinuity. A “commonsense” discursive unity is in fact a dispersion of elements displaying discontinuity, i.e. groups of statements may occur in any order, with any function and be correlated in any way with other groups of statements, since all of the objects, forms and <?page no="235"?> The Black Death and the Development of English 217 themes of discursive statements are historically linked to the external conditions of discourse production. In the Archaeology of Knowledge (1972), Foucault takes knowledge to be constructed historically through the production of discourses which can only be pieced together from the surviving textual and non-textual evidence. The discursive evidence “unearthed” displays some of the discontinuities and lack of coherence in discourses, and it may even lead to a sudden shift in the way in which a particular topic is constructed discursively (as we shall see in the following section). It is within this perspective on discourse that Foucault introduces the term “archive”, which he defines as “[t]he general system of the formation and transformation of statements” (1972: 127) or, alternatively, “the law of what can be said, the system that governs the appearance of statements as unique events” (1972: 129). 1 He also uses the term “positivity of a discourse”, by which he means “that which characterizes its particular unity throughout a particular discursive time.” The “positivity” of a discourse is thus to be found in the archive to which the discourse belongs, to “the law of what can be said.” It is the archive that determines how certain statements can be grouped together to form an apparent unity and how certain statements appear to us as historical events. 2 In the following section I shall describe and challenge one strand of a medical/ epidemiological discourse which is taken as “true”, i.e. that the Black Death was a pandemic of bubonic plague. It needs to be challenged for the simple reason that accepting this explanation encourages historians to consign the pandemic to a black box of knowledge in which no further questions need be asked, and in which some uncomfortable questions receive discursive solutions that do not fit all the facts. For the sociolinguist, a fresh look at the Black Death and its truly horrendous effects on population growth and internal migration within England after 1350 until the last two decades of the 15 th century might also create openings to ponder on the linguistic effects of contact between different forms of dialect in a situation in which there was as yet no overarching, regulating standard language. 1 I shall henceforth use the term “discourse archive” in order to avoid misunderstandings, e.g. that an archive is equivalent to a library or a stored collection of documents. 2 Blommaert (2005: 102) suggests that the archive consists of “the macro-sociological forces and formations that define and determine what can be said, expressed, heard, and understood in particular societies, particular milieux, particular historical periods.” <?page no="236"?> Richard J. Watts 218 3 The Black Death as part of a discourse archive The term “Black Death” refers to the devastating outbreak of plague that first appeared in Sicily in 1347 and swept through the whole of Europe (with the puzzling exception of a few small geographical areas), reaching as far as Iceland and Greenland by 1351. The expression is thought to have been derived from the colour of the so-called tokens of the disease (peticchiae), “large subcutaneous spots, often red and able to change colour between orange and black and between blue and purple” (Scott & Duncan 2001: 380), which appeared largely on the chest but also on the throat, arms and legs of the victim. These, rather than the buboes which, frequently but not always, formed under the armpits, in the groin or on the neck 3 were taken as sure signs (tokens) of impending death. The disease appeared to contemporaries to come from nowhere and to be fatal in the vast majority of cases, which gave rise to a number of stories to account for or explain its appearance. It is significant that one of these, concerning the provenance of the pandemic, has been absorbed into a medical discourse archive on the causes of plague, i.e. that it was an outbreak of bubonic plague. The explanation of the Black Death as a pandemic of bubonic plague can be seen as a strand in a scientific discourse archive, “the law of what can be said” about the plague that struck Europe in 1347. But when looked at in more detail (see Scott & Duncan 2001), the explanation appears to lack coherence and even to become implausible. Scott & Duncan (2001: 7) define bubonic plague as follows: [I]t is a bacterial disease of rodents depending on the rat flea for its spread and on a reservoir of resistant rodent species for the maintenance of the disease. Only occasionally does the disease spread to humans when one is bitten by a rat flea but, in the days before antibiotics and modern medicine, this was usually fatal and serious epidemics could be established. At a later point in their book Scott & Duncan describe the symptoms of bubonic plague and the course of the disease: The bubo is the characteristic symptom of bubonic plague; it is a lump formed by a swollen lymph node and is of variable size. It is found most commonly in the groin but its location depends on where the flea bites […] The bubo appears early in the illness, on the first or the second day […] Usually the bubo is very painful and tender and, in patients who live long enough or survive, it 3 Buboes are a swelling of the lymph-nodes and are certainly not unique to bubonic plague. They also occur in cases of gonorrhea, tuberculosis, or syphilis. <?page no="237"?> The Black Death and the Development of English 219 breaks down and discharges pus. […] The incubation period 4 is usually 2-6 days after exposure. (2001: 67) Scott & Duncan have looked at this part of the “story” of the Black Death as a pandemic of bubonic plague in scrupulous detail, and have come to the conclusion that it flies in the face of the evidence we have both from contemporary 14 th century sources and also from more modern research on the biology and epidemiology of pandemics. What, therefore, are the facts? As we have seen, the incubation period for bubonic plague is between 2 and 6 days, the end being almost invariably the death of the sufferer. The story of the provenance of the plague is that it was carried in ships from the port of Caffa on the Sea of Azov to Sicily, but the story is invalidated by the fact that the journey by boat in the first half of the 14 th century would have taken a minimum of five weeks or a maximum of three to four months to complete. The story even has it that the sailors came ashore in Sicily in apparently perfect health. The construction of a medical discourse archive on the nature of the plague began only at the turn of the 20 th century when a serious outbreak of bubonic plague in India and the Far East spurred micro-biologists, at the request of the Plague Commission of India, to investigate into and unravel the causes. A French micro-biologist by the name of Alexandre Yersin was successful in discovering the complex biology and epidemiology of the disease. Scott & Duncan explain the discovery as follows: Yersin demonstrated conclusively that bubonic plague is a disease of rodents, spread by one rodent to another by fleas. He showed that the course of an epidemic in humans is dependent on many factors because of the multiple hosts involved, and that it is completely different from an epidemic of a simple infection, which is spread from person to person. Yersin demonstrated that the infectious agent that is transmitted from rat to flea to human is a bacterium, which was named Yersinia pestis - Yersinia after the man who discovered it and pestis after the disease (the pestilence) that it was thought to have caused. (2005: 167) Bubonic plague was revealed to be a complex disease set off in humans by the bacterium yersinia pestis when transmitted by the bite of a rodent flea (xenopsylla cheopsis), whose stomach was blocked by the bacteria. The blocked stomach caused the blood drawn from biting a human to flow back into the incision together with a number of the bacteria. The reason for the flea xenopsylla cheopsis 5 transferring to humans as a source of food was the extinction of the host population through infection by yersinia 4 The incubation period is composed of the latent phase from the moment of infection on and the infectious stage, during which no signs of the disease are in evidence but the patient can pass the infection on to others. 5 The flea’s natural host was rodent species such as tarabagans on the Asian steppes and black rats in hot coastal areas. <?page no="238"?> Richard J. Watts 220 pestis. The flea could no longer get a blood meal nor could it ingest the blood. Hence it, too, was doomed to die, although not before having passed on the bacterium. The bacterium survives for the simple reason that certain species of rodent, off which the flea feeds, are immune to infection. Bubonic plague, therefore, was (and still is) an erratic disease that only struck human beings by chance. It was presaged by a mass mortality of rats or other rodents (e.g. gerbils, tarabagans). As we have seen, its incubation period was roughly 2-6 days before the symptoms appeared, and it spread relatively slowly and inconsistently through a human population. But the discovery of yersinia pestis offered a logical, scientific explanation for the plagues in England from the Black Death to the end of the 17 th century. It gave humankind clues as to how to avoid outbreaks of bubonic plague and how to treat patients presenting the relevant symptoms medically. Anyone who henceforth wrote about the plague pandemics in Europe in the late Middle Ages and later, whether historians, demographers, educationists, writers of popular history or others, simply assumed that they were outbreaks of bubonic plague. The “scientific proof” had been categorically provided, and all one needed to do when writing about plagues in England was to assume an overpopulation of rats, an abundance of rodent fleas, 6 a secondary rodent population that was immune, an appalling level of human hygiene, and the stories could be told. One book after another has been written about the Black Death in the last 30 years, and, with very few notable exceptions, they have all assumed the cause to have been bubonic plague. 7 We have here a classical case of a Foucaultian discourse archive waiting for heretical voices in the wilderness to spread the kinds of “explicit challenges to accepted ways of thinking and believing” that would constitute an anti-discourse. Yersin’s discovery of the plague bacterium was a brilliant, groundbreaking and painstaking piece of scientific detective work which has enriched our knowledge of plague pandemics immensely. But this does not alter the fact that much of what has been written from a less academic perspective has embellished the story to such a degree that the Black Death has attained mythical proportions going beyond the data and the explanatory theories set up by Yersin and others. 6 Nota bene: xenopsylla cheopsis and not the common-or-garden human flea (pulex irritans), since these were very rarely indeed associated with the ingestion of yersinia pestis. 7 I mention a brief selection of such works here, e.g., Ziegler (1965); Gottfried (1978); Horrox (1994); Kelly (2006); Hatcher (2009). The notable exceptions are Cantor (2001), who suggested that the implausible cause was anthrax in cattle, Shrewsbury (1970), who, despite recognising the paradoxes in assuming bubonic plague to have been responsible, consistently places the casualty figures from plague at as low a figure as 5% of the population and passes the buck to other diseases such as typhus, and the work by Graham Twigg and Scott & Duncan, to which I refer in this paper. <?page no="239"?> The Black Death and the Development of English 221 There is a wealth of evidence in Horrox (1994) 8 that those places which instituted a quarantine period for the disease soon settled at a figure of 40 days, despite the period of only a few days during which the victim visibly suffered from the symptoms before dying. Obviously, people must have had a rough and ready knowledge of the aetiology and pathology of the disease that was sufficient to determine the length of the quarantine period. There was no medical remedy to prevent an outbreak of plague and no cure once a victim had contracted it. Stories of the plague - where it had come from, how one could recognise it, what one could reasonably do to avoid infection, how it affected whole families and broke down the fabric of social life, what had caused it, and so on - preceded the pandemic in its relentless and by no means erratic South to North trajectory across Europe. The nature of these stories is admirably reconstructed in Hatcher (2009), in which he attempts convincingly, on the basis of the contemporary written documents available, 9 to create a “literary docudrama” rather than a conventional history of one such community, the village of Walsham le Willows in Suffolk in the period from 1345 to 1350. When the Black Death struck, it hit a naïve population with no experience of this kind of extensive pandemic. In England, repeated epidemics from the time of the Black Death in 1348-9 right through to the third quarter of the 17 th century exposed people in different parts of the country to the horrors of death by plague, until it mysteriously disappeared from Britain in the last third of the 17 th century. But by this time the stories had solidified and become a dominant discourse till the new medical discourse archive emerged at the end of the 19 th century. The first heretical voice in the wilderness to challenge the archive was that of Shrewsbury (1970), who, while admitting that some of the plague attacks could have been bubonic plague, recognised flaws in assuming uncritically that the causative infective element in the Black Death was bubonic plague. Instead he proposed other diseases such as typhus, dysentery, smallpox, tuberculosis, to explain the mortality that undoubtedly occurred. 8 Rosemary Horrox (1994) has provided a useful collection of firsthand reactions to the pandemic taken from various authors across Europe at the time of or shortly after the Black Death, and it is clear from reading these texts that the general belief was that the disease was transferred from person to person. 9 Local court records, notices of deaths, marriages and illegitimate births, church records and records pertaining to supralocal and national issues. In his preface Hatcher admits that there are no personal records of what rural life was like at the time of the Black Death, but he posits that “[i]n order to move from the general to the particular, and construct the lives of some of the individuals who lived in Walsham, our narrative will have to be extended beyond the fragmentary and often prosaic facts contained in surviving documents. In other words, this book will have to contain speculation as well as specifics, fiction as well as fact” (2009: xv). <?page no="240"?> Richard J. Watts 222 The work of Graham Twigg (1978, 1984, 1989, 1993), who has unflinchingly challenged the explanation that the Black Death was a pandemic of bubonic plague, is exemplary scholarship of the highest order. His work has been continued by Scott & Duncan (1998, 2001, 2005), who have worked on data from the fields of epidemiology, social demography, statistical modelling and biology, and have provided a convincing alternative scenario, which deserves more attention than it currently receives. The arguments presented by Twigg and Scott & Duncan indicate that much of what is put forward in support of the story of bubonic plague as the cause of the Black Death is a product of a new, medical 20 th century discourse archive, notwithstanding the undoubted excellence and accuracy of Yersin’s discovery of the bacterium yersinia pestis. The first challenge concerned the dependence of the bacterium on the black rat (rattus rattus), which was imported to England by ships sailing from the Continent and is generally restricted to urban areas associated with ports and harbours. As an adept climber, it makes its nest in the roofs of houses and does not expand its territory much beyond those areas. It may sometimes have been transported into rural areas, but it would have had much more difficulty in creating an environmental niche for itself there. Twigg has researched extensively on the black rat and has collected a wealth of as yet unpublished information on the extent of its spread to rural areas during the period of the Black Death. One small but intriguing piece of evidence is that farmers must have experienced little to no damage done to grain stored in corn ricks at ground-level before the migration of the brown rat (rattus norvegicus) to England in the 1720s. Twigg’s evidence is the emergence of mushroom-shaped straddles on which the cornricks were stored from the 18 th century on, the brown rat not being a particularly good climber. In addition - and again in the early 18 th century - dovecotes had to be redesigned to prevent the brown rat from stealing the eggs. If the black rat had been numerous in rural areas of England, measures such as these would have been necessary as early as the 14 th century, since, as we saw, the black rat is an excellent climber. The second challenge in Scott & Duncan’s theory is that there are no other species of rodent in England which are resistant to yersinia pestis. The third challenge concerns the fact that, since the range of the black rat does not extend to Norway, Iceland and Greenland, but the plague did reach those outlying northern areas of Europe, it could not possibly have been bubonic plague. The fourth challenge concerns the 40-day quarantine period agreed upon by the city states in Northern Italy during the Black Death and <?page no="241"?> The Black Death and the Development of English 223 applied throughout Europe in the following centuries. 10 Scott’s detailed study of parish registers in England from the second half of the 16 th and the 17 th centuries reveals that the plague that struck Europe during the Black Death and returned regularly through a 300-year period had, on average, a latent period after infection of 10-12 days, an infectious period of 20-22 days, and a period of 5 days during which the symptoms became evident. The total infectious period was thus 25-27 days and the total time from infection to death was, on average, 37 days. If, as all the contemporary reports indicate, the disease was spread from person to person, we can conclude the following: 1. An infected person could walk around and mingle with others for a period of roughly 3 weeks without anyone knowing, least of all him/ herself, that s/ he was infected at all. 2. The disease could spread extremely rapidly, much more rapidly than could bubonic plague. 3. Settlements situated on transportation routes (roads, tracks, rivers, the sea) were more open to infectives plying their wares and offering their labour, but all settlements connected in some way to the surrounding countryside were open to possible infection. 4. For this reason, far more people could be infected than is ever the case with bubonic plague, which is evidence that the disease was extremely effective and not at all erratic. 5. There were no climatic or seasonal limitations on the infection being spread. 6. There were no ways of preventing its spread apart from self-imposed quarantine periods (of 40 days! ). What was this form of plague? Scott & Duncan (2001) suggest that it might have been a virus infection, similar to Ebola, that caused haemorrhaging in the vascular system by entry into the white blood corpuscles, including the lymph nodes, which explains why several patients had buboes, others displayed only the tokens (peticchiae) and still others a combination of both. Scott & Duncan call this form of plague haemorrhagic plague, and although there is no longer any means of discovering the virus that caused it, their arguments against bubonic plague as the cause of the Black Death and the string of epidemics that followed on from it during the next 300 years are so convincing as to challenge the discourse archive of the Black Death as a huge pandemic of bubonic plague. 10 Cf. Scott & Duncan’s description of the self-imposed quarantine in the village of Eyam, Derbyshire to counteract the spread of plague in 1665-1666 (2001: chap. 10). <?page no="242"?> Richard J. Watts 224 For their pains in deconstructing the explanation of the Black Death as bubonic plague - which in no way invalidates the story of Alexandre Yersin’s discovery - they earn disrespectful comments from the writers of populist versions of the story. As an example, consider the following quotation taken from John Kelly’s The Great Mortality: An Intimate History of the Black Death, the Most Devastating Plague of all Time: The Deniers’ [i.e. those who deny that it was bubonic plague, RJW] case against Y. pestis can be reduced to two basic points. The first and weaker part of their argument involves their pet theories about what else could have caused the medieval plague. Anthrax, zoologist Twigg’s candidate, has never struck a human community in epidemic form and does not produce buboes (although anthrax victims do develop black boils). Professor Cohn’s candidate 11 is a mysterious Disease X, which he does not name, but believes has probably gone extinct. Scott and Duncan go furthest of all, arguing improbably that many of the worst epidemics in Western history, from the fifth century B.C. Plague of Athens to the Black Death, were caused by an Ebola-like illness they call hemorrhagic [sic] plague. (2006: 296) Those who wish to support conventional wisdom feel free, as Kelly does here, to use negatively evaluative expressions (e.g. “pet theories”) in reference to the results of serious, interdisciplinary, long-term academic research. Calling the researchers “Deniers” with a capital D reveals disdain for those who choose systematically and scientifically to challenge the discourse archive. Kelly has obviously not taken notice of the fact that viruses can disappear without trace, as easily as they can mutate consistently, so why should the disease not have “gone extinct”. He has also not realised from all the descriptions given in Horrox (1994) that, whatever the causal agent of the Black Death was, it was perceived as a disease spread in close contact between people - fleas or no fleas. Scott & Duncan themselves respond to Kelly’s comment on their speculation that the causal agent may have been an “Ebola-like illness”: We do not suggest that haemorrhagic plague in Europe during 1348-1670 was because of Ebola or any of the present-day viral haemorrhagic fevers that have been identified, but the close similarities with their symptoms and pathology described above suggest that a filovirus may have been the causative agent. (2001: 389) Challenging a discourse archive, even with the most meticulous data one can amass from the evidence available and with faultless methodology and argumentation, hardly earns praise from the scientific community. A discourse archive is “the law of what can be said, the system that governs the appearance of statements as unique events” and is universally accepted as 11 Samuel K. Cohn. 2003. The Black Death Transformed: Disease and Culture in Early Renaissance Europe. New York: Oxford University Press. <?page no="243"?> The Black Death and the Development of English 225 the “truth”, and it tends to become the property of “everyman”. The common contemporary reaction is to brand those who challenge orthodox beliefs as “Deniers”. People during the plague years conceptualised life as having the qualities of a gamble, a burden, or a prelude to paradise/ hell, so it stands to reason that they would translate these conceptualisations into narratives to explain, justify and ratify their behaviour under the constant threat of plague. As soon as bubonic plague was given a rational explanation through Yersin’s discovery, the need for the narratives that were developed on the basis of such negative conceptualisations of life, at least with respect to the threat of plague, seemed to disappear. Plague was no longer a threat. It no longer imposed the burden of insecurity on human beings, and it no longer turned the course of life into a lottery in which one could more easily lose than win. Without sticking meticulously to the matter-of-fact details of the biology and epidemiology of bubonic plague, historians have claimed the right to be able to provide a fully “rational” explanation of the events by assuming that the Black Death was a pandemic of bubonic plague. But those who went through the horror of that pandemic and a succession of more local but no less lethal outbreaks in its wake instinctively knew that the disease, whatever it was, was transmitted through close human contact, i.e. in situations of emergent social practice. It was brought by strangers moving into the community or by community members returning to the community from outside. I turn now to an important strand of a discourse archive governing “the law of what can be said, the system that governs the appearance of statements as unique events” on the history of English. 4 The funnel view of the history of English My major concern in Watts (2011) is to unearth and deconstruct a complex set of myths that lie at the basis of a powerful dominant discourse on the history of English that has now taken on the status of a discourse archive. The “commonsense” statements that form a naturalised discourse determining what can be said about the history of English constitute a canonical way of presenting the “facts” of language change in English. First, one starts by presenting “Old English”, a term which I assiduously try to avoid using in Watts (2011), as the precursor of the modern language and then to move through the familiar periodisation of Middle English, Early Modern English and Late Modern English. The end-product of the process is almost invariably a focus on the modern standard language, <?page no="244"?> Richard J. Watts 226 and the starting point a range of varieties of Anglo-Saxon. 12 The perspective on the history of the language is thus rather like a funnel, in which a number of varieties are poured in at the wide top of the funnel and standard English comes out of the narrow neck (cf. figure 1). The fate of the original varieties poured in at the top and others which may have arisen at a later stage are generally not taken into consideration: Figure 1. The funnel view of the history of English (Watts 2011: chap. 12) The funnel view constitutes a modern archive of “what can be said, expressed, heard, and understood” about the history of English. The contributions to Watts & Trudgill (2002) argue for a wider range of histories to cover other varieties than the standard, and one way to challenge the archive is to focus more consciously on a range of alternative histories. But how can challenges to these two archives be brought to bear on possible explanations for the changes undergone by varieties of English from before the Black Death to the end of the 15 th century? What might the effects of the Black Death be, conceptualised as haemorrhagic plague that could be transferred unwittingly from one person to another in close social contact, on changes in the varieties of the language? One thing at least can be said, viz. that haemorrhagic plague could be transferred 12 Even in the period of “Old English” there is a tendency to assume that the literary West Saxon variety of the 10 th -11 th centuries was a “West Saxon” standard language, and this is all too often generalised to cover the whole area in which Anglo-Saxon was spoken. <?page no="245"?> The Black Death and the Development of English 227 between two people as could language in instantiations of emergent social practice - although to posit a cause and effect relationship between plague and language change on that basis is somewhat absurd. Nevertheless, from a sociocognitive point of view, human language as a cognitive faculty emerges from countless instantiations of social practice between individuals through time. What I intend to argue in the following section is that language change, for more than a century following the Black Death, should be contextualised against the background of an astonishing population decline. The discourse archive governing the emergence of standard English, the funnel view, focuses almost exclusively on London and migration to London. While not wishing to deny that London, a powerful socioeconomic magnet, attracted speakers from other dialect areas, 13 there were other, socioeconomically less significant but linguistically potentially interesting centres in which language contact involving processes of speech accommodation and new dialect formation through koïneisation must have taken place. 5 Social and economic effects of the Black Death in England I begin this section by stressing a set of demographic facts that simply cannot be ignored in evaluating the economic, political, cultural and linguistic changes that took place in England from 1350 to 1500. At the beginning of the 18 th century the population of England, Wales and Scotland together was not much more than 6,000,000, in 1801 it was roughly 10,500,000 and in 1851 it had increased to roughly 20,800,000. In the space of 150 years it had trebled, and within the space of 100 years between 1700 and 1800 it had almost doubled. In the period immediately prior to the appearance of the Black Death in England in 1348, however, the population of England and Wales alone (without considering Scotland) is estimated to have been somewhere between 4,500,000 and 6,000,000 (see Hatcher 1996, who puts the figure closer to the upper limit) and the estimated post-Black Death figure lies between 2,500,000 and 3,000,000. If we subtract the estimated figures for Scotland in 1700 (ca. 1,300,000), we arrive at the astounding fact that the population of England and Wales in 1700 (roughly 4,700,000) had barely reached what it had been prior to the Black Death in 1348. It may even have been smaller if, like Hatcher, we take the upper limit of c. 6,000,000. The population figures had taken a full 350 years to recover. 14 The major reason for this 350 year gap 13 In particular from the Central and East Midland areas of England. 14 Information accessed on 05.08.09 from http: / / www.tacitus.nu/ historical-atlas/ population/ british.htm. <?page no="246"?> Richard J. Watts 228 between the end of the Black Death in England and Wales in 1349 and the end of the 17 th century was the fact that the plague (and other lethal epidemics) regularly, and almost without respite, struck England and Wales either on a local or a national scale until 1667, and there are even sporadic reports of plague as late as 1679. In all, there were 33 nation-wide plague epidemics between 1361 and 1665, 15 not counting many more on a local level. Plague was virtually endemic in England during the late 16 th century and through two thirds of the 17 th century with an outbreak, large or small, in one location or another on an annual basis. 16 Raw data of this kind become interesting if we link up population growth with factors such as economic growth, population movement, causes of mortality, and birth rate. Hatcher (1996) estimates an overall loss of population in 1348-49 of between 35 to 45% (although it may well have been closer to 50%), and argues that the recurrent visitations throughout the rest of the 14 th and most of the 15 th century continually led to population loss rather than growth. He quotes other historical demographers and economic historians as arguing that the country was feeling the economic pressure of overpopulation, even at the assumed level of between 4,700,000 and 6,000,000. This was mainly due to the rigid feudal system in operation prior to 1348, which hampered the movement beyond the demesne of those of the rural population who were dependent on the institutions and authority of that demesne and which effectively prevented increased agrarian production. The closer we move to the maximum figure, the more serious the economic situation would have been. The Black Death effectively cut the population by almost 50%, but rather than effect the overall situation adversely, the sudden demographic decrease actually stimulated economic growth from 1349 till towards the end of the 14 th century. At one fell swoop there was no shortage of land, and although manorial lords attempted to use their authority under the feudal system to retain their tenant farmers at the rents payable prior to the Black Death, those attempts were, in the long run, doomed to failure. To begin with, it became increasingly difficult to find the rightful “heirs” to tenanted landholdings that had become vacant through the wholesale death of nuclear families farming the land, and even if an heir could be found it was virtually impossible to force her/ him to enter that inheritance. Second, labourers realised very quickly that their labour had become a valuable resource, and this led to wholesale internal migration of farm workers off their traditional demesnes to look for work where labour was needed. 15 Cf. information accessed on 05.08.09 at http: / / urbanrim.org.uk/ plague%20list.htm. 16 Although Scott & Duncan (2001) argue that, rather than being endemic to England, plague was continually renewed by means of new infectives from the continent of Europe. <?page no="247"?> The Black Death and the Development of English 229 Wages and prices increased as the cost of land decreased. The initial stimulation of the economy, leading to the growth of new industries and an increase in the size of certain small villages to the status of town-sized industrial settlements 17 effectively led to a new distribution of wealth and consumption patterns across the country - all of this contingent on internal migration and the breakdown of the feudal system. From a linguistic point of view, such wholesale changes in social and economic structure must inevitably have led to less local isolation of speakers of dialectal varieties of English and to a greater propensity for dialect contact situations of various kinds leading to koïneisation processes and new dialect formation. But the downside to the prosperity that could be sustained by a reduced population was further population loss resulting from repeated national and local epidemics of plague. Consistently high wages, rising prices and the falling price of land led to an economic depression in the 15 th century that was only overcome when population figures finally began to rise in the last twenty years of that century. The relationship between the persistent decline in population figures largely as a result of repeated epidemics of plague (and other lethal diseases) and the dramatic increase in real-wage levels after 1350 in the period between 1250 and 1750 is represented in Hatcher’s figure 2, adapted here as figure 2 in the present paper: Figure 2. English population and the real wage-rates of craftsmen, 1250-1750 (from Hatcher 1996: 77, figure 2) 17 Hatcher (1996: 51) mentions Totnes and Tiverton in the south-west; Hadleigh, Maldon, Lavenham, Nayland, Long Melford, Sudbury and Coggeshall in Suffolk and north Essex (as a consequence of the thriving wool industry); Leeds, Bradford, Halifax and Wakefield in the West Riding; and Castle Combe and Stroudwater in the Cotswolds. <?page no="248"?> Richard J. Watts 230 Hence, while the short-term effects of the Black Death were positive, the long-term effects were nothing short of catastrophic. Hatcher (1996: 61) makes the point that the overall economic situation with land and food in abundance and affordable “while labour was scarce and well-rewarded” should have stimulated population growth. The fact that it did not, however, can only be put down to the level of mortality after 1377 in the wake of “at least fifteen outbreaks of plagues and/ or other epidemic diseases of national or extra-regional proportions; namely, 1379-83, 1389-93, 1400, 1405-7, 1413, 1420, 1427, 1433-34, 1438-39, 1457-58, 1463-64, 1467, 1479-80 and 1485” (1996: 63-64). He concludes from these facts that “[f]requent visitations of epidemics which not only killed substantial numbers but also left in their wake distorted ageand/ or sex-structures would severely inhibit the ability of the population to reproduce itself” (1996: 68). Clearly, migration of any kind, internal or external, will lead to new language contact situations in which koïneisation is likely to take place. This in turn will ultimately lead to new dialect formation. In the epilogue to his docudrama on Walsham le Willows in northwest Suffolk, Hatcher (2009: 320) traces from the available records signs of social unrest, dispute, and relatively large-scale migration by villeins from the village to other villages and towns. Unfortunately, we have no idea where they migrated to, but London must have attracted a large number. To reach London, however, migrants needed to walk over roads little better than muddy carttracks that quickly became impassable in rainy weather, and migration to find work in Norwich or one of the developing wool centres in south Suffolk and north Essex (Hadleigh, Maldon, Lavenham, Nayland, Long Melford, Sudbury and Coggeshall) must have been much easier. Evidence for the greater efficiency of sea and river transportation to London when compared with road transportation by horse and cart is given by Keene (2000: 96). Keene focuses his attention on the significance of London as the locus of the emerging standard in the 15 th century, but he does so from the point of view of an economic historian researching on urban centres. Nevertheless, the data that he gives support my argument that historical sociolinguists should focus on other centres than London in their efforts to trace internal migration movements across the country. Figure 3 (figure 6.1 in Keene’s article) shows the relative costs of transporting goods to London around 1300. To quote Keene: “Each isopleth line indicates the cost of carrying a quarter of wheat from places on that line to London by the cheapest means available, including costs of cartage, shipping, and handling” (2000: 96). Thus transporting goods from Cuxham in Oxfordshire would have cost 6.3 pence whereas transporting goods from Cambridge (roughly the same distance from London as the crow flies as Cuxham) via the river system and the sea would have cost 9.3 pence. <?page no="249"?> The Black Death and the Development of English 231 Walsham lies in a belt between 9.3 pence and 13.3 pence and is only marginally further from London than both Cambridge and Cuxham, which is a clear indication of the very poor state of the roads between Walsham and London. This kind of information is a pointer towards the relative difficulty in travelling from one place to another in the post-Black Death period, and is significant in choosing possible centres to which migration took place in the late 14 th and early 15 th century: Figure 3. Transport costs to London c. 1300. Adapted from Keene (2000: 96) and showing the routes from Walsham to Norwich, London and the north Essex and south Suffolk wool centres. 18 18 Plague may have helped to push up costs but it would not have affected the relative costs of transportation to London from different places. <?page no="250"?> Richard J. Watts 232 I shall give one final example (adapted from Keene 2000: 104) to indicate the kind of information that might offer clues concerning the direction of internal migration. One method for measuring the attraction of a central place to those living within its hinterland is to plot the number of debts owed to people in the central place and the residences of the debtors (see Keene 2000). Keene presents information gathered by Galloway & Murphy as part of the “Metropolitan market networks, 1300-1600” project in the late 1990s on the number of debts owed to Londoners at the turn of the 15 th century (c. 1400). 19 The data were gathered from a study of 7,806 cases from the Common Plea Rolls in the Public Record Office. What figure 4 (figure 6.6 in Keene 2000) shows is that certain minor central places had relatively strong credit relationships with the capital, indicating that they, as localised central places, must have had similar credit relationships with their own hinterlands. Although this is not direct evidence for migration tendencies, it could indicate that, although the pull of the capital city was much stronger than that of the smaller central places, they too must have attracted trade and migration from surrounding areas: Figure 4. Residences of debtors to Londoners, c. 1400 (adapted from Keene 2000: 104) 19 See at the following website: http: / / www.history.ac.uk/ projects/ metropolitan-marketnetworks. <?page no="251"?> The Black Death and the Development of English 233 6 Conclusion In conclusion, there are many other ways of creating synergies between the research activities of historical demographers and economic historians, on the one hand, and historical sociolinguists researching into the patterns of migration that might eventually have led to new dialect formation in England on the other. From the point of view of demography, economics, and socio-political and social structure, the Black Death created a historical caesura in English history that cannot and should not be ignored. While there is no ultimate proof that this was a form of plague in which a filovirus, transferred from person to person, caused “a haemorrhagic disease of high case fatality” (Scott & Duncan 2001: 385), the relentless progress of the disease across Europe between 1347 and 1351 - carried as it was by unwitting infectives along most transportation routes - the high level of fatality, the long incubation period, its occurrence in winter as well as summer and the wealth of contemporary evidence all challenge the medical archive in which the disease is taken to be bubonic plague. From a linguistic point of view, the principal focus of research should be less on the nature of the disease itself and more on the effects, lasting well over a century, on social, economic and political structure, patterns of migration, and forms of emergent social practice. Once a clearer idea of these factors is obtained, historical sociolinguists need to be able to trace the movement of people from one part of the country to another and to compare the written evidence of the late 14 th and 15 th centuries with what we know of Middle English dialects from important sources such as Kristensson’s Survey of Middle English Dialects (1988) and McIntosh, Samuels & Benskin’s A Linguistic Atlas of Late Mediaeval English (1986). We need to be able to trace out possible koïneisation processes during the 100-150 years following the Black Death. What was the linguistic situation in geographical area X prior to 1347? Who migrated into area X in the years following 1348, and where did the migrants come from? How extensive was that migration, and to what extent was it repeated after each new outbreak of plague? If, for example, the provenance of significant numbers of speakers migrating to London was from south Yorkshire and Lincolnshire, we would expect, given the numerical and social significance of those speakers in their new place of residence, to discover traces of their original dialects in documents produced in London, e.g. the use of they and their in place of hi/ he and hire as new third person plural pronouns, or lexical items deriving from Old Norse. This is indeed what we do find throughout the 15 th century. But the knowledge we have of these influences is used primarily to delineate the development of the standard language rather than to investigate into the history of new dialect formation, and it leads to the funnel view of the history of English which constitutes part of the second <?page no="252"?> Richard J. Watts 234 discourse archive dealt with in this paper. There is no direct causative relationship between outbreaks of plague and language change, but it is no exaggeration to posit a number of significant indirect influences which are well worth serious consideration in conjunction with other historical disciplines. Biographical sketch Richard J. Watts is emeritus professor of English Linguistics, now retired from the University of Bern. His major work has been in the areas of linguistic politeness, historical sociolinguistics and the standardisation of English. Among his book publications are Power in Family Discourse (1991, Mouton de Gruyter), Politeness (2003, Cambridge University Press) and Language Myths and the History of English (2011, Oxford University Press). He is editor of the journal Multilingua and co-editor (with David Britain) of the book series Language and Social Processes with De Gruyter Mouton. References Blommaert, Jan. 2005. Discourse. Cambridge: Cambridge University Press. Cantor, Norman. 2001. In the Wake of the Plague. New York: Harper Perennial. Cohn, Samuel K. 2003. The Black Death Transformed: Disease and Culture in Early Renaissance Europe. New York: Oxford University Press. Foucault, Michel. 1972. Archaeology of Knowledge. (Translated from French by Rupert Swyer. Original title: L’Archéologie du Savoir. 1969). London: Tavistock. Galloway, J. A., D. Keene and M. Murphy. 1996. “Fuelling the City Production and Distribution of Firewood and Fuel in London’s Region, 1290-1400”. Economic History Review 49, pp. 447-472. Gottfried, Robert S. 1978. Epidemic Disease in Fifteenth Century England: The Medical Response and the Demographic Consequences. Leicester: Leicester University Press. Hatcher, John. 1996. “Plague, Population and the English Economy, 1348- 1530”. (Second printing, first printing 1977). In Michael Anderson (ed.). British Population History: From the Black Death to the Present Day. Cambridge: Cambridge University Press, pp. 9-93. Hatcher, John. 2009. The Black Death: The Intimate Story of a Village in Crisis, 1345-1350. (Second edition, first edition 2008). London: Phoenix. Horrox, Rosemary. 1994. The Black Death. Manchester: Manchester University Press. <?page no="253"?> The Black Death and the Development of English 235 Keene, Derek. 2000. “Metropolitan Values: Migration, Mobility and Cultural Norms, London 1100-1700”. In Laura Wright (ed.). The Development of Standard English 1300-1800: Theories, Descriptions, Conflicts. Cambridge: Cambridge University Press, pp. 93-114. Kelly, John. 2006. The Great Mortality: An Intimate History of the Black Death, the Most Devastating Plague of all Time. New York: Harper Perennial. Kerswill, Paul. 2002. “Koïneization and Accommodation”. In Jack Chambers, Peter Trudgill and Natalie Schilling-Estes (eds.). Handbook of Language Variation and Change. Oxford: Blackwell, pp. 669-702. Kristensson, G. 1988. Survey of Middle English Dialects 1290-1350. (Second edition, first edition 1967). Lund: Lund University Press. McIntosh, Angus, M. L. Samuels & Michael Benskin. 1986. A Linguistic Atlas of Late Mediaeval English. Aberdeen: Aberdeen University Press. Scott, Susan & Christopher J. Duncan. 2001. Biology of Plagues: Evidence from Historical Populations. Cambridge: Cambridge University Press. Shrewsbury, J. F. D. 1970. A History of Bubonic Plague in the British Isles. Cambridge: Cambridge University Press. Twigg, Graham. 1978. “The Role of Rodents in Plague Transmission: A Worldwide Review”. Mammal Review 8, pp. 418-423. Twigg, Graham. 1984. The Black Death: A Biological Reappraisal. London: Batsfrod Academic. Twigg, Graham. 1989. “The Black Death in England: An Epidemiological Dilemma”. Maladie et Société, siècles XII-XVII, pp. 75-98. Twigg, Graham. 1993. “Plague in London: Spatial and Temporal Aspects of Mortality”. In C. Creighton (ed.). Epidemic Disease in London. London: University Centre for Metropolitan History, pp. 1-17. Watts, Richard J. & Peter Trudgill (eds.). 2002. Alternative Histories of English. London: Routledge. Watts, Richard J. 2011. Language Myths and the History of English. New York: Oxford University Press. Ziegler, Philip. 1965. The Black Death. London: Collins. <?page no="255"?> Narr Francke Attempto Verlag GmbH+Co. KG • Dischingerweg 5 • D-72070 Tübingen Tel. +49 (07071) 9797-0 • Fax +49 (07071) 97 97-11 • info@narr.de • www.narr.de NEUERSCHEINUNG OKTOBER 2011 JETZT BESTELLEN! Deborah L. Madsen / Mario Klarer (eds.) The Visual Culture of Modernism Swiss Papers in English Language and Literature 26 2011, 265 S. €[D] 49,00/ SFr 65,50 ISBN 978-3-8233-6673-7 The Visual Culture of Modernism offers a wide-ranging exploration of intertextual relations that bring together artists, artistic forms and artistic periods in response to the question: what is the relevance of early twentiethcentury American Modernism to our present historical moment? Scholars from Europe and America develop responses to this question based on the philosophical heritage of modernity and in the context of the range of Modernist cultural praxis. The essays collected here explore links between literary and cultural Modernism, the relationship between the concepts of modernity and Modernism, and the legacy of Modernism in the late twentieth century and the contemporary period. Cinema, cinematic paratexts, television, the visual arts of painting and photography, poetry, fiction, and drama are among the artistic forms discussed in terms of issues ranging from cinematic and stage reinterpretations of Modernist literary texts to the genre of televisual melodrama and the trope of racial passing. The essays argue that visuality remains an urgent concern, from the Modernist period to our present age of media revolution. <?page no="256"?> Narr Francke Attempto Verlag GmbH+Co. KG • Dischingerweg 5 • D-72070 Tübingen Tel. +49 (07071) 9797-0 • Fax +49 (07071) 97 97-11 • info@narr.de • www.narr.de NEUERSCHEINUNG OKTOBER 2011 JETZT BESTELLEN! Guillemette Bolens / Lukas Erne (ed.) Medieval and Early Modern Authorship Swiss Papers in English Language and Literature 25 2011, 325 Seiten €[D] 49,00/ SFr 65,50 ISBN 978-3-8233-6667-6 Reports of his death having been greatly exaggerated, the author has made a spectacular return in English studies. This is the first book devoted to medieval and early modern authorship, exploring continuities, discontinuities, and innovations in the two periods which literary histories and institutional practices too often keep apart. Canonical authors receive sustained attention (notably Chaucer, Gower, Shakespeare, Jonson, Milton, and Marvell), and so do key issues in the current scholarly debate, such as authorial self-fashioning, the fictionalisation of authorship, the posthumous construction of authorship, and the nexus of authorship and authority. Other important topics whose relations to authorship are explored include adaptation, paratext, portraiture, historiography, hagiography, theology, and the sublime. “This rich, challenging and exceptionally well conceived collection addresses the construction of authorship in medieval and early modern England, and revises received opinion in important ways. All the essays are worth attention; several should be considered essential reading.” Stephen Orgel, J. E. Reynolds Professor in the Humanities, Stanford University