Symposium on Lexicography X: Proceedings of the Tenth International Symposium on Lexicography May 4-6, 2000 at the University of Copenhagen 9783110933192, 9783484391093

Citation preview

njmTOdäiMPmoi ££

LEXICOGRAPHICA Series Maior Supplementary Volumes to the International Annual for Lexicography Supplements ä la Revue Internationale de Lexicographie Supplementbände zum Internationalen Jahrbuch für Lexikographie

Edited by Sture Allen, Pierre Corbin, Reinhard R. K. Hartmann, Franz Josef Hausmann, Ulrich Heid, Oskar Reichmann, Ladislav Zgusta 109

Published in cooperation with the Dictionary Society of North America (DSNA) and the European Association for Lexicography (EURALEX)

Symposium on Lexicography X Proceedings of the Tenth International Symposium on Lexicography May 4-6, 2000 at the University of Copenhagen Edited by Henrik Gottlieb, Jens Erik Mogensen and Arne Zettersten

Max Niemeyer Verlag Tübingen 2002

Die Deutsche Bibliothek - CIP-Einheitsaufnahme I Lexicographica / Series maiorj Lexicographica: supplementary volumes to the International annual for lexicography / publ. in cooperation with the Dictionary Society of North America (DSNA) and the European Association for Lexicography (EURALEX). Series maior. - Tübingen : Niemeyer. Früher Schriftenreihe Reihe Series maior zu: Lexicographica 109. International Symposium on Lexicography : Symposium on Lexicography Χ. 2002 International Symposium on Lexicography : Symposium on Lexicography Χ : proceedings of the Tenth International Symposium on Lexicography, May 4 - 6, 2000 at the University of Copenhagen / ed. by Henrik Gottlieb .... - Tübingen : Niemeyer, 2002 (Lexicographica : Series maior; 109) ISBN 3-484-39109-X

ISSN 0175-9264

© Max Niemeyer Verlag GmbH, Tübingen 2002 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. Printed in Germany. Gedruckt auf alterungsbeständigem Papier. Druck: Weihert-Druck GmbH, Darmstadt Einband: Nädele Verlags- und Industriebuchbinderei, Nehren

Table of Contents

Introduction English German Acknowledgements

VII XI XVII

Norman Blake

Towards a Dictionary of Shakespeare's Informal English

Wolfgang Viereck

The Atlas Linguarum Europae

1

and its Insights into the Cultural History of Europe

19

Tove Bjarneset

The Dictionary Project NORDLEXIN-N

31

Jane Bottomley

From Print to Disc: Creating the Electronic Version of the Cambridge International Dictionary of English (CIDE)

43

Ulrich Busse

Lexicography as a Sign of the Times: A Study in Socio-Lexicography

49

The Contragram Verb Valency Dictionary of Dutch, French and English

63

The Representation of Figurative Senses in Learner's Dictionaries

77

Old French Loanwords of Germanic Origin Borrowed into English

91

Timothy Colleman

Janet DeCesaris & Victöria Alsina

Bernhard Diensberg

Ken Fare

Henrik Gottlieb

Rolf Hesse

Lars Holm

Jean Hudson

Somatismen als Problem der dänischen und deutschen Lexikographie

107

Four Germanic Dictionaries of Anglicisms: When Definitions Speak Louder than Words

125

Grammatische Schwierigkeiten bei der zweisprachigen Lexikographie

145

35 Questions and Answers about Editing Jesper Swedberg's Swensk Ordabok

155

Categorizing Chaos: Text Types in CANCODE

163

VI Olga Kaipova & Svetlana Manik

Public Political Vocabulary: Model of a Dictionary

173

Ksenija Leban

Towards a Slovene-English False-Friend Dictionary

185

Anatoly Liberman

The Length and Breadth of an Entry in an Etymological Dictionary

199

Sändor Martsa

Homonymy vs Polysemy: Conversion in English

211

Geart van der Meer

Metaphors: How Do Dictionaries Scramble out of this Morass of Meaning? Japanese Learners' Problems in Using English-Japanese Dictionaries

Tadamasa Nishimura

Vilja Oja

231 243

Some Semantic Problems in the Translation of Colour Terms

253

Kurt Opitz

The Dictionary of Connotations: A Viable Proposition?

261

Yoshiaki Otani

Who Uses English-Japanese Dictionaries and When? Their Bidirectional Working

267

Gunnar Persson

From 'Spinning Woman' to 'Old Maid' to What? On the Sense Development of Spinster

273

The Colour Spectrum in Language: The Case of Czech. Cognitive Concepts, New Idioms and Lexical Meanings

285

Wlodzimierz Sobkowiak

On the Phonetics of Trans- in EFL Dictionaries

293

Andrejs Veisbergs

Euphemisms in General Monolingual Dictionaries

303

Hideki Watanabe

A Thesaurus of Old English Revisited

313

From Projection to Reception On the Process of Bilingual Dictionary Making

325

Vlra Schmiedtovä & Barbara Schmiedtovä

Arne Zettersten & Hanne Lauridsen

Introduction

The Tenth International Symposium on Lexicography at the University of Copenhagen, the proceedings of which are hereby published, took place on 4-6 May 2000. In this volume, we present 28 papers from lexicographers in 17 countries: Belgium, the Czech Republic, Denmark, Estonia, Germany, Great Britain, Hungary, Japan, Latvia, the Netherlands, Norway, Poland, Russia, Slovenia, Spain, Sweden and the USA. Two of the papers are based on plenary lectures held on the first day of the Symposium at the Royal Danish Academy of Sciences and Letters. Since 1994 the International Symposium on Lexicography has opened with the Otto Jespersen Memorial Lecture, and the year 2000 Otto Jespersen speaker, Norman Blake, presented his plans for a new Shakespeare dictionary. In his paper, Blake observes that Jespersen noted many writers' claims about Shakespeare's vocabulary, though these claims chiefly focus on the number of words he used or his inventiveness in creating new ones. In outlining the methods for collecting the appropriate material, Blake also emphasizes the fact that very few scholars have systematically explored register in Shakespeare before. In the second plenary paper from the Symposium, Wolfgang Viereck presents the "Atlas Linguarum Europae", an interpretative word atlas using both traditional methods such as onomasiology and semasiology as well as innovative ways of interpreting geolexical data, like for instance motivational mapping. With the aid of a great number of examples, Viereck demonstrates how the study of European word history offers important insight into cultural history and religion. Tove Bjerneset offers a presentation of the Norwegian NORDLEXIN-N project which has been based in Bergen since 1996. In the early 1990s, all Nordic countries were offered free use of the Swedish LEXIN database to facilitate the production of similar dictionary series. The dictionaries will hopefully become a useful tool for, among others, immigrants with limited reading proficiency and will be published in both a printed and an electronic version. Jane Bottomley discusses ways of exploiting the full potential of the electronic medium, with reference to the "Cambridge International Dictionary of English Online" (CIDE Online), a free Internet service, and the "Cambridge International Dictionary of English CD-ROM" (CIDE CD), whose development she has been involved in. Her paper covers four main areas: editing the content for an electronic format, managing the database and integrating feedback, the user interface, and, finally, the extra features outside the running text. Ulrich Busse explores the social dimension of lexicography by offering a contrastive analysis of two special types of foreign-word dictionaries. By comparing some of the most recent dictionaries of Anglicisms in German to the English hard-word dictionaries of the 17th century he demonstrates that lexicography, at least to a certain extent, has always reflected sociological and sociolinguistic phenomena. Timothy Colleman reports on the present project of the Contragram research group at the University of Ghent (Belgium), the compilation of the "Contragram Verb Valency Dictionary". It is stressed that the CVVD will be the first valency dictionary that is fully con-

VIII trastive, and that the traditional distinction between source language and target language(s) will not apply to the CVVD. As a result, the CVVD should ideally complement grammars, traditional translation dictionaries and existing valency lexicons in a number of respects. In their paper, Janet DeCesaris and Victoria Alsina discuss the representation of figurative senses in English, Spanish, and Catalan dictionaries. In English learner's dictionaries it is common not to label figurative senses as such, whereas in Spanish and Catalan lexicography there is a long-standing tradition of using a label to identify such senses. It is argued that proper use of labels can help represent coherent sense development, which constitutes vital information for language learners. Corpus data from English verb-preposition combinations support a presentation of polysemous words in learner's dictionaries that relate figurative senses to their literal counterparts. Bernhard Diensberg's paper deals with the variety of problems connected with Old French loanwords of Germanic origin in the English language. In analysing a number of these problems, Diensberg distinguishes between Old French loans of Germanic origin which have reflexes in (nearly) all Romance languages and those that are restricted to Gallo-Romance, which make up the majority of Old French loanwords with Germanic etymons. Ken Faro explores the lexicographical handling of so-called 'polysemic somatisms', a special category of polylexical phrases consisting of body sememes, which occur in actual language in both an idiomatic and a free syntagmatic variant. Fare demonstrates how these phrases are dealt with in a number of German and Danish dictionaries and concludes that this type of phrase is ambiguous in a strict sense and should therefore be monosemized in the dictionaries. In comparing four Germanic dictionaries of Anglicisms, Henrik Gottlieb warns against concluding from lexicographical evidence (the entries in the works examined) to lexicological realities (usage in Germany, Denmark, Sweden and Norway). He argues that the differences between the four dictionaries - and their internal discrepancies - mean that exact comparative statements concerning the influence of English on other Germanic languages cannot be made on the basis of the four works alone. Rolf Hesse describes the difficulties encountered in the translation of sample sentences, collocations and idiomatic expressions out of context in bilingual dictionaries, which can be attributed to grammatical differences between source and target languages. Hesse suggests how these difficulties might be dealt with, reduced or even removed. Lars Holm gives an account of his project of editing an early 18th-century dictionary manuscript, Bishop Jesper Swedberg's "Swensk Ordabok" and points out that there is no tradition on which the solution of many editorial problems can be based. He argues that none of the manuscripts written by Swedberg himself can serve as the basic manuscript, since they are all incomplete, and he offers suggestions as to how to tackle the resulting editorial difficulties. Jean Hudson discusses some methodological issues in corpus design with reference to CANCODE, a five million word corpus of naturally occurring spoken English. It is assumed that an influential variable on linguistic choice is the relationship that holds between speakers. The paper describes the categorization of the CANCODE data and offers some linguistic evidence in support of the validity of the further subcategorization of the important category that is generally classified in corpora as 'Conversation'.

IX The paper by Olga Karpova and Svetlana Manik is devoted to reflections on how processes in social and public political life lead to changes in language. They analyse different types of dictionaries and examine methods of dictionary criticism. The authors underline the necessity of a dictionary which offers objective presentation and labelling of political vocabulary. Ksenija Leban gives a detailed outline of the structure and functions of a SloveneEnglish dictionary of false friends, aimed at language professionals, a segment of dictionary users often neglected. False-friend dictionaries should help preclude errors based on the assumption that a target-language lexeme shares meaning and/or form with its sourcelanguage lookalike, a notion which - as demonstrated - is very often unfounded. The discussion in Anatoly Liberman's contribution turns on three theses: 1) An etymological dictionary should offer a detailed critical survey of the conjectures about the origin of the words included. 2) Closely related words should be treated in one entry rather than distributed according to the letters with which they begin. 3) As a general rule, the prehistory of loan words should be left out of the etymological dictionaries of borrowing languages. Sändor Martsa demonstrates that English conversions defy the traditional distinction made between polysemy and homonymy. The notions of semantic relatedness and predictability are considered, followed by the analysis of two groups of verbs representing meaning extension: animal verbs and verbs converted from instrument terms. Looking at a number of examples of both types, it becomes obvious that animal and instrument verbs are in fact word-metaphors or word-metonyms. This means that the semantic relationship between them and the corresponding parent nouns can only be interpreted as a kind of polysemy, not homonymy. Geart van der Meer addresses the treatment of conventional metaphors in the four main English learner's dictionaries. Focusing on the lexical field around the word 'morass' ('bog', 'swamp', 'quagmire', 'mire', 'quicksands', 'dregs'), it turns out that all these words behave in largely the same way, despite what the dictionaries report. This study suggests that demetaphorisation, i.e. loss of metaphorical content, should not be assumed too quickly and that we should speak of metaphorical use rather than metaphorical meaning. Tadamasa Nishimura presents the results of a study of Japanese learners' problems in connection with the use of learner's dictionaries. Nishimura claims that lexicographers have too often ignored dictionary users in Japan. He presents the feedback from a questionnaire given to high school and university students, with the conclusion and recommendation that lexicographers should use such surveys much more in the future. The two main issues in Vilja Oja's paper are, first, how to provide equivalents for colour terms in bilingual dictionaries, and second, how to define the meaning of a word in a computer database designed to provide for a multilingual dictionary. The discussion is based on examples from bilingual dictionaries, covering various languages (Estonian, English, Finnish, German, Russian) and from the Database of Estonian Colour Terms. Kurt Opitz challenges lexicographers' traditional preoccupation with ascertaining and listing the semantic charge of words as a paramount and immutable quality attributable to concepts rather than their representation in language. Such emphasis on this denotational aspect of lexemes neglects the vital role of speaker-oriented connotation so characteristic of actual language use. Consequently, a dedicated dictionary of connotations is proposed. This is outlined and illustrated by an examination of the lexeme 'garden': how it is treated in

χ conventional dictionaries, and how it could be presented for purposes of connotationsensitive decoding of text. Yoshiaki Otani draws attention to the great importance of English-Japanese dictionaries in modern Japanese society. Such bilingual dictionaries help to initiate Japanese users into the world of English while providing a frame of reference for the reception of English in the world of Japanese. In either process equivalents play a key role as interface between the two languages which are so different in syntax, concept and cultural background. In his paper, Gunnar Persson delineates the sense development of the word 'spinster' over the past 1000 years. He introduces a lexicological model based on semantic frames and a prototype analysis that does not distinguish between denotation and connotation in the same way as traditional semantic analyses. Based on a thorough analysis of the SYN2000 subcorpus in the Czech National Corpus, Virα Schmiedtovä and Barbara Schmiedtovä consider the linguistic wealth and diversity of the words for 'black', 'white' and 'red' in Czech. The authors point out that no existing dictionary is capable of embracing all the semantic and conceptual nuances of these words, and they express the hope that future dictionaries will address this issue. Wlodzimierz Sobkowiak presents a critical overview of the phonetic treatment of the English morpheme 'trans-' in seven leading EFL dictionaries. It is shown that the phonetic representation of 'trans-' lacks consistency both within and across dictionaries. Some variation can be explained by underlying linguistic factors hidden from the eyes of the learner. Some is due to the intuitive nature of the representations, often based on the individual preferences of compilers. Finally, phonetic representations in dictionaries are apparently not as thoroughly cross-checked as other data, which leads to further variation, this time completely haphazard. Andrejs Veisbergs considers euphemisms and their treatment in general monolingual dictionaries. He offers a linguistic analysis of euphemism formation (based on English material) and the treatment of euphemisms in several English and Latvian monolingual dictionaries. Serious inconsistencies in euphemism treatment (editorial decisions, subjective solutions in individual cases, absence of or inconsistency in the use of labels, etc.) as well as problems in representing the transitory nature of euphemisms in dictionaries are revealed. Hideki Watanabe observes in his paper that "A Thesaurus of Old English", published in 1995, has received less attention than it deserves. It is important both as the first period thesaurus of its kind and as a pilot study for the publication of "The Historical Thesaurus of English" which is to appear in a few years' time. Watanabe incorporates the points discussed by reviewers and appraises it once again, with the acknowledgements properly due to the most important contribution to Old English word study in the last decade of the twentieth century. Watanabe concludes his paper by proposing some revisions, particularly with a view to compounds, set phrases and idiomatic expressions. Arne Zettersten and Hanne Lauridsen's paper deals with the preparatory process of their new English-Danish dictionary and some aspects of the reception of the dictionary after its publication. Special attention is paid to the way the editors made use of a professional assessment company in deciding on priorities in the planning. After publication, the editors learnt from users that this bilingual dictionary is to be regarded as a production dictionary as well as a reception dictionary.

Vorwort

Das Zehnte Internationale Symposion zur Lexikographie, dessen Akten hiermit veröffentlicht werden, fand in der Zeit vom 4. bis 6. Mai 2000 an der Universität Kopenhagen statt. Wir freuen uns darüber, in diesem Band 28 Beiträge von Lexikographen aus 17 Staaten vorstellen zu können: Belgien, Dänemark, Deutschland, Estland, Großbritannien, Japan, Lettland, den Niederlanden, Norwegen, Polen, Russland, Slowenien, Schweden, Spanien, Tschechien, Ungarn und den USA. Zwei der Beiträge basieren auf Plenarvorträgen, die am ersten Tag des Symposions an der Akademie der Wissenschaften zu Kopenhagen gehalten wurden. Seit 1994 wird das Internationale Symposion zur Lexikographie durch die "Otto Jespersen Gedenkvorlesung" eröffnet. Der Otto Jespersen-Redner des Jahres 2000, Norman Blake, stellte seine Pläne für ein neues Shakespeare-Wörterbuch vor. Blake bemerkt u.a., dass Jespersen die Kommentare vieler Autoren zum Wortschatz Shakespeares notierte, obwohl diese hauptsächlich aus Aufzählungen der von Shakespeare benutzten Wörter sowie seiner lexikalischen Neuerungen bestehen. Ausführlich beschreibt Blake die Methoden bei der Einsammlung des einschlägigen Materials und betont dabei den Umstand, dass die sprachlichen Register bei Shakespeare von der bisherigen Forschung nur in ganz wenigen Fällen systematisch berücksichtigt wurden. Im zweiten Plenarvortrag des Symposions stellt Wolfgang Viereck den "Atlas Linguarum Europae" vor, einen interpretierenden Wortatlas, der sich bei der Interpretation geolexikalischer Daten sowohl traditioneller Methoden wie Onomasiologie und Semasiologie als auch innovativer Ansätze bedient, bei denen es vorrangig um den motivationeilen Aspekt geht. Anhand einer Vielzahl von Beispielen weist Viereck nach, dass das Studium der Wortgeschichte Europas mit wichtigen Einsichten in Kulturgeschichte und Religion verbunden ist. Tove Bjorneset stellt das norwegische Projekt NORDLEXIN-N vor, welches seit 1996 in Bergen läuft. Anfang der 90er Jahre wurde allen nordischen Ländern die schwedische LEXIN-Datenbank im Hinblick auf die Herstellung ähnlicher Wörterbuchreihen frei zur Verfügung gestellt. Mit diesen Wörterbüchern, die in sowohl einer gedruckten als auch einer elektronischen Version erscheinen werden, soll ein u.a. für Immigranten brauchbares Werkzeug erstellt werden. Jane Bottomley diskutiert verschiedene Möglichkeiten zur vollen Ausnutzung des elektronischen Mediums, indem sie insbesondere auf das "Cambridge International Dictionary of English Online" (CIDE Online), einen freien Internetdienst, sowie das "Cambridge International Dictionary of English CD-ROM" (CIDE CD) eingeht, an dessen Entwicklung sie selbst beteiligt war. In ihrem Beitrag werden vier Hauptgebiete behandelt: 1. die Redaktion des Inhalts für ein elektronisches Format, 2. die Datenbankpflege bzw. die Integration des Feedbacks, 3. die Benutzerschnittstelle und 4. die Organisation zusätzlicher Merkmale außerhalb des laufenden Textes. Ulrich Busse thematisiert die soziale Dimension der Lexikographie, indem er eine kontrastive Analyse zweier spezieller Typen von Fremdwörterbüchern unternimmt. Durch den Vergleich einiger der neuesten Anglizismenwörterbücher des Deutschen mit den englischen Wörterbüchern des 17. Jahrhunderts zu "schweren Wörtern" kann er nachweisen, dass die

XII Lexikographie - zumindest bis zu einem gewissen Grade - schon immer soziologische und soziolinguistische Erscheinungen reflektierte. Timothy Colleman berichtet über die Herstellung des "Contragram-Wörterbuches zur Verbvalenz" (CVVD), das gegenwärtige Projekt der Contragram-Forschungsgruppe der Universität Gent (Belgien). Er betont, dass das CVVD als erstes Valenzwörterbuch vollständig kontrastiv sein wird, und dass die traditionelle Unterscheidung zwischen Quellensprache und Zielsprache(n) für ein solches Wörterbuch nicht relevant ist. Im Endergebnis will das CVVD Grammatiken, traditionelle Übersetzungswörterbücher und existierende Valenzwörterbücher in mehrfacher Hinsicht ergänzen. Janet DeCesaris und Victoria Aisina diskutieren in ihrem Beitrag die Darstellung figurativer Bedeutungen in englischen, spanischen und katalanischen Wörterbüchern. In englischen Lernerwörterbüchern ist es üblich, figurative Bedeutungen nicht eigens zu markieren, wohingegen in der spanischen und katalanischen Lexikographie eine lange Tradition für die Markierung solcher Bedeutungen besteht. Es wird die Auffassung vertreten, dass ein reflektierter Gebrauch von Markierungen dazu beitragen kann, die kohärente Bedeutungsentwicklung zu verdeutlichen, anhand derer entscheidende Auskünfte für die Lerner einer Sprache konstituiert werden. Durch Korpusdaten aus dem Englischen werden Vorschläge zur Darstellung polysemer Wörter in Lernerwörterbüchern unterbreitet, u.a. im Hinblick darauf, wie die figurativen Bedeutungen zu ihren litteralen Gegenübern in Beziehung gebracht werden können. Der Beitrag von Bernhard Diensberg behandelt die mit den altfranzösischen Lehnwörtern germanischen Ursprungs in der englischen Sprache verbundene Vielzahl von Problemen. Bei der Analyse einer Reihe dieser Probleme unterscheidet Diensberg zwischen altfranzösischen Entlehnungen germanischen Ursprungs, die in (fast) allen romanischen Sprachen Reflexe haben, und solchen, die auf das Gallo-Romanische begrenzt sind, d.h. die Mehrzahl der altfranzösischen Lehnwörter auf germanischem Substrat. Ken Faro untersucht die lexikographische Berücksichtigung so genannter 'polysemer Somatismen', eine besondere Kategorie polylexikalischer Syntagmen, die aus körperbezogenen Sememen bestehen, und die im aktuellen Sprachgebrauch sowohl in einer idiomatischen als auch in einer freien syntagmatischen Variante vorkommen. Fare weist nach, inwieweit solche Syntagmen in einer Reihe deutscher und dänischer Wörterbücher behandelt werden und stellt fest, dass dieser Syntagmentyp im strikten Sinne ambig ist und daher in den Wörterbüchern monosemiert werden müsste. Aufgrund eines Vergleiches vier germanischer Anglizismenwörterbücher warnt Henrik Gottlieb vor der Gefahr, die darin bestünde, die lexikographischen Befunde (d.h. die Eintragungen in den untersuchten Wörterbüchern) unkritisch auf die lexikologischen Realitäten zu applizieren (aktueller Sprachgebrauch in Deutschland, Dänemark, Schweden und Norwegen). Er betont, dass die Feststellung von Unterschieden zwischen den vier Wörterbüchern - und deren internen Diskrepanzen - zur Folge hat, dass genaue Angaben zum Einfluss des Englischen auf andere germanische Sprachen auf der Basis der vier behandelten Wörterbücher nicht möglich sind. Rolf Hesse beschreibt die Schwierigkeiten, die bei der Übersetzung von Beispielsätzen, Kollokationen und idiomatischen Wendungen außer Kontext auftreten, und die auf grammatische Unterschiede zwischen Quellen- und Zielsprache(n) zurückgeführt werden können. Hesse unterbreitet eine Reihe von Vorschlägen zur Lösung bzw. Reduktion oder Entfernung dieser Probleme.

XIII

Lars Holm berichtet über den Fortgang seines Projektes zur Edition eines frühen Wörterbuchmanuskriptes aus dem 18. Jahrhundert, und zwar das "Swensk Ordabok" des Bischofes Jesper Swedberg. Holm weist auf den Umstand hin, dass keine Traditionen bestehen, auf deren Folie man editorische Fragestellungen lösen könnte. Er vertritt die Auffassung, dass keines der von Swedborg selbst verfassten Manuskripte als Editionsgrundlage dienen kann, da sie alle unvollständig sind. Abschließend schlägt er Wege zur Lösung der daraus entstehenden editorischen Schwierigkeiten vor. Jean Hudson diskutiert einige mit der Erstellung von Korpora verbundene methodologische Fragen, indem sie sich auf CANCODE bezieht, ein aus fünf Millionen Wörtern bestehendes Korpus des tatsächlich gesprochenen Englisch. Dabei wird vermutet, dass die Beziehung zwischen den Sprechern eine wichtige Variable in sprachlichen Wahlsituationen ausmacht. Anschließend wird die Kategorisierung der CANCODE-Daten beschrieben, und zur Unterstützung der Validität einer weiteren wichtigen Subkategorisierung, die in Korpora gemeinhin als 'Konversation' bezeichnet wird, werden sprachliche Belege angeführt. Olga Karpova und Svetlana Manik gehen in ihrem Beitrag der Frage nach, inwieweit gesellschaftliche und politische Prozesse zu sprachlichen Veränderungen führen. Sie analysieren eine Auswahl von Wörterbüchern und erörtern dabei verschiedene Methoden zur Wörterbuchkritik. Die Verf. betonen die Notwendigkeit der Erstellung eines Wörterbuches, welches eine objektive Darstellung und Markierung des politischen Wortschatzes bietet. Ksenija Leban gibt einen detaillierten Abriss der Struktur und Funktion eines Slowenisch-Englischen Wörterbuches zu falschen Freunden mit professionellen Sprachbenutzern als Zielgruppe - ein oft stiefmütterlich behandelter Wörterbuchtyp. Wörterbücher zu falschen Freunden sollen zur Ausschließung von solchen Fehlern beitragen, die auf der Vermutung basieren, dass ein zielsprachiges Lexem dieselbe Bedeutung und/oder Form hat wie seine scheinbaren Entsprechungen in der Quellensprache - eine in der Realität vielfach unbegründete Vorstellung. Die Diskussion im Beitrag von Anatoly Liberman konzentriert sich auf drei Thesen: 1) Ein etymologisches Wörterbuch sollte einen detaillierten, kritischen Überblick der Mutmaßungen über den Ursprung der aufgenommenen Wörter vermitteln. 2) Nah verwandte Wörter sollten in ein und demselben Wörterbuchartikel behandelt und nicht auf verschiedene Stellen je nach dem Anfangsbuchstaben des betreffenden Wortes verteilt werden. 3) Die Vorgeschichte von Lehnwörtern sollte in der Regel aus den etymologischen Wörterbüchern der Empfängersprachen ausgelassen werden. Sändor Martsa weist nach, dass englische Konversionen nicht im Rahmen der herkömmlichen Unterscheidung zwischen Polysemie und Homonymie beschreibbar sind. Begriffe wie semantische Verwandtschaft und Vorhersagbarkeit werden diskutiert, wonach zwei Gruppen von Verben analysiert werden: auf Tiere bezogene Verben sowie Verben, die von Instrumentausdrücken konvertiert sind. Durch die Analyse einer Reihe von Beispielen beider Typen wird klar, dass tierbezogene Verben und Instrumentverben in Wirklichkeit Wortmetaphern oder Wortmetonyme sind. Dies bedeutet, dass die semantische Verwandtschaft zwischen diesen und den entsprechenden Basissubstantiven lediglich als eine Art Polysemie - und nicht Homonymie - interpretiert werden kann. Geart van der Meer thematisiert die Berücksichtigung konventioneller Metaphern in den vier wichtigsten Lernerwörterbüchern des Englischen. Die Analyse konzentriert sich auf das Wortfeld um das englische Lexem für Morast, 'morass' ('bog', 'swamp', 'quagmire', 'mire', 'quicksands', 'dregs'). Es zeigt sich, dass sich all diese Wörter grob gesehen in

XIV derselben Weise benehmen, und zwar unabhängig davon, was in den Wörterbüchern steht. Es wird im Beitrag davor gewarnt, vorschnell mit Demetaphorisierung, d.h. dem Verlust metaphorischen Inhalts, zu rechnen, und weiter heißt es, dass anstelle von metaphorischer Bedeutung eher von metaphorischem Gebrauch die Rede sein sollte. Tadamasa Nishimura stellt die Ergebnisse einer Studie über Probleme von japanischen Lernern in Bezug auf den Gebrauch von Lernerwörterbüchern vor. Nishimura ist der Ansicht, dass Wörterbuchbenutzer in Japan von den Lexikographen zu oft außer Acht gelassen werden. Er stellt das Feed-back eines von Schülerinnen an Gymnasien und Studentinnen an Universitäten beantworteten Fragebogens vor und empfiehlt schluss-folgernd, dass Lexikographen sich künftig viel häufiger solcher Fragebogenuntersuchungen bedienen sollten. Die zwei Hauptanliegen des Beitrages von Vilja Oja sind erstens die Frage, inwieweit Äquivalente für Farbausdrücke in zweisprachigen Wörterbüchern implementiert werden sollen, und zweitens, wie die Bedeutung eines Wortes in einer elektronischen, im Hinblick auf die Herstellung eines mehrsprachigen Wörterbuches erstellten Datenbank definiert werden soll. Die Diskussion basiert auf Beispielen aus zweisprachigen Wörterbüchern. Dabei geht der Verf. auf mehrere Sprachen ein (Deutsch, Estnisch, Englisch, Finnisch, Russisch) und beschreibt abschließend die Datenbank der estnischen Farbausdrücke. Kurt Opitz stellt die traditionelle Beschäftigung der Lexikographen mit der Ermittlung und Auflistung der semantischen Ladung von Wörtern in Frage, welche eher Konzepten zuschreibbar sei als ihrer Repräsentation in der Sprache. Diese Betonung des denotativen Aspektes der Lexeme vernachlässigt die vitale Rolle der sprecherorientierten Konnotation, die für den aktuellen Sprachgebrauch so charakteristisch ist. Folglich wird ein besonderes Wörterbuch der Konnotationen vorgeschlagen. Dieser Vorschlag wird durch eine Untersuchung des Lexems 'garden' veranschaulicht und begründet: Wie wird dieses Lexem in konventionellen Wörterbüchern behandelt, und wie könnte es für den Zweck einer konnotationssensitiven Dekodierung von Texten dargestellt werden? Yoshiaki Otani macht darauf aufmerksam, wie wichtig englisch-japanische Wörterbücher fur die moderne japanische Gesellschaft sind. Mit Hilfe solcher zweisprachigen Wörterbücher werden japanische Benutzer in die Welt des Englischen eingeführt, und es wird ein Referenzrahmen bezüglich der Rezeption des Englischen in der Welt des Japanischen hergestellt. In beiden Fällen kommt Äquivalenten eine Schlüsselfunktion zu: als Nahtstelle zwischen den beiden in Bezug auf Syntax, Begriffsinhalt und den kulturellen Hintergrund so unterschiedlicher Sprachen. In seinem Beitrag stellt Gunnar Persson die Bedeutungsgeschichte des Wortes 'spinster' im Laufe der vergangenen 1000 Jahre dar. Er führt ein auf semantischen "Frames" basierendes lexikologisches Modell ein, in dem - im Gegensatz zu traditionellen seman-tischen Analysen - nicht zwischen Denotation und Konnotation unterschieden wird. Auf der Basis einer sorgfältigen Analyse des zum tschechischen Nationalkorpus gehörigen SYN2000-Subkorpus wird im Beitrag von Vira Schmiedtovä und Barbara Schmiedtovä die sprachliche Vielfalt der tschechischen Wörter für 'schwarz', 'weiß' und 'rot' thematisiert. Die Verf. weisen darauf hin, dass kein existierendes Wörterbuch all die semantischen und konzeptuellen Nuancen dieser Wörter beschreibt. Abschließend wird die Hoffnung zum Ausdruck gebracht, dass künftige Wörterbücher sich dieser Aufgabe zuwenden werden. Wlodzimierz Sobkowiak stellt eine kritische Übersicht über die Behandlung der Phonetik des englischen Morphems 'trans-' in sieben führenden Wörterbüchern zum Englischen als

XV Fremdsprache vor. Es wird aufgezeigt, dass die phonetische Darstellung von 'trans-' sowohl innerhalb einzelner Wörterbücher gesehen als auch der Wörterbücher untereinander unter fehlender Konsistenz leidet. Teils lässt sich die Varianz über tiefer liegende sprachliche Faktoren erklären, welche die Augen des Lerners nicht sehen, teils ist sie auf die intuitiv-individuellen Vorlieben des jeweiligen Lexikographen zurückzuführen. Außerdem werden phonetische Angaben in Wörterbüchern offensichtlich nicht so sorgfältig kontrolliert wie andere Daten. Dies führt nur zur weiteren Varianz und letztendlich zu völliger Willkür. Andrejs Veisbergs thematisiert Euphemismen und ihre Behandlung in allgemeinsprachlichen einsprachigen Wörterbüchern. Nach einer auf englischem Material basierenden sprachlichen Analyse von Euphemismenbildungen untersucht er die Behandlung von Euphemismen in mehreren englischen und lettischen einsprachigen Wörterbüchern. Dabei wird eine Vielzahl ernsthafter Inkonsequenzen der editorischen Praxis aufgedeckt. Beispielsweise werden Entscheidungen des Öfteren von Fall zu Fall getroffen, Markierungen werden inkonsequent oder gar nicht benutzt etc. Es wird außerdem auf eine Reihe anderer besonders euphemismenbezogener Probleme in den analysierten Wörterbüchern hingewiesen. In seinem Beitrag beklagt Hideki Watanabe, dass dem im Jahr 1995 erschienenen "Thesaurus of Old English" weit weniger Aufmerksamkeit zuteil geworden ist, als er verdient. Wertvoll ist das Werk nicht zuletzt deshalb, weil es sich dabei sowohl um den ersten periodenbezogenen Thesaurus dieser Art handelt als auch um eine Pilotstudie, die für die Erstellung von "The Historical Thesaurus of English", das in wenigen Jahren erscheinen wird, große Bedeutung haben könnte. Unter Berücksichtigung der von den Rezensenten hervorgehobenen Punkte bewertet Watanabe das Buch aufs Neue, indem er den Thesaurus als den wichtigsten Beitrag zur Erforschung der altenglischen Wortgeschichte im letzten Jahrzehnt des 20. Jahrhunderts bezeichnet. Abschließend unterbreitet Watanabe einige Änderungsvorschläge, besonders in Bezug auf Komposita, Kollokationen und idiomatische Wendungen. In ihrem Beitrag beschreiben Arne Zettersten und Hanne Lauridsen die mit ihrem kürzlich erschienenen englisch-dänischen Wörterbuch verbundenen Vorüberlegungen sowie einige Aspekte der Rezeption des Wörterbuches. Insbesondere wird darauf eingegangen, wie die Autoren bei der Festlegung des Arbeitsplans von einer professionellen Ratgeberfirma Gebrauch machten. Nach dem Erscheinen des Wörterbuches haben die Autoren von den Benutzern erfahren, dass dieses zweisprachige Wörterbuch zur gleichen Zeit als ein Produktionswörterbuch und Rezeptionswörterbuch charakterisiert werden muss.

XVI Die Herausgeber danken den Autoren sehr herzlich für ihre Bereitschaft, die Manuskripte ihrer Beiträge für den Druck zur Verfügung zu stellen. Wir danken auch allen - alten und neuen Freunden - für ihre Teilnahme am Symposion. Ebenso sind wir als Organisatoren dem Englischen Institut, dem Institut für Deutsch und Niederländisch und dem Zentrum für Übersetzungwissenschaft (alle Universität Kopenhagen) sehr zu Dank verpflichtet. Besonders dankbar sind wir Shirley Pollak für ihr begabtes, sorgfältiges und mehrsprachiges Korrekturlesen sowie für ihre Mitarbeit bei der Fertigstellung der Druckfassung des Manuskripts.

Kopenhagen im Mai 2001

Die Herausgeber

Acknowledgements

The editors wish to thank the authors of the contributions for placing their manuscripts at our disposal. Thanks also to all participants - old friends as well as newcomers - for joining the symposium. We are indebted for financial support to the English and German Departments and the Centre for Translation Studies, University of Copenhagen. Finally, we are deeply grateful to Shirley Pollak for her intelligent and meticulous multilingual proofreading and preparation of the papers for the printers.

Copenhagen, May 2001

The editors

Norman

Blake

Towards a Dictionary of Shakespeare's Informal English1

In the Department of English Language and Linguistics at the University of Sheffield are some photographs of distinguished philologists. One is of Otto Jespersen. No documents record how his photograph came to be there, but I suspect it was donated by Professor Moore Smith, for the two clearly knew each other well. Jespersen refers to Moore Smith in his autobiography and in several prefaces of his books, both while Moore Smith was in Oxford and later when he was professor at Sheffield. Jespersen never wrote a major study of Shakespeare's language, though Shakespeare's works are an important source for examples in his Modern English Grammar on Historical Principles (1909-49). He also wrote a brief chapter on "Shakespeare and the Language of Poetry" included in his Growth and Structure of the English Language (1982). In this chapter, a broad-brush exploration of some features of Shakespeare's language, he claimed that Shakespeare used "language to individualise the characters in his plays. In this he shows a much finer and subtler art than some modern novelists, who make the same person continually use the same stock phrase or phrases. Even when he resorts to the same tricks as other authors he varies them more" (p.203). He also noticed that words can "have another colouring than their present signification" (p. 205), for some occur only in the mouths of vulgar or affected persons. He did not develop these insights or produce much evidence to support them. For my own part I have been invited to compile a dictionary of informal English in Shakespeare. Given the links between Jespersen and Sheffield, and the interest I share with him in Shakespeare's language, this lecture seemed an appropriate opportunity to review how one might approach the task of deciding what to include in a dictionary like this. As Jespersen noted, many writers make claims about Shakespeare's vocabulary, though these claims focus on the number of words he used and his inventiveness in creating new ones. As for register, most commentators limit themselves to his use of either bawdy or colloquialism, often regarded as the exclusive feature of lower-class characters. The diction of Elizabethan "common life" was a familiar topic fifty years ago, though most essays were largely impressionistic and based on little real evidence. Few scholars have explored register in Shakespeare systematically, though Cusack (1970), Brook (1976) and Salmon (1965, 1967) have made a start. But Brook saw register mainly in terms of repartee and comedy, and Salmon confined her enquiries to the Falstaff plays. Otherwise, literary scholars restrict their sparse comments on Shakespeare's language to what they describe as his colloquialism. This comment about The Taming of the Shrew by Park Honan in his recent biography of Shakespeare is not untypical: Petruccio's and Kate's talk is colloquial, earthy, often bawdy, sharp as a slap in the face, enriched by snippets from country folk-tales and legends. (1998, p. 135)

Based on the Otto Jespersen lecture delivered on 4 May 2000.

2

Norman Blake

What Honan means by this is not clear and he gives no examples of this colloquial talk. The implication is that the language of these two characters is different from that found in other plays because of its colloquial and earthy nature. But the genre of drama demands that, if characters are to appear reasonably lifelike, their talk must not be too formal or wooden, a demand which applies to characters of high social standing as much as to Kate and Petruccio. In principle Hamlet may be as colloquial and bawdy as Petruccio. Philologists have often noted how difficult it is to trace colloquialism in language written four hundred years ago, and in this paper I would like to explore some possible approaches to isolate elements of informal language in Shakespeare's vocabulary and syntax. I shall focus on the problem presented by words and phrases not normally designated as informal by editors or lexicographers rather than on those forms beloved of commentators like dialect words, bawdy expressions or variant pronunciations. The tone of those words and phrases at the periphery of informality often escapes readers and they present difficulties to lexicographers. There are two ways of approaching this investigation: by considering material within a play on its own merits and by evaluating material in the play against external information. The two may overlap. The former is implied in the quotation by Park Honan, namely if characters are comic or of a low status, their language will be colloquial. In fact neither Petruccio nor Kate is specifically low in status, and it is the way they behave and the situations they find themselves in rather than their position in society which might determine whether their language is informal. If register is the use by individuals of a different level of language in accordance with the discourse situation, then colloquialism is not confined to a particular class of individual, though people in lower social classes may use it more because of their social environment. As sociolinguists tells us, we all, whatever our social class, indulge in colloquialisms when we are in our informal conversational mode; we all have a linguistic code of informality. Hence one way of approaching this question of register, as Jespersen suggested, is to consider words which appear to have an informal tone not so much by who uses them as by the context in which they appear. It may be that lowerclass characters find themselves more often in situations which favour informal language, but it is not invariably true. To make any conclusions convincing, it will be necessary to examine the contexts in some depth, following the principles of pragmatics. This method was adopted by Arthur King in his analysis of the language of Poetaster (1941). In that play he considered words as criticised if they were used either by characters who the author clearly satirises or by other characters when they use parody, irony or humour. The same approach may be applied to Shakespeare. Although some words or abbreviated forms may always be colloquial or informal, other words which appear to be neutral or even polite may in some discourse situations take on a different register. I start by considering this exchange in Much Ado About Nothing in which Don John and Borachio, two of the villains, are discussing the marriage being arranged between Claudio and Hero: lohn. What is hee for a foole that betrothes himselfe to vnquietnesse? Bor. Mary it is your brothers right hand. lohn. Who, the most exquisite Claudio? Bor. Euen he. lohn. A proper squier, and who, and who, which way lookes he? (1.3.43-9) 2 2

The quotations are from Hinman (1996), but the lineation is from Wells & Taylor (1988).

Towards a Dictionary of Shakespeare's Informal English

3

Here two phrases are ironic or derogatory, most exquisite and proper squier, and they will be examined in greater detail. In addition, the repetition of and who might suggest that this type of repetition was either informal or ironic, partly because of its syntactic form and partly because it appears to express curiosity mixed with disbelief. We need to discover whether these words and phrases are merely adapted by Don John to this register on this one occasion or whether they are more generally informal, even if informal in this connection means little more than "not normally used in a neutral context". Exquisite belongs to the language of courtly extravagance as a term of praise, though it became a vogue word, which aroused a reaction of disapproval in many who heard it. Its tone of exaggeration meant that it had lost its usefulness. The word occurs only in Shakespeare's plays, not his poems. It occurs in the plays twelve times, apart from this example in Much Ado, though one must remember that quartos and First Folio sometimes have different readings. The most revealing occurrences are those in Twelfth Night, where it occurs four times. When Viola as Cesario first meets Olivia, she embarks on a prepared speech as from Orsino in praise of Olivia. This starts Most radiant, exquisite, and vnmatchable beautie. (1.5.163), at which point Viola breaks off in case it is not Olivia she is addressing. This is an example of hyperbole in which the three adjectives may be regarded as appropriate to the extravagant praise of a woman, the beloved. This is the normal context for the word, confirmed in this case by the other adjectives which accompany it. The occurrence of radiant here is matched by its use by Francis Flute as Thisbe in the artisans' play in A Midsummer Night's Dream when he praises Pyramus in terms more appropriate to a woman, which is indeed part of the joke: Most radiant Piramus, most Lilly white of hue, (3.1.87). Other examples of radiant are usually used of the sun, which may suggest why it could pass over into hyperbole when applied to humans. It is applied to Cymbeline to exalt him when he is portrayed as a sun shining in the west (Cymbeline 5.6.476-7) to match the Roman Caesar in the east. It may not be surprising that it is used in The Merry Wives of Windsor in the final scene with the fairies, when Pistol says Our radiant Queene, hates Sluts, and Sluttery. (5.5.45). This example matches radiant against sluts and sluttery suggesting that the word may well be regarded as hyperbolic and thus needing to be deflated. The three other examples of unmatchable in Shakespeare are, however, unambiguously laudatory, but they are applied to men like Antony. Of these three adjectives in the example from Twelfth Night it is exquisite which taints the others as hyperbolic, though the opening of a letter to a beloved is by its own nature a place where exaggeration might be expected. Hamlet's letter is addressed To the Celestiall, and my Soules Idoll, the most beautified Ophelia. (Hamlet 2.2.110-11). But such openings where they occur in Shakespeare are often inflated as though tongue in cheek. The second and third examples from Twelfth Night come in the same piece of dialogue between Sir Toby Belch and Sir Andrew Aguecheek: To. What for being a Puritan, thy exquisite reason, deere knight. An. I haue no exquisite reason for't, but I haue reason good enough. (2.3.137-40) These represent a playful use of words by the two knights, who are straining to appear courtly and fashionable both to each other and to the others present. Nevertheless, the exchange suggests that exquisite is a word which lent itself to this usage, and Sir Andrew picks it up from Sir Toby because he strives to be fashionable. It may not be colloquial, but

4

Norman Blake

it suggests a register which is not as formal as it at first appears. The final example from Twelfth Night is also spoken by Sir Toby, who is hugely enjoying the teasing of Malvolio by Feste in his role as Sir Topas. To Malvolio's anguished cries of Sir Topas, sir Topas., Sir Toby can only blurt out in joy My most exquisite sir Topas. (4.2.62). These examples by Toby and Andrew together with the one used hyperbolically by Viola as Cesario suggest that exquisite was a loaded word; it was fashionable and part of courtly language but it could easily pass over into the informality of comic or ironic exaggeration. The only other comedy in which this word occurs is The Two Gentlemen of Verona, in which Valentine expresses his love for and admiration of Sylvia to Speed, who mocks him for his expression of love by misunderstanding his words in a way Valentine had not intended. The dialogue includes this exchange: Val. What dost thou know? Speed. That shee is not so faire, as (of you) well-fauourd? Val. I meane that her beauty is exquisite, But her fauour infinite. Speed. That's because the one is painted, and the other out of all count. (2.1.49-54)

One may accept that Valentine uses the word exquisite in sincere praise of Sylvia, but Speed mocks the word because it is fashionable and exaggerated, and because those who use it cannot see beneath the surface of what is described by it. A lady who uses sufficient cosmetics may be exquisitely beautiful to some, but not to those who know what it covers up. Neither uses the word neutrally: for one it represents the highest praise, for the other self-deception. Otherwise, the word occurs in several tragedies. The same sense of exaggeration is found in Romeo and Juliet, for it is a word used in the opening scene by Romeo to describe his first love, Rosaline. The expressions he uses are stereotyped and traditional, because this is not real love, but the love of a distant lady as part of the love game. Romeo does not use this type of vocabulary, which includes exquisite, when he really falls in love, with Juliet. Benvolio, like Speed in Two Gentlemen, mocks this vocabulary. He advises Romeo to open his eyes to look at other women: Ben. By giuing liberty vnto thine eyes, Examine other beauties. Ro. 'Tis the way to cal hers (exquisit) in question more, (1.1.224-6)

Although exquisite is marked off by brackets here, it is not otherwise unusual in that it fits into the pattern of Romeo's employment of the traditional vocabulary of love at this stage in his development. This vocabulary is appropriate for the game of love, but not to represent true feelings. Perhaps more significant are its uses in other tragedies. In Othello it is used three times in the scene which witnesses Cassio's disgrace and downfall. The first time is when Iago and Cassio discuss Desdemona, the former seeing her more as a sex object and the latter using language more suitable for a lover: Iago. ... he hath not yet made wanton the night with her: and she is sport for loue. Cas. She's a most exquisite Lady. Iago. And lie warrant her, full of Game.

Towards a Dictionary of Shakespeare's Informal English

5

Cas. Indeed she's a most fresh and delicate creature. (2.3.15-20) Although Cassio means exquisite to express his admiration for her in a positive sense without any ambiguity, the context in which it occurs suggests that far from being a distant beauty she is little more than an overcharged sex symbol. Cassio himself later in this scene undermines the status of this word, which he uses twice more when drunk. Iago, having persuaded Cassio to drink more than he should, indulges in nonsense talk. First he talks about the drinking abilities of various nations, where he claims of the English they are most potent in Potting. (2.3.70-1). To this Cassio responds Is your Englishmen so exquisite in his drinking? (2.3.74). Few commentators comment on exquisite in this context, though some editors follow the Quarto in reading expert. However, Andrews (1991) glosses exquisite "choice, distinguished" on the basis of the other examples in this scene. Although this may be the word's basic meaning, here the sense is "elegant, fastidious" or even "notorious". That it describes outrageous drinking habits puts the word into a different context from its normal one, and it implies excess. Later in this scene Cassio enthuses about the trite and probably traditional song on King Stephen's trousers sung by Iago with Why this is a more exquisite Song than the other. (2.3.91-2). Once again exquisite is put into a different context where it implies no more than general approbation, but in both cases of things which for most people would not merit any approval whatsoever. If one accepts in vino Veritas, one appreciates how Cassio really understands this word. There is a single example in Timon of Athens where it is used by Lucius, one of those who befriend and flatter Timon when he is wealthy and distributing his wealth to all and sundry. Lucius uses inflated language to describe Timon and his own friendship for him. It is hardly surprising, therefore, that he should instruct Servilius to greet Timon with commend me to thy Honourable vertuous Lord, my very exquisite Friend. (3.2.29-30). The number of adjectives and the modifying of exquisite by very all suggest how hypocritical this expression of love is. That exquisite should be addressed to a man is also potentially significant. Two final examples of exquisite occur in Cymbeline. The first is used by Giacomo, the duplicitous Italian, when he asks Imogen to store a trunk full of valuables in a safe place for him. They are described as Plate of rare deuice, and lewels Of rich, and exquisite forme, their valewes great, (1.6.190-1). On the one hand, this is a neutral, if somewhat inflated, description but, on the other hand, given Giacomo's character and the fact that there are no valuables anyway, we accept that he exaggerates their pretended worth to make sure Imogen stores the trunk in her own bedchamber; the word implies excess. The other is used by Cloten who hopes to marry Imogen. In the play he is portrayed as a braggart who uses courtly language which he has picked up without always knowing exactly what it may imply. When he is discussing Imogen with his mother, he describes her ...she hath all courtly parts more exquisite Then Lady, Ladies, Woman, from euery one The best she hath, and she of all compounded Out-selles them all. (3.5.71-4). Once again, although no doubt meant sincerely, the word occurs in a passage of hyperbole which may undercut its use. Although the language is not elaborate, the structure of Then

6

Norman Blake

Lady, Ladies, Woman implies a groping after effect without the speaker knowing how best to achieve this. There is, therefore, sufficient evidence to suggest that exquisite was a word fashionable in courtly contexts usually in praise of women, but for Shakespeare it had become a word of little more than general approbation which had lost its usefulness through misuse. It could be adopted for irony or comedy in appropriate situations. Words like this were relatively new and, as Gladys Willcock has commented, "By the end of the neo-classical period words like elegancy, fancy, invention were commonplace enough. We miss the whole point if we imagine they were so in the early fifteen-nineties." (1934, p. 12). S-L glosses exquisite as "excellent" and quotes a few examples without further comment. OED Exquisite a. and sb. 4 defines it as "Of a person, etc.: accomplished either in good or bad things; consummate, excellent, perfect". Few users of either dictionary will understand that Shakespeare uses this word to mean "over the top, excessive". In this period when many words were being incorporated into the language, it is the context rather than the word itself which indicates its level of formality. In the original passage from Much Ado its modification by most merely emphasises this state of affairs, because a word which hardly warrants any indication of degree is given a superlative modifier. Other examples employ this or other modifiers in their striving for an effect. The other phrase in the original passage used by Don John in Much Ado which matches most exquisite in its informal and ironic use is A proper squier. This example is of a different kind since it is the only occurrence of this phrase in Shakespeare, but the linking of proper with squier may be significant. Proper occurs frequently in Shakespeare. It has several senses and its principal one "own, conformable, belonging to a particular person or state" is not relevant here, for that is its neutral or acceptable sense. In that way this word differs from exquisite because it is most often used neutrally. Proper also occurs in the phrase a proper x, where χ is a noun of human status, such as man, maid etc. S-L glosses the word in these contexts as "honest, respectable (used of women)", but "fine, nice, pretty (used of men)" though one of the latter examples is translated as "handsome". S-L suggests the word is used ironically when modifying nouns like saying, jest or peace, without any indication that this might also apply when modifying nouns like man, though the separation of meanings when proper is used with male or female persons suggests that this is something worth exploring. There is variety in the tone of many examples of this type of phrase. In many instances there is no reason to doubt that the phrase a proper man is a compliment, as in Two Gentlemen 4.1.10-11, where the outlaws describe Valentine as such before they choose him as their leader; in Twelfth Night where Olivia can see the time when Viola/Cesario will become a proper man: (3.1.132, though there may be some irony here since that is not possible for Viola); and in Othello where Desdemona refers to Lodovico, who brings news that Cassio is to be the new governor of Cyprus, as a proper man. (4.3.34). A little more ambiguous, since the phrase comes in a passage of witty exchanges, may be Le Beau's description in As You Like It of Three proper yong men, of excellent growth and presence. (1.2.112-13), who are all thrown by Charles the wrestler so that they are likely to die. More ironic is Claudio's description of Benedick as a very proper man. {Much Ado 2.3.175) in response to Don Pedro's claim for the man (as you know all) hath a contemptible spirit. (2.3.173-4), both of which occur in a passage where Claudio and Don Pedro are tearing Benedick's character to shreds. The same applies in Richard III where Richard expresses astonishment that Anne findes (although I cannot) My selfe to be a

Towards a Dictionary of Shakespeare's Informal English

1

maru'llous proper man. (1.2.240-1). Clear contempt is evident in The Taming of the Shrew in Grumio's description of Lucentio as A proper stripling, and an amorous. (1.2.141); though stripling signifies immaturity, it is not otherwise a derogatory word, so it may be its combination with proper which gives this utterance its edge. There is no reason to doubt that proper can vary from being complimentary, through being ambiguous, to having an ironic tone. Unlike exquisite it is not usually derogatory, although it can be. It depends both on the context and on the company of other words it keeps. This leads us to focus on squire a word which by itself or as part of a compound occurs eighteen times in Shakespeare's works, all of them in his plays. It is a word which falls into two distinct fields, one the title of those gentlemen just below the rank of knight and the other referring to a man of any class, used both familiarly and contemptuously. We are not concerned with the former, which is usually linked with other words of rank. The latter is still an informal, even slang, word in Present Day English meaning "fellow, chap, lad". S-L glosses this sense as "a familiar title, given sometimes in tenderness, and sometimes in contempt; almost = fellow". OED Squire sb. 1 d refers to the word being in contemptuous use, though now obsolete. However, Holdsworth (1986) believes that squire also had a slang meaning in Shakespeare's English in the sense of "pimp, whoremaster", though he does not refer to this passage in Much Ado. The use of squire to refer to a child is affectionately familiar: Titania in A Midsummer Night's Dream refers to the mother of the boy who is the subject of the dispute between her and Oberon with (her wombe then rich with my yong squire) (2.1.131); Leontes in The Winter's Tale refers to his young son Mamilius So stands this Squire Offic'd with me: (1.2.172-3). However, squire can refer to a man, possibly a young man, though with contempt. In Othello Emilia refers to whoever betrayed Othello first as some most villanous Knaue, Some base, notorious Knaue, some scuruy Fellow. (4.2.143-4) and in her next speech as some such Squire he was That turn'd your wit, the seamy-side without, (4.2.149-50). In other examples the reference to status is present, but the connotations of squire are negative. In 1 Henry VI Talbot suggests Sir John Fastolf should be stripped of his rank because at the Battle of Patay he Like to a trustie Squire, did run away. (4.1.23); in 1 Henry IV Falstaff refers to vs that are Squires of the Nights bodie, (1.2.25) in a context which makes it clear they are thieves; and in King Lear the compound squire-like in To knee his Throne, and Squire-like pension beg, To keepe base life a foote; (2.2.387-8) paints the role of squires as unflattering, to say the least. The portrayal of squires is of men who are young and on the lowest rank of the aristocratic hierarchy who have constantly to defer to their elders and who, perhaps through inexperience, are likely to prove unreliable in battle. Although Claudio is young, inexperienced and still junior in rank (for Don Pedro goes out of his way to woo on his behalf), he has proved himself in the recent war. Familiarity can turn into contempt. Don John's A proper squier picks up the least favourable elements of the words proper and squier to highlight Claudio's youth and inexperience. The sense of squire in this context must refer not to his rank, but to his youth and inexperience. It may also suggest that Claudio is no better than a pimp. Given the later accusation that Hero has been unfaithful, the suggestion that Claudio is her pimp may look forward to later developments in the plot. The phrase suggests an informal register not fully recorded in either dictionary. The discussion of these words has been long, but they are rarely commented on in editions and it is far from certain that they are understood by readers of Shakespeare. Of these words exquisite is glossed with the single sense "supreme", but within most of the contexts in

Norman Blake

8

which the word occurs in Shakespeare that sense is made to seem extravagant or excessive, and thus a vogue word passes over into the informal register implying the opposite of supreme. The praise which the word conveys is artificial, immature and frequently ironic. Proper and squire have at least two meanings each. Proper is most often a word with neutral or favourable connotations; it is only when it is paired with words of ambiguous or unfavourable connotations that a sense of irony is attracted to it so that it changes register. It is a difficult word to pin down, but its unfavourable sense occurs principally in the phrase a proper x, where χ is a noun indicating a human being, for in this phrase it may suggest irony. Squire has two quite distinct fields of meaning, one is formal and refers to a given rank, and the other is informal, almost indeed in the slang register, with a range of connotations from familiarity to contempt. These connotations can be discovered only through a comprehensive analysis of their occurrences in different contexts in Shakespeare's works, and this is why this survey has been so detailed. The evidence indicates that exquisite and the phrase a proper squire may appropriately be included in a dictionary of informal terms. It is time to turn now to other ways in which we might be able to decide on words which are informal either regularly or in particular contexts. In the original passage from Much Ado I noted the occurrence of the repeated phrase used by Don John and who, and who, and I suggested that this repetition might be significant. Let us pursue this further. By repetition I mean only those instances where a word or phrase is repeated within the same grammatical framework and not those which through the rhetorical figures like antimetabole occur side by side because of the mirror reversal of phrases or clauses. Because repetition is so common in rhetorical figures, we must be careful how we interpret its use. Repetition often expresses some emotion like amazement or despair, and such expression is likely to involve verbal strategies suggesting colloquialism. There is nothing in either and or who which by itself is anything other than a neutral tone but, when they are joined together and repeated, the combination attracts attention to itself for it appears to have shifted from a neutral to an informal level. It reminds one of the repetition of Farthee well by a lord to Apemantus in Timon of Athens 1.1.265, for Apemantus picks up the repetition of this over-effusive departure formula and turns it into a "cynical jest". Apemantus says the lord should not have used this parting formula twice; he should have kept one for himself since Apemantus does not intend to use the formula to him for he must leave without any reciprocal gesture. The repetition is ambiguous in tone, so it is worth looking at other examples. Repetition of a word or phrase comes in two forms: one when the words are intensifiers of some sort (usually adjectives or adverbs) or pseudointerjections (either adverbials or nouns), and the other, as in the examples just quoted, involving a group of words where a verb is either explicit or implicit through ellipsis, for and who in the first example implies is she? Let us start with the first category. Some examples of intensifiers are certainly colloquial. If we take Hamlet, we can see how the little verse Hamlet recites to Horatio which ends: This Realme dismantled was of Ioue himselfe, And now reignes heere A verie verie Paiocke. (3.2.270-2)

contains the repetition of verie. Nobody knows exactly what Paiocke means, though the verse is clearly intended to conclude bathetically. If repetition of this sort is informal, the

Towards a Dictionary of Shakespeare's

Informal

English

9

introduction of verie verie lowers the tone to prepare us for the introduction of the final word Ρaiocke. Some editors (Wells & Taylor 1988 and Jenkins 1982) unnecessarily place a dash in front of this word to serve the function which is fulfilled by verie verie if they did but realise it. The tone of repeated verie may be strengthened by the word's ambiguous nature for it was often regarded as a mark of exaggeration. At least that is how one may interpret Mercutio's comment: The Pox of such antique lisping affecting phantacies, these new tuners of accent: Iesu a very good blade, a very tall man, a very good whore. (Romeo and Juliet 2.3.26-8)

Another example of this type of repetition occurs in Hamlet's first soliloquy which opens in the Folio Oh that this too too solid Flesh, would melt, (1.2.129), where one might at first suppose that too too was introduced to heighten the style of the soliloquy. However, OED Too adv. 4 shows that too too was a common phrase in the second half of the sixteenth and first half of the seventeenth centuries. The dictionary does not suggest what field it may fall into, though many of the examples it quotes indicate that its usage was informal partly because of the nature of the texts it occurs in and partly because of its association with other colloquial words and frequent alliteration. It was often spelt as one word tootoo or hyphenated too-too in the texts of this time as though it had adopted reduplicative status rather like Present Day English so so\ a number of modern Shakespearean editions retain the hyphen in too-too in some of its occurrences. Shakespeare himself uses this repetition on several occasions, mainly in prose passages in the comedies. In The Merry Wives of Windsor, Master Ford disguised as Brooke, when bargaining with Falstaff about seducing his own wife, says a thousand other her defences, which now are too-too strongly embattaild against me: (2.2.239-40). In Love's Labour's Lost it occurs in a passage of bombast spoken by Armado: the Schoolmaster is exceeding fantasticall: Too too vaine, too too vaine. (5.2.525-6), where the repetition of the phrase is significant as to its status. In the other two plays it occurs in verse. In The Merchant of Venice, Jessica while joking with Lorenzo in a semi-serious way as she is about to flee her father's house says: What, must I hold a Candle to my shames? They in themselues goodsooth are too too light. (2.6.41-2),

where both the witty tone of her comment and the association of too too with goodsooth and light suggest an informal level of speech. In 3 Henry VI Queen Margaret uses the expression to taunt the captured Duke of York, who had taken the crown from Henry VI, with: Oh 'tis a fault too too vnpardonable. Off with the Crowne; and with the Crowne, his Head, And whilest we breathe, take time to doe him dead. (1.4. 107-9).

Margaret's wickedness expresses itself in her language throughout her speech in which she goads York and belittles him, and her use of too too before vnpardonable may well indicate a lowering of tone before the final coup de grace. Too too also occurs once in The Rape of Lucrece describing Tarquin's hesitation before acting: But honest feare, bewicht with lustes foule charme,

Norman Blake

10 Doth too too oft betake him to retire, Beaten away by brainesicke rude desire. (173-5).

It is not certain that all these examples are informal, though those from the comedies certainly are and the one from 3 Henry VI exemplifies a mixture of anger and an almost vulgar playing with words. Even the example in Lucrece is somewhat ambiguous in its tone though, if it is not informal, it does at the very least illustrate Tarquin's troubled state of mind. Other repeated words act to deflate pomposity as when in Love's Labour's Lost Holofernes, having been praised for his alliterative verse on The prayfull Princesse, then claims This is a gift that I haue simple: simple, (4.2.66), where simple: simple, is more humorous than he intends because of its lowering of the tone after the pomposity of his verse. In Hamlet the repetition of a single word occurs more than one might expect. For example, there can be little doubt that, when Osric takes his leave with I commend my duty to your Lordship, and Hamlet replies Yours, yours: hee does well to commend it himselfe, there are no tongues else for's tongue. (5.2.142-5), Hamlet uses this repetition to make fun of Osric which is partly informal and partly exaggeration. Osric's inflated language is shown up for what it is by Hamlet's use of a different register. Hamlet's too too in his first soliloquy has not in recent times been interpreted as informal, but if these two words were represented as hyphenated in modern editions, as other examples of too-too often are, our attitude to its discourse level might be different. Indeed, James Orchard Halliwell (1844, 1.39) claimed it was a provincial form and cited Ray's A Collection of English Words (1674) as his authority, but since then it has been regarded simply as an intensive or not discussed at all.' As is well known, readings in the early quartos are different in this line. The first quarto has Ο that this too much grieu'd and sallied flesh, and the second quarto Ο that this too too sallied flesh would melt, for the line. The traditional reading in modern editions was the First Folio's solid flesh until Dover Wilson (1934, pp. 151-152) proposed the emendation sullied which, he argued, had been misprinted as sallied in the quartos. However, no editor has considered the implications of too too in the reading of the first line. If too too is informal, this may have some bearing on the reading of solid/sallied. Neither solid nor sullied (the preferred modern reading) gives completely acceptable sense here, since flesh is not exceptionally solid nor is Hamlet's flesh particularly sullied. Perhaps like verie verie, the introduction of too too with its informal register is preparing us for a different sort of word. Sallied must relate to the verb sally "to rush forth, to be thrust out" and may be used here in the informal sense of Hamlet's flesh being thrust out into a new world (Denmark which is a prison to him) where it is under assault and exposed to the temptations of the immoral and greedy court. This explanation might explain the first quarto's grieu'd "afflicted". Hamlet would rather his flesh metaphorically melted away by his return to Wittenberg. Although this may not be the right interpretation, we should exploit the information provided by such repetitive phrases in understanding the possible meaning of the text. It may be significant that in this soliloquy there are a number of other repeated words and phrases at strategic points: Ο God, Ο God! (line 132), Oh fie, fie (135), and even Why she, euen she, (149). Just as the first quarto has

GTSW defines too-too as "extremely, very", and notes it is "common".

Towards a Dictionary

of Shakespeare's

Informal

English

11

only a single too, it has none of these other repeated words or phrases; the second quarto has too too but only ο God God of the others. It seems as though either these more informal expressions were omitted in the quartos or that the emotional tone of the soliloquy was being more clearly pointed by introducing words of an informal level at significant points. It is worth recording that other soliloquies also have words of a similar informal level. A later one opens: Now might I do it pat, now he is praying, And now lie do't, (3.3.73-4), where pat both here and in other quotations from Shakespeare is colloquial, though interestingly neither the first nor second quarto has this word in this soliloquy in Hamlet. But all have the echoing of now and do in successive clauses, though that may be more rhetorical than informal. Hamlet's use of such expressions is echoed by repetitions uttered by other high-born speakers. I will quote one final example, from Love's Labour's Lost 5.2.268. Rosaline, one of the ladies-in-waiting to the Princess of France, caps Boyet's assessment of the gentlemen's wit with Wel-liking wits they haue, grosse, grosse, fat, fat. The apparent approval of the men's wits is undercut by two words each repeated, where the repetition (tagged on as it is at the end) gives the preceding words a definitely informal, even vulgar, tone." There would seem no reason to doubt that too too was informal in register and that it and similar repetitions are often used to deflate what might otherwise be overly pretentious utterances. The second category of repetition referred to previously consists of phrases which include a verb, either explicitly or implicitly. The first examples I wish to focus on are imperatives. There may be a structural distinction in the various types of imperative which allowed one group to develop a more informal level. Imperatives may be addressed to individuals or to groups of people. There is a tendency when they are addressed to a group to omit any pronoun or form of address. With individuals the opposite tendency may be noted; a pronoun, a form of address or a politeness formula like I pray thee is often included. Examples with plural addressees include guard with Halberds. (Comedy of Errors 5.1.186, said by the Duke to his officers), Guard her till Ccesar come. (Antony and Cleopatra 5.2.36, said by Proculeius to the soldiers); and those with individual addressees Abhominable Gloster, guard thy Head, (/ Henry VI 1.4.85), henceforth guard thee well, (Troilus and Cressida 4.7.137, said by Hector to Achilles). However, there are examples where a single imperative may not have the subject expressed, particularly where that imperative like go acts as a quasi-auxiliary, but this applies only when the verb is simple and not when it is a phrasal verb. When the simple verb is repeated, it tends not to have a subject expressed, and it creates a different tone for the imperative is not expressing a command, but more an emotional statement. It seems likely that both the simple imperative used as a pseudo-auxiliary and the repeated imperative, whether a simple or a phrasal verb, are informal in usage. Such forms are common in prose passages, but they are by no means confined to them. Let us consider some examples. Verbs used as pseudo-auxiliaries are common verbs like come and go. When used in this way they lose their primary meanings and are equivalent to Present Day English forms like go on, as in go on tell me what happened, where go on rather like OK expresses encouragement or scepticism or even both. Thus when Celia in As You Like It says to Rosalind come lame mee with reasons. (1.3.5-6), it is an informal way of saying "I am waiting to be convinced by your reasoning". When later in the play Duke Senior says Come, shall we goe and kill vs venison? (2.1.21), come has the nature of an '

For other examples see Joseph (1966), pp. 87-89.

12

Norman Blake

interjection or call for attention, which could be replaced by words like Now, What or some asseveration. In these cases, especially when come and go are pseudo-auxiliaries, modern editors find it difficult to decide how to punctuate the resulting expression, some putting a comma after come or go and others having no punctuation. This itself signals that these words have an ambiguous grammatical status characteristic of informal language which makes the conventions of standard grammatical punctuation hard to apply. When the imperative is repeated there may be a word in the vocative, though often for other reasons. Thus when Orlando is threatened by his older brother Oliver, he says Come, come elder brother, you are too yong in this. (1.1.50-1), but elder is introduced as a contrast to yong. Most often the repeated expression come come expresses dismay or surprise at what another character has said or done. Thus in The Comedy of Errors Angelo the goldsmith says to Antipholus of Ephesus who has denied any knowledge of the gold chain Come, come, you know I gaue it you euen now. (4.1.55). And in Much Ado in the scene of the masked dance Ursula does not believe Antonio's claim that he is someone other than Antonio and says Come, come, doe you thinkel doe not know you by your excellent wit? (2.1.111-12). Examples with go are just as frequent as pseudo-auxiliaries, but as an expression of surprise or disbelief it is much more common in the form go to, whether alone or repeated. In the example just quoted from Much Ado where Ursula rebukes Antonio for pretending to be someone else, she carries on goe to, mumme, you are he, (2.1.112-13), where the association of goe to and mumme makes the informal nature of both expressions clear enough. Other examples of go to are common. When it is repeated, it expresses particular frustration, even anger, as in: Go too, go too, thou art a foolish fellow, Let me be cleere of thee. (Twelfth Night 4.1.3-4), but in The Winter's Tale it expresses Leontes' jealousy: Goe too, goe too. How she holds νρ the Neb? the Byll to him? (1.2.183-4), with his distress emphasised by words referring to Hermione's mouth like Neb and Byll, which suggest animality or even bestiality. Go as a pseudo-auxiliary also occurs with verbs in contexts which suggest colloquialism: Go hang your self you naughty mocking Vnckle: (Troilus and Cressida 4.2.28), You may go walk, and giue me leaue a while, (Taming of the Shrew 3.1.57), Go sleepe, and heare vs. (Tempest 2.1.195 said by Antonio to Gonzalo). Indeed the verb go was one which had many meanings, most of which suggest that it was frequently used at an informal level. In the previous set of examples I referred to go to and this introduces the question whether phrasal verbs, of which this is one, were at this time more informal than formal in register. They occur much more frequently in the comedies than in the other plays and they are more likely to be found in prose than verse. However, it is doubtful whether one can say that all phrasal verbs were informal, but this was a productive area for new words, as it still is, and many of these new forms were and are created at a colloquial level. Phrasal verbs generally consist of monosyllabic verbs and prepositions, though some bisyllabic ones occur. Hence the majority of phrasal verbs consist of words of Anglo-Saxon origin and often contrast with polysyllabic words of French or Latin origin. Even today phrasal verbs like go up and come down are less formal variants of ascend and descend. It is appropriate that the plebeian in Julius Ccesar should say Let him go vp into the publike Chaire, Wee 7 heare him: (3.2.64-5) since we would expect a person of his status to use informal language. Perhaps Hamlet intends his language to be understood as informal and thus belittling to Claudio when he says you shall nose him as you go vp the staires into the Lobby. (4.3.35-6), since he links go vp with nose, a new verb in English, in a series of utterances

Towards a Dictionary of Shakespeare's Informal English

13

that hardly maintains the deference due to a king. The same may apply to go on, for when Othello says of Desdemona she can turne, and turne: and yet go on And turne againe. (4.1.255-6), the run of monosyllables suggests he has adopted a colloquial, even vulgar, level to destroy his wife's reputation. A few other examples from As You Like It must suffice; all chosen from witty exchanges in prose: Well said, that was laid on with a trowell. (1.2.99), Then there were two Cosens laid νρ, when the one should be lam 'd with reasons, and the other mad without any. (1.3.7-9), Now lie stand to it, the Pancakes were naught, (1.2.62-3), My better parts Are all throwne downe, and that which here stands νρ Is but a quintine, (1.2.239-40). The example of the verb to nose mentioned above, which is a new word in the sixteenth century but not a Shakespeare coinage, raises the question of how far words created by functional shift or by some other means such as back-formation were introduced at a colloquial level before, in some cases, being adopted more widely into the language. Although various commentators offer accounts of functional shift in Shakespeare, they rarely seek to describe the register which such words might occupy. In Present Day English functional shift is common and not necessarily colloquial, though that depends on the type of shift. Using a noun as a modifier is common in all types of English today and is rarely colloquial, and the same may be true in Shakespeare. When in Love's Labour's Lost Berowne says full of maggot ostentation. (5.2.409) we do not regard maggot which has been shifted from noun to modifier as colloquial; it is rather part of the heightened style of Berowne's discourse in what takes the form of a sonnet in the play. But functional shift and back formation are often used in Shakespeare as part of witty and humorous exchanges, especially the interchange between noun and verb, though the shift may involve morphological adaptation. Although frequent in comic scenes and the comedies, they are also found in more serious scenes where their possible colloquial nature adds to the implicit threat which may be involved. Ford in The Merry Wives of Windsor can turn a personal name, Mrs Prat, into a verb: lie Prat-her: Out of my doore, you Witch, (4.2.170-1), though as prat was part of thieves' cant for "arse" an obscene sense may be intended for this verb (Gotti 1999, p. 24). Even Roman senators can play this game, for Menenius in Coriolanus makes a verb out of the name Aufidius, the Volscian general defeated by Coriolanus, when he says I would not haue been so fiddious'd, for all the Chests in Carioles, (2.1.128-9). Less playful and more angry is York's Grace me no Grace, nor Vnkle me, {Richard 112.3.86, Q1 adds no Vnckle). Examples of back-formation include 1 Henry IV when Falstaff makes a verb iure out of Iurer in his you are Grand Iurers, are ye? Wee'l iure ye ifaith. (2.2.88-9), and The Merry Wives of Windsor when he exclaims I must conicatch, I must shift. (1.3.29-30 from conycatching), and later when Ford says lie coniure you, He fortune-tell you. (4.2.172-3 from fortune-telling)', a similar form occurs in Antony and Cleopatra where Iras says Go you wilde Bedfellow, you cannot Soothsay. (1.2.45 from soothsayer). Not all of these forms were coined by Shakespeare, but that is not significant for the point I wish to make. Many examples of functional shift and back-formation, whether coined by Shakespeare or not, fall within the colloquial register and some (though hardly those isolated here) may have been common in ordinary speech. Most of the examples which are of an informal register are marked by the fact that the words from which they are formed also appear in the vicinity of the new formation; our attention is drawn to their register in this way. It is time now to consider what might be informal from the evidence of external evidence, and I shall draw on two possible sources. The first examples might not be considered exter-

14

Norman Blake

nal by some, since I take them from differences between the texts which exist in Quarto and First Folio. Some quartos are described as "bad" quartos because they may have been memorially reconstructed by one of the actors, but even the good quartos may have been printed without authorial approval. It is possible that the quarto texts exhibit either a higher degree of colloquialism or, conversely, a higher degree of formality compared with what is found in the First Folio. Each example must be judged on its merits. The problem of deciding which is which may be illustrated by two relatively straightforward examples. At the end of Othello Lodovico, the new arrival from Venice, says in the First Folio: (belike) Iago in the interim Came in, and satisfi'd him. (5.2.325-6), whereas the Quarto has belike, Iago, in the nicke Came in, and satisfied him. Some modem editors prefer nick to interim in their editions, and Honigmann, a recent editor, notes of nick "Colloquial and 'low', hence 'sophisticated' in F[olio] (where interim gives a long line) or revised by Shakespeare" (Honigmann 1997, p. 328). The compression makes his note difficult to understand, though the implication seems to be that if the text was revised by Shakespeare this would affect the assessment of nick as colloquial and interim as sophisticated. Whether Shakespeare revised this passage or not seems immaterial, for there is a huge difference between these two words. Interim is used frequently by Shakespeare, but only on one other occasion in the phrase in the interim. On the other hand, nick does not occur in this sense elsewhere in the plays, except in the 1634 Quarto of The Two Noble Kinsmen attributed to John Fletcher and Shakespeare. The phrase in the nick was also used by others at about this time; it is recorded in OED from 1577 and was presumably just emerging from its colloquial status to a more polite level. However, the evidence suggests that interim is the more likely word to have been used by Shakespeare in Othello, partly because he does not use the phrase in the nick elsewhere, for the example in The Two Noble Kinsmen was possibly introduced by Fletcher, and partly because its register makes it an unsuitable expression for Lodovico. In other words, the Othello quarto probably contains a word of an informal register which may have been introduced into the text by someone other than Shakespeare. The other example tells a different story. In King Lear the Fool says of the eels in the First Folio: she put 'em i'th' Paste aliue, she knapt 'em o'th'coxcombs with α sticke, (2.2.294-5), whereas the Quarto reads she put vm ith past aliue, she rapt vm ath coxcombs with a stick. Both knapt and rapt are used in the sense "to hit, strike". In this case rap is and was the more common word, but knap is a word from a lower register used in this period which may thus be a suitable word for the Fool. The verb knap is used once elsewhere by Shakespeare, but with a different meaning. Solanio in The Merchant of Venice says I would she were as lying a gossip in that, as euer knapt Ginger, (3.1.8-9), where knap means "to bite, nibble". In Shakespeare's plays rap meaning "to hit, strike" is used elsewhere only in The Taming of the Shrew, but always with an ethic dative; but knap in this sense is not found elsewhere in the plays. In this case it is probable that Shakespeare chose a less common word of informal register for the Fool and that this was turned into a more common one by someone else, for rap is found more widely and is still used today. It would not be difficult for a copyist or compositor to memorise a line and mistakenly introduce for the word in the text one with roughly the same sound but more frequently heard. These examples show two contradictory tendencies: first, a frequently used formal word is replaced by an informal word which is emerging from the informal register into wider usage, and secondly, an informal but less frequently heard word specially chosen for a character like the

Towards a Dictionary

of Shakespeare's

Informal

English

15

Fool is replaced by a word of similar phonetic make-up which was more widely heard at the time. One further example, this time involving syntax, may suffice. In Hamlet, a passage regarded as a crux occurs at 1.3.107-9 where Polonius admonishes Ophelia. The First Folio reads: Tender your selfe more dearly; Or not to crack the winde of the poore Phrase, Roaming it thus, you'l tender me a f o o l e .

This reading is rarely reproduced in modern editions. This passage is not found in Quarto 1, Quarto 2 does have it but in a different form: tender your selfe more dearely Or (not to crack the winde of the poore phrase Wrong it thus) you'l tender me a foole.

Modem editors follow neither Folio nor Quarto and prefer to start line 109 with Running "this generally accepted emendation" (Jenkins 1982, p. 205), indicating the running of a broken horse with no wind. But there may be something to be said for the Quarto reading. The use of an imperative as a conditional is found in colloquial English, and survives today. It would be quite possible to hear an angry parent say to a naughty child "Do that once more, and you'll go straight to bed", where the imperative Do implies the conditional " I f ' . Despite the brackets in the quarto, the sense of line 109 could be "Misbehave in this way and you'll present me with a bastard". This would be a syntactic colloquialism which, having never entered the formal language, was misunderstood and so replaced in later versions. That the compositor of Quarto 2 did not understand it is suggested by his inclusion of brackets which destroys the syntax. The other type of external evidence is that found in other plays of this time and the one I have chosen for comparative purposes is Ben Jonson's Poetaster, written c 1602, which satirises other playwrights, including John Marston. There are several points already discussed which are confirmed as informal through this play, such as some types of repetition. Other expressions are attacked by Jonson and can be applied to Shakespeare.5 One is the use of the subject pronoun at the beginning and end of a clause: I'll not be guilty, I. (1.1.39). This structure is also found in Shakespeare as in: / am no Baby I, said by Aaron after his capture (Titus Andronicus 5.3.184), and I know it I: said by Juliet when dawn breaks (Romeo and Juliet 3.5.12). These have not aroused attention among scholars as possible colloquialisms, though to accept them as such adds an extra dimension to our understanding of the scenes in which they occur. Juliet's I know it I suggests a certain simplicity in trying to prove that dawn has not yet arrived; Aaron's I am no baby I smacks of bravado in the face of certain punishment. Conversely, the omission of the subject pronoun is also highlighted as a possible colloquialism in Poetaster. Hermogenes is given to saying 'Cannot sing (2.2.107 etc), and but 'will not sing (2.2Λ22). This omission in Shakespeare is not considered a colloquialism so much as a hangover from an earlier period when the inflectional system allowed such omissions to occur. But the infrequency of this omission

5

Poetaster

is quoted from Cain 1995.

16

Norman Blake

suggests that it was something which had been retained in some registers allowing it to be exploited by dramatists. This may be the explanation of such phrases as when Mistress Quickly in The Merry Wives of Windsor says Giue your worship good morrow. (2.2.33), and when Nathaniel in Love's Labour's Lost replies to an invitation And thanke you to: (4.2.158). These are people for whom an informal register is not inappropriate, though these expressions in their language have not been identified as such previously. Some of the words highlighted earlier in this paper as evidence of an informal register in Shakespeare do not figure as such in Poetaster, and equally some words pilloried by Jonson are not singled out by Shakespeare. Exquisite is not a word found in Poetaster, whereas a word like earnest which is attacked by Jonson appears to have no unfavourable connotations for Shakespeare. This is hardly surprising, for not every writer will want to attack the same word. There are nevertheless a number of words and phrases, which are used by both authors and whose status as informal is confirmed, though not always recognised in Shakespeare editions. Such words include swaggerer, used both by Rosalind once in As You Like It 4.3.15 and several times by Mistress Quickly and Falstaff in 2 Henry IV 2.5, and setter used by Poins once in J Henry IV 2.2.50. Although setter is recognised as thieves' cant by most editors, swaggerer is simply glossed "bully, blusterer" (if it is glossed at all) by editors who take their cue from S-L. But in As You Like It the connotations of this word used by Rosalind may be significant for what it tells us about her adoption of the role of the male Ganymede. Among phrases one can point to the expression gentleman bom. It is clear from its use by both Jonson and Shakespeare that this had become a vogue expression for those who claimed the status of gentlemen through birth, but did not show the breeding of a gentleman in their behaviour. In Shakespeare it is used by Slender in The Merry Wives of Windsor, who says and a Gentleman borne (Master Parson) who writes himselfe Armigero, in any Bill, Warrant, Quittance, or Obligation, (1.1.7-9), and in The Winter's Tale by the old shepherd and his son, who had brought up Perdita and were suddenly transformed into gentlemen; they then claim themselves to be gentlemen born and snub Autolycus for his previous behaviour towards them (5.2.126 f f ) . It increases the humour of this passage if one realises that gentleman born was a stock expression with these connotations rather than simply a phrase which these two had invented for their own purpose. Poetaster does, however, satirise many oaths and asseverations and these fall into two groups: those which are rather vapid like forsooth, on my word, and which is more, and those which are potentially blasphemous like 'Slid (for "God's eyelid"). It is the first of these two groups which deserves a brief comment. In Jonson phrases in this group frequently come after the rest of the utterance (i.e. they refer back rather than forward), and they often appear in twos and threes: Yes in truth, forsooth, (2.1.84), the coaches are come, on my word (2.1.150), we are new turned poet too, which is more; (3.1.23-4). They are the sort of oath which Hotspur criticised his wife for using in 1 Henry IV when she had said Not mine, in good sooth. (3.1.242). Their weakness may be caused not merely by their structure, but also by their position in the clause as though a kind of afterthought. A recognition of the significance of this positioning might help editors decide how to punctuate examples of these expressions, which they often place at the beginning of a second clause rather than at the end of the first. When the Hostess in 2 Henry IV says: you are both (in good troth) as Rheumatike as two drie Tostes, you cannot one beare with anothers Confirmities. What the good-yere? One must beare, and that must bee you: (2.4.53-7), editors often put the phrase

Towards a Dictionary of Shakespeare's Informal English

17

What the good-yere?, which stands by itself in F, with what follows rather than with what comes before it. But the evidence of Poetaster suggests that it is part of the previous clause and strengthens the other oath in good troth. It has not been possible to do more than suggest some methods which might be followed in collecting material for a dictionary of Shakespeare's informal English. However, one last point could be made. Many of the features I have commented on are ones which either survive or have their equivalent in contemporary spoken English, as can be recognised from the corpora of spoken language which have been assembled. Vogue words today, however, are not likely to be those with a Latinate origin, but are those associated with the teenage pop culture; they include words like great and wicked, both of which like exquisite express general approval and are overused. The repetition of intensifiers or modifiers is common enough: that's really, really wonderful news. But conversely the placing of adverbials at the end of a sentence so that they seem to weaken to become little more than emotional comments on what has gone before are characteristic of the modern spoken language: I like him really; and that's a silly thing to say, what's more. Phrases like a proper squire and a gentleman born have their equivalent in a real man and a true gentleman, both of which are stock idioms that are or were overused. The use of go as a pseudo-auxiliary is paralleled today by the use of got in the same role, he got drowned off the coast of Zanzibar. The formation of new phrasal verbs, many of which have not achieved a more formal status, is very productive in the spoken language; one need only think of all the verbs which can be used for the meaning "go away", like buzz off fuck o f f , push o f f , and shove o f f , to name only a few. The repetition of the subject pronoun at the end of a clause is also common enough in spoken English, though the repeated form may today take the non-subject case: he's a real bastard him, or the subject and auxiliary may be repeated: he's coming home, he is. On the other hand, different forms of ellipses such as the omission of the subject pronoun or the subject pronoun and auxiliary are frequent: Coming instead of I'm coming. And so the list could go on. What this might suggest is that the spoken language continues on its way almost untouched by the teachings of the grammarians and the efforts of educationalists. Criticisms of informal English by the speakers of Standard English and the satire on the pretensions of those who strive to enrich their language appear to have little effect on what is going on at the spoken level. Evidence of this informal register in the past is often only possible to extract from attacks by others and from dramatists like Shakespeare. But it is obviously robust enough to survive whatever we can throw at it; all we can do is to record it.

18

Norman Blake

References

(a) Cited Dictionaries GTSW = A GLOSSARY OF TUDOR AND STUART WORDS. Ed. W. W. Skeat. Oxford: Clarendon Press 1914. OED = THE OXFORD ENGLISH DICTIONARY. Ed. J. A. H. Murray et al. Oxford: Oxford University Press 1930. S-L = SHAKESPEARE-LEXIKON. Ed. A. Schmidt, rev. by G. Sarrazin. Berlin: Reimer 19023.

(b) Other Literature Andrews, John F. (1991): William Shakespeare Othello. London: Dent. (= Everyman Library). Brook, G.L. (1976): The Language of Shakespeare. London: Deutsch. (= Language Library). Cain, Tom. (1995): Poetaster: Ben Jonson. Manchester and New York: Manchester University Press. (= The Revels Plays). Cusack, B. (1970): Shakespeare and the Tune of the Time. Shakespeare Studies 23, 1-12. Gotti, Maurizio. (1999): The Language of Thieves and Vagabonds. Tübingen: Niemeyer. (= Lexicographica Series Maior 94). Halliwell[-Phillipps], James Orchard. (1844-1849): The Shakespeare Society Papers. 4 vols. London: Shakespeare Society. Hinman, Charlton. (1996): The First Folio of Shakespeare. 2nd ed. New York and London: W. W. Norton. Holdsworth, R. V. (1986): Sexual allusions in Love's Labour's Lost, The Merry Wives of Windsor, Othello, The Winter's Tale and The Two Noble Kinsmen. Notes and Queries 231, 351 -351. Honan, Park. (1998): Shakespeare; A Life. Oxford: Oxford University Press. Honigmann, Ε. A. J. (1997) Othello. London: Nelson. (= Arden third series). Jenkins, Harold. (1982) Hamlet. London: Routledge. (= Arden second series). Joseph, Sister Miriam. (1966): Shakespeare 's Use of the Arts of Language. New York and London: Hafner. Jespersen, Otto (1909-49): A Modern English Grammar on Historical Principles. 7 vols. Copenhagen: Munksgaard. - (1982): Shakespeare and the Language of Poetry. In his Growth and Structure of the English Language, 10Ihedn, 199-221. Oxford: Blackwell. King, Arthur H. (1941): The Language of Satirized Characters in Poetaster: A Socio-stylistic Analysis 1597-1602. Lund: Gleerup. (= Lund Studies in English 10). Salmon, Vivian. (1965): Sentence Structure in Colloquial Shakespearian English. Transactions of the Philological Society, 105-140. - (1967): Elizabethan Colloquial English in the Falstaff Plays. Leeds Studies in English n.s. 1, 3770. Wells, Stanley and Gary Taylor. (1988): William Shakespeare, The Complete Works, Compact Edition. Oxford: Clarendon Press. Willcock, Gladys. (1934): Shakespeare as a Critic of Language. London: Oxford University Press.

Wolfgang

Viereck

The Atlas Linguarum Europae and its Insights into the Cultural History of Europe

1. Europe and the Atlas Linguarum Europae: some background information

The Atlas Linguarum Europae (ALE) has been in existence for three decades now. Recently a new introduction was published (Alinei, Viereck 1997 [1998] a) as well as the fifth fascicle of commentaries and maps (Alinei, Viereck 1997 [1998] b). The sixth fascicle is in press (Viereck forthcoming a) and the seventh will be finished in manuscript in the course of 2000. The ALE can be called a linguistic atlas of the fourth generation, being preceded by regional and national atlases as well as by atlases of language groups. Atlases of the fifth type, i.e. on entire language families such as Indo-European, or of the final type, namely a world linguistic atlas, do not exist as yet. The ALE is the first continental linguistic atlas, its frontiers being neither political nor linguistic but simply geographic. The choice of the continent has nothing to do with Eurocentrism but only follows from the present state of research. The linguistic situation in Europe is quite complex. No fewer than six language families are present here: Altaic, Basque, Indo-European, Caucasian, Semitic and Uralic. In these language families, a total of 22 language groups, such as Germanic and Romance, can be counted. These, in turn, consist of many individual languages. It thus becomes apparent that the demands on scholars to interpret the heterogeneous data collected in 2,631 localities from Iceland to the Ural mountains are very high indeed. It is always the oldest vernacular words that are looked for in the various languages. These are, then, put on symbol maps and interpreted either synchronically or diachronically as the cases require. The ALE is, primarily, an interpretative word atlas. It uses both traditional and innovative methods. Among the former, onomasiology and semasiology must be mentioned. In onomasiology one elicits from informants the designations of certain objects etc., while in semasiology one asks for different meanings of a single form. Motivational mapping, however, is an innovative manner of interpreting geolexical data. It goes beyond an interest in etymology and asks for the causes or the motives in designating certain objects. Only in a large-scale project such as the ALE can this approach be successfully pursued. In national, let alone regional linguistic atlases, the area is too small for the approach to be productive. This may be one reason why it has aroused so little interest prior to the ALE. Another may be seen in de Saussure's dominance in modern linguistics. The arbitrariness of the linguistic sign, important as it is for the functional aspect of language, left hardly any room for the genetic aspect of language, i.e. for the serious study of motivation. Seen more narrowly, however, the motivation of a linguistic sign is not in opposition to its arbitrariness, as the choice of a certain motive itself is not obligatory.

20

Wolfgang Viereck

The latter aspects point to the past and it comes as no surprise that insights into the ethnolinguistic origins of Europe are also expected from the ALE. This is a most lively and controversially debated field at present where archaeologists and geneticists join forces with linguists, as the works of Renfrew (1987), Cavalli-Sforza et al. (1984), Sokal et al. (1992) as well as of Gamkrelidze and Ivanov (1995) reveal. In this connection mention should also be made of the nostratic theory (on which see, for example, Shevoroshkin 1989a, 1989b, 1990, 1992), of the typology of linguistic universals, and of the Sprachbünde. In the Uralic area the Continuity Theory, first advanced by archaeologists, then by linguists, seems now accepted by the majority of specialists. According to this theory Uralic peoples and their languages have lived in their present historical territories since the Mesolithic Age. Alinei (1996) adopts a similar approach for Indo-European, arguing that there was never an IndoEuropean invasion and that Indo-European languages followed the same diffusion pattern as the Uralic languages, thus diverging sharply from the many solutions to the IndoEuropean homeland problem postulated in the literature (cf. Mallory 1989). If people did not migrate, how did conceptions wander, one might ask. We find the same conceptions in every society irrespective of the languages peoples speak. I will address this aspect below. As regards the ALE, insights into Europe's cultural past follow less from loanwords and from reconstructed roots, although the project also has important contributions to its credit in these two areas - loanwords usually belong to the historical period and are thus too young, while reconstructed roots involve very early periods but are usually motivationally opaque and thus not very revealing for a cultural analysis - but rather from motivations in so far as they are transparent. This is an important point, as formal differences between languages can thus be ignored and the focus is solely on identity or similarity of cultural representations of reality. In what follows I will draw mainly on ALE data but also on my own research in illustrating Europe's cultural history. Of the manifold motivations for designating realia not all are relevant in this specific context. To give one example: the plant Leontodon Taraxacum is called dandelion in English, Löwenzahn in German and pissenlit in French. The first two names are motivated by the shapes of the leaves of the flower (French dent-de-lion 'lion's tooth' > English dandelion), while the third name refers to the medical consequences of the plant. This standard French term was taken over by neighbouring German and Dutch dialects as Bettpisser or Bettseicher, both 'bed-wetter' (cf. Sterck 1987 on Dutch).

2. Cultural History and Religion

As religion is the basis of every culture, the frame of reference here is the history of religions. The religious historian Donini (1977, 1984) has convincingly shown that in a classless society everything is natural and supernatural at the same time. The distinction between 'sacred' and 'profane' came later. As any class of realia, such as plants, animals and natural phenomena including the planets, is magic, it thus has a magico-religious character whose earliest form manifests itself in totemism, in totemic relationships with various classes of

The Atlas Linguarum Europae and its Insights into the Cultural History of Europe

21

realia. In so-called primitive societies this is still observable today. This relationship assumes different manifestations, as will be shown later. The first to have proved that modern folk literature preserves very ancient myths and conceptions was Propp (1946/1987). His insights as well as Riegler's (1937/1987) are of great importance in interpreting the geodialectal data. These data show that the cultural history of Europe is not made up of random elements and events but follows a unified, well-structured pattern where three separate layers can be distinguished. The layer that can be recognised and dated most easily belongs to history, namely to Christianity and Islam. As this is the most recent level, it also occurs most frequently in the data. Within this layer Christian motivations appear much more often than Muslim ones, thus mirroring the difference in the areal spread of the two religions in Europe. In the pre-historical period two levels can be distinguished, one characterised by 'supernatural', 'superhuman' pagan figures and, leaving anthropomorphism, the other by still earlier zoomorphic and kinship representations. As early as 1929, Leo Frobenius, the founder of cultural morphology, noted "that a period of anthropomorphism must have been preceded by an older one of zoomorphism" (p. 248 f., translated from German). The basic structure has remained the same from pre-historical to historical times. While dating the first-mentioned layer is unproblematic, Alinei assumes "that the anthropomorphic representations of reality are connected with socially stratified societies, typical of the Metal Age, while zoomorphic and kinship representations are connected with more primitive societies of the Stone Age" (1997:27). Unlike vertical dead archaeological stratigraphies, linguistic stratigraphies as presented on ALE motivational maps are horizontal and all the above layers are still alive.

2.1. The Christian/Muslim Layer Curiously enough, it is not from a map such as 'Christmas' [...] that we can draw linguistic evidence for a Christian culture in Europe (except of course for the obvious diffusion - or absence in the Islamic areas - of the referent). The strongest evidence for the identification of a Christian (and Muslim) layer, spread all over Europe, comes from ALE motivational maps which have nothing to do with Christendom or Islam: those of the 'lady-bird' [...], o f other insects [...] and of the 'rainbow' [...]. Why these? Because the referents belong to two notional categories - wild animals and natural phenomena - which are often associated with magico-religious names (Alinei 1997 [1998]a: 3).

From the many examples only the following were selected due to restrictions of space. Among animals, designations of the smallest and weakest pig of a litter can be mentioned in this category. In England and Wales, apart from Daniel, Anthony(-pig) was elicited, sometimes as Tanthony, an incorrect separation of Saint Anthony. He was the patron saint of swineherds to whom the smallest pig of each litter was usually dedicated. The butterfly, too, is interesting. The OED surprisingly notes "The reason of the name is unknown" (s. v. 'butterfly'). However, in the Germanic area the belief was widespread that witches in the guise of butterflies stole butter, milk and cream. Compounds with butteroccur most often, see Dutch botervlieg 'butterfly', German Butterfliege 'butterfly' and English butterfly. Dutch boterhex, boterwijf, both 'butterwitch', clearly point to the belief in

22

Wolfgang Viereck

witches. However, the butterfly is christianised in Europe, too, mainly in the South, as 'little angel', 'little Easter', 'the pope's wife' show, but also in Finland as 'Brigit's bird'. Also the ladybird yields a rich harvest everywhere in Europe. Most commonly a Christian or Islamic religious being or notion is associated with another animal, such as a bird (cf. English lady-bird), a hen (Danish marihene 'Mary hen', Frenchpoulette au bon Dieu 'good God's young hen'), a cow (English lady-cow or cow-lady, Russian bozhia korovka 'God's cow', French vache a Dieu 'God's cow', Italian vacheta de la Madona 'young cow of the Holy Virgin'), an ox (Spanish buey de Dios 'God's ox'), a beetle (German Marienkäfer 'Mary beetle', English lady-bug) or, more generally, a little animal (Dutch [Onze] Lieveheersbeestje 'Our Lord's little animal'). The religious being or notion can be 'God' - only additional examples are listed below - (Spanish area de Dios 'God's chest', Breton [aotrou] doue 'God', 'The Lord'), 'angel' (Breton elik doue 'God's little angel'), 'Jesus' (Swedish Jesu vallflicka 'Jesus' shepherd'), '(Virgin) Mary' (Swedish jungfru marias nyckelpiga 'Virgin Mary's key servant', Italian anima de la Madona 'soul of the Holy Virgin', French bete de la vierge 'animal of the Holy Virgin', English Virgin Mary, Sunny Mary), or the names of saints such as, in Italy, San Martino, San Gioani, San Nicola, in France, Saint Jan, Saint Jacques, Sainte Catherine, in Spain, San Anton, and in Ukraine, Patrick.' In the Muslim area we find 'Allah', 'mosque' and 'Fatimah'. The glow-worm appears as 'Saint John', 'Saint John's bonfire', 'Saint John's light', 'Saint John's sparkle', for example, in Czech, Slovak, Slovenian, Corsican, Danish, German, Hungarian, Estonian, Latvian, Lithuanian, Nakho-Dagestanian, French and Spanish. For plants the religious motivations are more numerous. The pansy (Viola tricolor) may be called Heiliges Dreifaltigkeitsblümchen 'little Holy Trinity flower' in German and 'Anne's eyes' in Russian. The daffodil (Narcissus Pseudo-Narcissus) is Saint Peter's bell in Wales, and Saint Peter's herb is an expression for the cowslip (Primula veris) in parts of England. Among the plants named after Christian saints may also be noted Latin herba sancti Johannis 'Saint John's wort', German Johanniskraut 'John's wort', English Saint John's wort (Hypericum); Saint Bridget's anemone (Anemone coronarea), Saint Bruno's lily (Paradisea liliastrum), Saint Bernards lily (Anthericum liliago), Saint Dabeoc's heath (Dabaecia cantabrica), Saint George's beard (Sempervivum tectorum) and Saint Barnaby's thistle (Centaurea solstitalis). In English quite a number of plant names refer to the Virgin Mary, such as Our Lady's bedstraw (Galium), Lady's candlesticks (Primula), Our Lady's cushion (Armeria), Our Lady's hands (Lamium maculatum), Lady's mantle (Alchemilla), Lady's seal (Polygonatum multiflorum), Lady's smock (Cardamine pratensis), Our Lady's gloves (Digitalis purpurea), Our Lady's purse (Portulaca oleracea) and Our Lady's ribbons (Phalaris). The milk thistle (Silybum marianum) is Lady's thistle, Marian thistle, Mary thistle in English and Mariendistel 'Mary thistle' in German. Plant names can also change in the course of time. Examples in point are the wild clary (Salvia verbenaca) that was named Oculus Christi by William Turner, who published the first scientific botanical treatise in England in 1538 or Holy oke - also Turner's name - for today's marsh-mallow (Althaea officinalis). Many plant names in the British Isles occur with the motivation 'devil' as the EDD, s.v. 'devil' II. 2, reveals. The great reedmace, Typholia latifolia, and the milk thistle, Silybum The saints were (and are) very numerous. Quite a few of them were only of regional, if not local importance. The adoration of saints was the Christian concession to heathen polytheism.

The Atlas Linguarum Europae and its Insights into the Cultural History of Europe

23

marianum, are attested with 'holy', as is the Glastonbury thorn, Crataegus Oxyacantha, whereas the dandelion, Leontodon Taraxacum, and the rose-root, Rhodiola rosea, show a motivation with 'priest' in Britain and the woollv-headed thistle, Carduus eriophorus, one with 'friar' there. Archangel is a name applied in English to several species of the dead nettle and allied plants, such as red archangel (Lamium purpureum) and yellow archangel (Lamium galeobdolon). Alinei (1997) found basically the same motivations in Italy and provides many examples. Natural phenomena as well as planets also testify to a Christianisation and Islamisation in Europe. The classic example of the ALE is the rainbow - and not only for the most recent level but for the whole geolexical stratigraphy. Everywhere in Europe we find compounds with, for example, 'belt', 'bow', 'bridge', 'ribbon', 'ring' plus a religious motivation such as 'God's belt', 'Noah's bow', 'St. Baraaby's crown' or 'Allah's bow'. An example from Latvian (dieva juosta 'God's belt') must suffice here. Once the basic structure of the classificatory system had been worked out, it became clear that the rainbow had been considered sacred by European peoples and that with the advent of the new religions lexical innovations were coined expressing the same relationship that had existed earlier. Also the moon was once the object of religious veneration, still discernible in Hungarian istenkaläcsa 'God's cake'. The investigator labelled this form as 'humorous' and thus modern, which, of course, it is not. For thunder a Christian motivation was noted for Karelian, namely 'Holy Elias'. 2 Even Christmas belongs to the natural phenomena as the pre-Christian winter solstice underlies the Christian feast. Not unsurprisingly, therefore, only a few names for Christmas have a Christian motivation, such as 'Christ mass', 'Christ birth', 'Christ day', attested in German, English, Dutch, Spanish, Basque, Sardinian, Greek and Albanian. We even meet supernatural powers in christianised form. Thus the corn spirit appears in Switzerland as Michel (after Saint Michael); it is also named Oswald (after Saint Oswald) in Lower Bavaria.

2.2. The Anthropomorphic Layer The same notions that provide examples for the other layers can be drawn upon for this middle layer, which is both pre-Christian/pre-Islamic and post-zoomorphic. Animals provide quite a number of magico-religious names. For the weasel there is 'fairy' in English, 'witch' in French and Italian, 'Diana' in Sardinian, wildes Fräulein 'wild demoiselle', a member of the magic wilde Leute, in German and 'domestic genius' in Russian. Also taboo motivations belong here, such as Italian donnola 'little woman' or French belette 'little beautiful woman', both names for the weasel. The ladybird is associated with the Finno-Ugrian god Ukko 'the Old Man', in Frisian with the elf Puken 'puck', in Southern Italy with the elf Monachello, in Rumanian with Paparuga and 'witch' and in Greek with the Moira. The butterfly appears in Austria as 'the forest's elf and in Russian as babochka (from the goddess Baba 'Old Woman'). The grasshopper may be 'pregnant mother' and 'lady' in Italian and 'demoiselle' in French. These names point to "an earlier, no longer recognizable sacred female being" (Alinei 1997 [1998]a: 6). The motivation for the smallest pig of the litter in Ireland is 'fairy' (siog and siabhra). The same motivation is attested In Karekare in Africa, thunder can be analysed as 'God's cry' (Dymitr Ibriszimov, pers. com.).

24

Wolfgang

Viereck

in England for the glow-worm {fairy lanthorri), while 'witch' is noted there for the swallow (witch-hag). The pig is mucc in Old Irish and moch in Old Breton. These words derive from Moccus, a Gallo-Latin name of a god, and ultimately from Mercurius. In the case of plants, the motivation 'fairy' occurs in England for the cowslip (Primula veris - fairy cups), the dwarf or purging flax (Linum catharticum - fairy flax) and the purple foxglove (Digitalis purpurea) as fairy petticoats, fairy bell, fairy cap and fairy thimble. The latter is also attested in Ireland and Scotland together with fairy finger. The bending of the tall stalks of this plant is believed to denote the unseen presence of supernatural beings. 'Witch' is noted in English dialects, apart from the purple foxglove (witchf's] thimble), for the mountain ash, Pyrus Aucuparia (witch-beam), the harebell, Campanula rotundifolia, and the corn bluebottle or bluebonnet, Centaurea cyanus, both called witch bells, and the dandelion (Leontodon Taraxacum) is called witch-gowan in England and Scotland. In Ukrainian the dandelion is named kulbaba, referring to the goddess Baba O l d Woman' and to the figure old women may have. The EDD notes 'Jupiter' for the common house-leek (Sempervivum tectorum) as Jupiter's beard - the same motivation is attested in Italian for this plant. In both languages we find the motivation 'Venus' for the Dipsacus sylvestris (Genus's basin, Venus's bath in English). The monkshood (Aconitum napellus) is named Venus ' doves in English dialects. In Ukrainian and Russian the daffodil (Narcissus PseudoNarcissus) is called nartsis, in German Narzisse. More examples of mythical beings as name-givers of plants are, for example, Adonisröschen 'Adonis' or 'helleboraster' (Adonis vernalis) or Daphne (Daphne mezereum) in German. Also in English this plant is called daphne. In English Adonis and Andromeda are designations of whole genera of plants, N.O. Ranunculaceae and N.O. Ericaceae, respectively. In Grape Hyacinth (Muscari atlanticum Boiss and Reut.) the British recall the beautiful lover of Apollo from whose blood, the myth says, sprang the flower, and Apollo was also connected [...] with Mount Parnassus [...], whence Grass of Parnassus (Parnassa palustris). Iris, goddess of the rainbow, daughter of Electra, messenger of Zeus and his wife Hera, has a whole botanical genus to herself and various irises grow wild in the British Isles in addition to the many varieties imported for the gardens (Ashley 1974: 117 f.).

For the supernatural powers such as the corn spirit we also encounter anthropomorphic motivations such as, in Ireland, carlin, seanbhean, both meaning 'Old Woman', old maid, (old) hag, cailleach 'old hag', also meaning 'Old Woman'. Cailleach and carlin occur in Scotland too. Moreover, maid(en) is noted in both countries for this notion, as is Kornjungfer 'corn maid' in German. A mythical 'Old Man' (der Alte, der Kornalte, der Kornmann) is widespread in Germany, as is a mythical O l d Woman' (die Alte, altes Weib, Gerstenweib 'barley woman') (for more examples cf. BeitI 1933/1987). In parts of the adjoining Slavic area the corn spirit is called baba 'Old Woman' and stary 'Old Man' respectively. Among natural phenomena and planets, the rainbow has anthropomorphic representations everywhere in Europe. In the Turkic area they are associated with Kalmuk Taengri, in the Uralic area with Ukko and Tiermes, in the Indo-European area with Laume (in the Baltic region), Iris, 'Old Woman' (in the Romance and Slavic regions) and Soslan in Ossetian, often together with 'bow', 'belt' or 'ribbon'. For thunder as well as for lightning one encounters the Celtic storm god Taranus, Germanic Thorr, Lithuanian Perkunas, the FinnoUgric Ukko, the Lapp Tiermes and the Slavic Perun. Names for cloud can be motivated by 'Old Man', as, for example, in Swedish. For the moon we find 'Old Man' in the Nenets

The Atlas Linguarum Europae and its Insights into the Cultural History of Europe

25

area and 'hoary Old Man' in Ostyak and for the sun there is the sun god Yarilo in Russian and in Ukrainian.

2.3. The Zoomorphic Layer In the most archaic layer, i.e. the zoomorphic and totemic layer characteristic of egalitarian societies, the realia investigated appear in the form of either an animal or a kinship name. As this is the oldest layer, the evidence is, quite naturally, less overwhelming than for the two younger layers. Starting with supernatural magico-religious beings, an appropriate example would be the last com sheaf cut at harvesting into which the vegetation demon, it was believed, retreated. In Ireland we find granny 'grandmother' and in Germany Mutter 'mother', also in the compounds Erntemutter 'harvest mother', Kornmutter 'corn mother', moreover Kind 'child', also as Erntekind 'harvest child', Braut 'bride' and Große Mutter 'grandmother' as designations for this notion.The last corn sheaf was also named after those animals in the appearance of which this demon was imagined, namely Bär 'bear', Bock 'buck', Hase 'hare', Rind 'ox', Kuh 'cow', Geiß 'goat', Hahn 'cock', Wolf ' w o l f and Kater 'tomcat' in German. We find the same picture in the neighbouring Slavic area (see Beitl 1933/1987). 'Cat' is also attested in French (chat). Gi(o)rria 'hare', hare's bite/sheaf/seat/tail, rabbit, cow, hog, piardog 'crayfish' and swallow occur in Ireland. Hare is also noted in parts of England and Scotland for the last corn sheaf. Coming to animals, Riegler (1937/1987) had already interpreted wild animals and insects as relics of a totemistic view of the universe in which they would be our closest relatives. This relationship, similar to kinship, is consequently expressed by kinship terms. Propp (1946/1987) noted that the totem animal in its original form is embodied by the 'mother' and by matrilinear kin. This is indeed what we most often find in European dialects.' Many kinship names were recorded for the ladybird: 'grandmother' in, for example, Polish, Russian, Serbo-Croatian (Bosnian, Serbian, Croatian), Mordvinian, Udmurtian, Finnish and Komi-Zyrian, 'mother' in, for example, Rumanian, Belorussian, Tatar, Bashkirian and Livian, 'aunt' in German and Italian, 'bride and spouse' in, for example, Turkish, Albanian, Macedonian, Italian, 'sister-in-law' in Bulgarian. 'Grandfather' occurs in Swedish, Komi-Zyrian and Maltese and 'uncle' in Albanian for the same notion. The butterfly as a relative appears as 'grandmother' in Rhaeto-Romance mammadonna, as 'mother' in German and Sardinian and as '(grand)father' in the Uralic area. Kinship names for the weasel abound: ('little') 'bride' is attested, for example, in Turkish, Bulgarian, Rumanian, Italian, Greek, Albanian and German, 'godmother', for example, in Galician and in Spanish, 'daughter-in-law' in Portuguese, Occitan, Italian, Turkish and '

Thus the E D D attests twenty plant names with 'mother' and six with 'grandmother', but only one with 'father' and none with 'grandfather'. We find the same picture with names of diseases as the following examples show: fever appears as 'godmother' in Serbo-Croatian (Serbian, Croatian, Bosnian) and Russian, moreover in Russian as 'mother', 'little sister' and 'aunt', a disease of the skin is 'little grandmother' in Russian, plague is 'godmother' in Bosnian, Serbian and Croatian and smallpox appears as 'grandmother' in Russian, in Yakut as 'little aunt' and in Yukaghir as 'little mother' (cf. W. Viereck, Κ. Viereck 1999). On the controversy over patriarchy and matriarchy in antiquity cf. Wagner-Hasel (1992).

26

Wolfgang Viereck

Hungarian, 'mother' in English (see Old English beomodor 'beemother') and 'godfather' in German. Many more examples of this type can be cited. Thus the bear is called 'mother', 'father' and 'grandfather' by Turkic and Tartar peoples and 'dear grandfather' by the Swedes. The Hungarians call it 'godfather' and the Lapps 'clever father'. The wolf appears as 'little brother' in Ukrainian and Russian and the fox as 'godfather' in German (vaddermann voss 'Mr godfather fox' or Herr gevatter 'Mr godfather'), as 'little sister' in Ukrainian and Russian and as mon cousin 'my cousin' in French. Parent 'relative' is a name for cuckoo in French and the toad is called grossmudder 'grandmother' in Low German. It must be interpreted as a sign of prehistoric totemism when tribes or their leaders were given names of animals. The leaders of the Jutes Hengist 'stallion' and Horsa 'horse' or the leader of the Goths Berige 'bear' are cases in point, as are the Germanic Wylfmgas ' w o l f , the Italic Hirpi (from Latin hirpus ' w o l f ) and the Piceni (from Latin picus 'woodpecker'). Compared with animals, plants do not seem to play the same role in totemism. Some plants are given kinship names, others are associated with animals. The pansy (Viola tricolor) is called Stiefmütterchen 'little stepmother' and Stiefkind 'stepchild' in German (Vetvichki 1987) and the dandelion (Leontodon Taraxacum) in Swedish mammpap(p)or 'dad of m a m m a ' = 'grandpa' and in Belorussian 'grandma' and 'grandpa'. In Ukrainian the pansy has both kinship and animal names: 'brothers', 'brother and sister', 'orphan' and 'cuckoo birds' (Mamchur 1987). For England the EDD lists many plant names with, for example, the motivation 'pig', 'fox', 'goat', 'toad', 'cat' and 'horse'; see Alinei (1997: 25 f.) for corresponding Italian plant names. However, the magico-religious belief is not always clear. Thus, the resemblances between the appearance of the leaves and the claws of birds seem to have led to the English designations bird's foot (Ornithopus), crowfoot (Ranunculus) and goosefoot (Chenopodium). As to natural phenomena and planets, the moon is called 'grandfather' in Nenets and thunder is called 'father' and 'grandfather' in the Finno-Ugric area. These relationships are clearly totemic. In this class of realia animals occur rather often. For the rainbow we have 'dragon', 'snake', 'ox', 'cow', 'fox', 'drinking animal' in many European languages and dialects." Other zoomorphic representations appear with thunder, namely 'dragon' and 'serpent' and with lightning ('whale' and 'dolphin'). Mist is associated with the ' f o x ' and the ' w o l f in France and Germany, the 'eagle' with storms in Northern Europe and the 'cat' with the air trembling with heat in parts of Germany.

It is worth mentioning that the rainbow as a drinking animal is attested also b e y o n d Europe in Japan and in China. In Chinese culture it is a double-headed dragon drinking water o n both s i d e s o f a river.

The Atlas Linguarum Europae and its Insights into the Cultural History of Europe

27

3. C o n c l u s i o n

In the process of the cultural development of Europe we thus find recurrent structural patterns: the same reality was first given kinship and zoomorphic names to be followed by anthropomorphic names and finally by Christian and Islamic names - and this across all language and dialectal borders. Also names of diseases and designations for bread follow the same pattern (cf. W. Viereck, Κ. Viereck 1999 and Viereck forthcoming b). 5 The collected data could be the core of a comparative motivational dictionary. The three periods mentioned, of course, do not end and begin abruptly. Each one of them lasted for thousands of years. Archaeological finds show that also between the Stone Age and the Metal Age there were fluid transitions and that anthropomorphic representations also from the Neolithic Age were found (Müller-Karpe 1998). That the transitions and overlaps between the pagan and Christian layers can be documented much better is due to their being much closer to our time. Up to the early 4th century AD the early Christian church was an underground church and it took many centuries until the Christian faith had penetrated everywhere in Europe. Only in the 8th century AD was the Indiculus superstitionum et paginiarum written in either Fulda or Mainz with instructions on how to deal with pagan cults, spells and fortune-telling (Müller-Kaspar 1996: 419). With new religious beliefs a wave of new designations followed, yet the old conceptions often remained the same. To take just one example out of many: When Christianity came to Britain, the bright yellow flowers of the plants in the Hypericum family that had been associated with the golden brightness of Baldur the sun-god came to be called St. John's wort, as Baldur's Day became St. John's Day. The plant continued to be thought a cure for wounds and on St. John's Eve good Christians wore a sprig of it to ward off evil spirits and especially to protect themselves against the stray thunderbolts of the gods (Ashley 1974:116).

Saint John's Day is the Christian equivalent of the summer solstice, one of the most important events in prehistoric times. The quotation shows that pagan thought was alive and well in the early Christian period. However, examples of this can easily be found today: The initials of Caspar/Kaspar + Melchior + Balthasar + the year are still written on the front doors of people's houses in Catholic areas in Germany and in Poland on Epiphany, January 6, to protect the people from evil of any kind and drivers hang small pictures of Christopher in their cars for protection in many countries, such as Ukraine and Germany. Apparently Enlightenment had no effect on people's piety. 6 The ALE is naturally based on European dialects. The adopted motivational approach has uncovered some important pieces in the mosaic of the cultural development of Europe. Their implications, no doubt, transcend the frontiers of the European continent. In light of the complementary nature of world cultures it would be highly desirable if the picture presented were complemented by insights gained in other cultures.

5

''

Unfortunately the ALE has only one notion, a cold, in the first-mentioned area and none in the second. The explanation, sometimes advanced, that this custom refers to the biblical passage "Christus mansionem benedicat" rather than to the three holy kings is highly improbable. This possibility is not even alluded to in such reference works as Erich, Beitl (1981), i.v. 'Dreikönig'.

28

Wolfgang Viereck

References

(a) Cited Dictionaries EDD = THE ENGLISH DIALECT 1905. OED = THE OXFORD ENGLISH Clarendon Press 2 1989.

DICTIONARY.

Ed. Joseph Wright. 6 vols. London: Henry Frowde 1898-

DICTIONARY.

Comp. J. A. Simpson, E. S. C. Weiner. 20 vols. Oxford:

(b) Other Literature Alinei, Mario (1996): Origini delle Lingue d'Europa. Vol. 1: La Teoria della Continuitä. Bologna: II Mulino. - (1997): Magico-religious Motivations in European Dialects: A Contribution to Archaeolinguistics. Dialectologia et Geolinguistica 5, 3-30. - (1997 [1998]a): The Atlas Linguarum Europae after a Quarter Century: A New Presentation. In: Mario Alinei, Wolfgang Viereck (eds.): Atlas Linguarum Europae. Perspectives nouvelles en g6olinguistique, 1-40. Rome: Poligrafico. - ,Wolfgang Viereck (eds.) (1997 [1998]b): Atlas Linguarum Europae. Vol.1 - cinquifeme fascicule: Commentaires and Cartes. Rome: Poligrafico. Ashley, Leonard R.N. (1974): Uncommon Names for Common Plants'. The Onomastics of Native and Wild Plants of the British Isles. Names 22, 111-128. Bächtold-Stäubli, Hanns, Eduard Hoffmann-Krayer (eds.) (1927-1942, reprinted. 1987): Handwörterbuch des deutschen Aberglaubens. 10 vols. Berlin: de Gruyter. Beitl, R. (1933, repr. 1987): Korndämonen. In: Bächtold-Stäubli, Hoffmann-Krayer. Vol.5, 250314. Cavalli-Sforza, Luigi, Albert J. Ammerman (1984): The Neolithic Transition and the Genetics of Populations in Europe. Princeton, N.J.: Princeton University Press. Donini, Ambrogio (1977): Enciclopedia delle religioni. Milano: Teti. - (1984): Lineamenti di storia delle religioni. Roma: Editori Riuniti. Erich, Oswald Α., Richard Beitl (' 1974, repr. 1981): Wörterbuch der deutschen Volkskunde. Stuttgart: Kröner. Frobenius, Leo (1929): Monumenta Terrarum. Der Geist über den Erdteilen. 2nd edition of Festland kultur. Frankfurt: Buchverlag. Gamkrelidze, Thomas V., Vjaceslav V. Ivanov (1995): Indo-European and the Indo-Europeans. 2 vols. Berlin: de Gruyter. [English translation of Indoevropejskij jazyk i indoevropejcy. 2 vols. Tbilisi: Izdat. Tbilisskogo Univ. 1984.] Mallory, J. P. (1989): In Search of the Indo-Europeans: Language, Archaeology and Myth. London: Thames and Hudson. Mamchur, Fedir (1987): Zviv na polonyni veres. Kiev: Molod. Müller-Karpe, Hermann (1998): Grundzüge früher Menschheitsgeschichte. 5 Bde. Darmstadt: Wissenschaftliche Buchgesellschaft. Müller-Kaspar, Ulrike et al. (1996): Handbuch des Aberglaubens. 3 Bde. Wien: Tosa. Propp, Vladimir (1987): Die historischen Wurzeln des Zaubermärchens. München: Hanser. [German translation of Istoriceskie korni volsebnoj skazki. Leningrad'. Izdat. Leningradskogo Gosud. Univ 1946] Renfrew, Colin (1987): Archaeology and Language: The Puzzle of Indo-European Origins. London: Cape. Riegler, R. (1937, repr. 1987): Tiergestalt and Tiernamen. In: Bächtold-Stäubli, Hoffmann-Krayer. Vol. 8, 819-842 and 863-901.

The Atlas Linguarum Europae and its Insights into the Cultural History of Europe

29

Shevoroshkin, Vitaly V. (ed.) (1989a): Reconstructing Languages and Cultures. Bochum: Brock meyer. - (ed.) (1989b): Explorations in Language Macrofamilies. Bochum: Brockmeyer. - (ed.) (1990): Proto-languages and Proto-cultures. Bochum: Brockmeyer. - (ed.) (1992): Nostratic, Dene-Caucasian, Austric, and Amerind. Bochum: Brockmeyer. Sokal, Robert R., Neal L. Oden, Barbara A. Thomson (1992): Origins of the Indo-Europeans: Genetic Evidence. Proceedings of the National Academy of Sciences of the United States of America 89, 7669-7673. Stearn, William T. (1975): History of the British Contribution to the Study of the European Flora. In: S.M. Walters (ed.): European Floristic and Taxonomic Studies, 1-17. Faringdon: Classey. - (1992, repr. 1996): Steam's Dictionary of Plant Names for Gardeners. London: Cassell. Sterck, A.A. (1987): Paardebloemen: planten zonder vader; variatie, evolutie en toepassingen van het geslacht paardebloem (Taraxacum). Utrecht: Stichting Uitgev. KNNV. Turner, William (1538): Libellus de re herbaria novus. Ed. with translation into English by Mats Ryden, Hans Heiander, Kerstin Olsson. Uppsala: Swedish Science Press 1999 (= Acta Societatis Litterarum Humaniorum Regiae Upsaliensis 50). Vetvichki, Vazlav (1987): Rasteniä lesov ipolej. Prague: Artia. Viereck, Wolfgang (Hg.) (forthcoming a): Atlas Linguarum Europae. Vol.1 - sixieme fascicule: Commentaires and Cartes. Rome: Poligrafico. - (forthcoming b): Bemerkungen zur Kulturgeschichte Europas. Juznoslovenskifdolog 56. [Issue in memory of Pavle Ivic] - Karin Viereck (1999): 'Die seltzamen namen all'. Zu einigen Ergebnissen des Forschungsprojekts Atlas Linguarum Europae. - In: Eun Kim et al. (eds.): Aktive Gelassenheit. Festschrift für Heinrich Beck zum 70. Geburtstag, 711-723. Frankfurt: Peter Lang (= Schriften zur Triadik und Ontodynamik 17). Wagner-Hasel, Beate (Hg.) (1992): Matriarchatstheorien der Altertumswissenschaft. Darmstadt: Wissenschaftliche Buchgesellschaft (= Wege der Forschung 651).

Tove Bjorneset The Dictionary Project NORDLEXIN-N

1. Introduction

In Sweden, bilingual dictionaries suitable for teaching the Swedish language to immigrants have been developed since the late 1970s. The initiative to this was taken and funded by the Swedish government. The background for the project was a decision made by the Swedish parliament that Sweden should promote active bilingualism and education in the Swedish language for immigrants. The project was named Spräklexicon för innvandrare (LEX-IN) and the basic development work was completed in 1984. The Swedish National Agency for Education has been in charge of the project since 1991. The LEXIN dictionaries are lucid and user-friendly, making them a useful tool also for immigrants with limited reading proficiency even in their own mother tongue and perhaps no previous experience in using a dictionary. The entry words are selected with the special needs of the target group in mind. In addition to words selected from the general vocabulary, the dictionaries contain more than 1000 so-called social terms and expressions accompanied by exhaustive definitions (encyclopedic information). These entry words are considered to be of particular importance for immigrants and refugees in their encounter with an unfamiliar environment. The dictionaries are published in various sizes: MINI (approx. 4500 entries), MIDI (approx. 8000 entries), STORA (approx. 17000 entries) and MAXI (approx. 28000 entries). The translations made so far include approx. 20 languages. Among these are Albanian, Arabic, Armenian, Bosnian, English, Finnish, Greek, Croatian, Macedonian, Persian, Polish, Serbian, Somali and Tigrinya. The number of lemmas in these dictionaries vary, and more dictionaries are forthcoming, both as printed versions and Internet versions. The dictionary entries contain information on parts of speech, inflectional forms written out in full, definitions and/or synonyms, formalised syntax notation and phonetic representations. Many entries also include comments on style and grammar and examples of actual usage, set phrases and idioms. There are illustrations accompanying approximately 1700 lemmas. These are divided into 33 thematic classes, such as "family and relations", "the human body", "weights and measures", "city and traffic" and "school and education". The illustrations are included in the dictionaries in the form of an appendix. They are also published as a separate booklet, Bildtema. In the early 1990s all Nordic countries were offered free use of the Swedish LEXIN database to facilitate the production of similar dictionary series. In the autumn of 1996 the Ministry of Education, Research and Church Affairs (KUF) assigned the Norwegian NORDLEXIN-N project to the Norwegian Term Bank, a section of the HIT Centre (Humanities Information Technologies) at the University of Bergen. In the initial project phase the mate-

'Dictionaries for immigrants'.

32

Tove Bjorneset

rial from the MINI edition was transferred to Norwegian. In June 1997 work started on transferring the STORA edition, a process that lasted throughout 1998. In 1999 KUF left the overall responsibility for the project to the National Centre for Educational Resources (NLS). At the end of 1999 the Norwegian database equalled the Swedish database in size. The Swedish illustrations were made quite a few years ago, many of them reflecting features peculiar to Swedish society. To ensure an up-to-date and culturally correct result, the illustrations for the Norwegian version will be made from scratch, but in accordance with the thematic subdivisions in the Swedish version.

2. Transfer of Swedish Source Material to Norwegian

The most characteristic feature of the N O R D L E X I N - N project is perhaps that it is the actual source language material which was transferred from Swedish to Norwegian. This work could roughly be divided into two phases: - Establishing equivalences. - Making the source material suitable for Norwegian conditions.

The transfer of the Swedish material was carried out by mapping and establishing equivalences between the Swedish and the Norwegian dictionary records. The central part of the work consisted in clarifying and establishing the equivalence for the lemma field in the records. In those cases where no equivalent could be found, the dictionary record was inactivated.

2.1. A Few Words about Equivalence Equivalence can be defined as semantic and functional correspondence between words or expressions in two or more languages. Equivalence is a very central concept in bilingual lexicography, where the relationship between a lexical quantity in the source language and the corresponding semantic and functional quantity in the target language is described. A bilingual dictionary aims at providing the best possible target-language equivalents for words and expressions in the source language. It is customary to distinguish between three levels of equivalence: Full equivalence, i.e. complete correspondence in sign content and semantic range, is a rare phenomenon in the general language, as opposed to LSP ("language for special purposes"). The reason is that the conceptual world is different from language to language, among other things as a result of differences in historical conditions and culture. Despite this fact, full equivalence has been achieved to a considerable extent for the Swedish material. In the case of partial equivalence, source words and target words only correspond to a certain extent. The difference often manifests itself in the fact that the underlying concept of the word in the target language has fewer characteristic features than the underlying

The Dictionary

Project

33

NORDLEXIN-N

concept for the word in the source language. The concept in the source language will sometimes function as a superordinate concept. Zero equivalence also occurs, and then usually with culture-specific concepts, such as "De Aderton" (i.e. members of the Swedish Academy).

2.2. Establishing Equivalences between Swedish and Norwegian The objective of establishing equivalences to the Swedish material is to utilise the translations to the minority languages already electronically available in the Swedish database. A major part of the translation work from Norwegian can then be carried out automatically. In order to achieve this, all Norwegian equivalence fields must be linked to the Swedish dictionary records and the fields they are related to. According to the contract, we should concentrate on establishing equivalences for the lemmas and the idiomatic expressions, as these fields are the only ones which are always translated to the entire set of target languages in the Swedish dictionaries. Despite the fact that not all fields have been translated to the target language in all LEXIN dictionaries, we have chosen to establish equivalences in all fields whenever possible. There are two reasons for this: there are considerable discrepancies as to which fields have been translated in the various Swedish dictionaries. Moreover, it is important for a possible future Nordic joint database that the translation of the Swedish source material to Norwegian should be as complete as possible. During the first phase, establishing equivalences, we made use of the Swedish material to the fullest possible extent. Consequently all lemmas, phrases, idioms and compounds were translated. All equivalated and inactivated dictionary records were checked and approved by a native Swedish speaker and philologist with competence also in Norwegian. Initially all equivalences were established. At a later stage, however, equivalences for lemmas which were considered to be irrelevant for Norwegian conditions, were inactivated in the Norwegian database. In adapting the database to Norwegian conditions, we had to consider things independently of the Swedish material. This was essential in order to obtain dictionary material which reflects the peculiarities of Norwegian society. In this phase the equivalated database was systematically supplemented with additional Norwegian records. Finally, a selection of inflectional forms was included in the records, as well as phonetic representations of the lemmas.

2.3. Quantitative Dimensions and Equivalence Percentage The Norwegian database NORDLEXIN-N covers the same number of concept as the Swedish LEXIN, maybe even more, despite the fact that LEXIN covers more lemmas. The difference is mainly a consequence of synonymy and concepts peculiar to Swedish society with no counterpart in Norwegian. Allowing for possible inaccuracies in the list below, there can be no doubt that the percentage of equivalence between the Swedish and the Norwegian databases is high: Equivalence, lemmas Equivalence, compounds

76% 79%

Equivalence, examples Equivalence, idioms

88% 87%

34

Tove

Bjorneset

3. T h e N o r w e g i a n Dictionary Records

A dictionary record in the Swedish LEXIN database contains, in addition to the actual lemma, many fields describing and exemplifying the lemma, such as definitions, phrases, idioms and compounds. The stylistic level of a large number of entries is indicated in separate fields. In NORDLEXIN-N these categories have been standardized to e.g. Formal, Informal, May cause offence, Swearword and Child language. The Norwegian record format is based on the Swedish one. Although we have elaborated a little on this format, the Norwegian and Swedish formats are still fully compatible. All equivalated dictionary records have in principle the following structure: Field Name

Content

2

S-lem

Swedish lemma

E-lem

Equivalated lemma

S-def

Swedish definition/synonym

E-def

Equivalated definition/synonym

N-def

(Replacive Norwegian definition)

S-eks

Swedish phrase

E-eks

Equivalated phrase

N-eks

(Additional or replacive Norwegian phrase)

S-idi

Swedish idiom

E-idi

Equivalated idiom

N-idi

(Additional or replacive Norwegian idiom)

S-sms

Swedish compound

E-sms

Equivalated compound

N-sms

(Additional or replacive Norwegian compound)

S-mor

Swedish inflected forms

E-mor

Norwegian inflected forms

S-utt

Swedish phonetic representation

E-utt

Norwegian phonetic representation

All Norwegian equivalents are registered in the Ε-fields. These fields will at a later stage be electronically matched with the corresponding Swedish fields. Where no equivalence relationship has been established for the Swedish lemma (the S-lem field), the dictionary record 2

The letter S serves as a prefix to all Swedish fields. The letter Ε serves as a prefix to all Norwegian equivalated fields. The letter Ν serves as a prefix to all Norwegian fields established from scratch.

The Dictionary Project

NORDLEXIN-N

35

has been inactivated, which means that the record has been removed from the Norwegian version of the database. Lacking Norwegian equivalences for other fields have been inactivated by the insertion of a hyphen. Alternatively, additional or replacive Norwegian fields (N-fields) have been established, with a content which is not linked to any Swedish field, and which therefore will have to be translated manually to the target languages. The dictionary record below illustrates a rather frequent phenomenon: the lemma has been equivalated, but the Swedish definition field has been inactivated. We considered the ability to be smashed or broken as an essential characteristic of glass, and wanted to include this as part of the definition. Whenever we have made semantic changes of this kind, a separate N-field has been established: Field Name

Content

S-lem

glas

E-lem

glass ('glass')

S-kat

subst.

E-kat

subst ('noun')

S-def

ett härt och genomskinligt ämne

E-def ('a hard and transparent substance') N-def

et hardt, krtusbart og gjennomsiktig materiale ('a hard, breakable and transparent substance')

The field S-def merely gives "hard" and "transparent" as essential characteristics, whereas the N-def has added the word "knusbart" to the aforementioned characteristics to include the characteristic "breakable".

4. A d j u s t m e n t s for N o r w e g i a n Conditions

Swedish and Norwegian are not as similar as we are inclined to believe. Apparently identical words in the two languages may well turn out to be so-called false friends or conceal important shades of meaning and semantic differences. Words that are synonymous in one language may well have clearly distinct meanings in the other. Differences of these and similar kinds have, as already mentioned, in some cases made it necessary to establish new Norwegian records.

36

Tove Bjernesel

4.1. Norwegian Dictionary Records established from Scratch Social and administrative institutions are often not directly compatible with the Swedish ones. A number of terms and words relating to Swedish society (institutions, social structure, culture, etc.) have been inactivated in the Norwegian database. To compensate for this, records for comparable and related Norwegian concepts have been added to the database. Furthermore, we have found it necessary to include a number of terms and expressions referring to phenomena which either did not exist or were only of peripheral interest at the time when the Swedish LEXIN database was established. Among these are expressions relating to social aspects from the last twenty years or so. We have also included terms and expressions from miscellaneous disciplines and social areas which we felt should receive more emphasis. This material was collected from pamphlets and leaflets issued by municipal institutions, pharmacies, banks, post offices, employment agencies, public communication companies, etc. This has been done in order to capture words and expressions frequently encountered in everyday life. We have also included important words and terms relating to the various disciplines taught in compulsory school, the structure of the Norwegian educational system and similar matters. Some expressions covering the most common leisure activities have also been included. With the exception of the major political parties and organisations, we have decided to omit information which is clearly transient. In 1999 we drew up a list of approximately 800 terms covering information technology and data processing. 200 of these have so far been incorporated in the database. As NORDLEXIN-N is not meant to be an LSP dictionary, terms which are relevant for specialists only have not been included. The emphasis has been on useful terminology for the non-specialist end users of text-processing systems, spreadsheets, e-mail and the Internet. This terminology was collected from various sources on the Internet, among these a list from the Norwegian Language Council, and from the latest edition of the largest Norwegian data/computer dictionary, Norsk dataordbok (Hofstad, Leland and Scott 1997). Much effort was invested in collecting synonymous terms for the various concepts. Synonymy is a wellknown phenomenon for all subject fields characterized by rapid changes and far-reaching development. For each set of synonyms we have selected a preferred term according to terminological principles. References have been made from the remaining synonyms to the preferred terms.

4.2. Examples Examples are often overlooked in the discussion of lexicography. The reason may be that they are considered to be less important than definitions. Examples will not only supplement the information given in the definition of the lemma, but also display the lemmas in their natural contexts. Examples also show how different meanings may be distinguished from each other. In this respect examples are illustrative, but their number should not be too small, lest problems relating to analogical thinking should arise. Examples illustrate grammatical patterns, and may also show other typical collocations. Examples are particularly useful in bilingual dictionaries. Lemmas can rarely be translated in isolation, without a context - certain verbs, for instance, require a preposition etc.

The Dictionary Project

N0RDLEX1N-N

37

The objective is partly to show the contrasts between different languages, ranging from subtle differences in e.g. the use of articles to the use of different parts of speech to express the same content in the two languages. In such cases examples are an indispensable aid in discovering translation equivalents which can be used in real-life situations, and not only in meta-language, in definitions, etc. The examples should show that the translation equivalent is not applicable in all contexts. There should also be examples showing how the lemma should be translated in those cases where no satisfactory equivalent can be found. All these considerations contributed to the decision to give priority to and room for as many examples as possible in the NORDLEXIN-N dictionary project. There is a natural distinction between dictionaries with the user's mother tongue as target language and dictionaries which go the other way. Bilingual dictionaries for Norwegians with Norwegian as target language are primarily an aid for understanding text and speech. These are often referred to as passive dictionaries. The Norwegian user will most likely already know the context of the word he is trying to look up; what he needs is a listing of as many translation variants as possible in order to be able to pick out an optimal equivalent. As a rule there is no need for examples in dictionaries of this kind, with the exception of fixed word combinations and idioms - more or less fixed expressions the meaning of which does not reflect the meaning of the separate words. Bilingual dictionaries for Norwegians with Norwegian as source language are primarily an aid for producing text or speech. These are often referred to as active dictionaries, and are perhaps the most important tools for translators. Translators will often also use monolingual dictionaries, but there will always be information that can only be found in a bilingual dictionary. The reason is that a comprehensive bilingual dictionary is in fact a contrastive grammar with more grammatical information than ordinary grammar books. The dictionary is, however, an implicit grammar; there are neither rules nor systematic descriptions. It is from the examples given in the dictionary entries that the contrastive nature of the dictionary emerges. The NORDLEXIN-N dictionaries will be bilingual dictionaries with Norwegian as source language and the mother tongue of the users as target language. This is because our starting point is the electronic material underlying the Swedish LEXIN dictionaries. There is a strong wish both in Norway and Sweden for developing reversed versions of the dictionaries. This will, however, depend on available resources and will be considered at a later stage.

4.3. Set Phrases and Idioms We have systematically selected and registered additional set phrases and idioms, as such linguistic phenomena are considered to be of great importance for achieving fluency in a foreign language. Such expressions also give clues to the understanding of different cultures and ways of thinking. We have so far included more than 1000 extra expressions of this kind.

38

Tove

Bjerneset

4.4. The Norwegian Textbook Standard The Norwegian textbook standard is, roughly speaking, a collection of alternative written forms prescribed by the Norwegian Language Council, thus being the forms with the highest status in Norwegian spelling and inflection. There are also other forms which are acceptable as far as spelling and inflection go - the so-called alternative or square bracket forms. All public affairs, such as compulsory school and the Norwegian Broadcasting Corporation (NRK) are obliged to follow the textbook standard. There are separate textbook standards for Nynorsk and Bokmäl. These standards are by no means static; they are being continuously evaluated by the Norwegian Language Council. The perhaps most important target group for NORDLEXIN-N, is pupils in compulsory schools. According to the contract, all alternative forms of the lemmas which comply with the textbook standard for Bokmäl shall be included in the records. Despite this explicit statement in the contract, there were several options as to how this should be implemented in practice. In which order should the various forms be presented? Should an internal project standard be established for the contents of the lemma and definition fields? And if so, what should it be? There was almost no end to possible solutions. Any solution we would end up with would still be unsatisfactory and open to criticism due to individual preferences. We decided that we could not stick to only one variant of Bokmäl. In dictionary projects like this it is important not to suppress the wide range of variation within the written aspect of Bokmäl; our aim should be to make these variations as visible as possible. Nevertheless, we decided to establish a project standard to be used in the lemma field and in the fields where the lemma is explained. This project standard is much in line with relatively moderate or conservative Bokmäl. In the remaining fields we have deliberately alternated between the permissible forms within the textbook standard. Inconsistencies have not been seen as a problem, but as for Bokmäl generally, concrete nouns and certain stylistic contexts tend to motivate the choice of feminine nouns and other "radical" forms. In the last resort, this is just a matter of individual judgement. All alternative forms within the textbook standard are included in the field for morphological information. In the case of alternative forms of equal status, which is a quite typical phenomenon, these are placed in parentheses. In the Swedish database there are approximately 1000 records with one or more alternative forms, whereas the corresponding number for the Norwegian database is more than 6000. The number of alternatives is also significantly higher, especially concerning the compounds. The multitude of alternative forms in Norwegian unfortunately makes this form of representation somewhat less well-presented as compared with the Swedish version.

5. Phonetic Representation

There is no official standard for Norwegian pronunciation, nor any particular pronunciation with a higher status than others. Speaking one's own dialect has become widely accepted,

The Dictionary

Project

39

NORDLEXIN-N

irrespective of the social or geographical context or the social standing of the speaker. What is perhaps closest to what we might call standardized pronunciation, is the pronunciation used by actors performing in theatres and similar places, but this pronunciation will often sound a bit stiff and unnatural to most people. Presenters working for the Norwegian Broadcasting Corporation are obliged to speak either Bokmäl or Nynorsk, all the same the pronunciation differs considerably as a consequence of the dialectal background of the individual presenters. For these reasons we found it impossible to establish an accurate phonetic representation of Norwegian. To be a functional aid for the target group, such representation must necessarily be rather general in order not to be an outright misrepresentation of a great number of dialects. All the same, the phonetic representation must provide more information than what can be deduced from spelling conventions. Whenever we have had to choose between two different, geographically conditioned pronunciations, we have chosen the one predominant in the eastern part of Norway, where the greater part of our target group actually live. Whenever two different pronunciations are found more or less in the entire country, both have been included. We have not used IPA characters for the phonetic transcriptions, as these may be felt to be too unfamiliar and difficult for our primary target group. To compensate for the lack of these special characters, the various sounds have been represented by means of the most common letter or letter combination used for the individual sounds in ordinary spelling. Each phone is usually transcribed as one letter.

5.1. Diphthongs The diphthongs (e.g. a\i and o\y) will be marked by an arch indicating that the two vowel characters are used to represent one sound.

5.2. Quantity Quantity is represented by a colon succeeding the long vowel. The quantity character is not used for consonants because consonant length is predictable from vowel length: jeg mä lete etter boka mi ( Ί have to look for my book') flyet skal snart lette ('the plane is about to take o f f )

[le:te] [lete]

5.3. Stress Primary stress is indicated by a dot underneath the vowel. Secondary stress and stress in monosyllabic words have not been indicated.

Tove Bjorneset

40 5.4. Tönernes There are two tönernes in Norwegian. The pronunciation is not the same for: 1. tanken (definite singular form of the noun 'tank'), and 2. tanken (definite singular form of the noun 'thought')

This difference has not been formally marked in the transcription. The reason is that there are great differences between the various dialects as to which toneme is used for a particular word. There are also dialects with no tonemic opposition.

6. Parts of Speech

The traditional division of words into parts of speech has a long tradition in Norway. This division is, however, to a great extent based on semantic criteria at the expense of morphological and syntactical criteria, making it less attractive for linguistic purposes. Norsk referansegrammatikk (Faarlund et al. 1997) attempts a consistent approach by applying identical criteria for establishing the various parts of speech. Semantic criteria are avoided whenever possible. For some parts of speech there are no changes. The category ' V e r b ' includes exactly the same words as previously, but the criterion has been that these words are inflected for tense, not that they express e.g. an action. Nouns and interjections also remain unchanged. NORDLEXIN-N makes use of the following parts of speech: Part of speech

Comment

Example

Verb

Unchanged

Jeg bor i Norge ( Ί live in Norway')

Interjection

Unchanged

Huff!, AEsj/, Au! ('Ouch!' 'Ugh!' Oh!')

Noun

Unchanged

En ny bil ( Ά new car')

Pronoun

Prev.: mostly pers. pron.

Han smilte til henne {'He smiled at her')

Adjective

Unchanged + ordinal nos. En gammel mann ('An old man')

Adverb

Mostly unchanged

Det regner ikke nä ('It is not raining now')

Preposition

Mostly unchanged

De bor i Bergen ('They live in Bergen')

Conjunction

Prev.: coordinators

Mann og kone ('Husband and wife')

Subjunction

Prev.: subordinators

Han sa at jeg tok feil ('He told me that I was wrong')

Determinative

See comment below

Et lite hus (Ά small house')

Determinatives constitute a new class comprising all words inflected for gender and number, and which can precede a noun like this: mann ('man/husband'), but (as a rule) not

The Dictionary Project NORDLEXIN-N

41

like this: en mann ('a man/husband') (words which can fill the empty slot here are classified as adjectives). The class of determinatives also excludes words which can stand alone, like jeg ( ' m e ' ) og han ( ' h e ' ) (these are classified as pronouns). Determinatives are subdivided into three types: a) Possessives (previous: possessive pronouns): min mann ('my husband') hennes mann ('her husband') deres mann ('their m a n ' ) b) Quantifiers (previous: indefinite pronouns): en mann ('one m a n ' ) tre menn ('three m e n ' ) mange menn ('many/several men') c) Demonstratives (previous: demonstrative pronouns): den mannen {'that m a n ' ) disse mennene ('these men') samme mann ('the same m a n ' ) Demonstratives can also stand alone, the presence of a non-expressed noun being implied.

7. C l o s i n g R e m a r k

In 1999 the National Centre for Educational Resources (NLS) undertook a comprehensive survey in order to evaluate the needs of the various minority language groups for a Norwegian LEXIN dictionary. On the basis of a priority list of ten languages, NLS decided on Bosnian as the first target language in the NORDLEXIN-N project. In April that year, however, we were informed that the Norwegian-Serbo-Croatian dictionary published by Universitetsforlaget in 1995 will be supplemented with Bosnian and published in a new edition by another publisher in a year or so. Considering this fact, NLS decided to abandon Bosnian as the first target language in the NORDLEXIN-N project. It has now been decided to use the English translation of the Swedish material in a pilot project for automatic linking of translations to the Norwegian material. After that, priority will be given to Tamil. The dictionaries will be published both in printed and electronic versions. The work will be carried out at the HIT Centre.

42

Tove Bjorneset

References

Faarlund, Jan Terje, Svein Lie and Kjell Ivar Vannebo (1997): Norsk referansegrammatikk. Oslo: Universitetsforlaget. Hofstad, Knut, Stäle Loland and Per Scott (1997): Norsk dataordbok. 6 th Edition. Oslo: Universitetsforlaget.

Jane Bottomley From Print to Disc: Creating the Electronic Version of the Cambridge International Dictionary of English (CIDE)

In the past, electronic versions of dictionaries were little more than searchable versions of the printed text. As online dictionaries grow in popularity, and dictionary CD-ROMs become more widely available and affordable, quality and functionality are improving all the time. In this paper, I will discuss ways of exploiting the full potential of the electronic medium, with reference to the Cambridge International Dictionary of English Online (CIDE Online), a free Internet service, and the Cambridge International Dictionary of English CDROM (CIDE CD), whose development I have been involved in. The four main areas covered are: 1. 2. 3. 4.

editing the content specifically for an electronic format managing the database and integrating feedback the user interface extra features outside the running text

1. Editing

1.1. Space Virtually unlimited space means a more expansive display is possible, as well as a more liberal inclusions policy. This means there is no need to compress text into dense paragraphs: each definition and example can start on a new line. Nor is there a need for obscure codes: η can be expanded to noun and ν to verb etc., and these terms can themselves be further explained via hyperlinks. This is the kind of information which usually lies dormant in the front and end matter of a dictionary, rarely disturbed by dictionary users. Space also means that editors have more scope for updating the dictionary with new words: almost five hundred words were added to CIDE CD.

1.2. Design Screens are not as easy to read as print, so it is vital to take full advantage of the abundance of space in order to make the entries easier to navigate. However, we need to examine closely the merits of lumping and splitting information: too much information written out in full with lots of new lines can be just as confusing as compressed paragraphs. Similarly,

44

Jane Bottomley

while colour is useful for distinguishing the different elements within a text, too many colours can be distracting. Learners' dictionaries pride themselves on their simplicity and clarity, with lots of examples, and definitions only using a core defining vocabulary, and the same needs to be true of the user interface. Too many links and options can end up being confusing. The purpose of any link or option should be self-evident. Simply because something is possible in multimedia, it doesn't mean we should necessarily do it.

1.3. Defining Style As far as how we define something is concerned, multimedia presents us with many more options, and the most appropriate defining style can be employed, such as illustrations, animation, video, photographs or sound. CIDE CD has 70,000 British and American pronunciations, and there is the potential to demonstrate actual sounds, such as crash, bang and crackle, which are hard to define in words.

2. C o n t e n t M a n a g e m e n t

2.1. Database A flexible database can be adapted to multiple formats. Tagging the content of the database allows different elements of the text to be identified and displayed in the most appropriate way for each particular medium. For example, all compounds and idioms are identified on the database, so we were able to colour-code them for the CD, making them easier to locate in the search panel. Tagging of content allows for more advanced electronic searches. With the CIDE CD filter system, a user is able to search on part of speech or grammar code, for instance. With the printed medium there can be a lapse of several years between writing an entry and seeing it in print. But with electronic products, data can be changed and processed at the last minute. This means we can continuously update the online dictionary, and we were able to add new words to CIDE CD, such as GM, e-commerce and ecotourism, and new senses of existing words, such as the "junk e-mail" sense of spam. It also allows us to deal with revisions. This is especially relevant to the online version, but of course, CDs can also be updated from the Internet.

2.2. Online Feedback With over four million searches to date on CIDE Online, there is an abundance of data to inform the development of the Cambridge dictionaries. Some of this is direct feedback, as online users are encouraged to comment on entries and suggest new words. If they spot

From Print to Disc: Creating the Electronic

Version of the CIDE

45

inconsistencies or omissions, we can correct these, not just online, but on the database as a whole, so benefiting all future versions, electronic or paper. In addition to this, all searches are logged. This means we can find out which words are most commonly searched for: there is a (slightly censored) "Top Twenty" posted on the Cambridge dictionary website, which is updated every month. Furthermore, we can find out which words are not found and, if they warrant inclusion, add them to the database. Common errors can also be logged, so we can help users find difficult-to-spell words like millennium and accommodation. We can use this feedback to promote interaction between the online, CD, and paper versions: the online version can be a prototype for the CD, and changes may feed back into print, when it is practical.

3. T h e U s e r I n t e r f a c e

3.1. Text Manipulation With the electronic medium, users can have much more control over the amount of material they are viewing and the way it is organised. On CIDE CD, the user is able to customise the display, choosing which information appears on the screen. For example, learners who are not familiar with IPA may find phonetic transcriptions simply clutter up the screen and decide to suppress them. Customisation can also be used to adapt a product for a different set of users, so grammar codes could be suppressed for native speaker schoolchildren, or more sound options could be added for visually impaired users. User annotations, whereby the user attaches a note to an entry, are another way of personalising the dictionary. As mentioned above, detailed tagging of the database allows for tailored searches through the filters, such as verbs followed by -ing, slang words, or idioms featuring a particular word such as eye or cat. In addition, "related word" coding, whereby every entry on the database is categorised according to meaning or subject area, means the functionality of the CD dictionary can be greatly extended, transforming it into a thesaurus and vocabulary builder. Clicking on the "related words" field in the entry for small will generate synonyms spanning a range of registers, including dinky, minuscule, itsy-bitsy and petite·, the entry for environment will lead the user to a list of other words and expressions related to the topic, including up-to-date entries such as Euroland. It is true that searches using filters may also highlight omissions and tagging inconsistencies, but as explained, this information will feed back into the database and benefit all future editions, electronic and paper.

3.2. Searching the Dictionary The electronic medium is a versatile one. With CIDE CD, the user can look up a word from any software, for example while word processing or using e-mail or the Internet, and copy or print an entry or part of an entry.

46

Jane Bottomley

Colour-coding of headwords, compounds and idioms makes it easier to locate run-ons, which are easy to miss in a packed entry. Electronic searching makes the cross-references or indexes of paper dictionaries redundant. Also, in the Cambridge dictionary it turns out that a feature that was used in the print edition actually works better in the electronic format: "guidewords" work well as they can be seen as a list, whereas numbered definitions are more difficult to navigate on-screen. Electronic searches mean "wildcards" can be used with difficult-to-spell words, and also to find words with a particular affix ("proof generating childproof, foolproof etc.). Also, with CIDE CD, as you type in a word, all the words beginning with that letter or combination of letters appear in the search panel, so the users can find the word they need with only an approximate spelling.

3.3. Hyperlinks Hyperlinks are immediate and the user can click on any word anywhere in the text. Consequently, people are more likely to use these than cumbersome cross-references and indexes. Hyperlinks also make illustrations, pronunciation, study sections, and exercises more immediate and interactive.

4. E x t r a F e a t u r e s on the C D

4.1. Pronunciation Being able to hear the words is a function that can complement or replace phonetic transcriptions. There is also the potential to cover more varieties of English. With CIDE CD, users can listen to British and American pronunciation. They can also record themselves and compare their own pronunciation with a recording of a native speaker.

4.2. Illustrations Colour pictures from the print version can be used electronically with no extra expense, and colour can be added to black and white pictures. The use of "hotspots" means that labels do not clutter, and hyperlinks provide immediate two-way access, from entry to picture, and vice versa.

From Print to Disc: Creating the Electronic

Version of the CIDE

47

4.3. Exercises Exercises can be automated, generating a random selection of questions from a larger pool each time the user chooses an exercise. They can also be directly linked to relevant study sections.

4.4. Study Sections Study sections are more interactive. They are also less intimidating because the information can be broken up into manageable chunks. Furthermore, the absence of space restrictions means that comprehensive lists of examples can be included.

5. Conclusion

In conclusion, it would seem that the process of transforming a paper dictionary into an online or CD-ROM dictionary is facilitated by a flexible database and an awareness of the potential, and potential pitfalls, of the electronic format. These powerful reference tools are by no means a replacement for paper dictionaries, but an alternative means of accessing and using reference material. Moreover, the development of a CD-ROM or online dictionary, and the attendant feedback from users, can directly affect the content and format of the print version.

References

CIDE = CAMBRIDGE DICTIONARY OF ENGLISH, Cambridge University Press, 1995. C I D E O n l i n e = CAMBRIDGE DICTIONARY OF ENGLISH ONLINE, C a m b r i d g e U n i v e r s i t y Press,

http://dictionary.cambridge.org/ CIDE CD = CAMBRIDGE DICTIONARY OF ENGLISH CD-ROM, Cambridge University Press, 2000.

1999.

Ulrich Busse Lexicography as a Sign of the Times: A Study in Socio-Lexicography'

1. Introduction

Apart from questions of objective, scope, presentation of data and, first and foremost, the potential users, which pertain to any serious lexicographical project, a distinction needs of course to be made between commercial lexicography and academic projects, since the socio-economic conditions are quite different. To take an example, the four competing British learner's dictionaries all appeared with major new editions in the very same year of 1995. On the one hand, this gives evidence to the fact that there is obviously a demand for this type of dictionary but, on the other hand, this also proves that there is serious economic competition, as no publishing house wants to fall behind its competitors; the dictionary has become a commodity. By contrast, the sudden proliferation of dictionaries of Anglicisms in Germany and other European countries during the 1980s and 90s cannot in all likelihood be explained by such socio-economic constraints; it rather documents that the compilers felt that such documentation was still wanting at this point in time. In order to show that the same topic can be dealt with in quite different ways I shall focus on three dictionaries of Anglicisms in present-day German, namely: 1)The Anglizismen-Wörterbuch [AWb] (1993-96) compiled by Broder Carstensen and myself. 2) A Usage Dictionary of Anglicisms in Selected European Languages [UDASEL] by Manfred Görlach (forthcoming), in which 1 participated as second reader for German, and 3) the Wörterbuch überflüssiger Anglizismen [Dictionary of superfluous Anglicisms] (1999) edited by Reiner Pogarell and Markus Schröder. Typologically these dictionaries constitute a special type of foreign word dictionary in that they cover foreign wordstock in present-day German that stems exclusively from English. For example, the AWb perceives its special role as follows: the dictionary is not a monolingual dictionary because it describes loan processes in detail. Nor is it a bilingual dictionary that simply establishes semantic equivalences between lexemes in the recipient language and their donor language (cf. AWb 1993: 25*). The ubiquitous presence of the English language in present-day German and other languages of the world can be felt by language users in many different walks of life. For instance, Gnutzmann (1998: 130) finds that "it is most conspicuous in the business world, the My colleague Oliver Grannis (Osnabrück) has suggested the term socio-lexicography that investigate the repercussions of social processes in dictionaries.

for studies

Ulrich Busse

50

pop industry, communications technology, in scientific communication and advertising." As a reaction to this, in the recent past many European languages have devoted specialised dictionaries to this topic, e.g.: Croatian, Danish, French, Italian, Latvian, Norwegian, Polish, Spanish and Swedish (cf. references and Gottlieb, this volume). Apart from these specialised dictionaries, Anglicisms are also covered by a number of national dictionaries of neologisms or vogue-words respectively. As far as German is concerned, the growing lexical influence of English from the Second World War onwards has been extensively described and analysed in articles and monographs as the bibliographies in the AWb (1993: 131*-193*) and the annotated bibliography compiled by Busse/Görlach (forthcoming) amply document. However, only with the completion of the AWb has the amount of lexical influence been covered in detail. The dictionary compiled by Neske (1972) may be credited as the first German dictionary of Anglicisms, but it has a number of drawbacks as it does not provide documentation of its sources and principles for the selection of its lemmata. Furthermore, the lexis of specialised fields, in particular that of economics, is overrepresented. To my knowledge, there exists only one other German dictionary devoted entirely to Anglicisms, viz.: Ilse Sörensen Englisch im deutschen Wortschatz, Lehn- und Fremdwörter in der Umgangssprache (1995). The work, which is unfortunately out of print, consists of a short introduction to the forms and functions of English loans in present-day German and a dictionary listing some 2,500 Anglicisms ranging from old and fully accepted loans to modern ones current only in specialised domains and jargons. However, the dictionary does not offer documentation in form of citations, dates and socio-pragmatic information.

2. German Dictionaries of Anglicisms

2.1. The

Anglizismen-Wörterbuch

For a dictionary of Anglicisms in present-day German two basic options are principally possible: either to set up a dictionary designed for the user interested in information on certain specific problems or queries, such as spelling, pronunciation or uncertainties of meaning and usage or, on the other hand, a dictionary written for linguists and other scholars working in related fields of interest and seeking comprehensive information on the complex problems concerning the integration of an item, its socio-historical context, etc. Although not explicitly stated in the introduction to the AWb, the dictionary entries themselves with their detailed structure testify to the fact that this question has clearly been resolved in favour of the latter approach, though efforts have been made to keep the entries clearly organized and largely free of linguistic jargon so that the interested layman, with rather more practical questions, can also make use of the dictionary.

Lexicography as a Sign of the Times: A Study in Socio-Lexicography

51

2.1.2. The scope of the dictionary The Dictionary of Anglicisms is based on a corpus of 100,000 citations from German newspapers, periodicals, catalogues, advertisements, booklets, pamphlets and from high as well as trivial literature, including a few samples of spoken German, especially from television and radio. The dictionary deals with the 3,500 most common and current Anglicisms in present-day German. The term Anglicism is used here as a generic term for Briticism, Americanism, Canadianism, etc. The main focus of the dictionary lies on those Anglicisms which have come into German after 1945. Older loans are included only if there have been changes in meaning or new meanings added. A central criterion for the currency of an item - and its subsequent inclusion in the dictionary - was its documentation within the corpus. The cut-off margin for potential entries was set at five citations from different sources over an extended stretch of time. However, these numbers were not strictly adhered to at all costs, especially, when the competence of the compilers served as counter-evidence for under-representation of the respective item in the corpus. In cases like this supplementary documentation was looked for. Together, these factors should ideally guarantee a solid foundation and prove a certain reliability of an item as a prospective dictionary entry or, to put it the other way around, help exclude nonceformations and short-lived vogue-words. On the other hand, this does not mean of course that Anglicisms designating concepts which due to changes in the socio-historical setting have become obsolete e.g. Airlift, Luftbrücke, Skylab and others, are excluded because the things they refer to no longer exist. Words like this are included but given a special usage label and often also a detailed commentary in a special position within the framework of the dictionary entry. Furthermore, in this special section of the entry structure many headwords are also provided with detailed additional information on the origin and history of a term and its integration into the German language. In many instances the dictionary also features encyclopaedic information and references to other authors, who have dealt with the item.

2.1.3. The treatment of technical terms Technical terms have been included in so far as they were documented in the corpus. With the corpus consisting mainly of press publications, only those elements from technical languages are covered that normally appear in newspapers and periodicals. In this way, highly technical terms are eliminated, but those more central to modern ways of life, viz. terms from computing Cursor, Diskette, Laptop, Maus, online, Software, business administration and economics such as Break-even-point, Buyout, Cash-flow, Franchise, Leasing, Windfall-Profits etc. were included in the dictionary because the average reader is likely to encounter them in newspapers or magazines.

52

Ulrich

Busse

2.1.4. The terminology of linguistic borrowing The terminology delineating the various loan processes at work has been developed exclusively for this dictionary, thus discarding the terminology put forward by Betz and others. This means that the items have been categorized into two large groups: those words deriving directly from (aus) English, i. e. Job, Show, Baby, recyclen, surfen; clever, cool; High Tech, Product Placement; Big Brother is watching you, last (but) not least', and those modelled on {nach) English, i.e. Erste Dame, Urknall, Wasserglätte; grünes Licht geben/erhalten, etc. A third, but minor category consists of those cases where an English model has sparked the borrowing process and is still recognizable as such, but where German has morphologically altered, mostly truncated, the word to an un-English form, such as the shortenings Deo (for deodorant), Pulli (short for pullover), Profi (for professional - the English short form being pro) and probably the most well-known member of this class Twen ('people in their twenties'). These rare cases are labelled as related to (zu) English. The AWb hence subscribes to a formal typology based on the dichotomy of native vs foreign wordstock and the various degrees of its integration, whether imported (without formal integration), or partially or completely substituted by native material. Such formal classification of Anglicisms leaves out questions of usage as it does not account for the place of an Anglicism within the German wordstock nor its use by the language users. This is the point of departure for the UDASEL. While Görlach admits that large-scale sociolinguistic studies on the use of Anglicisms are still wanting and that questions of frequency, acceptance as well as social and dialectal variation have always posed a serious stumbling block for both prescriptive and descriptive lexicographers, he turns usage into the central category of his dictionary.

2.2. The UDASEL - Anglicisms and their usage Despite the fact that several national dictionaries exist, the UDASEL tries to cover the lexical impact of English on 16 European languages, because "the aims and methods of the compilers of these dictionaries diverge a great deal, and the projects date from different decades; as a consequence, no proper comparison of the data of the individual languages is possible on the basis of these published books" (Görlach 1998: 209 f.). The dictionary is comparative in scope and gives evidence on selected European languages from different language families, that is 1) Germanic (Icelandic, Norwegian, Dutch and German), 2) Romance (French, Spanish, Italian and Rumanian), 3) Slavonic (Russian, Polish, Croatian and Bulgarian) and 4) four other languages (Finnish, Hungarian, Albanian and Greek). In contrast to the AWb the data presented in the UDASEL are not corpus-based, because for some of the languages under observation no corpora exist. Instead, data was taken from existing recent dictionaries and, additionally, educated native speakers were consulted. The dictionary covers the time span from 1945 into the mid 1990s with a cut-off date in 1995, laying special emphasis on the more recent loans. All in all, it contains some 3,000 to

Lexicography as a Sign of the Times: A Study in Socio-Lexicography

53

4,000 entries. Unlike the AWb it omits all types of caiques or other substitutions as entries. Loan translations and the like are only mentioned if there is an entry for the English word. In such cases as with German Airbag a 'larger-than sign' (>) signals that Airbag is more frequent than the creation Prallsack. Whereas the AWb categorises its entries according to formal criteria, the UDASEL classifies them according to the categories of acceptability and currency paired with elements of morphological and phonological integration. The Anglicisms are arranged on a cline of increasing integration ranging from 0 to 5. Görlach (1998: 215) writes that "these currency values are probably the most important piece of information in the entry." The words not forming part of the language in question are subcategorised in three groups: — : The word is not known (no entry for the individual language, but a caique or other native equivalent may be provided): thus, for example, Cleverness, Gully and Fitness are absent from all languages except German. 0 : The word is known mainly to bilinguals, but is felt to be English as Weekend in German. 0 : The word is known but used only with reference to British or American situations, i.e. it is a foreignism, e.g.: Acre, Barrister, College, County, Earl, Lord, etc. If a word is accepted by native speakers it can be regarded as part of the linguistic system. However, acceptance can vary in degree depending upon a number of sociolinguistic variables. In many cases restricted use in particular registers coincides with incomplete morphological and phonological integration. 1 : The word is restricted in use: the nature of the restriction (age, style, technical, regional distribution, etc.) is indicated by diachronic, diatechnical, diatopic, etc. markers, e.g.: Ballyhoo, Event, Kids, Scoop are all examples of journalese. 2 : The word is fully accepted and found in many styles, but still marked as English in its spelling, pronunciation or morphology, e.g.: Jeans or Thriller. 3 : The word is no longer recognized as English; the fact can only be established etymologically, e.g.: Frack, Humor, Keks, Pudding, Rum, Sport. 4 : The word is identical with an indigenous item in the receiving language, so that the contact only resulted in a semantic loan, e.g.: Maus for computers. 5 : The word is not of English origin. This category covers items of neo-classical provenance (in particular Latin or Greek) whose English origin is impossible to determine from a morphological or phonological point of view. For these reasons such words as German Fotografie and Grammofon are excluded. But if the same etymon proves to be an anglicism in a particular language, it is included. Thus, for example, Aktion in German is an older loan from Latin not featured in the dictionary, but Action as a more recent one from English is included (for the above categories see Görlach 1994b, 1997, 1998).

54

Ulrich Busse

2.3. The Wörterbuch überflüssiger

Anglizismen

In 1997 a new society for the preservation of the German language was founded. Under the auspices of this circle this Wörterbuch überflüssiger Anglizismen was compiled in the autumn of 1999, and within just a few months saw a second edition which its editors in their preface interpret as having published at the right point in time, because in their opinion many people are fed up with the present anglicisation of the German language. On its Internet pages the society divides the Anglicisms in present-day German into the following categories: 1) annoying, because the German word is better or older 2) borderline cases 3) accepted Anglicisms 4) pseudo Anglicisms It is admitted that there is no strict dividing line between these categories and that there is a broad spectrum ranging from Event and Highlight, which are deemed extremely irritating to fully integrated items such as Pullover and Keks. In contrast to this classification the dictionary presents a collection of about 3,500 Anglicisms, which in the opinion of the editors are altogether superfluous. In the scanty preface the authors state that the objective of the dictionary is twofold in that it wants to document Anglicisms and act as a reference book for those who care for good German and want to produce texts with German words. The words are said to have been chosen from all domains with which ordinary people are likely to come into contact. Specialised vocabulary and trendy neologisms have been incorporated more sparingly. However, a glance at the lemmata illustrates that the product cannot live up to these standards because in terms of frequency, synchronic and diachronic degree of foreignness, their function and place within the vocabulary of German, etc. the Anglicisms are extremely heterogeneous, a fact clearly documented by such entries as: anti-dim glass, biassed [sic], bid-ask-spread, bombast, ear, fannings, fetish, grillen, happy-end, hobby, leg pulling, railcard, railway card, rallye, ranch, sweat-shirt, weather, wrap industry. Specimen entries look like this: bias: Verzerrung biassed: voreingenommen bid-ask-spread: Geld-Brief-Spanne bidirectional: ambivalent, in zwei Richtungen gehend big: groß, mächtig, auch in big-band: Tanzorchester big bang: Urknall big boss: Magnat, Bonze, Chef, Vorgesetzter

These samples clearly illustrate that the label Wörterbuch is rather a misnomer for this work. This so-called dictionary should rather be labelled a glossary, because it lacks all features of modern lexicographical microstructure such as German spelling - the lemmata are spelled according to the American Heritage Dictionary - pronunciation, grammatical

Lexicography

as a Sign of the Times: A Study in

Socio-Lexicography

55

information, pragmatic markers and, above all, documentation. Any documentation that would prove the superfluous character of the Anglicisms is missing. In my opinion these flaws invalidate the whole undertaking because a distinction of loans into those which are deemed necessary to fill lexical gaps in the recipient language and those which are not has never been a valid one. For example Feier, Fete and Party are not synonymous, because they cannot be used under the same conditions; they belong to different registers and convey different connotations. For just the same reasons Baby and Säugling or Kleinkind cannot be freely exchanged. To exchange Cowboy with Rinder- or Viehhüter would be to miss out on the local colour of Anglicisms; to supplant Festival by Festspiele means to ignore the socially completely different settings of the Salzburger- or Wagner Festspiele in Bayreuth or the Filmfestspiele in Cannes and those of the music festivals of Woodstock, Altamont or the Isle of Wight. Galinski (1975) and Pfitzner (1978) have provided valuable stylistic criteria of Anglicisms in German. More recently, Leisi & Mair (1999: 220) have drawn attention to the fact that Anglicisms and other loans can be necessary not only for structural reasons but also for sociolinguistic ones, so that in a given social group they act as a means of identification and demarcation to the outside, much in the same way as slang terms have always done. So if this word list has so many shortcomings why consider it in the first place? In my opinion it is nonetheless worth consideration for a number of reasons. On the one hand, it can hold a place in the long-standing purist tradition in Germany. The ubiquitous role of English and its function as lingua franca, has again given rise to fears that the huge influx of Anglicisms could in the long run endanger the continuity of a fully-fledged German language. On the other hand, the question of whether the incorporation of Anglicisms into the German language is to be viewed as potentially dangerous, because it creates communicative barriers or adds to stylistic enrichment, reminds me of the discussion of the so-called inkhorn terms of 16th century England and the hard word controversy of the 17th century. Manfred Scheler (1977: 104) has put the problem this way: Bei allen negativen Begleiterscheinungen, die die Überfremdung des englischen Wortschatzes durch die massenhafte Entlehnung rom.-lat. Wörter mit sich bringen mußte, bedeutete die Aufnahme fremden Wortguts in erster Linie eine Bereicherung und Verfeinerung der sprachlichen Ausdrucksmittel. Der Latinismus trug wesentlich zur hohen Kunst der englischen Literatursprache bei.

3. C a n p r e s e n t dictionaries o f A n g l i c i s m s b e c o m p a r e d to 17th c e n t u r y E n g l i s h hard-word dictionaries?

Admittedly, many of the most recent Anglicisms currently used especially in the German language of advertising and in computerese are of a different nature than the latinisms in 16th andl7th century Britain. Toughes und trendiges Hip-hopping or Plattpreising and Extremsparing do not really add to copia verborum, but then again, many of the hard words were also rather short-lived and peripheral. Nevertheless, it is interesting to observe certain similarities between the most recent surge in the production of dictionaries of Anglicisms and the hard-word dictionaries. In the 16th century the English language was on its way

56

Ulrich Busse

from a vernacular to a standard language. In many areas of learning Latin still reigned supreme. Some four hundred years later, English with its global spread has long supplanted Latin as a modern lingua franca, and in many respects the other national languages of Europe have taken in large numbers of borrowings and have also retreated as official languages in politics, diplomacy and scholarship. Hence, to a certain extent, there seems to be a socio-historical connection between the English hard-word dictionaries and present-day dictionaries of Anglicisms in that both could be interpreted as typical signs of their times. However, with the exception of the Dictionary of superfluous Anglicisms, which in scope and treatment equals the scanty Early English word lists or glossaries, the AWb and the UDASEL prove that lexicography has made huge progress over the past centuries. In early modern England the huge influx of Latinate neologisms was not universally welcomed either. On the one hand, the neologisms from Greek and Latin were regarded as a means to enrich and refine the English language, to make up for its alleged shortcomings and through augmentation provide stylistic variation. On the other hand, purists opposed the Latinate aureate diction of the 15th century and the inkhorn terms of the 16th century (cf. Barber 1997: 42-79, Gray 1988). The difficulties connected with the proper understanding and usage of the Latinate terms led to a series of dictionaries which tried to explain these hard words (cf. Stein 1985, Starnes/Noyes 1946). In sociolinguistic terms the learned Latinisms can be seen to constitute a social class distinction and a language or lexical bar to those who cannot handle them. Misuse (named malapropism after the character of Mrs Malaprop in Sheridan's play The Rivals, 1775) and folk etymology may result from this. Social consequences of the misuse of foreignisms are outlined by Grove (1950), Leisi (1985: 80-83), and Corson (1985, 1995). Beside the huge influx of Latin words, borrowing from French also continued. On the basis of the Chronological English Dictionary [CED] (1970), Wermser (1976: 45) has shown that from 1510 to 1724 the percentage of Latin loans declined from 48% to 38% and those from French from 41% to 26%. In comparison, the loans from other living European languages such as Dutch, Spanish, Italian, Portuguese, and from non-European languages manifested themselves to a greater extent only from the 16th century onwards, but have remained far less numerous (cf. Scheler 1977: 63-70). A graph given in Crystal (1995: 72) based on the data of the OED shows that vocabulary growth in Renaissance England clearly peaked at around 1600 with about 6,000 new words per year. Because of its bias in favour of "great English writers", Schäfer (1989) has estimated that if all corrections were taken into account, the total number of discrepancies in the OED database could amount to half a million. The advanced dissociation of the English wordstock makes English different from, for example, German. Leisi (1985: 75) comes to the conclusion that among the modern European languages English and French have a far more dissociated lexis than German and Italian. This characteristic of Latinisms can make them difficult or hard words for speakers lacking classical education. However, this does not seem to have led to severe misunderstandings or a breakdown in communication, because many Latinisms in English do not belong to the high-frequency register of every-day discourse. Gabriele Stein (1988) has shown in detail that "the beginnings of dictionary-making for most of the European languages [...] lie in language contact situations. This means that for these languages monolingual lexicography was preceded by bilingual and multilingual

Lexicography

as a Sign of the Times: A Study in

57

Socio-Lexicography

lexicography" (29). Hence, in the 15th and 16th centuries in Britain there were a great number of these dictionaries around, which mostly included Latin. The first purely English dictionary issued was Cawdrey's A table alphabeticall. However, this dictionary does not really constitute a truly English monolingual dictionary but introduces a special type known as 'hard-word dictionary' at a time when the influx of foreign words reached an unprecedented peak. "Many of these, especially from Latin and Greek, were incomprehensible to the unlearned reader. Without exception, the seventeenth-century English dictionaries are dictionaries of hard words, providing a selection of English words likely to be found obscure" (Barber 1997: 76). Barber describes the history of the English dictionary in the 17th century as one of increasing size and complexity. In their account of the Early Modern English Dictionary Corpus, Lancashire & Patterson (1997) provide data on dictionary growth. While the bilingual dictionaries range from 18,500 (Palsgrave 1530) entries to 47,000 (Cotgrave 1611), the hard-word dictionaries are much smaller in scope. The following table taken over from Lancashire/Patterson shows the relevant figures: Table 1. English hard-word dictionaries based on Lancashire/Patterson (1997: 48; shortened) Author's Name Edmund Coote Robert Cawdrey John Bullokar Henry Cockeram

Date 1596 1604 1616 1623

Type Ε hard-w Ε hard-w Ε hard-w Ε hard-w

Entries 1,350 2,500 4,200 9,900

OED Citations 2 200 5 1,600

Since the history of these dictionaries is well known and covered accordingly (cf. Starnes & Noyes 1946), it shall not be repeated here. Nevertheless, three of these dictionaries will be presented briefly because their varying objectives somehow echo those described earlier on for the modern German dictionaries of Anglicisms. The lengthy subtitles of these dictionaries indicate their objective. A table Alphabeticall, conteyning and teaching the true writing, and vnderstanding of hard vsuall English wordes, borrowed from the Hebrew, Greeke, Latine, or French. & c. With the interpretation thereof by plaine English words, gathered for the benefit & helpe of Ladies, Gentlewomen, or any other vnskilfull persons. Whereby they may the more easilie and better vnderstand many hard English wordes, w h i c h they shall heare or read in Scriptures, Sermons, or elsewhere, and also be made able to vse the same aptly themselues (Cawdrey 1604: title-page).

The pedagogical purpose of John Bullokar's An English Expositor (1616) is clearly expressed in the second part of the work's title: "the interpretation of the hardest words vsed in our Language". Apart from obsolete or obsolescent vocabulary from previous centuries, especially archaisms from Chaucer's times, which have been included for decoding purposes but which, in the author's opinion, should not be used in everyday speech, and specialised vocabulary from "diuers termes of art, proper to the learned in Logicke, Philosophy, Law, Physicke, Astronomie, etc. yea, and Diuinitie itselfe, best knowen to the seuerall professors thereof', the dictionary includes, first of all foreign loans "the great store of

Ulrich Busse

58

strange words, our speech doth borrow, not only f r o m the Latine, and Greeke, (and s o m e f r o m ancient H e b r e w ) but also from forraine vulgar Languages round about vs" which had caused the outbreak of the " i n k h o m controversy" (cf. Gotti 1998: 63). T h e Dictionary of Cockeram (1623) The English Dictionarie: or, an Interpreter of hard English Words has added " m a n y words that have never appeared anywhere else - a d p u g n e , adstupiate, bulbitate, catillate, fraxate, nixious, prodigity, vitulate, and so o n " (Read 1989: 386). It consists of three parts, the first of which contains "the choicest w o r d s " with w h i c h the language has been enriched (= inkhom terms), but it also includes a second bilingual part which provides copious " m o r e refined and elegant" Latin equivalents for plain native " v u l g a r " words, and as a third part it also includes a dictionary o f ' m y t h o l o g y . The first Booke hath the choisest words themselues now in vse, wherewith our Language is inriched and become so copious, to which words the common sense is annexed. The second Booke containes the vulgar words, which whensoeuer any desirous of a more curious explanation by a more refined and elegant speech shall looke into, he shall there receiue the exact and ample word to expresse the same (Cockeram 1623). Specimen entries taken f r o m the very beginning of the Second Part of The English tor look like this:

Transla-

To Abate or take away. Deduct, Deduce. Abominable, Sacrilegious, Detestable. To abound, Exuperate. To abound overmuch, Superabourtd Too great abundance, Superfluity, profluence, uberty, confluence Unlawfull Absence from ones place or charge, Nonresidency. Thus, in order to meet public d e m a n d this dictionary was not only tailored to the necessities of text reception but also to text production, providing its users with fashionable words. H o w e v e r , it is quite conceivable that the air of erudition often resulted in inappropriate use or involuntary misuse and blunders.

4. S u m m a r y and conclusion

Today, the English language with its various m o d e s of existence as mother tongue, second language and foreign language m a y indeed best be described as a global language or lingua franca in its function as a means of communication between non-native speakers. Attitudes towards this development vary strongly. For s o m e people, the English language is uniting the world whereas for others the world-wide presence of English is regarded as linguistic imperialism: " D e p e n d i n g on their experience with and attitude to A n g l o - A m e r i c a n language and culture, for s o m e people, English is a linguistic m e a n s of oppression, for others it is an ideal and universal m o d e of c o m m u n i c a t i o n " (Gnutzmann 1998: 130). A b o u t four hundred years ago, the English language was only on the brink of emancipating itself f r o m a written vernacular to a fully-fledged national standard for all d o m a i n s o f

Lexicography as a Sign of the Times: A Study in Socio-Lexicography

59

written communication, and in many areas still struggling against the dominating influence of Latin. The above presentation of modern dictionaries of Anglicisms and English hard-word dictionaries of the 17th century revealed certain parallels. Both have their origin in intense linguistic and cultural contact of one dominating language with another, with the difference that English has come full circle from having absorbed huge numbers of mainly Latinate loanwords in the,\7th century to being the world's most prolific open reservoir of words being borrowed quite extensively by many (European) languages. From the present point of view, however, it has to be added that this special type of foreign-word dictionary has a completely different position in the lexicographical landscapes of the two countries under observation. Whereas in Britain the hard-word dictionaries were soon absorbed by major monolingual dictionaries and dictionaries of foreign words have always led a rather marginal existence (cf. Kirkness 1989), dictionaries of foreign words have been going strong in Germany ever since their introduction. Nevertheless, in both cases the languages-in-contact-situation has given rise to special types of dictionaries, the hard-word dictionary and the dictionary of Anglicisms, two subtypes of foreign word dictionaries. Whether in Germany foreign-word dictionaries in general, and dictionaries of Anglicisms in particular, will one day merge into a general, scholarly, well-documented monolingual dictionary as in the English tradition from Dr Johnson (1755) to the OED remains to be seen. A comparison of the macro- and microstructures of the German dictionaries of Anglicisms shows that scope and objectives vary considerably. On the one hand the A W b and the UDASEL are descriptive, informative works that take a neutral scholarly position. On the other hand, the Dictionary of superfluous Anglicisms takes a critical, emotionally engaged stance towards foreignisms and thus moves in the direction of a normative puristic dictionary. By contrast, the English hard word dictionaries were inspired by a pedagogical and enlightening impetus to educate and thus emancipate those who had difficulties in understanding and using these words. Cockeram even goes so far as to propagate the use of modish foreignisms in cases which could be worded simpler by using less fashionable native words. The fact that Cawdrey's dictionary had four editions, and that of Bullokar even saw 14 editions, the last as late as 1731, also proves that socio-economic factors also played a part in those days. Thus, the dictionaries discussed may not only be taken as typical signs of their times, but also as reactions to public demand, and, moreover, as expressions of lexicographical credos, as the different conceptions and treatment of the subject matter indicate the underlying attitude towards the linguistic matters at stake. In this respect they document the repercussions of both sociological and sociolinguistic issues.

60

Ulrich Busse

References Dictionaries o f Anglicisms A W b = ANGLIZISMEN-WÖRTERBUCH. D E R EINFLUSS DES ENGLISCHEN AUF DEN DEUTSCHEN WORT-

SCHATZ NACH 1945. Founded by Broder Carstensen, continued by Ulrich Busse with the assistance of Regina Schmude. Berlin/New York: de Gruyter 1993-1996, 3 vols. DTV-WÖRTERBUCH ENGLISCHER UND AMERIKANISCHER AUSDRÜCKE IN DER DEUTSCHEN SPRACHE. H g g .

Fritz Neske, Ingeborg Neske, München: Deutscher Taschenbuch Verlag 2 1972 ['1970]. ENGLISCH IM DEUTSCHEN WORTSCHATZ, LEHN UND FREMDWÖRTER ΓΝ DER UMGANGSSPRACHE. H g . I l s e

Sörensen. Berlin: Volk und Wissen 1995. U D A S E L = A USAGE DICTIONARY OF ANGLICISMS IN SIXTEEN EUROPEAN LANGUAGES. H g . M a n f r e d

Görlach. Oxford: OUP forthc. (Published 2001 as A Dictionary of European Anglicisms. Oxford.) WÖRTERBUCH ÜBERFLÜSSIGER ANGLIZISMEN. Hgg. Reiner Pogarell, Markus Schröder. Paderborn: 1FB Verlag 2 2000 [' 1999], A DICTIONARY OF ANGLICISMS ΓΝ DANISH. Ed. Knud Serensen. Copenhagen: Munksgaard 1997. ANGLICISMI LATVIESU VALODA. Ed. J. Baldunciks. Riga: Z i n a t n e 1989. ANGLICIZMI U HRATSKOM ILI SRPSKOM JEZIKU: PORIJEKLO - RAZVOJ - ZNACENJE. E d . R u d o l f F i l i p o v i d .

Jugoslavenska Akademija Znanosti i Umjetnosti - Skoska Knjiga 1990. ANGLISISMEORDBOKA. ENGELSE LANORD I NORSK. Eds. Anne-Line Graedler, Stig Johansson. Oslo: Universitetsforlaget 1997. DICCIONARIO DE ANGLICISMOS. Ed. R. J. Alfaro. M a d r i d : G r e d o s 2 1 9 7 0 .

DICTIONNAIRE DES ANGLICISMES. Eds. J. Rey-Debove, G. Gagnon. Paris: Robert 3 1984. DICTIONNAIRE DES ANGLICISMES. Ed. Manfred Höfler. Paris: Librairie Larousse 1982. DIZIONARIO DEGLI ANGLICISMI NELL' ITALIANO POSTUNITARI. R d . G . R a n d o . F l o r e n c e : O l s c h k i 1 9 8 7 .

NY SVENGELSK ORDBOK. Ed. Bo Seltin. Lund: Studentlitteratur 1993. TENDENCJE ROZWOJOWE WSPOLCZESNYCH ZAPOZYCZEN ANGIELSKICK W JEZYKU POLSKIM. E d .

E.

Manczak-Wohlfeld. Cracow: Universytet Jagiellonski 1995.

Works Cited Barber, Charles (1976; 2 1997): Early Modern English. Edinburgh: Edinburgh University Press. Bullokar, John (1616): An English expositor: teaching the interpretation of the hardest words in our language. Busse, Ulrich; Manfred Görlach (2001): An annotated Bibliography of Anglicisms in Selected European Languages: German. - In: Manfred Görlach (ed.): An annotated Bibliography of European Anglicisms. Oxford: OUP. (Published 2001 as An annotated Bibliography of European Anglicisms. Oxford.) Cawdrey, Robert (1604): A table alphabetical!, conteyning and teaching the understanding of hard usuall English wordes. Cockeram, Henry (1623): English dictionarie: or, an interpreter of hard English words. [The dictionary texts are available for general use on the World Wide Web at http:// www.chass.utoronto.ca.8080/english/emed/patterweb.html.] Corson, David (1985): The Language Bar. Oxford: Pergamon Press. Corson, David (1995): Using English Words. Dordrecht: Kluwer. Cote, Edmund: English Schoole-maister. Imprint: Widow Orwin (1596). Cotgrave, Randle: A dictionarie of the French and English tongues. Imprint: A. Islip (1611). Crystal, David (1995): The Cambridge Encyclopedia of the English Language. Cambridge: CUP. C E D = A CHRONOLOGICAL

ENGLISH DICTIONARY.

LISTING 80,000

WORDS IN ORDER OF THEIR

EARLIEST

OCCURRENCE. Eds. Thomas Finkenstaedt, Ernst Leisi, Dieter Wolff. Heidelberg; Winter 1970. Galinsky, Hans (1963; 3 1975): Stylistic Aspects of Linguistic Borrowing. A Stylistic View of American Elements in Modern German. - In: Broder Carstensen / Hans Galinsky: Amerikanismen der

Lexicography

as a Sign of the Times: A Study in

Socio-Lexicography

61

deutschen Gegenwartssprache. Entlehnungsvorgänge und ihre stilistischen Aspekte. Heidelberg: Winter, 35-72. Gnutzmann, Claus (1998): English as a Global Language: What does it mean? Neusprachliche Mitteilungen aus Wissenschaft und Praxis 51/3, 130-137. Görlach, Manfred (1994b): A Usage Dictionary of Anglicisms in Selected European Languages. International Journal of Lexicography 7/3, 223-246. - (1997): Usage in the Usage Dictionary of Anglicisms in Selected European Languages. - Studia Anglica Posnaniensia 31, 67-77. - (1998): The Usage Dictionary of Anglicisms in Selected European Languages: a Report on Progress, Problems and Prospects. - Links & Letters 5, 209-221. Görlach, Manfred, Busse, Ulrich (forthc.): English in Europe: German. - In: Manfred Görlach, (ed.): English in Europe. Oxford: OUP. Gotti, Maurizio (1998): John Bullokar's "Termes of Art". - In: Jacek Fisiak, Marcin Krygier (eds.): Advances in English Historical Linguistics. Berlin: de Gruyter, 63-78 (= Trends in Linguistics: Studies and Monographs 112). Gray, D. (1988): A Note on sixteenth-century Purism. - In: E. G. Stanley, F. T. Hoad (eds.): Words. For Robert Burchfield's Sixty-Fifth Birthday. Cambridge: CUP, 103-119. Grove, Victor (1950): The Language Bar. London: Routledge & Kegan Paul. Kirkness, Alan (1989): Das Fremdwörterbuch. - In: Franz Joseph Hausmann et al. (eds.): Wörterbücher: Ein internationales Handbuch zur Lexikographie. Vol. 2. Berlin: de Gruyter, 1168-1178 (= HSK 5.2). Lancashire, Ian; Katharine Patterson (1997): Learning from the Early Modern English Dictionaries Corpus. - In: Raymond Hickey, Merja Kytö, Ian Lancashire and Matti Rissanen (eds.): Tracing the Trail of Time. Proceedings from the Second Diachronic Corpora Workshop New College, University of Toronto, Toronto, May 1995. Amsterdam: Rodopi, 47-61. Leisi, Ernst ( 7 1985; 8 1999 Leisi, Ernst, and Christian Mair) Das heutige Englisch. Wesenszüge und Probleme. Heidelberg·. Winter. Palsgrave, John: Lesclarcissement de la langue francoyse. Imprint: R. Pynson (ca. 1524), completed by J. Hankyns (1530). Pfitzner, Jürgen (1978): Der Anglizismus im Deutschen. Ein Beitrag zur Bestimmung seiner stilistischen Funktion in der heutigen Presse. Stuttgart: Metzler (= Amerikastudien 51). Read, Allen Walker (1989): Encyclopaedias and Dictionaries. - In: Encyclopaedia Britannica. Macropaedia. Vol. 18. Chicago: The University of Chicago, 365-394. Schäfer, Jürgen (1989): Early Modern English Lexicography. Oxford: Clarendon Press, 2 vols. Scheler, Manfred (1977): Der englische Wortschatz. Berlin: Schmidt. Starnes, De Witt T., Gertrude Ε. Noyes (1946): The English Dictionary from Cawdrey to Johnson 1604-1755. Chapel Hill: The University of North Carolina Press. Stein, Gabriele (1985): The English Dictionary before Cawdrey. Tübingen: Niemeyer. (= Lexicographica. Series Maior 9). - (1988): The emerging Role of English in the Dictionaries of Renaissance Europe. - Folia Linguistica Historicam, 29-138. OED = THE OXFORD ENGLISH DICTIONARY ON COMPACT Disc. Hgg. Edmund Weiner, E. S. C. Weiner. Oxford: Clarendon Press 1992. Wermser, Richard (1976): Statistische Studien zur Entwicklung des englischen Wortschatzes. Bern: Francke (= Schweizer Anglistische Arbeiten 91).

Timothy Colleman The Contragram Verb Valency Dictionary of Dutch, French and English

1. Introduction

In this paper I will report on the present project of the Contrastive Grammar Research Group of the University of Ghent (known as Contragram), viz. the compilation of a trilingual contrastive verb valency lexicon, the Contragram Verb Valency Dictionary of Dutch, French and English (henceforth: CVVD). This project is funded by the Research Fund of the University of Ghent ('Bijzonder Universitair Onderzoeksfonds')·' The second section sketches out the rather long history of contrastive grammar research at the University of Ghent, focussing on the reasons why verb valency and verb complementation became Contragram's main area of research. In the third section, some key theoretical notions underlying the methodology of the dictionary will be outlined and the most important stages of this step-by-step methodology will be illustrated by means of an example. In section 4, I shall briefly consider the advantages we believe the CVVD will have over existing bilingual verb valency lexicons and over traditional translation dictionaries. Section 5 finally presents a state of affairs of the project.

2. History

The Contrastive Grammar Research Group of the University of Ghent consists of members from the departments of English, French and Dutch Linguistics. In 1988, members of these departments started joint research on the grammars of their respective languages, their first objective being to compile a descriptive contrastive grammar using a common terminological framework. Such a grammar, it was felt, would contribute to remedying the lack of agreement between terminologies and analyses adopted in the teaching of French as a second language and of English as a foreign language in Flanders (Simon-Vandenbergen et al. 1996: 7). At the end of this first project, partly funded by the Flemish Ministry of Education, a two-volume contrastive grammar was published under the title Nederlands, Frans en Engels in contrast ('Dutch, French and English in contrast'). The first volume (Devos et al.

A first grant, G O A (Geconcerteerde Onderzoeksactie - Concerted Research Action) contract no. 12052095 covered the period 1995-1998. The present grant ( G O A contract no. 12050399) covers the period 1999-2004. The supervisors of the project are Anne-Marie Simon-Vandenbergen (English linguistics - spokesperson), Johan Taeldeman (Dutch linguistics) and Dominique Willems (French linguistics). The current staff are Bart Defrancq (French linguistics), Dirk Noel (English linguistics) and Timothy Colleman (Dutch linguistics). Filip Devos was the collaborator for Dutch during the first phase of the project (1995-98).

Timothy Colleman

64

1991) dealt with the noun phrase, the second volume (Devos et al. 1992) with the sentence. More information about this contrastive grammar project is to be found in Devos (1992), Devos (1995), Devos and Defrancq (1995). While they were working on the second volume on the sentence and preparing a third volume on the verb phrase, the researchers were faced with the enormous difficulties the combination of a sentence relator and the inherent elements to this relator poses in foreignlanguage learning (Devos and Defrancq 1995: 10). Indeed, languages differ widely in the field of verb valency and verb complementation. Some classic examples are mentioned under (1) and (2). ( 1 ) a) Dutch: Ik kijk naar de televisie. b) French: * Je regarde ä la television. > Je regarde la t i l e v i s i o n . c) English: * I'm watching at the television. > I'm watching television. ( 2 ) a) English: I doubt his professionalism. b) French: * Je doute son professionnalisme. > Je doute de son professionnalisme. c) Dutch: * Ik twijfel zijn professionalisme. > Ik i e t w i j f e l zijn professionalisme. OR: > Ik twijfel a a n zijn professionalisme.

Not surprisingly, foreign-language learners experience great difficulty translating verb constructions; not only do they have to be able to select the most appropriate verb as an equivalent of a verb in their mother tongue, they also have to know how to use that verb in the correct or most appropriate construction. The above examples clearly demonstrate the need to complement the syntactic and morphological descriptions offered in (contrastive) grammars with data from the verbal lexicon, i.e. with exhaustive lexical descriptions of verb complementation patterns. Simon-Vandenbergen et al. (1996) elaborate on the lack of interest for lexicalised grammar in traditional linguistic description. Traditionally, in language teaching as well as in other areas of (foreign) language use, two tools are used side by side: a grammar and a dictionary. Neither of these deal with the grammatical aspects of the lexicon. Grammars only outline general grammatical rules and patterns, and as for dictionaries, most of them do not include much grammar: usually, grammatical information is only implicitly and fragmentarily given in the examples (see Simon-Vandenbergen et al. 1996: 8). As far as monolingual dictionaries are concerned, exploratory research by Dirk Noel (1996) has indicated that the recent so-called 'learner's' or 'pedagogical' dictionaries of English show an improvement in this regard, but this new approach has not yet found its way to bilingual dictionaries. In other words, neither traditional (translation) dictionaries, nor grammars offer an (exhaustive) description of the grammatical aspects of the verbal lexicon, and yet the need for such a description is sharply felt, in (foreign) language teaching as well as in other areas of (foreign) language use. For this reason, lexicalised grammar, or more specifically the area of verb valency and verb complementation became Contragram's new field of interest. First, in the period 1993-1994, an exploratory two-year project funded by the National Fund for Scientific Research (Belgium) was carried out, focussing on some of the most frequent verbal constructions in Dutch, French and English (see Defrancq and Martens 1994). Since 1995 then, we have been working on the compilation of the CVVD (in two phases: 1995-98 and 1999-2005, see note 1).

The Contragram Verb Valency Dictionary of Dutch, French and English

65

3. M e t h o d o l o g y

3.1. Theoretical prerequisites The CVVD will be quite revolutionary in that it will be the first valency dictionary that is fully or really contrastive. In an overview article of 1996, Filip Devos distinguishes three criteria for contrastive verb valency research: the approach to compiling a contrastive valency dictionary should be multilingual, multilevel and multidirectional (Devos 1996a: 32-33). In sections 3.1.1 to 3.1.3, I will deal with these three prerequisites in somewhat more detail, following the description in Devos (1996a: 33-40).

3.1.1. Multilingual approach A contrastive verb valency dictionary should systematically describe verb complementation patterns of two or more languages, using a uniform apparatus. Most existing valency lexicons are ruled out by this criterion, for they are monolingual, describing verb complementation in just one language, typically German or French. In fact, for Dutch, no valency lexicons exist as yet, so the CVVD will have the additional advantage of containing the first systematic description of the verb complementation patterns of Dutch.

3.1.2. Multilevel approach In a contrastive verb valency dictionary, both structures and meanings should be systematically contrasted. In other words, the user should be provided with structural/syntactic as well as lexical/semantic information concerning the verb entries of the source language and the target language(s). In doing so, one may find identical structures in languages A and Β which show a difference in meaning (as in (3)), or one may find the same meaning being expressed through different structures (as in (4)). Both examples are taken from Devos (1996a: 34). (3) same structure (kijken/regarder ä/naar NP) but different meaning: - Dutch kijken naar NP = 'sensory looking' e.g. Hij draagt een wit pak en kijkt naar de fotograaf. 'He wears a white suit and is looking at the photographer.' - French regarder ά NP = 'to pay attention to1 e.g Si I'on regarde trop aux principes, on n'y arrive pas. 'If we pay too much attention to principles, we'll never get there.' (4) same meaning ('to have a view, to look out on') but structural difference - Dutch uitkijken op NP e.g. De nieuwe woningen zullen uitkijken op de binnenplaats en op het winkelcentrum. 'The new houses will look out on the courtyard and on the mall.' - French regarder NP e.g. Autrefois, l'Auberge des temples regardait Angkor-Vat. 'Formerly, the temple's inn looked out on Angkor-Vat.'

Timothy Colleman

66

These examples suffice to make clear that a thorough account of verb complementation should contain both semantic and syntactic information. In valency lexicons that restrict themselves to enumerating the formal patterns a particular verb can enter into, the semantic difference between for instance Dutch kijken naar NP and French regarder a NP would be entirely lost.

3.1.3. Multidirectional approach Finally, the approach should be multidirectional, that is, equal attention should be paid to all languages involved. In other words, the traditional distinction between source language and target language(s) should be transcended, in order to make the dictionary equally accessible for different groups of native speakers. We know of no existing bilingual valency lexicon that answers this criterion. Most restrict themselves to enumerating the patterns a particular verb can enter into in the source language and providing translations for the target language. Their macrostructure is thus copied from traditional bilingual translation dictionaries: they are »«/directional instead of multidirectional. The methodology that has been developed for the CVVD however, allows us to treat the three languages involved on a par (see below).

3.2. The methodology of the CVVD The principles outlined in the preceding section determine the methodology of the CVVD. In this section I shall describe the different stages in the building of a CFKD-lemma, illustrating them by means of the entry for the Dutch verb betekenen (roughly equivalent to English to mean).

3.2.1. Formal analysis First, on the basis of large electronic corpora, an exhaustive list is drawn up of all the formal patterns a particular Dutch, French or English verb can enter into.2 In this formal analysis, different types of grammatical information are distinguished: distributional, categorical, relational, lexical and transformational/verb specificational information. In principle, we prefer categorical notions like NP (noun phrase), AdjP (adjective phrase), Pfin (finite clause) et cetera above functional-relational notions like Direct Object, Indirect Object, et cetera. This adds to the user-friendly nature of the dictionary, since categorical notions

1

The corpora used are the 27 and 38 million words corpora of the Instituut voor Nederlandse Lexicologie ('Institute for Dutch Lexicology'), the British National Corpus and the Le Monde 1994 corpus on CD-ROM. If necessary, the corpus data are complemented with information derived from grammars and dictionaries.

The Contragram Verb Valency Dictionary of Dutch, French and English

67

involve less interpretation and are more transparent.' It is clear however, that at a secondary level, relational labels too will have to be introduced in the analysis, for categorical notions alone do not suffice to describe verbal structures exhaustively and unambiguously. Consider sentences (5) and (6), both instances o f to feel taking an N P complement. (5) John felt an insect (on his arm). (6) John felt a fool. When necessary, relational notions will be mentioned: the pattern in (6) will be rendered as NP C|, with C| denoting a Subject Complement.

'

After my presentation, someone passed the remark that these categorical labels may well be very transparent to professional linguists, but not necessarily to language students at secondary or even undergraduate levels. We are aware that notions like noun phrase, prepositional phrase, et cetera might strike students who were never taught to use them before as rather confusing at first sight, but we believe that this should not be an insurmountable problem. After all, we will of course provide a straightforward introduction to the dictionary. Moreover, in addition to the fact that we believe a categorical approach to be more 'honest', since it involves less interpretation by the researchers, there are some very good pragmatic reasons why we prefer categorical notions as well. Firstly, if we should start from functional-relational labels, these should be specified in categorical terms anyway, for a Direct Object for instance may occur in the form of a noun phrase, a finite clause, an infinitival clause et cetera (see a, b, c respectively). a) Trading so far this year is up and the board expects "another year of good growth." b) He had obviously been sleeping out the daylight in the gloom of the kirk, never expecting that some idiot with a camera, rucksack and heavy boots would come crashing into his bedroom. c) He hadn't expected to see the girl on a racecourse so soon after Ascot. Secondly, functional-relational labels would be far more confusing than categorical labels, because of the many terminological differences. In French grammatical practice for instance, no difference is made between Indirect Object and Prepositional Object. The italicized constituents in (e) and (f) would both be labelled "Compliment d'Objet Indirecte" (see Devos et al. 1992: 27ff.). e) Cela ne I'a pas empechi d'icrire sur le cricket. f) Pour ma part, je vais, dans un premier temps, ecrire ä l'ensemble des medecins.

Timothy Colleman

68 Table 1: betekenen:

formal analysis 4

pattern

NP

NP

NP

NP aan NP

NP

NP[i'e dire' for instance would take one to the entry for zeggen/dire/say. Anyone who would like to try this out is advised to pay a visit to the collection of draft entries on our website (http://bank.rug.ac.be/contragram).

Timothy Colleman

74 (14) (a) "Not if it means us going off without you," said Snodgrass (BNC). (b) "Niet als dat betekent dat we zonder jou vertrekken," zei Snodgrass.

(15) (a) And he added under his breath: "More than enough!" without being quite sure what he meant (BNC). (b)*En hij voegde er fluisterend aan toe: "Meer dan genoeg!" zonder te weten wat hij betekende.

The CVVD lemma explicitly states that such a translation would be inadequate: in the Dutch column, a cross-reference to the verb bedoelen is mentioned under V . l . Furthermore, the approach taken in the CVVD is also characterized by non-selectivity, i.e. both differences and similarities between the languages are mentioned in the analysis (whereas many existing bilingual valency lexicons only highlight the differences). Not only does the CVVD explicitly warn learners which patterns cannot be translated by their prototypical translations, it also lists all the patterns that can safely be translated by their protoequivalents. Noel (1995) argues that a lot of traditional bilingual dictionaries seem to suffer from a "false friends syndrome": dictionary makers tend to offer a lot of different translations for a given verb, without exploring all structural and semantic possibilities of those verbs that would intuitively seem the best matches. Whether or not this is due to "an exaggerated anxiety about false friends", the fact remains that this approach results "in an incomplete picture of the semantic and syntactic correspondences of pairs like considerconsiderer, and this may lead the naive user to conclude that they are indeed, to a greater extent than is actually the case, false friends" (Noel 1995: 9). The CVVD avoids this pitfall by sticking to a non-selectivist approach: the proposed translation of a particular verb in each of the other languages is maintained throughout the whole analysis, and only when that proposed translation would not be appropriate, a reference to another verb will be given. A second major advantage to foreign language learners consists in the explicit grammatical information: this will warn the user when verbs that are prototypically equivalent meaningwise, like icouter and to listen for example select different complementation patterns (see above, (8)). In addition, this grammatical explicitness can help learners to disambiguate meanings. To a certain extent, it is possible to predict the meaning of a verb from its grammatical context. The entry for betekenen/signifier/mean will tell the user that whenever the Dutch verb betekenen is used followed by a pronominal phrase and a prepositional phrase, it can only mean 'to be of importance to somebody'. Whereas the English sentence under (16) is ambiguous between Ί do not know the meaning of the term verb valency' and Ί do not consider verb valency a matter of importance' (meaning groups I and III respectively), the corresponding Dutch sentence under (17) can only mean the latter. (16) Verb valency means nothing to me. (17) Verbale valentie betekent niets voor mij.

I hope these examples suffice to make clear that the CVVD can complement existing monolingual and bilingual dictionaries in several respects. In addition, the research that is being conducted within the scope of the project should lead to interesting theoretical insights too. The data gathered for the compilation of the dictionary have already been used for fundamental research into the relationship between syntax, semantics and the lexicon in the three

The Contragram Verb Valency Dictionary of Dutch, French and English

75

languages involved (see e.g. Colleman 2000, Defrancq 1996, Devos et al. 1996, Νοδί 1997, 1998 and Simon-Vandenbergen 1998).

5. State o f a f f a i r s

In the course of the first GOA-project (see note 1), the researchers developed the methodology described in section 3 (and in far greater detail in Devos 1996a) and made a useful start in compiling the dictionary. First, experimental entries were designed for the most frequent verbs of perception, communication and cognition. In a second stage, some slightly less frequent verbs that were referred to in the first set of entries were tackled. This first set of draft entries can still be accessed through the Contragram homepage (http:// bank.rug.ac.be/contragram). These entries are outdated in a number of respects - we have made some minor modifications to the microstructure of the lemmata and the general organization of the dictionary since then - but they will do to give an idea of what the CVVD could eventually look like. The ambitious objective of the Contragram team is to compile a contrastive verb valency dictionary that covers the 500 most frequently used non-periphrastic verbs of each of the three languages, as well as some very frequently used periphrastic verbs (e.g. aanspraak maken op, faire attention, to make a start). On completion of the project, that is by the end of the year 2004, the CVVD should contain at least 700 contrastive entries, covering an estimated 2,100 verbs. At the time of writing, August 2000, it contains some 150 contrastive entries. In addition, monolingual analyses for some 100 verbs of each of the three languages have been compiled and are awaiting contrastive treatment, so we may consider ourselves on schedule.

6. C o n c l u d i n g r e m a r k

In this paper, I have tried 1) to outline the methodology of the CVVD 2) to touch upon its advantages and applications and 3) to report on the present state of affairs of the project. Parts of it, especially sections 3 and 4 on the methodology and the advantages of the CVVD, are based upon earlier overview articles, most notably Devos (1996a) and Devos and Defrancq (1995). Some issues that I could only touch upon (the notion of protoequivalence, the deficiencies of existing (bilingual) verb valency lexicons, etc.) are dealt with in far greater detail in Devos (1996a) and some other titles from the bibliography. Practical information about Contragram (supervisors, publication lists, contact address, et cetera) is to be found on our website, which also offers a selection of draft entries.

Timothy

76

Colleman

References

Colleman, Timothy (2000): Zullen, gaan of presens? De toekomstaanduiders in het (Belgische) N e derlands. In: V. De Tier, M. Devos and J. Van Keymeulen (eds.), Nochtans was scherp van zin. Een bundel artikelen aangeboden aan Hugo Ryckeboer voor zijn 65ste verjaardag. Gent: Vakgroep Nederlandse Taalkunde, pp. 51-64. Defrancq, Bart (1995): The proto-equivalent. Contragram. Quarterly Newsletter of the Contrastive Grammar Research Group of the University of Gent 3, 4-7. - (1996): Object complements in English, French and Dutch: some observations. In A.-M. SimonVandenbergen, J. Taeldeman and D. Willems (eds.): Aspects of Contrastive Verb Valency, 125143 (= Studia Germanica Gandensia 40). Defrancq, Bart and Heidi Martens (1994): Kontrastief valentieonderzoek [Contrastive valency research] Internal reports of the F K F O project no. 8.0009.92, February 10th and June 29th, 1994. Devos, Filip (1992): Nederlands, Frans en Engels in contrast: een contrastieve grammatica voor het vreemdetalenonderwijs. Spieghel Historiael 33: 115-123. - (1995): Contrastive grammar: tenets and criteria. Papers and Studies in Contrastive Linguistics 30, 17-29. - (1996a): Contrastive verb valency: overview, criteria, methodology and applications. In A.-M. Simon-Vandenbergen, J. Taeldeman and D. Willems (eds.): Aspects of Contrastive Verb Valency, 15-81 (= Studia Germanica Gandensia 40). - (1996b): Types of information in the CVVD. Contragram. Quarterly Newsletter of the Contrastive Grammar Research Group of the University of Gent 6, 4-9. Devos, Filip, Rik De Muynck and Lutgart Martens (1992): Nederlands, Frans en Engels in contrast. II. De zin. Leuven: Peeters. Devos, Filip, Rik De Muynck and Mieke Van Herreweghe (1991): Nederlands, Frans en Engels in contrast. I. De nominale constituent. Leuven: Peeters. Devos, Filip and Bart Defrancq (1995): Contrastive Linguistic Analysis: towards a Dutch-FrenchEnglish Grammar and Valency Dictionary. In: L. Beheydt (ed.): Toegepaste lingu'istiek in de jaren 90/ Linguistique appliquee dans les annees 90/ Angewandte Linguistik in den 90er Jahren, 5-19. Gent: A B L A (= A B L A Papers 16). Devos, F., Β. Defrancq and D. Noel (1996): Contrastive verb valency and conceptual structures in the verbal lexicon. Language Sciences 18, 1-2: 319-338. Also in: K. Jaszczolt and K. Turner (eds.) (1996) Contrastive Semantics and Pragmatics. I. Meanings and Representations. Oxford: Pergamon. Noel, Dirk (1995): Do bilingual dictionaries suffer from a false friends syndrome? Contragram Newsletter of the Contrastive Grammar Research Group of the University of Gent 3, 8-9. - (1997): The choice between infinitives and that-clauses after believe. English Language and Linguistics I, 2: 271-284. - (1998) Infinitival copular complement clauses in English: Explaining the predominance of passive matrix clauses. Linguistics 36, 6: 1045-1063. Simon-Vandenbergen, A.-M. (1998): I think and its Dutch equivalents in parliamentary debates. In S. Johansson and S. Oksefjell (eds.) Corpora and Cross-linguistic Research: Theory, Method, and Case Studies, 297-317. Amsterdam: Rodopi. Simon-Vandenbergen, Anne-Marie, Johan Taeldeman and Dominique Willems (1996): Introducing Contragram or why whe need contrastive verb valency research. In: A.-M. Simon-Vandenbergen, J. Taeldeman and D. Willems (eds.): Aspects of Contrastive Verb Valency, 7-13 (= Studia Germanica Gandensia 40). VAN D A L E GROOT WOORDENBOEK NEDERLANDS-ENGELS.

Antwerpen: Van Dale Lexicografie 1986.

Ed.

W.

Martin

and

G.

Tops.

Utrecht/

Janet DeCesaris/Victoria Alsina The Representation of Figurative Senses in Learner's Dictionaries*

1. Introduction

Learner's dictionaries are the focus of much recent work in lexicography. One need only look at recent issues of the International Journal of Lexicography or the special volume published in the Lexicographica series entitled The Perfect Learner's Dictionary? to become convinced that the field is committed to the study of this particular type of dictionary. To date, most of the discussion surrounding learner's dictionaries has been concerned exclusively with English pedagogical lexicography, in particular with the four British dictionaries LDOCE, OALD, CIDE, and Cobuild. English, however, is not the only language for which learner's dictionaries have been produced. Spanish now has three learner's dictionaries aimed specifically at the foreign language learner, and there are comparable dictionaries for French and German as well. In this paper we discuss an issue that is of particular importance to the non-native speaker of a language, namely the representation of figurative senses of words in dictionaries. We will compare lexicographic practice for English with that for Spanish and, to a lesser extent, with that for Catalan in relation to this issue to question whether the practice currently dominant in English learner's dictionaries best serves the needs of learners. Learner's dictionaries were first developed in conjunction with English, and until the publication of CIDE they were exclusively associated with the major publishers of Englishlanguage dictionaries in Britain. Although the origins of these dictionaries and their differences as compared with other English monolingual dictionaries are well known, the place of learner's dictionaries within the lexicographic tradition of the Engish language, as opposed to the lexicographic traditions of other languages, has not been widely discussed in the literature. We feel that the point deserves some mention, because no matter how revolutionary or different these dictionares may claim to be, they still fall squarely within a lexicographic tradition that has been in existence for a very long time, and that tradition has certain characteristics which distinguish it from others. For example, dictionaries of English are not classified as being "usage dictionaries" as opposed to "normative dictionaries" because the English language does not have an official language academy that is authorized by society to set standard usage. Consequently, any dictionary may claim to define the standard, and all typically make statements to that effect. In contrast, in the Romance languages there is a clear distinction between usage dictionaries on the one hand, and prescriptive, standard-language dictionaries on the other because the presence of official language academies is deeply rooted and their role well defined. Another example of differences in lexicographic tradition is the representation of pronunciation. A phonemic representation is

Work on this paper was supported by a grant (BFF 2000-0834) from the Spanish Ministry of Education to the Institut Universitari de Lingüistica Aplicada, Pompeu Fabra University, Paz Battaner, principal researcher, which we gratefully acknowledge.

78

Janet DeCesaris/Victoria

Alsina

considered essential in general purpose dictionaries of English but rarely found in comparable dictionaries of Spanish or Catalan. Its absence is not simply a matter of linguistics because arguments based on transparency of pronunciation to educated native speakers hold little water if we bear in mind that general purpose dictionaries are consulted by an extremely wide range of people with varying backgrounds, not just by educated native speakers. Furthermore, in the case of Catalan, the orthography does not provide a single possible pronunciation so that a phonemic transcription is potentially helpful information. Phonemic representation in these dictionaries is thus also a matter of tradition. Because we feel that the role of tradition should not be underplayed in lexicography, we start our discussion of the representation of metaphorical senses in learner's dictionaries with a brief review of how metaphorical senses are treated in the general-purpose monolingual dictionaries of the three traditions we are concerned with in this paper.

2. English

2.1. Oxford English Dictionary The OED uses two different labels to flag non-literal meaning, figurative and transferred use. Figurative is applied to the use of words or phrases in which the meaning is not literal or concrete, and suggests a comparison whereas transferred is used for words or senses used outside their normal contexts. Berg (1993: 183) notes: There is some indication in a quotation in the Dictionary under the verb 'transfer' that James Murray may have originally thought of transferred in more general terms as a superordinate category of which non-literal applications such as figurative, allusive, etc. were subordinate categories. However, it seems apparent that this concept was not uniformly applied and that, for example, figurative and transferred are seen in most instances as two distinct categories.

The example Berg gives to illustrate the use of these two labels is the following: For example, the quotation from Ogden Nash - 'they will give you a look that implies that your spine is spaghetti and your soul is lard' - is clearly an example of the figurative use of the noun 'spaghetti'. In contrast, a quotation referring to the same word as a name for 'insulating tubing used over bare wire' illustrates the way in which the characteristic tubular form of spaghetti is applied in a 'transferred' or extended sense to wiring that resembles it. (Berg 1993: 124).

Berg's discussion implies that the difference between extended sense (which in theory should carry the label transferred) and comparison to the literal sense of the word (which should merit the label figurative) is usually quite discernible, but this may not always be the case. Osselton (1988: 248) notes that the distinction "generally seems to involve the opposition "physical/non-physical", giving the following examples: stream 'flow of water' = literal sense stream of persons (attached to a physical object other than water) = transferred sense stream of words, stream of events = figurative sense.

The Representation

of Figurative

Senses in Learner's

79

Dictionaries

The fact that the OED's dual system of labelling has not been adopted by most publishers of English dictionaries is a clear indication that many lexicographers feel uncomfortable with the distinction.

2.2. General-purpose monolingual dictionaries There are several approaches to the representation of non-literal meaning found in this type of dictionary. On the whole, American dictionaries use few labels and do not explicitly label figurative sense. Some British dictionaries, such as the Longman Dictionary of the English Language (1991) and Collins English Dictionary follow American practice and use no figurative label, whereas others, notably The New Oxford Dictionary of English (NODE) (1998), include the label figurative, although it is difficult to ascertain what criteria have been used to determine its presence. In the following definitions from NODE, one sense of inferno is labelled as figurative, whereas the parallel sense of hell is not: inferno 2 (usu. Inferno) hell (with reference to Dante's Divine

Comedy)

figurative a place or situation that is too hot, chaotic, or noisy: the inferno rush hour.

of the Friday

evening

hell a place regarded in various religions as a spiritual realm o f veil and suffering, often traditionally depicted as a place o f perpetual fire beneath the earth where the wicked are punished after death. a state or place o f great suffering; an unbearable experience: I've been through life hell.

hell \ he made

her

Furthermore, NODE has decided to list "the meaning accepted by native speakers [...] as literal and central" first, followed by subsenses that are either figurative extensions, specialized cases, or other types of extensions or shifts in meaning (1998: ix). The New Oxford Dictionary of English has opted for an approach emphasizing semantic coherence with explicit marking of figurative senses, rather than ordering senses based on historical grounds or on frequency, in the belief that it will help users to develop "an understanding of how senses in the language relate to one another" (1998: ix). It is worth noting, however, that even though this approach is applauded by many (e.g. Landau 1999), it does not represent the prevailing practice in English lexicography today.

2.3. Learner's dictionaries The four main British dictionaries differ in their representation of metaphorical senses. Cobuild and LDOCE do not use the figurative label and list senses according to frequency as determined by the data found in their corpora. Because frequency overrules other criteria in establishing the order of senses, a metaphorical sense of a word may be listed before the literal sense it derived from, as can be seen in the following entry for bloodletting from LDOCE (1995:127):

80

Janet DeCesaris/Victöria

Aisina

bloodletting η 1 killing people; BLOODSHED: The violence was a foretaste of the bloodletting to come. 2 the medical practice in former times of treating people who were ill by removing some of their blood. 3 the reduction of the number of people working for an organization.

A metaphorical sense may also be listed in the dictionary even though no literal sense is given, as can be seen in Cobuild's entry for the same word (1995: 168): blood-letting 1 Blood-letting is violence or killing between groups of people, especially between rival armies. Once again there's been ferocious blood-letting in the township. 2 Journalists sometimes refer to a bitter quarrel between two groups of people, usually people from within the same organization, as blood-letting, especially when it involves members of each group publicly condemning members of the other. His comment came as the blood-letting in Conservative ranks over their Eastbourne defeat continued.

CIDE includes the label fig. among the labels used in the dictionary, but it is used sparingly in conjunction with idiomatic expressions and even more rarely for the extended meaning of a lexical item. CIDE contains entries like the following (1995: 795): landslide FALLING EARTH, landslip [pron] η a mass of rock and earth moving suddenly and quickly down a steep slope · Landslides can be caused by earthquakes andfloods. landslide VICTORY η the winning of an election with an extremely large number of votes · The opinion polls are predicting a Liberal landslide in next week's election. · His popularity has fallen dramatically since he won a landslide victory last year. decant obj ν to pour (a liquid) from one container into another, or (fig. Br infml) to move (a person or thing) from one place to another · Wines like port and sherry are often decanted from their bottles into more attractive containers for serving. · (fig. Br infml) Visitors have to decant themselves from the buses into small boats for the tour. (1995: 353)

From a historical point of view, it seems clear that there has been a definite trend away from labelling figurative uses as such, so that at the present time most dictionaries, for native speakers and learners alike, do not make use of the figurative label. At the same time, learner's dictionaries have increasingly moved towards entries that are organized according to frequency of use because of the widespread belief that users are best served by placing the senses that they are most likely to deal with at the beginning of the entry. Frequency of use, of course, is independent of whether a sense is literal or metaphorical, and thus dictionaries in which frequency is the overriding criterion for ordering senses are likely to de-emphasize the relevance of relationships among the various lexicalized senses a word may exhibit. The tendency towards disassociation of senses can be seen in two dictionaries as wildly disparate with regards to the representation of lexical meaning as Cobuild, which favors polysemy, and CIDE, which favors homonymy. Landslide has two main entries in CIDE, whereas in Cobuild landslide is a single headword with two subsenses (1995: 93132): landslide 1 A landslide is a victory in an election in which a person or political party gets far more votes or seats than their opponents. He won last month's presidential election by a landslide...The NLD won a landslide victory in the elections five months ago. 2 A landslide is a large amount of earth and rocks falling down a cliff or the side of a mountain. The storm caused landslides and flooding in Savona.

The Representation

of Figurative Senses in Learner's

Dictionaries

81

As represented in both of these dictionaries, the two senses of landslide would appear to be unrelated to one another. The disassociation of related senses in relation to the four Biritish learner's dictionaries has been discussed by van der Meer (1999). We will come back to his comments later; we now turn our attention to the representation of metaphorical sense in Spanish and Catalan lexicography.

3. Spanish and Catalan

Spanish and Catalan shall be discussed in a single section for two reasons: (1) with reference to the point at issue in this paper, the two lexicographic traditions are very similar to each other and contrast with the situation described above for English lexicography; and (2) our comments will focus on Spanish and will only touch on Catalan because there are currently no monolingual Catalan dictionaries for foreign learners. The label fig. (figurative) has traditionally been abundantly used both in Spanish and Catalan lexicography. In Spanish, the 21st edition of the DRAE (1992), which is available on CD-ROM, contains 17,161 occurrences of this label. This information is significant because the DRAE is the main dictionary on which Spanish lexicography has been based for the last two centuries and, although it has been continuously revised and updated, the basic conception on which it was built has not changed (we might note that this point is precisely the source of most of the criticism levelled against it). As a result, it is not surprising to find that the label fig. is deeply rooted in the Spanish lexicographical tradition and is used in all Spanish general purpose dictionaries, much as it is in the DRAE. The situation is very similar for Catalan: the label appears frequently in the first Catalan normative dictionary, the DGLC (1932), on which all dictionaries of Catalan today are based. It continues to be used in the recent revision of the DGLC, the DIEC (1995); in short, the label fig. is just as firmly established in Catalan lexicographic tradition as it is in Spanish. In spite of this long-standing, widespread use of fig., this label, unlike others, has not been the object of academic study in either language, so that neither the criteria for using or omitting it nor the precise dividing line between this label and the label por ext. ("por extensiön", which is more or less equivalent to transferred) are known.1 In a review of the Diccionario para la ensenanza de la lengua espanola, G. Corpas mentions in passing that "figuradamente is used for metaphors and por extension for metonymies" (Corpas 1998: 372). Although this may indeed be the case in the DRAE, this principle is not strictly followed in most Spanish or Catalan dictionaries; rather, it seems to be a general tendency. In spite of the lack of detailed studies concerning this issue, the main tendencies in the use of the label fig. appear to be as follows2. The label (fig.) is clearly always used for non-literal senses, most of which are based on a sense that has been previously defined in the entry:

1

For example, the labels ironico and despectivo have been studied by Garriga (1996a, 1996b). We are grateful to Paz Battaner for pointing out several examples to us.

82

Janet DeCesaris/Victöria Aisina plata 1 Elemento quimico, metal noble muy düctil y maleable; buen conductor del calor y la electricidad; se alea con el cobre para la fabricaciön de moneda. [...] 5 figurado. Moneda de plata. 6 figurado. Dinero, riqueza. 7 figurado. Alhaja que conserva su valor intrinseco aunque pierda la hechura ο adorno. (DALE) [silver 1 Chemical element, a precious metal which is very soft and malleable; a good conductor of heat and electricity; it is alloyed with copper in order to produce coins. [...] 5 figurative. Silver coin. 6 figurative. Money, riches. 7 Jewel which maintains its intrinsic value, even if it has lost its shape or beauty.]

Senses 5 and 6 are metonymies, and sense 7 is derived from the previous two. As this example shows, the change of sense can be based either on metonymy or metaphor. A parallel Catalan example would be the following: absorbir v.tr. Fer penetrar i retenir a dins seu. L 'esponja absorbeix I'aigua. La sorra absorbi iota I 'aigua. Els gasos han estat absorbits pel carbö vegetal. Absorbir els teixits orgänics un veri un medicament. | Fig. Fer desaparöixer, consumir enterament (una cosa) com englotint-la. Un cos que absorbeix la llum, la calor. El joc ha absorbit tota la seva fortuna. | Emparar-se (d'algu) ocupantlo enterament. Els negocis I 'absorbeixen. (DGLC) [absorb intr.v. To cause to penetrate and to retain inside oneself. The sponge absorbs water. The sand absorbed all the water. The gases have been absorbed by the charcoal. A poison or a medicine absorbs organic tissues.| Fig. To cause to disappear, to entirely consume (something) as if by swallowing. A body absorbs light, heat. Gambling has absorbed his entire fortune. | To take hold (of somebody) occupying him entirely. His business absorbs him.] Furthermore, the change of sense can take place by omitting some of the semantic traits contained in the literal sense while maintaining others: abismo 1 m. Cualquier profundidad grande, imponente y peligrosa, como la de los mares, la de un tajo, la de una sima, etc. l i t . en sent. fig. Se sumiö en el ABISMO de la desesperaciön. 2 Infierno, lugar de castigo eterno. 3 fig. Cosa inmensa, insondable ο incomprensible. 4 fig. Diferencia grande entre cosas, personas, ideas, sentimientos, etc. (DRAE) [abyss 1 m.n. Any large, imposing and dangerous depth, such as that of the sea, a precipice, a gap, etc. It is also used figuratively. He sunk into the ABYSS of hopelessness. 2 Hell, a place of eternal punishment. 3 fig. Something enormous, unfathomable or incomprehensible. 4 fig. An enormous difference between things, persons, ideas, feelings, etc.] In this entry the label appears in the example listed in the very first sense, along with the literal meaning. Sense 2, which is also derived, is not marked with the label because it has been judged to be so lexicalized as to not invoke the physical idea of abyss.

The Representation

of Figurative Senses in Learner's

Dictionaries

83

The label fig. is at times, but not always, informative: specifically, it provides information when it relates traits from previous senses to the definition given for the new sense, as shown in the entry for Sp. dguila 'eagle': Aguila 1 Sustantivo femenino. Ave rapaz accipitriforme, de 8 a 9 dm de altura, de vista perspicaz, fuerte musculatura y vuelo rapidisimo (gdn. Aquila) [...]. 9 figurado. Persona de mucha viveza y perspicacia. (DALE) [eagle 1 Feminine noun. Accipitriform bird of prey, of between 80 and 90 cm of height, having sharp eyesight and strong muscles, and very swift of flight (gen. Aquila) [...]. 9 figurative Person of great liveliness and sharpness.] The label is also easy to understand for the user (especially the native user) when the figurative sense is based on one o f the basic cultural metaphors such as those studied by Lakoff and Johnson (e.g. generosity is openness). The following definition for Cat. abrasar 'bum, parch, consume with passion' actually contains two metaphorical senses: abrasar v.tr. Convertir en brasa ardent. Un munt de carbo abrasat. \ Comunicar una escalfor ardent. L 'aire abrasat pel sol. | Fig. Lliurar a les ardors d'una passio. L 'amor que I 'abrasa. Abrasarse d'ira. (DGLC) [burn up tr. v. To turn into burning embers. A pile of burning coal. | To give off a burning feeling of heat. The air which is burnt up by the sun. | Fig. To give up to the flames of passion. The love which burns inside him. To be burning with hate.] The second sense, 'to give o f f a burning feeling o f heat', has not been labelled fig. because it is so lexicalized as to be considered a completely independent meaning. Nevertheless, there are many entries in which the label fig. is uninformative, as the sense in which it is found is so thoroughly demetaphorized as to not bear any relation to any o f the other senses in the entry. In the entry for Sp. plancha 'metal plate, iron', it is impossible to deduce what characteristics associated with plancha acted as the source for the metaphorical sense. plancha 1 Sustantivo femenino. Lamina de metal liana y delgada [...]. 2 Utensilio de metal con forma triangular, cuya cara inferior, muy lisa y acerada, se calienta generalmente por una resistencia eldctrica; en la parte superior tiene un asa por donde se coge para planchar [...]. 12 figurado. Desacierto ο error que hace quedar en situaciön desairada ο ridicula al que la comete: hacer una plancha. (DALE) [iron 1 Feminine noun. Flat thin metal sheet [...] 2 Metal appliance having a triangular shape, whose lower part, which is very smooth and made of steel, is heated, generally by means of an electical coil; it has a handle on its upper part which is used to pick it up in order to iron. [...] 12 figurative. Mistake or error which makes the person who committed it look awkward or ridiculous: to commit such a mistake]

84

Janet DeCesaris/Victöria

Aisina

Another case in which this label can be uninformative, or even misleading, is when it is used in conjunction with a sense that has become dominant to the extent that the original sense is no longer regularly used by speakers. In this case the label establishes a diachronic relationship between two senses, but the fact that the original sense has disappeared from synchronic usage makes the label devoid of meaning for today's dictionary users: aconseguir v. tr. Arribar caminant, corrent, nedant, etc., a ajuntar-se (a una persona ο cosa que va al davant). Si no camineu mis de pressa, no el podreu pas aconseguir. Si no us torbeu pel cami, els aconseguireu abans d'arribar a la collada.\\ [...] Fig. Arribar (al fi que hom es proposa, a obtenir allö que hom desitja). Volien obrir la porta, pero no ho aconseguiren. No aconseguireu pas que desisteixi. Mai no aconseguireu el seu consentiment. Aconseguir honors, riqueses, aplaudiments, I 'aprovaciö. (DIEC) [overtake intr.v. To manage, by walking, running, swimming, etc., to overtake (a person or thing which is ahead). If you don't walk faster, you won't overtake him. If you don't stop on the way, you'll overtake them before you get to the pass. || [...] Fig. To accomplish (the purpose one is aiming at), to manage to obtain (what one wishes). They wanted to open the door, but they didn't manage to do so. You won't manage to make him give up. You'll never obtain his consent. To obtain honours, riches, applause, approval.]

In modem Catalan aconseguir is used in the sense of 'reach, achieve', not 'overtake', thus marking the sense 'reach' as being a figurative meaning of the verb would seem counterintuitive. We have seen, then, that the label fig. is widely used in general-purpose Spanish dictionaries, and that it often serves to identify transferred senses and metaphorical comparisons to a literal sense. Metaphors that are easily interpreted by a competent native speaker are often not as understandable to foreign speakers. Given that this is the case for the target users of learner's dictionaries, it is curious to observe that the three recent Spanish learner's dictionaries have broken with a well-established tradition in their language and have by and large omitted the label fig. In this respect they are probably following the trend of English learner's dictionaries.

4. Discussion

4.1. Association of senses and frequency Since use of the fig. label is accepted in Spanish and Catalan lexicography, we might ask why it has been criticized in the English tradition. The following two reasons come to mind: (a) The label fig. does not add anything to the meaning; and (b) it is impossible to develop clear-cut criteria to determine when to use the label, so that including the label increases the amount of inconsistency in the dictionary. While it is true that labels do not supply additional meaning, they do provide other types of information, most of which concerns usage. Part of the usage information that is pertinent in learner's dictionaries is surely that which is relevant to the contexts in which a given word or phrase occurs. Learner's dictionaries typically include many examples because

The Representation of Figurative Senses in Learner's Dictionaries

85

examples always provide some context. The fact that the label itself does not add information to the semantics of the entry is not, we believe, the source of the problem because much of what is present in dictionaries does not, strictly speaking, add semantic content to definitions. It is the second issue, lack of consistency, that is much more difficult to resolve. Metaphor in language is so pervasive that conventionalization of metaphorical use, often known as lexicalization or, in Zgusta's terminology, stabilization, quite naturally is at varying stages of development along a continuum for which only the end points are well-defined. Van der Meer (1999: 203) identifies three stages in sense development: A: literal meaning and literal interpretation B: literal meaning, but then metaphorical transfer C: new, independent meaning, loss of relevant association with literal (original) meaning

Use of the label fig. would seem to apply to stage Β in a synchronic sense. It could also be used in stage C, if the dictionary is designed to show the historical relationship between a transferred sense and the original sense, which we do not believe is the best approach in a learner's dictionary, although it may be valid for a general-purpose dictionary for native speakers. The rise of a new, independent sense involves loss of a part of the original meaning, as can be seen in the entry for abismo, and the problem in associating that change with a label is that labels are only understood as binary oppostions, whereas the amount of literal meaning that has been lost is rarely quantifiable in binary terms. In other words, use of a label may not lend itself to a phenomenon which is gradient in nature, and semantic change occurring in a languge community over generations is definitely gradient from the lexicographer's point of view. Of course, this issue not only exists for the label fig. but rather is inherent to all markings in dictionaries, which users tend to interpret as 100% applicable if present, 0% applicable if absent. If the lexicographer were to take a broad view of what constitutes stage B, we might easily imagine a dictionary in which fig. appears on every page. This is, more or less, the case of the DRAE, because the Spanish Royal Academy is interested in providing an historical perspective on sense development. The question is whether that broad view is the best way to represent such extension of meaning in the context of a learner's dictionary, and whether the broad role for the label is the only one possible. We like to suggest that this label can be applied more systematically than it has been in the past as it is more delimited in scope and thus could be a resource to resolve the incoherent representation of sense development. As we mentioned earlier, corpus lexicography has made the determination of frequency of use more objective, and a result of organizing entries on the basis of frequency is often the disassociation of senses. Van der Meer has cogently pointed out that such practice is not conducive to enhancing the "non-native learner's awareness of [...] synchronic etymology: the realisation that meanings may be related to other, more basic meanings" (1999: 196). He proceeds to demonstrate the problems of disassociation in definitions of the word morass, a noun which is used in all three stages of sense development: in its original, literal sense, in a figurative sense that still brings to mind the literal sense as evidenced by collocations, and in a more recent, independent sense. Van der Meer states, "It is an interesting question to what extent the full source meaning (the literal meaning), as 'backdrop', keeps lurking in the background even there." (1999: 205). Perhaps because he deals with a noun,

86

Janet DeCesaris/Victoria Alsina

his arguments for organizing dictionary entries to show the relationship between literal and non-literal senses appeal to learners' needs to grasp "subtle shades of meaning." We agree and would like to show that such reorganization has benefits for learning other, important information as well. We will now look at how the metaphorical senses associated with two verbs are represented in learner's dictionaries.

4.2. The representation of metaphorical senses of verbs in learner's dictionaries Verbs are complex because they can involve several arguments, and in English they are particularly tricky for learners because of the role of prepositions. The 'backdrop' van der Meer refers to is not only related to the meaning but in the case of verbs related to preposition choice. The first verb we consider is sweep. Again, the entries discussed are taken from Cobuild and CIDE because they are at the two poles on the polysemy-hofnonymy axis. Cobuild lists 21 senses, not all of which are verbs. The first sense given, the physical sense of sweeping an area of floor or ground, contains an example of sweep in conjunction with the preposition into: She was in the kitchen sweeping crumbs into a dust pan. This same preposition is present in one of the examples in sense 9 ('swift movement into a room'), She swept into the conference room, and could have been used in sense 14 ('return to power'), in which the dictionary lists the example sweep back to power although it could also have included sweep back into power. We might note that for this sense, use of the preposition into is much better than to when the noun that is the object of the preposition can also be interpreted as a physical place, at least in American English': Clinton swept into office / ??Clinton swept to office Clinton swept into the Oval Office /??Clinton swept to the Oval Office.

The fact that sweep in its figurative sense is not shown to be derived from its literal sense makes it impossible for learners to see a connection between the two, yet the fact that the two senses share the preposition into is hardly coincidental. CIDE includes four main entries for sweep, which are related to the concepts of 'clean,' 'remove,' 'move', and 'win'. Under the first sense, that of'clean', there is an example with sweep into. CIDE lists the expression to sweep under the carpet/rug/mat under this heading because it clearly derives from the literal sense of cleaning. Under the heading 'move', the third main sense, we see another example with into, The National Party swept into power..., although it is not clear to us that this example is well placed because the next main heading for sweep is 'win' and the dictionary states that you can say a sweeping victory (i.e. why is The National Party swept into power ... an example of movement and not one of winning?). What is clear from the entries both in Cobuild and in CIDE is that the dictionaries do not explicitly show that the range of prepositions associated with figurative uses of sweep is related to the range of prepositions that is possible when the verb is used in its literal sense.

5

British English and American English often differ in terms of preposition choice.

The Representation

of Figurative Senses in Learner's

Dictionaries

87

We will examine another case, that of the verb weave, to see if the same situation holds. Cobuild lists six senses for this word, and only gives an example of the phrase weave into with a derived sense of the verb in sense 6: If you weave details into a story or design, you include them, so that they are closely linked together or so that they exist alongside each other; used in written English. She weaves imaginative elements into her poems.

If we consider the literal use of weave, we note that strands or threads are woven into cloth, so it appears that the combination weave into is another case in which the choice of the preposition in the figurative usage depends on the preposition used in the literal sense. This link cannot be inferred from the information in Cobuild. The entry for weave in CIDE, on the other hand, includes the combination of weave with into in the first sense listed, the literal sense. CIDE also highlights the combination of weave with together (as can be seen in the second sense, under the heading 'twist'). weave Imake cloth] [...] to make cloth by repeatedly crossing a single thread through two sets of long threads on a loom [...] This type of wool is woven into fabric which will make jackets. weave [twist) [...] to twist (long objects) together, or to make (something) by doing this · We were shown how to roughly weave ferns and grass together to make a temporary shelter. [...] · (fig.) The biography is woven from (= combines,) the many accounts which exist of things she did.

CIDE, however, does not include an example of the figurative use of weave into.

4.3. Evidence from the British National Corpus To confirm that preposition choice in figurative uses is dependent upon preposition choice in literal contexts, we consulted a sample set of data taken from the British National Corpus. Five hundred examples of weave were chosen at random. Of these, 61 (12.2%) contained the preposition into·, 12 examples were of weave in the literal sense of weaving cloth and 48 were of weave in a figurative context. Together was found in a total of 12 examples, 5 of which exhibited the literal sense and 7 of which a figurative sense. The fact that weave into, understood metaphorically, is found in close to 10% of the total sample clearly establishes that the occurence of the preposition into is not coincidental, and we may attribute its presence to the fact that the core meaning of weave also chooses into. Preposition choice in English is not simply a matter of collocations; rather, learners are well aware of the fact that verbs and prepositions combine to form semantic units. It would seem that learner's dictionaries should strive to represent as much information relative to preposition choice as possible because it is so opaque to non-native speakers. We suggest that some opacity would subside if the fig. label were used regularly to identify metaphorical senses. In our opinion the labels used in the CIDE entry for weave, then, are well placed, and we wonder why the editors of this dictionary have used the label so sparingly.

88

Janet DeCesaris/Victöria

Aisina

4.4. Limiting application of the figurative label Metaphor is so characteristic of language use that one may rightly ask where the limits on the use of the figurative label should be placed. We believe lexicographers working on learner's dictionaries must bear the following two points in mind: (a) the main purpose of the label is to associate related senses with a view to dependent lexical choice (e.g. preposition choice, collocations); and (b) such relationships will only be understood if the literal sense is still widely in use. Following these guidelines would eliminate the label from entries in which no literal/figurative contrast is possible, as in the entry for desliz 'mistake' from the Vox Spanish learner's dictionary: desliz 1 m. fig. Falta ο equivocaciön poco grave: por ese pequeho ~ se estropearon todos los cälculos; cometiö un ~ al mencionarlo delante de ellos. 2 fig. Falta moral, especialmente la que tiene relaciön con el sexo: se habla de sus frecuentes deslices con hombres del mundo deI espectäculo. [slip 1 m. fig. Minor fault or mistake: because of that little all our calculations went wrong; he committed a ~ when he mentioned it in front of them. 2 fig. Moral error, especially one related to sex: People speak of her frequent ~ with men in show business.]

It would also eliminate the marking from entries like that for plancha discussed earlier because the label does not clearly associate one meaning to another.

5. Conclusion

We have seen that English-language dictionaries differ in their treatment of the metaphorical extension of meaning, and marking such extension of meaning by means of a usage label is not consistent. Such a label, however, is commonplace in the other lexicographic traditions we have discussed, those of Catalan and Spanish. The availability and increased technological sophistication associated with the use of large corpora has made determining frequency of use a more objective matter, and because dictionaries aim to represent the language of a speech community, and not that of an individual lexicographer, such empirical data is used by several publishers to determine the order of senses. This approach results in dictionary entries that do not attempt to relate the various senses of a single word. Dictionary entries that aim to show sense development point learners in the right direction in their efforts to grasp the wide range of contexts many words are used in. Importantly, in the case of verbs such sense association can also help them come to grips with preposition choice in English. In other words, the use of into in Clinton swept into office is not coincidental because in the literal use of the word sweep the depository of whatever is being swept is introduced by into (swept into the dustbin). One way to explicitly relate senses is that which has been chosen by traditional Catalan and Spanish lexicography - by use of a label. By contrast, we would suggest that the fig. label is unnecessary for multi-word expressions for which a literal compositional semantic interpretation is not possible because the very inclusion of the expression in the dictionary should be enough to alert users to the

The Representation

of Figurative Senses in Learner's

Dictionaries

89

unpredictable meaning. The phrase be in the same boat (e.g. Don't worry too much; we 're all in the same boat) merits inclusion in a learner's dictionary because a literal interpretation is highly unlikely. Excluding lexicalized idioms would represent a more limited presence for this label in comparison with traditional Spanish and Catalan lexicography but might prove to lend itself to systematic application. The fig. label has been criticized in the past because it has been applied erratically, but if it is assigned a specific role, w e believe it can prove helpful to non-native learners. Whether by means o f labelling or otherwise, metaphorical extension o f sense cannot be separated from the contexts in which the literal sense o f the word is used, and w e believe learners would be better served by these dictionaries if that relationship were made more explicit.

References

(a) Cited Dictionaries CIDE = CAMBRIDGE INTERNATIONAL DICTIONARY OF ENGLISH. Cambridge: Cambridge University Press 1995. CLAVE, DICCIONARIO DE USO DEL ESPANOL ACTUAL. Madrid: Ediciones SM. 1996. COBUILD = COLLINS COBUILD ENGLISH DICTIONARY. London: Harper Collins 1995 [1987], DALE = DICCIONARIO ACTUAL DE LA LENGUA ESPANOLA. Director: Manuel Alvar Ε querra. Barcelona: Vox 1993. DRAE = DICCIONARIO DE LA LENGUA ESPANOLA. Real Academia Espaflola. Madrid: EspasaCalpe 1992 [1780], DGLC = Fabra, Pompeu. DICCIONARI GENERAL DE LA LLENGUA CATALANA. Barcelona: Edhasa 1993 [1932], DIEC = DICCIONARI DE LA LLENGUA CATALANA. Institut d'Estudis Catalans. Barcelona: Enciclopedia Catalana, et al. 1995. LDOCE = LONGMAN DICTIONARY OF CONTEMPORARY ENGLISH. Harlow: Longman 1995 [1978], LONGMAN DICTIONARY OF THE ENGLISH LANGUAGE. Harlow: Longman 1991. OALD = OXFORD ADVANCED LEARNER'S DICTIONARY. Oxford: Oxford University Press 1995. OED = OXFORD ENGLISH DICTIONARY. Oxford: Clarendon Press 1991 [1884-1928], SALAMANCA. DICCIONARIO DE LA LENGUA ESPANOLA. Madrid: Santillana 1996. THE NEW OXFORD DICTIONARY OF ENGLISH. Ed.: Judy Pearsall. Oxford: Clarendon Press 1998. VOX-ALCALÄ DE HENARES. DICCIONARIO PARA LA ENSENANZA DE LA LENGUA ESPANOLA. Barcelona: Bibliograf 1995.

(b) Other Literature Berg, Donna Lee (1993): A Guide to the Oxford English Dictionary. Oxford: Oxford University Press. Corp as Pastor, Gloria (1998): M. Alvar Ezquerra. Diccionario para la ensehanza de la lengua espahola. International Journal of Lexicography 11.4, 369-373.

90

Janet DeCesaris/Victoria

Alsina

Garriga, Cecilio (1996a): La marca de irönico en el DRAE: de Autoridades a 1992. In: Forgas, E. (ed.) Lexicoy Diccionarios, 105-131. Tarragona: Departament de Filologies Romäniques, Universität Rovira i Virgili. - (1996b): Las marcas de uso: despectivo en el DRAE. Revista de Lexicografta I, 113-147. Lakoff, George & Mark Turner (1980): Metaphors we live by. Chicago: University of Chicago Press Landau, Sidney (1999): Review of: Judy Pearsall (ed.) The New Oxford Dictionary of English. International Journal of Lexicography 12.3, 250-255. Van der Meer (1999): Metaphors and Dictionaries: The Morass of Meaning, or How to Get Two Ideas for One. International Journal of Lexicography 12.3, 195-208. Osselton, Ν. Ε. (1988): Figurative Words: Modern Practice and the Origins of a Labelling Tradition. In: Hyldgaard-Jensen, K. & A. Zettersten (eds.) Proceedings of the Third International Symposium on Lexicography May 14-16, 1986, 239-49.

Bernhard Diensberg Old French Loanwords of Germanic Origin Borrowed into English

1. Introduction

The fact that there are quite a few common Germanic borrowings from Latin (though rarely found in Gothic) which reflect early contacts between the Roman Empire and the adjoining Germanic tribes has become a kind of received wisdom and is widely accepted.

2. Latin Influence on Old English

Long before the Germanic tribes (Angles, Saxons, Jutes and Frisians) left their homesteads on the continent about 450 AD and took possession of the southern and eastern parts of Britain, they must have been in contact with the Romans who had occupied considerable parts of western and southern Germany. Consequently, their dialects borrowed quite a few words from Latin that are found in almost all Germanic languages - the so-called continental borrowings (cf. Baugh/Cable4, 1993: 77-79). As a rule the following examples are frequently quoted: OE strwt 'street', G Straße (< L via strata 'paved road'); OE win 'wine', Gothic wein, G Wein (< L vmum)\ OE mynet 'mint', G Münze (< L mon&a); OE cycen 'kitchen', G Küche.(< L coquind). Apart from mint, ModE pound belongs to the earliest Germanic borrowings from Latin: OE pünd, ModE pound, OHG pfunt, HG Pfund (cf. Gothic, Old Norse, Old Low Franconian pund). The etymological dictionaries propose a derivation from Latin pondö, related to pondus 'weight', itself derived from L pendere 'to weigh' (e.g. BDE: 825b, s.v. pound1 \ cf. also Kluge/Seebold23, 1999: 628a, s.v. Pfund, 8th c.). Very few Latin words were borrowed from the Britons (or Celts) who had lived under Roman rule for almost four centuries: OE ceaster, a place-name element ( 'semer' > 'acquerir par le travail'. Furthermore, in this and many other cases both geographical and chronological considerations ultimately exclude Old Low Franconian as the donor language unless we are prepared to assume a borrowing from Gallo-Romance.' Not only is there OF gaaignier, ModF gagner (> Modern English gain v.), but nearly all Romance languages - both in the Eastern and Western part of the Roman Empire - do exhibit continuations of a much older Germanic root; e.g. Italian guadagnare, Spanish guadanar 'to mow' and ganar 'to gain', Catalan guanyar, Occitan/Portuguese gazanhar, Portuguese ganhar (see OED, s.v. & ODEE: 385a, s.v.; cf. also Corominas/Pascual III: 234a, s.v. guadana 'scythe' - cf. guadanar 'to mow' - Corominas/Pascual III: 65b, s.v. ganar 'to gain1). - Cf. Medieval Latin ganäre which is due to the vernacular form (Corominas/Pascual III: 65b, s.v. ganar 'to gain'). The Romance languages presuppose either *gwadanäre (Spanish ganar) or *gwadaniäre (French, 4

6

M.K. Pope, From Latin to Modern French with Especial Consideration of Anglo-Norman. Phonology and Morphology, Manchester, 1934/52, pp. 1-13; W. von Wartburg, Evolution et Structure de la Langue Fran a is occasionally attested in (Gallo-)Romance, e.g.: L domitäre - ModF dompter/AF donter - danter - ModE to daunt; AF/ME dame (< L dominam), CF dan(s), AF/ME daun(s) (< L dominus), CF dangier, AF/ME daunger (< L *dom[i]-niarium). Cf. Rothwell, William. 1999."Aspects of Lexical and Morphosyntactical Mixing in the Languages of Medieval England", in: Multilingualism in Later Britain. Proceedings of the conference held at Aberystwyth in 1997. Edited by D.A. Trotter (Cambridge: Brewer, 2000), pp. 213-232.

96

Bernhard

Diensberg

2.1.2 regain v.1" borrowed from AF regainer, rewainer 'to retake, win back' (AND: 614a), allegedly from *waiöanjan (anfrk.) 'auf die Weide fuhren' (FEW 17: 461a); cf. DEAF G, 1974: 11: regäaignier, gain sb. 2 & vb.2 (for the etymology); may be related rowen n. (cf. BDE: 418a. s.v. gain n. ; ODEE: 385a, s.v. gain n.)

2.1.3. rowen n. second crop of grass/hay (dial. & US)" ME rewayn (14th c., OED) and ryweyn (15th c., OED), rowayne (15th c , OED), probably related to AF rewainer, a variant of regainer 'to retake, win back'. OED, s.v., proposes ONF *rewain (cf. Modern Picard rouain, Norman revouin), OF/ModF regain, cf. AL rewaynum (14th c„ OED), see gain sb.2 & vb.2 for the etymology of the second element (DEHF, 1998: 326a, s.v. gagner/gain: of Germanic origin). ModE /u:/ is due to the alternation /ΰ ~ iu/ in a smaller subgroup of French loanwords (see Diensberg 1985: 65). On the other hand, borrowings of Germanic origin may be restricted to Gallo-Romania, i.e. (Old/Middle) French and its dialects, (Old and Modern) Proven9al - also referred to as Occitan (= Modern Provencal) - including Franco-Provencal. It stands to reason that the items in question must have been borrowed after the fall of the Western Roman Empire (476/480 AD). Thus, both Frankish (5th to 9 th cies. AD) or Old Low Franconian (clOOO1200 AD) could have been the donor languages. As a rule, both the BDE and KDEE editors - though not always consistently - postulate an unattested Frankish etymon (where the FEW editors propose an Old Low Franconian origin; see above), e.g. abandon, ambush and affray (see below). However, the ODEE (OED) editors generally adopt the label Germanic to denote the ultimate origin of an item, regardless whether the word is restricted to Gallo-Romania or found in all Romance languages and/or dialects. At a later date Old Norse comes into the picture with the conquest of and the settlement in the maritime province of Gallia which was consequently called Normandy after its Northern Germanic occupants (10th c. AD). - I will use the label Gallo-Romance for the postulated etymons with all these cases, e.g. affray, defray, dismay (see below).

2.2. A Gallo-Romance etymon of Germanic origin The FEW editors, III 293a, s.v. *exfridäre 'jemanden in seinem Frieden stören' ['(literally) to take s.o. out of his peace'] is a good example. The FEW editors (III: 293b) point out that the continuations of Germanic *frijju are restricted to Gallo-Romance exlusively. Therefore the label Gallo-Romance may be both appropriate and fully justified with *exfridäre affray and similar cases.

"' The etymons of regain v. seem to be restricted to Gallo-Romance. The etymons of rowen n. seem to be restricted to Gallo-Romance.

Old French Loanwords of Germanic Origin Borrowed into English

97

2.2.1. affray v. quarrel noisily, brawl (obs.) from Anglo-French affraier, effre(i)erlesfre(i)er 'to terrify, frighten1, originating from Gallo-Romance *exfredäre (instead of Vulgar Latin *exfridSre) 'jemanden in seinem Frieden stören' [(literally: to take s.o. out of peace] which is restricted exlusively to GalloRomance. The Vulgar Latin verb is formed of ex- 'out of and Frankish *frithu 'peace', cf. Old High German fridu, (High German Friede), Old Saxon frithu, Old English frif), fri(o)Jju. See also afraid adj., fray n. and defray v. - cf. also array, astray, bray, dismay, pay, ray, stray.

2.2.2. affray n. noisy quarrel, brawl, disturbance, alarm, fright (ca. 1303) borrowed from Anglo-French effrei, affrai(e), esfrei dismay, consternation, from affraier, effre(i)erlesfre(i)er to terrify, frighten, from Gallo-Romance *exfredare (instead of Vulgar Latin *exfridSre), literally: 'to take s.o. out of peace'. The Gallo-Romance verb is formed of ex- 'out of and Frankish *frithu 'peace', cf. Old High German fridu, (High German Friede), Old Saxon frithu, Old English frip, &i(o)pu. (KDEE: 21b; BDE: 18a; ODEE: 17b)

2.2.3. defray v. pay, settle, from Middle French defraier, desfraier (< OF des- + fraier 'to cover expenses'; from OF frais costs, expenses, not related to affray (of different etymology; see, however, Gamillscheg's proposal below). FEW III: 755b, s.v. frayer 'faire les frais' (= to cover/defray costs/expenses), gives OF desfraitier, desfretier (see Godefroy II: 589a, s.v.; cf. T/L II/2: 1594) and OF defraier (see Godefroy IX: 336b, s.v.; cf. T/L II/2: 1594) as derivations from Latin *fractiäre, from fractus, p.ppl. o f f r a n g e r e (FEW III: 752b, s.v.). Actually, it was the intention of Walter von Wartburg (and of the subsequent editors) of the Französisches Etymologisches Wörterbuch (FEW) to provide a (comprehensive) survey of Gallo-Romance vocabulary (i.e. Eine Darstellung des gallo-romanischen Wortschatzes). Basel: Zbinden; Leipzig: Trübner and Tiibingen: J.C.B. Mohr, 1922ff. - A revised edition of the FEW is now under way - to cover all vocabulary of Gallo-Romance origin, as its subtitle makes clear (cf. W. v. Wartburg, FEW I/A-B: Vorwort, p. VI). However, Gamillscheg, EWFS 2 : 449b, s.v. frayer 2 & 447b, s.v. frais, proposes a Germanic origin for OF fret, frait obl.sing.Ifres pi., namely Frankish *fridu, quoting Medieval Latin fredum 'fine for the breach of peace'. Unless we assume a derivation from OF fret, frait obl.sing. for the variant desfraitier, desfretier, this form cannot be accounted for by Gamillscheg's etymology. (BDE: 260b, s.v. defray, accepts the FEW etymon; ODEE: 252a, s.v. defray, agrees with Gamillscheg's etymology). - Gamillscheg's etymology (EWFS 2 : 447b, s.v. frais) would presuppose a Gallo-Romance verb of the type dis-, de- + *fredSre. -Cf. above affray v. 'to quarrel noisily'.

98

Bernhard

Diensberg

2.2.4. dismay v. trouble greatly. About 1300 demayen\ earlier dismayen (probably before 1300), borrowed from desmai-, strong stem of Anglo-French desmaier, desmaer to dismay (AND: 173a), from Latin intensive de- + Anglo-French esmaier, esmaer to dismay (AND: 262b), from Romance *desmagäre, ultimately derived from Romance des- (< L dis-) + Germanic *maj- "to be able', related to OE mcej majon, ModE may)·, Anglo-French/Old French esmaier, esmaer to dismay, from Romance *exmagäre, cf. DEHF: 250b, s.v έηιοί. - (BDE: 286a; ODEE: 274a; KDEE: 372a).

3. Phonological problems concerning (Gallo-)Romance borrowings from Germanic dialects

Latin (syllable-) initial v- rendered by Romance gu- and w- (in peripheral dialects), e.g. AF gaster, guaster 'to lay waste, damage' (AND: 331a) (< L vastus, vastäre), ModF gäter; cf. ModE waste v./adj; OF gai'ne n. (< L vagina 'fourreau d'epee'; DEAF G, 1974: 53; cf. DEHF, 1998: 326b). Germanic (syllable-)initial w- rendered by Romance gu- and w- (in peripheral dialects), e.g. gain v., allegedly from *waiöanjan (anfrk.) 'auf die Weide fuhren', see FEW 17: 461a & the discussion above); ModE war n., from Anglo-French werre/Central French guerre (< Frankish *werra 'Wirren' < Germanic *werso), see FEW 17: 567a); cf. ModE wage/wager, ward/guard; warranty/guarantee, cf. ME warente 'to protect'; ME guerdon/gerdon - ModE reward; ME warisshe/warice to heal' - ModF guerir/guerison - ModE warison; ModF garenne - ModE warren.12 Both Middle and Modern English show variation between the velar semi-vowel /w-/ and the voiced stop /g-/ as do both Anglo-French and its Central French counterpart. The problem to be addressed concerns the representation of Latin (syllable-) initial v- as for instance with Latin vastus, vastäre) which appear as Romance gu- (sometimes reduced to g-) and w- (in some Medieval French dialects), Italian guasto, guastare, Spanish/Portuguese gasto, gastar, Central French/Anglo-French gaster, guaster (> ModF gäter), Old Northern French/Anglo-French waster (AND: 331a, s.v. gast, guast, vast, wast adj. & gaster, guaster, waster v.), ME waste, wasten ~ ModE (to) waste·, cf. the latinisms devastate/devastation. Whilst traditional Romance scholarship assumed (a wholesale series) of blends with or influence from a phonologically and semantically similar Germanic etymon, recent research has taken recourse to phonological considerations. In the above case the incorrect Germanic forms *wastan, OHG (sic!) *wastjan, Germanic *wöstjan (Lausberg, Rom.Sprachwiss. II: § 303; cf. Meyer-Lübke, REW 5 , s.v. 9168 vastäre), were postulated.

12

Meier, H., "Lat. v- und die alten Germanen", Zeitschrift für deutsche Wortforschung, N.F. 16, 1960: 32-46; cf. Meierl986: 47; Lausberg, Heinrich. 1967. Romanische Sprachwissenschaft II: Konsonantismus, Sammlung Göschen, Bd. 250. Berlin: Walter de Gruyter, §§ 303, 350.

Old French Loanwords

of Germanic

Origin Borrowed

into

English

99

A rather simplistic hypothesis had Latin initial v- replaced by Germanic w-, whilst the labio-dental v- arose rather late as a continuation of Germanic ß- (< I-E bh-) (cf. Krahe/Meid, German. Sprachwiss. I, §§ 77-80) and still later in (West) Germanic dialects due to the voicing of medial -f- in contact with voiced segments (cf. Krahe/Meid, German.Sprachwiss. I, § 73). Discarding the psychologically oriented approach which had led to the assumption of a wholesale series of blends with or influence from a phonologically and semantically similar Germanic etymon, H. Weinrich, Phonologische Studien zur Romanischen Sprachgeschichte, 1969: 103, starting from sentence phonetics, assumed both a weak variant v(occurring between and after vowels) and a strong variant g(u)- (occurring in initial position and after consonants). These variants were eventually phonemicised and generalised to any position." Harri Meier elaborates on Weinrich's hypothesis and, whilst rejecting the decisive role of sentence phonetics, he points out that the Latin phonemic system did not allow the cluster -nv- to occur within one and the same morpheme which was therefore replaced by -mb- (cf. ModF d'emblee, based on Latin involäre) or by -ng(u)-, the latter already attested in Latin inguis, lingua, extinguere/distinguere, languor. As a result both /gu/ and /v/ lost their phonemic status and were later generalised to any position so that Latin vastäre appeared as Romance *guastar, cf. Spanish/Portuguese gastar and French gater." However, both Weinrich and Meier fail to account for Romance variants with w- and their continuations in Middle and Modern English. Referring again to the continuations of Latin vastus, vastäre, all attested forms can be easily explained by the above-quoted formula concerning the development of the (Germanic/Romance) initial labio-velar cluster gw- = gw- (preserved), gw- > g- (loss of the bilabial element), > w- (loss of the velar stop) > v- (reduction to labio-dental). The above variation can be observed in a couple of word families whose Germanic origin is beyond reasonable doubt, e.g. ModE guard n./v. & guardian n. and ward n. & warden n. (< Germanic wardön 2nd weak class), OF guar-der/garder, AF garder, guarder, gwarder, warder (AND: 329a), ModF garder, Italian guardare, Spanish/Portuguese guardar, etc. (+ derived forms). ModE war n. & warrior n., ModF guerre & guerrier & guerroyer, from Germanic (Frankish?) *werra (OHG werrd) ModE warren - ModF garenne, AF garenne, garrenne, warenne (AND: 329b), from Germanic *wer- (cf. Frankish *werra), but cf. Gaulish *warenna (MedL varenna 'game park'); ModE gage and wage (+ related/derived forms, e.g mortgage and wager), derived from Germanic *wadi 'pawn' (Pfand), ModF gage·, AF gager 'to give pledge' (AND: 326a); the above discussed OF gaaignier, AF gainer, gahaingnier, etc. (AND: 326b), ModF gagner (> Modern English gain v.) and the continuations of Germanic *weidenön in other Romance languages and/or dialects. "

"

H. Weinrich, Phonologische Studien zur Romanischen Sprachgeschichte, 1969:103. Meier 1960:43-44 & 45-46; 1986:47. " Germanic (Frankish?) *werra is unattested in Romanian, which has razboi of Slavonic origin. 14

Bernhard Diensberg

100 4. The case of

maim!mayhem

Contrary to what most (etymological) dictionaries assume for maim v. (and related mayhem v.): the ultimate etymology is uncertain and a Germanic etymon is commonly postulated, but no conclusive identification has been made. The occurrence of variant forms both in Romance languages (including French which together with insular sources shows the earliest attestations) and in medieval Latin might speak in favour of a common Romance etymon instead of simply a Gallo-Romance origin. The latter hypothesis (supported by FEW16: 501b) would imply borrowings from Old French into Old Occitan and into the remaining languages of (Western) Romania. To my mind, this is undoubtedly one of the most complicated cases and the FEW 16: 500b, s.v. *maiöanjan (anfrk.) 'verstümmeln' (to mutilate), develops a rather elaborate hypothesis of two Germanic stems of the same meaning, i.e. to mutilate, namely *maiöanjan and *hamjan which had been partially merged with OFr. mahaignier (cl 160) and me(s)haignier (cl 160 A.D.) as the main results. Obviously, OFr. mahaignier is earlier than OFr. mehaignier and meshaignier, which show the negative prefix me(s)- (< Gmc mis), ModF me- and are undoubtedly due to popular etymology. This is equally true of OFr. malhaignier for mahaignier.(FEW 16: 502a). I am going to start from the hypothesis put forward by the editors of the FEW (501b502a), but I will modify or supplement it on some counts. First of all a new classification which includes both Anglo-French and medieval Latin material found under OED3, s.v. maim v. (revised entry). My classification is going to be threefold: 1. the primary OFr./AFr. variants mahaignier/mahaigner Latin and ME continuations maynhe, meygne (type I), plus the Scottish reflex manyie (type III); 2. the primary AFr. variants maheimer/mehainer that show the -m ~ -n alternation, typical of Anglo-French (cf. EME rancun, ME ransome, from AFr. rangun, ransun, *ransum, CF rangon, ModE ransom); 3. the secondary variants such as Old Occitan/Old Provensal magagnar (cl 190- AD), Catalan maganyar (c 1160- AD), Italian magagnare (1251- AD) which have been borrowed from Old French and transmitted to Italian and Catalan via Old Occitan/Old Provensal. Moreover, the two main primary forms display contracted variants (l.b.) & (2.b.). La.)

2 a.)

OF mahaignier

AF maheimer

Oprov maganhar

OF/MF me(s)haignier

AF maheiner

Catalan maganyar

OFr malhaigniö

AFr mahimer

Italian magagnare

OFr mahaigni6

AFr mehainer

Italian (dial.) macagnare

MedL mahaingniare

MedL mahemiare

MedL maganiare

3 a.)

AFr mahaignier

ME mahayme, etc.

AFr mahaignier

ModE mayhem

AFr mahaignier AFr mahaignier

Old French Loanwords of Germanic Origin Borrowed into English

l.b.) contracted forms

2 b.)

AFr maigner Scottish manyie

cf. APr maym n. MedL maimare ModE maim

101

contracted forms

cf. EME amainet, amained ME maynhe ME meygne The FEW editors, 16: 500b, s.v. *maiöanjan (anfrk.) 'verstümmeln' (to mutilate), can adduce proof of the existence of an Old Low Franconian root *maiö- later *maid- which denotes the castration of stallions and, ultimately, any sort of mutilation. The corresponding Gallo-Romance *madaniäre would have lost the medial -d- and contraction would have been the result. Cf. the case of English GAIN v., borrowed from from AF gainer, gaagnier, gaai(g)nier 'cultivate, till' (AND: 326b), cf. OF gaaignier 'cultiver la terre, labourer', which allegedly reflects Old Low Franconian *waidanjan, later *waidanjan, (Gallo-)Romance *gwaidaniSre. However, for reasons unknown, the hiatus has been preserved and an -hinserted that (if pronounced) functioned as a glide (cf. Greive 1970: 212). The spelling of ModFr. trahison, from OF traisun, from Vulgar Latin *tradltiönem, with loss of medial -d-, signals the preservation of the hiatus by means of medial -h-. However, Anglo-French traisun has not kept the hiatus and thus has yielded ModE treason. Thus, forms listed under l.a.) such as OFr mahaignier were the rule, whilst contractions such as AFr maigner were quite rare. The preponderantly Anglo-French forms under 2.a.) show the depalatalisation of medial -an- with -ain- (later -aim-) as its result, a process frequently observed in that language. The Old French verb was borrowed into Old Occitan/Old Proven9al as maganhar that found its way into other Romance languages/dialects. The medial -g- obviously owes its existence to popular etymology, having been inserted on the model of Old Provenpal magorn 'mutilated leg' (cf. FEW 16: 502a). maim v. 'to mutilate' ME maymen/mahaymen testify to the borrowing of maheimer, a variant of AF mahaigner, mahainer (AND: 397b). From Gallo-Romance *madaniäre which reflects Germanic *maiöanjan 'to mutilate', cf. Gothic *ga-maips adj. respectively ga-maidans adj. acc. pi. 'crippled' (Feist: 191b, vgl. 340a, s.v. maidjan 'to alter, forge'). The BDE: 624a & the KDEE: 842b, s.v. maim v., also quote Gothic ga-maidans adj. acc. pi. 'crippled' and Old Icelandic meidha 'to hurt' (BDE) and Old Norse meiöa 'to hurt' (KDEE) respectively. The FEW editors, 16: 500b, s.v. *maiöanjan (anfrk.) 'verstümmeln' (to mutilate), assume the existence of an Old Low Franconian root *maid- later *maid- which denotes the castration of stallions and, ultimately, any sort of mutilation. For the alternation AF -ain (final) and -aim- (medial) as found in the related noun mahain - mahaime (AND: 397b), which is at the base of ModE mayhem «.(q.v.); cf. also aim v. (KDEE: 842b; BDE: 624a; ODEE: 546b: Romance *mahagnäre of unknown origin - a form which contradicts all available evidence). mayhem n. old law (OED) ME mayme/mahayme 'mutilation', a term of French law now obsolete, reflects AF mahain - mahaime (AND: 397b); see above MAIM v. for the etymology. (KDEE: 874b; BDE: 643b; ODEE: 564a)

102

Bernhard Diensberg

amaim v. 'to mutilate' (obs.). EME amainen/amaimen testify to the borrowing of AFr. maigner, a contracted variant of AFr. mahaigner (AND: 397b), prefixed by native a-; see below MAIM v. for the etymology. There are two occurrences in two early versions of the Ancrene Riwle, namely (ca 1230, al200?) MS Corpus amainet p.ppl. 'crippled' (78b:23) M. 288/and MS Caius & Gonville (2nd half of the 13th c.) amamed p.ppl. 'crippled' (83/16) which may be an error either for amained (see Zettersten 1965: 236, s.v. ameaset) or rather for amaimed. All other MSS replaced this lectio difficilior by MS Cleopatra (A) (ca. 1230) amaset (131:8), MS Vernon (ca 1390) amaset (fol. 385c; typescript 183/28) and MS Nero (al250) amased (\2QI\0). The MED, s.v. amased, wrongly interprets MS Corpus amainet as an error for amaied 'stunned, overwhelmed'(< AFr amaier, a variant of esmaier 'to dismay' AND: 262, s.v.) which does not fit the context neither in form nor in meaning (see Zettersten 1965: 236-37). For the alternation AF -ain (final) and -aim- (medial) as found in the related noun mahain - mahaime (AND: 397b), which is at the base of ModE mayhem «.(q.v.).

5. The case of march/mark

march v. English form is based on the Old French verb marchier (T/L V: 1126), cf. Anglo-French marcher (AND: 407a, s.v. marcher1), cf. Dauzat: 445a, s.v. marcher (AF marchier 'trample', 1155). This verb is undoubtedly of Germanic origin and thus related to/or derived from Frankish markm ( 'mark one's steps' > 'walk' (13 th c., DEHF) > to march (15 th c., DEHF; later in military use) does not seem to pose any greater problems. Referring to Philip N.R. Durkin, "Root and Branch: Revising the etymological component of the Oxford English Dictionary", TPhS 97, 26 - march v.2 (draft revised etymology). mark v. the English form is based on Middle French marquer (its vocalism due to related marcher, less likely influenced by Italian marcare) cf. Old French merquier (cf. T/L V: 1483, s.v. merche, merque 'boundary'), cf. Anglo-French marc, mark, marqe 'mark (coin)' (AND: 406b); cf. Dauzat: 448a, s.v. marquer (OF merquier, 1190); EWFS 2 : 603b, s.v. marque 1. "Spur", "Kennzeichen".; cf. FEW 16: 550b, s.v. an. merki. (BDE: 634b, s.v. mark}·, ODEE: 556b, s.v. mark3). mark n. the English form is based on Anglo-French marc, mark, marqe 'mark (coin)' (AND: 406b), itself derived from the Old French verb marqu(i)er (cf. T/L V: 1483, s.v. merche, merque 'boundary') which is ultimately of Germanic origin. (BDE: 634b, s.v. mark1; ODEE: 556b, s.v. mark2).

6. A b b r e v i a t i o n s

AF Anglo-French, the French spoken and written in England after 1066 until the early 15th century, conveniently documented in the AND, less well in Godefroy. Anglo-French might be taken to refer to the later period: ca. 1250 until first half of 15th century. AL AN CF

CL EME GalloRomance F G HG I-E

Anglo-Latin - Medieval Latin as spoken in the British Isles Anglo-Norman should be avoided because it may be misleading; yet could be used to refer to the early period (1066 to ca. 1250 AD). Central French, mainly (but not exclusively) the medieval language of the Paris region. Also called Francien, a rather vague and poorly-defined term. Except for the cases where there is no ambiguity Central French should be used instead of Old French! Classical Latin Early Middle English The period after the fall of the Roman Empire in France.(ca. 500-800) French German High German Indo-European

104 L LL Mdu ME

Bernhard

Diensberg

Latin generally used to refer to both Classical Latin (CL), Late Latin (LL) and Vulgar Latin (VL). Late Latin, also named Popular Latin, largely identical to Vulgar Latin (VL), 1sl century our era to ca. 500. Middle Dutch Middle English, ca. 1000-1500.

MedL

Medieval Latin

MF

Middle French (ca. 1340-1600)

MHG

Middle High German

ModE

Modern English, also termed PDE, ca. 1700 to the present.

ModF

Modern French, ca. 1600 to the present.

obs.

Obsolete

OE

Old English, 7th century to ca. 1000.

OF

Old French: all the historical and etymological dictionaries referred to in this article regard the medieval language of the Paris region (eventually the home of standard French) as the typical representative of Old French. Except for the cases where there is no ambiguity (e.g. items borrowed after 1250/1300) Central French should be used instead of Old French. To put it otherwise, Central French will be preferred whenever the medieval language of the Paris region is contrasted with the French spoken and written on English soil after 1066. Central/Old French 9 th century to ca. 1340.

OHG Oicel OLF Opic ON OProv OS PDE Romance

Old High German Old Icelandic Old Low Franconian Old Picard, the medieval dialect of Picardy in northern France Old Norse Old Provencal Old Saxon Present Day English, also termed ModE, from ca. 1700 to the present. Romance, i.e. the ancestor of the Romance languages denoting the period after the fall of the Roman Empire. Vulgar Latin - a popular or debased variant of Classical Latin, widely attested during the period of Imperial Rome.

VL

7. Abbreviations of Dictionaries and Manuals

AND BDE B/W 6 DALF DEAF DEHF EWFS FEW KDEE MED

— — — — — — — — — —

Stone et al.; see References Barnhart, see References Β loch/Wartburg, 6'im' edition Godefroy, Dictionnaire de l'Ancienne Langue Franchise, see References Baldinger, see References Dubois, Mitterand, Dauzat, see References Gamillscheg, see References von Wartburg, see References Terasawa, see References Kurath et al., Middle English Dictionary see References

Old French Loanwords of Germanic Origin Borrowed into English ODEE OED Pope REW T/L

105

— Onions, see References — Murray et al., see References — From Latin to Modern French, see References — Meyer-Lübke, see References — Tobler/Lommatzsch, Altfranzösisches Wörterbuch, see References

References

Dictionaries Baldinger, Kurt (ed.), Dictionnaire Etymologique De L'Ancien Franqais (DEAF) G 1-G 9/10, Quebec, Canada: Les Presses de L'Universiti Laval; Tübingen: Niemeyer, Paris: Klincksieck, 19741995. Baldinger, Kurt & Frankwalt Möhren (eds.), Dictionnaire Etymologique De L'Ancien Frangais (DEAF) Hl - H3, Quebec, Canada: Les Presses de L'Universite Laval; Tübingen: Niemeyer, Paris: Klincksieck, 1997-1999. Barnhart, Robert K. & Sol Steinmetz (eds.), The Barnhart Dictionary of Etymology (BDE), Bronxville/N.Y.: H.W. Wilson, 1988. Battisti, Carlo & Giovanni Alessio (eds.), Dizionario Etimologico Italiano (DEI), 5 vols. Firenze: G. Barbara Editore, 1975. Oxford University Press. (New Ed. 1975.) Corominas, J. and J.A. Pascual (eds.). 1980/83. Diccionario Critico Etimologico Castellano e Hispanico (Corominas/Pascual). 2nd Ed. 5 vols. Madrid. (2nd Ed. 1983.) Dubois, Jean & Henri Mitterand & Albert Dauzat (eds.), Dictionnaire Etymologique et Historique du Frangais (DEHF): Larousse-Bordas, 1998. Feist, Sigmund. 1939. Vergleichendes Wörterbuch der Gotischen Sprache mit Einschluß des Krimgotischen und sonstiger zerstreuter Überreste des Gotischen, 3. Auflage, Leiden: Brill (see Lehmann, W.P. 1986) Gamillscheg, Ernst (ed.), Etymologisches Wörterbuch der französischen Sprache (EWFS 2 ). 2nd edition, Heidelberg, 1969. Kluge, F., Etymologisches Wörterbuch der deutschen Sprache, 23. erweiterte Auflage, bearbeitet von Elmar Seebold et al., Berlin/New York: Walter de Gruyter, 1995, unveränderter Nachdruck 1999. Kurath, H. and S.M. Kuhn and R. E. Lewis (eds.). 1952ff. Middle English Dictionary (MED). Ann Arbor/Mich.: The University of Michigan Press. (So far Letters A through Ρ have been published.) Lehmann, W.P. (ed.). 1986. A Gothic Etymological Dictionary. Based on the Third Edition of Vergleichendes Wörterbuch der Gotischen Sprache with Bibliography Prepared under the Direction of H.-J. H. Jewitt. Leiden: Brill Schützeichel, R., Althochdeutsches Wörterbuch, 5. Auflage (Tübingen: Niemeyer, 1995) Stone, L.W., William Rothwell & T.B.W. Reid (eds.), Anglo-Norman Dictionary (AND), London: Publications of the Modern Humanities Research Association, 1977-1992. Terasawa, Y. (ed.), The Kenkyusha Dictionary of English Etymology (KDEE). Tokyo: Kenyusha, 1997. Tobler, A. and E. Lommatzsch (eds.). 1925ff. Altfranzösisches Wörterbuch (T/L). Berlin and Wiesbaden: Steiner. Von Wartburg, W. (ed.), Französisches Etymologisches Wörterbuch (FEW), Basel: Zbinden; Leipzig & Tübingen: J.C.B. Mohr, 1922ff. (A revised edition of the FEW is now underway.)

Manuals, Monographs and Editions Baugh, A.C. and T. Cable, A History of the English Language, 4th edition. London: Routledge, 1993

106

Bernhard

Diensberg

Gamillscheg, E. 1934-36. Romania Germanica. Sprach- und Siedlungsgeschichte der Germanen auf dem Boden des alten Römerreichs. 3 Bände. Berlin. Leipzig. Goossens, J.1974. Historische Phonologie des Niederländischen. Sprachstrukturen. Reihe A: Historische Sprachstrukturen, Band 2. Tübingen: Niemeyer. Greive, Artur. 1970. Etymologische Untersuchungen zum Französischen h aspire. Romanische Etymologien, Band 3. Heidelberg: Carl Winter. Krähe, Hans und Wolfgang Meid. 1969. Germanische Sprachwissenschaft. Teil I: Einleitung und Lautlehre. 7. Auflage, Sammlung Göschen, Bd. 238/238a/238b. Berlin: Walter de Gruyter. - (1969): Germanische Sprachwissenschaft. Teil II: Formenlehre. 7. Auflage, Sammlung Göschen, Bd. 780/780a/780b. Berlin: Walterde Gruyter. - (1967): Germanische Sprachwissenschaft. Teil III: Wortbildungs-lehre, Sammlung Göschen, Bd. 1218/1218a/l218b. Berlin: Walter de Gruyter. Lausberg, Heinrich. 1967. Romanische Sprachwissenschaft II: Konsonantismus, Sammlung Göschen, Bd. 250. Berlin: Walter de Gruyter. Meier, Harri. 1986. Prinzipien der etymologischen Forschung. Romanistische Einblicke. Heidelberg: Carl Winter, 1986. Meyer-Lübke, Wilhelm and J.M. Piel. 1966. Historische Grammatik der Französischen Sprache. Zweiter Teil: Wortbildungslehre. 2. Auflage. Heidelberg: Carl Winter. Pope, M.K., From Latin to Modern French with Especial Consideration of Anglo-Norman. Phonology and Morphology, Manchester, 1934/52. (Reprinted 1952.) Schützeichel, R., Althochdeutsches Wörterbuch, 5. Auflage (Tübingen: Niemeyer, 1995). Von Wartburg, W., Evolution et Structure de la Langue Franfaise. Huitieme edition. Bibliotheca Romanica. Series Prima: Manualia et Commentationes, vol. I. Berne: Editions Francke, 1967 (first published 1946). Weinrich, Η., Phonologische Studien zur Romanischen Sprachgeschichte, Forschungen zur Romanischen Philologie, Heft 6. Münster: Aschendorffsche Verlagsbuchhandlung, 2. Auflage 1969 (first published 1958). Wollmann, Α. 1990. Untersuchungen zu den frühen lateinische Lehnwörtern im Altenglischen. Phonologie und Datierung. Münchener Universitätsschriften, Band 15. München: Fink. Zettersten, Arne. 1965. Studies in the Dialect and Vocabulary of the Ancrene Riwle, Lund Studies in English 34, Lund: Gleerup & Copenhagen: Munksgaard.

Articles and Reviews Diensberg, Bernhard. 1994."Towards a Revision of the Oxford Dictionary of English Etymology". In: Hyldgaard Jensen, K. and V. Hj0rnager Pedersen. (eds). Symposium on Lexicography VI. Lexicographica Series Maior 57. Tübingen: Niemeyer, 207-233. - (2000):"Towards a revision of the etymologies in the Oxford English Dictionary (OED 3rd edition, in progress)". In: Mogensen, J.E., V. Hjemager Pedersen and A. Zettersten (eds): Symposium on Lexicography IX. Lexicographica Series Maior 103. Tübingen: Niemeyer, 203-223. Dietz, Klaus, "review of Alfred Wollmann, Untersuchungen zu den frühen lateinischen Lehnwörtern im Altenglischen (München, 1990) in: Kratylos 37 (1992), 142-151. Durkin, Philip N.R., "Root and Branch: Revising the etymological component of the Oxford English Dictionary", TPhS 97(1999), 1-49. Meier, Harri, "Lat. v- und die alten Germanen", Zeitschrift für deutsche Wortforschung (ZdW), N.F. 16,1960,32-46. Rothwell, William. 1999."Aspects of Lexical and Morphosyntactical Mixing in the Languages of Medieval England", in: Multilingualism in Later Britain. Proceedings of the conference held at Aberystwyth in 1997. Edited by D.A Trotter (Cambridge: Brewer, 2000), 213-232.

Ken Fare Somatismen als Problem der dänischen und deutschen Lexikographie

1. Einleitung

"In der Phraseologie bezeichnet man herkömmlich alle Phraseologismen, die einen Körperteil als Komponente enthalten, als Somatismen", so Burger (1998:88). Das Phänomen ist auch unter teilweise alltagssprachlichen Bezeichnungen wie "Körper-Phraseologismus" bekannt, wie z.B. bei Braun/Krallmann (1990:74). Die Kategorie umfasst Idiome wie jmdn. auf den Arm nehmen und Auge um Auge, Zahn um Zahn. Somatismen in diesem weiten Sinn bereiten der Lexikographie sicherlich nur insofern Schwierigkeiten, als sie überhaupt zur Kategorie der Idiomatik gehören. Auf das Problem der Idiomatik in der Lexikographie haben in den letzten 30-40 Jahren schon zahlreiche Forscher aufmerksam gemacht - speziell zum Thema Idiomatik in der bilingualen Lexikographie mit Dänisch und Deutsch (Fara 2000b). 1 Über Idiomatik generell soll deshalb dieser Beitrag nicht handeln. Stattdessen soll die idiomatische Subkategorie der So-matismen besprochen werden, und zwar unter einem anderem Blickwinkel, als dies bisher geschehen ist. Meine Haupthese ist, dass es sich für die Lexikographie lohnen würde, eine neue idiomatische Kategorie einzuführen, und zwar eine Kategorie, die ich Somatismen im engeren Sinn (S.i.e.S.) nenne. Das Ziel des Beitrags ist zu zeigen, wie diese bisher scheinbar unreflektierte Kategorie der Lexikographie Schwierigkeiten bereitet, sowie nicht zuletzt wie man diese Probleme lösen kann. Dabei ist die Perspektive auf die dänische und deutsche Lexikographie beschränkt, und zwar aus dem Skopus eines dänischen Mutter-sprachlers. Methodisch ist die Untersuchung empirisch ausgerichtet, weshalb die Wörterbücher vor allem auf der Grundlage von Korpusuntersuchungen 2 überprüft werden. Der Beitrag gestaltet sich so, dass zunächst der Begriff Somatismus im engeren Sinn definiert wird. Gleichzeitig wird motiviert, warum diese Kategorie ein idiomatisches "Sonderproblem" darstellt. Im nächsten Schritt wird untersucht, wie die einzelnen Wörterbücher mit den Problemen umgehen, die diese lexikalische Kategorie der Lexikographie bereitet. Abschließend werden verschiedene Lösungsvorschläge für die dargestellten Probleme gegeben.

1

2

In der Germanistik v.a. Wissemann (1961), Burger (1983), Petermann (1983), ScholzeStubenrecht (1988), Schemann (1989), Schemann (1991), Burger (1992), Burger (1998:168-192). Wichtige Einfuhrungen in die Korpuslinguistik sind: Sinclair (1997), Mcenery/Wilson (1997), Kennedy (1998) og Biber et al. (1998).

108

Ken Fare

2. Grundlage, Definition und Empirie

2.1. Grundlage Die Inspirationsquelle dieser Untersuchung waren nicht die spärlichen und m.E. nicht besonders ergiebigen Auseinandersetzungen der phraseologischen Forschung mit dem Begriff Somatismus. Sondern sie bestand zunächst in Frustrationen, die bei meiner eigenen, usuellen Wörterbuch-benutzung (siehe Wiegand 1998) entstanden sind. Außerdem wurden Burgers (1976, 1982) Arbeiten über die sogenannten "Kinegramme" exzerpiert.' Burgers Kinegramm-Begriff hat gewisse Affinitäten zu den hier introduzierten S.i.e.S., doch die Unterschiede sind eher auffallend. Zu den wesentlichsten Unterschieden gehört, dass Burgers Kinegramme sowohl poly- als auch monolexikalische Lexik umfassen, d.h. auch Lexeme wie lächeln, blinzeln u.ä., und dass der Begriff "Kinegramm" impliziert, dass nur von solchen Lexemen die Rede ist, die eine Bewegung bezeichnen (vgl. griech. hinein 'bewegen' (Kluge 1989:370)). Der in dieser Arbeit introduzierte Somatismus-Begriff beschränkt sich nicht auf Körperbewegungen, denn dies würde Syntagmen wie rote Ohren haben und jmdm. steht der Schweiß auf der Stirn ausschließen, die logisch zur Kategorie der S.i.e.S. gehören (vgl. unten). Vor allem aber wird bei Burger das Problem der lexikographischen Kodifizierung von Kinegrammen nicht besprochen.

2.2. Definition Folgende Auswahl an polylexikalischen Syntagmen: die Stirn runzeln - rynke panden; rote Ohren bekommen/haben - fa/have rode arer; die Zähne zusammenbeißen - bide tcenderne sammen; sich die Haare raufen - rive sig i höret; sich an den Kopffassen - tage sig tii hovedet; den Kopf schütteln - ryste pä hovedet; krumme teeer; sich an die Brust schlagen - slä sig for brystet; die Hände über dem Kopf zusammenschlagen - 0

hat eine Reihe von Merkmalen gemeinsam: 1) Erstens sind sie Idiome i.e.S., d.h. lexikalisierte metaphorische polylexikalische Lexeme (vgl. Fare 2000a). 4 2) Zweitens bestehen sie oberflächensemantisch (vgl. ibid.) ausschließlich aus somatischen Sememen. 3) Und drittens handelt es sich um solche Syntagmen, die nicht nur theoretisch, sondern auch real - d.h. empirisch nachweisbar - polysem sind. Sie werden sowohl idiomatisch, als auch litteral, d.h. kompositioneil (vgl. Pinkai 1991), verwendet. 5

Der sich wiederum auf den eher volkskundlich ausgerichteten Arbeiten RÖHRIGS (Z.B. 1967) stützt. In dieser Arbeit, die eine empirisch basierte Konfrontation des dänischen und deutschen idiomatischen Inventars durchfuhrt, wird eine positiv definierte Idiom-Kategorie vorgelegt.

Somatismen als Problem der dänischen und deutschen

Lexikographie

109

W e n n m a n die Somatismen i.e.S. von den Somatismen i.w.S. abgrenzt, dann müssen die beiden letzten M e r k m a l e herangezogen werden. Somit sind mit der Faust auf den Tisch schlagen und etwas ist jmdm. ein Dorn im Auge, die von der Forschung (z.B. Lundh (1992) und Braun/Krallmann (1990)) als " S o m a t i s m e n " bezeichnet werden, keine Somatismen im engeren Sinn. Denn ersterer enthält ein nicht-somatisches Lexem, Tisch, und der andere ist nicht p o l y s e m im engeren Sinn." Dies gilt fiir die große Mehrheit der Somatismen, sowie für die Idiome allgemein: Ihre sogenannte " d o p p e l t e L e s a r t " (vgl. Burger 1998:59ff.), die von manchen Darstellungen der Idiomatik generell nachgesagt wird, ist oft ziemlich theoretisch. D e n n selbst wenn ein foreign speaker ein Idiom wie jmdm. auf der Nase herumtanzen nicht unmittelbar verstehen w ü r d e , w ü r d e er sicherlich zu d e m Schluss k o m m e n , dass das Idiom nicht litteral interpretiert werden soll und deshalb nicht unkritisch in L 2 übertragen werden kann, und zwar aufgrund eines Analogieschlusses von seinen muttersprachlichen Idiom-Kenntnissen. 7 Deshalb sind die Somatismen im weiteren Sinn, anders als die S.i.e.S., keine besonders problematische Lexemkategorie für die Lexikographie.

2.3.1. Empirische Beispiele f u r die Polysemie der Somatismen i.e.S. Ein wichtiges M e r k m a l der S.i.e.S. ist also ihre Polysemie, und dass diese nicht bloß theoretisch, sondern in der Tat empirisch verifizierbar ist. Unten seien einige Beispiele fiir S.i.e.S. angeführt, deren sowohl idiomatische als auch litterale Realisierung anhand von Korpusbelegen nachgewiesen werden kann: 1) die Stirn runzeln a) litteral "Sensoren messen den elektrischen Hautwiderstand, den Blutdruck oder die Herzschlagfrequenz. Bewegungsmelder registrieren, wenn der Proband die Stirn runzelt, Kameras halten seinen Gesichtsausdruck fest" (DIE ZEIT 26.02.98, Auf dem Jahrmarkt der Forschung) b) idiomatisch "er erzählt Witze, bei denen sein Chef die Stirn runzeln und die Beförderungschancen neu überdenken würde" (STERN 05.11.87:32) 2) den Kopf schütteln a) litteral "Kein kuscheliger Ort. Den Namen eintragen, das Alter, und dann schüttelt er den Kopf, nein, er sei nicht im Methadonprogramm'' (DIE ZEIT 19.11.98, Auf Messers Schneide) b) idiomatisch "Was immer das Motiv der BBC-Spitze sein mochte, ob Angst vor dem einflußreichen Minister oder die edle Absicht, seine Intimsphäre vor weiteren Übergriffen zu schützen - man setzte sich in

'' 7

Burgers (1998:13) Einwände gegen diesen Begriff - ich habe seine Orthographie übernommen halte ich für nicht stichhaltig. Denn "wörtlich" ist eine gemeinsprachliche Bezeichnung, deren Verwendungsbereich breiter ist, als Burger annimmt - genauso wie "buchstäblich". Aus diesem Grund ist "litteral" ein geeigneterer fachsprachlicher Terminus. Eine litterale Lesart von etwas ist jmdm. ein Dorn im Auge würde die semantische Logik verletzen. Ob die Idiomatik eine universale Kategorie ist, ist m.W. ungeklärt. Zumindest scheinen aber alle indoeuropäischen Sprachen über idiomatische Inventare zu verfügen.

110

Ken

die Nesseln. BBC-Mitarbeiter sind wütend, die Öffentlichkeit schüttelt den K o p f ' 05.11.98,

Fare

( D I E ZEIT

Piep)

3) rote Ohren bekommen a) litteral "Skifahren auf griechische Art ist ohnehin mit dem alpinen Pendant schwerlich zu vergleichen, obwohl es für rote Ohren und blaue Nasen allemal reicht" ( D I E ZEIT 1 3 . 0 2 . 9 8 , Wedeln bei den Göttern) b) idiomatisch " D a s war kein Ruhmesjahr für sie: Sachverständige und viele Prognostiker der Forschungsinstitute haben die Konjunktur 1994 so schlecht vorausgeschätzt, daß sie noch jetzt rote Ohren haben m ü ß t e n " ( D I E ZEIT 30.12.94, Aufschwung ohne Masse)

2.3.2. Exkurs: Analysenprobleme und "Idiomatizitätsindikatoren" Wenn man potenzielle S.i.e.S. empirisch untersucht, d.h. in authentischem Text, so sind die beiden Realisierungsarten nicht immer so säuberlich voneinander zu trennen, wie es bei den drei obigen Beispielen der Fall war. Manchmal lässt sich nicht eindeutig feststellen, ob das Syntagma im jeweiligen Kontext idiomatisch oder litteral dekodiert werden soll, vgl. z.B.: " R e c h t verdutzt war ein Zecher, der sich morgens um 7.30 die Augen rieb und sich allein in einem zugeschlossenen Lokal der Innenstadt b e f a n d " (MK 16.01.96, In der Kneipe eingeschlossen)

Das Syntagma sich die Augen reiben kann in diesem Text teils als ein rein extralinguistisches Morgenritual interpretiert werden, teils als ein idiomatischer Ausdruck fiir die Überraschung des " Z e c h e r " s, als er herausfindet, dass er in der Kneipe ein-geschlossen ist. Im letzteren Fall hat er sich kaum wirklich die Augen gerieben. Man kann den Text aber auch als eine bewusst zweideutige Formulierung auffassen, die beide Les-arten gleichzeitig realisiert, aber dies kann eben nicht eindeutig entschieden werden. Der Text ist ambig. Es besteht aber nicht selten die Möglichkeit, dass man sich an etwas, was man "Idiomatizitäts-Indikatoren" nennen könnte, orientieren kann. Das sind hypothetisierende, futurisierende, potenziellster ende und generalisierende Elemente wie dürfte, würden, mögen, werden, man, mancher, jeder und immer. Solche Elemente kollokieren häufig mit S.i.e.S., die idiomatisch realisiert sind, wie z.B: a) " D o c h einige Fans dürften sich verdutzt die Augen reiben, wenn die 24jährige als lispelndes, ungemein blondes Dummchen auf der Leinwand erscheint" (MK) (potenzialisierender Idiomatizitätsindikator) b) " m e i n Hausarzt (rümpft) immer die Nase, wenn ich Leberwurst... Doch das ist eine andere Geschichte" (MK) (generalisierender Idiomatizitätsindikator)

Hält man daran fest, dass S.i.e.S. polysem sind, dann muss die Funktion der Idiomatizitätsindikatoren sein, den ambigen Text zu monosemieren (vgl. Fleischer 1982:37). Denn die Idiomatizitätsindikatoren bewirken, dass das Syntagma in der konkreten Distribution ziemlich eindeutig als ein Idiom interpretiert werden muss.

Somatismen

als Problem der dänischen

und deutschen

Lexikographie

111

2.4. Zusammenfassung vom Abschnitt 2 Diese empirischen Belege machen deutlich, dass Somatismen i.e.S. in der Sprachwirklichkeit sowohl idiomatisch als auch litteral realisiert werden. Sie haben prinzipiell Kopräsenz von Idiomatizität und Litteralität (vgl. Wotjak 1986:186). Und dadurch unterscheiden sie sich von den Somatismen i.w.S. Ein S.i.e.S. wird hier also als ein Idiom definiert, dessen Konstituenten-Semantik ausschließlich körperbezogen ist, und das idiomatisch und litteral kodiert ist, was empirisch nachweisbar ist.

3. Somatismen i.e.S. kontrastiv

Für einen Dänen sind die S.i.e.S. den Kopf schütteln, die Stirn runzeln und rote Ohren haben nicht besonders auffällig. Denn auch ihre dänischen Äquivalente sind polysem. Aus diesem Grund scheint die Frage berechtigt, ob diese lexikalische Kategorie überhaupt kontrastiv relevant ist? Denn wenn die Somatismen eine "Verbalisierung körperlichen Verhaltens" (vgl. Burger et al. 1982:57) darstellen, dann müsste man doch annehmen, dass es sich um Universalien handelt? Und müssen sie dann überhaupt von der bilingualen Lexikographie behandelt werden? Dieser Einwand wäre unberechtigt, denn S.i.e.S. sind zwar möglicherweise als Kategorie eine universelle Erscheinung, doch im Einzelnen können kontrastive Unterschiede beobachtet werden:

3.1. Deutsch vs. Dänisch Das Deutsche verfügt z.B. über den Somatismus i.e.S. die Hände über dem Kopf zusammenschlagen. Wie die Definition erfordert, kann es sowohl litteral als auch idiomatisch verwendet werden, obwohl das Syntagma sicherlich in idiomatischer Realisierung am häufigsten vorkommt." Es hat im letzten Fall eine Semantik, die man mit 'sehr erstaunt, überrascht, entsetzt sein' (Wahrig 2000) beschreiben kann. Betrachtet man dagegen das Dänische, so wird man feststellen, dass eine rein konstituentenbasierte Übersetzung von die Hände über dem Kopf zusammenschlagen ins Dänische ein Syntagma ergibt, das zwar sinnvoll ist, das aber dennoch keinen lexikalisierten semantischen Inhalt hat.

*

Korpusuntersuchungen (ZK) ergeben, dass von 8 Syntagmen die Hände über dem Kopf schlagen 5 eindeutig idiomatisch sind, während der Rest ambig ist.

zusammen-

112

Ken Fare

3.2. Dänisch vs. Deutsch Kehren wir nun die Optik (Kromann 1995:118) um und machen das Dänische zum Ausgangspunkt, so kann das gleiche Phänomen anhand des Somatismus slä ud med armene beobachtet werden, das auf Deutsch etwa mit *die Arme herausschlagen wiedergegeben werden kann. Außer seiner litteralen Semantik hat das Syntagma auf Dänisch eine Reihe verschiedener lexikalisierter Bedeutungen als Idiom, darunter z.B. 'nicht zurückhaltend sein'", vgl.: "Reuter er styringsfreak. Der hvor jeg er god til at slä ud med armene, serger han benhärdt for, at der er daekning", beklager Preisler (BA) " M e n jeg har ingen formel for, hvordan man naermer sig det poetiske. Poesien er en sjaelden og meget sky fugl, hvis man kasfter for meget op og slär ud med armene, er den for laengst f l e j e t " (BA)

Im Deutschen findet man offensichtlich nicht dieses Syntagma, weder als Idiom noch als litterale Kollokation. Es scheint, dass es sich um ein dänisches lexikalisches Schibboleth (vgl. Bußmann 1990:666) handelt. Man könnte weitere Beispiele hervorheben, doch es soll hier kein vollständiges dänischdeutsches Inventar an somatischen Null-Äquivalenzen geliefert werden. Dagegen soll exemplarisch gezeigt werden, dass auch bei stark motivierten (siehe Burger 1998) Idiomen wie den polysemen Somatismen kein naives l:l-Verhältnis auf bilingualer Ebene gelten kann. Die funktionale Äquivalenz und formale Parallelität zwischen einigen dänischen und deutschen polysemen Somatismen, sowohl auf litteraler als auch auf idiomatischer Ebene, darf nicht dazu führen, dass sie von der Lexikographie als eine kontrastiv unproblematische Kategorie betrachtet werden. Polyseme Somatismen sind, mit einer Formulierung von Kromann/Kjaer (1995:2), "interlingual unvorhersagbar".

4. S o m a t i s m e n i.e.S. in der dänischen und deutschen Lexikographie

Wir wollen jetzt zur lexikographischen Behandlung der polysemen Somatismen übergehen. Die polysemen Somatismen sind eine in mehreren Hinsichten homogene Kategorie. Deshalb scheint es berechtigt davon auszugehen, dass in der Lexikographie standardisierte Prinzipien für ihre Behandlung gelten. Im Folgenden soll eine Reihe deutscher und dänischer Wörterbücher im Hinblick auf diesen Punkt untersucht werden. Zunächst soll aber der Frage nachgegangen werden, inwieweit diese lexikalische Kategorie überhaupt von der Lexikographie berücksichtigt wird. Wir fragen damit nach dem Anteil der polysemen Somatismen an der Lemmaselektion der Wörterbücher.

'

Siehe auch Reder (1998) und Becker-Christensen (1999).

Somatismen als Problem der dänischen und deutschen Lexikographie

113

4.1. Quantitativ: Zur Selektion von S.i.e.S. in der Lexikographie Fleischer (1982:19) kritisiert Phraseolexika wie Röhrig (1991 [1973]), weil hier die idiomatischen Kinegramme, und, kann man hinzufügen, auch die S.i.e.S., ausgeklammert werden. Er bemerkt, dass die Argumentation, diese Syntagmen hätten sich nicht weit genug von der ursprünglichen Bedeutung entfernt, " f ü r die moderne Phraseologieforschung nicht akzeptabel" sei. Es könne nicht die Aufgabe eines Wörterbuchs sein, auch nicht die eines idiomatischen, "das Schwergewicht auf die kulturgeschichtlich und volkskundlich 'interessanten' " (Fleischer, ibid.) Idiome zu legen, während andere Idiome aus nicht nachvollziehbaren Gründen unberücksichtigt bleiben. Von dieser Kritik ausgehend kann man fragen, inwieweit die dänische und deutsche Lexikographie die polyseme somatische Lexik berücksichtigt. Dies ist zugleich ein guter Ausgangspunkt für weitere Detailfragen zum Thema Somatismen in der Lexikographie (siehe 4.2.). Zur Beantwortung dieser Frage wurden jeweils für Dänisch und Deutsch etwa 20 S.i.e.S. zusammengestellt um feststellen zu können, inwieweit diese lexikalische Kategorie von den Wörterbüchern kodifiziert ist. Es handelt sich um folgende Wörterbücher: (Dänisch): Politikens Nudansk (= NU10), Danskordbogen (Systime) (= SY), Danske Talemäder (= DA); (Dänisch-Deutsch): DanskTysk (Gyldendal) (= GYda.,y), Dansk-Tysk Ordbog (Munksgaard) (= MU), Gads Large Dansk-Tysk (GAda.ty), Talemäder i dansk (= TA); (Deutsch-Dänisch): Tysk-Dansk (Gyldendal) (GYty^ja), Gads Large Tysk-Dansk ( G A ^ ) ; (Deutsch): Duden Deutsches Universalwörterbuch (= DUW), Wahrig Deutsches Wörterbuch (WAHR), Langenscheidts Großwörterbuch. Deutsch als Fremdsprache (LA); Duden 11: Redewendungen (DUl 1) und schließlich Wörterbuch Deutsch als Fremdsprache (de Gruyter) (GRUY). Will man einen Überblick darüber erhalten, welche Wörterbücher welche Somatismen enthalten, kann man die Ergebnisse der Suchfragen in einer Matrix darstellen. Zwei solcher Matrizen, jeweils mit Dänisch und Deutsch als Ausgangssprache, sind im Folgenden aufgeführt:

"' Alle Wörterbücher in dieser Untersuchung werden aus Platzgründen durch ein Signum wie dieses repräsentiert.

Ken Faro

114 a) D ä n i s c h D ä n i s c h e r S.i.e.S.

GY

GA

da-ty

da-ty

MU

TA

•

• -

-

+ +

•

+

+

+ +

+ +

•

•

+

+

+

+

-

-

+

•+

+

—

+

+

—

_

+

+

11. äbne munden

—

+

+

19. fä/have rede 0rer

-

+ -

+

+

+

—

+

+

+ •

+

+

+ +

+

+

+

+

+

+

+

+

+

+

+

•

+

+

+

+

+

—

•

+

•

•

+

—

•

-

+

—

+

•

-

-

16. have sved pä panden

17. vende tommelfingeren nedad 18. trille tommelfingre

+ +

-

+

9. laegge haenderne i skedet 10. slä sig pä lärene

15. traskke pä skuldrene

+

-

+ +

-

NU

•

7. vride haender 8. ryste pä haendeme/händen

12. rynke pä naesen 13. rynke panden 14. lefte pegefingeren

DA

+

1. knibe sig i armen 2. laegge armene over kors 3. slä ud med armene 4. stramme balderne/ klemme balderne sammen 5. ryste pä hovedet 6. tage sig til hovedet

SY

•

+

+

Deutsches Äquivalent sich in den Arm kneifen die Arme kreuzen (?) 0

den Arsch zusammenkneifen (?) den Kopf schütteln sich an den Kopf fassen die Hände ringen jmdm. zittern/zittert die Hände/Hand / jmds.... die Hände in den Schoß legen sich auf die Schenkel schlagen den Mund aufmachen die Nase rümpfen die Stirn runzeln den Zeigefinger heben die/mit den Schultern zucken jmdm. steht der Schweiß auf der Stirn den Daumen nach unten halten Daumen/Däumchen drehen rote Ohren bekommen/haben

Somatismen als Problem der dänischen und deutschen Lexikographie

115

b) Deutsch D e u t s c h e r S.i.e.S.

GY

GA

tv-da

da-tv

DUW GRU Y + +

LA

DUl 1

WAH

Dänisches

R

Äquivalent

-

-

+

gnide sig i ejnene

-

-

+

-

0

+

+

+

+

+

trille tommelfingre

-

-

-

-

-

-

+

-

+

+

+

+

+

fa kolde fodder

+

+

+

-

+

+

+

rive sig i häret

-

-

+

+

+

+

+

laegge haenderne i skedet

8. sich die Hände reiben

+

+

-

+

-

+

+

gnide sig i haenderne

9. die Hände ringen

+

-

+

-

-

-

+

vride hasnder

-

-

+

+

+

-

+

ryste pä haenderne/ händen

+

-

+

+

+

+

+

0

+

+

+

-

+

+

+

tage sig til hovedet

13. den Mund aufmachen

-

-

+

+

+

+

+

abne munden

14. die Nase rümpfen

+

+

+

+

+

+

+

rynke nassen

-

-

+

-

-

+

-

have/fä rede 0rer

-

-

+

+

+

-

-

have sved pä panden

17. die Stirn runzeln

-

-

+

+

+

-

+

rynke panden

18. die Zähne zusammenbeißen

+

-

+

+

+

+

+

bide taenderne sam-

19. den Zeigefinger

-

-

-

+

+

-

-

1. sich die Augen reiben

-

-

2. sich an die Brust

+

-

+

+

+

-

5. kalte Füße bekommen 6. sich die Haare raufen 7. die Hände in den

schlagen 3. Daumen/Däumchen drehen 4. den Daumen nach unten halten

Schoß legen

10. jmdm. zittern/zittert die Hände/Hand / jmds... 11. die Hände über dem

vende tommelfingeren nedad

Kopf zusammenschlagen 12. sich an den Kopf fassen

15. rote Ohren haben/bekommen 16. jmdm. steht der Schweiß auf der Stirn

men l0fle pegefingeren

heben

Die Matrizen sind wie folgt zu lesen: In der äußersten linken Spalte stehen die untersuchten Somatismen, und in der obersten die Signa der Wörterbücher. Wenn ein Somatismus im Wörterbuch kodifiziert ist, wird dies mit einem Pluszeichen markiert, Nichtvorhandensein markiert ein Strich. Ganz rechts sind, wenn vorhanden, das jeweils deutsche oder dänische Äquivalent aufgeführt. Fragezeichen markieren nicht ganz sichere Äquivalente. Aus den Matrizen geht hervor, dass die polysemen Somatismen vielfach nicht kodifiziert sind: Tatsächlich sind in etwa 40 % der Fälle die jeweiligen Somatismen in weniger als der Hälfte der Wörterbücher aufgeführt. Es gibt sogar S.i.e.S., die in keinem der untersuchten Wörterbücher vorkommen, nämlich sich in den Arm kneifen und den Daumen nach unten halten. Dass beide Somatismen im Deutschen keine Exotismen, sondern durchaus lexi-

116

Ken Faro

kalisiert sind, kann anhand von "Korpora" wie Die Zeit auf CD-ROM 1995-1999 (ZK) und Frankfurter Allgemeine auf CD-ROM 1995-1997 (FK) nachgewiesen werden. Interessant ist dabei, dass kein signifikanter Unterschied zwischen der gemeinsprachlichen Lexikographie und der Speziallexikographie (den drei Phraseolexika, TA, DA und DUl 1) festgestellt werden kann: Der durschnittliche Deckungsgrad dieser Idiom- Wörterbücher liegt knapp unter dem Gesamtdurchschnitt von 57 %. Die Priorität der Somatismen i.e.S. scheint also für beide Lexikographietypen gleich (niedrig) zu sein. Wenn ein so relativ großer Teil dieser lexikalischen Kategorie in der mono- und bilingualen Lexikographie mit Dänisch und Deutsch nicht kodifiziert ist, so liegt dies sicherlich an ihrer Zwitterstellung zwischen einerseits den "klassischen Idiomen" (vgl. Clausen 1994:48) wie das Zünglein an der Waage sein und jmdm. das Messer auf die Brust setzen und andererseits den Kollokationen. D.h. häufig vorkommenden Syntagmen ohne eine idiomatische Dimension, wie etwa sich das Haar waschen und zu Abend essen. (Zum Problem der Kollokationen: siehe z.B. Cop 1991 und Burger 1998:50ff.). Die S.i.e.S. haben an beiden Kategorien teil, und dies verursacht offensichtlich Unsicherheit beim Lexikographen. Obwohl diese Syntagmen zur "Kernlexik" (Bergenholtz/Mugdan 1990:1618) der beiden Sprachen gehören, werden sie im Allgemeinen von der Lexikographie quantitativ nicht gebührend berücksichtigt. Dass dieses Problem auch eine qualitative Komponente hat, darüber soll im Folgenden die Rede sein.

4.2. Qualitativ: Identifikation und Markierung

4.2.1. Fehlende Identifikation als polysemes Syntagma a) Unten ist der Wortlaut aus einem Artikel eines deutsch-dänischen Wörterbuchs aufgeführt: zusammenbeißen tr. bide sammen (fx die Zähne) (GYty.da)

Auf den ersten Blick scheint dieser Artikel unproblematisch zu sein, doch der aufmerksame Benutzer wird wissen, dass das deutsche Ausgangssyntagma polysem ist: es kommt sowohl in einer idiomatischen als auch einer litteralen Variante vor. Doch auf welche dieser Varianten bezieht sich das Äquivalent? Vielleicht auf beide? Das Wörterbuch gibt keine Hinweise darauf. Das Problem an dieser Stelle besteht also darin, dass das Syntagma nicht eindeutig identifiziert wird: als bloßes Idiom, als bloß litterale Kollokation, oder als polysemer Somatismus mit beiden Dimensionen, was hier am angemessensten scheint. b) Ein Eintrag aus einem dänisch-deutschen Wörterbuch: rynke panden die Stirn runzeln (GA Spanish

potato(es) (1565) tobacco (1577)

Η

potatis (1725)

Spanish

batata > patata tabaco

Tabak [—]

tobak (1629)

[not borrowed] tobak [—] (1720)

Persian

pilav

Η

Η

Hindi

bangla

Ojibwa

uisge (beatha) nindoodem

Bungalow (1941) Whisky (1808)

bungalow

Gaelic

pilau / pilaff (1612) bungalow (1676) whisky (1715) totem (1760)

Hindi

jangal

jungle (1776)

Russian > Aleutian

parka

parka (1780)

Persian > Urdu French

pay jamah

Afrikaans

trek(ken)

Urdu Chinese name Japanese

kaki

jinrikisha

Chinese

zhilaohu

French

apres-ski

Italian name Italian

(Candido) Jacuzzi graffito / graffiti tri + athlon

"Greek" (English)

pique-nique

Shanghai

pyjama(s) (1800) picnic (1826) trekking (1849) khaki (1857)

Η Dschungel (1909) Parka (1968)

whisky (1798) Η djungel parkas (1982)

Pyjama (1915) Picknick (1930) Trekking (1979)

pyjamas

Η Η

khaki

Η

rickshaw

paper tiger (1952)

Papiertiger (1963)

Η

apresski(ing) (1954) jacuzzi (1966) graffiti (1967) triathlon (1973)

Η

after ski (1980)

Jacuzzi (1982) Graffiti (1986)

jacuzzi (1988) graffiti (1985)

Η

triathlon (1984)

shanghai (v) (1871) rickshaw (1886)

picknick Η

Η

Norwegian (AO) djunke (1960) yacht (1852) potet (1755) tobak [—]

pilau / pilaff [—] (1831) bungalow (1948)

pilaff (1992)

whisky (1828)

bungalow (1960) whisky (1894)

totem (1937)

totem (1938)

jungle (1837)

jungel (1921)

parka (coat)

parkas (1958)

Η pyjamas (1913) picnic (18c)

pyjamas piknik (I960)

trek(king) (1984) kaki (1950)

trekking (1990) kaki (1960)

shanghaje (1920) rickshaw

sjanghaie (1916) rickshaw

papirtiger (1963) afterskiing (1969) [—]

Η Η

jacuzzi (1986) jacuzzi (1985) Η graffiti (1962) graffiti [-} (1976) triatlon triathlon (1984) (1984)

In the table, all dates in the A W , A O and N S O c o l u m n s are those listed in the respective dictionaries, w h e r e a s in the case o f the D A D , the incomplete dating s y s t e m has been supp l e m e n t e d b y data from other sources: the n e o l o g i s m dictionary Ordbog dansk

1955-1998

(Jarvad 1 9 9 9 ) and the authoritative Ordbog

over det danske

over

nye ord

Sprog

i

(1918-

56). A non-entry is indicated by a [ — ] , but in cases where I am absolutely certain that a

140

Henrik

Gottlieb

non-entry was indeed introduced in due time before the editorial deadline of the dictionary in question, I have entered the word before the 'no-entry sign', as with graffiti in Danish. Although we see a more uniform picture than in tables 6 and 7, one still finds it hard to believe that for instance trekking still had not entered Swedish in 1992, or that the word Triathlon was not found in German before the editorial deadline of the AW in 1989. At the other end of the diachrony, we must accept that possible German versions of totem and khaki, for instance are excluded because of their age, cf. the "Einfluß ... nach 1945" clause in the title of the AW. However, it seems odd - even regarding the lengthy explanation concerning "Alte und neue Anglizismen" (pp. 35*-37* in the front matter of the AW) - that early and stable loans like Yacht and Whisky are included in the German dictionary. Be that as it may; by comparing the four columns of Anglicisms with the English dating (cited from The Oxford English Dictionary with supplements (1928-86), and from The New Oxford Dictionary of English) one gets an idea of the often surprising time lag between the introduction of a (foreign) term in English, and the re-introduction of the term in German, Danish, Swedish and Norwegian. Whereas items introduced in English until the 19th. century show no documented use in the (other) Germanic languages until some hundred years later (cf. trekking and khaki), the twentieth-century loans in English tend to pop up in German, Danish, Swedish and Norwegian some ten or twenty years later. Naturally, with improved methods one might find the actual time lags of relay Anglicisms, and of Anglicisms in general, to be shorter. This is demonstrated by the Anglicism expert Hans Galinsky (Galinsky 1991, 216-217), who analyzed the English-German time lags of four selected modern Anglicisms. Whereas quality of life and snowmobile had time lags of 33 and 40 years, respectively, the (relay) Anglicism kiwi (meaning 'kiwi fruit') reached Germany after only five years, with the term co-generation establishing an incredible record of only two days. But, as Galinsky adds, it "has stayed an expert's term".

6. T h e f o u r dictionaries as tools f o r assessing the impact of English

One of the cornerstones in the foundations of linguistics is the fact that "linguistic science must formulate its findings in terms provided by its object of study: the theorist's language [...] is either the same as the language under study, or it is another language, but a language nevertheless. The language of the theory of language is, we might say, peculiarly contaminated. Other sciences are not afflicted in quite the same way: medical science is not, as it were, writ in blood." (Malmkjaer 1999:13)

With this monumental statement in mind, my ambition of comparing and evaluating a number of dictionaries dealing with the influence of a fifth language on four different languages, of which only one is my own mother tongue - well, even the idea of writing it all in that foreign language, English - may be deemed risky. Still, I hope I have been able to shed some light on an important issue, and on how difficult it is to establish a common ground in interlingual lexicography. A groundbreaking venture listing up to 4000 cognate Anglicisms in each of sixteen European languages (Görlach (ed.) (in print)) still has not materialized.

Four Germanic

Dictionaries

of Anglicisms:

When Definitions

Speak Louder than

Words

141

However, although this mega-dictionary will definitely fill a void, there should be little doubt that, considering the lexicographical differences encountered in the present study, the venture of bringing together lexicographical traditions from culturally and linguistically very diverse countries will produce dissimilarities on paper which are not found in reality, and vice versa. Still, lexicographical research on the impact of English in contemporary societies is of utmost importance, not least empirically based research. Although in this study we have discussed dictionaries listing thousands of Anglicisms, in sheer quantitative terms Anglicisms may not seem conspicuous in European languages today. Yet, a considerable part of the present growth and development of Western languages is triggered by English. In Danish, for instance, according to Pia Jarvad (1995, 135) the vocabulary is being reshuffled these years, and not only are ten percent of all new words of English origin; these loans typically nouns - tend to carry significant semantic weight. "Those are the words that are instrumental in creating our world view, [...] and this means that to an increasing extent we let another culture with its language govern our reality." (Jarvad 1995, 135, my translation). In the same vein, Finnish linguist Paavo Pulkkinen notices that after WW2, the number of new semantic loans in Finnish has increased more than the number of loan translations. He suggests that the reason for this shift (from direct to indirect Anglicisms) is that "numerous Finns have recently begun thinking partly along Anglo-American lines." ([...] talrika finnar nufortiden redan har börjat tänka delvis pä angloamerikanskt sätt.) (Pulkkinen 1989, 92). Until the twentieth century, the number of Anglicisms - even in cognate languages such as those investigated here - was very limited. In the preface to his (historical) dictionary, Knud Serensen states that with regard to Anglicisms in Danish, "before the eighteenth century there is very little to record - a mere sprinkling of words like hird and kirke (from the Old English period), the verb kime from the late Middle Ages, and words like daggert and pingvin from the period 1550-1700. My material contains 79 words adopted during the eighteenth century and 319 words borrowed in the course of the nineteenth century." (Serensen 1997, 3)

The number of Anglicisms in early-twentieth-century Norwegian was of the same magnitude. In one of the first studies of Anglicisms ever (Stene 1940), which includes a dictionary section, Norwegian linguist Aasta Stene commences her conclusion with these words: "Numerically, the body of English loan-words in modern Norwegian is not impressive. In this material 531 loan-words have been entered [...]" (ibid., 210). Stene later almost prophetically adds: "Present-day Norwegian borrowing from English is so considerable that it is justified to say that we have entered a period of predominant English influence on the vocabulary." (ibid., 210). Doubtless, the number of English-inspired lexical - and presently, other - items in Norwegian, as in most other languages, has grown since Stene's seminal work." Interestingly, the publication of Stene's book was interrupted by the war that led to even greater anglophone (viz. American) dominance than one could have envi-

*

For a contemporary discussion of the situation in Norway, see Graedler (1998). A historical survey of English loans in the smallest Germanic speech community, the Faroe Islands, is found in Joansson (1997).

142

Henrik

Gottlieb

saged back then, in the heyday o f fascism in Europe, w h e n non-anglophone nations were determined to conquer the world, not just the minds o f people. Finally returning to the four dictionaries under scrutiny here, w e must realize that although a lot can be gained from consulting the four works together, one should be careful not to conclude from lexicographical evidence (the entries in the four works) to lexicological realities (usage in the four speech communities). In other words, the differences bet w e e n the dictionaries - and the discrepancies found within the individual works - mean that one should be cautious w h e n trying to produce exact comparative statements concerning English influence on other Germanic languages.

References

Carstensen, Broder, Ulrich Busse and Regina Schmude (1993-96): Anglizismen-Wörterbuch. Der Einfluß des Englischen auf den deutschen Wortschatz nach 1945. Band 1 1993 (A-E), Band 2 1994 (F-O), Band 3 1996 (P-Z). Berlin / New York: Walter de Gruyter. Collins COBUILD English Dictionary (1995, second edition). London: HarperCollins Publishers. Galinsky, Hans (1991): "Americanisms, Briticisms, Canadianisms, New Zealandisms, and Anglicisms in contemporary German", in: Ivir, Vladimir and Damir Kalogjera, (eds.) Languages in Contact and Contrast. Berlin / New York: Mouton de Gruyter. Görlach, Manfred (ed.) (in print): A Usage Dictionary of European Anglicisms. [Published 2001 as A Dictionary of European Anglicisms.]. Oxford: Oxford University Press. Gottlieb, Henrik (1994): Tekstning- synkron billedmedieoverscettelse. Copenhagen: Center for Oversasttelse, Engelsk Institut, Kebenhavns Universitet. — (1997): Subtitles, Translation & Idioms. Copenhagen: Center for Translation Studies, Department of English, University of Copenhagen. — (1999): "The Impact of English: Danish TV subtitles as mediators of Anglicisms", in: ZAA, Zeitschrift für Anglistik und Amerikanistik 47 (2), 133-153. Tübingen: Stauffenburg. — (2000) "In video Veritas: Are Danish voices less American than Danish Subtitles'?", in: Screen Translation 2000: Six Studies in subtitling, dubbing and voice-over. Copenhagen: Center for Translation Studies, Department of English, University of Copenhagen, 139-159. — (forthcoming): Sample chapters on the Web as a basis for empirical studies of translations. Paper read at the Research Models in Translation Studies Conference, Manchester University 28-30 April 2000. Graedler, Anne-Line (1998): Morphological, semantic and functional aspects of English lexical borrowings in Norwegian. Oslo: Scandinavian University Press / Universitetsforlaget AS (=Acta Humaniora 40). Graedler, Anne-Line and Stig Johansson (1997): Anglisismeordboka. Engelske länord i norsk. Oslo: Universitetsforlaget. Gyldendals Tibinds Leksikon (1977): Copenhagen: Gyldendalske Boghandel. Hedlund, Lars-Erik and Birgitta Hene (1992): Länord i svenskan. Om spräkßrändringar i tid och rum. Wiken. Höfler, Manfred (1982): Dictionnaire des anglicismes. Paris: Librairie Larousse. Jarvad, Pia (1995): Nye ord - hvorfor oghvordan? Copenhagen: Gyldendal. — (1999): Nye ord. Ordbog over nye ord i dansk 1855-1998. Copenhagen: Gyldendal. Joansson, Tordur (1997): English loanwords in Faroese. Torshavn, Fannir Larsen, Fritz (1994): "More than loan-words. English influence on Danish", in: RASK. Internationalt Tidsskrift for Sprog og Kommunikation 1, 21-46. Odense: Institute for Languages and Communication, Odense University. Longman Dictionary of Contemporary English (1995; third edition). Harlow: Longman

Four Germanic Dictionaries of Anglicisms:

When Definitions Speak Louder than Words

143

Malmkjaer, Kirsten (1999): Descriptive Linguistics and Translation Studies: Interface and Differences. Utrecht: Platform Vertalen & Vertaalwetenschap. Meier, A.J. (2000): "The status of "foreign words" in English: The case of Eight German Words", in: American Speech 75 (2), 169-183. The New Oxford Dictionary of English (1998; first edition). Oxford: Oxford University Press (Clarendon). Ordbog over det danske sprog (1918-56). Copenhagen: Det danske sprog- og Litteraturselskab, Gyldendal. Oxford Advanced Learner's Dictionary of Current English (2000; sixth edition). Oxford: Oxford University Press The Oxford English Dictionary with supplements (1928-86). Oxford: Oxford University Press. Picone, Michael D. (1996): Anglicisms, Neologisms and Dynamic French. Amsterdam / Philadelphia: John Benjamins Publishing Company. Pulkkinen, Paavo (1989): "Anglicismerna i finska spräket", in: Bojsen, Else et al., (eds.) Spräk i Norden 1989. Nordisk Spräksekretariats Skrifter 10. Oslo: J.W. Cappelens Forlag, 89-93. Random House Webster's Concise College Dictionary (1999). New York: Random House Rey-Debove, Josette and Gilberte Gagnon (1980): Dictionnaire des anglicismes. Les mots anglais et americains en franqais. Paris: Le Robert. Selt6n, Bo (1993): Nysvengelsk ordbok. Lund: Studentlitteratur. Sicherl, Eva (1999): The English Element in Contemporary Standard Slovene. Phonological, Morphological and Semantic Aspects. Ljubljana: Znanstveni inätitut Filozofske fakultete. Stene, Aasta (1940) (printed 1945): English Loan-words in Modern Norwegian. A study of linguistic borrowing in the process. London and Oslo: Oxford University Press and Johan G. Tanum Forlag. Svartvik, Jan (1999): Engelska - öspräk, världssprak, trendspräk. Stockholm: Norstedts Ordbok. Scrensen, Knud (1973): Engelske lan idansk. Copenhagen: Dansk Sprognaevns skrifter 8. - (1995): Engelsk i dansk: Er det et must? Copenhagen: Munksgaards sprogserie. — (1997): A Dictionary of Anglicisms in Danish. Copenhagen: The Royal Danish Academy of Sciences and Letters.

Rolf Hesse Grammatische Schwierigkeiten bei der zweisprachigen Lexikographie'

0. Einleitung

Bei der Erarbeitung von zweisprachigen Wörterbüchern kommt es immer wieder zu Übersetzungsschwierigkeiten, die mit dem einzelnen Wort oder Ausdruck nichts zu tun haben, sondern auf allgemeine, oft außersprachliche Faktoren zurückzuführen sind, wie z.B. kulturelle, geographische oder klimatologische Unterschiede zwischen den beiden Sprachgebieten.' Aber auch Unterschiede in der grammatischen Struktur der beiden Sprachen können eindeutige Übersetzungen in den immer kontextfreien Beispielsätzen und Ausdrücken verschiedener Art verhindern oder erschweren. Zunächst ein einfaches Beispiel: Wenn der dänische Ausdruck: den 1. oktober ins Deutsche übersetzt werden soll, gibt es, wenn kein Kontext vorliegt, mindestens zwei Lösungen: Der 1. Oktober - oder: Den 1. Oktober.

Der Grund dafür ist, dass hier eine fast kasusfreie Sprache einer Sprache mit Kasus gegenübersteht. Deshalb ist es unumgänglich, den Ausdruck im Wörterbuch durch einen Minikontext (in beiden Sprachen) zu disambiguieren, z.B.: Den 1. oktober er en mandag. Kebenhavn, den 1. oktober.

Der 1. Oktober ist ein Montag. Kopenhagen, den 1. Oktober.

Dieses Problem entsteht dadurch, dass Beispiele, idiomatische Ausdrücke usw. in einem Wörterbuch so weit wie möglich kontextfrei sein müssen, um dem Benutzer den Zugang zu schneller und unmittelbar anwendbarer Information zu ermöglichen. Hierbei spielt die Unterscheidung zwischen Hinübersetzen und Herübersetzen (aktivem und passivem Wörterbuch) keine besondere Rolle und wird deswegen hier außer Betracht gelassen. Zwischen fast kasuslosen Sprachen wie z.B. Dänisch und Englisch oder zwischen Kasussprachen wie Deutsch und Griechisch sind ähnliche Beispiele eindeutiger: der 1. Oktober den 1. Oktober

1

-

η 1η Οκτωβρίου την 1η Οκτωβρίου.

Bibliothekarin Monika Jensen, Roskilde, hat mir bei der sprachlichen Revision des Textes sehr geholfen, wofür ich ihr ganz herzlich danke. So z.B. zwischen Dänisch und Neugriechisch: a.) kulturell: Die große Mehrheit der Dänen ist evangelisch, die meisten Griechen sind orthodox; b) geographisch: im bergreichen Griechenland haben Ausdrücke für Hügel und Anhöhen eine andere Referenz als im flachen Dänemark; c) klimatisch: 'kühl' ist eindeutig positiv in einem warmen Klima, 'warm' eindeutig positiv im Norden.

146

Rolf Hesse

1. S p r a c h e n o h n e Infinitiv

In den heutigen Balkansprachen (Griechisch, Bulgarisch, Makedonisch, z.T. auch in den übrigen) gibt es keinen Infinitiv. Da aber der Infinitiv in der westeuropäischen lexikographischen Tradition die zentrale und neutrale Hauptform des Verbs ist, entsteht hier fllr den Lexikographen ein Konflikt. Als ich mit der Erarbeitung des Dänisch-Neugriechischen Wörterbuches begann und auf diese Problematik aufmerksam wurde, habe ich andere Wörterbücher daraufhin untersucht und auch Benutzer solcher Wörterbücher befragt, ohne aber eine befriedigende, geschweige denn eine konsequente Lösung zu finden. Die Benutzer scheinen sich dieses Problems kaum bewusst zu sein, dafür aber stellt es den Lexikographen vor viele schwierige Wahlen. Ein Beispiel: Das griechische Verb είμαι bedeutet eigentlich 'ich bin', wird aber als Lemma mit 'sein' übersetzt. Das ist natürlich nicht 100% korrekt, aber die "korrekte" Übersetzung 'bin' hätte andere Nachteile: Wer würde z.B. das Verb 'sein' unter dem Lemma 'bin' suchen? Trotzdem wird dieses Prinzip des finiten Lemmas in Latein-dänischen und grönländisch-dänischen Wörterbüchern konsequent angewandt, bei den Balkansprachen aber ist das nicht üblich. Beim Lemma funktioniert die Ungenauigkeit ohne Probleme: Deutsch-griechisch: Griechisch-deutsch:

geben = δίνω (wörtlich: 'ich gebe') δίνω = geben.

Wenn wir aber zu den Beispielen, Kollokationen usw. kommen, wird es komplizierter, vorausgesetzt, die Beispiele usw. sollen überhaupt übersetzt werden (was j a nicht in allen Wörterbüchern der Fall ist): Griechisch (ohne Infinitiv)

Deutsch (mit Infinitiv)

LEMMA: (1) δίνω [δίηο]

geben

('ich gebe') BEISPIELE: (2) του δίνω το κλειδί

ich gebe ihm den Schlüssel/ ihm den Schlüssel geben

(3) δίνω το λόγο μου

ich gebe [jmdm] mein Wort/ [jmdm] sein Wort geben

(4) του δίνεις το κλειδί;

gibst du ihm den Schlüssel?

Beim Lemma (1) funktioniert die Widergabe der Präsensform durch den Infinitiv reibungslos; auch beim ersten Beispiel (2) kann der Infinitiv des Lemmas als Übersetzung beibehalten werden. Bei (3) aber muss dazu auch noch das Possessivpronomen der ersten Person (μου = mein) durch sein übersetzt werden, wenn man am "neutralen" Infinitiv festhalten will. Beim Fragesatz (4) ist eine Übersetzung mit dem Infinitiv ausgeschlossen. Soll also ein breites Spektrum an Gebrauchsbeispielen übersetzt werden, ist ein konsequentes Fest-

Grammatische Schwierigkeiten bei der zweisprachigen Lexikographie

147

halten an der Gleichsetzung: Ί . Person im Griechischen = Infinitiv im Deutschen' daher kaum möglich. Das Beispiel mit dem Infinitiv ist natürlich ein Extremfall. Aber jede grammatische Kategorie bietet dem Lexikographen ähnliche Schwierigkeiten, wie wir es zuerst bei den Verben und dann bei den Substantiven zeigen werden.

2. P e r s o n u n d Z a h l

Dänisch

Deutsch

du ser

du siehst

De ser

Sie sehen

I ser

ihr seht

Englisch

Französisch

Griechisch

tu vois

βλέπεις

vous voyez

βλέπετε

you see

Wie man sieht, stimmen die dänischen und deutschen Anredeformen scheinbar überein. Aber selbst das täuscht, denn ein dänisches du, das oft auch gegenüber Unbekannten und sogar Prominenten als informelle oder freundlich gemeinte Anredeform benutzt wird, muss oft durch Sie ins Deutsche übersetzt werden. Es ist aber in der lexikographischen Praxis kaum möglich, solche Feinheiten zu berücksichtigen, und das dänische du wird in Beispielen wohl immer mit du übersetzt. Natürlich können in den betreffenden Artikeln (du, Sie usw.) viele Nuancen erklärt und demonstriert werden, und ein schön eingerahmtes Kästchen kann über den Gebrauch von Anredeformen im Allgemeinen informieren. Das Problem bei den übrigen Artikeln wird dadurch jedoch nicht beseitigt. Die Personalformen, die sich bei der Übersetzung als mehrdeutig erweisen, sollten am besten nur dann benutzt werden, wenn der Kontext den genauen Sinn verdeutlicht, z.B. durch Possessivpronomina, oder wenn andere Zwänge vorliegen (z.B. Zitate, Beispiele aus einem Textcorpus usw). Das Englische bildet mit seinem you einen Extremfall fiir die 2. Person. In einem Englisch-Deutschen Wörterbuch (Sawers 1990) finden sich folgende typische Beispiele für die Übersetzung von you:

Rolf Hesse

148 Englisch

Deutsch

LEMMA: get where did you get that? I don't get you you've got it when you get to know him how are you getting on?

woher hast du das? ich verstehe Sie nicht du hast es kapiert wenn man ihn (erst mal) kennenlernt wie geht's? Insgesamt LEMMA gel:

11 0 1 1 (unübers.y. 2

you = du: ihr: Sie: man:

LEMMA: sagen Was Sie nicht sagen!

you don't say!

wem sagst du das!

you're telling me!

er sagt du zu mir

he calls me 'du'

Insgesamt LEMMA sagen: du: ihr: Sie: man:

2.1. Verbalendungen/Subjektspronomen Griechisch

Deutsch er

βλέπει

sie

sieht

es

In 2.3. haben wir die Gegenüberstellung von zwei Sprachen: in der ersten wird das Subjekt durch die Verbalendung gezeigt, während die andere ein explizites Subjektpronomen verlangt. So kann ein griechischer Satz in der 3. Person ohne explizites Subjekt oder ein genusspezifisches Prädikativ entweder durch 'er', 'sie' oder 'es' übersetzt werden, während das deutsche 'er sieht' entweder nur mehrdeutig im Griechischen wiedergegeben werden kann oder eindeutig durch Hinzufiigung eines Demonstrativpronomens, was wegen der Emphase jedoch nicht ganz korrekt wäre - und auch mehr Platz im Wörterbuch fordert. Zusammenfassend lässt sich das Problem der Personalformen so illustrieren:

Grammatische Schwierigkeiten bei der zweisprachigen

Griechisch

Deutsch

βλέπω βλέπεις βλέπει

sehen/ich sehe du siehst/man sieht er/sie/es sieht

βλέπουμε

wir sehen

βλέπετε

ihr seht/Sie sehen

βλέπουν

sie sehen

149

Lexikographie

Zwischen Griechisch und Deutsch gibt es aus verschiedenen Gründen nur drei einigermaßen "übersetzungsgeeignete" Personen: 'du', 'wir' und 'sie' (PI.), wobei du jedoch nicht ganz unproblematisch ist und bei sie Eindeutigkeit nur durch Kleinschreibung zu erreichen ist. Die übrigen Formen bedürfen alle eines disambiguierenden Kontextes.

3. Aspekt, T e m p u s , M o d u s

Die aufgrund von Unterschieden in der Aspekt-, Tempus^ und Modusstruktur entstehenden Probleme können in sehr vereinfachter Form so angedeutet werden: Dänisch

Deutsch

Französisch

jeg har givet

Griechisch έχω δώσει

ich habe gegeben j'ai donnd έδωσα jeg gav

ich gab je donnais

έδινα

je donnerais

θα έδινα

ich gäbe jeg ville give

ich würde geben

Deutsch und Griechisch: Deutsch ich schrieb ich habe geschrieben

Griechisch έγραφα = 1 wrote/was writing/used to write έγραψα = I wrote/have written έχω γράψει = I have written

Den beiden wichtigsten deutschen Präteritumsformen entsprechen drei griechische: das Imperfekt, der Aorist (Perfektiv, die einfache und häufigste Präteritumsform) und das peri-

150

Rolf Hesse

phrastische Perfekt. Eine befriedigende Übersetzung setzt voraus, dass sowohl Beziehungspunkt als auch Einmaligkeit/Wiederholung aus dem Minikontext hervorgeht. So kann der Unterschied Einmaligkeit/Wiederholung im Deutschen z.B. durch die Kon-junktion zum Ausdruck gebracht werden: Als ich das Lied hörte ... Wenn ich das Lied hörte ...

Ό τ α ν άκουσα ί ο τραγούδι... (Aorist) Ό τ α ν άκουγα το τραγούδι... (Imperfekt).

3.1. Nomina Bei den Substantiven kann es wegen Numerus, Kasus, Genus oder des bestimmten Artikels zu ähnlichen Schwierigkeiten kommen.

4. N u m e r u s

So kommt es im Gegensatz zum deutschen oder dänischen Sprachgebrauch in den Balkansprachen häufig vor, dass nicht zählbare Substantive auch in der Mehrzahl benutzt werden: Deutsch

Englisch sg. pl.

Griechisch pi. sg·

Schnee Holz (Hölzer) Liebe

snow (snows) wood love

χιόνι ξύλο αγάπη

χιόνια ξύλα αγάπες

Im Dänisch-Griechischen Wörterbuch habe ich solche griechischen Wörter mit der Angabe "ogsä pl" (= kann auch in der Mehrzahl benutzt werden") versehen. Im GriechischDänischen zeigen Beispiele den Gebrauch der Pluralformen.

5. K a s u s

Siehe in der Einleitung das Beispiel mit dem 1. Oktober.

Grammatische

Schwierigkeiten

bei der zweisprachigen

Lexikographie

151

6. G e n u s

Hier geht es vor allem um weibliche Berufsbezeichnungen, wobei es wiederum kaum möglich ist, 100% konsequent zu sein. Im Dänischen gibt es z.B. kein männliches und weibliches Genus und nur ausnahmsweise Sonderformen fur Frauen (sangerinde = Sängerin). Für das dänische minister kan man dann auf Deutsch: Minister m., Ministerin f . schreiben, aber sollen wir auch "Stukkateur/Stukkateurin" oder "Henker/Henkerin" an-fuhren? Und wie heißt eine junge Frau, die sich zum Zimmermann ausbildet? Die einzig praktikable, wenngleich inkonsequente Lösung dieses Problems besteht wohl darin, dass die wichtigsten weiblichen Formen (Ärztin usw.) in das Wörterbuch aufgenommen und seltene oder konstruierte Formen weggelassen werden.

7. D e r bestimmte Artikel

Beim bestimmten Artikel geht es vor allem um den generischen Gebrauch, wie z.B. im Französischen: les enfants ne payent pas

=

die Kinder haben freien Eintritt - oder: Kinder haben freien Eintritt.

Der Sinn kann nur durch Kontext oder andere Elemente disambiguiert werden, wie z.B. durch Änderung des Tempus: les enfants n'ont pas ραγέ

=

die Kinder hatten freien Eintritt (oder: dieKinder haben nichts bezahlt!).

8. Lösungen

Die oben angeführten Schwierigkeiten können auf verschiedene Weise beseitigt oder reduziert werden. Fünf miteinander kombinierbare Lösungsmöglichkeiten seien hier erläutert: 8.1.

Die Beispiele werden nur in der einen Sprache angegeben, typisch nur in der Ausgangssprache, wenn das Wörterbuch für "Zielsprachler" geschrieben wird, und um-gekehrt. Aber das setzt voraus, dass die Zielgruppe entweder über ziemlich gute Sprach-kenntnisse verfugt, oder aber die fremde Sprache den Muttersprachlern leicht verständlich ist (z.B. Deutsch für Dänen, aber nicht Deutsch für Griechen). Dazu kommt, dass das Wörterbuch

152

Rolf Hesse

für die andere Sprachgruppe weniger brauchbar wird, was vor allem dann ein großer Nachteil ist, wenn es um Wörterbücher zwischen kleineren Sprachen geht.

8.2.

Man muss auch mit der menschlichen Vernunft rechnen (wenn das Buch für Menschen und nicht für elektronische Verarbeitung bestimmt ist). Z.B. kann man davon ausgehen, dass der Benutzer das deutsche Pronomen er auf eine männliche Person bezieht: er steht am Fenster = he stands (is standing) at the window, (und nicht etwa, dass 'der Stuhl1 oder 'der Hund' am Fenster steht).

Dazu kommt, dass auch gewisse grundlegende Sprachkenntnisse beim Benutzer vorausgesetzt werden dürfen, z.B., dass das englische you im Deutschen drei bis vier Äqui-valente hat, oder dass der deutsche Infinitiv als Lemma im Griechischen oder Bulgarischen durch die 1. Person Singular wiedergegeben wird.

8.3. Die Probleme meiden: Formen, wie das englische you, die sich bei der Übersetzung als mehrdeutig herausstellen, können so weit wie möglich umgangen werden. Das führt aber unweigerlich zu Konflikten: a) mit Frequenzansprüchen: wenn hochfrequente Ausdrücke dadurch ausgeschaltet und immer wieder durch wenig benutzte Formen verdrängt werden; b) mit streng angewandtem corpusbasiertem Material.

8.4. Das Problem der Mehrdeutigkeit durch Doppelübersetzungen lösen: (s)he comes; ein(e) Mitarbeiter(in). Aber auch diese Lösung bringt Konfliktmöglichkeiten mit sich: a) die Übersichtlichkeit kann durch Tausende von Klammem, Schrägstrichen, Kursivierungen usw. leicht verloren gehen. b) 100% Konsequenz ist praktisch ausgeschlossen (siehe oben Punkt 6). c) Doppelübersetzungen fordern viel Platz und einen sehr großen Zeitaufwand. Außerdem wird das Buch dadurch teurer, aber nicht unbedingt besser. 8.5. Die Probleme einfach liegen lassen: Wir müssen uns meiner Meinung nach damit abfinden, dass Mehrdeutigkeiten, darunter auch grammatischer Art, in Wörterbüchern unvermeidlich sind. Daher müssen die Benutzer lernen, dass Übersetzungswörterbücher keine schlüsselfertigen Lösungen liefern, sondern nur Vorschläge und sprachliche "Bau-materialien" anbieten. Selbst noch so große Mühe und Aufmerksamkeit von Seiten des Lexikographen

Grammatische

Schwierigkeiten

bei der zweisprachigen

Lexikographie

153

können nicht verhindern, dass viele Sätze in irgendeinem Zusammenhang auch anders als in der vom Lexikographen (oder seiner Quelle) beabsichtigten Bedeutung aufgefasst werden können. Ein Vegetarier, der den Satz "das ist weder Fisch noch Fleisch" liest, wird vielleicht sofort an Gemüsebuletten denken - aber deshalb muss nicht unbedingt die Angabe "in übertragenem Sinne" beigefügt werden. Dasselbe gilt auch für Übersetzungsprobleme grammatischer Art. Bei der Erarbeitung eines Wörterbuches sollte man bestrebt sein, die wichtigsten Problemkomplexe dieser Art so früh wie möglich in den Griff zu bekommen, damit sowohl die Wahl der Beispiele als auch die Übersetzung mit einem hohen Grad an Konsequenz erfolgen kann.

Literatur

Hesse, Rolf (1995): Dansk-nygrcesk ordbog. Kebenhavn: Munksgaard. - (1998) Nygrcesk-dansk ordbog: ordbog over moderne grcesk / Rolf Hesse; under medvirken af Kerasia Marousou og Giannis Vaenas. Athina: Patakis Publications. Sawers, Robin (1990): Harrap's Concise Dictionary: English-German and German-English. Edited by Robin Sawers. London: Harrap.

Lars Holm 35 Questions and Answers about Editing Jesper Swedberg's Swensk Ordabok (c. 1725)

1. Who was Jesper Swedberg - and what is Swensk Ordabok! Jesper Swedberg was one of the most dynamic and versatile figures of the Swedish Caroline era. He took active part in politics on a national level. He was a theologian: professor, bishop (his diocese included also the Swedes in London and North America), hymn writer and hymn-book editor (counteracted by hostile fellow clergymen). But for most of his life he also was a philologist, as much prolific as polemic. He published books on orthography and grammar - and completed his lifework with a comprehensive Swedish dictionary (Swensk Ordabok). This unpublished work is probably - and ironically enough - his most valuable contribution to the description of Swedish.

2. Should SO really be published - at last? Yes! I try to explain that in my doctoral dissertation Jesper Swedbergs Swensk Ordabok bakgrund och tillkomsthistoria (Jesper Swedberg's Swedish Dictionary - background and origin), Uppsala 1986. In my opinion SO is "the best Swedish dictionary of its period and perhaps the only good one before Sahlstedt's 1773". And now Svenska Vitterhetssamfundet (The Swedish Society for belles-lettres, whose principal aim is to republish old Swedish texts) also seems to be interested in publishing dictionaries such as Stiernhielm's and Spegel's.

3. Would not a facsimile be the best form? Arguments for: (1) especially the Uppsala manuscript is rather legible; (2) printing could start almost immediately. Arguments against: (1) the critical apparatus would be hard or impossible to arrange; (2) the practice of SVS does not include facsimiles. (An electronic version, however, would be a possible complement.)

4. Must the editorial work emanate directly from the manuscripts? Of course it ought to do so, but it is not feasible in practice. For the time being I am working with Xerox copies. Naturally a total check against the originals must take place before publishing.

156

Lars Holm

5. Which one of the five manuscrpits should be selected as the basic manuscript? (I refer to the chapter "Swedbergs Swensk Ordabok - handskrifterna" (... the manuscripts) in my dissertation. 1 name the manuscripts after the cities to whose scholarly libraries they belong.) The answer is, paradoxically enough, Uppsala, written by Swedberg's son Jesper Swedenborg and provided with additions by at least three other hands of which Jesper Senior's is the most assiduous one. But the other four manuscripts should be allowed to make their contributions: Skara 1 is of greatest value as an original and prototype; Stockholm is the latest complete version - but unfortunately and irrevocably half destroyed; Skara 2 was written outside the Swedberg family - but by a hand that has left its traces in Uppsala; Linköping is a fragmentary original.

6. What working method is suitable? First I transcribe all Uppsala, letter by letter, typing it into an iMac in Word 98. Then I manually enter all deviations between Uppsala and the other manuscripts on separate copies for each manuscript. Finally I compile a list of certain deviations (see below) in footnotes page by page. Now (April 2000) the letters A - C are ready, F and Τ are being transcribed.

7. How should the basic manuscript be reproduced? Diplomatically, of course.

8. Is diplomatic reproduction a practical principle or an ideal to aim at? Mostly the latter, I am afraid.

9. Does the diplomatic principle exclude every kind of normalization? No, reasonably it should not, I hope. See below.

10. Can the logical law of identity "A is equal to A " always be maintained? Maybe in the best of all possible worlds - but that world knows no manuscript that boasts of two or more ways of writing e. g. k or s - also in black-letter or roman type. Which allographs should one choose (from the crammed keybord of the computer), and what information would these choices convey to today's and tomorrow's readers of SO?

35 Questions and Answers About Editing Jesper Swedberg's Swensk Ordabok

157

11. Should the layout of the manuscript be reflected in the layout of the edition? Yes. My ambition is to maintain the division in pages, columns, lines, as well as all indentations (with some difficult borderline cases). There are two reasons: (1) the method gives a feeling of "authenticity"; (2) an arbitrary division would palpably obstruct the editorial work - and still more the control of it.

12. Can fidelity to the original layout be total? No, not reasonably. Examples: (1) added words within the column frames get their proper placing; words outside the frames, with or without "scrolls", are placed at sensible alphabetical positions with a footnote stating their real placing - to act in any other way would imply pseudofacsimile. (2) Lack of space in the manuscript can result in the type: "Col:2v / v8"; I normalize to "Col:2:8."

13. How to act when the size of type chosen for the edition runs counter to the desired fidelity to lines and pages? Preliminarily I have chosen a smaller type - right down to 9 points (instead of 14). This method might be misleading, and I am rather doubtful of it; I will submit the question to professional printers.

14. Naturally all signs will be reproduced as exactly as possible - but how to reproduce vacant space? A certain degree of normalization is probably inevitable. It might be easily acceptable that distances between columns and between lines are fixed in print even if they change in the manuscript (so the harmony in height between the columns on a page is often given up). The compounds present a greater problem; the risk is evident that the editor's personal practice guides him to either put a space or not between the parts of the word or not. References to the Bible are very frequent in SO. I have decided to normalize them strictly: e.g. "Act: 1:2", without any space. The same rule applies to all other kinds of references to literature.

15. Should general typographic variations be represented in the edition? Yes and no. Underlining is kept. Boldface and italics are scarcely found in the manuscripts; in print they are given other functions: additions by Jesper Swedberg and expanded abbreviations.

158

Lars Holm

16. Should the (hardly consistent) distribution between black-letter (Swedish) and roman (Latin) in the manuscript be kept in the edition? Doubtful! Black-letter would hardly lure younger linguists and philologists into studying SO. On the other hand: had SO been printed in its own time, the Swedish would no doubt have been in black-letter, the Latin in roman (cf. e.g. Spegel's Glossarium, 1712). The valid practice of SVS is: black-letter is changed into upright roman type, roman type into italics. I leave the question open for the present.

17. How to deal with the (rather few) Greek words? They are reproduced literally in Σψμβολ (ought to be Συμβολ!). No explanations are required since they are always provided by the context. The words could, however, be transcribed into roman in footnotes.

18. How should unmistakable errors in the manuscript - marked and unmarked - be dealt with in the edition? Diplomatically and silently. No emendations or conjectures are made to Swedish and Latin words, biblical quotations, margin placings, etc. Errors in Uppsala are often revealed in Skara 1 and then they are evident from the footnotes. If, for example, a biblical quotation is incorrect in all the manuscripts, information will be given in a footnote. Cancelled words and letters - it is mostly impossible to find out who cancelled them - are represented: idem. Overwritten letters (by Jesper Senior?) that have totally obscured the underlying ones: bold; if the result is illegible: (xxx).

19. Is the difference between upper and lower case letters a question of shape or size - or both? A crucial problem! The answer is important for the form of the edition.

20. How did Jesper Jr (the hand of Uppsala) cope with upper and lower case letters? Maybe he had some gaps in his writing education - which were hardly filled during his time as an elementary-school teacher for North American Indian children. He is simply a little anarchic in his choices between small and capital letters - and he was certainly not unique for his time. F and S are particularly hard to interpret. I have until now - though with hesitation - followed the principle: "on the line = capital, under the line = small letter". For certain allographs of c, m, η, ο, ν etc. there is a sliding scale from upper to lower case letters. Should the surrounding letters or a micrometer be decisive? I interpret the medium-sized letters as small. A radical solution of the whole problem would be to normalize to lower case generally after Jesper Sr's handwriting in Skara 1.

35 Questions

and Answers About Editing Jesper Swedberg's

Swensk

Ordabok

159

21. Are there letters that cause problems of interpretation at a graphemic level? Yes, especially s - z, and I - J. I and J seem to share character in SO Uppsala. One of them must be chosen, even if it might seem unfamiliar to write "Ieremia" or "Jcke" (not). (The system for u, v, and w, on the other hand, was stabilized in the 1720s, though different from the one of our time.) The question of full stops and commas is also relevant here. It is often very hard to decide what was intended by the writer, so there is a great risk of unconscious normalization.

22. Which (clusters of) characters should be interpreted as ligatures - and how should they be represented in the edition? There is a sliding scale for ck, ff, fft etc. from a ligature to two (three) separate characters; I therefore always type ck, ff, fft etc. ß, Professional > Socio-cultural > Intimate The linguistic items that were investigated are all well-known indicators of informality, and it was assumed from the start that they would be prominent in the CANCODE texts. These results are encouraging not only in that they show a certain correlation in the expected direction but, and perhaps more so, because there are indications of unexpected features at the most informal end of the cline. The use of private verbs, for example, decreases considerably in the most informal text types, contrary to expectations. (In another investigation of four individual texts, one from each of the Relationship categories, it was found that the occurrence of hedges followed the same pattern.) It is also interesting to see that we do not find many occurrences of a word like okay at the most informal end of the cline. It hardly need be repeated that the investigation described here cannot be considered conclusive evidence in any direction, but the results of the pilot study suggest that a more comprehensive investigation of the proposed text types will provide interesting insights into linguistic correlates of spontaneous speech along the formality-informality gradient.

5

Rayson et al (1997:146), for example.

Categorizing

Chaos: Text Types in CANCODE

171

Acknowledgement

I am g r a t e f u l to the Directors of C A N C O D E , P r o f e s s o r s R o n a l d Carter and

Michael

M c C a r t h y , and to R e b e c c a H u g h e s and A n n e Finell, f o r n u m e r o u s c o m m e n t s and a d v i c e on earlier drafts.

References

Biber, D. 1988. Variation in speech and writing. Cambridge: Cambridge University Press. Carter, R.A. & M.J. McCarthy. 1997. Exploring spoken English. Cambridge: Cambridge University Press. McCarthy, M.J. 1998. Spoken language and applied linguistics. Cambridge: Cambridge University Press. Rayson, P., G. Leech, & M. Hodges. 1997. Social differentiation in the use of English vocabulary: Some analyses of the conversational component of the British National Corpus. International Journal of Corpus Linguistics, 2(1) 133-152. Sigley, R. 1997. Text categories and where you can stick them: A crude formality index. International Journal of Corpus Linguistics, 2(2) 199-237. Ventola, E. 1987. The structure of social interaction: A systemic approach to the semiotics of service encounters. London: Francis Pinter.

Olga Karpova /Svetlana Manik Public Political Vocabulary: Model of a Dictionary

1. Introduction

The role of social and public political factors in modern societies has intensified. This process is closely connected with the development of social and political vocabulary. The latter has resulted in both a considerable extension of such vocabulary, and in qualitative changes in the lexicon. Therefore there is no reason to present social and public political language in artificial isolation. On the one hand it is a product of its social and cultural environment and, on the other, it is an element of no mean importance in the development of language in general. To sum up, events of social political life constantly find reflection in any national language and in changes in the lexicon which are being registered in dictionaries. An examination of the public political vocabulary should be carried out, taking into consideration social development. One can describe such lexico-semantic subsystems within a linguistic framework, but referring to extra-linguistic realities is unavoidable when analyzing semantic change. Historical semantics is often limited to the description of the notion as such. We would suggest a shift of focus to the history of political, socio-economic and psychological ideas and notions. Lexicographers face a lot of problems when investigating public political vocabulary even from the synchronous point of view. Elements of such vocabulary are, as a rule, polysemous in any socially and hence ideologically heterogeneous linguistic group. The question of how subjectively words are interpreted by different social groups arises here. It is especially vital for a general-purpose dictionary, which tends to be seen as normative - as a lexicographical authority. The role of the dictionary in society has already been emphasized by many Russian and foreign scholars: A. Golovanevsky, V. Dubitchinsky, O. Ermakova, O. Karpova, H. Bejoint, J. Haynes, J. Lye, T. McArthur, E. Nida, K. Varantola. A. Wierzbicka, and others. It is generally admitted that dictionaries are social entities, and therefore lexicography cannot be separated from ideology.

2. Ideology and Public Political Vocabulary (PPV)

Ideology is significant for understanding and using PPV in language. But what is ideology? The term has undergone changes from Destutt de Tracy's definition as "science of ideas" to the modernist sense. It refers to any "system of ideas that aspires both to explain the world and to change it" (Cranston 1986). Such an ideology is "regarded as justifying

174

Olga Karpova/Svetlana

Manik

actions", is "held implicitly or adopted as a whole", and is "maintained regardless of the course of events" (OED). This conception of ideology dates from the Enlightenment. The evolution of the modernist sense has been complex (as demonstrated in the writings of Hegel, Marx, Weber, Camus, Popper and Sartre) and usually refers to totalitarian secular systems such as Fascism and Communism. During the 20 th century, however, a universalist sense has emerged. It covers any sociopolitical system, modern or ancient, defunct or surviving, European or other. Islam and Hinduism may therefore have ideological aspects insofar as Muslims or Hindus engage in sociopolitical theory-building and in kinds of mental and social control. Tom McArthur believes that "their ideologies may derive as freely from and interact as freely with their worldviews, hierarchies and canons as any in Europe and its diasporas. In this sense, ideology need not be consciously contrived, totalitarian, secular, or even coherent" (McArthur 1995). Ideology is considered "to be rather a matter of how the dominant institutions in society work through values, conceptions of the world, and symbol systems, in order to legitimize the current order" (Lye 1997). This legitimization is managed through the widespread teaching (the social adoption) of ideas about the way things are, how the world "really" works and should work. These ideas direct people's thinking in such a way that they accept the current way of doing things, the current sense of what is 'natural', and the current understanding of their roles in society. We do not set ourselves to examine the concept of ideology. But the main task of the work - examination of social and political vocabulary and its ideological coloring requires particular knowledge and attitude. Our preference in interpreting ideology is for the broader terms belief system and idea system, proposed by Tom McArthur (McArthur 1995). Although these terms overlap (both relating to worldviews, hierarchies, canons, laws, propaganda, publications, media, and the like), they usefully cover conditions that develop organically in whole communities rather than as the intellectual exercises of a selfselected elite (who may well develop specific ideologies within them). Many people wish to gain a more profound understanding of belief and idea systems. That is why the latter are the focus of lexicographers' attention and find reflection in works of reference. Achieving clarity of word definition may well be a prime function of dictionaries. If so, such an aim and undertaking is ideological in the sense that it is intended to influence how people perceive certain things (and perhaps also how they then behave in relation to them). Traditional belief and idea systems are likely to influence reference materials only in implicit and indirect ways, but not explicitly (for instance to justify a particular course of action or way of life). Thus, registration and ideological coloring of social and public political vocabulary in modern dictionaries should be analyzed keeping in mind compilers' traditional implicit approach.

Public Political Vocabulary: Model of a Dictionary

175

3. Evaluation of PPV in modern Russian and English dictionaries

This paper is based on the following statement by H. Bejoint: "the ideology of a society shows through in some aspects of macrostructure and of microstructure, particularly the choice of words, the wording of the definitions, and the choice of examples, so that the dictionary can be used, in reverse, as a means of access to the values of a society: it can give a reasonably accurate idea of what is valued, or on the contrary stigmatized, by the more prestigious social groups." (Bejoint 1994: 124)

We absolutely agree with H. Bejoint that in the choice between inclusion and exclusion, as well as in the type of definition, lexicographers may feel the desire to be truly descriptive, which prompts them to record all the reasonably frequent words and to say exactly how they are used. On the other hand, they are well aware that the simple fact of including and defining "nasty" or generally objectionable words might result in a certain loss of respectability. Let us analyze the ideologically colored dictionary, the only dictionary of this type ever published in Russia: A.L.Golovanevsky "Ideologically Colored Dictionary of the Russian Language of the XIX - beginning of XX century" (1995) (rojioeaHeBCKHfi A.JI. (1995) H,ne0Ji0rHHecKH-0ueH0HHbifi cjiOBapt pyccKoro «3biKa XIX - Hanana XX ΒΘΚΟΒ). It was compiled on the basis of examination of the public political vocabulary registered in dictionaries in the aforementioned period and thus systematizes material already known. Its author and compiler sets certain principles for lemma selection: - the sociopolitical meaning of each item should result from the development of that word's semantic features; - each item should be an object of evaluation by a particular socio-ideological grouping; - each item should be registered by at least two dictionaries of different types; - each item should be listed in a Soviet dictionary and its usage should be confirmed in contemporary scientific and learned literature. These criteria show that A. Golovanevsky took into account historical development, the social nature of the word and actual usage. He designed the dictionary to display semantic change as well as the reasons for change. This dictionary lacks cross reference. Words and their derivatives are supplied with separate definitions. Thus, "proletariat", "proletarian", "proletarianization" and "proletarianize"; "progressive", "progress", "progressist(s)" and "to progress" are separate entries because they gain importance as phenomena of social and public political life. A. Golovanevsky always include evaluations in the definitions, borrowed from the dictionaries quoted. For example, he registers a French word without transliteration: AgentProvokateur. To our mind such treatment underlines the negative evaluation expressed by people who correlated their unfavourable emotions towards strangers (cf. 'aliens') with foreign (i.e. strange) language. In another of the dictionaries cited, the similar entry is spelled «azeHm-npoeoxamop». It probably reflects historical development of events when the French word was borrowed and assimilated. Although the entries of the dictionary are not labeled one can find etymological information in them. Historical development of word meaning is traced by comparing

Olga Karpova/Svetlana Manik

176

definitions in the dictionaries o f different periods, and its ideologically colored character by comparing definitions in the dictionaries o f the same age. For example, «pHjiHcrep 1. ripe3pHTejibHoe

H33BäHHe

HHTepecaMH, n p o c r o r o

nejioeeKa

MajioayuiHoro,

3aHjrroro

MemaHHHa; cpeAH cryaeHTOB 3THM hmchcm

ΟΛΗΗΜΗ

ÖYAHHHHBIMH

Ha3biBaioT BcsKoro,

He

npHHaanewauiero κ cryneHHecKofi c p e a e . 2.

EypjKyii,

Boo6we

nejioeeK

oßbieaTe^b,

CKJUUIKH, C MEMAHCKHMH

«MemaHCKofi»

6ioprep.

CaMoaoBOJibHbift,

NPHBUHKAMH

orpaHHMeHHbiti

Η CKJIMOM

He/toeex

c

yMa.

MemaHCKH

Kpyro3opoM Η xaH)KecKHM noee/ieHMeM; o6biBarejib. [Philistine I. Contemptuous term for a faint-hearted man, busy only with everyday interests, interests o f a c o m m o n petty bourgeoisie; among students anyone who does not belong to the student sphere is called philistine. II. Bourgeois, a man o f a " b o u r g e o i s " feature, with bourgeois habits and mentality. C o m m o n l y : inhabitant, burgher. Self-satisfied, narrow-minded person with bourgeois range o f interests and sanctimonious behavior; inhabitant.]

The first meaning is likely to have formed before the Revolution, the second one in Soviet times when 'petit bourgeoisie' was equated with 'enemy' ('bourgeois'). The student term (an outsider, stranger, alien) has gradually gained negative ideological coloring. Such treatment o f the entry is considered to be useful for the dictionary user; it fully shows the historical development o f the word. Thus A. Golovanevsky's dictionary is mainly a historical dictionary. It reveals the first attempt to reflect (through the impartial

lexicographer)

on the most vital

political

phenomena o f Russian and world history. It can be regarded as an experimental step towards the public political dictionary. Another approach is found in the "Explanatory Dictionary o f the Russian Language o f the End o f the X X century. Language Changes" edited by G. Sklarevsky (St. Petersburg, 1 9 9 8 ) (TojiKOBbifi cjioeapb pyccKoro jrauica KOHua X X BeKa. ifatiKOBbie Η3ΜβΗβΗΗΛ ( 1 9 9 8 ) . ΠΟΛ pea. Γ.Η. CKJiapeBCKofi). Its aim is to display changes in the Russian language from 1 9 8 5 to 1997. As it attempts to describe the dynamics o f the lexical system, the dictionary includes various categories o f vocabulary: -

basic vocabulary which has undergone semantic, stylistic, connotative and other changes during the examined period (for example, duccudenm

(nonconformist)

is registered in

the following way: 1. Pen. BepyiomHii, OTCTynaiouiHii οτ otJjHiwajibHoro pejinrH03H0ro yneHiw Η Jin He nojiHHHfliomHHCH qepKOBHOH ÄHciwnjiHHe. 2. Β coeeTCK. BpeMJi: TOT, KTO He pa3flejwji KOMMyHHCTHHecKoft HfleonorHH, npoTHBocroHJi

cymecTByiomeMy

peJKHMy Η noflBepranca 3a 3το npecjieflOBaHHSM Η penpeccmiM. CM. AHTHCoeeTHHK. OrmeneHeu. [7. Religious. ecclesiastical ideology,

discipline.

were in opposition

and repressions. -

Believer

2. In Soviet

deviating from times:

to the present

Cf. Antisoviet.

an official

religion

those who did not share

regime

and, therefore,

exposed

or

disobeying

the

communist

to

persecution

Renegade.]

new vocabulary which includes both absolutely new words registered for the first time in the

dictionary:

aHTHnpaeo

(antiright),

couHajiHCTHHecKHfi (antisocialistic), lady),

6oM3caTHHK

(house

for

the

aHTWipaBOBOH

6e3HajioroBMfi (taxfree), homeless),

(antirightful), 6H3HecMeHKa

ÖKWKCTHHKH

(budgetary

ΒΗΤΗ-

(businessofficials),

Public Political

Vocabulary: Model of a

177

Dictionary

TocayMa (State Duma), peHHicapHamM (reincarnation) Η Τ.Λ) and relatively new words already described in the dictionaries of new words (ajiKaui (drunk), aaneprpayHfl (underground), aHTHapMeficKHft (anti-army), 6ami4>opMnpoBaHne (band formation), ßecnpeae/i (unthinkable, inconceivable or unheard-of situation), rpaHT (grant), ÄeflOBiuHHa (bullying harassment), HMHÄ>K (image), HHBaniOTa (foreign currency), CHJIOBOH (law enforcement)\ - words with the label 'Obsolete' or with the comment "In the old times", "In prerevolutionary Russia": axura (excise duty), 6jiaroTB0pHTejibH0CTb (charity), MmiocTbiH* (alms), ay^HTop (auditor), Ka3Ha (treasury, exchequer), npecTOJi (throne), npmoT (asylum), H H H O B H H K (official); - words associated with the categories of the bourgeois society and rooted in today's Russian reality: HHJMUHH (inflation), Maijma (mafia), Koppynmw (corruption), 6e3pa6oraua (unemployment), 6H3Hec (business), Kamroji (capital), ßarncHp (banker); also words referring to social and cultural structures of foreign states: Nop (mayor), napjiaMeHT (parliament), MyHHUHnajiHTeT (municipality), οφκς (office), φβρΜβρ (farmer), ΓΗΜΗΕ3ΗΛ (grammar school), JiHueft (lyceum), ryöepHaTop (governor), ryeepHep (tutor); - words and word combinations which are not in active use but which reflect categories of Soviet times widely used in the official language: BaxTa (watch, allotted period of work), aeaHrapa (advance-guard, vanguard)). The dictionary also includes names of people and places found in contemporary social debate, often names that have become symbolic: XpncToc (Christ'), C H r (CIS), CCCP (USSR), Β A3 (VAS), JlyÖHHKa (Lubjanka), KpeMJib (Kremlin). Definitions are meant to reflect the demands of the dictionary user, as seen below: Cneujiarepb. Β coeeTCK. epeMo. CoKp. CneunajibHbiH jiarepb (jiarepb una oco6o onacHbix

npeCTynHHKOB, Κ KOTOpbIM OTHOCHJIHCb Η nOJIHTHHeCKHe 3aKJHOHeHHbie). [Specicamp. In the Soviet times. Abbrev. of 'Special camp' (a camp for especially dangerous criminals, including political prisoners).]

Moreover, definitions may contain information of an encyclopedic nature: I I p a B o . . . 2 . Y3AK0HEHHA)I Η LOPNANQECKH Ο Φ Ο Ρ Μ Π Β Η Η ^ ΒΟ3ΜΟ>ΚΗΟ?Π> N0JIB30BATBCA, BJIAAETB

ocymecTBJiaTb KaKOii-JI. Β NA aeaTejibHOCTH, NPOH3BO^HTB OTAe^bHbix rpaamaH, ceMefl, opratuuauHH, npeanpHarafl Η Τ.Λ ).

HTO-JT.

Η Τ.Π.

(KacaeTCH

[Right ... 2. Legitimized and legalized possibility to use, possess sth., carry out some kind of activity, produce sth., etc. (concerns individual citizens, families, organizations, enterprises, etc.).]

The above examples show the explanatory character of this dictionary, as it is inclined to describe phenomena of public political life, to familiarize its reader with lexical information, i.e. with possible variants of the entry. That is why definitions are sometimes supplied with a synonym: flpHSbiJib...; ΛΟΧΟΛ. floanojibee... ; aaaerpayHa. KoMMepcaHT...; npeanpnHHMaTeiib

or with an antonym with the label opposite:

[ P r o f i t . . . ; income.] [Cellar ...; underground.] [ M e r c h a n t . . . ; businessman.]

Olga Karpova/Svetlana Manik

178 EeAHbie ...; προτΗβοπ. 6oran>ie AeMOKpaTHti...; προτΗβοπ. TOTajiHTapmM

[Poor ...; opposite rich] [Democracy ...; opposite totalitarianism]

The entries come with different labels: 1. Labels of functional and social qualification demonstrate that the word belongs to a particular social sphere or group. For example, Ocpuif. (oHUHanbHoe) [Offic. (official)] marks the vocabulary of the official sphere, of documentation, etc. which is inappropriate in everyday communication (3aHflT0CTb employment, MMOHMymne needy or indigent, MHoroMaHflaTHbifi multi-mandated, HajioronjiaTejitmHK taxpayer, cpeflHecTaTHCTHHecKHfi statistical average)·, )Kapz. (οκαρεοπ) [•Jarg. (jargon)] characterizes socially or professionally defined vocabulary: 6a6KH dough, brass, 6axcbi bucks, TaiuHTbca to be wild about sth., KcepHTb to xerox, TycoBaTbCH to gather together, 3arpamca travelling the seas, φπρΜ3Η foreigner), etc. 2. Stylistic labels that place words on the scale High / Low. BbicoK. (Bbicoicoe) [High] is used for words of texts for ceremonial occasions. In the material used it usually marks the vocabulary of Soviet times (CBCTOH luminary (of a person), 3apHHua summer lightning). CHUOK. (cHHweHHoe) [Low] is used with words considered inferior, typically victims of harsh evaluation. Such words are normally registered with the label Colloquial: c6auaTb to go stroll; nip down to, nopHyxa pornography, ö a n a e T b to get high, c e a j i H T b " y f i T H " to take off. 3. Evaluative labels which mark steady emotional coloring or evaluation of a word. HpoH. (HpoHHHecKH) [Iron. (Ironically)]: noJiHTiycoBKa political gathering, HapoflHbifi H30paHHHK people's elect, ceKcyajibHO 03a60HeHHbiH sexually frustrated. Heodoöp. (HeoaoöpHTejibHo) [Disap. (Disapproving)] is a sign of negative evaluation: ΜΗΤΗΗΓΟΒΙΙΧΗΗ3 mass meeting, (JjyHKUHOHep office clerk, COBKOBMÜ Soviet. Ilpe3pHT. (npe3pHTejibHo) [Scorn. (Scornfully)]: HHTejiJinreHTHK intellectual, egghead. LLlyTJi. (uiyTJiHBo) [Hum. (Humorous)]: 6apa6auiKa poltergeist, BKJNOHHTB neMaTHbiii CTBHOK switch on a print machine (to print money). The authors of the dictionary implement icons to improve user-friendliness. There are different icons for the following functions and pieces of information: New word (meaning); first dictionary listing Relatively new word (meaning); Dictionary listing during the examined period Revitalization of a word "Π" Revitalization of a certain word meaning W Devitalization of a word Separator of statements and quotations Separator of a series of statements reflecting semantic shifts Peculiarities of usage Set expressions A Phraseology An explanation of all icons is given at the bottom of each page, which facilitates dictionary use. Below one of the articles from the dictionary is quoted:

Public Political Vocabulary: Model of a Dictionary

179

MacopyÖKa, η, ΜΗ. pod. 6οκ, Jtc. fJepen. IJyÖA. bo Η Ha, 6ohh«; τ ε ρ ρ κ τ ο ρ ι « , H a κ ο τ ο ρ ο Β O H a npoHcxoflHT. BOÜHO c ΗρακοΜ yyce Hananacb. Odnaxo naiuu jtcmpeöbi U3 eoenuo-npoMbiuineHHOZo xoMWiexca, emxHyeutue β ceoe epeMH cmpany β αφεαπαψιο MHCopyöxy, mamaiomcx ucnonb3oeamb anmueoeHHue HacmpoÜKU dm npoeedenux nonumum HeeMetuamenbcmea β ÖJiaeopodHoe deno oßysdaHU* aepeccuu. CMeHa, 22.01.91. /Jecxmxu xeadpamubix Kwiojuempoe HenpoxoduMbix yufenuü eoxpye cmojiuifbi föeo-OcemuHCKOÜ aeniOHOMHOÜ oönacmu npeepamwiucb β zuaanmcxym Mftcopyöxy, ade npomueo6opcmeyiou{ue cmopoHbi npuMeHRtom aemoMamu, 6poHempaHcnopmepu u dcotce CHHmbie c öoeebix eepmonemoe Heynpa&nxeMbie paKembi. Co6., 1991,5.

Ό " 1. EeccMbicjieHHax

2. tezo,

κακαΛ. HacH/ibCTBeHHbie üeftcTBHx no othouichhio κ (penpeccHH, yÖHficTBa η Τ.Π.).

«MxcopyÖKa» momanumapHoeo peMCUMa He exonocmyio npoeepnyjtacb. CMeHa, 16.01.91. ß penpeccuenyio MacopyÖKy nonwiu ne mojibKO me, Kmo oKa3biean npomueodeücmeue u Kmo nomeHtfuaiibHO MOZ 0KO3AMBCR UHOKOMUCAKIHUM, HO U coeepuieuno nocnyuiHbie, aMopcpHbie. Hh)K, 1991,2. 3. BecnejiOBeMHbie ycjiOBHX «H3HH, )KH3HeHHbic 06cT0«TejibCTBa. Hauiu Hbtneuinue

«tacmnuKu»,

daü UM βΟΛΚ, nponycmxm nac iepe3 MacopyÖKy yoxacoe 3noxu nepeonaianbHoao HaKoiwenux u «omMbieanuH yxpadeHHoeo». Βεκ XX, 1989, 12. .p. |Mincing-machine, s, pi. Figurative. Public. ^ Useless war, slaughter; territory where it takes place. The war with Iraq has begun. But our

hawks

from the military-industrial complex who have pulled the country into Afghan mincing-machine try to use anti-war mood to hold the non-interference policy in the decent matter of aggression restraining. Smena, 22.01.91. Tens of square kilometers of impassable ravines around the capital of the South-Osetinian autonomous region have turned into a huge mincing-machine, where the fighting sides use tommy-guns, armored troop carriers and even non-controlled ballistic missiles taken from battle helicopters. Sob., 1991, 5. Forcible actions with regard to people (repressions, killings, etc). "The mincing-machine"

of the

totalitarian regime has not run idly. Smena, 16.01.91. Those who opposed and who potentially could be differently minded as well as absolutely obedien, found themselves in the repression mincing-machine. NiZ, 1991,2. Inhuman living conditions, living circumstances. Today's "private traders", if given vent, will pass us through the horror mincing-machine of the primary accumulation and "laundering of the stolen" epoch. Vek XX, 1989, 12.] The dictionary entry covers all aspects of word description. The elaborate icon and labeling system seems very useful. The "Explanatory Dictionary" supplements the existing similar dictionaries of Russian as it reflects a more mobile part of the vocabulary. Furthermore, it can be regarded as a model of lexicographic description of language dynamics. In our opinion public political dictionaries combine the characteristics of political and explanatory dictionaries. That is why it is worth examining the role of ideological factors found in an explanatory dictionary such as the "Oxford Dictionary of New Words". New words (neologisms) often belong to the public political sphere: assisted suicide; basehead; bodyism; canteen culture; church planting; clear blue water; codependency; decommunization; ethnic cleansing; feminazi; glass ceiling; iron John; lipstick lesbian; nanny state; new world order; no-fly; Ossi; one nation; sleaze factor; spin doctor; velvet divorce; Wessie, etc. The words supplied with the icons "Politics" and "People & Society" form the basis of the entries under discussion. The former are words relating to political events (Politics) and the latter are used by particular social groups and by people with definite characteristics (race, age, religion, etc.)

180

Olga Karpova/Svetlana Manik

The dictionary compilers state that the word combination "Political Correctness" is widely used. The abbreviation PC (politically correct) can possess both disparaging and approving coloring: for instance, in the terms of social life the notion "New Man" has been doubted (and often compared with the absolutely politically incorrect word combination "New Lad"). The language of politics consists of the private argot which politicians use when talking to other insiders, of the vocabulary about politics, and finally of public language used by political figures. That is why we find notions as "Blairism", "Clintonite", "Clintonomics", "Majorism", as the policies and personality of a politician are fundamental in the public life of a country, and in forming public opinion. The Oxford dictionary also includes words which reflect, for instance, the state of affairs in the ex-communist camp, the gradual reformation of these states, and minor ethnic groups' struggle for justice or recognition of their independence: Decommunisation; Ethnic cleansing; First Nations; One Nation; New World Order; codependency; dependency culture; multicult; single currency; subsidiary; West Lothian Question. The "Oxford Dictionary of New Words" is characterized by its neutral coloring of the entries. For instance, Contra noun ... A member of any of the guerrilla forces which opposed the Sandinista government in Nicaragua between 1979 and 1990; often written in the plural contras, these forces considered collectively.

The compilers define each phenomenon without taking sides. Each entry is a mobile system of meanings: it may register peculiarities of usage and inform about language sphere (where the word is applicable and in what context), which is a great advantage of this dictionary: Refusenik noun... Colloquially, any person who has been refused official permission to do something or who has refused to follow instructions, especially as a form of protest.

Handbagging noun... In media slang, a forthright verbal attack or volley of strident criticism, usually delivered by a female politician (especially Margaret Thatcher, British Prime Minister 1979-90).

In short, social factors play an important role in the ODNW. But it does not attach any ideologically colored labels to words. The emphasis is on the user's background knowledge. Relations within social groups and subcultures sharing certain views on life or political theories are sure to create ideological coloring of a number of words. That is why the main part of the 'colored' social and public political vocabulary is related to the spheres of "Politics" and "People & Society". Evaluation of words is found either in semantics, being a component of the notion itself, or in the description of its usage, its pragmatics. We will now get back to a discussion of political dictionaries. As it has been mentioned, public political dictionaries should borrow some peculiarities from political dictionaries. Let us analyze one of the political dictionaries, the "American Political Dictionary" by J. Piano and M. Greenberg (1993). The dictionary articles unite definitions and detailed historical descriptions in a condensed format. Each political term is examined from the point of view of the events or

Public Political Vocabulary: Model of a Dictionary

181

situation in the country which caused the active use of the word. One may call it a hidden etymological label. But the emphasis is on public life and politics, not on linguistic factors. We should note that one of the ways in which lexicographer can make offensive words appear less offensive is to define a word by a synonym, which is usually a more respectable word. The authors of this political dictionary implement this method. Thus, there are no disparaging statements or word combinations which could reflect negative attitudes to a phenomenon. For example, Affirmative Action A plan or program to remedy the effects of past discrimination in employment, education, or other activity and to prevent its recurrence. Various federal and state statutes require affirmative action to redress past discrimination against or promote the employment of racial or religious minorities, women, the handicapped, disabled veterans of the Vietnam era, and to some extent, the aged. Affirmative action usually involves a work-force utilization analysis, the establishment of goals and timetables to increase use of underrepresented classes of persons, explanation of methods to be used to eliminate discrimination, and establishment of administrative responsibility to implement the program. Good faith and a positive effort to remedy past discrimination must also be shown. As it is seen the definition is rather extensive. It includes a lot of extralinguistic material. Such description is essential to understand the heart of the matter. At the same time that kind of exposition does not mention the discrimination cases or charges of violations of the law which would insert an evaluative element in the article. Thus it aims at providing complete and exact information about political vocabulary, which in turn makes it an explanatory dictionary. But on the other hand it differs from an explanatory one as it thoroughly analyzes and investigates the facts and events which led to the rise of the word. Such features make this political dictionary a quality reference book for learners of English. Analyzing different types of dictionaries, we have worked out the following principles for a lexicographical description of public political vocabulary: - ideological factors or evaluation should influence reference materials only in implicit and indirect ways, but not explicitly; - the problem of exclusion / inclusion of the examined vocabulary should be solved in the direction towards greater liberalty in the inclusion of taboo words in order to prevent a distorted picture of the language's lexical resources; - to make an offensive word appear less offensive it should be defined by a synonym, usually a more respectable word; - to prevent potential trouble modern lexicographers should employ usage labels to warn the reader against the indiscriminate use of such words or meanings (opprobrious, disparaging, offensive, etc.); - in general-purpose dictionaries definitions should reflect society's general attitude to all sorts of ideological, political or moral issues; • many words in a language can be defined in different ways according to one's ideological or moral values (cf. abortion·, poor, woman·, etc.); - the choice of the semantic content of the examples is relatively free, since examples are mainly meant to provide additional semantic information or to illustrate syntactic behavior; - lexicographers should stay neutral; - definitions should have an explanatory and encyclopaedic character. All these principles will have be taken into consideration while compiling a dictionary of public political vocabulary. We propose our own model of a public political dictionary free from the compilers' ideological standpoints - on the basis of the examination of

Olga Karpova/Svetlana

182

Manik

macro- and microstructure of different dictionaries and our principles for the description of the public political vocabulary. It should be an explanatory dictionary, as one of the most important tasks of such a work is to render the meanings of the public political words and register the role of evaluation. Moreover we are talking of a special dictionary type encoding sociopolitical words. It is a dictionary of modern word usage (of new words) since it will primarily include units of recent usage. In our opinion, lemma selection should be defined by frequency in actual usage as documented in printed sources, in Internet sources and in data from the research process of Cyberlexicography. These works are known as The Internet Dictionary Project. It has worked out a new principle of dictionarymaking, based on the involvement of users, who are requested to add and supplement entries. We think that such an approach will guarantee topicality of the ideological issues, and it will also help to cover a wide spectrum of ideological evaluation markers in the definitions. Below we suggest a sample page of such a Public Political Dictionary. We are working on creating our own vocabulary database. Meanwhile, the sample page is made on the basis of the "Oxford Dictionary of New Words", which we consider to be a suitable role model. But we have changed the lexicographical structure. To our mind the definition that has five sections (entry, definition, etymology, history and usage, and illustrations) is not convenient for a user who would be obliged to read the whole article to get the required information.

Ethnic cleansing noun

The mass expulsion or extermination of people from a minority ethnic or religious group within a certain area.

EU see EUROPEAN UNION

E.C. was reported in 1991 as conflict spread in the former Yugoslavia; recently associated with the bitter fighting between BOSNIAN Serbs and BOSNIAN Muslims, who had apparently lived in amity as neighbours for generations. It is not yet clear that E.C. in the Balkans can be considered a thing of the past.

The world still seems helpless to stop ethnic cleansing in Bosnia. - Imprimis Sept. 1994, p. 3 The area has a large number of towns and villages, many emptied of Muslims and Croats in three years of ethnic cleansing. - The Times 9 June 1995, p. 11.

Public Political Vocabulary: Model of a Dictionary

183

Euro noun

also written EURO The designated name,

The euro is likely to be broken

The name agreed for the future

proposed in 1971, for the

into 100 units, like pence to a

European currency unit, to

future European currency unit

pound The European

replace the ecu after monetary

has formed a specific area of

Commission expects the euro

union and establishment of a

controversy in the wider issue

to be known colloquially

SINGLE CURRENCY.

of a SINGLE CURRENCY. It

the pound in Britain, the mark

avoids giving preference to

in Germany and so on.

as

any of the existing national

-TheTimes 17 Feb. 1996, p.

currencies (like the euromark

34.

or the eurofranc). The name was agreed on at a summit conference in Madrid in December 1995 and designs for the banknotes were unveiled at a meeting of finance ministers in Dublin in December 1996.

The left column will contain the entry, a pronunciation label and a grammar label. To make the search quicker and more convenient, there should be a graphical label - as in the ODNW - showing the usage sphere of the entry (eg. "Elections", "Executive organ", "Legislature", "Negotiations", "Peacekeeping affairs", "Human Rights"). The definition is also found in this column. Bearing in mind the specific character of the dictionary, the definition should cover only the meaning(s) of the public political sphere. Finally, this column should contain synonyms or antonyms and set expressions involving the headword. The middle column is completely devoted to etymological information, usage points and also cross-references. We think the user's attention should be focused on the usage data since it provides extralinguistic information essential for both native speakers and learners. Various facts or events which coined new words or new connotations are listed here, since our dictionary will provide users not only with definitions, but also with pragmatic and contextual information. The right column lists authentic examples of word usage. Examples thus constitute a repository of the common values and interests of the society whose language is described. In our opinion examples cited without mentioning their sources should be avoided: to avoid subjectivity we shall list all quoted sources. Special dictionaries of public political vocabulary seem to be a necessity nowadays, due to the needs of modern society. Users want such vocabulary presented and labeled objectively, without the ideological outlook of the compilers being imposed on them.

184

Olga Karpova/Svetlana Manik

References

B6joint, H. (1994) Tradition and Innovation in Modern Dictionaries. Oxford. Cranston, Μ. Ideology (1986) II The Encyclopaedia Britannica. Dubitchinsky,V.V. (1998) Teoreticheskaya i practitheskaya leksikographiay. Vena-Harkov. [Dubitchinsky,V.V. (1998) Theoretical and Applied Lexicography. Vienna-Harkov.] Golovanevsky, A.L. (1995) Iz istorii formirovanijya ideologitchesky-otzenochnoj leksiky ν russkom literaturnom yazike // Golovanevsky, A.L. Ideologitchesky-otzenotchnij slovar russkogo yazzika XIX - natchala XX vekov. Bryansk. S. 120-127. [Golovanevsky, A.L. From the History of the Ideologically Colored Vocabulary Formation in the Russian Literature Language // Golovanevsky A.L. (1995) Ideologically Colored Dictionary of the Russian Language of the XIX - Beginning of XX century. Bryansk. P. 120-127]. — (1995) Ideologically Valued Dictionary of the Russian Language of the XIX - beginning of XX century. Bryansk. Haynes, J. (1989) Introducing Stylistics. London. Lye, J. (1997). Ideology: A Brief Guide. Brock. McArthur, T. (1995) Culture-Bound and Trapped by Technology: Centuries of Bias in the Making of Wordbooks // Lexicographica SM., 64. P. 381-389. Nida, E.A. (1995) Lexical Cosmetics // Lexicographica SM., 64. P. 69-72. Piano, J.C.; Greenberg M. (1993).7Vie American Political Dictionary. Fort Worth. Russkijyazik kontza XX stoletiay (1985-1995) (1996) Moskva. S. 32-67. [The Russian Language of the End of the XX century (1985-1995) (1996). Moscow. P. 32-67.] Sklarevsky, G (ed.) (1998): Explanatory Dictionary of the Russian Language of the End of the XX century. Language Changes. St.Petersburg. Tulloch, S. (1992) Oxford Dictionary of New Words. Oxford. Varantola, K. (1997) On the Information Needs of Dictionary Users. II Aktualnije problemy teoretitcheskoj i prokladnoj leksikographii. Ivanovo. [Actual Problems of Theoretical and Applied Lexicography. Ivanovo. P. 98-110.] Wierzbicka, A. (1995) Dictionaries and Ideologies: Three Examples From Eastern Europe // Lexicographica SM., 64. P. 181-195.

Ksenija Leban Towards a Slovene-English False-Friend Dictionary

1. Introduction

Lexicographers are becoming more and more aware of the importance of dictionary users in the compilation of lexicographical products. Their wish to meet the demands and specific needs of dictionary users has given birth to different types of dictionaries. Although most of the restricted and specialised dictionaries available on the market are monolingual, mainly because they are produced in "lexicographically developed" countries and aim at an international audience, the situation has been improving also in the production of bilingual dictionaries. As early as 1979, Tomaszczyk reported that a number of his respondents believed that "a wider range of highly specialised dictionaries would be a better solution than the increasingly bulkier general dictionaries" (Tomaszczyk 1979: 115). Since then, several authors have noticed that dictionary users prefer consulting monolingual dictionaries to bilingual ones or constantly check the information provided by bilingual dictionaries in the available monolingual lexicographical works. It has been suggested that the main reason for the inadequacy of general bilingual dictionaries is their desire to satisfy the needs of all prospective dictionary users (Tomaszczyk in Hartmann 1983). The reference needs of dictionary users are simply too diverse and incompatible to be met efficiently in one dictionary. In recent years, tendencies towards the production of specialised bilingual dictionaries can be observed in Slovenia as well, where, for the time being, the compilation of restricted bilingual dictionaries has not become part of standard lexicographical practice. One of the possible restricted-dictionary projects might consist of the production of a Slovene-English false-friend dictionary.

2. The Definition of Slovene-English False Friends

Slovene-English false friends could be described as pairs of words of non-native origin which have preserved their foreign-looking and/or -sounding form in the Slovene language, and maybe in the English language as well, and in which the misleading relationship between meaning and/or form may induce us to commit interlingual errors.

186

Ksenija Leban

3. The Typology of Slovene-English False Friends

According to the relationship between form and meaning, false friends can be divided into: • semantic false friends, i.e. pairs of words which are identical or similar in form, but (partly or wholly) dissimilar in meaning (organ v. organ)·, • pairs of words which are identical or similar in meaning, but dissimilar in form. Three types may be established: a) morphological false friends, i.e. pairs of words which are identical or similar in meaning, but of different morphemic structure (avtoportret v. self-portrait)·, b) phonological false friends, i.e. pairs of words with which dissimilarity in pronunciation, sometimes accompanied by differences in spelling as well, is restricted to the phonological level (penicilin v. penicillin), and c) orthographical false friends, i.e. pairs of words with which dissimilarity in form is restricted to spelling (laso v. lasso)·, • zero-equivalent false friends, i.e. words of non-native origin whose foreign-looking and/or sounding form might induce us to believe that there exist corresponding words, identical or similar in form, in the target language, when in fact they do not (avtostopar v. hitchhiker). Semantic false friendship may be established on the conceptual, connotative, stylistic and collocational levels. It should be pointed out that semantic false friends are not necessarily established on one of the above-mentioned levels only. Very often differences between the so-called conceptual false friends occur on other levels as well. Considering the dögree of treacherousness and the number of false friends in separate categories, a hierarchy of semantic false friends may be introduced. With semantic false friends, one could establish that conceptual false friends are more treacherous (and numerous) than stylistic, collocational and, finally, connotative ones. Conceptual false friendship is established when there exists only a partial overlap between two lexemes or when there is no overlap at all between the two items. Accordingly, conceptual false friends are divided into partial and total false friends. While total false friends are pairs of words whose meanings are completely disparate, partial false friends only partially overlap in meaning. They are more treacherous than total false friends, for their partial overlap often induces us to believe that the vocabulary items concerned coincide completely. Three types of partial false friends ought to be mentioned. a) The first group consists of pairs of words in which the LI lexical item is narrower in meaning than its L2 counterpart, i.e. LI < L2. This is by far the largest group of partial false friends, with new items entering the group almost every day (e.g. leming ν lemming). b) The second group comprises pairs of words in which the LI lexical item is broader in meaning than its L2 counterpart, i.e. LI > L2 (e.g. limonada ν lemonade). c) The third group is composed of pairs of words which are at the same time broader and narrower in meaning, i.e. LI L2 (e.g. lazaret ν lazaretto /lazaret/lazarette).

Towards a Slovene-English False-Friend Dictionary

187

Apart from differences occurring on the conceptual level, false friendship may (also) be established on the connotational, collocational and stylistic levels. Although there are hardly any purely connotational false friends (lasciven ν lascivious), there is a significant number of conceptual false friends with different connotations ( a f e r a ν affair). Sadly, the connotational meaning is difficult to define as it so often appears to be highly subjective and subtle because it is very context-dependent and therefore unstable. Collocational false friends are pairs of words with different ranges of collocation (aparat ν apparatus), while stylistic false friendship results from differences occurring on the stylistic level. Stylistic differences between LI and L2 lexical items seem to be restricted to the following areas: • field of discourse, when one of the items is restricted to a particular technical field while its counterpart belongs to core vocabulary (antena ν antenna tech / aerial), • tenor, especially when the contrast between the two items is that of formal v. informal

(eksakten (-a, -o)Jml ν exact), • frequency of use (nataliteta ν natality rare, birth rate) and • geographical varieties, especially the differences between American English and British English vocabulary items (plastelin ν plasticine BrE. play dough AmE). In 1998, I examined internationalisms, i.e. "words which are used internationally" (Ivir 1988: 93), beginning with the letter L' in Slovene and English with a view to establishing how significant the phenomenon of false friendship really is. The Slovene corpus consisted of 895 Slovene lexical items. On the basis of dictionary information, it was established whether the lexical pairs differed on the semantic, morphological, phonological and/or orthographical levels. When no divergences could be found or when they proved to reflect systematic differences between the two languages, lexical pairs were classified as true friends. Often, lexical pairs differed on more than one level. A hierarchy of false friends was therefore developed. This partly reflects the conclusions of other analyses and definitions of false friends (Crystal 1987, Golöbiö 1988 and 1989, Granger and Swallow 1988, Hayward and Moulin 1984, Ivir 1968 and 1988, Limon 1983, Malone 1982, Partington 1993, Topalova 1997, Van Roey 1990 and Welna 1977), and the results of the comparison of the lexical items included in the corpus. While analysing the above-mentioned lexemes, it could be observed that with certain types of false friends, differences may occur on various levels. Semantic false friends, for example, may differ in meaning, morphology, pronunciation and spelling. With morphological false friends, the meanings of the lexemes are the same - the lexical pairs would differ in their morphological structure. Often morphological differences would be accompanied by phonological and orthographical divergences as well.

'

Edward L. Thorndike, on the basis of a study of the lexicon, divided the English alphabet into 105 approximately equal units, called blocks. The letter L occupies four blocks or almost 4 per cent of the English vocabulary. Since a considerable number of lexical items of foreign origin begin with this letter, it was thought that internationalisms beginning with the letter L might be considered a representative sample.

188

Ksenija Leban

Finally, while differences in pronunciation with phonological false friends may also be supported by different spelling, divergences among orthographical false friends are restricted to orthography. In the above-mentioned corpus of 895 Slovene lexical items and their English counterparts, there were 84 true friends, 14 orthographical false friends, 184 phonological false friends, 236 morphological false friends, 232 semantic false friends and 115 zeroequivalent false friends. Out of 232 semantic false-friend pairs, 179 entered the conceptual false-friend category, 37 the stylistic, 14 the collocational and 2 the connotational false friend group. The conceptual false-friend group consisted of 45 total false friends and 134 partial false friends. Out of these 134 cases of partial false friendship, the Slovene lexical item was narrower in meaning than its English counterpart (LI < L2) in 78 instances, broader in meaning (LI > L2) in 37, and at the same time broader and narrower in meaning (L1L2) in 19 instances. The above-mentioned numbers prove that the phenomenon of false friends is more widespread and that the differences between lexical items are more complex than is generally assumed.

4. The Prototypical False-Friend Dictionary

The prototypical false-friend dictionary may be described as a bilingual restricted dictionary, usually arranged according to the alphabetical macrostructure and covering up to several thousand false-friend pairs, in an attempt to preclude errors which are due to the assumption that a target-language lexeme has both the same or similar meaning and/or form as its corresponding source-language counterpart. Since false-friend dictionaries mainly cater for dictionary users engaged in encoding activities, with a high level of mother-tongue and foreign-language competence, accompanied by considerable linguistic training, the differences between LI and L2 lexemes may - or should - be dealt with on all levels, but most of all, very thoroughly.

5. T h e Slovene-English False-Friend Dictionary and Its Characteristics

a) The Target Group This type of dictionary should meet the needs of a very specific target group characterised by a high level of both LI and L2 competence. In other words, the false-friend dictionary target group would primarily include translators and interpreters, together with language teachers and language students. Although the dictionary might be of interest to others as well, it is those aiming at impeccable language performance who should be the lexicographer's primary concern.

Towards a Slovene-English False-Friend Dictionary

189

b) The Word List The lexicographical team should decide which types of false friends to include in the dictionary and where. It has already been suggested that a hierarchy of false friends be established and decisions regarding the lexicographical treatment of the lemmata be based on it. It should be added that the analysis of internationalisms beginning with the letter L between Slovene and English showed that misleading differences occur with all the false-friend types, which means that a thorough false-friend dictionary should not a priori neglect or exclude any false-friend type. On the other hand, however, an all-inclusive false-friend dictionary does not seem to be a feasible dictionary project since it would probably surpass the restricted-dictionary size by far. Should the lexicographer decide on a maximising lemma selection, the L section of the Slovene-English false-friend dictionary would be covering no less than 811 lemmata, which means that this particular section would include only 172 lemmata less than the largest Slovene-English bilingual dictionary, with 983 dictionary articles in its L section. I therefore propose that the Slovene-English false-friend dictionary cover the most frequently occurring internationalisms resulting in false friendship. It has been suggested that false friendship may be caused by the misleading relationship between form and meaning. Although semantic false friends are hardly ever identical, but usually similar in form, they differ from other types of false friends. With morphological, phonological, orthographical false friends and even zero-equivalent ones, where dissimilarity in form reaches the very extreme of non-existence - the misleading differences are restricted to form. Dictionary users therefore only need to be given the information about the correct form of the target-language lexicographical equivalent. Perhaps it would be advisable to include only those most frequent internationalisms resulting in the abovementioned false-friend types where discrepancies may also be observed in writing. In other words, only those phonological false friends where dissimilarity in pronunciation is accompanied by differences in spelling as well should be covered by the dictionary. Whether to include morphological, phonological, orthographical and zero-equivalent false friends in the dictionary, depends also on the quality and reliability of other, especially general, bilingual dictionaries available on the market. With semantic false friends, however, the lexicographical team should opt for a detailed, even meticulous dictionary treatment covering all the differences and similarities between the lexemes. This is very important, for it is the exhaustive treatment of the contrasted lexemes that should captivate the dictionary buyer. It goes without saying that the treatment should be more thorough than that of a general bilingual or multilingual dictionary. The allinclusive treatment of the lemmata should compensate for the limited number of dictionary articles and convince the user to purchase the lexicographical work. The principal word list should then consist of frequently occurring internationalisms resulting in either partial or total false friendship. Other types of semantic false friends may appropriately be treated by general bilingual and multilingual dictionaries as well. With stylistic and connotational false friends, this may be done through careful labelling, while with collocational false friends, all the relevant collocations may be listed in the respective dictionary articles either in general dictionaries or in dictionaries of collocations.

190

Ksenija Leban

c) The Macrostructure The above-mentioned suggestions will be reflected in the dictionary's macrostructure. The lexicographical team may, for example, decide to treat the selected partial and total false friends between Slovene and English in the principal word list supplemented with a reversed word list functioning as an index referring the dictionary user to the page(s) where the respective dictionary articles may be found. The dictionary back matter may also include several appendices comprising morphological, phonological, orthographical and zero-equivalent false friends respectively. Each of the appendices should be accompanied by a reversed list of the above-mentioned false friends. Although such a complex macrostructure might be approved by dictionary users fascinated by language, such as linguists, it would not make for fast information retrieval. The best solution would probably be a Slovene-English false-friend dictionary with the main word list consisting of selected partial and total false friends between Slovene and English, followed by a mixture between an index and an appendix. This should consist of an alphabetically arranged list of the most frequently occurring internationalisms between Slovene and English resulting in false friendship accompanied by the respective lexicographical equivalents in the target language. All the lexical items included in the principal word list should be supplemented with references to the pages on which they are treated more thoroughly. Since the lexicographical team should aim at bidirectionality, the principal word list ought to be followed by two dictionary components of this type, one of starting with the Slovene lexemes provided with their English lexicographical equivalents and the other consisting of an alphabetically arranged list of the English lexemes complemented with Slovene lexicographical equivalents. Whichever the approach, it should be explained, together with other lexicographical decisions, in the metalexicographical introduction to the dictionary. Apart from the list of contents, which is not obligatory, although it remains a very userfriendly dictionary component, especially if occurring in the front matter, and the dictionary grammar, which, in the case of a false-friend dictionary, is not necessary as the work is intended for dictionary users with a high level of LI and L2 competence, the front matter of a Slovene-English false-friend dictionary should consist also of a preface, a user's guide, a list of abbreviations and an encyclopaedic section. The latter should, in a user-friendly way, introduce and discuss the phenomenon of false friends. A false-friend typology constituting the basis of the dictionary ought to be explained. Furthermore, the dictionary users should be enlightened about the complexity of this interlingual phenomenon in order to be able to make full use of the information provided by the dictionary. Since the lexicographical work should aim at bidirectionality, all the above-mentioned components, including the informative label, should be given in both dictionary languages, i.e. in Slovene and English.

Towards a Slovene-English False-Friend Dictionary

191

d) Bidirectionality False-friend dictionaries should aim at bidirectionality. 2 There are several reasons for selecting this approach. First of all, false-friend dictionaries cater for a very specific target group characterised by a high level of both LI and L2 competence. Second, dictionaries covering languages of limited distribution cannot count on breathtaking sales figures. In other words, one false-friend dictionary between languages of limited distribution or between a language of limited distribution and a language of wider distribution is usually supposed to ftilfil the needs of both LI and L2 native speakers. Third, a contrastive microstructure provides for a clearer view of the similarities and differences of false-friend pairs. Fourth, false friends do not pose problems only when translating into foreign languages, but, surprisingly enough, also when translating into the mother tongue. 3 Bidirectionality does not only influence the macrostructure, but also, or especially, the design of the microstructure. 4

e) The Microstructure In order to ensure maximum user-friendliness of a false-friend dictionary's microstructure, the microstructure designed by Van Roey, Granger and Swallow would, with several relatively minor modifications, best serve the purpose. Thus false friends should be treated in two juxtaposed columns on the same page, with each column being dedicated to one language:' LI

L2

Such a contrastive microstructure would allow for a clearer view of the similarities and differences between the lexical items analysed as well as for fast information retrieval. Each dictionary article should be introduced with the false-friend pair. Articles should be subdivided into three parts, each section being introduced by, in the case of a Slovene-English false-friend dictionary, Slovene and English lexicographical equivalents.

2

' 4

5

If compared with multilingual projects, bilingual false-friend dictionary projects can treat false friends more thoroughly; the dictionary structures, and especially microstructure, can be better exploited, but most of all (with bilingual false-friend dictionaries) the lexicographical team may rely on the high LI and L2 language competence of the potential dictionary users, which is hardly ever the case with multilingual false-friend dictionaries. This has been acknowledged by authors such as Hartmann (quoted by GabrovSek 1998) and Browne (Browne 1987: the informative label). In order to ensure that a dictionary will equally benefit LI and L2 native speakers, all the dictionary components should be written in both languages, the exception being the main word list. However, a reversed word list, supplemented with page references to the respective dictionary articles would direct L2 dictionary users. This approach differs from the approach adopted by van Roey et al. (1995) who decided to shift the languages in the third section (see below).

Ksenija Leban

192

The first section should list all the senses in which the Slovene and English lemmata coincide (LI = L2). The second section should cover all the senses of the Slovene lemma which are rendered into English by means of lexemes other than the English lemma. The second section is therefore dedicated to the cases in which the Slovene lemma is broader in meaning than its English counterpart (LI > L2). The third section, on the other hand, should cover all the senses of the English lemma which are rendered into Slovene by means of lexemes other than the Slovene lemma. This section is then dedicated to the senses in which the English lemma is broader in meaning than its Slovene counterpart (LI < L2). A FALSE-FRIEND PAIR LI =L2 LI >L2 LI to shelve, calf -)· to calve.6 The term "zero-derivation", on the other hand, is usually restricted to instances of full conversion (see Adams 1973, Marchand 1969, Kastovsky 1977). Since the present discussion is concerned only with one

4

5

6

Lipka (ibid.), along with Quirk et al. (ibid.), argues against Marchand's claim that the term "conversion" (first used in the above sense, to the best of my knowledge, by Kruisinga (193132)) should be avoided in synchronic treatments, since word class in contemporary English is neutralized or latent and to a great extent context-dependent. Similarly to Aitchison, he sees conversion as a basically synchronic process now available for the extension of the lexicon. Due to clear grammatical differences in word class and the close parallelism with overt suffixal derivatives (see the equations above), Lipka argues that conversions should be included with homonymy. Note that the direction of the synchronic derivation sometimes may not reflect the actual diachronic development. Relying on a set of criteria for determining the direction of derivation proposed by Marchand (1964), Lipka suggests that an item should be regarded as derived, if in its paraphrase the other item is applied; thus to nail is derived from nail, since its paraphrase is 'fasten with a nail' (see Lipka 1990: 84ff).

214

Sändor Martsa

type of full conversion, this difference in scope and use will be ignored and the two terms will be used synonymously. 7 3. Word-formation and semantic transfer; evidence for polysemy Generally, it is assumed that the meaning of a derived item is not arbitrary, since it can be predicted from the meaning/s of the motivating base/s and/or the affix/es being attached to it. Thus the meaning of writer is predictable from the meanings of write and the agentive suffix -er, or the meaning of the endocentric compound steamboat is predictable from the meanings of its constituents steam and boat. Non-arbitrariness of the meaning of the majority of derived lexemes is prima facie evidence for the intrinsic relationship between word-formation and semantic transfer. Although the nature of this relationship may vary depending on actual processes of word-formation, there is every reason to believe that in different types of word-formation morphological or morpho-syntactic derivation goes hand in hand with semantic derivation. In other words, in word-formation an onomasiological act is necessarily accompanied by a semasiological act. What has been said about the relationship between word-formation and semantic transfer evidently applies to noun-verb (and perhaps to other types of) conversions too. Consequently, the meanings of to pig 'to eat a lot, overeat' or to chain 'fasten with a chain' appear to be respectively derivable, that is to say, to a large extent predictable from the meanings of pig or chain, or more precisely from the knowledge (or frame) speakers possess about pigs and chains. It has to be noted, however, that the relationship between a noun and a verb zero-derived from it is in many respects specific and differs from the relationship described in the previous paragraph in general. The formal difference is obvious: the lexemes involved in the process of conversion are identical phonologically and/or graphemically. The more important and essential difference is that zero-derived lexemes in fact are word metonyms or word metaphors, which is inconceivable in the case of other derivatives. This then means that in noun-verb conversions metonymic or metaphorical transfer takes place, since zero-derived verbs are metonymically or metaphorically derived from the corresponding parent nouns. If this is so, a claim can be made that the semantic relationship between lexemes involved in the process of conversion is to be interpreted as polysemy and not homonymy. As has been mentioned earlier, this view defies the traditional distinction made between homonymy and polysemy, though it is not entirely unprecedented. Similar to Lakoff s above opinion on the genesis of polysemy, other linguists also have come close to expanding the scope of polysemy beyond the boundaries of single lexemes belonging to different syntactic classes. Examining regular polysemy, Miller (1978: 104) postulated a so-called construal rule which applies to a set of English nouns that can be used as verbs, where the verbal use is derived from the nominal use in a regular way. Dealing with systematic or conventional polysemy, Copestake and Briscoe (1995: 15) maintain that when the senses of formally identical but syntactically different lexical items are predictably related, they represent categorial polysemy as opposed to categorial homo7

The terms, "functional change/shift" or "functional derivation", also used for conversion, are found infelicitous by Marchand (1969), since zero-derived words do not enter other functional categories.

Homonymy vs Polysemy: Conversion in English

215

nymy, where predictability is irrelevant. In keeping with this, sink and to sink should be considered homonyms, since the meaning of the latter is not predictably related to the former, whereas the meaning of to respect is predictable from the meaning of respect, so they display categorial polysemy. Panman (1982: 105f!) notes that homonymy and polysemy are lexico-semantic relations between word-tokens rather than between lexical items and the only valid criterion that distinguishes them in the case of conversions is meaning similarity recognized by the speakers. Investigating the phenomenon of polysemy in a wider, cultural perspective, Lehrer (1990: 207f) claims that the term "polysemy" understood in a non-standard way can be applied to all cases where there is phonological and orthographic identity of semantically related forms (cf. comb and to comb). Drawing partly on the classic study of Clark and Clark (1979) on nouns surfacing as verbs, she classifies zero-derived verbs into groups according to the semantic (theta) roles the corresponding parent nouns typically fulfill in utterances (1990: 230ff). These groups are as follows: 1. Agents: to butcher, to father, to model·, 2. Instruments: to bicycle, to hammer, to mop; 3. Patients: to button, to farm, to light; 4. Goals: to group, to loop, to pile; 5. Locations: to bench, to porch, to tree', 6. Body parts: to arm, to hand (in), to head. This classification is based on conceptually justifiable generalizations, conceptual frames which, very much in the manner of Lakoff s understanding of polysemy, underlie the predictability and thus the polysemous nature of the meanings of zero-derived verbs. Finally, mention must be made of a recent study in which, also within the cognitive semantic framework, Kövecses and Radden (1998: 54) discuss the so-called metonymyproducing relationships, such as the whole idealized cognitive model (1CM) and its parts and parts of an ICM1 With respect to the latter they claim that in an action ICM (as being a type of event ICMs) the relationship between the predicate (relational element) and one of the participants (i.e. agent, instrument, time, etc.) of the event may give rise to metonymy. In the action ICM the following types of metonymic relationships are involved (I will mention only those relationships that are relevant for the present discussion): (1) (2) (3) (4) (5) (6) (7)

INSTRUMENT FOR ACTION: to ski, to shampoo (the hair) AGENT FOR ACTION: to butcher, to author (a book) OBJECT INVOLVED IN AN ACTION FOR THE ACTION: to blanket (the bed) MEANS FOR AN ACTION: to sneeze (the tissue off the table) MANNER OF ACTION FOR THE ACTION: to tiptoe TIME PERIOD OF ACTION FOR THE ACTION: to summer (in Paris) DESTINATION FOR MOTION: to porch (the newspaper)

This classification of metonymic relationships is fully consistent with both Lakoff s above view on the genesis of polysemy and the idea of non-standard polysemy suggested by Lehrer. For it is the systematic relationships between the elements (i.e. the predicate and the participants) of the same idealized cognitive model (i.e. action ICM) that generate polysemy reflected in a number of metonymic applications of verbs zero-derived from nouns.' This classification also sheds light on how (cognitively) salient functions of entities ' '

ICM here is taken to be similar to, though not identical with, the notion of frame. Zero-derived verbs may indicate other metonymic relationships in other ICMs too. To pity and to hunger, for instance, can be taken as instantiations of the metonymic relationship FEELING/EMOTION FOR EXPERIENCING FEELING/EMOTION.

216

Sändor Martsa

motivate the use of the names for these entities as verbs indicating actions associated with them. Moreover, it also implies that the predictability of derived senses requires considerable language processing from the hearer to compute the derived senses. In this connection Aitchison (ibid.) notes that, as elsewhere in word-building, to arrive at the proper understanding of the sense in which a zero-derived verb is used, active computation is needed, integrating existing knowledge with the current situation. For an illustration of this see the two distinct senses of to pizza in the following two sentences: Sammy pizza-ed (= dropped it on) the floor. Felicity pizza-ed the dough (= converted the dough to a pizza).

Aitchison has found that adults usually use these conversions in simplifies the problem of interpretation for the hearer.10 The pragmatic role of the context (situation) and mutual innovative senses for denominal verbs is exhaustively discussed Relying partly on Grice's cooperative principle (1975), they convention for innovative denominal verbs (ibid., 787):

conventional ways, which knowledge in computing in Clark and Clark (ibid.). formulated the following

In using an innovative denominal verb sincerely, the speaker means to denote a) a kind of situation b) that he has good reason to believe c) that on this occasion the listener can readily compute d) uniquely e) on the basis of their mutual knowledge f) in such a way that the parent noun denotes one role in the situation, and the remaining surface arguments of the denominal verb denote other roles in the situation.

In keeping with this convention, hearers are supposed to understand innovative uses such as to bottle in the sentence: They bottled (= threw bottles at) the demonstrators

To conclude this part, it can be postulated that the semantic relationship between lexemes involved in the processes of conversion can be viewed as a kind of polysemy that can be called non-standard or non-default polysemy (as contrasted to standard or default polysemy that holds between the senses of the same lexeme). In respect of noun-verb conversions semantic relatedness means that the sense/s of denominal verbs are regularly related to and thus predictable from the senses of the corresponding parent nouns and/or the situations in which the derived verbs are employed (see the uses of to pizza and to bottle above). Senses of zero-derived verbs are related to the senses of the respective nouns in two major ways: metonymically and metaphorically. In the examples below to knife relates to knife metonymically, whereas to hare is related to hare metaphorically. Compare: The victim had been knifed in the chest. I called out to him but he hared off in the opposite direction. (S. O'Connell)

10

It is important to emphasize, especially from the point of view of second-language acquisition, that innovative denominal or deadjectival verbs never accept irregular forms, even if they sound like an irregular verb: cf. ring ('circle') —> to ring (ringed) ('surround'). (For further information see Pinker 1999: 158ff.)

Homonymy vs Polysemy: Conversion in English

217

In the following sections I turn to the analysis of noun-verb conversions. Given constraints on space, however, it is impossible to provide a comprehensive analysis of the types of metonymic and metaphorical relatedness within this type of conversion. Therefore, in sections 4 and 5 I will examine only those denominal verbs that are zero-derived from names for animals and instruments. I will point out that both groups of denominal verbs enter various polysemy relations due to the permanent interplay between metaphorical and metonymic patterns."

4. Animal verbs' 2

Examining English verbs zero-derived from animal names (animal verbs, for short), I have found that with respect to the types of semantic transfer operating in the process of morphosemantic derivation these verbs can be classified into two groups. To one of them, within which three subgroups can be identified, belong verbs that can be considered metaphorical extensions of the corresponding animal names (cf. hare -» to hare (off) 'to run o f f ) . Verbs constituting the other group, on the other hand, can be viewed as metonymic extensions of the respective animal names (cf. foal —> to foal 'to give birth to a baby horse'; worm -> to worm 'to give an animal medicine in order to kill the worms in its intestines'). In other words, considering the types of semantic transfer, English animal verbs can be considered wordmetaphors and/or word-metonyms." In what follows, I will discuss only word-metaphors, that is to say denominal verbs that are to be treated as metaphorical extensions of names from animals. The examples below are intended to demonstrate that animal verbs as word-metaphors form three distinct subgroups in accordance with three different morpho-semantic patterns underlying their derivation from the corresponding animal names. One group includes verbs that are perceived as straightforward metaphorical extensions of animal names, though these animal names (e.g. hare) as a rule are not used metaphorically as nouns. Another group of verbs are derived from animal names (e.g. ape), which have metaphorical applications, but these applications are not related to the metaphorical senses of the derived verbs. Finally, there is a third group comprising verbs that are evidently the metaphorical extensions of the metaphorical senses of the corresponding animal names (e.g. beaver). Schematically, these three patterns are as follows: (the arrows here as well as in the examples below show the assumed direction/s of metaphorical transfer)

" For further information on the interaction of these two types of polysemy in general see Warren (1992: 51ff) 12 A more detailed discussion of verbs zero-derived from animal terms can be found in Martsa 1997, 1999. " Note that animal verbs belonging to either of the two groups have fully adapted to the system of English verbs by taking particles that have primarily aspectual functions: e.g. they indicate the iterativity (e.g. to monkey about, to horse about/around) and, more typically, the perfectivity or completion of the action or state in question (e.g. to beaver away, to chicken out, to pig on, to wolf down, to squirrel away).

218

Sändor

Martsa

Pattern 1: Noun —> Verb (metaphorical) E.g. squirrel —> to squirrel (away) ('hide') Pattern 2: Noun, -> Noun 2 (metaphorical) E.g. wolf i—> wolf Verb (metaphorical)

2

('a womanizer')

to wolf (down) ('eat a lot')

Pattern 3: Noun! —> Noun 2 (metaphorical) Verb (metaphorical) E.g. chickent—» chicken2 ('a coward') —> to chicken (out) ('decide not to do sth, because it's too dangerous')

The presentation of the examples illustrating each pattern is designed in the following way: in each example the animal names as parent nouns are given first with short explications attached to them. The aim of the explications is to demonstrate what type of world- or folk knowledge is associated with particular real-world animals and, more importantly, how this knowledge provides the cognitive ground for the metaphorical applications given next to the arrows either in the form of nouns (patterns 2, 3) or denominal verbs (patterns 1,2,3).14 Both the explications and the glosses on metaphorical applications were adapted from the dictionaries listed at the end of this paper.

4.2. Pattern 1 hare [animals of this kind are known to be able to run fast, especially when they are frightened]

ferret [animals of this kind are known to be fierce hunters; sometimes they are kept by people for hunting rabbits and rats]

->

->

to hare (off) 'run off (about animals)' E.g. I called out to him but he hared off in the opposite direction, at top speed. (S. O'Connell) to ferret\ 'search busily for a thing (in not a very organized way) (about humans)' E.g. The General Director ferreted in his breast pocket for his reading glasses. (CEGM) to ferret2 (out) 'find sth by thoroughly searching' E.g. Ο 'Connor was the person who ferreted out the truth in this case. (CEGM)

squirrel [animals of this kind are known to hide or store food in order to eat it later] —>

14

to squirrel (away) 'hide sth (about humans)' E.g. As soon as they get paid they squirrel the money away so they won't be tempted to touch it. (CIDE)

For a detailed discussion of what constitutes folk knowledge concerning animals see Martsa 1999.

Homonymy VJ Polysemy: Conversion in English

219

Other examples may include: carp to carp 'complain'; drone -> to drone ι (on) 'talk boringly for a long time in a low, monotonous voice' —> drone2 'make a low buzzing or humming noise'; parrot -» to parrot 'repeat exactly without understanding'; rabbit —> to rabbit (on) 'continue to talk too much about sth or nothing'.

4.3. Pattern 2

ape ι

ape2

[animals of this kind are known to have a clumsy build, an ugly face; they are -> also known for their ability to mimic human gestures, especially in captivity]

'an ugly, clumsy, stupid person' E.g. Show me your identity card instead of shooting at my feet, you ape. (COBUILD)

I to ape 'imitate, mimic, usually unsuccessfully (about humans/things)' E.g. He called the new building unoriginal and said that it merely aped the classical traditions. (CIDE)

dogi [some species of animals of this kind are fierce and dangerous; they are believed to be faithful to their masters] -> i to dog\ 'follow closely (about animals)' E.g. He's been dogging me all day. (COBUILD) to dogi 'be with sb all the time (about things)' E.g. We were dogged by bad luck throughout the journey. (LDELC)

dog2 'an unpleasant, evil person' E.g. He let her down, the dirty dog. (COBUILD)

Sändor Martsa

220

wolf χ [animals of this kind usually hunt in packs; male species are also known to roam and hunt alone; they eat quickly and greedily]

(lone) wolf2 'a lonely or solitary person' -> wolfs 'a womanizer' E.g. He had the reputation of being a bit of a wolf. (CIDE)

to wolf(dov/n) 'eat a large amount quickly and greedily (about humans/animals)' E.g. The boys wolfed the sandwiches down and then started on the cakes. (CIDE) Other examples belonging to this pattern are: duck\-• duck-i 'a likable person' I to ducki 'go quickly under water' —> to duck2 'avoid sth' ; snake ι -> snake2 'a dangerous person'

i to snake 'move in long, winding curves'; worm\ -» worm2 'a person not deserving respect'

to worm (one's way through/into sth) 'move slowly because of difficulties'.

4.4. Pattern 3 beaver \ [animals of this kind are known to be very busy]

beaver2 ->

'a hard-working person' E.g. Eager beaver.

to beaver (away) ->

'work hard (about humans)' E.g. I've been beavering away at my typewriter all day and I've nearly finished Chapter 2. (S. O'Connell)

221

Homonymy vs Polysemy: Conversion in English

chickeri\

chicken2

to chicken (out)

[birds of this kind are shy and timid; they are easily scared]

'a cowardly person' E.g. What a chicken he won't jump. (COBUILD)

'decide not to do sth, because it's too dangerous' E.g. A group of us went hang-gliding, but a couple of people chickened out of it at the last moment. (CIDE)

Pigi

Pigi

to pig (out/on)

'a dirty, greedy, —> unkind, oppressive person' E.g. What a pig! He refused to help, even though he could see we were having trouble. (CIDE)

'eat a lot (on a particular occasion)' E.g. She is always pigging herself on a chocolate. (CIDE)

[animals of this kind are usually dirty, filthy; they eat a lot greedily and noisily]

-»

Other verbs to be considered are: fox ι —> fox-ι 'a shrewd, cunning person' -> to fox χ 'cause a lot of difficulty; baffle' —> to fox2 'deceive'; monkey\ -> monkey2 'a lively, naughty child' to monkeyi (about) 'behave in a silly way; be playful' —> to monkey2 (about) 'interfere with sb's affairs' —> to monkey3 'do sth to sth in a careless way'; rati -> rat2 'a disloyal, dishonest man' to ratx (on) 'act in a disloyal way; betray';

5. Instrument verbs

The term "instrument" will be applied broadly to the name of whatever entity by means of which an action can be carried out. Consequently, it will subsume not only hammer and padlock, identified respectively as instrument and patient or as instrument and object involved in the action in Lehrer's and Kövecses and Radden's above classifications, but also elbow or head, bus or helicopter, butter or sugar, drum or trumpet. These entities can in some way or other be conceived as instruments, with the help of which specific actions can be performed. Considering the aforementioned classifications and relying on the available linguistic data, instrument nouns seem to fall into two groups that are based on the roles instruments play in the accomplishment of the actions in question. Generally speaking, an instrument can be either dispensable, non-canonical (e.g. whistle) or

Sändor Martsa

222

indispensable, canonical (e.g. ink) from the point of view of the given action (see Kiparsky 1983: 11). Besides, linguistic data reflecting everyday experience suggest that the use of instruments, whether dispensable or indispensable, often leads to some change in or of the physical state or location of the target entities. The use of instruments, particularly dispensable, non-canonical ones, also implies that the target entities are manipulated in some way. Based on these considerations, two groups of instrumental verbs can be set up in English, representing as a matter of fact two metonymic patterns: 1. non-canonical instrument for action pattern, 2. canonical instrument for action pattern.

5.2. Instrument verbs of the non-canonical instrument for action pattern Denominal verbs instantiating this pattern indicate activities in which the canonical instruments are replaceable by other instruments employed in the same function. Thus, for example, the act of hammering can be performed not only by a hammer, but also by a piece of stone or by a rock. This fact, as a rule, is reflected in the dictionary descriptions of these verbs too. To hammer, for example, is defined in the ALD4 as 'to hit or beat (sth) with a hammer or as if with a hammer'; or the definition for to string (up) in the same dictionary runs as follows: 'to hang or tie (sth) with a string, rope, etc.'. If the instrument is other than the expected basic one, it can be expressed by an additional prepositional phrase, as is demonstrated in the following examples borrowed from Jensen (1990: 93):'5 He hammered the nail with a rock. He brushed the clothes with his hands. I paddled the canoe with a copy of The Financial Times. She strung up the picture with a wire. Can you whistle with a blade of grass? He had to saw the bread with a knife. They anchored the ship with a rock. He combed the grass with a rake. We wedged the door open with a brick.

By contrast, if the denominal verb indicates an action performed by the primary, canonical instrument incorporated in the meaning of the verbs, its use in the sentence will be felt redundant (cf. Lehrer 1990: 231). Compare: (?) Lee ironed the shirt with an iron.

Redundancy is removed, however, if the instrument is modified. Compare: Lee ironed the shirt with a new GE iron.

As regards the processes of semantic transfer, this group is rather complex. In my opinion this is due primarily to the permanent interaction between metonymic and metaphorical transfers permeating the whole group. We cannot go into all details, so two examples will suffice. The fact that zero-derived verbs constituting this group can indicate activities carried out by canonical as well as non-canonical instruments signals that these verbs are 15

Some of these examples are adapted from Kiparsky (ibid.)

Homonymy

vs Polysemy:

Conversion

in English

223

likely to undergo partial or full metaphorization. For the illustration of this consider some of the possible uses of to hammer below: ( 1 ) 1 could hear him hammering (away) in the house next door. He hammered the nail into the wall. (2) He hammered on the table with his hand. (3) Manchester United were hammered 5-1. They had English grammar hammered into them. After much discussion the negotiators hammered out a compromise settlement.

Whereas in (1) and (2) to hammer respectively means 'to hit or beat with a hammer' and 'to hit or beat as if with a hammer', in (3) these meanings metaphorically shift to 'to defeat (sb) utterly', 'to force to learn sth by repeating', and 'to achieve sth by great effort'. These metaphorical transfers are motivated by salient, noteworthy aspects of hitting or beating such as application of force and repeated strikes on the target. Note that these metaphorical uses of to hammer are already not related directly to the parent noun hammer, for the secondary metaphorical verbal senses are derived from the metonymic verbal sense, thus evoking non-standard polysemy. The complex nature of this group is also confirmed by to whistle being related metonymically to both senses of the noun whistle ('a clear shrill sound made by forcing breath through a small whole between partly closed lips', e.g. He made a whistle·, and 'an instrument used to produce a clear shrill sound, especially a signal', e.g. The referee blew his whistle), which in turn are also related metonymically." Consequently, to whistle also has two senses: 'to make a sound of a whistle' (Cf. The boy was whistling (away) cheerfully) and 'to make a sound of a whistle using an instrument' (for the illustration of the second sense see J. Jensen's example above).17 Although only the second sense is relevant from the point of view of the present discussion, the first sense cannot be ignored either because of the nature of the given activity: namely, whistling without producing a specific sound is unthinkable. Finally, with regard to the above examples it seems that in spite of the complexity and heterogeneity characterizing this group the metonymic senses of denominal verbs are relatively easily accessible, whereas the interpretation of metaphorical senses needs more language processing and contextual support. This is in line with the basic assumptions concerning the nature of metonymy and metaphor in the cognitive semantic framework, according to which the former is viewed as a referential function, whereas the latter is associated with the function of understanding (Lakoff & Johnson, 1980: 36).

16

17

I will ignore the fact that the first sense itself contains further metaphorical senses found in the phrases the whistle of a steam engine, a blackbird's whistle. Further metaphorizations or metonymizations are also possible. Cf. The bullet whistled past his headJ He whistled his dog back.

224

Sändor Martsa

5.3. Instrument verbs of the canonical instrument for action pattern Metonymie senses of the zero-derived verbs exemplifying this pattern1" are relatively easily accessible, since the instruments denoted by the nouns directly determine the activities that can be carried out with them. Thus to chain, to button, to screw respectively mean 'to fasten with a chain' (cf. prisoners chained to the wall), 'to fasten with a button or buttons' (cf. to button (up) one's coat), 'to fasten with a screw' (cf. to screw a lock on the door)." Jensen (ibid.) provides some examples to prove that using nouns indicating instruments other than those incorporated in the meaning of verbs yields anomalous sentences.20 •She taped the picture to the wall with pushpins. •They chained the entrance off with a rope. •She buttoned up her dress with snaps. •They screwed the lamp to the ceiling with glue. •The artist inked his drawings with crayons. •He stapled the papers together with nails. •He riveted the parts together with nails.

Similarly to the previous group, making the implicit instrument explicit will lead to redundancy, unless the instrument is modified. Cf. (?) They chained the entrance off with a chain. (?) She buttoned up her dress with the buttons. (?) They screwed the lamp to the ceiling with a screw

Semantically, this group is more homogeneous nevertheless the interaction of the metonymic and metaphorical patterns can be observed here as well. The metaphorical use of to chain for example in Too many women feel chained to the kitchen sink

is derivable from the metonymic sense of chain found in the phrase be in chains ('being kept as a prisoner'), which refers to the fact of being imprisoned metonymically. In fact, being chained to the kitchen sink may be conceived as being in chains, that is to say, being kept in prison. However, no metaphorization is expected to occur if the instrument is a specific one, usually expressed by a derived or compound noun. Thus, metaphorization is highly unlikely for the following verbs zero-derived from specific instrument terms: to pitchfork ('to lift or move (sth) with a pitchfork') to sweeper ('to clean the carpet with a carpet-sweeper'), to snowplow ('clear snow from the roads, railways with a snowplow'), to atom-bomb ('to destroy with an A-bomb'). Within this group several thematically more or less homogeneous subgroups can be identified, some of which will be presented briefly below. These subgroups are not

" In Lehrer's above classification they are treated as patient verbs. " Note that in the ALD4 to chain is given the definition 'fasten or confine sb/sth with or as if with a chain', suggesting that it should be assigned to the non-canonical instrument for action pattern. However, definitions in other dictionaries appear to support Jensen's treatment of to chain. " Some of these examples are also taken from Kiparsky (ibid.)

Homonymy vs Polysemy: Conversion in English

225

particularly productive, potential conversions are often pre-empted by other established lexemes or phrases. Generally, it can be stated that vehicle terms in English are easily converted to verbs, unless there are other conventionalized forms, verbs or phrases, that block the process of conversion (cf. Neagu 1997). Thus we can find to bicycle ('to ride on a bicycle'), to bus ('to travel/transport by bus'), to sledge/to sled ('to travel/carry on a sledge/sled'), to truck ('to transport sth in a truck'), but to car and to airplane, to train are pre-empted by to drive and to fly, or by the expressions to go/travel by train or to take a train. Lehrer (1990: 231) observes that to bus is more frequent as a transitive verb (cf. bus children to school), its intransitive use is still rejected by some speakers. The ALD4 provides the following example for the intransitive use: I usually bus (it) to work in the morning.

It is important to observe that similarly to vehicle terms brand names are also frequently turned into colloquial verbs of travelling and sending (cf. Pinker ibid., 61ff). E.g.: We Chevy 'd up and down the Main Street. I FedExed the package last night. Down to their last thirty dollars, they Greyhounded home. Because of his fear of flying he Amtrack'd to New York.

Some other verbs with established meaning also belong here . E.g.: Don't forget to mail these letters. We ship goods to any part of the world.21

Note that in some cases the zero-derived verb drifts away from its parent noun. To taxi for instance refers to the movement of an aircraft on ground or on water before or after flying. In this case the meaning of the instrument as it is understood in the present discussion is suppressed. Cf. The plane was taxiing along the runway.

Since body-part terms are in many ways polysemous, the verbs zero-derived from them also have multiple senses not always readily derivable from the salient functions of the respective body-parts. Roughly speaking, the following verbs can be considered here: to elbow, to hand, to finger, to thumb, to eye, to nose, to head, to mouth, to neck. They usually mean specific actions carried out with the help of the given body parts. Cf. He elbowed (= pushed) me out of the way. She fingered (= touched) the silk to feel its quality. He was thumbing through (= turning over the pages of) the dictionary. He eyed (= watched) me with suspicion. He could head (= strike) a ball faster than he could kick it. (COBUILD) Only yesterday she had been mouthing (= saying) platitudes to the sixth form. (COBUILD) 21

To ship may as well belong to the non-canonical instrument for action pattern, since the act of shipping can be carried out not only by means of a ship. Cf. Fresh supplies were shipped out by lorry.

226

Sändor Martsa

The two of them were necking (= hugging and kissing) on a park bench. The dog nosed (= smelled) out a rat.

The specific actions indicated by the denominal verbs in the above examples are put in brackets. It is interesting to notice that these verbs actually are double metonyms for they are related to the corresponding body-part terms metonymically, instantiating non-standard polysemy, and simultaneously designate specific activities also metonymically, exemplifying standard polysemy. It also has to be mentioned that this behaviour of verbs zeroderived from body-part terms may be attributed to blocking or language economy. For instance, to eye does not merely mean 'to watch' but 'to watch/observe in a specific way', and this is presumably so because of the existence of to watch. For the illustration of the specific metonymic senses of to eye consider the following examples and their paraphrases (note that in the paraphrases the additional stylistic effect of the denominal verb is neutralized): The children were eyeing the sweets. / The children were looking at the sweets with longing. Did you see that creep eyeing up every woman at the party? / Did you see that creep lecherously looking at every woman at the party?

Some verbs that are zero-derived in a regular way from mass nouns indicate various substances used predominantly to cover the surface of some target entity fully or partially. Cf. Rosamund went on buttering her potato pancake. (COBUILD) His little girl was busy colouring in her picture. (COBUILD) She painted the house blue. ... a wall that was poorly plastered. The floor has been sanded smooth. She powders her face. Some women varnish their toe-nails.

Closely related to the above verbs are a few more verbs that are also derived from mass nouns. However, the substances designated by these nouns are used to bring about some chemical or biological change in or of the target entities. The verbs below express exactly this. Cf. Is this tea sugared? Salted peanuts are my favourite. The owner of the shop was accused of watering the beer.

Finally, mention must be made of denominal verbs derived from musical instrument terms. These verbs as a rule indicate actions of performing on musical instruments denoted by the corresponding parent nouns. Cf. He learned to fiddle as a young man. He will drum later this evening. He piped a jig so we could dance.

Similarly to to taxi discussed earlier, to trumpet also seems to drift apart from its parent noun, since it is related to to trumpet metaphorically and not metonymically. Cf.

Homonymy vs Polysemy: Conversion in English

111

He's always trumpeting (= proclaiming loudly and forcibly) his own opinions. The two elephants stretch up their necks and trumpet a celebratory chorus. (COBUILD)

6. Conclusions

Relying on the notions of semantic relatedness and predictability, an attempt has been made in the above discussion to justify the assumption that in noun-verb conversions and particularly in animal verbs and instrumental verbs it is the lexico-semantic relation of polysemy that holds between the parent nouns and the denominal verbs zero-derived from them. Examining various verbs within each of the two groups, it has become evident that the animals and instruments the parent nouns respectively denote, or more precisely the knowledge which speakers possess about real-world animals and instruments and which is encapsulated in the meanings of parent nouns to a large extent determine the nature and the type of the actions that can be associated with them. If this is so, then it comes as no surprise that the meanings of the animal and instrument terms and the meanings of the verbs zero-derived from them are also closely related in the sense that the meanings of the verbs are or can be predictable from the meanings of the parent nouns. I have found that animal verbs in most of the cases can be regarded as word-metaphors, whereas converted instrumental verbs in their primary meanings yield almost exclusively word-metonyms. This in turn explains why the lexico-semantic relation between the nouns and the denominal verbs is treated here as polysemy, rather than homonymy, since in the case of the latter semantic relatedness, at least on the synchronic plane, is out of the question. It has also been demonstrated that the above understanding of polysemy, called here non-standard or non-default polysemy, is not entirely unprecedented. This extension of the scope of polysemy to lexemes belonging to separate word-classes, for instance, appears to comply with the findings of recent cognitive semantic research concerning the genesis of polysemy within idealized cognitive models (ICMs) and between different ICMs. This also entails that, at least in the case of noun-verb conversions discussed here, there is a need for maximizing polysemy at the expense of homonymy (see Taylor 1995: 105). The analysis of linguistic evidence has revealed that the zero-derived metaphorical and metonymic verbs often undergo secondary metaphorization, which can be attributed to the interaction between simultaneous processes of non-standard and standard polysemy in the case of conversion of nouns into verbs.. As we have seen in section 5.2., it is in the case of noncanonical instrument for action pattern that this interaction is the most intensive. Finally, a tentative claim can be made that what has been said about the relation of polysemy with respect to the noun-verb conversions of the types discussed here in all probability can be extended not only to other types of noun-verb conversions (cf. butcher -> to butcher, kennel —> to kennel), but also to word-formation processes of conversion in general.

228

Sändor Martsa

References Adams, Valerie (1973): Introduction to Modern English Word-Formation. English Language Series 7. London: Longman. Aitchison, Jean (1990): Words in the Mind. An Introduction to the Mental Lexicon. Oxford: Basil Blackwell. Bauer, Laurie (1983; 51991): English Word-Formation. Cambridge: Cambridge University Press. Clark, Eve and Herbert Clark (1979): When Nouns Surface As Verbs. Language 55: 767-811. Clark, Eve V. (1982): The Young Word Maker. A Case Study of Innovation in the Child's Lexicon. In: E. Wanner and L. R. Gleitmann (eds.): Language Acquisition: The State of the Art, 390-425. Cambridge: Cambridge University Press. (Quoted in Aitchison, 1990: 156.) Copestake, Ann and Ted Briscoe (1995): Lexical Disambiguation in a Discourse Context. Journal of Semantics 12: 15-67. Cowie, Anthony P. (1982): Polysemy and the Structure of Semantic Fields. Nottingham Linguistic Circular 11: 51-64. Grice, H. Paul (1975): Logic and Conversation. In: P. Cole and J.L. Morgan (eds.) (1975): Syntax and Semantics, 3: Speech Acts. New York and London: Academic Press. Jensen, John T. (1990): Morphology. Word Structure in Generative Grammar. Amsterdam: John Benjamins Publishing Company. Kastovsky, Dieter (1977): Word-Formation, or: at the Crossroads of Morphology, Syntax, Semantics, and the Lexicon. Folia Linguistica 10: 1-33. Kiparsky, Paul (1983): Word-Formation and the Lexicon. In: F. Ingemann (ed.), Proceedings of the 1982 Mid-America Linguistic Conference, 3-22. Lawrence: University of Kansas. Kövecses, Zoltän and Günter Radden (1998): Metonymy: Developing a Cognitive Linguistic View. Cognitive Linguistics 9/1: 37-77. Kruisinga, E. (1931-32): A Handbook of Present-day English. Accidence and Syntax. Groningen: P. Noordhoff. Lakoff, George (1987; 21990): Women, Fire, and Dangerous Things. What Categories Reveal about the Mind. Chicago-London: The University of Chicago Press. Lakoff, George, Mark Johnson (1980): Metaphors We Live By. Chicago-London: Chicago University Press. Lehrer, Adrienne (1990): Polysemy, Conventionality, and the Structure of the Lexicon. Cognitive Linguistics 1-2: 207-246. Lipka, Leonhard (1990): An Outline of English Lexicology. Tübingen: Max Niemeyer Verlag. Lyons, John (1977): Semantics 1-2. Cambridge: Cambridge University Press. Marchand, Hans (1964): A Set of Criteria for the Establishing of Derivational Relationship Between Words Unmarked by Derivational Morphemes. Indogermanische Forschungen 69: 10-19. — (1969): The Categories and Types of Present-day English Word-Formation. Munich: C.H. Beck. Martsa, Sändor (1997): Animal Verbs. In: Märia Kurdi and Jözsef Horväth (eds.): HUSSE Papers 1997. Proceedings of the Third Biennial Conference, 314-322. Pdcs: University Press Pecs. - (1999) On Exploring the Conceptual Structure of Folk Knowledge: The Case of Animal Terms. Linguistics e Filologia, 973-988. Universita Degli Studi Di Bergamo. Miller, George (1978): Semantic Relations among Words. In: George Miller (ed.): Linguistic Theory and Psychological Reality, 60-118. Cambridge: The MIT Press. Neagu, Mariana (1997): Metonymie and Metaphorical Patterns in Regular and Referential Polysemy. (unpublished paper presented at ESSE/4 Congress in Debrecen, Hungary) Panman, Otto (1982): Homonymy and Polysemy. Lingua 58/2: 105-136. Quirk, Randolph; Sidney Greenbaum, Geoffrey Leech, Jan Svartvik (1985): A Comprehensive Grammar of the English Language. London-New York: Longman. Pinker, Steve (1999): Words and Rules. The Ingredients of Language. London: Weidenfeld & Nicolson. Taylor, John R. (1989; 21995): Linguistic Categorization. Prototypes in Linguistic Theory. Oxford: Clarendon Press.

Homonymy vs Polysemy: Conversion in English

229

Warren, Beatrice (1992): Sense Developments. A Contrastive Study of the Development of Slang Senses and Novel Standard Senses in English. Stockholm: Almqvist & Wiksell International. Examples were taken from the following

sources:

ALD4 = OXFORD ADVANCED LEARNER'S DICTIONARY OF CURRENT ENGLISH. Ed. Anthony P. Cowie. Oxford: Oxford University Press 41990. ['1948 Comp. A. S. Hornby]. CIDE = CAMBRIDGE INTERNATIONAL DICTIONARY OF ENGLISH. Ed. Paul Procter. Cambridge: Cambridge University Press 1995. CEGM = COLLINS COBUILD ENGLISH GUIDES 7: METAPHOR. Alice Deignan. London: HarperCollins Publishers 1995. COBUILD = COLLINS COBUILD ENGLISH LANGUAGE DICTIONARY. Ed. John Sinclair. London and Glasgow: Collins 1995. LDELC = LONGMAN DICTIONARY OF ENGLISH LANGUAGE AND CULTURE. Ed. Delia Summers. Longman 21993 ['1992 Ed. Delia Summers] O'Connell, Sue (1993): Focus on Proficiency. Walton-on-Thames/Edinburgh: Nelson. 1993.

Geart van der

Meer

Metaphors: How Do Dictionaries Scramble out of this Morass of Meaning?

1. Introduction

Over the past few years I have published three papers on the treatment of conventional metaphors in dictionaries (Van der Meer 1996, 1997 and 1999). In the first two I tried to show that the tendency, based on frequency considerations, of the main learner's dictionaries to present metaphorical sense definitions before the literal meaning, was misguided because it did not teach the learner unambiguously that - to varying extents, it is true - the metaphorical meaning is derived from the literal meaning (and should therefore logically follow it) and still frequently depends on it for its proper interpretation, as well as showing in the behaviour of its collocates that the literal sense is far from forgotten. In 1997 I wrote that [i]n the case of numerous words the non-figurative sense is still there as a synchronic fact, enriching the figurative use with a "by-way-of-speaking" dimension. This is, in fact, a precondition for the figurative use to be called such at all ("figurative use" always means that there is a "literal" use as well). To mention an example: in CC^ flak is defined as severe criticism only,2 which in effect means that this use has been stripped of its figurative, by-way-of-speaking overtones. To all intents and purposes the same holds for CIDE, which gives flak ("OPPOSITION") and flak ("FIRING OF GUNS") two separate "guidewords", thus creating an impression that the two meanings are semantically unrelated. The CIDE treatment o f f l a k does not differ from that of real homonyms like bud "PLANT PART" and bud "MAN", or homonyms like bear ANIMAL and bear CARRY (also cf. CIDE viii). In OALD, which still recognises a literal meaning, this comes first, whereas LDOCE defines the figurative sense first. Here we have the problem in a nutshell (itself another word where treatment in the four dictionaries differs!): is the literal sense relevant - to the (foreign) learners - and, if so, how should the literal and figurative senses be defined, linked and presented? (Van der Meer 1997: 556-7) The reasons for the present widespread practice are, as already stated, a strong tendency among lexicographers to give prominence to the most frequently occurring sense, which is the sense which the learner is most likely to look up (Rundell 1998). However (Scholfield 1999) vocabulary development and awareness of meaning extension are just as important functions of the learner's dictionary as meaning retrieval. In my opinion, the (genuinely) figurative uses of a specific word cannot be fully understood except by reference to its literal meaning. Treating this basic meaning first sharpens the non-native learner's awareness of what I have called synchronic etymology: "the realisation that meanings may be re-

The abbreviations used here are as follows: 2nd ed. (London, 1995). CIDE = Cambridge International Dictionary of English (Cambridge, 1995). LDOCE = LONGMAN DICTIONARY OF CONTEMPORARY ENGLISH, 3rd ed. (Harlow, 1995). OALD = Oxford Advanced Learner's Dictionary, New Edition, 5th ed. (Oxford, 1995), NODE: The New Oxford Dictionary of English (Oxford, 1998) Obviously, of course, due to a frequency-based ordering principle. C C = COLLINS COBUILD ENGLISH DICTIONARY,

1

232

Geart van der Meer

lated to other, more basic meanings" (Van der Meer 1997: 559). The word defuse, for example, is in CC treated in this way: "1 If you defuse a dangerous or tense situation, you calm it. 2 If someone defuses a bomb, they remove the fuse from it so that it cannot explode" (idem: 562). In my view this should be rewritten (van der Meer 1997: 567) as: DEFUSE: When someone defuses a bomb, they take away or destroy the device (the FUSE) that makes it explode (illustrative example(s)). When someone defuses situations or sentiments that are viewed as bomb-like in that they too are dangerous ("explosive"), such as criticism, disputes, danger, threats, arguments, anger, it means that they take away the immediate cause of the danger (illustrative example(s)). Remarks: here I have tried to make the figurative definition run parallel - both syntactically and lexically - with the literal, so that the reader cannot escape the comparisons (as the basis of the metaphor): "bomb": "situations or sentiments that are viewed as bomb-like in that they too are dangerous ("explosive")" and "the device (the FUSE) that makes it explode": "the immediate cause of the danger". Further, the list of collocates (criticism, disputes, danger, threats, arguments, anger) serves to give some idea of the dangerous situations and sentiments that may be defused. Also note the words "viewed as", which are a signal of the metaphorical character of the nonliteral meaning.

In my 1999 article I returned to this subject, delving deeper into the problem of the phenomenon of metaphor itself. In this paper I used the example of morass to illustrate my point. In the example {someone) has got stuck in a morass of procedure and paperwork dictionaries (cf. Van der Meer 1999) do not mean by morass in this particular context "a dangerous area of soft, wet ground" (LDOCE, sense 2 (!)) but rather something more like "a complicated and confusing situation that is difficult to get out of' (idem, sense 1). However, words like got stuck in fit the literal sense of morass very well, though procedure and paperwork are of course in no way combinable with this literal sense. Even clearer is And what a morass he had found there, that cesspit of her uncle's words, where everyone pushed his fellow down. while trying himself to climb ahead; he might have been left to drown, had not this same Charles Brandon come to his aid.4 It appears that the metaphorically used words are still often combined with collocates that without any further context would be interpreted literally and that would allow morass to be interpreted literally as well (stuck in a morass, left to drown...). These collocates are part of the metaphor and are, I believe, evidence that in the mind of the speaker the metaphorical sense of morass is still strongly associated with, and derived from, its literal sense. This literal sense remains active in the background and, inter alia, attracts and supports the proper collocates (like struggle through, being sucked into, flounder in, follow into, sinking into, falling into, etc. - all used metaphorically). The result is what might perhaps somewhat inaccurately be called "extended metaphors". In the 1999 article I proposed the following model:

5

4

Unless stated otherwise, examples are taken from the free sample of the British National Corpus as made available on the Internet. Interestingly, LDOCE here defines the metaphorical sense of cesspit after the literal: "a place or situation in which people behave immorally". Clearly, this definition does not fit the present example, which may either mean that the word is here used in a more original and less conventional sense or that the putative metaphorical dictionary senses are in principle less stable than is often believed. I will return to this issue below.

Metaphors:

How Do Dictionaries

Scramble

out of this Morass of Meaning?

233

Model of metaphor: METAPHOR

TRIGGER: SEMNTIC

INCOMPATIBILITY OF SOURCE AND TARGET IN THE INTERPRETATION

metaphor inducing context

^




Est. sinine

As the conceptual field of the Russian CUHUÜ is narrower than that of the Estonian sinine, it was necessary to include the word eonyöoü in the Estonian-Russian dictionary to cover the lighter shades of blue. In the Russian-Estonian dictionary we find that the Estonian equivalent to the Russian word zonyöoü is helesinine. Now, as for the equivalents for the Estonian word helesinine, we find the Eng. light blue and bright blue, Fin. vaaleansininen and heleänsininen, Ger. hellblau, Rus. ceermo-cuHuü and zony6oü. The fact that two terms are given for helesinine in the Estonian-English, EstonianFinnish and Estonian-Russian dictionaries might be interpreted by a user in two ways. The semantic field of the Est. helesinine is either broad enough to encompass both of the notions 'light blue' and 'bright blue', or the two equivalents are synonyms the semantic fields of which coincide with that of the Estonian word. The reality is, of course, that the semantic fields of the words expressing the notions of 'light', 'bright' and 'dark' are not alike in different languages. Similar problems emerge when translating other Estonian colour terms with an adjectival attributive component. Estonian helehall and helepruun have been directly translated into English by light grey and light brown. The other compounds consisting of hele- and a basic colour name have been translated by a combination of bright and the basic colour term (e.g. helekollane = bright yellow, heleroheline = bright green). Smaller dictionaries usually do not handle such compounds or phrasal terms. The concepts of dark as well as light have proved to differ semantically in several languages, for example, when Hungarian and American English colour terms are compared (Barratt and Kontra 1996). Additionally, in Estonian there are terms in which the highest degree of whiteness or blackness is emphasised by complementing the usual words for 'black' and 'white' with the adjectival attributes hele 'light, bright' and tume 'dark', respectively. An analogous combination being impossible in many other languages, the Estonian compound helevalge should

Some Semantic Problems in the Translation of Colour Terms

255

be translated by a term for intense white and tumemust by a word (combination) for an intense black tone.

1.2. Red A comparison of the equivalents of 'red' as another basic colour results in an ever more complex diagram. The general Estonian term for 'red' is punane. Dictionaries give its equivalents as the Eng. red, Fin. punainen,

Ger. rot, Rus. Kpacnvtü, ajibiü (see Figure 2).

Dictionaries with Estonian as the target language match the Finnish and German words as well as the Russian Kpacubiü with the Estonian punane. The English-Estonian dictionary, however, gives two equivalents: punane and verev. Actually the latter is a synonym of punane only in South-Estonian dialect, whereas in northern Estonia as well as in standard language the adjective verev is associated with blood (< veri 'blood'). That is why dictionaries with (standard) Estonian as the source language match verev with words meaning 'blood-coloured'. The Estonian equivalents for the Russian anwü are erepunane and helepunane. The Estonian erepunane is matched with such dictionary equivalents as the Fin. kirkkaan punainen, Ger. grellrot, Rus. Άρκοκραοπαύ, anbiü. The Russian apKOKpacHbiü is supplied with as many as fifteen Estonian synonyms. Translating all those, in turn, we arrive at even more specific terms. The Estonian tulipunane, for example, has been translated into English as fiery-red, fire-coloured and red-hot. When consulting dictionaries for equivalents of the Estonian helepunane we find the English bright red, scarlet, vermilion and the Finnish vaaleanpunainen, heleänpunainen. Again, the English scarlet (Est. sarlakpunane) and vermilion (Est. kinaverpunane) seem to refer to more specific colours. The Finnish vaaleanpunainen has taken roosa (Eng. ρ ink, rose, rosy, Ger. rosa(farberi)\ Rus. po3oebiu) as the Estonian equivalent. A literal Estonian translation of the components of vaaleanpunainen would be 'hele' (Eng. 'bright, clear, light') and 'punane' (Eng. 'red'). Yet, as the semantic field of the Estonian punane is narrower than that of the Finnish word punainen, the lighter (more whitish) tones of those fields are not quite equal notions either.

256

Vilja Oja

Figure 2

Eng. red. Eng blood-red

> Est. veripunane

Fin. helakan punainen

Fin. punainen

veripunainen,

Ger. rot

Rus. KpacHbiü anbiü

Ger. blutrot

helepunane

S

Eng. bright red

Fin. kirkkaan punainen 4

Rus. npKOKpacHbiü ajibiu 15 synonyms,

Λ

Rus. κροβαβο KpacHbiü

Est. erepunane,

Ger. grellrot

-ruskea

scarlet

Est. sarlakpunane,

vermilion

helepunane Est. kinaverpunane

Fin. vaalean punainen

Est. roosa

heleän punainen

e. g. Est. tulipunan

ng. pink, rose, rosy Eng. fiery-red, fire-coloured, red-hot

Fin.

vaaleanpunainen,

Fin. tulipunainen Ger. feuerrot

Ger. rosa, -färben

Rus. KpacHbiü

OBMHHOZO

ifeema,

Rus. ρ030βblÜ

nymfoebiü Although the Estonian helepunane and veripunane, as well as tulipunane, sarlakpunane, and other words for more specific shades all denote a red colour, they are not full synonyms to the general word punane. If we drew a circle to illustrate the semantic field of the word punane, this huge circle would encompass all (over 200!) terms ever used for 'red' in Estonian, yet none of them would cover the whole circle. Consequently, the word punane is a hyponym for them all. The word helepunane, in turn, is a hyponym for the words sarlakpunane, veripunane, etc.

1.3. Purple, violet, lilac Finding an English equivalent of the words marking a mixed tone of blue and red (Est. lilla) seems to be complicated for Estonians as well as for some other Finno-Ugric nations (e.g. Fin. sinipunainen, Hun. lila) (see Koski 1983: 228-230; Barratt and Kontra 1996). As standard Estonian employs two parallel terms lilla and violett ~ violetne, there have been attempts to keep them semantically differentiated (see Saareste: 584). Both loanwords seem

Some Semantic Problems in the Translation of Colour Terms

257

to have been naturalised as general terms for mixed tones of blue and red (although Ulla is the older one of the two and fits the Estonian system better). Bilingual dictionaries treat them as follows (see Figures 3 and 4). Figure 3

Est. lilla Est. sirelililla, sirelikarva Eng. (hele-) lilac (kahvatu-) mauve (punakas-) purple (tume-) violet Fin. sinipunainen, sinipunerva, lila, violetti

Est. kahvatulilla -> Est. purpurne, punakasviolett, tume punakaslilla > Est. sinipunane Est. violetne Est. lilla

Ger. lila lilafarben, -farbig fliederfarben

Est. lillavärviline, lillat värvi

Rus. nunoebiü

Figure 4

Eng. violet, purple

Est. purpurne punakasviolett tume punakaslilla

Fin. violetti, sinipunainen Ger. violett Rus. (puojiemoebiü, minoeuü The colour of the blossoms of lilac and violets have motivated such compound terms as sirelikarva, sirelililla (< sirel 'lilac') and the rare word kannikesesinine, dial, kannikasinine (< kannike 'violet'). Consequently, the Estonian sirelililla is indeed a correct equivalent of the English adjective lilac.

258

Vilja Oja

The English word mauve (< Fre. mauve) has been translated into Estonian as kahvatulilla (< adj. kahvatu 'pale')· As colour words with such a structure (kahvatu + a colour adj.) are not often found among dictionary headwords, it was entirely legitimate to include it in the article of the generic term lilla together with an explanation (see Saagpakk). However, the compound kahvatulilla is not too informative as to the actual shade. The colour of the Estonian species of the mallow family (Lat. Malva) have been described as roosa 'pink', valkjasroosa 'whitish pink', lillakas 'purplish', valge 'white', roosakasvalge 'pinkish white', and sinakasvalge 'bluish white' (Kuusk et al. 1972). The plant has not lent its Estonian name kassinaeris to any colour. It is likely that the French or English mauve as well as the German malvenfarbig originate from the colour of the larger blossoms of some southern European subspecies. As for the word purpur 'purple', it has most likely found its way into Estonian as well as into Finnish from the Bible. There is no agreement among users as to what the words purpurpunane, purpurivärviline, and purpurne really mean. The Defining Dictionary of Literary Estonian describes it as lillakaspunane 'purplish red' (EKSS). Not one of our dialectinformants has suggested lilla (or violetne) as a full synonym of colour terms with the stem purpur-. In most cases it is believed to be a specific shade of red or even bright red.

2. Principles of defining the meaning of the words in the Database of Estonian colour terms

In our database the semantic information is divided between two fields. The field Meaning (M) contains the explanation of the meaning of the term and the field Object (O) carries information on its sphere of usage. The information for those two fields has been extracted from the word explanations and case examples available in the files of the original speech material. The field Μ contains the following data: 1) standard colour terms that have been mentioned as full or partial synonyms of the word in question (beginning from the closest synonym). Partial synonyms are presented in brackets as derivatives with a modifying suffix (e.g. lilla M: lilla, violetne \punakas 'reddish' sinakas 'bluish']). 2) other features of the colour reflected in the term (brightness / darkness, richness etc.). The most general terms without any additional nuances are labelled "neutr." (neutral) (e.g. helesinine M: sinine 'blue1, hele 'light'; neutr.) 3) etymological and other explanatory notes (e.g. sirelikarva M: lilla sireli värvi, hele 'light' < sirel 'lilac') If the colour term is derived from the way in which the colouring agent is produced, its package, the dyeing process ,etc., a corresponding explanation is entered in the field M, e.g. kaselehe/kollane M: helekollane 'light yellow' (rohekas 'greenish') - dyed with young birch leaves; poti/sinine M: indigo sinine 'indigo blue' - dye made of copper vitriol < pott 'the pot in which the liquid component of the colour solution was collected and prepared'. The objects which the colour word applies to are recorded in the field O. In the case of a large number of examples, more general terms (class names) are used (e.g. animals, tex-

Some Semantic Problems in the Translation of Colour Terms

259

tiles). Very diverse material is generalised under the label "generic". Examples of the usage are available as a special field. The database contains additional information on the dialectal distribution of each colour word and on its interpretation according to some dictionaries of standard or dialectal Estonian.

3. Conclusion

Consequently, while preparing a database the task is to differentiate between the full and partial synonyms of the colour term, adding specifying explanations. It would be advisable to provide an etymological reference. In addition, the range of application of the term should be pointed out. The user of a printed dictionary would appreciate the provision of a colour term with several equivalents. However, partial equivalents should be differentiated from full synonyms. Any explanatory notes and references are no doubt welcome, as most of the users can hardly be expected to have a linguistic background. As for a bilingual electronic dictionary, it should initially present the most appropriate equivalent found in the target language and the user should have an option to view a longer explanation, secondary equivalents, or a reference list of all partially synonymous headwords as well as examples of the usage of the colour term. Due to its capacity and its multitude of user options an electronic dictionary can be easier to use and provide more information than a printed one.

References

(a) Cited Dictionaries E D W = ESTNISCH-DEUTSCHES WÖRTERBUCH/ EESTI-SAKSΑ SÖNARAAMAT. Comp. Κ. Kann; Ε. Kibbermann;

F. Kibbermann; S. Kirotar. 4., verbesserte und veränderte Auflage. Tallinn: Valgus 4 1987 ('1964). E E D = ENGLISH-ESTONIAN DICTIONARY. Tallinn: Kinex 1992.

EKSS = EESTI KIRJAKEELE SELETUSSÖNARAAMAT I-V. Eesti NSV Teaduste Akadeemia. Keele ja Kirjanduse Instituut. Tallinn 1988-1999. EMS = EESTI MURRETE SÖNARAAMAT I-II. Eesti Teaduste Akadeemia. Eesti Keele Instituut. Tallinn 1994-1999. ESS = EESTI-SOOME SÖNARAAMAT/EESTILÄIS-SUOMALAINEN SANAKIRJA. Comp. Paul Kokla, Helga Laanpere, Mart Mäger, Arno Pikamäe. (Ed. Valdek Pall) Tallinn: Valgus 1972. EW = EUROWORD. Espoo: Eurokustannus Oy 1995. MWT = MULTI WORD TRANSLATOR. (English, German, French, Spanish, Italian, Estonian, Swedish, Finnish.) Electronic dictionary. Logicom Media 1997. Parts, Clarisa. EESTI-INGLISE SÖNARAAMAT KOOLIDELE./ESTONIAN-ENGLISH DICTIONARY FOR SCHOOLS.

Tallinn: Valgus 1969.

260

Vilja Oja

Rauk, Melanie. INGLISE-EESTI SÖNARAAMAT SCHOOLS. Tallinn: Valgus 1966. RFS V.

=

PYCCKO-OHHCKHH

Ollikainen)

KOOLIDELE./ENGLISH-ESTONIAN

CJIOBAPB/VENÄLÄIS-SUOMALAINEN

SANAKIRJA.

DICTIONARY

(Ed.

M.

FOR

Kuusinen,

AKANEMHS HayK C C C P . KapejibCKHÖ ΦΗΛΗ&ΓΙ. ΗΗΤ j n u i c a , jiHTepaTypti Η

HCTopwH. M o c K e a : T o c y a a p c T B e H H o e H3,naTejibCTBo HHOCTpaHHbix η HauHOHajibHbix c n o e a p e f t

1963. Saagpakk, Paul F. EESTI-INGLISE SÖNARAAMAT/ESTONIAN-ENGLISH DICTIONARY. Tallinn: Koolibri 2 1992 ('1982). Saareste, Andrus. EESTI KEELE MÖISTELINE SÖNARAAMAT/DICTIONNAIRE ANALOGIQUE DE LA LANGUE ESTONIENNE, II. Eesti Teadusliku Seltsi Rootsis väljaanne nr. 3. Stockholm: Vaba Eesti 1959. S E S = SAKSA-EESTI SÖNARAAMAT/DEUTSCH-ESTNISCHES WÖRTERBUCH. Comp. Ε. Kibbermann; S. Kirotar; P. Koppel. Tallinn: Valgus 1975. Silvet, Johannes. INGLISE-EESTI SÖNARAAMAT/AN ENGLISH-ESTONIAN DICTIONARY, I - I I (Third revised and enlarged impression.) Tallinn: Valgus 3 1989-1990 ['1939], SoES = SOOME-EESTI SÖNARAAMAT/SUOMALAIS-EESTILÄINEN SANAKIRJA. Comp. Kalju Pihel, Arno Pikamäe. (Ed. Paul Alvre) Tallinn: Valgus 1971. Tamm, Johan. EESTI-VENE SÖNARAAMAT/3CTOHCKO-RYCCKHK CJiOBAPb. Tallinn: Valgus S 1 9 8 1 ['1955],

VES = VENE-EESTI SÖNARAAMAT/PYCCKO-3CTOHCKHH CJiOBAPt. I-IV. (Ed. Helle Leemets, et al.) Eesti NSV Teaduste Akadeemia. Keele ja Kiijanduse Instituut. Tallinn: Valgus 1984-1994. Wiedemann, F. J. EESTI-SAKSA SÖNARAAMAT/ESTNISCH-DEUTSCHES WÖRTERBUCH (revised by Jakob Hurt.) Tartu: Eesti Kirjanduse Seltsi kirjastus 31923 ['1869].

(b) Other Literature Barratt, Leslie B.; Kontra, Miclös (1996): Matching Hungarian and English color terms. International Journal of Lexicography. Vol. 9, 2, pp 102-117. Koski, Mauno (1983): Värien nimitykset suomessa ja lähisukukielissä. Suomalaisen Kirjallisuuden Seuran toimituksia 391. Savonlinna. Kuusk, Vilma (comp.), et al. (1972): Taimede välimääraja. Käsiraamat körgemate taimede tundmaöppimiseks. Tallinn: Valgus. Oja, Vilja (1998): Database and computer analysis of the Estonian colour adjectives. Caron, Bernard (6d.) Actes du Congres International des Linguistes (Paris 20-25 juillet 1997). Oxford: Elsevier Sciences (CD-ROM).

Kurt Opitz The Dictionary of Connotations: A Viable Proposition?

1. Introduction

The non-term dictionary has traditionally been used to designate a dazzling array of concepts and implementations that have little more in common than a vague notion of 'collected information' on any particular area of knowledge or experience, usually presented in alphabetical order of key words. Its connection to the language field has occasionally been a tenuous one at best and restricted to the medial role of language in the delineation of diverse systems of concrete and abstract realia\ there have thus been compiled dictionaries of art and music, biography and film, of classical reference in English poetry and of metaphors, not to mention the vast area of technical specializations. Why not, then, a dictionary of connotations as well, bearing upon a crucial area of language behaviour? The idea of such a dictionary must surely have been about for some time; it is all but inevitable in the context of modern semantic theory which holds that the meaning of words is composed of a generally shared part - their denotation - and a less precise, individually or socially determined aura of association and significance: their connotation. Language dictionaries ordinarily aim at denotation in their attempt to present lexemes as elements of the abstract language system; as soon as those lexemes are applied to a language act and turn into constituent parts of parole, their denotational purity becomes contaminated, as it were, by circumstantial conditions: external influences, mental state and emotional disposition of the speaking individual on the one hand, as well as corresponding forces in any addressee. Connotation, as Georges Kassai who suggested the idea of making it accessible in dictionaries for translation points out, is one of the consequences of the instability of the linguistic sign which in turn is caused by the varying needs of its users (Kassai 1992).

2. Language vs dictionary: a conflict of interests?

The question that immediately needs to be addressed is, how can properties of linguistic signs that do not enjoy universal validity - which is to say, that are not generally shared and hence cannot be considered to possess a clearly circumscribed area of signification - how can such properties be gathered and glossed within the framework of a published dictionary? Our general understanding of dictionaries tends to require the very opposite; we expect glosses with the utmost degree of reliability which presupposes near-universal validity within the adopted linguistic scope and a reasonable immunity from chronological variation. And, of course, we assume dictionaries to supply information based on - somehow proven facts. To lexicographers, such expectations are commonplace demands, accepted

262

Kurt Opitz

and confirmed with a straight face and, for the most part, in good faith. On closer inspection, however, we may find such assurances to be aim rather than fact. Intentionally or not, dictionaries are often ambiguous about meaning, hedging rather than straightforward and, generally, likely to be committed to an approach more in keeping with obeisance to connotation than with the strict observance of the denotative principle. Illustrative examples of that approach can usually be found in the treatment of supposed synonymy and in the way style markers and other tags are handled in specific dictionaries, although their compilers would probably disclaim any such intention. The admission of so-called synonyms, which leads to multiple equivalents for single headwords, is common lexicographic practice and represents the unacknowledged application of the principle of connotation. It might indeed be argued that, from the point of view of phonology, polysemy and connotation are mutually interdependent concepts, although direct influence may be difficult to prove in a practical way in the interpretation of different lemmata of the same lexeme. Yet there can be no doubt at all about the contextual limitations governing different translations of the same headword, which represents one of the greatest challenges to lexicography and one of the chief pitfalls for the unsuspecting dictionary user whose ignorance or neglect of the connotational dimension of supposed 'equivalents' is forever causing confusion and exasperation with the dictionary as a worthwhile instrument in general. If 'equivalence' is a conceptual trap, so is the designation of synonym - a misleading term if there ever was one: although perhaps being used to describe the same concept, synonyms of an identical phenomenon do so from decidedly different vantage points, thus yielding starkly different perspectives upon one and the same object; this is also the reason why synonyms are in most cases non-interchangeable. It is preferable that lexicographers should indicate and clarify such semantic complexities rather than hide them under a scheme of deceptive simplicity. What is a matter of course in polytechnical dictionaries treating polysemous entries, namely, the practice of tagging, could be consistently applied as a minimal step towards addressing connotation without too much trouble in a good many cases in other dictionary types as well. Once it has become common procedure, the foundations for a proper attitude towards connotation and for its effective and meaningful handling have been laid: the recognition will have taken hold that connotation is a necessary and universal condition of language use, subject only to the same rules and restrictions as any of its other elements. If connotation has thus far not been recognized in that manner and accorded its proper place in the dictionary article, such neglect may not least be due to the mistaken notion of systemic purity believed to be necessary for maintaining the semantic charge of headwords quasi-abstract, defined and definable as only denotation promises to deliver it. Such reasoning must be countered by the reminder that even in a language dictionary, it is realia that are in the end described and classified. Applied to our present context this means that existing connotations need to be identified in terms of their actual semantic roles in real language activity, so as to be decoded in an unambiguous and functional manner. Yet it cannot be denied that the conventional dictionary and actual language - parole are in a fundamental sense at cross purposes: while dictionaries are designed to supply universally valid semantic information, language users are interested in the ad hoc signification of words. In order to gain access to the latter, they therefore rely instinctively on the connotative dimension of speech. Dictionaries, on the other hand, have no use for instinct; they adopt the pigeonhole approach of referral to fixed, science-oriented perceptions, dis-

The Dictionary of Connotations: A Viable Proposition?

263

claiming anything that is dynamic in an uncontrolled spontaneous way. A dictionary willing and able to consider the vital but unruly dimension of connotation must on the contrary find a grid of categorization suitable to filter precisely that semantic charge of its contents.

3. Categories of ad-hoc signification: a realistic idea?

Let us consider the problem by looking at the word garden. As it does in English, it transports in most European languages, beside the more or less technical sense, "piece of ground devoted to growing flowers, fruit, or vegetables," such affective connotations as "protection," "security," "man-made," "beauty", etc., depending on the speakers and the given context. None of the latter are immediately revealed by the wording of the dictionary article quoted above (COD), which goes on to list quite a number of lemmata that are in turn described denotationally: e.g. as "specially fertile region"; "(pi. with name prefixed as Onslow, Spring, C~s) set of houses in street, square, etc." " ... (attrib.; with or without hyphen) cultivated, not wild (~ plants, ~ cress; common or si. ordinary)". If we compare this with the more recent LDCE in the hope of finding a disposition more sympathetic to connotation, the result is disappointing: in its simplified microstructure and metalanguage virtually the same information comes across, if possible in a still more abstract, technically "clean" way that in no manner suggests an affective dimension to our headword - this despite the fact that the dictionary is intended for learners of English as a second language whose need for help in precisely such matters of semantic discrimination is a documented fact. A poor record of past performance is no reason, however, to consider attention to connotation as alien to lexicography. The real question is rather one of feasibility. To begin with: how can specific connotations be transformed into functional headwords? And what would be the size of a dictionary devoted entirely to connotations? Obviously the major part of the general lexicon will fall outside its useful range, leaving as key components to be treated the class of nouns, plus a certain number of adjectives, but also including various types of proper names (historical, biographical, literary, geographical, etc.) that are frequently not listed in the average language dictionary. Their inclusion will not turn the dictionary of connotations into an encyclopedia; the information it provides will be restricted to the semantic charge of its entries, and for any background information about the origin and significance of references and allusions users will have to turn elsewhere. The seeming inconvenience of this procedure loses some of its weight when we realize that the dictionary will be consulted by fairly sophisticated users on very specific points posing a problem to translators and interpreters, literary scholars, journalists - people who are generally familiar with the ways of gaining access to non-routine information. A body of users with such characteristics facilitates a reasonably economical design of the dictionary's microstructure. Basically, each headword would be followed by a number of single-term entries, each representing a particular connotation but not necessarily expressed through the same word class. Whether those entries should - or even could - be tagged so as to associate them with certain contextual areas (which would have to be very

264

Kurt Opitz

carefully delineated), is a question that can find its answer only through experience; it may be wise to leave users to make their own judicious choice ad hoc, based on the text sample at hand. What appears to be useful and feasible, however, is a scheme by which the relative currency and/or frequency of each connotation would be indicated. This could be achieved by different letter type for different frequency brackets, similar to the method used to list remedies in homeopathic repertories. What would be more difficult to achieve, and to spell out, are objective criteria of classification - provisionally, one might think of a simple listing of connotations in alphabetical order. In addition, a positive (+) or negative (-) mark could be used to characterize the general valuatory drift of headwords if this can be reliably established on the basis of social consensus about a given concept or phenomenon. The potential difficulty of any such attribution over a longer period of time results from the notoriously unstable social substratum of all language and, hence, "the perpetual need to update (it) in the fluctuating rhythm of events and fashions," as Kassai notes in his paper; words like liberal, profit, synthetic illustrate well the effect of time and place upon connotation and serve to underline its ad hoc validity.

4. From theory into practice

Having considered the various theoretical and practical arguments connected with the project of a dictionary of connotations, let us see what it would look like, and how it would differ from existing language dictionaries. Here is the potential article for the headword garden: GARDEN(+),

beauty, innocence, man-made, protection

Example for a proper name: CROMWELL(-)

civil war, religious fanaticism, cruelty, self-righteousness

Obviously the list of existing connotations for any one headword must be considered as essentially open and subject to differing individual sensibilities; their number must therefore be curtailed in the dictionary without, however, being reduced to a cursory standard, which would create havoc in the weighting system and vitiate the very purpose of the whole undertaking. In the course of this discussion I have repeatedly referred to certain problems or difficulties of a lexicographic nature. One question that remains to be asked is whether a dictionary of connotations can be commercially feasible. To find an answer we have to take two aspects into account. As far as its potential market is concerned, the present and future outlook for transnational information and communication should be reassuring. Peoples and their cultures are going to be forced into ever closer contact, which in turn will require mutual tolerance born of insight and understanding; for that reason dictionaries, and those of the hidden messages of language in particular, can only be expected to grow in importance. Thus demand seems to be certain. The second aspect concerns production costs; for

The Dictionary of Connotations: A Viable Proposition?

265

while the finished dictionary - no matter whether on paper, CD, or online - has been shown to be possible in a very modest format, its compilation may prove immune to most of the usual modern labour-saving techniques. Unlike denotation-based dictionaries, a dictionary of connotations cannot be automatically compiled on the basis of an existing corpus of text without undergoing a highly complex, and therefore highly unlikely, additional data-processing step designed to put the implicit connotations of specific words into explicit verbalized form for electronic handling. This means that most of the work connected with the preparation of the dictionary's contents will probably fall upon the human lexicographer, from the original choice of entries to the final decision on connotations to be included and how to rank, or even tag, them properly. The role of the computer in this process will, in other words, be strictly that of a mechanical tool easing access to, and handling of, material that has been selected and analysed beforehand by qualified live agents. Will a dictionary of connotations be worth the human effort and the money spent on it? On balance, I believe it would, always considering it as a project in specialized lexicography which does not intend to compete with standard language dictionaries. As a novel supplement, it would fill a conspicuous and regrettable gap in the line of available language tools that has existed all too long.

References

Kassai, Georges (1992): Pour un dictionnaire des connotations. Prague, 284-287. COD = The Concise Oxford Dictionary of Current English, 4th ed. Oxford, 1951. LDCE = Longman Dictionary of Contemporary English, 3rd ed. 1995.

Yoshiaki Otani Who Uses English-Japanese Dictionaries and When? Their Bidirectional Working

1. Peculiarities of English-Japanese Dictionaries (E-Js)

The answer to the title of this paper may seem obvious. Such dictionaries are both for Japanese who want to study English and for those speakers of English who want to study Japanese, just as English-Danish dictionaries are both for Danish people who want to study English and for speakers of English who want to study Danish. However, there is something unique about E-Js. English-speaking people who use E-Js are actually very few, since the Japanese entries in them are written in a mixture of Chinese characters and Japanese syllabic script, kana, and not in the Roman alphabet. Only advanced learners of Japanese can read them, and since there are not so many such learners, E-Js are virtually only of use to Japanese people. To illustrate this point, I will show how the lst-person pronoun Τ is entered in E-Js. Generally, the grammatical information 'lst-person pronoun, subjective' comes first, and then one or two representative Japanese equivalents for Τ are given: watashi, boku, etc. The problem here is that in Japanese the subject is not syntactically obligatory, and the pronoun indicating 1st person is especially prone to be absent, since the identity of the speaker is generally obvious in the situation. For example: I will do it. (watashi ga yarimashou). (ga is a particle attached to the noun/pronoun in the subjective case) Now, since the subject is not obligatory in Japanese, watashi is expressed only when it is necessary to make the meaning of the sentence clearer, as when the speaker implies Ί, in contrast to you'. When watashi is not expressed, Τ has a zero equivalent. Another function of the first-person pronoun is to show the stance the speaker takes toward the hearer, and accordingly there are a number of equivalents in this case: watakushi (dated, polite), watashi (neutral and representative), atashi (female), ore (macho), boku (male, familiar), wagahai (male, pedantic), temae (male, humble or vulgar), etc., a varied group. A Japanese is familiar with all these variations, and the representative watashi is enough as an entry here. However, in a self-contained dictionary entry, this equivalent alone gives only partial information and is not user-friendly to foreign learners of Japanese who might use an E-J.

268

Yoshiaki Otani

2. Nature of Equivalence in E-Js

'I' is a special case and, besides, it is a function word and concerns grammar more than meaning. Dictionaries serve primarily as reference books on meanings of entry words, especially the meanings of content words, such as nouns, verbs, adjectives and adverbs, which will be the main concern of this paper. In a bilingual dictionary, word meanings are usually given in the form of equivalents, which are affected by the relation between the two languages. For example, Danish and English belong to the same language family, and so there are a good number of words which are quite similar in both languages, such as system-system, centrum-center, lillelittle, etc., and therefore an English-Danish dictionary can be more compact than an E-J. On the other hand, between English and Japanese there is no such kinship; the two languages are basically different in syntactic structure, concept and background culture. Therefore word-for-word equivalence is not always easy. English words have, so to speak, to pass through syntactic, collocational or cultural filters before they find suitable partners in Japanese. Of course, some words can surmount equivalence barriers more easily than others, but as I have demonstrated in the previous section, the barriers between English and Japanese are not very low. Still, bilingualism is the current of the times, and various kinds of E-Js are published and sold in large numbers. At this point it would be relevant to consider the nature of equivalence in E-Js.

3. Diversity of Equivalents

3.1. Concrete Words Among words, there are some which are more independent of syntactic or cultural difference than others. For example, natural elements such as 'rain' ame and 'sea' umi are cases in point. However, there are some whose equation is not exact: for example, the distinction between 'fog', 'mist', and 'haze' does not exactly correspond to that between kiri, moya, and kasumi. Bodily parts are common to all humans, and there is no problem about such pairs as 'nose' hana, but while 'neck' is kubi, kubi is 'neck & head'. Demarcations and appellations are sometimes specific to a culture or language. Products of modern scientific technology, such as information technology (IT), computers, and jet planes are some of the examples which are almost the same by whatever name they are called. Due to standardization and globalization, the number of such words is on the increase. However, some concrete words are similar in meaning, but not exactly the same. For example, 'bath' is huro in Japanese, but in English-speaking culture a bath is taken in an individual bathtub and the water is washed away by each bather, while huro is used by the members of a family in turns without changing the water. Of course, in this international

Who Uses English-Japanese Dictionaries and When?

269

age, each meaning can include the other, but the semantic range of the two words is not entirely identical.

3.2. Words of Abstract Nature First, let me consider 'idea' as an example. The following definitions are from two English monolingual dictionaries (EMDs). "any content of the mind, esp. the conscious mind." CED "something one thinks, knows, or imagines; mental conception or image; notion." WNWD

The workings of the mind are universal to all humans, but their representation in words varies from language to language, and we have a scheme of metaphysical concepts based on our experience in Japanese. The definitions in EMDs alone are not sufficient for conveying the sense of the word unambiguously to us. We need to refer 'idea' to the mental paradigm set in Japanese by such equivalents as kangae, kannen, and omoitsuki. We sometimes use the loan word ai-de-a, but even this derives its meaning from the position it occupies in the scheme of Japanese concepts. However, equivalents of abstract words tend to vary according to collocation and context, and E-Js generally give a number of them to choose from. Then there is the problem of how detailed they should be in E-Js. In the case of 'idea', kangae is fairly representative and fits in with many collocations, but even if kangae represents the core meaning of 'idea', it would hardly do for the user to expand on kangae on his/her own: hence a number of other equivalents are usually given. It is one of the hardest tasks for the editor to choose the most basic and suitable equivalents. Then there is a group of adjectives indicating middling quality, such as fair, ordinary, average, moderate, etc. They have a common concept of middling, each with fine shades of meaning of their own. They are cross-referenced to each other in EMDs, which somehow help define their relations among themselves, but these adjectives have delicate rhetorical implications, and perhaps even native speakers make the subtle distinction, because they have been exposed to these adjectives in context for years. In Japanese, too, adjectives of this kind form a cluster of similar words whose meanings do differ slightly, but it is not easy to clearly distinguish their subtle variation in meaning. So this is a case of group equivalence. The approximate concept of the group is not difficult to grasp. E-Js generally give nami-no (= average-neutral, sometimes derogatory) and mazumazu-no (= so-so) common to each, plus some other equivalents which are meant to distinguish them from each other. But even if we read the entries in E-Js very carefully, we do not feel we can tell one from the other clearly. It will take thorough exposure to examples in context to get the feel of these English adjectives. I will give one more example, the verb 'reduce'. Verbs are usually more involved in syntactic structure, and 'reduce' is no exception. I will illustrate this point by comparing the quotation of a definition of the verb from Random House Unabridged Dictionary and the corresponding one from Shogakukan Random House E-J Dictionary. It would be too long if I quote the whole entry from the latter, so I will only give some of the illustrative examples here.

270

Yoshiaki Otani

reduce:"6. to bring to a certain state, condition, arrangement, etc.: to reduce glass to powder." Random House Unabridged. :"5. [...] reduce grapes to wine [...] be reduced to an exact form [...] be reduced to desperate measure [...] reduce glass to powder [...]."(12 examples in all, each with Japanese translation). Shogakukan Random House E-J.

In the American edition, there is only one example, while in the Japanese edition there are 12 examples, each with Japanese translation. This shows that, in the case of verbs, wordfor-word equivalence is often difficult or insufficient, and the equivalents must be given in collocation and context. The examples cited above are minuscule, but they give a hint as to how diverse actual equivalents can be.

4. Bidirectional W o r k i n g of Equivalents

4.1. Initiators into the World of English As we have seen above, equivalents are diverse in nature, but there is one element common to all: almost all of them are more or less approximations. This is inevitable, but then of how much use are they? If we are in an English-speaking environment, the language is directly related to what we experience there. But most Japanese learn English in a non-English-speaking environment, and accordingly English is merely a system of symbols or signs for us. Thus it is equivalents that mediate between English and Japanese, and substantiates English words for us. Let us take an example: "The boy runs fast." When we consult an E-J and learn that 'boy' is shounen, 'run' is hashiru and 'fast' is hayaku, the above sentence makes sense with the help of a knowledge of English syntax and the use of the definite article. The three English words are no longer mere symbols, but are related, through the equivalents, to what we actually experience. A language is an autonomous system of linguistic symbols or signs, each complete in itself. However, it does not exist in a void but represents the real world and the real world is projected into the language. So an ordinary grown-up Japanese is in possession of a framework of knowledge, reasoning and perception molded in his vernacular, Japanese. We may call this a frame of reference, paradigm, or matrix, and it is to this framework that we refer the meanings of English words through equivalents. With their help, the learner is initiated into the world of English semantically. However, as I have stated, equivalents are in most cases approximate. In the above example, I equated 'boy' with shounen, but actually, in most E-Js, the first equivalent is otokono-ko (= male child or adolescent). However, otoko-no-ko is a colloquial expression, and I, for one, hesitate to use it as a representative equivalent. Shounen may mean simply a young person of both sexes, but in everyday use shounen is a boy and shoujo is a girl. This example shows how rare exact equivalents are.

Who Uses English-Japanese Dictionaries and When?

271

Still, as a learner's vocabulary grows larger, his/her world of English will gain wider scope and greater depth, and he/she will become less dependent on equivalents and increasingly move into the semantic network of English. Gradually, EMDs will become as useful as E-Js to him/her. Equivalents have served as initiators into the world of English.

4.2. Frame of Reference for the Reception of English If the learner renounces the world of Japanese and remains in the world of English for good, he/she will have no more use for E-Js. But, as a matter of fact, most of us have no intention of going over to the other side never to return, but only make temporary forays into the world of English from the vantage point of our home and base, with the intention of bringing our acquisition back home. Seen in this light, the world of English we acquire is a superstructure laid on the foundation of our vernacular. The foundation supports the superstructure, while foreign elements seep down from above into the foundation, making it broader and more cosmopolitan. The process is reciprocal. In this sense, E-Js work bidirectionally. First they help us to be initiated into the world of the target language, and then they serve as an objective and comprehensive reference book for the transfer of English into our own world. In today's Japan, learning English is an important accomplishment as well as a skill for international communication. Indeed, a very large proportion of the Japanese population are more or less involved in bilingualism between English and Japanese. They need reliable, comprehensive E-Js, in which English is made to correspond with Japanese through the whole gamut of the two languages, and which provide us with extensive and objective information an individual can hardly hope to amass on his/her own. Of course, E-Js are not a Japanese learner's sole recourse when he/she learns English as a foreign language. He/she can make up for the deficiencies of E-Js by using EMDs and other reference books, have direct experience in an English-speaking environment, etc., and his/her knowledge and intelligent judgement will help the learner to assimilate the information found in E-Js. But, as I have said above, the role E-Js play in the study of English is of vital importance to learners in Japan.

5. E-Js, their Present and Future

A perfect E-J is a contradiction in terms, but, within their intrinsic, limits, E-Js are at a fairly advanced level now, as they have a history of well over a century behind them, and the feedback from users has helped to refine them a great deal. But this does not mean there is no room for improvement. For one thing, we can try to make equivalents more adequate by seeking them in as many different contexts or collocations as possible, making use of the large corpora now available in both English and Japanese.

272

Yoshiaki Otani

Besides, E-Js need endless revision since new words and senses evolve constantly and both English and Japanese keep changing independently of each other. Their relation is never static but dynamic. But the trouble is that it requires a great deal of money and time to compile or revise a large, authentic E-J, which, alas, is unlikely to sell in great numbers. This is not a very attractive commercial enterprise for publishers, who prefer to make profit by publishing handy pocket E-Js. However, reliable and comprehensive E-Js are one of the most important mainstays of bilingualism between English and Japanese. Bilingualism is demanding at the individual level, and good E-Js can make the learner's burden much less heavy than it would be otherwise.

References

CED = Collins English Dictionary. Ed. Marian Makins. Glasgow: HarperCollins Publishers 1994. WNWD = Webster's New World Dictionary. Ed. Victoria Neufeldt. New York: Webster's New World 1988. Random House Unabridged = Random House Unabridged Dictionary, 2nd Edition. Ed. Stuart Berg Flexner. New York: Random House 1993. Shogakukan Random House English-Japanese Dictionary, 2nd Edition. Ed. Tomoshichi Konishi, et al. Tokyo: Shogakukan 1994.

Gunnar Persson From 'Spinning Woman' to 'Old Maid' to What? On the Sense Development of Spinster

Terms for various aspects of womanhood have always been exposed to socially and culturally motivated changes of meaning, very often in a negative direction (cf. Kleparsky 1990:91-92). In this paper I will try to illustrate such a development with the aid of prototype theory and frame semantics.' In Persson (1990:18) I argued that componential analysis and prototype theory are each other's opposites. The basis of this contrast is that a 'hardcore' componential model makes a sharp distinction between conceptual or denotative meaning on the one hand, and connotative or referential meaning on the other (cf. Leech 1981:9-13), while the theory of prototypes does not make such a distinction at all (cf. Aitchison 1994:51-72). The properties that single out central examples of a category from peripheral ones would seem to be referential or connotative by definition, as they do not serve as minimal criteria of category membership. To take one example, woman may be defined analytically in terms of the complex concept 'human female adult'. This is taken to be the "categorial concept" on which every native speaker of English can agree. However, in order to ensure that we identify central examples of the category, we will probably have to include a number of properties or attributes traditionally termed "non-criterial" or "connotative". Such properties might be 'having breasts and a womb', 'being able to bear children', 'having feminine features', 'having a high-pitched voice', etc. A barren woman who has undergone both a uterectomy and a double mastectomy, and has a bass voice and masculine features, will of course still be called a woman. A woman lacking all these so-called connotative characteristics of womanhood is, however, likely to be regarded a peripheral rather than central example of the category by many (prejudiced) language users. We can illustrate this relationship between the two models as in Figure 1.

This is a revised version of a paper called 'Social Change, Prototypes and Word Meaning' published in Todtenhaupt, Μ. & I. Valfridsson (eds.) (1994).

Gunnar Persson

274 woman

women

'can bear child

'has a highpitched voice'

'human female adult'

'has breasts and a

The left-hand part of Figure 1 is a 'conceptual container' with criterial concepts, in the traditional sense, stored in the central box and non-criterial ones in the peripheral one. The right-hand part symbolises a set of women with 'good attributes' in the centre and 'worse attributes' in the periphery. The peripheral attributes need not fulfil more than the central criteria of the left-hand model, while the central are taken to possess some or all of the properties listed as connotative concepts in the periphery of the left-hand box. In other words, the marginal connotations of the hardcore model become factors of central importance for the identification of prototypes. Thanks to the development of prototype theory and frame semantics within various cognitive frameworks (see for example Langacker 1991 and Lehrer and Kittay 1992), it has been recognised that so-called "connotations", which have been decried by some traditional semanticists as useless, fleeting, and idiosyncratic associations, may in fact be some of the most important criteria of word meaning.2 One need only look at the sense development of many everyday words to realise that very often a so-called "connotation" has served as the basis of a new major sense.5 One may ask how this could be possible if the original connotation had just been a marginal association. One of the best examples of this is probably the sense development of spinster, the etymological history of which is outlined in the OED as follows:

2

3

Leech (1981:13) takes a disparaging view of connotative meaning, which, as opposed to conceptual meaning, he claims "is relatively unstable" and "indeterminate and open-ended". It is therefore interesting to note that John Stuart Mill (1879:33-6, quoted in Ullmann 1962:74)) claimed that "whenever the names given to objects convey any information, that is, whenever they have properly any meaning, the meaning resides not in what they denote, but in what they connote. The only names of objects which connote nothing are proper names; and these have, strictly speaking, no signification." Mill's view is not unlike that held by cognitive linguists. A case in point is gay, whose 'connotation' "addicted to social pleasures and dissipations" {OED) eventually gave rise to the now established meaning 'homosexual. Of a place: frequented by homosexuals' (OED).

From 'Spinning Woman' to Old Maid' to What? On the Sense Development of Spinster

275

I. A woman (or, rarely, a man) who spins, esp. one who practises spinning as a regular occupation. The first recorded example, shown in (1) below, is from 1362. (1) And my wyf at Westmunstre pat wollene clop made, Spak to pe spinsters for to spinne hit softe. II. Appended to names of women, originally in order to denote their occupation, but subsequently (from the 17th century) as the proper legal designation of one still unmarried. (OED) The first unmistakable instance of the latter designation is from 1617 (2). (2) A Spinster, a terme, or an addition in our Common Law, onely added in Obligations, Euidences, and Writings, vnto maids vnmarried. (OED)

A hundred years later spinster had acquired the still extant meaning 'a woman still unmarried; especially one beyond the usual age for marriage, an old maid'; cf. (3). (3) W. Irving Alhambra II. 140 The vigilant Fredegonda was one of the most wary of ancient spinsters. (OED, 1832)

The dictionary makes no attempt to trace the causes of this development, but it seems reasonable to assume that there was a pragmatic reason. Most women employed as spinsters are likely to have been unmarried, perhaps even required to be so, because they could more easily work long hours if they had no husband and children to take care of. Gradually, 'unmarried' came to be more and more strongly associated with the meaning of the word and in the end it took over completely, making the original sense obsolete. We could thus explain the development in the following way; the terms in brackets are regarded as optional concepts: Stage Stage Stage Stage Stage Stage Stage

1: 2: 3: 4: 5: 6: 7:

'human being, esp. a woman who practises spinning (as a regular occupation)' 'woman who practises spinning as a regular occupation' '(unmarried) woman who practises spinning as a regular occupation' 'unmarried woman who practises spinning as a regular occupation' 'unmarried woman (who practises spinning as a regular occupation)' 'unmarried woman' 'old maid'

It is very interesting to observe the interplay between the optional and obligatory concepts in this chain of development. 'Woman' became obligatory because very few men were employed as spinners. For similar reasons 'unmarried' was introduced and gradually made obligatory. It is still not quite clear, however, how the leap from stage 5 to stage 6 came about, but it may be assumed that spinning was a very common trade for unmarried women, so common that being a spinster became obligatorily associated with being unmarried. So pragmatic factors created 'connotations', which gradually developed into 'denotations', or so traditional lexicography would probably have it. But it is in fact quite unnecessary to think of this development in terms of connotations. It can equally well be seen as a socially and pragmatically conditioned change of prototypes, as illustrated in the semantic frames below, in which ovals represent concepts and attributes of concepts, while values of attributes are represented either by circles symbolising prototypical values of attributes or

276

Gunnar Persson

by rectangles symbolising non-prototypical values/ An exclusive prototype (i.e. without any alternative value) is also symbolised by a circle. STAGE 1

STAGE 2

The transition from stage 1 to stage 2 was brought about by two non-prototypical values, 'male (sex)' and 'occasional (occupation)', disappearing from the semantic frame of 'spinster'.

'

This is my own version of a semantic frame system developed by Barsalou (1992).

From 'Spinning Woman' to Old Maid' to What? On the Sense Development of Spinster

277

STAGE 3

As we can see, the major change from stage 2 to stage 3 is the introduction of 'civil status' as a salient attribute with 'unmarried' as its prototypical value. It is not unlikely that spinster came to serve as a metaphor for 'unmarried woman' during the period of transition from Stage 2 to Stage 3. Since the prototypical spinner was an unmarried woman, it was probably easy to start using spinster figuratively about any unmarried adult female. STAGE 4

In stage 4 'unmarried' has become the exclusive prototypical value of 'civil status'. This is of course an important step towards the dominant sense that developed later on (Stage 5).

278

Gunnar Persson

STAGE 5

Stage 5 is close to the sense development illustrated in stage 6, as 'spinning' is no longer a prototypical value of the occupational attribute. STAGE 6

In stage 6 the connection with 'spinning' has been completely lost and the lexeme has therefore become entirely demotivated with regard to its original sense. As we all know, the development did not stop here. Stage 6 soon became the source of yet another prototypical sense, that of the frumpish and prudish woman who has been left on the shelf, the social motivation being that spinsterhood was highly undesirable and must therefore indicate that an unmarried woman had flaws in her character or appearance that

From 'Spinning Woman' to 'Old Maid' to What? On the Sense Development of Spinster

279

made it impossible for her to catch a husband. In a traditional model these aspects would again have to be represented as peripheral connotations, while in a prototype model they would come out as markers of central examples, as illustrated in Stage 7. STAGE 7

The sense illustrated in stage 7 has been the dominant one for more than two hundred years; examples (4) and (5) are from the 20th century (1961). (4) [...] the dining-room of an establishment run by an amiable and fluttery spinster. (LOB) (5) [...] Kerstin Meyer struck exactly the right note of hysterical devotion as the Poet's spinster secretary. (LOB)

It is tempting to speculate about what will happen to this word now. For a long time it seemed as if it was becoming obsolete (in fact, that is exactly what my native informants have told me). There are only a couple of instances of the word in LOB and none at all in BC. The demise of the word might come about for a number of reasons: (a) in present-day society the importance of marriage as a desirable or even indispensable institution for women has decreased to a great extent, and as a result spinster may be losing its descriptive salience; (b) the predominantly unfavourable sense might alienate women, which will hasten its departure as a lexeme of current relevance. However, the word is far from dead yet. There are 47 British and 9 American instances in the Cobuild Corpus, many of which collocate with negative attributes such as sex-starved, frustrated, ugly and lonely (UK). On the basis of the above speculations a scenario such as that in Stage 8 might be envisaged (cf. Persson 1994:176).

280

Gunnar Persson

STAGE 8

The only addition made in stage 8 is the introduction of 'obsolete' as the prototypical value of the usage of spinster. The inclusion of 'current' as an alternative is due to the fact that developments of this kind take time and are also linked with the ages of the language users of a community (what sociolinguists refer to as "apparent time"; cf. Fasold 1990:213). Words are, however, known to undergo unexpected developments for particular reasons. One example is gay, which was reclaimed from its negative usage and became the positive in-group term used by homosexuals. Recent findings on the Internet seem to indicate that spinster may be undergoing a similar development in the USA. The following is a quotation from the site Spinster Elders: Our ciders are leading the way for spinster acceptance by openly acknowledging themselves as spinsters in a society that punishes for being different. There was a time women were banished for being unmarried at any age, regardless to the circumstance they were branded with the Sword. For nearly a century that word has held women by the throats pressuring many to marry and live in hopeless and unhappy situations. Spinster elders know what it's like to run and hide whenever the marital question was raised and that's the reason for living openly for young Leather Spinsters to see that a woman can live a fulfilling life alone. This seems to indicate that American women are reclaiming spinster and consciously trying to use it as a favourable term for unmarried (career) women. Of interest is also the fairly recent coinage of the compound leather spinster, which, according to the site LEATHER SPINSTERS and Their Degrees of Asexuality?

'

This is how the Leather Spinster Fellowship presents itself on the Internet: "Knowing who they are brings forth peace of mind and comfort, it also gives them the desire to share their new found awareness of self with others. People who could appreciate their excitement of self-acceptance as happily single women, other women who have endured the ridicule and dumbfounded looks from strangers, friends, and family. No one wants to be an island so a need to know others that value their separateness rather than submit to societal demands to conform is desired. Spinster fellowship has increased with more women acknowledging their social diversity with the intent on living

From 'Spinning Woman' to 'Old Maid' to What? On the Sense Development of Spinster

281

[...] is used to describe a happily unmarried careerwoman (sic) under the age of 55. She usually falls into one of two categories: (1) Heterosexual woman who is contented with life without a mate; (2) Asexual woman with no need for sexual stimulation or intimate relationships. The motivation for the metaphorical use of leather is said to be that "Leather signifies toughness or defiance of authority". So perhaps it is too early to count out spinster after all.6 Instead of the scenario presented in stage 8, it might be more accurate to sketch a frame such as that in stage 9 below, in which spinster is seen as a positive in-group term, at least in American English.7 STAGE 9

Stage 9 represents an attempt to outline a prototypical profile of the new meaning of spinster that seems to be emerging. One important change is that unmarried, with its negative

6 7

in a "roundabout fashion" (openly) for all to see. Unity with others that share a common belief makes for strength and agility in times of pressure from the greater society at hand. No more secret rendevous (sic) at isolated locations and mouth covered whispers, these women are meeting in public places such as libraries, restaurants, taverns, coffee shops, parks, and so on with little care if they're suspect in their motives. The first thing outsiders assume when seeing a group of attractive, sophisticated women is they must be available for dates and relationships with the nearest man or secondly they're gay, both assumptions are false. The more leather spinsters make their presence known the more their image in the media and societal perceptions of them will change. Being open about their lifestyle but not forcing others to accept them will help future generations of women that don't see committed relationships, intimacy, and marriage as belonging on their priority list. It will give them another option from which to choose without the same cultural bashing experienced by spinsters of past and present. Unlike other groups, leather spinsters refuse to force others to accept them, their ultimate goal is to be happy living as they please, it is believed that with time and visibility more women will find solace in knowing they belong somewhere other than committed relationships and marriage. It must also be stated that leather spinsters tend to be of the thought that all future options should be left open, so if the desire to add new dimensions to their living arrangements should arise they're not locked into decisions made in the past." There are no less than 24,225 sites containing the word spinster(s) on the Internet. According to the same site, British women are also joining this movement.

282

Gunnar Persson

implications, has been replaced by single, which carries no such association. Whether this will become an established usage or whether the extinction sketched in stage 8 will take place is not possible to tell at the moment. At any rate it is interesting to be able to follow a historical development while it is happening. The sense development of spinster is a good example of the intimate relationship between culture, social conditions, and language (cf. Hughes 1988). Culture and social change will always influence word meanings, albeit not always in the expected direction. As we have seen, the answer to the question in the title might therefore be either spinster = 'obsolete and negative term for an old maid' or 'unmarried, independent career woman' - or both. The person who lives will see.

References

Dictionaries and Corpora BC = A Standard Sample of Present-Day Edited American English, for Use with Digital Computers (= "the Brown Corpus"). eds. Francis, W.N. et al. Providence, Rhode Island: Dept. of Linguistics, Brown University 1964. LOB = The Lancaster - Oslo - Bergen Corpus of British English (= "the LOB Corpus"), eds. Johansson, S„ G.N. Leech & H. Goodluck. Oslo: Dept. of English, University of Oslo 1978. OED = The Oxford English Dictionary. 2nd ed. Oxford: Oxford University Press 1989.

Other literature Aitchison, J. (1994) Words in the Mind. An introduction to the mental lexicon. 2nd ed. Oxford: Basil Blackwell. Barsalou, L.W. (1992) 'Frames, Concepts, and Conceptual Fields'. In: Lehrer, A. & E. Kittay (eds.) 1992:21-74. Fasold, R. (1990) The Sociolinguistics of Language. Introduction to Sociolingustics, Volume II. Oxford: Basil Blackwell. Hughes, G. (1988) Words in Time. A Social History of the English Vocabulary. Oxford: Basil Blackwell. Kleparsky, G.A. (1990) Semantic Change in English. A Study of Evaluative Developments in the Domain of Humans. Lublin: Redakja Wydawnictw Kul. Langacker, R.W. (1991) Concept, Image, and Symbol. The Cognitive Basis of Grammar. Berlin and New York: Mouton de Gruyter. Leech, G.N. (1981) Semantics. The Study of Meaning. 2nd ed. Harmondsworth: Penguin Books. Mills, J.S. (1879) A System of Logic, Ratiocinative and Inductive. Vol. I. 10th ed. London. Persson, G. (1990) Meanings, Models and Metaphors. A Study in Lexical Semantics in English. Stockholm: Almqvist & Wiksell International. - (1994) 'Social Change, Prototypes and Word Meaning'. In: Todtenhaupt, Μ. & I. Valfridsson (eds.) (1994). Todtenhaupt, Μ. & I. Valfridsson (eds) (1994) Sprache als lebendiger Kulturspiegel. Festschriftßr Astrid Stedje. Stockholm: Almqvist & Wiksell International. Ullmann, S. (1962) Semantics. An Introduction to the Science of Meaning. Oxford: Basil Blackwell.

From 'Spinning Woman' to Old Maid' to What? On the Sense Development of Spinster

283

Internet Sources (2000) LEATHER SPINSTERS and Their Degrees ofAsexuality. http://www.leatherspinsters.com/preview.html Sinclair, J. et al.(2000) The CobuildDirectCorpus. Birmingham: The University of Birmingham/ London: Harper Collins Publishers, http://titania.cobuild.collins.co.uk/form.html Spinster Elders, http://leatherspinsters.com/elders.html

Vira Schmiedtovä/Barbara

Schmiedtova

The Colour Spectrum in Language: The Case of Czech. Cognitive Concepts, New Idioms and Lexical Meanings

1. Introduction

The representative corpus SYN2000 in the Czech National Corpus (CNK) project contains 100 million word forms taken from different types of texts.1 All project analysis is based on this corpus, containing word forms taken from newspaper texts (70%), poetry and literature (15%) and scientific texts (25%). As might be expected, the language material drawn from these sources may be distorted; word frequency and word combinations are influenced by journalistic slang. Czech is an inflectional language and for that reason various forms of words have to be taken into consideration. First of all, we have tried to determine the breadth and depth of the linguistic material in the SYN2000 corpus. Then we chose the adjectives indicating the basic colours of the spectrum. Of course, there are numerous other words derived from these adjectives. Accordingly, nouns and adverbs are also included to a limited extent. Verbs, names for shades of colour, and phrases and expressions linking colours with other words, not derived from them, have not yet been processed in the corpus. This work is in the planning stages.2 The basic colour terms, from a linguistic point of view, occur with the following frequency: eerny (black)

bily (white) eerveny (red) rudy (deep red) zeleny (green) modry (blue)

17 698 16717 8 646 2 104 7 220 6 569

zluty (yellow)

5 571

grey (Sedy, Sedivy)

3 779 3 135 2817 922

hnidy (brown)

ritzovy (pink) oranzovy (orange) flalovy (violet)

876

An interesting linguistic fact can be observed in the proportional frequency of different entries. Greater frequency means greater phraseological load and semantic wealth. The observation, based on frequency dictionaries of different Indo-European languages such as Czech, English, Italian, Polish, Rumanian, Slovak and Ukrainian, is that in all observed languages the colour white takes first place.' Our corpus is quite contrary (see table above). For reasons of time and space, we have chosen to focus only on the three most frequently occurring colours: eerny 'black', bily 'white', and eerveny 'red'.

The representative corpus was created in winter 1999 - 2000. Kocek J., Kopfivovä M„ Schmiedtovä V.,(2000). Pawlowski (1999).

286

Vira Schmiedtovä/Barbara

Schmiedtovä

Black and white are opposites and literature on colours suggests that all languages distinguish between these two colours. Because they are contrasting colours their description cannot be entirely separated and the entries for these two colours overlap. Therefore, we have additionally chosen red as the most striking of the colours for our analysis.

2. Black

In contemporary Czech a new expression has appeared: eerny küo 'black horse' (65). Not in the sense of the chess figure but as an indicator of someone or something, in whom or in which great hopes and expectations are placed, a favorite: „ the 'black horse' of the league", „ the 'black horse' of the season". In the CNK we have not actually analyzed the entry eerno- 'black' when it forms part of composite words, but the most frequent occurrences of this component in the SYN2000 are: Eernobyl 'Chernobyl', eernobylsky 'Chernobyl'; eernomorsky 'Black Sea' adj.; eernobily 'black-and-white'. The most frequent derivatives are eernoch 'black man' and eernossky black - in the sense of 'pertaining to black people'. If we compare instances of the word eerny in the SYN2000 with the entries in the Czech language dictionaries, we see no major differences. The word eerny belongs to the core of the lexicon. It is contrasted with the word bily 'white' and vice versa. These two are contrasting colours. As their meaning structures have developed, the word eerny has taken on a semantic character that is negative, dirty, difficult, whereas the word bily has acquired positive, clean and unadulterated associations. The basic meaning of the word eerny is to express a colour, the Dictionary of Literary Czech Language (SSJE) uses the following definition for this word: 1. Being the colour of soot: Writing a dictionary definition for a colour is difficult. Books which analyze the meanings of colours suggest that the meanings of the words black and white have developed from existential human feelings, which see these two words as night (black, dark) and day (white, light), that is light is seen in positive terms and dark in negative terms. We can document this perception of black with the expression noeni mitra 'nightmare' (576) "an unpleasant problem causing a person some concern". In this association, the words night and black are often interchangeable. For example the expression eernä mura '* black mare' (42) can also be found. Even within this initial meaning of the word black, a further meaning or designation can be pinpointed, where blackness as single semantic feature is used to represent a wider whole. Eerny 'black' = eernoch 'black man', eernossky 'black' adj. „'the Black Continent' = Africa", „eernä kräska = dernoäskä kräska" - 'a beautiful black woman'. As far as new expressions are concerned contemporary language has the combination eerny ier;/'black sheriff (9) = employee of a security firm with a black uniform: „they decided to hire black sheriffs in order to get back what was rightfully theirs"; eernä skrioka 'black box'(85) = 1. record of an airplane's flight „President Yeltsin is to hand over to his Korean counterpart the black box, a recording of the flight information of the Korean Air-

The Colour Spectrum in Language: The Case of Czech

287

lines Boeing shot down in 1983", 2. security anti-theft device in a car „when a police patrol receives the coded signal sent out by the black box, the police contact the firm Secar..." 2. Something which is dark, but not exactly black: black eyes (which are brown) (178); dark eyes (124); black beer (2); dark beer (18) - antonym light beer; black coffee (without milk) (150), black bread (rye bread) (39). 3. Illegal, not permitted, dishonest, secret: eerny trh 'black market' (387), eerny fond 'black fund' (45), udilat nico na eerno (63) for doing something unofficially. When we further narrow this meaning, we reach the sense of 'not paying' as in the Czech expression jet na eerno (13) 'traveling without paying', lit. 'black' eerny pasazer (150) 'black passenger', i.e. without a ticket. 4. Bad, evil, unsuccessful: eerne svidomi 'black conscience' (18), eernä listina 'black list' (120), eerny rok 'black year' (2), eernä kniha 'black book'(7). 5. Not optimistic: vidim to eernl 'the future looks black to me' (101), lieit nico ν eernych barväch (6) 'to paint something black', 6eme mySlenky 'black thoughts' (45) „I am known as a person who paints everything black". 6. Insufficient light: eernä noc 'black night' (34), eerny les 'black forest' (47) = 'deep', eernä hodinka 'the darkest hour' (45) 7. Difficult, laborious, soiling: eernä metalurgie 'black industry' (4), eernä präce 'black work' (17), eerni nädobi 'heavy black pans' (2) = 'pans not made of porcelain'. 8. Harsh, cruel: eerny humor 'black humor' (223) 9. Difficult, dangerous: eerny kasel 'black cough' (21), eerne nestovice 'black pox' (16) Black as the colour of mourning probably came to our culture with Christianity, which took advantage of and reinforced the existential relationship to black as the colour of unpleasant sensations, of insufficient light, and gave it a negative sense, while white was the colour of light, of positive sensations and was given positive connotations. According to some linguistic evidence this was not always the case. The word combination bily rubäs 'white shroud' (4) is used for "the clothes in which a dead person is dressed", bilä ραηί 'white lady' (46) is an expression for a ghost, that appears in the white shroud of a dead person. A number of castles have their own "white lady", suggesting the very opposite of the general rule.

288

Vira Schmiedtovä/Barbara

Schmiedtovä

There are also numerous phraseological expressions, which the SYN2000 indicates to be still in common use: trefit (se) do eerniho 'to hit the black' (6) = "to say or do the right thing"; E(e)erny Petr 'Black Peter' (188) - 1. "a children's card game", 2. figuratively, something unwanted „ he gave the Eerny Petr to the Palestinians, the Eerny Petr remained in the hands of angry France". The Czech language also often uses expressions of a terminological nature: eerne uhli 'black coal' (251), figuratively eerne zlato 'black diamonds' 'anthracite', but also crude oil 'black gold'(44); eernä zvir 'black game' = 'wild boar' (416); eerny rybiz 'blackcurrant' (82); eerny bez 'elderberry' (68); eernä dira 'black hole' (266) 1. physically "place in space where gravity is so strong that it pulls everything to it, including light 2. figurative somewhere where everything, especially money, disappears for good „Russia is a black hole for the Germans", eerne divadlo 'lit. black theater' (167) (the semantic feature of insufficient light and black dress is also relevant here a style of theatrical performance, often in Prague, using visual deception, with actors dressed in black velvet costumes, illuminated in such a way that they cannot be distinguished from the background. This makes it possible to create on the stage an illusion of movement that is not physically possible). A distinctive transition from black to white is seen in the combination eernobily 'black-and-white' (1476), which, apart from indicating colour, also means "an insensitive superficial way of seeing or experiencing something" 'seeing things in black and white'; or 'having something' eerne na bilem 'black on white' (90), „having written proof of something".

3. White

The adjective bily 'white' is less structured in terms of meaning than black. It would be interesting to know whether the negative characteristic present in the word black is in some way psychologically more attractive than the positive characteristic present in the word white, whether the human psyche chooses to focus more on black and whether this is reflected in the rich structure of meanings. The positive characteristic of the word white is currently seen in the name of an organization, which takes an interest in the fate of people who have become victims of a crime; the organization is called the Bily kruh bezpeei 'The White Circle of Security' (48). Another example of the positive character of the word white is in the word combination bilä kniha (28), see below. In the Czech language a new phrase that has recently emerged is bily kuo 'white horse'(135). This phrase is not used in the sense of the knight on a chess board, but as a term for someone who is used for some kind of illegal business transaction and then removed, often killed, „e.g. He found himself in the typical position of a white horse for the Romany gangs." It is interesting that the phrase bily küo 'white horse' has generally negative connotations, while the phrase eerny küd 'black horse' is positive. If we compare the occurrences of the word white in the Czech National Corpus with the meaning structure of all our main dictionaries, it has to be acknowledged that there is no major divergence. The word white seems to be a core component of vocabulary.

The Colour Spectrum in Language: The Case of Czech

289

The basic meaning of the word is to express a colour; the Dictionary of the Written Czech Language uses the following definitions: 1. Being the colour of snow, milk, etc. Just as with eerrty 'black', where the adjective can be used to mean "pertaining to a black person", bily 'white' can be used to mean "pertaining to a white person", „e.g. de Klerk yesterday appealed to the white inhabitants of South Africa...; ...he is calling on all whites to gang up on blacks." 2. Snlhovy (of/in/like snow) is a synonym occurring very frequently in the Czech National Corpus „e.g. bila stopa 'white track' (142) = "a track in the snow for cross-country skiing", bile hry 'white games' (4) = 'Winter Olympics'; bile velikonoce 'white Easter' (1) = when it snows at Easter; bile svahy 'white slopes' (6) = snow covered slopes; bily sport 'white sport' interestingly does not refer just to skiing (6) but also tennis (22). Surprisingly, this meaning is not included in available dictionaries. 3. Something that is bright, but not exactly white white hair, gray (87), white coffee "with milk" (46), white wine, (420) antonym red, white meat (90) - fish, chicken, veal, lean meat. 4. Sufficient light bily den (bright, lit. "white" daylight). 5. Blank bily list papiru (5), bile misto 'lit. a white place' (143) = something about which we know nothing, a frequent Czech phrase in this sense is bile misto historie 'a blank page "white place" in history', and antiquated meaning, surviving in the sense of the phrase Rdd bileho Iva 'the Order of the White Lion' (32); the Czech state medal has the meaning of the word silver. 6. Meaning of the word silver This is where the meaning structure of the word white ends. What is lacking here is the contrast to the word black in the sense of illegal, not permitted, dishonest, secret. But a neologism with this sense can be found in: bily dovoz 'white imports' (0) meaning legal imports. The phrase eerny trh 'black market' is well known - illegal, but there is also a neologism bily obchod 'white market' (0) - legal, another neologism being seda ekonomika 'the gray economy' (61) - semi-legal. We can also note the terminological link - bily jogurt 'white yogurt' (50) = without flavoring, bild mouka 'white flour' (27) = wheat flour, which also forms the basis of the expression bile peeivo (30) white bread, white rolls, etc. Phraseological expressions bear their own meanings - bile limeeky 'white-collar workers' (84) = non-manual workers; bile maso 'white meat' = prostitutes (59). The most common phrase in the SYN2000 in this sense is obchod bilym masem (the white slave trade). Bild vräna 'a white crow' (51) = something or someone exceptional „e.g. As if that would make the Czech Republic... one of the "white crows" among the post-communist countries". A non-dictionary but very transparent use of

290

Vira Schmiedtovä/Barbara Schmiedtovä

white is in the term bile plästi 'white coats' = doctors „e.g. This strikes me as a typical example of "white coat" solidarity. They never admit publicly that a colleague might have made a mistake even when he has sewn a hat into someone's stomach...". We may also notice the expression bilä kniha 'white book' (28) „e.g. After introducing "black and white books", in which the city councils of Moravskä Ostrava-Piivoz could record the good and bad deeds of local business people, she had another idea". Here we see an example of the polarization of black as negative and white as positive. Next to the black and white books, the expression, eervena kniha 'red book' (9) also exists (see analysis of the colour red).

4. Red

This leads us to the colour red. After black and white it is the third most interesting colour linguistically. Red is perceived as the colour of fire, sun, blood but also as a colour symbolizing danger, and therefore also the need for special care. It is a very striking colour. An interesting thing about the colour red in Czech is that it is generally expressed through one of two basic words, eerveny and rudy. Both these words have their own particular associations. In Czech we call the main square in Moscow Rudi nämlsti 'Red Square'. Red used to signify, and evidence suggests that it continues to signify, a revolutionary left-wing position. In post-revolutionary Russia, Bolscheviks were red rudi, there were Red Divisions, see Babel's Rudd jizda (in English - The Red Cavalry). In contemporary Czech, we encounter the word rudy in this sense in connection with the Red Army Faction and the Khmer Rouge in Cambodia. In reference to American Indians, the word rudy is used as a signifier of race in the same sense as with black and white, „e.g. rudy indiänsky närod 'the Red Indian nation'; rudy bratr 'red brother; redskin' (7). Rude prävo 'Red Right', the former communist daily, has now changed its title to Prävo. 'Right' has recently begun to appear in Czech in connection with brown. Brown symbolizes fascism, red communism „e.g. They warn against the red-brown danger; ...wearing a red or brown shirt". Linguists have shown that the more familiar people are with objects or ideas, the more detailed and richer the system of meanings they create for them. How can we explain, for example, the way that Czech distinguishes the shades of red in such detail? Not only does Czech have two words to express this concept, but also a large number of expressions to express shades of red: bordovä 'bordeaux', purpurovä 'crimson/purple', nachovd 'crimson', karminovä 'carmine', granätovä 'garnet-red', cihlovä 'brick-red', νίηονά 'wine-red', jahodovä 'strawberry', krvavä 'blood-red', sarlatovd 'scarlet'. Bruwtny 'russet/deep red' in Czech can only refer to a person. Dictionaries use concrete notions for the definition of colours. The Dictionary of the Written Czech Language uses the following definition for eerveny 'red': 1. Being the colour of blood or of field poppies For rudy 'deep red' the definition in the same dictionary is 'being the colour of blood, (full, dark) red. However, this is a tautological definition. Or should we understand this definition

The Colour Spectrum in Language: The Case of Czech

291

in such a way that eerveny is the colour of oxygenated blood and rudy the colour of nonoxygenated blood? If we compare the breadth of meanings for red eerveny and black, we have to conclude that for the adjective red there is only one basic meaning: to indicate a colour. The word eervenä as a noun exists only in the sense of blood or to indicate the colour of a traffic light, „e.g. the "red" was gushing out of him; he drove through a red." From the connotations red evokes we can also observe certain phraseological expressions: eervenä nit 'red thread' (76) - something striking, easy to follow, a logical sequence „e.g. This idea was like a "red thread" going through the entire program." The pairs red and white blood cells, red and white wine, eervene and bile zeli 'red and white cabbage' (red cabbage is purple, just as in reality eervenä repa 'beetroot' is not really red.) 2. Danger We have yet to look at red as a colour symbolizing danger. A rather old-fashioned expression eerveny kohout 'it. red rooster' (17) indicates fire „e.g. Only the firemen were able to put a stop to the havoc being wreaked by the 'red rooster'". On traffic lights, red was chosen as a colour to warn against danger. 3. Red is also the colour of love What does the use of the colour red in the expressions Red Cross and Red Crescent mean? In this case it is red on white, it is a link between two strikingly different, contrasting colours, frequently used on various signs, including the Czech national flag. It is interesting to see the ways this colour is linked to the word "book". This is seen in the expression eervenä knihovna 'lit. red bookcase' (80). The expression comes from the colour of an edition of books that came out during the First Czechoslovak Republic (19181938) - reading for women and girls. If we say that something is as if from the "red bookcase" in Czech today it means that it is in the same spirit as the books that came out then. They are romantic novels, where a poor girl always ends up marrying a rich man. But the expression eervenä kniha 'red book' (9) means something else. It is a list of endangered species. Here the word is used in the sense of warning and danger. And how about the rudä knizka 'little red book'? This was the Czechoslovak Communist Party membership card (2) or Mao's red book (5). There are also further expressions in Czech: eervene tväre (41), rty, (12) iista (2), pusa (5). 'red cheeks, lips, mouth' symbolizing good health, whereas eerveny nos 'red-nose' means a drunk (24) or a nose which has been exposed to the cold for a long time (1), the red nose of a clown (9), indicate symptoms of crying (4) eerveni oei 'red eyes' indicate symptoms of crying (2), or strong wind or an illness (12), or the eyes of some animals (22).

292

Vira Schmiedtovä/Barbara

Schmiedtovä

5. Conclusion From an analysis of these three examples one can see the extent of linguistic wealth and diversity we are dealing with. Available lexical resources hide enormous information not only on the frequency but also on the users of a language and on their linguistic conceptualization of the world. Due to limitations in size and lexicological concepts, no existing dictionary is capable of embracing all these nuances. Currently we can only hope that the next dictionary of contemporary Czech that comes out on the basis of the SYN2000 and the Bank of Czech, corpora created in the Czech National Corpus project will address this issue.4 This would mean that without traditional size limitations we would be able to include such fine nuances of language as well.

References

Kocek J., Kopfivovä M„ Schmiedtovä V. (2000): The Czech National Corpus. In Euralex 2000 Proceedings, Stuttgart, 127-132 Pawlowski, A. (1999): The Quantitative Approach in Cultural Anthropology: Application of Linguistics Corpora in the Analysis of Basic Colour Terms, Journal of Quantitative Linguistics 6, 1999/3 222. The Dictionary of the written Czech language (Slovnik spisovneho jazyka eeskeho), Vol. I-III, Academia, Prague 1960-1989.

4

The Bank of Czech is a cover term for the sum of all the converted and marked-up texts of contemporary written language which can be examined using our software.

Wlodzimierz

Sobkowiak

On the Phonetics of Trans- in EFL Dictionaries

0. Introduction

In my 1997 contribution to the Accents'97 Conference I argued that: (a) "learning a foreign language is the acquisition of rules", (b) dictionaries, even if not normally perceived as such, are repositories of such rules, including phonological ones, (c) EFL dictionaries are often seriously inconsistent in their treatment of pronunciation, both within and across dictionaries, (d) this inconsistency jeopardises the role which such dictionaries play in EFL pedagogy, either explicitly or implicitly, (e) as well as - particularly in the case of machinereadable dictionaries (MRDs) - the heuristic (indexical) and serendipitous function of phonetic representation in such dictionaries. The case in point was sonorant syllabicity which was examined in three MRDs, namely Chambers, OALDCE and OED2, and found wanting on account of poor consistency. An extended version of that contribution is now part of my book devoted to pronunciation in EFL MRDs (Sobkowiak 1999). Thus, indirectly, in Sobkowiak 1997 I took issue with Wells's 1985 statement that "The purpose served by pronunciation indication [in dictionaries] is [...] to advise the user who is unsure of the spoken form of a word by recommending a suitable pronunciation for it" (ibidem:45; my emphasis). I believe that the user of an EFL dictionary has the right to expect advice not only on the pronunciation of a given word, but also on sublexical units on the one hand, such as common letter/phoneme clusters (e.g. word-final syllabic sonorants) or morphemes, and supralexical units on the other, such as idioms (see Broeders 1987, for example), collocations, and possibly even larger linguistically salient entities. In this paper I will present a short critical overview of the phonetic treatment of one particular English morpheme in leading EFL dictionaries (see Chevillet 1993 for a similar treatment of homo-). I will show that, in addition to the didactic dysfunctions of phonetic inconsistencies underlined in my 1997 paper, there are further problems having to do with the inconsistent treatment of morphology and its complex interaction with phonology, the field which certainly belongs to the most difficult ones to acquire from the perspective of a learner of English as a foreign language.

1. The phonetics of trans-

In terms of its functional load, semantics or morpho-phonology, the prefix trans- in English certainly does not belong to the most linguistically interesting 'building blocks' of English words (cf. Marchand 1969:147-8). Some of its tokens attach to some rather infrequent learned stems and have little circulation beyond very technical discourse, even if - in these

294

Wlodzimierz Sobkowiak

areas - the prefix is in fact potentially productive. Of the total of one to three hundred transwords (depending on the size of the dictionary) only thirteen occur on the lemmatised BNC word-list with any appreciable frequency. Of these, the most frequent transport (n) has rank 1363 (artificially boosted by the character of the majority of BNC texts), and transfer (v) 1612 (ditto). Yet, phonetically, the little prefix turns out to be surprisingly variable, both across dialects/accents and within. OED starts the discussion of the pronunciation of trans- like this: "In the pronunciation of trans- in combination [i.e. as a prefix], great diversity prevails locally and individually in cultivated speech. This diversity affects both the vowel a and the consonant s" (OED: vol. 10, p. 250).

Then, OED claims that: (a) due to analogical pressure from words such as chance or pass the prevalent pronunciation was /trains-/ in the London area, (b) in unstressed syllables the vowel may reduce to schwa through a series of intermediate stages, (c) the fricative is voiceless before voiceless consonants, (d) "also, of course, where s coalesces with initial s of the second element, as in transcend", (e) "In the south of England many use (trans-) in all trans- combinations, irrespective of what consonant or vowel follows", (f) "But many, even in the south, use (tranz-) before a liquid, or nasal, or any voiced consonant, and before a vowel". It is interesting to compare this OED evaluation of the state of affairs in the 1920s with Wells's exposition provided in his extremely influential LPD (Wells 1990), which is here quoted in extenso: "For EFL learners, the form traan's is acceptable in all contexts in all kinds of English. Actual usage preferences are fairly complex. - In the choice between s and ζ forms we can distinguish various phonetic contexts according to the sound with which the stem begins, as follows. (1) Before a voiceless sound (trans'form), s is usual. (2) Before 1 (trans'late) and before an unstressed vowel sound ('transit), s is usual though a minority use z. (3) Before other consonants (trans'gress, trans'mit), and before a stressed vowel sound (trans'act), the tendency is for BrE to prefer z, but AmE to prefer s. This also applies in any word where the prefix is felt as separate (,trans o continental). For the vowel, RP prefers as, although a substantial minority use a:, and some words have variants with a; AmE always has ae. Before a stem beginning with s the final consonant sound is often lost (trans+scribe giving transcribe traen'skraib). (Wells 1991: 728-9).

It would seem that the main change which occurred over the last seventy years in the pronunciation of trans- in RP affected the vowel: "RP prefers as" now, where the "prevalent pronunciation" was / trains -/ (Marchand lists both, /ae/ first). What is important from our EFL vantage point is that from LPD, as well as from a number of other similar sources, the learner will be able to formulate the following practical rules for RP: (1) the preferred vowel is /ae/, (2) the voicing of the fricative assimilates to the following segment, with the exception of Wells's case (2) above, (3) the subjective strength of the morpheme boundary affects the assimilation: the stronger it is, the better the chance for ItJ. For American English the pronunciation is invariably /traens-/. Doubtless under the impression of the complexity of the rule(s), Wells prefaced his exposition with the somewhat condescending permission allowing foreigners to use just one variant, effectively the American one. This in itself is an interesting side comment on

On the Phonetics of Trans- in EFL Dictionaries

295

"purpose served by pronunciation indication" in dictionaries (see Wells's quote above) 1 as well as on the global objectives of EFL pronunciation learning, as seen by one of the leading experts of the field. I am myself convinced, however, that the trans- rule, complex as it is, is no more inherently difficult than a hundred and one grammar rules found in thousands of EFL textbooks the world over. What is more important is that most learners would more or less consciously acquire such rules as trans- from their close phonetic encounters with the English lexis, including those mediated by the phonetic representation of entries in EFL dictionaries, rather than (or at least in addition to) explicit exposition of a pronunciation-textbook style. While it is true that research on dictionary use invariably shows that few learners specifically look up pronunciation in general EFL dictionaries (Bejoint 1981, Kipfer 1987, Nuccorini 1992 or Chi 1998)2, it would be hard to deny that in their search of spelling and meaning they must of necessity encounter phonetic transcription, if only thanks to its rather salient position in the dictionary entry. Additionally, my own research into pronunciation-dictionary habits of university students of EFL in Poland (Sobkowiak 1999) shows that very many such learners believe that "All English dictionaries should show pronunciation in some way" (93%) and that they "have recently checked pronunciation in an English dictionary" (95%). Considered from this point of view, the consistency of phonetic representation within and across dictionaries becomes an essential issue indeed. This is additionally underscored by the growing perception of dictionaries as more than simply reference sources, but rather fully fledged teaching/learning resources (see Nation 1989 or Meara 1993). This change is in turn strengthened by the dynamic development of hypertextual multimedia (hypermedia) MRDs for use on individual PC workstations in the learner's home. In what follows, I will analyse the (in)consistency of the phonetic representation of trans- in the four leading EFL monolingual dictionaries: OALDCE, LDOCE, COBUILD and CIDE, as well as in three widely used pronunciation dictionaries of English available in Poland, Kenyon & Knott's Pronouncing dictionary of American English (PDAE), Wells's Longman pronunciation dictionary (LPD) and Biedrzycki's Practical Pronouncing Dictionary (PPD).

' 2

In a lecture presented on November 13th 1998 in Poznan, entitled "RP - a model for teaching", Wells explicitly stated that one of these purposes is to simplify RP "by codification". But see Tomaszczyk 1979 for the surprising 65% of respondents who said they consulted dictionaries for stress and pronunciation. Tomaszczyk's methodology was faulty on a number of counts, however.

296

Wtodzimierz Sobkowiak

2. trans- in dictionaries

2.1. Data It might be objected that comparing pronunciation dictionaries with general ones is not fair to the extent that the former are naturally focussed on pronunciation and would, for example, provide a few variant pronunciations of one word, where the latter would only give one. Yet, as explicitly stated in the front matter of most pronunciation dictionaries, the first listed variants are regarded as the "main pronunciations (recommended as models for learners of English)" (LPD p. viii), which makes them directly comparable with the only pronunciation indicated in general EFL dictionaries. It is exactly these representations (rather than alternative or substandard ones) which I will consider below. To make the amount of data more manageable right from the very start, I decided to ignore some derivatives of trans- words: nouns ending in -ency/-ence, -able-final adjectives and the regularly derived adverbs in -ly. Of the remaining derivatives I only retained for further scrutiny those which showed some (morpho-)phonological differences in the shape of their trans- elements, compared to their bases, e.g. the derivatives of transit. Words where the trans- element is not felt to be a morpheme in current English (like in transom) were dropped, while those which are to some extent equivocal were retained if they reflected in an interesting way on the treatment of trans- in the given dictionary (like transient). Finally, words where trans- does not show any variation within or across the seven dictionaries were deleted. These words were: transcend /trasn'send/, transcript /'trasnskript/, transept /'transept/, transfer /traens'f3:/, transfigure /trans'figa/, transfix /trans'fiks/, transform /trasns'fooim/, transfusion /trans'fjugn/, transistor /traen'zsita/, transplant /traens'pla:nt/ transubstantiation ^traensafystasnji'eijn/. All these deletions left the thirty-one words appearing in Table 1 in the Appendix. Not all cells are occupied, of course, because some dictionaries do not list some words. Such is the case of transalpine or transient, for example. As seen in Table 1, only one dictionary, that of Wells, lists all words under analysis here. For economy of space, only the onset portions of transcriptions are entered in the table for longer words, such as transcontinental. Minor details of transcription were normalised across the dictionaries to enable easy comparison, such as the treatment of schwa+sonorant in syllabicity-prone contexts, the notation of rhoticity and epenthetic stops, the variant transcriptions of mid-back vowels and closing diphthongs, the marking of syllable boundaries, etc. With all this unavoidable normalisation, the trans- portions of words are faithfully represented to the extent of detail which is relevant for this discussion.

2.2. Consistency within dictionaries Wells's LPD is today the most highly regarded, and certainly the most extensive, pronunciation dictionary of English in existence. Also from the point of view of EFL didactics it fully deserves its high esteem (see Sobkowiak 1994). LPD and OED are unique among the

On the Phonetics of Trans- in EFL

Dictionaries

297

dictionaries which I consulted in providing a descriptive/advisory section on the pronunciation of trans-, as shown in section 1. It is fitting, then, to begin this overview with LPD. Wells's "usage preferences" are generally observed throughout the 31 words of Table 1. A more observant learner/user might, however, be baffled by certain apparent inconsistencies. For example: if the preferred pronunciation is /traenz-/ "where the prefix is felt as separate", why does the prefix itself, i.e. unattached to any stem, appear as /traens-/? Presumably, sheer statistics is the answer here: this is the globally prevalent pronunciation, so the default phonetic representation reflects this fact. But how does one explain /traens-/ in transship? How is it different phonetically from, say, transcontinental, used by Wells to illustrate the workings of clause (3) in his preferences? Certainly, the prefix is "felt as separate" in both cases? A point which is minor from our current perspective, but which may be quite confusing to an inquisitive learner, is the prefix boundary. This is particularly salient where the second syllable bears stress, and the stress mark is effectively a syllable boundary symbol. As can be seen in Table 1, it sometimes occurs after, and sometimes before, the trans-fmaX fricative. While it might be too much to expect from a pronunciation dictionary to go into the nuances of etymology and morpho-phonology to explain the differences between trans+parent and trans+spire, for example, it might still be questioned why the syllable structure of trans+port is shown as analogous to trans+spire or trans+scribe rather than to trans+pose. And why is it tran+sact but trans+alpinel In accordance with his preferences, Wells shows American trans- with the final /s/ in most cases, notably such as transverse or transpolar, but there are confusing items nevertheless. If AmE prefers /s/ "before a stressed vowel sound", the tendency is not particularly well represented by the data at hand, where the number of /s/-words with this context (itransact and transalpine; trans+cend is not counted as it falls under the last clause of Wells's preferences) more or less equals that of /z/-words (transistor and transition). Other dictionaries, both pronouncing and general, show similar types of inconsistencies. Biedrzycki's PPD, like LPD, lists both British and American pronunciation. Notice first that PPD is more (surface-wise) consistent in its trans- boundary placement than LPD: transcribe, trans'pire, transport are the relevant cases. Second, for some reason there is /s/ in transgress, even though the prefix shows as /traenz-/ in other pre-voiced contexts. Third, /traenz-/ in transliterate is clearly meant to emphasise the subjective 'separateness' of the prefix. Fourth, the metathesis of fricatives in BrE transition/al is unique /TD-internally and across all dictionaries (LPD shows 9% BrE poll-panel preference for this variant, which is consequently listed last)3. Finally, there are interesting inconsistencies across the BrE/AmE phonetic representations in PPD. Why should there be boundary placement differences between Br and Am transact, transcribe/ption and transpire, for example? Kenyon & Knott's PDAE is still widely used in Poland and elsewhere as the main pronunciation dictionary of American English, despite its age (notice the lack of transsexual)*, although Wells's LPD is gradually taking over, mainly thanks to its superb coverage. PDAE is quite internally consistent in its treatment of trans-: (a) /zJ only appears before stressed 3

Interestingly, the fourteenth edition of Jones's EPD agrees with Biedrzycki's PPD. A diachronic change may be in progress here, which EPD and PPD failed to capture. "PDAE is an historical, rather than a current, descriptive record of American English pronunciation" (Bronstein & Stewart 1990:159).

298

Wlodzimierz Sobkowiak

vowels, and (b) the stress mark is placed after /s/ in such words as transport and transact (but trans+s... words behave like in LPD). Finally, consider the four general EFL dictionaries. OALDCE shows IzJ in transit, although /s/ appears before unstressed vowels in other words. The strongly morphemically separate trans- occurs in transsexual, but not in transship. The pre-lateral trans- appears with Izl, which is counter to Wells's preferences, but regularises the behaviour of trans- in pre-consonantal contexts (consistent regressive assimilation of voice). There are some inconsistencies in syllable-boundary placement; compare transport and transpose with transplant / traens'pla:nt / and transparent, for example. LDOCE shows transitive with Isl, but transitory with Izl. Translate and transliterate are treated differently with respect to voicing, like in Br PPD but unlike in Br LPD. Boundary placement is again inconsistent: compare transparent and transport on the one hand with transplant on the other. In COBUILD the treatment of transitive and transitory is consistent, i.e. in favour of IzJ. But the transp... family of words is again problematic in terms of boundary placement: this time transparent sides with transplant against transport and transpose. Notice that the basic variant of the morpheme trans- unattached to any stem is given with Izl, which may send the reader wondering why there should be devoicing in translate. In CIDE transit appears with IzJ, but transitive with Isl. Syllable boundary is again unstable: compare translate, transparent or transport with their close neighbours.

2.3. Consistency across dictionaries If the mildly inconsistent treatment of trans- in particular dictionaries may at times be a nuisance to phonetically observant learners, the variation across dictionaries is bound to be positively baffling. The analysis in this section will be rather cursory because most inconsistencies have already become apparent from the preceding within-dictionary discussion. An intermediate EFL learner will understand that many English words have variant pronunciations, even within one accent, such as RP, and without wider sentential context. He/she might not then expect to obtain the one and only 'true' pronunciation of a given word from the dictionary, but would certainly desire to see some consistency in what is listed as the basic or recommended phonetic variant. And yet, most words in Table 1, i.e. a fair proportion of all trans- words in English exhibit wild inconsistency in this respect. The inconsistent treatment of syllable boundaries was mentioned many times above. While the practice of most dictionaries appears to be: "put it before Isl if it comes from etymological coalescence of /s+s/" (but the learner is not explicitly informed, of course), there is unmotivated variation: transcribe, transcription and transpire appear as trans+... in PPD. On the other hand, where one would expect trans+... across all dictionaries, one gets tran+s... in some: CIDE's translate, LDOCE's and CIDE's transparent, LPD's and the 'big four's treatment of transport and transpose are all cases in point. The morphological status of trans-, its subjective 'separateness', is ample source of further inconsistency in pre-consonantal contexts: while all surveyed dictionaries agree on transcontinental, only OALDCE has BrE tran/z/late, and only LPD has BrE tran/s/literate or tran/s/lucent. Transsexual is given with /z's/ in OALDCE, with isl in LPD and LDOCE, and with /(z)'s/ in CIDE.

On the Phonetics of Trans- in EFL Dictionaries

299

Wells's voicing preferences listed in LPD are not universally observed either, especially as they interact in complex ways with the 'separateness' of trans-·. PPD lists BrE tran/s/gress, OALDCE, COBUILD and CIDE give tran/z/it, COBUILD - tran/z/itive, COBLJILD and LDOCE - tran/z/itory. The pre-lateral inconsistency was mentioned above. Notice that even across the AmE word lists there is no complete voicing consistency: LPD gives 'tran/z/late (with stress on the first syllable), while ΡDAE and PPD give tran/s/'late. Summing up: neither within dictionaries nor across them is there complete consistency in the phonetic representation of trans-. Some of the observed variation can be explained by some underlying linguistic factors, which are, however, hidden from the eyes of the learner. Other is doubtless due to the intuitive nature of the representations, which are not regularised by reference to specific phonological/phonetic rules but rather to individual preferences of the compilers, as I emphasised concluding my 1997 paper. Finally, phonetic representations in dictionaries are apparently not as thoroughly cross-checked as other elements of the microstructure, such as definitions or grammar codes, which leads to further variation, this time completely haphazard.

3. Conclusion

Faced with the inconsistencies in the phonetic treatment of trans- words, a linguistically naive learner cannot fall back on a host of etymological, morpho-phonemic, phonotactic and pragmatic factors to appreciate the hidden underlying pattern. S/he can only consult what is explicitly provided on the phonetic surface, always remembering the multitude of notational conventions affecting the shape of phonetic representation, from strictly transcriptional ones to characteristically lexicographic (all this variation was normalised for the purpose of the present discussion). If this surface phonetic representation delivers conflicting evidence, the dictionary will positively be an obstacle in discovering coherent phonetic patterns or regularities. Especially as the dictionary ruling on lexical (and more generally linguistic) matters continues to be sacrosanct to learners of foreign languages generally, English in particular. And yet, this pattern creation or rule enhancement is one of the important - but often forgotten, little appreciated or deliberately ignored - functions of dictionaries. Lexicographers will have to find a way to accommodate homodialectal phonetic variation and idiosyncrasy in such a way as not to jeopardise the rule-implanting functions of EFL dictionaries. This is becoming increasingly easier with machine-readable dictionaries, where, on the one hand, the lexicographer has full access to computer-assisted regularisation tools in the process of compilation, and, on the other, the user can interactively query the contents of the resulting dictionary for interesting phenomena, such as the incidence of voicing in trans- conditioned by the phonetic context, accentual pattern of the word or morphological status of the prefix, for example. Any unmotivated inconsistencies in such querying will be all the more visible, of course.

300

Wlodzimierz Sobkowiak

References

Bdjoint, H. (1981): The foreign student's use of monolingual English dictionaries: a study of language needs and reference skills. Applied Linguistics 2.3, 207-22. Biedrzycki, L. (1995): A practical pronouncing dictionary of British and American English (PPD). Warszawa: Wydawnictwo Naukowe PWN. Breeders, T. (1987): The treatment of phonological idioms. In: A.P. Cowie (ed.): The dictionary and the language learner (= Papers from the EURALEX Seminar, Leeds 1-3 April 1985; Lexicographica Series Maior 17), 246-56. Tubingen: Niemeyer. Bronstein, A.B. & W.A. Stewart. (1990): Usage variants for a planned American English pronouning dictionary. In: T. Magay & J. Zigäny (eds.): BudaLEX '88 Proceedings. Papers from the EURALEX 3rd International Congress, 159-67. Budapest: Akad6miai Kiadö. Chevillet, F. (1993): Etudes des transcriptions phondtiques des dditions de 1 Oxford English Dictionary. Etudes Anglaises 46.3, 313-27. Chi, M.L.A. (1998): Teaching dictionary skills in the classroom. In: T.Fontenelle et al. (eds.): Euralex'98proceedings, 565-77. Li6ge: University of Li6ge. Hornby, A.S. (ed.) (1989): Oxford advanced learner's dictionary of current English (OALDCE). [4th ed.] Oxford: Oxford University Press. Jones, D. (1977): Everyman's English pronouncing dictionary (EPD). [14th ed.] London: J.M. Dent & Sons, Ltd. Kenyon, J.S. & T.A. Knott. (1953): A pronouncing dictionary of American English (PDAE). Springfield, Mass.: G. & C. Merriam Company. Kipfer, B.A. (1987): Dictionaries and the intermediate student: communicative needs and the development of user reference skills. In: A.P. Cowie (ed.): The dictionary and the language learner (= Papers from the EURALEX Seminar, Leeds 1-3 April 1985; Lexicographica Series Maior 17), 4454. Tübingen: Niemeyer. Marchand, Η. (1969): The categories and types of present-day English word-formation. Alabama: University of Alabama Press. Meara, P. (1993): The bilingual lexicon and the teaching of vocabulary. In: R. Schreuder & B. Wei tens (eds.): The bilingual lexicon, 279-97. Amsterdam: John Benjamins. Nation, P. (1989): Dictionaries and language learning. In: M.L. Tickoo (ed.): Learners' dictionaries: state of the art, 65-71. Singapore: SAMEO Regional Language Centre. Nuccorini, S. (1992): Monitoring dictionary use. In: H. Tommola et al. (eds.): EURALEX'92 Proceedings, 89-102. Tampere: Tampereen Yliopisto. Procter, P. (ed.) (1995): Cambridge international dictionary of English (CIDE). Cambridge: Cambridge University Press. Sinclair, J. (ed.) (1995): Collins COBUILD English dictionary (COBUILD). [2nd ed.] London: HarperCollins Publishers. Sobkowiak, W. (1994): Review of Longman pronunciation dictionary. Studia Anglica Posnaniensia 28, 221-3. — (1997): Consistency in EFL dictionary phonetics. In: E. Waniek-Klimczak (ed.): Teaching English phonetics and phonology II. Accents '97, 95-102. Lödz: Wydawnictwo Uniwersytetu Lödzkiego. — (1999): Pronunciation in EFL Machine Readable Dictionaries. Poznan: Motivex. Summers, D. (ed.) (1995): Longman dictionary of contemporary English (LDOCE). [3rd ed.] Harlow: Longman Group Ltd. Tomaszczyk, J. (1979): Dictionaries: users and uses. Glottodidactica 12, 103-19. Wells, J.C. (1990): Longman pronunciation dictionary (LPD). London: Longman. — (1998): RP a model for teaching. A lecture presented at an RP conference in Poznari, 13 November 1998.

On

the Phonetics

of Trans-

in EFL

Appendix: Table 1. Some

tU Q a ü

CA - N C C Μ Μ b t

Q £ ffl Ο U

trans-

c — Q, w

ο ο υ 4 Ν Ν Ν Ν Ν ~CA Β Β Β Β Β Β Μ Η Μ . Μ is is is I is !

£

cοo ο _> > ~N Ν C Β

Β CA 3

η Ν — ι ΠΙ 3 Ο 4-· Μ 3 Ο ö 'ε* ε CL C L C L CA a. CA "CA CA Ν Β Ί "a Β "β Β "c , Η . Η Μ .. Μ Η is i £ ! to is to ! is £

η > α> > a "a

£I

LΝ CA ι bo ο 3 ο CA «2 2 ο 3 rt υ 3 ο Ö Λ ε ε 'ε1 > > CL. CL α. CL Ο. ω CA CA ~Ν CA CA CA Ν Ν Ν ~N Ν Β Β Β "ε "β Β "ε "ε c Β Β W W . Η Η Η Η ti . Η Η is is is : is is £ to is is ! is is >

ω Χ

« Μ

< ο

, Eg : t!

Ν

__ Λ £> — Η Ο "C A j2 ε g « g

ti S £ g g

Χ» — Q, CA °> - u Μ -"> Μ b oA -"> ~C β ε ε « UU

tt Ju B-% - οε •S Μ ΠΙ ^ -g a a a § g g g

ι-3 c£• ϊ:S3 60 s-gg g i s

IQ

Oh 0.

03 Q Q. Oh

_ β ο Λ ε -Β τ-, ο 3 00 ™ η μ J/J "kj Ν Ν Ν Ν Ν c c c - c C C ΜΗ S « a g « μ is is to to to J3 J= to

is

*L

**.

to

to

to

Ε Λ Μ , aΝ« £ iJ Ν Ν J" "Ν ^ c ε e g a « ! to to is to to Ρ

Iα -S. » s c c i Β Μ «i

ο t-l 2 3 ΟA _ C α. C A C A Β ε "ε Κ Ν W is_ is is >

Λ ο _ 6b ca a ο 3 ο W 3 ο _CL τ Ν "CA _CA Ν Β Β g ε ε Μ a α to 4B S

c S *s "2 o. » > n 3, Sr =- S Μ - ω CO s ° a ο w Μ π ^ ^CO ^ " yboo 0 r ? t ? 0W ΙΛ C/3 C/l , '—> C/3 Ν _ N .. .. a S W M M g g M M M g g H M www j to to to Κ to to is Ρ Ρ is to

-

ω
•g a § g

_

a an

t I a ο I 3 a η a H Β Ξ. ε ε g S ' g ' g ' g S"g g g s s s g s J= J= is is is J=. is

CA

ο — ι 2 S 3, a- ο u D Μ 3 §§. eS 2 Ο CA L ο. C τ ^CA _> _> C A ΙΛ C A CA CA C Β ε "β Β Β Β Μ Η Β Η « £ £ is £

«

«

LO. U e· κ

3

re u 2 ca a ο 3 ο tΜ ο ο ο CL 3 ¥ C L CQ CA CA ~CA ~CA CA CA Ν Ν" c c ε ε "ε c c c g c c Β D c -c a g g -c "c a g g -c "c 2 Η Μ μ a g g g g Μ W W H g g M W W g 3 g W H to toa is ι f. Ε is is c . is is is is_ Ι - — is t js t . t Ε t J £ b « λ o. — c -

3 'ε1 CA ε Μ is

Ο ν— 2 ο W 3 CL ο. ο CA CA ε CA ε "ε Μ ν— α Um -t1-

* 3 ο ο -Ρ· CA Β Β η α is is

CA

4-·

^

_>

_>

"ca

Ν

£• η ο

α Ό ε α

σ ^ ε ο

β sν« S- Κδ ε Ο Ο «) ω

.2 -S

r S ε 3

•SPs: * α ιΐ ε ε ε ε ε 3 3 3

α ε α

Ν

Β Β β n u n is is is

Η Η

§

•5 ε ο υ Co ε ο

ω > "Μ ε 8 is

•S· -ε S ε3 α

Andrejs Veisbergs Euphemisms in General Monolingual Dictionaries

1. Euphemisms

Post-modern society produces a multitude of euphemisms. New power to euphemism production is provided by political correctness (Burridge 1996) and various "isms"(starting from ableism and ending with weightism). The widespread character of this phenomenon was confirmed by the coinage of a new term - euphemantics (Dodd 1962) in the 1960s. Many books (Pei 1970; Enright 1985; Allan 1991) and dictionaries (Spears 1981; Neaman & Silver 1983; Green 1984; Holder 1987; Rawson 1989; Ayto 1993) have been dedicated to the subject. Euphemisms are emotionally neutral words or expressions used instead of synonymous offensive, too direct or unpleasant words. Thus they deal with the ambiguous and the unpleasant. Traditionally euphemisms concentrate in the spheres connected with the human body, secretion, nakedness, genitals, sickness, but today they actively invade new spheres crime, military action, advertising, business, politics - anything "from private pleasure to public pain" (Rawson 1981). Like slang, most euphemisms are ephemeral (there is, however, the hard set which is rather stable). One of the causes of a rather rapid change of euphemisms is that after a certain period of use they tend to acquire the negative meaning of whatever they refer to and become contaminated and cumbersome; new ones are then sought. Hence the ever changing vocabulary items for bad/disagreeable/taboo concepts (Hughes 1991:13-15). Hence the 2000 words for "prostitute" in English (Allan 1991; Burridge 1996). Hence a multitude of present-day inventions for lay-off a few of which can be seen here: workforce adjustment, payroll adjustment, census reduction, negative employee retention, vocational relocation, career change opportunity, voluntary termination, involuntary termination, deselection, permanent downsizing, repositioning, restructuring, reshaping, realigning, reducing duplication, releasing of resources, dehiring, surplussing, excessing, redundancy elimination. This in turn affects the general usage as well, thus intercourse has almost disappeared from normal conversation because of its euphemistic sexual connotations, and gay can hardly be used in its primary meaning when referring to cheerful humans anymore.

304

Andrejs Veisbergs

2. E u p h e m i s m Creation Techniques

Euphemism creation techniques involve some of the traditional word-formation types as well as presenting some peculiarities. The techniques are broadly similar in most languages (Veisbergs 1997) yet we will operate here with English: • Loans borrowed mainly from Latin and Greek seem less offensive. They are more technical, sophisticated, longer and sometimes the meaning is not immediately apparent: to micturate (to piss), rebate (bribe), senior (elderly), sub-optimal (failed). • Abbreviations seem less dangerous or impolite: WC (toilet), KIA (killed in action), big C (cancer), Ο (orgasm). • Adaptations (distortions) - a specific word-formation pattern of euphemization: cripes (Christ), gosh (good gracious), Fanny Adams (fuck all), eff (fuck). • Truncation - deletion of some letters in writing is less widespread today than it used to be. Asterisks, or hyphens are usually used: G-d (God), F**k (fuck). • Widening of meaning - a word (usually a semantically more general superordinate term) is found and its meaning widened to include the meaning of the avoidable word. In extreme cases we deal with complete substitution of denotative meaning: growth (cancer), relationship (affair), inmates; clients; residents (prisoners), assets (enemy targets), manhood (penis). • Metaphoric transfers: blossom (pimple), theatre of operations

(battlefront).

• Ellipsis is used for euphemization as it transfers the meaning of the phrase onto another word, which is not directly associated with the avoidable subject: ladies (ladies' room), action (military action), intercourse (sexual intercourse), remains (mortal remains). • Use of negative. There is a distinct tendency in many languages to use negative prefixes for softening the effect of the word or making it extremely vague: the underprivileged (the poor), to disimprove (to make worse), to deselect (to exclude). • Longer periphrasis, where some lexemes are gaining semi-affix status: differently abled (disabled), visually impaired (blind), physically different/challenged (crippled), less developed (underdeveloped). There seem to be waves of euphemization when either a particular type of euphemism creation is enhanced (e.g. components -abled, -different) or a sphere of human activities (trendy movements, war, invalidity) undergoes serious euphemization. Thus, real war simulated the vocabulary of the technology of illusionary entertainment (Ebo 1995): surgical strikes (precision bombing), ordinances (bombs), to hit the jackpot (to hit a big target), involuntary conversion (crash landing). Here euphemistic use often borders on intentional blur, obfuscation and politically correct language, the boundaries of which overlap and are difficult to determine. Euphemisms in a way have a referent for which there are usually several synonyms. Yet their specificity is not determined by knowledge of some technical or professional domain but by the current agreement/vogue in the language - another peculiarity common to both. Yet another shared feature is the fact that, like slang, most euphemisms tend to disappear or move into colloquial and neutral layers.

Euphemisms in General Monolingual Dictionaries

305

It should be mentioned that here we do not consider situational/textual and individual (as well as ironic) euphemistic senses that can be easily created, but words/senses that have a relatively established character. Thus from a lexicographic viewpoint euphemisms are either 1) specific words/word combinations having euphemistic meaning; or 2) euphemistic senses in polysemous words or phrases from the general lexicon.

3. Euphemisms in Larger Dictionaries

Apart from specialised euphemism dictionaries which are not the subject of this paper, euphemisms do appear in general dictionaries. Analysis of various general monolingual dictionaries suggests that 1) there seems to be no consensus on the nature of euphemisms among lexicographers; 2) often there is no strategy of euphemism reflection/treatment (labelling, period of use), 3) it seems very difficult to draw the boundaries between euphemisms and politically correct items as well as deliberate blurs, also between euphemisms and stable jocular expressions. These conclusions seem to apply to monolingual dictionaries of English, German, Russian and Latvian, however, due to limited space we will focus on English dictionaries. Euphemisms have not been thoroughly discussed in lexicographic literature - they are mentioned in few books on lexicography, e.g. in Benson (1986), and no separate article is dedicated to them in the Encyclopaedia of Lexicography (Dictionaries 1989), which covers most of the issues connected with lexicography. The Encyclopaedia actually lists only 3 references to euphemisms in the index. A question might be asked - why a phenomenon so widespread and popular in practice, so much talked and written about in linguistics, has attracted so little specific lexicographical interest/reference? This question we will try to answer by looking at various dictionaries - samples of lexicographical practice. Theoretically, large-size historical dictionaries, such as the Oxford English Dictionary could document use and timing of euphemisms extensively, e.g. eff [Variant of EF, name of the letter F, euphemistically representing FUCK v. 2, 3.] 1. (Used as an expletive on its own account, as a milder alternative to the full form of the word fuck, or else as a euphemistic report of an actual use of the full word.) 1950 HEMINGWAY Across the River & into the Trees ix.78 'Eff Florence,' the Colonel said. 1958...

However, few euphemisms receive a thorough treatment. The OED actually provides relatively few euphemisms, - the latest online version, in fact, offers only 278 cases! Moreover, the descriptions are not always exhaustive, e.g. massage - used euphem. for: sexual activity. The expressions massage establishment, parlour etc. used as signs outside a building are frequently alleged to mean 'brothel' bathroom - a room containing a bath and often other toilet facilities. Hence also euphem. for lavatory

306

Andrejs Veisbergs

senior citizen - a term for an elderly person, esp. one who is past the age of retirement, orig. US. Frequently used in official communication and by the media as a euphemism for 'old-age pensioner'.

There are numerous other terms that are elaborated without the use of the label "euphemism", though the definitions are clearly euphemistic. These have often specialised labels, eg· loins 2. Chiefly Biblical and poetic. This part of the body regarded a) as the part of the body that should be covered by clothing and about which the clothes are bound b) as the seat of physical strength and of generative power possess spec, to have sexual intercourse with (a woman) knowledge sexual intimacy. Now only carnal knowledge (archaic and legal)

The problem is addressed in Berg's guide to the OED: "The OED confines its occasional use of the term to obviously oblique references to such things as sexual intercourse (a fate worse than death), or criminal behaviour ... or unpleasant truths... . As well as occurring as an italicized label, words 'used euphemistically' sometimes preface a definition involving an ironic reversal of meaning..." (Berg 1993:121). Naturally, one of the problems is the ephemeral, transitory nature of many euphemisms, e.g. gay has undergone various meaning changes and ranking within the entries, where sense development from the euphemistic "immoral" through slangy and then colloquial and finally neutral "homosexual" can be observed. It must be said that the use of gay meaning homosexual is much older than most dictionaries might suggest. (There is a citation of gay in this sense by Gertrude Stein in Vanity Fair of 1922 (Lighter 1994:871).) Gay 2a. Addicted to social pleasures and dissipations. Often euphemistically: of loose or immoral life b. hence in slang... c. of a person: homosexual. Of a place: frequented by homosexuals Slang (OED CD version)

In a more dynamic way the sense development and change of labelling can be seen in medium size dictionaries. Oxford Learner's Dictionary gay 2. (in a bad sense) immoral (Learner's, 1948) gay 4. (colloq.) homosexual (Oxford 3rd ed. 1974) gay 3. homosexual (Oxford 4th ed. 1989) gay 1. HOMOSEXUAL; of or for homosexual people (Oxford 5th ed. 1995) Concise Oxford Dictionary gay euphem. dissolute, immoral, living by prostitution (Concise 5th ed. 1964) gay 2. (euphem.) dissolute, immoral; (si) homosexual, frequented by homosexuals (Concise 7th ed. 1982) gay 4. colloq. a homosexual

Euphemisms in General Monolingual Dictionaries

307

(Concise 8th ed. 1990) gay 4. a homosexual; b intended for or used by homosexuals 5. colloq. dissolute, immoral (Concise 9th ed. 1995)

4. E u p h e m i s m s in Medium-size Dictionaries.

To get a more comprehensive picture of the situation four medium-size desktop dictionaries (The Concise Oxford, Oxford Advanced Learner's Dictionary, Chambers 20th Century Dictionary, Collins English Dictionary) were selected for a study. Generally very few euphemisms and euphemistic meanings were found to be included. This is no doubt due to the inability to decide on the semantic vagueness, longevity and spread of the concrete word/meaning. New euphemisms accordingly find it hard to get into dictionaries while old, archaic and rarely used euphemisms tend not to be eradicated from the following revised editions. The few euphemisms/euphemistic meanings that we encountered centred mainly in the sphere of sex, body, bodily functions, death - thus in the traditional, not to say oldfashioned, sphere. It should be mentioned that there are also dictionaries which do not use the label euphemism at all: COBUILD (general avoidance of labels), Webster's Dictionary (College edition). This may start a discussion on what sort of label euphemism is - is it a style or register marker? Can it be considered a positive expressive connotation and normsetting marker (Hansen 1985:194)? Or is it an evaluative label (Pueschel 1989)? How valid is the term itself? It seems to be important for learners, e.g. Longman Dictionary of Contemporary English (1995) (learner's type) does not give the label, but tries to suggest the connotations through the definition, e.g. Massage parlour - a word meaning a brothel (=place where people pay to have sex), used to pretend that it is not a brothel.

Collins English Dictionary (1995) does not suggest the use of the label euphemism, yet the word appears in quite a few lemmas. However, similar words and meanings are treated differently: Member - another word for penis Organ - a euphemistic word for penis Tool - Taboo, SI. another word for penis collateral damage - unintentional damage to civil property and civilian casualties caused by military operations.

Finally, to produce a more exact picture a list of 38 euphemisms was drawn up (the first part mainly containing the traditional euphemisms, the second part mainly euphemisms outside the traditional sphere). The euphemisms on the list were not all hard and fast cases as can be seen.

Andrejs Veisbergs

308 e word or meaning given as euphemism + word or meaning given without label "euphemism", "euphemistic" — word or meaning not given at all

eff Fanny Adams

OED e. e.

CONC. OXF. e. e.

CHAMBERS e. e.

massage (parlour) bathroom

e.

e.

—

e.

+'

cloakroom

e.

e. e.

+

OALD — —

COLL. e. e.

e.

e.

e.US

+US,C

e.Br.

e.Br.

~

+

adult

e.US

pass away

+

e.

+

sleep with/together

+coll. +

+

+

e. +

+

+

+

+

+

the remains depart

e.

+

+

+obs.

e.

one-night stand

+spec.

+coll-

+infml.

to relieve oneself prophylactic

+

+

+coll. +

e.

+infml. +

+

+US

+US

member posterior

+arch.

+

—

+US e.

+US +

+coll.

rest-room

+S1. +

+joc. e.

+

—

+

-

love-child

+

+

+ +

e.

e.

demise

+

+

+

+fml.joc.

e.fml.

hostess

+popularly + + +boil.

+

-

-

+

+

+

+

friend fallen woman

+

—

-

—

—

+

—

—

—

+

possess know

+spec.

+

-

—

+

+arch. +arch.

+arch.

+arch.

—

+arch.

knowledge

+

blessed senior citizen gosh

e. e.US +

e.

+arch. e.

e.

+

-

+

+obs. e. e.Br.NZ

e.

+coll.

e.

e.

holiday

e.

-

—

help oneself

e. +

+

+

+

-

+

+

+

+

+

+

+

intercourse

disadvantaged third-degree (treatment) institutionalise

—

-

+

+

+

+

-infml. +

appropriate

-

+

+

+

+

four-letter word

+

+

+

+

pacification action dehire

+

—

—

—

—

+

+

+

+

+

+US

--

-

—

Euphemisms in General Monolingual

309

Dictionaries

This small study clearly illustrates the fact that there is no agreement among lexicographers on whether a word is a euphemism or not. The dispersal is quite strong. In a way the most outstanding result of the study was that only one word (blessed) was characterized as a euphemism by all 5 dictionaries - a word most of us would not consider the most likely candidate for a euphemism. As can be seen, few are marked as euphemisms, also the number of euphemism senses differs considerably (one could exclude OED from the count as the number of entries and scope is so different).

OED Concise Oxford Chambers OALD Collins Dictionary

Labels 10 8 3 11 10

Euphemistic sense 36 33 26 26 30

5. Euphemisms in Smaller Languages and Lexicography

English in a way serves as the embodiment of euphemism spread in the large languages. An interesting topic would be euphemisms in lesser used languages, e.g. Latvian. Having been under both totalitarian regimes of the 20th century one could expect a large amount of political euphemisms (translations from the dominant languages. This is certainly true and has also been studied (Veisbergs 1999). Yet in Latvian lexicography euphemisms are in an even more bizarre situation. While much could be said about the use of euphemisms (there are the folklore euphemisms, traditional euphemisms, political propaganda and modern ones) little can be said of dictionaries. They generally contain neither the political nor the body-oriented euphemisms - they are victims of prescriptive and purist dominated Latvian lexicography as it was until the 90s. There are in fact only two monolingual dictionaries of Latvian in the second half of the 20th century - a multi-volume one (LatvieSu 1972) and a single-volume one (Latvieäu 1987), thus both are products of the Soviet period. Only one political euphemism was found in the bigger dictionary: Element - a bad, evil person (Latvieäu 1972). And even here one could wonder whether the above is a euphemism or a dysphemism. It is quite amazing that there is no mention of such often-used (and not always sinister) phrases/words of the period as: baltie plankumi pärvietot padarit nekaitigu militärs contingent

'white spots' 'to resettle' 'to turn harmless' 'military contingent'

nasty episodes in Soviet history to deport to kill Soviet Army in another state

Andrejs

310 notikumi

ärpusreglamenta tiecibas auto aizdzlsana atbrtvosana

'events'

at-

Veisbergs

'illegitimiate relationship'

events endangering the soviet system/lives, needing quick action, like events in Poland, events in Chernobyl. hazing

'driving away of car' 'liberation'

car theft, carjacking occupation

In a way the last euphemism was so often used by both occupying powers (also when speaking about the enemy's use of the word, when it would be put in inverted commas), that it certainly deserved a place in the dictionary. However on second thought one should be moderate in one's expectations and not ask the impossible. Expecting truthful explanation of these words in Soviet dictionaries would be the same as asking Nazis to explain the final solution. A modern dictionary of Latvian would have to include numerous units of the new fields where euphemisms are rife today - economic crime and citizenship issues, e.g. a multitude of terms for non-citizens (where one can see quite many euphemism formation types): nepilsoni ärpilsopi nelatviesi ärvalstnieki bezvalsnieki nevalstnieki krieviski runäjosie latviski nerunäjosie migranti imigranti apatrldi

non-citizens those outside citizenship non-Latvian foreigners stateless those without state Russian speaking non-Latvian migrants immigrants stateless.

In conclusion it can be stated that the study produced more questions than answers which suggests that the problem of euphemism treatment in dictionaries deserves more attention than it has received so far.

References

a) Dictionaries CHAMBERS 20TH CENTURY DICTIONARY ( 1 9 9 5 ) : C h a m b e r s .

COLLINS ENGLISH DICTIONARY (1995): 3d Edition, Harper Collins Publishers. THE CONCISE OXFORD DICTIONARY OF CURRENT ENGLISH ( 1 9 6 4 ) : 5TH ed. O x f o r d : C l a r e n d o n P r e s s . THE CONCISE OXFORD DICTIONARY OF CURRENT ENGLISH( 1 9 8 2 ) : 7TH ed. B o m b a y : O U P . THE CONCISE OXFORD DICTIONARY OF CURRENT ENGLISH ( 1 9 9 0 ) : 8TH ed. O x f o r d : C l a r e n d o n P r e s s . THE CONCISE OXFORD DICTIONARY OF CURRENT ENGLISH ( 1 9 9 5 ) : 9TH ed. O x f o r d , O U P LATVIESU LITERARÄS VALODAS VÄRDNTCA ( 1 9 7 2 ) : R i g a , Z i n ä t n e .

LATVIE5UVALODASVÄRDNICA(1987): Riga: Avots. OXFORD ADVANCED LEARNER'S DICTIONARY ( 1 9 7 4 ) : 3d. ed. O x f o r d : O U P .

Euphemisms in General Monolingual Dictionaries

311

OXFORD ADVANCED LEARNER'S DICTIONARY ( 1 9 8 9 ) : 4TH. e d . O x f o r d : O U P . OXFORD ADVANCED LEARNER'S DICTIONARY ( 1 9 9 5 ) : 5TH. e d . O x f o r d : O U P . OXFORD ENGLISH DICTIONARY ( 1 9 8 9 ) : 2 n d . e d . O x f o r d : O U P . LONGMAN MODERN ENGLISH DICTIONARY ( 1 9 7 6 ) : L o n g m a n . LONGMAN DICTIONARY OF CONTEMPORARY ENGLISH ( 1 9 9 5 ) : 3 d . ed. L o n g m a n D i c t i o n a r i e s . A LEARNER'S DICTIONARY OF CURRENT ENGLISH ( 1 9 4 8 ) : O x f o r d : O U P .

b) Other literature

Allan, K., Burridge, K. (1991): Euphemism and Dysphemism: Language Used as Shield and Weapon. New York: OUP. Ayto, J. (1993): Euphemisms. London: Bloomsbury. Benson, M., Benson, E., Ilson, R. (1986): Lexicographic Description of English. John Benjamins. Berg, D.L. (1993): A Guide to the Oxford English Dictionary. Oxford: OUP. Burridge, K. (1996): Political correctness: euphemism with attitude. In: English Today. Vol 12 No.3, 42-49. Dodd, A. R. (1969): Euphemantics. The new language of Science. In: ETC, A Review of General Semantics. San Francisco. Vol. 26. No. 2. 236-240. Ebo, B. (1995): The Gulf Conflict and the Technology of Illusionary Entertainment. In: Journal of American Culture. Vol. 18.3. Fall, 19-25. Enright, D. J. (1985): Fair of Speech: The Use of Euphemism. Oxford: OUP. Green, J. (1984): Newspeak: A Dictionary of Jargon. London: Routledge & Kegan Paul. Hansen, B. et al (1985): Englische Lexikologie. Leipzig: VEB Holder, R. W. (1987): A Dictionary of American and British Euphemisms. Bath: University Press. Hughes, G. (1991): Swearing. Oxford: Blackwell. Lighter, J. E. (1994): Random House Historical Dictionary of American Slang. Vol 1-2. New York: Random House. Neaman, J.S. & Silver, C.G. (1983): A Dictionary of Euphemisms. London: Hamish Hamilton. Pei, M. (1970): Words in Sheep's Clothing. London: Allen and Unwin. Pueschel, U. (1989): Evaluative Markierung im allgemeinen einsprachigen Wörterbuch. In: Dictionaries, Wörterbücher, DictionnairesA-3. Berlin, New York: Walter deGruyter, 693-699. Rawson, H. (1981): A Dictionary of Euphemisms and Other Doubletalk. New York. Crown Publishers Inc. Rawson, H. (1989): Wicked Words. New York: Crown Publishers Inc. Spears, R. (1981): Slang and Euphemism. Jonathan David Publishers, Inc. Veisbergs, A. (1997): Euphemisms in Latvian and English. In: Contrastive and Applied Linguistics 7. Riga, University of Latvia, 36-46. — (1999): Politkorekcijas. In: Rigas Laiks 9. 17-19.

Hideki Watcmabe

A Thesaurus of Old English Revisited

1. Introduction

Since A Thesaurus of Old English (hereafter referred to as TOE) was published in 1995, it has received less attention than it deserves. It is important as a pilot study for the publication of The Historical Thesaurus of English (hereafter HT) to appear in a few years' time. It is, moreover, important as the first period thesaurus of its kind. The editors of HT intended to "provide a new kind of source material for the study of English vocabulary" by incorporating "[d]ata from OED and from A Thesaurus of Old English" which are presented in the ordered semantic categories."' The few reviewers of the thesaurus, though they unanimously admit that this will be an indispensable "companion to Old English lexical studies",2 enumerate limitations thereof. Unless those limitations are recognized, and then overcome or reduced, TOE would not expand its readership to remain only serviceable to a very limited group of specialized readers, and HT itself would directly be influenced by those limitations. Now that TOE "has almost sold out its first edition, and a second is being planned," 5 1 have rewritten my review to incorporate or refer to the points discussed by the reviewers and appraise it once again, with the acknowledgements properly due to the most important contribution to Old English word study for the last decade of the twentieth century and with the hope as well that the editors would revise it in its second edition.4

2. Review of the Reviews

One of the editors of TOE (Jane Roberts) extracted about 48,000 word senses, mainly from A Concise Anglo-Saxon Dictionary edited by J. R. Clark Hall and Bosworth-Toller's An Anglo-Saxon Dictionary and with this "separable body of ancillary materials" apart from the materials from OED, they began to compile TOE as an independent work and as "a pilot thesaurus for" HT. So every reader of TOE should first note that "the TOE essentially

1 2

' 4

Christian Kay (1999: 8). Richard Dance (1997: 312). Christian Kay (1999: 9). In the 16th Conference of The Japan Society for Medieval English Studies, December, 2000, I chaired a symposium "Λ Thesaurus of Old English and Medieval English Studies" with four speakers. We talked about TOE in the history of thematic dictionaries, touching on research on binominals, double or multiple glosses, and hapax legomena in Old English literature.

314

Hideki Watanabe

reflects the information to be found in the Bosworth-Toller and Clark Hall dictionaries (TOE, p. xviii)."5 The first review of the thesaurus was made by Richard Dance on Medium /€vum in 1997. This is very short but very aptly summarizes what has been achieved and left to be desired as well. He puts in caveats against the arrangement of items of different dialects and/or origins without any label under the same heads and "the shortage of indication of its entries' original contexts (p. 313)".6 In the following year, two substantive reviews of the thesaurus appeared: one was by me in Studies in English Literature and the other by Manfred Görlach in Anglia. Mine appeared half a year earlier but for the sake of reference and coherence I sum up Görlach's discussion first. Görlach, who first rates the thesaurus as "the most important contribution on OE studies for years (p. 398)", goes on to present four main "limitations (p. 399)": 1) "The 18 categories [as main divisions] chosen are very different status, their sequence is not easily explained and the length of the chapters varies great deal" (See below); 2) "Locating concepts can be very time-consuming"; 3) "Most entries are organized under nouns"; 4) "The editors do not even discuss whether their system is based on semantic or encyclopedic/notional categories, or whether such a distinction makes sense in a historical corpus (pp. 399-400)." Concerning this last point, I will make some remarks in rebuttal later on.

2.1. Eighteen main categories in A Thesaurus of Old English7 01 02 03 04 05 06 07 08 09 10 11 12 13 14 s

6

The Physical World Life and Death Matter and Measurement Material Needs Existence Mental Faculties Opinion Emotion Language and Communication Possession Action and Utility Social Interaction Peace and War Law and Order

The editors of HT have built up slips, each of which has a headword from the OED, a definition of a particular meaning, the period within which the sense was used, and the corresponding Roget number. Since the materials on Old English in The OED record words which occur in literature after AD 1150, the editors used standard Old English dictionaries to supplement the information... It was published before the completion of Dictionary of Old English (DOE)·, only A, /E, B, C, and D parts were available to the editors. See Watanabe (1998: 132-133). Examples of this are given in the section 4. Just after my review was published, Jane Roberts visited Japan to give several lectures. In a lecture given at the University of Tokyo she mentioned the three superordinate classes to the 18 categories in A Thesaurus of Old English: I External World (01-05), II Mind (06-11), and III Society (12-18).

A Thesaurus of Old English Revisited

15 16 17 18

315

Property Religion Work Leisure

In my review (1998a: 131-8) attention was focused on the treatment and classification of poetic words in TOE, where I showed some misplacements of figurative expressions and dispersion of the words which have the same referent, especially when they are poetic compounds serving as kenning. I also pointed out jumbled orders of sections and sensehierarchies in each section. These considerable reservations notwithstanding, this thesaurus of Old English has many admirable qualities. First, the two-volume format with the independent volume of Index is handsome and time-saving. Secondly, the four kinds of flags, especially Flag Q, which signals the dubious status of the form, are of much use; as I show later, this Q is very instructive when it informs us of dictionary ghosts and hyper-corrections, unreasonable emendations, and, sometimes, fabrications in the edited texts. These flags, according to the editors, reflect signs and abbreviations employed in Clark Hall.

3. Four Flags and Treatment of Hapax Legomena and Poetic Words

So I would like to start my discussion here by looking at the four kinds of flags again in order to show the influence Clark Hall exerted on TOE. The editors of the thesaurus set four flags, O, Q, P, and G: Figure 1 0 0 P G

infrequent (nonceword or hapax legomena) questionable or dubious (invented by editors, dictionary ghost, etc.) poetic exclusively or particularly used in glosses

Let me introduce here the various names of and epithets for individual months in Old English. If we remember the compound weodmonaö, literally "month of leaves", an epithet for August, we can start to locate the section for the names of individual months in the thesaurus by looking at this compound noun on page 1,499 of Part II. There we find that section 05.11.03.03.01.03 is designated for specific months. One subdivision of 05. Existence is 05.11 A time, period of time, where we finally have a collection of the names of individual months:

Hideki

316

Watanabe

Figure 2 January February March April May June and July June July August September October November December and January December

ianuarius, se aeftera geola februarius, solmonab hlyda, hrasdd 8 , hrebmonab, martius aprelis, eastermonab maius, brimeolce, brimilcemonab q liba, libamonanb q lyba, midsumermonab q , se aerra lyjja, seremonab q giululing 08 , iulius, lyba, maedmonabq, se aeftera lyba agustus, hasrfest, rugern 0 , weodmonab hasrfestmonab0, haligmonab, September october, octobermonab 0 , winterfylledmonab, winterfylleb blotomonab, nouember, nouembermonab geola december, decembermonab, geola, geolmonab, midwintermonab q se aerra geola

Although the names o f the months appear in clusters in several manuscripts, to the best o f my knowledge, there is no readily accessible dictionary or article that lists as many equivalents as this." But while w e are impressed by the richness o f the items, w e are at the same time dismayed by the frequent use o f the flag Q: primilcemonapq, lipamonanpq , midq q q sumermonap , seremonap , mcedmonap , midwintermonap A s w e are not informed of the "original contexts"' of these "dubious" words in TOE, w e have to consult Clark Hall, where they are rendered as follows in Figure 3: Figure 3 M o n t h s with Flag Q

Clark Hall's R e n d e r i n g s

The Menologium

brimilcemonab

May, MEN (v. CHR p276)

in marginal note

libamonanb

Not Cited

in marginal note

midsumermonab

June, MEN (Hickes)

No Occurrence

seremonab

June, MEN. [v. 'siere']

No Occurrence

maedmonab

July, MEN (Hickes)

No Occurrence

midwintermonab

December, MEN (Hickes)

NoOccurrence

The abbreviation M E N refers to a metrical calendar generally called Menologium at the end o f Chronicle, or in Hickes' Thesaurus."' The first and second words o f Figure 4, primilcemonap and lipamonanp, do appear as the marginal notes, "in a much later hand"", in the 8

As for the manuscripts or the editions of them containing the names of the months, see N. R. Ker, Catalogues of Manuscripts Containing Anglo-Saxon. (Oxford, 1957). Names of months are gathered in F. Wormald, English Kalendars Before A.D. 1000 (Henry Bradshow Society, 1934). ' See Richard Dance (1997: 313). "' George Hyckes, philologist and Anglo-Saxonist, Compiled Antiquae Literaturae Septentrionalis Libri Duo (Oxford, 1703-5), which is generally known as Hickes' Thesaurus. " L. Earle and C. Plummer, eds., Two of the Saxon Chronicles Parallel. Oxford (1892: 276).

317

A Thesaurus of Old English Revisited

poem. And the following four do not appear in the poem Menologium but in Hickes' Thesaurus, which was compiled in the eighteenth century; they are not attested in any work in the Old English period (up to 1150). That is why these words appear with the flag Q. (In Menologium there also occur forma monap (1. 9) for January and se teodamonp (1. 181) for October, both of which are not registered in TOE because they are not in Clark Hall.) I would also like to invite your attention to the difference, or in fact no difference, between the cases with the single flag Ρ and the double flag OP. See Figure 4, where I list those compounds with scur- or -fyr. Figure 4 Compounds with scur- as the first element in Clark Hall and TOE Clark Hall

TOE Sections

Flags

Frequency

scurbeorg

f. roof., Ruin 5

11.10.01.02

OP

xl

scurboga12

m. rainbow,

01.03.01.06.02

OP

xl

Gen. 1540 scurfah

rainy, stormy

01.03.01.06.02

xl

A9..369; MFH 172 scurheardt

made hard by

05.08

Ρ

x2

11.10.01.02

OP

xl

blows (epithet of a sword) scursceadu

f. protection against storms, Gen. 813

heafmfyrt

n. cruel fyre, B.

03.01.09.02.01.01

Ρ

x2

waelfyrt

n. deadly fire: funeral pyre, B125

02.02.04.01

Ρ

x2

It seems that in TOE the double flag OP corresponds to the headwords with an indication of single occurrence in poetry, such as the fifth line of the poem Ruin indicated for scurheard, in Clark Hall. But the single flag Ρ defies such an easy explanation; Clark Hall sometimes adds something to indicate place of occurrence to the daggered headwords. Compare scurheard, heapufyr, and wcelfyr. Actually these three compounds are almost exclusively used in Beowulf, each only twice." There seems to be no difference between the words which are represented once and those which are only twice represented, as far as poetic or metaphoric compounds go. Clark Hall explains in his list of abbreviations that "[References to page and line have, as a rule, been restricted to words only occurring once" and that "[a headword with] t [a dagger] occurs only or mainly in poetical texts."14 It is very misleading that the lexicographer does not clearly mention the frequency of those words with this dagger 12

14

The prosaic counterpart to scurboga seems to be renboga, which /Elfric used in his Genesis. The compound scurheard occurs in Beowulf 1033 and in Andreas 1133. See page ix.

318

Hideki

Watanabe

symbol. In fact, it is often the case with those "poetic" words with the dagger that they are very infrequently represented. This unclear status of the dagger symbol in Clark Hall is directly reflected in TOE with the flag P. We should be very careful when we say whether a certain word is poetic or prosaic. Let us here take the equivalents for the concept of fear for example. Among other synonymous words, simplex and complex alike, ege and egesa are most frequently represented in the corpus of Old English literature. The former occurs 842 times in prose and glosses and 21 times in verse and the latter occurs 117 times in verse and 88 times in prose and glosses.15 It is clear that egesa is poetic as against the prosaic ege, but these distinctive uses of the two are not in Clark Hall nor in TOE simply because their frequencies, occurring over a hundred times, conceal their stylistic characters. Compare those equivalents for fear in Figure 5. The compound gryrebroga occurs only twice in Old English poetry: once in Beowulf 2,227 and the other in Christ 848. As this infrequent poetic word is labeled OP, such semihapax legomena as scurheard and wcelfyr - words which appear very few times in a single work or those which are once represented in a certain work with a few reminiscent thereof in another - might as well be registered with the double flag OP, too, to have a different status than more commonly-used words in TOE. Anyway the reader of TOE should note that the items with the flag Ρ are sometimes quite like those with OP in their frequency and use. Figure 5 The Frequencies and Status of Equivalents for 'fear' in Clark Hall and TOE Word ege egesa broga gryre gryrebroga oga fyrhtu forhtnes forhtung

Prose and Glosses

Verse

Clark Hall

TOE

842

21 117 21 23 2 1 8 0 0

-

-

-

-

-

-

88 60 23 0 84 169 8 6

t t

-

OP

-

-

-

-

-

-

-

-

4. Separations of the Equivalents

As we have seen, scurheard is an epithet of a good sword and Clark Hall gives a parenthesized note to the headword. As the epithets for good swords, fyrheard and regnheard are 15

Watanabe (1993: 25).

A Thesaurus of Old English Revisited

319

also represented in Beowulf, any specialized reader would expect these three compound adjectives to appear side by side in a certain section in TOE, but, in fact, the latter two hapax legomena are registered separately under the head 03.01.03 Hardness. More strangely, of these three only fyrheard is chosen to appear under the head 13.02.08.04.03 Sword, despite Clark Hall's parenthesized note. Turning to the section 05.08 Strength, where scurheard appears, we are no less surprised to see there is a subsection for "strength of sword-edge", where it sits all alone without its legitimate mates, regnheard and fyrheard. To sum up the situation I would say that there are at least three possible places for adjectives denoting hardness of a sword in TOE: 03.01.03 Hardness; 05.08 Strength; and 13.02.08.04.03 A sword. In order to see how the equivalents for the same referent or one idea are dispersed like this, let us once again see the poetic compounds. I would like to draw your attention as I did in my review, to those with hild(e)- "battle" as the first element and those with -fyr as the final element.16 Clark Hall lists fifty compounds with hild(e)-, most of which seem to be hapax legomena with the signs for the location of occurrence (see pages 182-3). Let us first see a nominal compound hildeleoma 'gleam of battle' employed twice in the poem Beowulf, first referring to a sword in line 1,143 and later on in line 2,583 a fire-flame spewed by the dragon. TOE rightly puts the first case into section 13 under the title of Peace and War (13.02.08.04.03 A sword) together with other nominal expressions denoting the weapon, and the second into the section 03 Matter and Measurement (03.01.09.02.01.01 A kind of fire), with heapufyr 'battle-fire' {Beowulf2,547), which is a variation of dracan lege in line 2,549. Here the editors classified the word according to its contextual meanings. Put differently, they noticed that it has very different types of referents. See Figure 6. The compound hildeleoma, which in the first case serves as a kenning for a sword and is rightly classified, should have been associated with a group of metaphoric compounds with the head hilde-, including hildegicel 'battle icicle', hildencedre 'battle-snake', and hildescur 'battle shower'. These compounds for weapons are expected to relate to one another somewhere in the tree of categorical hierarchy; they should be treated separately with such nonfigurative expressions as hildemece, hilderand, hildetux or hildewcepen. But in fact the editors decided to put hildegicel into section 02 Life and Death (02.02.04.01 Cause/occasion of death), where we find another compound wcelfyr (This wcelfyr is represented only twice in Beowulf, to refer to the fire flame in 1,119 and the dragon's fire in 2,582). The other possible home for the second case of hildeleoma and headufyr, then, might be 'Fire-flame from a dragon' under the section Dragon (02.06.10.01.01), one subsection of which is actually allocated for Dragon's coil: hringgewindla. In the same way hildescur, now in 02.08.03, could be linked with hildencedre (13.02.08.04.04.01 An arrow, dart, bolt) or scur (13.02.08.04.02 A shower of missiles).

16

Watanabe (1998: 136).

Hideki Watanabe

320

Figure 6. The homes for the metaphorical compounds containing hild(e)- and —fyr Compounds

Sections in TOE

Proposed Places

hildeleoma (Beowulf 1,143)

13.02.08.04.03 A sword

-

hildeleoma

03.01.09.02.01.01

02.06.10.01.01

A kind of fire

Dragon

02.02.04.01 Cause/occasion of death

13.02.08.04.03 A sword

13.02.08.04.04.01

-

(Beowulf2,583)

hildegicel (Beowulf 1,·606) hildenaeder (Elene 119, etc.)"

An arrow, dart, bolt hildescur (Gu^lac heajjufyr

1143)

(Beowulf2,547)

02.08.03 Pain, bodily discomfort

An arrow

13.02.08.04.04.01

03.01.09.02.01.01

02.06.10.01.01 Dragon

A kind of fire waelfyr (Beowulf 1119)

02.02.04.01 Cause/occasion of death

waelfyr (Beowulf 25S3)

02.02.04.01 Cause/occasion of death

-

02.06.10.01.01 Dragon

As for the second instance of hildeleoma comparison with wcelfyr 'deadly fire' might make the case even clearer. See Figure 6 once again. This compound wcelfyr also appears twice in Beowulf with very different meanings and referents. It first describes the funeral pyre which burns Hnaefs body in line 1,119, and later refers to the fire spewed by the fearful dragon in the line 2,583. So the second case of wcelfyr is expected to sit alongside with hildeleoma somewhere in the thesaurus to be related with the concept of "fire" or "dragon". But the editors neglected or disregarded this metaphoric use, putting the former case in 02.02.04.01 Cause/occasion of death with the flag Ρ under the specific head Deadly flame and the latter in 03.01.09.02.01.01 A kind of fire. Thus the two poetic compounds sharing the same referent are separated. I would like to propose another home 02.06.10.01.01 Dragon for these words, for the editors collect such metaphoric compounds as hringboga, inwitgcEst, and lyftfloga, which do not literally or necessarily mean a dragon, but contextually denote the fire dragon in Beowulf.

17

Elene 119, 141; Judith 222.

A Thesaurus of Old English Revisited

321

Figure 7 Section 03.01.09.02.01.01 in TOE

Proposed Addition or Deletion 03.01.09.02.01.01

A kind o f fire: fyrcynn An immense fire: fyrbryne Cruel fire: hea^ufyr

Delete

Destructive fire: manbryne, teonleg Destructive flame (from a dragon): hildeleoma

+ heajjufyr, waelfyr

A burning o f a house: husbryne Holocaust: ealloffrung Fire for burning dead or living: ad Flame o f funeral fire: adleg, blwylm

+ wslfyr

I might conclude this section by suggesting that the three poetic compounds, each used only once or twice to contextually refer to the dragon's fire should be gathered together in Destructive flame (from a dragon) under 03.01.09.02.01.01 A kind of Are, where hildeleoma alone is (See Figure 7). If not, a new subsection should be set up, possibly as Fireflame from a dragon under 02.06.10.01.01 Dragon.

5. The Full Range of Synonyms as a Companion to Old English Lexical Studies"

In his review of TOE Görlach regrets that "the reasons given for the classification used in the thesaurus are disappointingly brief' thereby making it "impossible to use their material to distinguish between notional and semantic developments in English (1998: 400)". But, on the contrary, I would like to encourage the students of OE lexical studies to make full use of their lists of equivalents. In my review I wrote, "if we want to see a group of verbs or idiomatic phrases meaning 'to die', we could either start from a familiar word such as steorfan or sweltan in the index, or go to Class II. Life and Death and read through the layers of subdivisions. Then we find that in section 02.02.03 To die, perish six common verbs are listed: acwelan, aswindan, gefeallan, forswealtan, forweordan and (ge)losian. These are followed by ten groups of synonyms classified according to their original senses. The classification here, apparently based on the fruits of vocabulary studies in Old English, is highly informative and provides us with new insights. The largest group, with the heading ..Involving journey, includes eighteen verbs, of which only one entry, forsidian, is flagged OP; it is clear how far and wide periphrasis and metaphor of this kind go. The group with the heading ..Involving separation from the soul lists three idiomatic phrases, gast agiefan, gcest ofgiefan, and gast onsendan set by sawlian. (Watanabe (1998a: 133-4))". Here the items are elegantly classified according to their figurative or extended senses. Of course that is not always the case with other sections; we should be very careful when we read through those lists, because, as the reviewers note, no indication is made among the 18

See Richard Dance (1997: 312).

Hideki Watanabe

322

words of different origins, dialects, or periods, and "[n]o attempt... is made to highlight the terms of greatest frequency within any sense-group"." There is no knowing, for example, which is the best represented word among the entries in each section for Face 02.04.03.01.03: andwlita, ansien, hleor, mud, neb(b), nebwlite, and wlita. Nor do we know whether they are readily interchangeable in any context. The list, however, poses the question why Jember in his concise English-Old English Dictionary or Buck in his copious Selected Synonyms in the Principal Indo-European Languages chose hleor and wlita and andwlita and ansien, respectively. Figure 8 Equivalents for 'face' in TOE, Buck (1949), and Jember (1975) TOE·. Jember: Buck:

andwlita, ansien, hleor, mud, neb(b), nebwlite, wlita hleor, wlita andwlita, ansien

It is instructive to browse the entries here in TOE because we can easily see that those words for 'face' are clearly divided into two groups: derivatives from verbs meaning 'to see, gaze' and the names of the parts of the face. Those included in the second group, hleor 'cheek', mud 'mouth' and neb(b) 'nose' come, through the process of metonymy, to enlarge their sense-sphere to denote 'face', their superordinate word; they could possibly be registered once again under a certain subsection of Face which collects such words as to denote 'face' in metonymic use, as the editors actually did for the equivalents for 'to die'.

6. T r e a t m e n t of Set Phrases and Idioms in the Index

Finally, I would like to see the treatment of set phrases and idiomatic expressions in the Index, which has not been discussed by any reviewer. As I mentioned, the two-volume format facilitates quick reference, but it is clearly meant for very specialized readers with extensive knowledge of Old English. In the Index there appear both simplex and complex forms with sporadic registrations of set phrases consisting of two or more words.

" See Watanabe (1998: 134).

A Thesaurus of Old English Revisited

323

Figure 9 Treatment of Set Phrases in the Index of TOE Registered Phrases

No Citation

'to (begin to) speak'

wordum

wordhord

'on earth, in the world'

undersunnan

under

under wolcna hrofe

underheofenes

under

under swegle

wrixlan

wolcnum

onlucan

heofenum

under swegles

hador/hwealf begong

While wordum wrixlan 'to converse' (p. 1529) is chosen, wordhord onlucan 'to begin to speak' is neglected after wordhord (p. 1528). The editors like under sunnan, under wolcna hrofe, and under wolcnum to be in the Index, but not under heofenum, under heofenes hador/ hwealf, under swegle, and under swegles begong. After such prepositions or adverbs of frequent occurrence as a (p.717), in (p. 1113), on (pp. 1225-8), to (pp. 1383-5), ut (pp. 1474-6), etc., there are long lists of set phrases, which, of themselves, are very informative for the students of Old English idiomatic expressions. Any reader who knows several phrases for 'for ever' in Old English will be impressed to see the items registered in the Index, for example, a butan ende after a, atfre to ealdre after atfre, or to ealdre, to ecum pingum, to widan ealdre, to widan feore, and to worulde after to. And of course ealne widan feorh is registered after eal.2" Thus we can collect set phrases and idiomatic expressions reading through the Index volume.

7. Conclusion and Call for Revision

Thus far I have shown that TOE in this format could only claim very limited readership. Görlach jocularly notes, "[f]or experts" the thesaurus will "make fascinating browsing", but "for other less specialized readers it will have... the interest of a telephone directory (1998: 398)". But I seriously doubt that those "less specialized readers" could even reach the items they need in the thesaurus just with the Index volume. To conclude my discussion here, I would like to propose the following revisions for the second edition of the thesaurus:

2

" Interesting citations of the idiomatic or metaphoric set phrases for the headwords of the Index include: agan wcelstowe gewald (Victory), bam handum twam (Zeal), clcene bcec habban (Truth of speech or thought, etc.), gefeallan on lufe (A loving relationship), heafdu up ahebban (Courage, boldness, valor), gehieran Godes dreamas (The service of the mass), in heortan and heortan (Deceitfulness, falseness, duplicity). Other set phrases or binominals cited include: gemengan on f>tem hcemede (Sex, generation), humbre streames gemcere (England), leofan men (Courteous forms of address), ban and fell,flcesc and fell, and ban, humbre streames gemcere (England), oft and gelome (Frequent, of common occurrence).

324

Hideki Watanabe

1) The flags and status of the items which appear as the headwords with a dagger ( t ) in Clark Hall should be double-checked; 2) Compounds containing the same initial or final element should be gathered once again for (more) proper placement and classification; 3) The treatment of set phrases and idiomatic expressions in the Index should be unified. With these in mind, the reader of TOE would always have to use Clark Hall as a supplement.

References

Buck, Carl Darling, (ed.) (1949): A Dictionary of Selected Synonyms in the Principal Indo-European Languages, (reprinted., 1988): The University of Chicago Press. Clark Hall, J. R., (ed.) (1894): A Concise Anglo-Saxon Dictionary. Cambridge University Press. (4th Edition enlarged by Herbert D. Meritt, 1960): Yale University of Toronto Press. Dance, Richard (1997): Review of A Thesaurus of Old English. Medium Mvum 44, 312-313. Earle, J., (ed.) (1892): Two of the Saxon Chronicles Parallel. Oxford. Frank, Roberta, and Cameron, Angus, (eds.) (1973): A Plan for the Dictionary of Old English. University of Toronto Press. Godden, Malcolm R. (1992): Literary Language. In: The Cambridge History of the English Language. 490-535. Cambridge. Görlach, Manfred (1998): Review of A Thesaurus of Old English. Anglia 116, 398-401. Jember, Gregory K., (ed.) (1975): English-Old English, Old English-English Dictionary. Colorado: Westview Press. Kay, Christian J. (1999): The Historical Thesaurus of English: A Millennium Report. In: Medieval English Studies Newsletter, 41, 8-10. Tokyo. Ker, N. R. (1957): Catalogues of Manuscripts Containing Anglo-Saxon. Oxford. Roberts, J. and Christian Kay with Lynne Grundy, (eds.) (1995): A Thesaurus of Old English. King's College London Medieval Studies XI. London: King's College London. Watanabe, Hideki (1993): Some Neglected Aspects of Meaning of the Old English Noun-Verb Combination egesa stod. Studies in Medieval English Language and Literature, 8, 25-37. Tokyo: The Japan Society for Medieval English Studies. — (1998a): Review of A Thesaurus of Old English. Studies in English Literature, English Number 1998, 132-138. Tokyo: The English Literary Society of Japan.

Arne Zettersten/Hanne Lauridsen From Projection to Reception On the Process of Bilingual Dictionary Making

The bilingual dictionary project referred to here was concluded in 1999, and we usually explain to people that our two-volume English-Danish Dictionary took six years to complete. However, when contemplating the whole process of production, we realize that we must take into account the life before as well as after the actual composition of all the entries of the dictionary. The embryo of the concept of the English-Danish Dictionary was created in Reykjavik, Iceland on 20 June 1992, when John Sinclair, the founder of the COBUILD Project and the Bank of English at Birmingham University, and Arne Zettersten met at the Nordic Conference of English Studies. The first COBUILD English Language Dictionary had been published in 1987 and the Student's COBUILD Dictionary had been finished in 1988. John Sinclair's vision of a series of bilingual European dictionaries based on a COBUILD dictionary captivated Arne Zettersten and the idea was brought forward to Politikens Forlag in Copenhagen, a publishing house experienced in dictionary making, particularly in the case of dictionaries of contemporary Danish. The discussions of the new concept were carried out at the publishers in 1993 and 1994 before the real practical work started. Before we discuss the various analyses that had an effect on the process, we should like to summarize the whole process in the following points: 1. The dictionary is based on the entries, definitions and authentic examples of the second edition of the Collins COBUILD English Dictionary of 1995. 2. The Bank of English at COBUILD, Birmingham, contributed about two hundred million words of English text for the work on the second edition. 3. By courtesy of HarperCollins, London, and COBUILD, Birmingham, the 1995 database was brought over to Copenhagen, where we set up a new bilingual database of our own at Politikens Forlag (= the publishing house owned by Politiken, one of the leading newspapers in Denmark). 4. Owing to the fact that the COBUILD dictionary was corpus based, we could draw on the vast collection of complete sentences. Therefore our bilingual dictionary contains circa 100,000 examples, illustrating the meanings and sub-meanings of the headwords. 5. For each meaning and sub-meaning we have provided Danish equivalents, sometimes up to six equivalents. 6. One other major feature of the dictionary is the fact that the definitions of all meanings of each headword are written in the characteristic interactive COBUILD style in full English sentences. 7. After the Collins COBUILD English Dictionary was published in 1995, the COBUILD database, The Bank of English, continued to grow. By 1996 it had grown to 323 million words.

326

Arne Zettersten/Hanne Lauridsen

8. After 1995 we have been able to supplement the COBUILD vocabulary with new data, particularly within areas like electronics, new media, business and politics. 9. All the 100,000 examples in the COBUILD Dictionary (1995) were equipped with grammar tags indicating the grammatical build-up of most examples. We translated all these word-classes and grammar patterns into Danish, which means that all the 100,000 examples in English can be used as model sentences, all grammatically analysed in Danish codes. This information provides the user with a new English grammar in the form of complete model sentences, all grammatically coded and equipped with relevant word definitions and equivalents. 10. Since a complete set of equivalents is given as well as a complete analysis of syntactic constructions expressed in Danish, we believe that maximum information on all Danish equivalents of all English entries will be obtained. This will be even more obvious from the fact that ample indications of pragmatic information have been included in the dictionary. At the beginning of our planning period back in 1993, we had as our starting point the Guidelines for the Bilingualization of COBUILD Dictionaries, as they had been formulated by the COBUILD Project, Birmingham. These guidelines outline a number of overall policies and procedures which were meant to be taken into consideration when preparing bilingualized versions of COBUILD dictionaries, also called Bridge Bilingual Dictionaries. These dictionaries should - according to COBUILD - have a number of distinctive features; these features are found in the Guidelines for the Bilingualization of COBUILD Dic-

tionaries. The most important feature is - according to the Guidelines - that a bridge dictionary provides definitions of the English headwords in the user's mother tongue. The principle is that an item which is assumed to be unfamiliar is best explained in terms of what is assumed to be familiar to the user, the notion being that for a large number of learners worldwide the move from using a traditional bilingual dictionary to using a monolingual dictionary of the foreign language can be difficult to cope with. The headword (word or expression) being explained should be inserted in English in the body of the definition/explanation, with enough context to show its syntactic and lexical patterning. In the definitions, the headword and any other text retained in English should be highlighted in bold in order to provide a visual distinction between English and the language into which the definitions are translated. The Guidelines offer a list of main elements, such as the indefinite article, possessive adjectives and reflexive pronouns, which should be kept in English in order to help the learner decode and encode English text successfully. In exceptional cases the definition may be superfluous, especially if the English headword has a straightforward ready equivalent in the target language. The point of view is that the users should not be irritated by lengthy explanations for items which they are perfectly familiar with. This is the most important feature. Of other distinctive features mentioned in the Guidelines, it is relevant to note: Grammatical terms should be left untranslated. Style labels (e.g. formal, informal, etc.) may or may not be translated.

From Projection to Reception - On the Process of Bilingual Dictionary Making

327

The Guidelines suggest that certain types of entry might be omitted in order to save space, for instance low frequency words, old-fashioned and literary words, offensive words. The Guidelines do not recommend a special COBUILD Dictionary as a basis but state that the choice of dictionary, Collins COBUILD English Language Dictionary (1728 pages), Collins COBUILD Essential English Dictionary (960 pages) or Collins COBUILD Student's Dictionary (704 pages), on which to base a bilingual version will depend on the range of prospective users and the nature of the market. The first and most important point to us was whether we should translate the definitions in accordance with the Guidelines, or whether we should retain the English definitions and offer a wide range of Danish equivalents. In the Guidelines two examples of the Bridge concept are given. These examples are from (a) a proposed Arabic Bridge Dictionary and (b) the Brazilian Portuguese Bridge Dictionary, published in 1995. Both bridge dictionaries are based on Collins COBUILD Student's Dictionary (704 pages). Our target groups were advanced learners of English (e.g. upper secondary education, universities) and advanced users of English (e.g. the business world). As a consequence and due to the level of competence in Denmark - we based our dictionary on the large COBUILD dictionary from 1995 (1728 pages). We suspected that our target groups would not accept definitions with a combination of an English headword and a definition in Danish - a definition which might sometimes appear naive when it was translated into Danish in the interactive, rather simple style, used by COBUILD. Furthermore, we thought the advanced user would appreciate having the grammatical terms in Danish (Latin) corresponding to the terminology normally found in Danish dictionaries. We also found that style labels were important information, especially for a foreign language speaker, and would definitely prefer to translate these terms. The Guidelines, however, were quite clear, especially on the first point, and we decided to work according to these Guidelines in what was to become a pilot project. With a fairly large group of highly qualified translators we made a first version of a small part of the dictionary. Thus this first concept of a bridge dictionary in Denmark was based on the idea that we would translate all the definitions into the target language, Danish. The next step was a professional evaluation of the pilot project. In collaboration with a commercial assessment company, the Copenhagen publishers, Politikens Forlag, carried out two evaluation programmes. Tests were carried out where extracts of the first version were shown to and evaluated by advanced learners and users such as teachers of English from schools, universities and business schools, students, representatives of commercial companies, and some general users. The result of this external quality and marketing analysis was clear - and in accordance with our expectations: due to the general level of competence of English in Denmark the test subjects did not accept the definitions where English headwords were combined with Danish definitions. They found them artificial, and in many cases naive. Another important aspect, which was discussed in these evaluations, was the language and exhaustiveness of the grammatical information. The test subjects preferred Danish/Latin grammatical terminology and to have as much grammatical information as possible. On the basis of this professional evaluation, the results of which supported our own preferences, it was decided to rethink the editorial rules introduced by COBUILD. First of

328

Arne Zettersten/Hanne Lauridsen

all we decided to retain the definitions in English. One reason for keeping the COBUILD definitions in English is that the level of English proficiency in Denmark is generally rather high and typical COBUILD-definitions are usually composed in a defining style that is rather easy to follow. The defining vocabulary normally adds up to about 2,500 words in COBUILD. This is quite normal for the new monolingual dictionaries, The Longman Dictionary of Contemporary English, for example, having a defining vocabulary of circa 2,000 words. Another reason for keeping the definitions in English is the fact that important collocations can be studied by the user both in the whole-sentence definitions and the complete examples, which are usually two or three in number. For each meaning of each headword one to five Danish equivalents were to be given. Furthermore, we decided to translate the grammatical terminology into Danish/Latin so that it would correspond to for instance the terminology used in Politikens Nudansk Ordbog. This meant that more than 100,000 grammatical codes had to be translated into Danish. Since we regarded style labels as a very important piece of information for foreign language learners, we decided that the information about style, pragmatics, illocutionary meanings, and varieties of English should be placed in a conspicuous way in the entry. Instead of letting this information be part of the definitions, often indicated in a rather indirect way, we have placed the results of editing of information on style, pragmatics, illocutionary meanings, sociolects, and dialects in a special paragraph called BRUG (= usage) at the end of an entry. Instead of omitting certain types of entry we increased the number of entries, for instance by adding entries on modern technology and media as well as geographical entries relevant for Danes. As a result of the various stages in the creation of an English-Danish Bridge Dictionary, the whole process turned out to be a very long one. But the pilot project was not an unnecessary stage, a waste of time. On the contrary: it made it possible for others and for us to test the original concept and gave us the necessary support for making the relevant changes. Bearing in mind that some typologists discuss dictionaries in terms of 'active' and 'passive' ones, meaning that the former are 'encoding' (writing and speaking) and the latter 'decoding' (reading and listening), our own monolingual-bilingual English-Danish dictionary is certainly more 'active' than most standard bilingual dictionaries. With such a wealth of information on definitions, meanings, examples, collocations, synonyms, antonyms and pragmatics, the user will be in a good position to select the most satisfactory equivalent in Danish. In addition, the user's knowledge of English for productive purposes will - provided he or she reads and contemplates everything under each entry - increase considerably. Our dictionary, which we have advertised as "two dictionaries in one", fits very well into the category that Henri Bejoint calls "monolingual-bilingual" in his Tradition and Innova-

tion in Modern Dictionaries'. "Among the dictionaries for foreign students, one interesting recent development is the monolingual-bilingual dictionary with its source-language definitions and its equivalents in the target language added to each meaning. These dictionaries are important, because they go beyond the traditional distinction between monolingual and bilingual dictionaries. N o study of their effectiveness has been carried out yet, but it will be interesting to see how they perform, particularly in the comparative study of the effectiveness of various means of indicating meaning in a foreign language,

From Projection to Reception - On the Process of Bilingual Dictionary Making

329

such as definitions and equivalents. These dictionaries deserve more study than they have received so far." (Bejoint 1994, p. 87) True enough, such dictionaries deserve more study, but there is also a great need for dictionaries that are similar to the monolingual-bilingual ones. We have come to realize that the Kernerman Semi-Bilingual Dictionaries are fairly close in principle to the ones described by Bejoint. Kernerman has published several so-called Password English Dictionaries based on Chambers Concise Usage Dictionary. There are such Password Dictionaries for speakers of Croatian, Lithuanian, Thai, Korean, Norwegian, and other languages. The Kernerman Password Dictionaries provide definitions and examples in English as well as equivalents in the target language. However, the definitions are much more condensed than the ones based on the COBUILD English Language Dictionary and they are not given in the same complete sentences. The examples provided are not as many and diversified, but the principle is nearly the same. Kernerman has also produced the corresponding English-Swedish Dictionary called The Ideal English Dictionary for Swedish Speakers. (Akademiforlaget Corona, 1999). Kernerman has further produced semi-bilingual dictionaries for languages such as Hebrew, Arabic, Italian, Greek, French, Polish, Spanish, Portuguese as well as Chinese and Swahili, and several others. In addition to realizing that our use of a large database like the Bank of English made it convenient to work in the direction of the proper monolingual-bilingual dictionary, we found that many users wrote to us afterwards explaining that they used the dictionary for English production purposes. We had certainly explained in advance that collocations and model phrases, idioms and complete English sentences could easily be found in the dictionary, but we did not know that these aspects would turn out to be so easily understood by users. In fact, it dawned on us slowly that we had to some extent produced three dictionaries in one, a monolingual COBUILD dictionary, an English-Danish one and in addition, a production dictionary, without actually calling this a Danish-English dictionary. The monolingual-bilingual dictionaries discussed above are virtually the same as those described by Bo Svensen in his Practical Lexicography (1993). He regards them as "mixed monolingual and bilingual" (p. 21) and explains that such dictionaries have begun to appear in Sweden as part of the LEXIN project ('Lexikon för invandrare'= dictionaries for immigrants). Svensin also suggests that dictionaries of this kind should be "supplied with a single index providing references from words in the native language to the entry words in the foreign language" (p. 13). We intended first to co-publish a Danish-English word list (index) at the end of Vol. II of the dictionary but realized that it would not be a complete dictionary and would in many cases be misleading to users. Instead we suggested the possibility of using the CD-ROM version for that purpose. The user can click on Danish words and get into equivalents for further selection, so the opportunity is there for the creative user. Since we used the large COBUILD version, we have so far an advantage over similar projects in other languages. Whatever competition will come from different corpora in the future, we are totally confident in claiming that the right decision was taken when marketing analysis was carried out at an early stage in the complicated production process.

330

Arne Zettersten/Hanne

Lauridsen

References

Bijoint, Henri (1994), Tradition and Innovation in Modern English Dictionaries. Oxford: Clarendon Press. Collins COBUILD English Language Dictionary (1987), London: Collins. Collins COBUILD English Dictionary, 2 nd ed. (1995), London: HarperCollins Publishers. Collins COBUILD Essential English Dictionary (1988), London: Collins ELT. Collins COBUILD Student's Dictionary (1990), London: HarperCollins Publishers. Collins COBUILD Student's Dictionary. Bridge Bilingual Portuguese (1995), London: HarperCollins Publishers. Svensen, Bo (1993), Practical Lexicography. Oxford University Press.

Principles and Methods of Dictionary-Making.

Oxford: