121 40
English Pages 503 [530] Year 2011
ISBN 978-3-0343-0761-1
www.peterlang.com
Peter Lang
STUDIES IN HISTORICAL LINGUISTICS VOL. 9
Wim Vandenbussche is Professor of Dutch Linguistics at the Vrije Universiteit Brussel. He teaches courses on Dutch and Germanic language history, as well as on various aspects of sociolinguistics. His research is situated in the domain of historical sociolinguistics, with particular attention to the language situation in Flanders during the eighteenth and nineteenth centuries.
Studies in Historical Linguistics
Steffan Davies is Lecturer in German at the University of Bristol. He studied History and German at the University of Oxford, writing his doctoral thesis on the literary, cultural and historiographical treatment of Albrecht von Wallenstein in the ‘long nineteenth century’.
Language and History, Linguistics and Historiography
Nils Langer is Reader in German Linguistics at the University of Bristol. His primary research interests lie in the area of historical sociolinguistics and he is currently working on language contact in Schleswig-Holstein in the nineteenth century.
Langer, Davies and Vandenbussche (eds)
What are the points of contact between the study of language and the study of history? What are the possibilities for collaboration between linguists and historians, and what prevents it? This volume, the proceedings of an international conference held at the University of Bristol in April 2009, presents twenty-two articles by linguists and historians, exploring the relationship between the fields theoretically, conceptually and in practice. Contributions focus on a variety of European and American languages, in historical periods from the Middle Ages to the present day. Key themes at the intersection of these two disciplines are the standardization and classification of languages, the social and demographic history of medieval and early modern Europe, the study of language and history ‘from below’, and the function of language in modern politics. The value of interdisciplinary collaboration is demonstrated in a wide-ranging set of case studies, on topics including language contact in Northern and Central Europe, the relationship between peninsular and transatlantic Spanish, and new approaches to the recent histories of Nicaragua, Luxembourg and Bulgaria. The volume seeks out the interdependencies between the two fields and asks why exchanges between linguists and historians remain the exception rather than the rule.
Nils Langer, Steffan Davies and Wim Vandenbussche (eds)
Language and History, Linguistics and Historiography Interdisciplinary Approaches
Peter Lang
www.peterlang.com
Peter Lang
STUDIES IN HISTORICAL LINGUISTICS VOL. 9
Wim Vandenbussche is Professor of Dutch Linguistics at the Vrije Universiteit Brussel. He teaches courses on Dutch and Germanic language history, as well as on various aspects of sociolinguistics. His research is situated in the domain of historical sociolinguistics, with particular attention to the language situation in Flanders during the eighteenth and nineteenth centuries.
Studies in Historical Linguistics
Steffan Davies is Lecturer in German at the University of Bristol. He studied History and German at the University of Oxford, writing his doctoral thesis on the literary, cultural and historiographical treatment of Albrecht von Wallenstein in the ‘long nineteenth century’.
Language and History, Linguistics and Historiography
Nils Langer is Reader in German Linguistics at the University of Bristol. His primary research interests lie in the area of historical sociolinguistics and he is currently working on language contact in Schleswig-Holstein in the nineteenth century.
Langer, Davies and Vandenbussche (eds)
What are the points of contact between the study of language and the study of history? What are the possibilities for collaboration between linguists and historians, and what prevents it? This volume, the proceedings of an international conference held at the University of Bristol in April 2009, presents twenty-two articles by linguists and historians, exploring the relationship between the fields theoretically, conceptually and in practice. Contributions focus on a variety of European and American languages, in historical periods from the Middle Ages to the present day. Key themes at the intersection of these two disciplines are the standardization and classification of languages, the social and demographic history of medieval and early modern Europe, the study of language and history ‘from below’, and the function of language in modern politics. The value of interdisciplinary collaboration is demonstrated in a wide-ranging set of case studies, on topics including language contact in Northern and Central Europe, the relationship between peninsular and transatlantic Spanish, and new approaches to the recent histories of Nicaragua, Luxembourg and Bulgaria. The volume seeks out the interdependencies between the two fields and asks why exchanges between linguists and historians remain the exception rather than the rule.
Nils Langer, Steffan Davies and Wim Vandenbussche (eds)
Language and History, Linguistics and Historiography Interdisciplinary Approaches
Peter Lang
Language and History, Linguistics and Historiography
Studies in Historical Linguistics Vol. 9 Edited by Dr Graeme Davis, Karl A. Bernhardt & Dr Mark Garner
PETER LANG Oxford • Bern • Berlin • Bruxelles • Frankfurt am Main • New York • Wien
Nils Langer, Steffan Davies and Wim Vandenbussche (eds)
Language and History, Linguistics and Historiography Interdisciplinary Approaches
PETER LANG Oxford • Bern • Berlin • Bruxelles • Frankfurt am Main • New York • Wien
Bibliographic information published by Die Deutsche Nationalbibliothek Die Deutsche Nationalbibliothek lists this publication in the Deutsche National bibliografie; detailed bibliographic data is available on the Internet at http://dnb.d-nb.de. A catalogue record for this book is available from the British Library. Library of Congress Cataloguing-in-Publication Data: Language and history, linguistics and historiography : interdisciplinary approaches / Nils Langer, Steffan Davies and Wim Vandenbussche (eds). p. cm. -- (Studies in Historical Linguistics; 9) Includes bibliographical references and index. ISBN 978-3-0343-0761-1 (alk. paper) 1. Historical linguistics. 2. Language and culture. I. Langer, Nils, 1969- II. Davies, Steffan. III. Vandenbussche, Wim. P140.L24 2011 417'.7--dc23 2011037901
ISSN 1661-4704 ISBN 978-3-0343-0761-1
E‐ISBN 978‐3‐0353‐0230‐1
© Peter Lang AG, International Academic Publishers, Bern 2012 Hochfeldstrasse 32, CH-3012 Bern, Switzerland [email protected], www.peterlang.com, www.peterlang.net All rights reserved. All parts of this publication are protected by copyright. Any utilisation outside the strict limits of the copyright law, without the permission of the publisher, is forbidden and liable to prosecution. This applies in particular to reproductions, translations, microfilming, and storage and processing in electronic retrieval systems. Printed in Germany
Contents
List of Figures
ix
List of Tables
xi
1
Language and History, Linguistics and Historiography: Theoretical Outlook and Methodological Practices
1
Steffan Davies, Nils Langer and Wim Vandenbussche
Language and History, Linguistics and Historiography: Interdisciplinary Problems and Opportunities
3
Patrick Honeybone
History and Historical Linguistics: Two Types of Cognitive Reconstruction?
15
Nicholas M. Wolf
History and Linguistics: The Irish Language as a Case Study in an Interdisciplinary Approach to Culture
49
Brian D. Joseph
Historical Linguistics and Sociolinguistics: Strange Bedfellows or Natural Friends?
67
Nicola McLelland
From Humanist History to Linguistic Theory: The Case of the Germanic Rootword
89
vi
Agnete Nesse
Editorial Practices and Language Choice: ‘Low German Language Monuments’ in Norway
111
2
127
Standardization and Authenticity
Robert Evans
Of ficial Languages: A Brief Prehistory
129
Tomasz Kamusella
Classifying the Slavic Languages, or the Politics of Classification
147
José del Valle
Linguistic History and the Development of Normative Regimes: The Royal Spanish Academy’s Disputed Transatlantic Authority
175
Juan R. Valdez
Colouring Language: Pedro Henríquez Ureña’s Representations of Spanish and Dominican Identity
193
Laura Villa
‘Because When Governments Speak, They Are Not Always Right’: National Construction and Orthographic Conf licts in Mid-Nineteenth-Century Spain 209 Gijsbert Rutten and Rik Vosters
As Many Norms as There Were Scribes? Language History, Norms and Usage in the Southern Netherlands in the Nineteenth Century 229
vii
Anneleen Vanden Boer
Language and Nation: The Case of the German-Speaking Minority in Belgium
255
3
269
Demographics and Social Dynamics
Richard Ingham
The Decline of Bilingual Competence in French in Medieval England: Evidence from the PROME Database
271
Rembert Eufe
Merovingian Coins and Their Inscriptions: A Challenge to Linguists and Historians
293
Remco Knooihuizen
The Use of Historical Demography for Historical Sociolinguistics: The Case of Dunkirk
323
4 Language History from Below
341
Judith Nobels and Marijke van der Wal
Linking Words to Writers: Building a Reliable Corpus for Historical Sociolinguistic Research 343 Helmut Graser and B. Ann Tlusty
Sixteenth-Century Street Songs and Language History ‘From Below’ 363
viii
Juan Manuel Hernández-Campoy
Mood Distinction in Late Middle English: The End of the Inf lectional Subjunctive
389
5
407
Language and Ideology
Lisa Carroll-Davies
Identifying the Enemy: Using a CDA and Corpus Approach to Analyse Sandinista Strategies of Naming
409
Krassimir Stoyanov
Ritualized Slogan Lexis in the Bulgarian Press during the Times of Violent Contradiction in Ideologies (1944–1947)
429
Kristine Horner and Melanie Wagner
Remembering World War II and Legitimating Luxembourgish as the National Language: Consensus or Conf lict?
447
Michela Giordano and Federica Falchi
John Stuart Mill and Salvatore Morelli: Language as a Social Tool in Nineteenth-Century Britain and Italy
465
Notes on Contributors
483
Index
489
List of Figures
Figure 1 The idol Tuisco (Verstegan 1605: 71; Warburg Institute) Figure 2 The frontispiece to Stieler (1691) Figure 3 From the dialogue De dudesche unde de Norman, Ms 1418, Bergenfahrerarchiv, City Archive, Lübeck Figure 4 Front page of Ms. Gl.kgl.S.2836, Royal Library, Copenhagen Figure 5 A. Schleicher’s genealogical tree of the Indo-Germanic (Indo-European) languages (1873: diagram after p. 33)
161
Figure 6 Percentage occurrence of feminine nouns overall, with del/du and al/au
282
Figure 7 Percentage of occurrence of feminine nouns with de and à, and with combinations ending in -l and those ending in -u
282
Figure 8 NeighborNet diagram showing the similarity of Fort-Mardyck French to other French varieties
335
Figure 9 Map of the Dunkirk area, with places mentioned in the text
336
Figure 10 The explicit reference to the writing process in letter 3–1–2008, 129–130, letters-as-loot-corpus Figure 11 Letter by Claeijs Pietersen, letters-as-loot-corpus Figure 12 Letter by Jan Lievensens, letters-as-loot-corpus Figure 13 An example of a signature in a larger handwriting than the body of the letter, letters-as-loot-corpus Figure 14 An example of a signature in a dif ferent and less experienced hand writing than the body of the letter, letters-as-loot-corpus Figure 15 Jacob van de Velde’s handwriting and signature in letter 3b-1–2008, 131–132, letters-as-loot-corpus
x
List of Figures
Figure 16 Jacob van de Velde’s handwriting and signature in a receipt kept in the archive of the Audit Of fice of Zealand Figure 17 Flow chart to determine the autographical status of letters Figure 18 The weaver’s song (Stadtarchiv Augsburg, Reichsstadt, Urgichten, 13.09.1591, Peter Müller/Matthäus Sporhan) Figure 19 The tailor’s song (Stadtarchiv Augsburg, Reichsstadt, Urgichten, 28.06.1594, Christof Halbritter) Figure 20 Percentages of usage of was and were in indicative and subjunctive in singular contexts
399
Figure 21 Percentages of usage of was per singular subject type and mood
399
Figure 22 Percentages of usage of was per plural subject type and mood
401
List of Tables
Table 1 Consonant alternations in Dutch spelling, compared with German 105 Table 2 Dual classification of Slavic
154
Table 3 Elaborated dual classification of Slavic
157
Table 4 Revised triple classification of Slavic I
157
Table 5 Revised triple classification of Slavic II
158
Table 6 Revised triple classification of Slavic III
159
Table 7 Dual-cum-quadruple classification of Slavic
163
Table 8 Spelling features and the normative framework
242–243
Table 9 Southern (left) and Northern (right) spelling variants per feature (distribution per cent) in a digitized corpus of handwritten documents from the Southern Netherlands
246
Table 10 Spelling variants per province (distribution per cent)
247
Table 11 Correlation between mother tongue and the sense of af filiation to Belgium
262
Table 12 Frequency distribution for the variables indicating the sense of af filiation to Belgium 262–263 Table 13 Frequency distributions for the language command
264
Table 14 Correlation between language command and the sense of af filiation to Belgium
265
Table 15 Frequencies of masculine and feminine nouns used with masculine preposition + article forms in PROME, 1280–1399 281 Table 16 Frequency of masculine and feminine noun types co-occurring with masculine premodifier forms cel, cest, nul and tut, late thirteenth century/early fourteenth century and 1370s 283–284
xii
List of Tables
Table 17 Frequencies of del with masculine and feminine nouns, 1280–1324 and 1377–1399 285 Table 18 Locals and immigrants in the Dunkirk marriage register, by gender and sample year
330
Table 19 Distribution of languages across marriage partners 1647–1697
330
Table 20 Proportions of French speakers among the population of Dunkirk, overall and by gender, and the preference for mono-ethnic marriage in the sampled years 332 Table 21 Subject type
393
Table 22 Levelling to was/were in contexts of Indicative Mood
397
Table 23 Levelling to was/were in contexts of Subjunctive Mood
398
Table 24 FSLN corpus documents
416
Table 25 Envío corpus contents
417
Part 1
Language and History, Linguistics and Historiography: Theoretical Outlook and Methodological Practices
Steffan Davies, Nils Langer and Wim Vandenbussche
Language and History, Linguistics and Historiography: Interdisciplinary Problems and Opportunities
For the past few years, the notion of interdisciplinarity has been a buzz word to be found in any programmatic research outline or grant proposal, understood to be a vital break away from scholarly isolation and too narrow a focus on one’s own methodology and research questions. This pertains to the humanities just as much as to the natural sciences, and it is hardly a new idea. The historian Marc Bloch stated in the 1940s that ‘it is indispensable that the historian possess at least a smattering of all the principal techniques of his trade’ (Bloch 1992: 57). More specifically with regard to the disciplines of linguistics and History, Bloch exclaims: ‘What an absurd illogicality that men who half the time can have access to their subject only through words, are permitted, among other deficiencies, to be ignorant of the fundamental attainments of linguistics’ (ibid.).1 Yet in practice, exchange between the two fields has often remained a desideratum rather than actual achievement. The fragmentation of traditional philology into a broad subset of highly specialized branches of linguistics from the 1960s (computational linguistics, feminist linguistics, language acquisition studies, etc.), all of which tended to focus on present-day language use only, meant that the study of language structure and texts in their historical context lost the prominence it had enjoyed since its emergence as a serious scholarly discipline in the early nineteenth century. The ‘linguistic turn’ much discussed by historians since the 1970s has produced, despite its name, 1
To distinguish between the two meanings of history, namely History as a scholarly discipline versus history to refer to the past, we use a capital letter for the former. Since no such problem arises to distinguish between language and linguistics, we do not use a capital letter for linguistics.
4
Steffan Davies, Nils Langer and Wim Vandenbussche
little engagement, critical or otherwise, with the discipline of linguistics as most linguists know and practise it, focusing instead on the mediate nature of access to the past, and the ‘literary’ treatment of that past in the present (summary in Fulbrook 2002: 18–24). Exceptions to this mutual isolation can be found on both sides of the disciplinary fence, of course, but largescale collaboration between our two disciplines has continued to be rare, not least because scholars lack the confidence to step outside their area of expertise and into territories where basic assumptions, technical terms and methodological practices may be alien. In particular, there is a general lack of opportunity to exchange ideas and basic knowledge of core ideas. This book is an attempt to facilitate such an opportunity. It contains studies first presented at a conference at the University of Bristol in April 2009, generously supported by a Scientific Network Grant from the Arts and Humanities Research Council (AHRC), which explicitly addressed issues of interdisciplinarity, that is, to what extent there is any actual scope for collaboration, common knowledge of ‘tools’, or exchange of ideas between linguists and historians.2 First steps towards a break away from disciplinary isolation had already been accomplished within the field of historical sociolinguistics over the last ten years. The aim of key conferences in historical sociolinguistics (Shef field 2001, Bristol 2003, Bristol 2005, and Bruges 2006), edited volumes (Linn & McLelland 2002, Deumert & Vandenbussche 2003, Langer & Davies 2005, Elspaß et al. 2007), and the thinking behind the foundation of the Historical Sociolinguistics Network (HiSoN, since 2005) has been to achieve a much greater awareness of similarities and dif ferences in the sociological histories of related and neighbouring languages, at first mostly restricted to Germanic languages for practical reasons. Success in bringing separate linguistic traditions and languages together created the impetus for branching out further, to establish how a direct exchange between pure historians and historical sociolinguists may provide further 2
This is an opportunity to give formal thanks to the AHRC, the University of Bristol, the Vrije Universiteit Brussels for their financial support, as well as our student helpers Evi Egger, Catherine Hillier, Stefanie Hundehege, Nadia Taylor, and, in particular, Stef fi John. Special thanks also to Sabine Lehmann, Juliane Fischer, and Patrick Beuge for their help with editing this conference volume.
Language and History, Linguistics and Historiography
5
benefits for our understanding of human societies. Yet the task of fostering interdisciplinarity between linguists and historians is hard. There is much less understanding of what the other discipline is doing, partly because of physical limitations: there are a limited number of hours in the day, and for many academics, the pressure to publish frequently leaves little time to learn about a whole new scholarly discipline, on the of f-chance that it might enrich their findings. Furthermore, we suspect, scholars feel most comfortable in the areas in which they were trained. In response to such challenges, the authors in this volume have written for a readership not necessarily expert in their discipline. Language is a primary means of access to the past, and the historian’s primary means of expression. The centrality of language as a research tool is most acutely clear to historians who work on countries other than their own. Most countries require their History students to have a good understanding of two foreign languages at least. In the United Kingdom, where this is not the case, Richard Evans, a historian of Germany, warned, upon taking up the Regius Chair of Modern History at Cambridge, that: ‘The rapid and continuing decline of language-learning in British schools is in many ways the most important single factor threatening the continuation of the long tradition of British historians’ engagement with the European Continent’ (Evans 2009: 201). But ‘domestic’ historians, too, can ill af ford to neglect foreign languages. Significant aspects of the history of the British Isles, too – requiring knowledge of Latin, Middle English, French, or the Celtic languages – will remain a closed book to the monoglot reader of modern English; the historian of modern East-Central Europe will sooner rather than later encounter sources in German or Russian. The growth of immigrant communities in twentieth-century Europe may require the historians of the future to learn a new set of foreign languages. Historians need to master languages, but, as Robert Evans argued in his inaugural lecture to a Regius Chair of Modern History, this time at Oxford (Evans 1998), they also need an appreciation of language that goes well beyond what simple words mean. In order to interpret their sources, historians need to understand discourse analytic methods, the impossibility of reading a text without context, and the sociolinguistic dynamics involved when writers use particular linguistic varieties and variants.
6
Steffan Davies, Nils Langer and Wim Vandenbussche
Such a list may, of course, sound like yesterday’s news. E.H. Carr’s What is History?, a staple of first-year university courses on historical method, pointed out a good half-century ago that historical facts ‘are always refracted through the mind of the recorder’, and noted: ‘The very words which [the historian] uses – words like democracy, empire, war, revolution – have current connotations from which he cannot divorce them’ (Carr 1964: 22, 24–25). Thus an interpretation of historical sources requires a sophisticated understanding of how language works socially – be this in terms of lexical semantics, historical pragmatics, or text linguistics. Providing the source with context is a fundamental, natural historical task (Evans 1997: 103–107). This volume, however, argues for more profound and explicit engagement on the historian’s part with the methods and findings of linguistics, and demonstrates the potential results of such work. On a pragmatic level, such engagement adds detail and guards against errors. Simple f luency in a living or dead language is arguably insuf ficient to unearth and interpret correctly the linguistic thrusts and nuances contained in a historical text. Where texts are written in a non-standard variety, an understanding of sociolinguistics is crucial for their correct contextualization, since not all non-standard variants are used by the prototypical uneducated speaker. Spelling variation in the seventeenth century carried a dif ferent social association in periods before the existence of a standard language than today; Macha (2004) shows how particular spelling conventions demarcated Catholics and Protestants in a corpus of war correspondence from the sixteenth century, whilst no social stigma was attached to the use of non-standard spelling then. Analysing sixteenth-century spelling variation enables us, following Macha, to say something about the author’s or printer’s deliberate or conventional positioning during the time of the confessional wars. Without understanding the sociological significance of such linguistic markers, historical textual interpretation would arguably be naïve at best, and skewed or simply wrong at worst. Not only language, but also linguistics, belongs in the historian’s toolbag, even if it is rarely treated thus: History students in Germany are instructed that heraldry, palaeography, numismatics and sigillography are among History’s ‘ancillary sciences’ (die historischen Hilfswissenschaften), but there is no mention of linguistics (e.g. Opgenoorth and Schulz 2001).
Language and History, Linguistics and Historiography
7
For some areas of History, language and linguistics plays a still more fundamental role. One such area is the ongoing interest in national and ethnic identities and their origins: identities that emerged, as Robert Evans points out, in ‘intimate conjunction with language’ (Evans 1998: 24; for a survey, Joseph 2006). Comparing linguistic evidence from the oldest languages on the British Isles, Trudgill (2009) draws a new and provocative picture of the extent and nature of language contact before the Norman Conquest. Given the particular social settings and types of interaction required for significant language-contact phenomena to emerge, Trudgill is able to present historians and archaeologists with a linguistically motivated account of living conditions in Great Britain before 1066. Another example is Ingham (this volume) who uses statistical evidence and language-contact examples to enhance our understanding of immigration and integration in medieval Anglo-Norman England. Greater interest in groups excluded from traditional narratives, and especially in day-to-day lives and horizons, has expanded still further the range of research techniques on which the social historian relies. As John Tosh remarks: ‘There is probably no other field whose primary sources are so varied, so widely dispersed, and so uneven in quality’ (Tosh 2006: 136). Broad and sparse evidence demands a diversity of method: Bloch, the advocate of teaching linguistics to historians, co-founded the inf luential journal Annales d’histoire sociale et économique in 1929 calling for the closer integration of History with other disciplines, notably the social sciences. ‘Linguistics’ in Bloch’s lifetime meant almost exclusively systemic linguistics: sociolinguistics was yet to be born and has trailed behind the historians’ turn to studying ‘history from below’. Elspaß (2005), however, has cemented linguists’ interest in the field, and with Elspaß et al. (2007), there is now a collection of studies focused on unearthing ‘language histories from below’. Linguistics techniques function here not only as a guardian against misapprehension, but in developing historical analysis: comparing common spelling conventions and grammatical phenomena in seventeenthcentury Dutch letters is just as useful as handwriting in the identification of scribes (cf. Nobels and van der Wal, this volume). The ‘social history of language’ is a final area in which the two disciplines are closely intermeshed: in calling for such a history, Peter Burke remarked in the late 1980s that ‘a
8
Steffan Davies, Nils Langer and Wim Vandenbussche
number of historians have recently come to recognize the need to study language as a social institution’, and posited ‘a gap between linguistics, sociology (including social anthropology) and history […] which can and should be filled by the social history of language’ (Burke and Porter 1987: 1). Burke’s own work, and projects such as the multi-volume Hanes Cymdeithasol yr Iaith Gymraeg / A Social History of the Welsh Language (e.g. Jenkins 1997–2000) have responded to the challenge. Similarly, linguists need History. The concentration of many linguistics curricula in European universities on present-day language study (and of sociolinguistics on contemporary language use) has regrettably implied a disregard for historical development, even though the principles and developments in language change are often most insightful for our understanding of how language ‘works’ as such. As Robert Evans points out, the sociolinguist’s interest in language change is essentially an interest in the past (Evans 1998: 11). Honeybone (this volume) reduces the principal dif ferences between the two disciplines to the following, provocative, core definition: linguistics studies languages, and History studies people. But, he argues, these things are not unconnected: languages live in people, and much historical evidence comes from records that people wrote using their languages. Honeybone identifies areas of historical linguistics where an understanding of social processes (historical or otherwise) are irrelevant, such as individual sound changes from one set of sounds to another (e.g. the Second Sound Shift in High German, shifting /p, t, k/ to /pf, ts, kx/ and /f, s, x/). Endogenous changes, arising from within linguistic systems themselves, may not be attributable to the social context of the people who innovate them, though as the products of past social activity, they are no less ‘historical’ for it. Exogenous changes, however, can only be explained if we understand the lives of the speakers involved: one such case is the formation of new dialects or languages where speakers of ‘old dialects’ mix in areas where no established dialect exists. The emerging new forms and varieties can only be understood, Honeybone claims, if we know the demographics of the speakers, requiring substantial collaboration between linguists and historians. For example, Knooihuizen (this volume) provides one such example of how a detailed demographic examination of marriage registers can inform our understanding of language contact and language loss in seventeenth-century Dunkirk. Linguists writing social histories of
Language and History, Linguistics and Historiography
9
language use need information from historians as to when a particular text was written, who the likely author was, at what occasions texts were produced, and, very importantly, which texts were not written or did not survive. In multilingual scenarios, dif ferent languages were used but not all were written down; the availability of material resources may have had an ef fect on education, which in turn af fected literacy. The general lack of texts written in Frisian, Low German, or Sønderjysk in eighteenth- and nineteenth- century Denmark gives a distorted view of the thriving actual use of these languages in the southern part of the country. Seen this time from the linguist’s perspective, the relationship again functions on three levels: an essential af finity (the common study of past activity), areas of study in which the two disciplines work side by side (such as the social history of language), and with History functioning as a linguists’ ‘tool’, where historical findings incidental to linguistics bear a direct relevance in informing its work. Two final examples show that the research methods and core wisdoms of both disciplines are embedded in a historico-linguistic minefield. Key terms such as ‘Germany’ and ‘German’ are known to historians as slippery concepts: the political and cultural entities known as ‘German’ have shifted over centuries, and even at the founding of the Kaiserreich in 1871, there was a deeply-felt mismatch between German ethnic and national identity and the geographical boundaries of the new state. Indeed, the need for historians to learn multiple languages indicates in itself that the concept of the nation, often defined by linguistic community, is shaky. When it comes to the names for languages, however, scholars are much more ready to assume continuity across centuries, even though it is by no means clear where the boundaries between, say, Standard German and Moselle-Franconian were on the one hand, and Luxembourgish and German on the other. The assumption that we can talk about ‘the history of German’ rests on the identification of German as a separate language – yet the beginnings of any language are always murky and the ‘birth’ of a language can be dated as either the day when a variety was significantly dif ferent from its neighbouring varieties, or the day when the speakers ‘felt’ that their linguistic variety was suf ficiently distinct to be seen as a separate language. The dif ference between ‘fact’ and perception is never clear: even key terms such as ‘language’ or ‘German’ are mired in significant confusion. This confusion cannot be resolved without
10
Steffan Davies, Nils Langer and Wim Vandenbussche
interdisciplinarity since it does not rest on misunderstanding the processes involved, but rather on the complexities and range of situations to which these terms are applied. The danger does not lie in these ‘wobbly’ foundations but rather in the fact that the terms (language, German) are used as real and hard notions in the scholarly discourse without always realizing that terms are historical creations. Kamusella (this volume) presents a similar discussion with regard to the politics of classifying Slavic languages, which in recent years have seen the politically motivated emergence of Czech and Slovak, as well as Serbian and Croatian, as distinct languages, even though their respective linguistic dif ference is no greater than that between major dialects in English or Italian. The same applies to terms which are much more clearly ‘technical’ as used by various disciplines but where the dangers – and the possibilities – lie in their dif ferent understanding by, say, linguists, historians, and literary scholars. The key notion of ‘politeness’ in the eighteenth century will be understood dif ferently by historians and linguists. For historians, politeness concerns sociability and is a mode of social coordination; literary historians with an interest in the periodical essay genre use the concept as a discursive medium for the development of sociability. In historical sociolinguistics, on the other hand, politeness is identified as a set of attributes that are commercialized and commodified via the identification of good grammar as the way to ‘polite’ conversation and the operationalization of politeness through the medium of humiliative discourse (Fitzmaurice 2010). In turn, both interpretations are dif ferent from the notion of linguistic politeness as a key concept in modern sociolinguistics, which talks about way speakers use the concept of ‘face’ (after Brown & Levinson 1987) to position their status as speakers in a particular discourse. Dictionary entries on politeness will provide neither the historian nor the linguist with a suf ficiently refined dif ferentiation of the changes in the word’s meaning and usage over the centuries. Appreciating its range guards us from an over-simplified understanding of the term, and allows us to overcome the anachronism pointed out by Carr. Reinhart Koselleck’s work demonstrates how profitably an appreciation of semantics could be turned to an elucidation of historical concepts (Koselleck 2004; also Brunner et al. 1972–1997).
Language and History, Linguistics and Historiography
11
There is thus considerable scope for benefits from a better mutual understanding of our conceptual premises, research methods and scholarly outcomes. In particular, the papers in this volume address these general areas of potential common interest: Language History from Below, with a new focus on informal sources such as pauper letters or emigrants’ diaries, which had been largely ignored in linguistics until recently yet which inform, and are informed by, research programmes working on ‘history from below’; Political Language, which investigates discourses relating to political debates and ideas and which demonstrate the importance of linguistic analyses of political discourse in order to understand how particular terms are used and political ‘realities’ are created and transmitted in a particular community; Language Contact, which focuses on multilingual communities to identify how a particular language is used or suppressed in order to create individual and group identities. Language contact is an important field for understanding the genesis of communities and their feeling of distinctiveness, since language is often one of the key components that separates a community from others; Historical Semantics, which examines how the meaning of words changes over time, partly, but not exclusively, due to a change in the social environment that require or permit a refocusing of how words are used; Attitudes to Language, which investigates what people think about language and how this relates to their evaluation of particular varieties as good languages, e.g. standard languages, and bad languages, e.g. particular sociolects and regional languages; Historiography, which – in this context – is concerned in particular with two areas, namely on the one hand, how the description of linguistic phenomena has changed during the last four centuries and, on the other hand, how editorial practices have changed over the years. Practices completely acceptable and sensible for historians can obscure vital information for a linguist wishing to interpret the same source.
The volume thus contains an outline of the possibilities of contact between the two disciplines, using both case studies from cutting-edge research and more general, theoretical discussions on interdisciplinary exchange between language / linguistics on the one hand and History / historiography on the
12
Steffan Davies, Nils Langer and Wim Vandenbussche
other. This volume is intended to serve both as a platform to show where we are, and as a starting point of instruction as to how research in the ‘other’ discipline is conducted and what principal questions and problems concern them. The chapters of this volume demonstrate how wide-ranging such issues are and how, nonetheless, a common thread in the advancement of our understanding of human societies in the past and humans – as actors and speakers – can be discerned.
References Brown, Penelope, & Stephen Levinson. 1987. Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Brunner, Otto, Werner Conze & Reinhart Koselleck (eds). 1972–1997. Geschichtliche Grundbegrif fe. Historisches Lexikon zur politisch-sozialen Sprache in Deutschland. 8 vols. Stuttgart: Klett. Burke, Peter, & Roy Porter (eds). 1987. The Social History of Language. Cambridge: Cambridge University Press. Carr, E.H. 1964. What is History? Harmondsworth: Penguin. Chambers, J.K., & Peter Trudgill. 2005. Dialectology. Cambridge: Cambridge University Press. Deumert, Ana, & Wim Vandenbussche (eds). 2003. Germanic Standardisations. Amsterdam: Benjamins. Elspaß, Stephan, Nils Langer, Joachim Scharloth & Wim Vandenbussche (eds). 2007. Language Histories from Below (1700–2000). Berlin, New York: Walter de Gruyter. Evans, Richard J. 1997. In Defence of History. London: Granta. Evans, Richard J. 2009. Cosmopolitan Islanders: British Historians and the European Continent. Cambridge: Cambridge University Press. Evans, R.J.W. 1998. The Language of History and the History of Language: An Inaugural Lecture Delivered before the University of Oxford on 11 May 1988. Oxford: Clarendon Press. Fitzmaurice, Susan. 2010. ‘Changes in the Meanings of Politeness in Eighteenth-Century England: Discourse Analysis and Historical Evidence.’ In: Jonathan Culpeper & Dániel Kádár (eds), Historical (Im)politeness. Bern: Peter Lang. 87–115.
Language and History, Linguistics and Historiography
13
Fulbrook, Mary. 2002. Historical Theory. London: Routledge. Jenkins, Geraint H. (ed.). 1997. The Welsh Language before the Industrial Revolution. Cardif f : University of Wales Press. Jenkins, Geraint H. (ed.). 1998. Language and Community in the Nineteenth Century. Cardif f : University of Wales Press. Jenkins, Geraint H. (ed.). 2000. The Welsh Language and its Social Domains 1801–1911. Cardif f : University of Wales Press. Jenkins, Geraint H., & Mari A. Williams (eds). 2000. ‘Let’s Do Our Best for the Ancient Tongue’: The Welsh Language in the Twentieth Century. Cardif f : University of Wales Press. Joseph, John E. 2006. ‘“The Grammatical Being Called a Nation”: History and the Construction of Political and Linguistic Nationalism.’ In: Love, Nigel (ed.), Language and History: Integrationist Perspectives. London: Routledge. 120–141. Koselleck, Reinhart. 2004. Futures Past: On the Semantics of Historical Time. Translated by Keith Tribe. New York: Columbia University Press. Langer, Nils, & Winifred Davies (eds). 2005. Linguistic Purism in the Germanic Languages. Berlin, New York: Walter de Gruyter. Linn, Andrew, & Nicola McLelland (eds). 2001. Standard Germanic. Amsterdam: Benjamins. Macha, Jürgen. 2004. ‘Konfession und Sprache. Zur schreibsprachlichen Divergenz um 1600.’ In: Mattheier, Klaus J., & Haruo Nitta (eds), Sprachwandel und Gesellschaftswandel. Wurzeln des heutigen Deutsch. München: iudicium. 161–174. Opgenoorth, Ernst, & Günther Schulz. 2001. Einführung in das Studium der Neueren Geschichte. Paderborn etc: Schöningh. Tosh, John, with Seán Lang. 2006. The Pursuit of History: Aims, Methods and New Directions in the Study of Modern History. London: Pearson Longman. Trudgill, Peter. 2009. Investigations in Sociohistorical Linguistics. Stories of Colonisation and Contact. Cambridge: Cambridge University Press.
Patrick Honeybone
History and Historical Linguistics: Two Types of Cognitive Reconstruction?
Abstract This article compares principles and practice in history and (structural) historical linguistics. I argue that the disciplines can be both connected and distinguished by the recognition that they engage in acts of cognitive reconstruction. I show that such reconstruction is fundamental to both disciplines, but that they do it dif ferently: historical linguistics reconstructs unconscious entities, while history reconstructs at the conscious level. For these arguments to go through, certain commitments are required from the disciplines’ philosophies: mentalism of the type associated with Chomsky’s linguistics, and idealism of the type associated with Collingwood’s history. Although cognitive reconstruction is important to both areas of study, it is not all they do: to provide the context for my arguments, I also consider a number of other connections and distinctions between the disciplines, in terms of the questions that they can ask, the evidence available to them, and their relationships to synchrony and diachrony.
1 Introduction The academic disciplines of history and historical linguistics clearly have some things in common – they both deal with aspects of the past, after all. There is also much that dif ferentiates them, however, in their aims and methodologies, and in their intellectual context and traditions. In this article I argue that we should recognize one particular point which both connects and distinguishes the disciplines, in ways which are not commonly
16
Patrick Honeybone
discussed (and I consider a few other things which link or dif ferentiate them, too).1 The main point has to do with the very nature of their objects of enquiry, and it will require us to entertain a set of controversial but compelling assumptions about the philosophy of the disciplines. The article can be seen as a contribution to the comparative philosophy of disciplines, and its main aim is to help us better understand what it is that we do when we do historical linguistics and/or history (or, at least, central parts of them) and also in what it is that we don’t. I do not compare everything that historical linguists and historians do here, nor everything that they are interested in. Rather, in seeking to answer the fundamental questions ‘what is the object of study in history?’ and ‘what is the object of study in historical linguistics?’ I argue that certain fundamental aims in the two disciplines allow us to remove the question mark in the title to this piece: history and historical linguistics do engage in two – interestingly connected, but also quite dif ferent – types of cognitive reconstruction. In section 2, I consider a number of basic connections and dif ferences between the two disciplines that are our focus; these are mostly beyond the main topic under consideration here, but they provide a context for the discussion in sections 3, 4 and 5. These latter sections consider historical linguistics and history in the light of the idea proposed in the article’s title: section 3 focuses on historical linguistics, and section 4 on history. Section 5 concludes.
1
For discussion of these ideas, I would like to thank Nils Langer, Michael and Diana Honeybone, and the audiences at the Workshop on History and Linguistics, Linguistics and Historiography at the University of Bristol in 2008, and at a talk in the Language and History Interdisciplinary Seminar at the University of Oxford in 2009. I would also like to thank two anonymous reviewers for comments on a draft of this chapter. Any errors remaining are, of course, my own.
History and Historical Linguistics
17
2 History and historical linguistics: Parallels and dif ferences One obvious dif ference between historical linguistics and history is that historical linguistics is a subdiscipline, or branch, of a larger area of study: linguistics. ‘History’ is equivalent to ‘linguistics’ as a superordinate disciplinary term, which itself has second-level subfields, such as social history and political history. Certain characteristics of historical linguistics are inherited because it is a ‘type of linguistics’, so some of what follows compares aspects of history with aspects of general linguistics.2 With this in mind, we can recognize a basic dif ference in the two fields’ terminologies: for linguistics, there is a handy distinction between (1a) and (1b), whereas the equivalent is absent for history, as (2a) and (2b) show. (1a) the name of the discipline: linguistics (1b) the discipline’s object of study: language (2a) the name of the discipline: history1 (2b) the discipline’s object of study: history2
2
As Lass (1997: 27) has it: ‘The primary constraint on a historical subject is its nonhistorical metasubject. Historical biology is part of biology, and hence constrained by biological knowledge and theory; historical linguistics is a branch of linguistics, constrained by non-historical linguistic knowledge and theory.’ We will see below (in sections 2.1 and 3) how constraints developed in general linguistics can play a role in historical linguistics. Incidentally, history, as a metasubject, is unusual in not having a historical sub-subject – there is historical sociology, historical geography and historical anthropology, for example – but, perhaps because we need to avoid going round in circles, there is no historical history (there is ‘the history of history’, but that’s a dif ferent matter). This could be seen as a further dif ference between the two disciplines in question here. On the other hand, historians do have the useful term historiography, to refer to the study of the writing of history as an academic discipline. There is no equivalent linguisticography, but this is simply a lexical gap, as the linguistic equivalent has long been studied. If we were to fill the gap, this article might even be described as a contribution to comparative historiography and linguisticography.
18
Patrick Honeybone
It could be claimed that (2b) should be ‘the past’, but there are reasons to reject this: history1 typically requires human agency (or even written records) for its subject matter to count, something considered in some detail in section 4. To be of interest to history1, these acts of human agency need to have occurred in the past, but that is not the same as simply being the past. The term ‘history’ is thus ambiguous (an obvious point, long made by others writing on the nature of history, such as Carr 1961), and as it is important to be clear about the terms used in a piece such as this, it might sometimes help to have the subscriptal distinction in (2). I use this in the current section, but − rather than cluttering the whole article with subscripts − I adopt the convention in later sections that, if no subscript is given, then ‘history’ should be taken to mean history1 (as has been implicit up till now). It might seem reasonable to expect ‘historical linguistics’ to be an interdisciplinary branch of enquiry existing in the overlap between history1 and linguistics. This is not, however, where much of historical linguistics lies. Considerable work in historical linguistics deals with autonomously linguistic structural entities (‘autonomous’ in the sense of Newmeyer 1986, for example), such as phonological segments and syntactic categories, and these entities are not subject to the conscious human will that history1 considers. This is not to say that no part of historical linguistics overlaps with history1 – some parts of it certainly do, and there is an intersecting area of study where precisely the same questions can be asked. It may well be that this is where much of ‘historical sociolinguistics’ lies (the fact that such work is also known as ‘the social history of language’ – as in Burke 2004 – corroborates this idea). Historical linguistics is thus not a unitary (sub)discipline, because it groups together any aspect of anything which connects language or language use with history2. Questions which need both historical and linguistic investigation include such basic issues as: how does the standardization of languages occur? There are clearly linguistic issues involved in standardization, but we also need to understand the social relations and context which motivated the individuals who enacted the process, and this requires historical methods. Specific questions which exist in the disciplinary overlap might include: what role did the idea that there was a Czech language play in political
History and Historical Linguistics
19
developments in the Central Europe? and what impact did this have on the Slavic dialect(s) spoken there? (see, for example, Törnquist-Plewa 2000, Evans 1998) or what were the demographics of the people involved in the early colonization of New Zealand? and how did this lead to the formation of the ‘new-dialect’ of New Zealand English? (see, for example, Trudgill 2004). Such questions can all be part of historical linguistics, but they are not the kinds of questions considered in this article. I focus here (mostly in section 3) on issues of autonomous linguistic structure which are primarily or only of interest to linguists. With this restriction, we can identify a number of parallels between history1 and (at least the structural part of ) historical linguistics. One seems obvious, so much so that we have noted it already: both are connected with aspects of (actions from, states that existed in) the past, as spelt out in (3). (3a) history1 aims to understand past actions (3b) historical linguistics aims to understand past linguistic systems
There are two reasons why this is not in fact as obvious as it seems: (i) it requires us to reject the postmodern approach to history1, and (ii) it requires us to tease apart the synchronic and diachronic approach to the study of the past. Postmodernism is a contentious and complex philosophical current in history1 (see, for example, O’Brian 2001 and Munslow 2001), as in many other fields. Some historians (such as Jenkins 1991 and Munslow 2001) avowedly subscribe to it, to argue, contrary to (3a), that we cannot ever rediscover or, therefore, truly understand history2. Postmodern history ties in with the general postmodernist distrust that anything could ever be objectively true, or that we can ever understand the thoughts of anyone other than ourselves, arguing that we are tied to our own contemporary linguistically-determined world-view. This means, the claim goes, that history1 just tells possible tales about history2, and other tales are always just as reasonable (for example, Munslow 2001 writes that ‘[m]y history is just another cultural practice that studies cultural practice’).
20
Patrick Honeybone
The approach that I consider here requires us to reject postmodern doubt, and to assume that we are indeed able to understand history2 in terms of how it actually was (following, for example, Evans 1998 and O’Brien 2001), just as we can hope to understand the truth about physics and medicine. It is noteworthy that postmodernism has not been inf luential in linguistics (as it is generally conceived) − another dif ference between the two disciplines.3 Perhaps this is because many linguists view linguistics as more of a science than a humanity (and postmodernism has found no place in science), but it may also be that the very existence of historical semantics and etymology – one branch of linguistics – contradicts or conf licts with postmodern tenets. The argument from etymology is that we are able to recover the meaning of words from the past, with some painstaking work, and hence we can understand what others mean, or meant. In its own way, this point f lags up a further connection between historical linguistics and history1 – the latter relies on the former to some degree, because historians do need to rely on a firm reconstruction of the meaning of the texts that they work with if they are to escape the postmodern current. Thanks to postmodernist arguments, historians are more conscious of the significance of language to historical study (taking a ‘linguistic turn’), but while we know that words can change their meaning – an example of this is given in (11), below – an awareness of the results of etymology (along with a judicious use of materials such as dictionaries which are contemporary with the texts that are being studied) allows us to reject the postmodern position, as is necessary for the argument to be made below to go through. For (3) to make sense, we also need to tease apart synchrony and diachrony. This distinction is inescapable in linguistics following its description in Saussure (1916), a fundamental text for twentieth-century linguistics. A 3
Rather oddly, some postmodern philosophers claim that they are doing linguistics, or at least are interacting with the concepts that linguistics deals with (e.g. Derrida 1967), or that linguistics was fundamental in the pathway of postmodernist ideas, through the development of structuralism. For a clear explanation of the baf f lement that this causes in linguists, see Dresher (1999), and for a demolition of some of the small amount of work in (applied) linguistics which has been inf luenced by postmodern ideas, see Borsley (2000).
History and Historical Linguistics
21
synchronic approach focuses on a particular system at one point in time, whereas a diachronic approach focuses on the changes that occur between chronologically successive instantiations of one particular system. This is quite straightforward in principle: it was recognized by linguists before Saussure named it, and the same distinction is made in other historical disciplines, even if under dif ferent names (for example, Warkentin 2009 talks of the ‘horizontal approach’ − that is, synchrony − and the ‘vertical approach’ − diachrony − in historical geography). A naïve view of historical linguistics and history1 might expect both to be purely diachronic disciplines (for example, Pei 1965 writes ‘[a] term often used as a synonym for historical linguistics is diachronic linguistics’), but for (3a) and (3b) to hold, they must study synchrony: a particular action can only occur at one specific point in history2, and past linguistic systems exist at specific points in history2. In fact, historical linguistics and history1 can and do focus on both diachrony and synchrony: for example, we might synchronically study the society or politics of the kingdom of Prussia in the middle of the nineteenth century, or we might study the diachrony of the changing organization of the German states during the nineteenth century; or we might synchronically study the phonology of mid-fifteenth-century Middle English, or the diachrony of the changes in the tense vowel system of English from the end of the Middle English period to the start of Late Modern English. The synchrony/diachrony distinction will prove important below. We can recognize one further parallel between historical and historical linguistic method, if only because it f lags up one way in which the two disciplines are similar (even if the terminology used in this area hides this). This is pointed out in (4). (4a) the historical linguist uses both direct and indirect evidence (4b) the historian uses both witting and unwitting testimony
The observations in (4) describe two types of sources available in the two disciplines. The distinction in (4a) is analogous to that in (4b), and it is implicit in (4) that both disciplines are limited in similar ways by the evidence that is available to them. If we take historical phonology as an example
22
Patrick Honeybone
of work in historical linguistics, the two parts of (4a) can be explained as follows: in order to do historical phonology, we need to know how past phonological systems worked (which sounds they used, how they were organized, etc.). Direct evidence for this includes overt comments on pronunciation from past orthoepists and spelling reformers (such as Hart 1569, who devised his own phonetic alphabets to better represent English phonology); they needed to explicitly describe how people spoke in the sixteenth century in order to be able to write about what they wanted. Indirect evidence for historical phonology includes the consideration of spellings in documents from the past, and of rhymes and puns from past linguistic states, none of which was intended to provide special evidence about phonology (see, for example, Dobson 1968, Jones 1989 and Lass to appear). Marwick (2001a, b) explains the (4b) distinction used in history1, writing: ‘“Witting” means “deliberate” or “intentional”; “unwitting” means “unaware” or “unintentional”. “Testimony” means “evidence”. Thus, “witting testimony” is the deliberate or intentional message of a document or other source; the “unwitting testimony” is the unintentional evidence (about, for example, the attitudes and values of the author, or about the “culture” to which he/she belongs) that it also contains.’ The parallel in (4) is that both disciplines consider a range of types of sources, which require parallel types of interpretative skill: some aspects of these sources aim to tell the reader something, and are hidden until research reveals them. 2.1 How can cognitive reconstruction be relevant? In order to show that history and historical linguistics are linked through cognitive reconstruction, we need to establish why we should want to discuss cognitive states at all. The distinctions made in (1) and (2) are worded as if they were very simple; like most things, though, they are not. The claim that cognitive reconstruction is involved in both branches of study derives from positions that have been developed in the disciplines’ philosophies concerning the true nature of (1b) and (2b) − of their objects of study. They are set out in (5) and (6).
History and Historical Linguistics (5)
in history1: the ‘idealist’ position (e.g. Collingwood 1946) = the aim of history1 is to rethink the thoughts of people from the past
(6)
in linguistics: the ‘mentalist’ orientation (e.g. Chomsky 1965) = language exists in the minds of speakers
23
The claims in (5) and (6) both link their respective areas of study to the cognitive realm, and will be crucial to the central argument in this article. They are bold and controversial claims, and I explain their bases in sections 3 and 4. In order to do that, one final connection between the two disciplines needs to be recognized: (7)
both historical linguistics and history1 rely on cognitive uniformitarianism
The assumption in (7) is that humans and the human mind have been qualitatively the same throughout history, so that the fundamental nature of the language that is investigated in historical linguistics and the thoughts that are considered in history1 have always been the same. The essential assumption is that both history1 and historical linguistics can only consider the history2 of modern humans and would not seek to consider anything beyond that (thus D’Oro 2006 writes that ‘[i]n ordinary usage history tends to be identified not with natural history but with the history of human af fairs’). If we assume that the brain has not evolved since the development of modern humans somewhere between 100,000 and 200,000 years ago, we can reasonably adopt a uniformitarian position on the mind: the ways in which our brains allow us to think have not changed, so uniformitarianism is realistic in history1, and as language is uncontroversially enabled in some way by what is in our minds (for Chomsky it is constrained by Universal Grammar which guides the acquisition of language), uniformitarianism in historical linguistics is reasonable, too. As Lass (1997: 28) shows, by quoting Whitney (1867: 24), a basic, non-cognitive uniformitarianism has been adopted by some in historical linguistics since at least the late nineteenth century, when it was in its prime:
24
Patrick Honeybone The nature and uses of speech […] cannot but have been essentially the same during all periods of its history […] there is no way in which its unknown past can be investigated, except by the careful study of its living present and recorded past, and the extension and application to remote conditions of laws and principles deduced by that study.
This practically spells out general uniformitarianism, which was first established in geology: the principles governing the world were the same in the past as they are today, and if we establish the principles that govern an object of enquiry by considering how it works in the present (or perhaps also, at Whitney says, in parts of the past that we understand well), we can restrict our ideas about how that object could have been in the parts of past that we need to understand. For our purposes, uniformitarianism means that (i) we should not allow for any kind of language during history2 which violates the linguistic principles that we have come to understand through our intuition- and corpus-based investigation of contemporary languages, and that (ii) we can assume that human actions and thoughts during history2 were produced on the basis of the same kinds of thoughts and impulses that we humans have today, which area available for investigation. If the claims in (5) and (6) are right, cognitive uniformitarianism is an important methodological guide. Understood this way, both history1 and historical linguistics are interested in things which existed in people’s minds during history2. The claim that both disciplines involve cognitive reconstruction is thus beginning to become clear. To justify it fully, we need to consider what it is that historical linguists and historians do. I turn to this in the next two sections.
History and Historical Linguistics
25
3 What do historical linguists do? In order to make my case, I need to show that there is an aim in (at least part of ) both historical linguistics and history to carry out cognitive reconstruction. The argument is easier to make, and more widely accepted, in historical linguistics. I explain its basis here. Two commitments from the philosophy of the discipline are needed to make the case. These are (i) the fundamental orientation towards mentalism in linguistics, mentioned in (6), and (ii) a commitment from historical linguistics concerning the status of the results of historical linguists’ work: we must be able to give them a realist, and not formulist, interpretation. I explain both of these brief ly below: the assumption of mentalism immediately, and the requirement of realism in section 3.1. The claim in (6) is that the object of study in linguistics is a mental, or cognitive, object. This is not uncontroversial, but we cannot consider all the arguments that have been had about the nature of language here; what is important, and is obvious to all, is that humans have knowledge of language stored in their minds – it is not contentious to say that we need to involve the cognitive realm when we speak. The idea that language is a cognitive entity is often associated with the ‘ideal speaker-hearer’ approach of generativism, one of the mainstream positions in linguistics (although not all aspects of the generative approach need be adopted for my fundamental claim to be accepted). From within this tradition, Chomsky has long argued that we need to recognize a distinction between ‘language’ and ‘the use of language’. This is essentially the same distinction that Saussure (1916) proposed between ‘langue’ and ‘parole’, so it has long roots. Building on an early (1965) distinction between competence and performance, Chomsky (1986) discusses a distinction between I-Language and E-Language. I-Language is an aspect of knowledge − it is the set of linguistic generalizations that we know in order to be able to speak our language (or languages). For example, we know that in English, the regular plural is formed by adding an alveolar
26
Patrick Honeybone
fricative which takes its voicing from the final segment in the word that it attaches to (so we say [kats] cats but [kabz] cabs), and we know that whphrases like what and how many occur at the start of clauses in questions. Like all types of knowledge, I-language is i-nternal to each i-ndividual. It is the grammar that exists in the mind/brain of speakers. ‘E-language’ refers to the utterances that speakers make when they talk or write. This connects with everything else that the ‘everyday sense’ of language involves, including everything connected to language which is e-xternal to the speaker, such as the communities that they belong to. The standard generative opinion is that linguists who aim to understand such purely (autonomously) linguistic generalizations about phonology, syntax and similar aspects of language should focus on I-language. E-language is subject to inf luence from many dif ferent areas: physical pressures af fecting linguistic performance, social and political pressures af fecting interactions, and historical contingency in terms of what speakers happen to talk about. The claim is that the set of utterances that speakers make is not a well-formed or coherent object, and it is not enough if we hope to discover the generative system that constitutes the language. We cannot wait for occurrences of everything that is possible in a language in corpora of utterances, and they cannot tell us what is impossible in a language, so E-language is not the right object of study if we aim to figure out the generative system that a speaker knows. This is not to say that E-language is uninteresting or unworthy of study: the sociology and history of languages and the interactions and perceptions of their speakers are major objects of study which we need in order to understand the human condition; corpus linguistics has told us a lot about collocation and frequency, but these things are facts about how speakers use their language, not about language itself. Mentalism need not be stripped of all aspects of linguistic variation, however (as it often is in generative approaches). In order to use their language appropriately, speakers need to know about the types of variation that the members of their speech community engage in, so this must be part of their knowledge about language – perhaps not precisely the same kind of knowledge that we have about purely structural linguistic, but it, too, points in the direction of mentalism for linguistics. The mentalist
History and Historical Linguistics
27
case that ‘language’ exists in the minds of speakers, as I-linguistic states of knowledge, has been clearly made in mainstream branches of linguistics. It gives linguists of a structural bent a coherent object of study, and I adopt it here. 3.1 Does historical linguistics involve cognitive reconstruction? We have seen that is it reasonable to claim that language is a cognitive entity. The case that historical linguistics engages in cognitive reconstruction is now not that dif ficult to argue. If language exists in the mind, and if one of the things that historical linguists do is to try to work out the structure of past synchronic linguistic states, then historical linguistics engages in cognitive reconstruction. In this section, I consider the ways in which historical linguists try to work out the structure of past synchronic linguistic states, and then I turn to issues that some linguists raise concerning the status of the reconstructions that are done in the discipline. One of the earliest and greatest achievements of historical linguistics was the proof that many currently existing languages are ‘genetically’ related to each other – that they derive in some real sense from the same ancestor proto-language. For example, we know that most of the languages spoken in Europe (and the countries to which they have been colonially exported in the past few centuries), and many spoken in Asia are derived from ProtoIndo-European (PIE). One of the central aims in nineteenth-century linguistics was to reconstruct proto-languages, especially PIE, about whose daughter languages much is known, and work on PIE continues today. The great tool for such work is the comparative method of reconstruction, in which forms for a particular linguistic feature are compared in related languages, to establish correspondences of form, which are then considered together to establish what is the most likely ancestral form from which they all could have derived (for a full description of the method, see Fox 1995). These forms are taken to be the reconstructed linguistic features of the proto-language. A classic example is the reconstruction of the PIE stop system. Although most European languages have two series of stops – often transcribed as
28
Patrick Honeybone
/p, t, k/ vs /b, d, g/ – we can be sure that PIE had three series, thanks to the evidence of languages like Sanskrit and Greek, which preserve multiple stop series. Also, we know that one of the series of PIE stops became fricatives in Proto-Germanic, so that contemporary English three, with an initial fricative, corresponds with Ancient Greek treis, Latin tres, Old Church Slavonic trije and Old Irish tri, for example, which all have stops (nominative masculine forms for three given here where the number declines in a language). To summarize some of this, (8) shows in its final column the classic reconstruction of the three series of PIE stops (with the coronal stop standing for the full series), which have been reconstructed on the basis of the oldest attested forms including those given in the other columns (taken from Job 1989). ‘OCS’ stands for Old Church Slavonic, the oldest attested Slavic language, ‘Vedic’ is a form of Sanskrit, and Gothic represents the Germanic languages, such as English, so it has a fricative (as in three) corresponding to stops in the other languages. The overwhelming preponderance of stops in non-Germanic languages shows that the PIE reconstruction of a stop is on firm ground. (8) Avestan
Vedic
Greek
Latin
Old Irish
Gothic
OCS
PIE
t
t
t
t
t
θ
t
t
d
d
d
d
d
t
d
d
d
dh
th
f
d
d
d
dh
The forms in the last column in (8) are an attempt to reconstruct the stop system of a language for which we have no records. Given that it was argued above that the only coherent locus to situate a language, in its technical sense, is in the minds of its speakers, this means that what is shown here is an attempt at cognitive reconstruction – a reconstruction of something that existed at some synchronic point in the minds of the speakers of PIE. Two points should be raised here. Firstly, it is common practice in historical linguistics to indicate forms which have been reconstructed in this way with an asterisk, to contrast them with forms which are attested in written documents and the like. This means that PIE /t/ is commonly
History and Historical Linguistics
29
represented as *t, for example. I do not adopt this practice here because all the forms given in (8) are reconstructed in some way, as discussed below, so the status of the PIE form is not fundamentally dif ferent from the other forms in (8). Secondly, we should note that the classic reconstruction of PIE stops has been subject to considerable challenge. It is not in doubt that there were three series of stops, but linguists have argued at length about the precise way in which they should be reconstructed. The best known alternative reconstruction is the ‘Glottalic Theory’ of Gamkrelidze & Ivanov (1973), which can be seen as contending that rather than voiceless/ plain, voiced and voiced aspirated series, there were voiceless aspirated, ejective and voiced aspirated series. Proposals such as the Glottalic Theory have been made for a number of reasons, in part centring around the idea that the classic reconstruction is typologically highly unlikely, because we do not expect to find languages with a breathy voiced (‘voiced aspirated’) series, like /dh/ if they do not also have a voiceless aspirated series, like /th/; that is, the uniformitarianism of (7) is applied. Far from casting doubt on the assumption that the final column in (8) is a case of cognitive reconstruction, however, the fact that linguists argue about the best way to reconstruct a language shows that they care about getting the cognitive reconstruction right. The uniformitarian constraints of general linguistics on what is a possible language only apply to a reconstructed language if the reconstruction is taken to be psychologically real. By PIE, then, we mean the I-linguistic grammar that a speaker of PIE had. In the same way, when historical linguists work to figure out any aspect of an attested language such as Old English (OE), they are trying to reconstruct the mental grammar of speakers of OE (at a particular point in its development, such as classical West Saxon). It may be that in certain cases we can never be successful in reconstructing complete past I-languages – if evidence in a particular case is scant, for example – but we can still hope to reconstruct at least parts of past languages fully, such as the plosive or fricative system of the language, or its nominal morphology, and these parts were mental (sub)systems in their own right. To consider an example of synchronic cognitive reconstruction from OE, we know from a range of types of evidence that OE had essentially only one series of fricatives, of the type /f, θ, s/; there were no voiced
30
Patrick Honeybone
fricatives of the type /v, đ, z/. We also know that OE fricatives were subject to a phonological process of intersonorant voicing. To exemplify with the labial-dentals, this means that the fricatives in fisc ‘fish’ and wulf ‘wolf ’ were voiceless [f ] on the surface, but that those in wulfas ‘wolves’ and drifan ‘drive’ were voiced [v] (as many have shown, such as Hogg 1992b, Lass 1994).4 This takes us beyond comparative reconstruction. However, while evidence in the form of written texts is available for this reconstruction, the issues for the historical linguist are essentially the same. This is why I do not adopt the asterisk to mark ‘reconstructed’ forms in PIE: all historical linguistics deals with reconstructed forms. Historical linguists aim to work out past I-languages, and it is clear that past I-languages, as objects of enquiry, cannot be observed. This is exactly the same for OE (Middle Dutch, nineteenth-century Korean, etc.) as it is for PIE: none of them can be observed. The ultimate object of study for historical phonologists who work with attested languages is not the written forms. Written forms can provide important evidence for historical phonology, but they need to be interpreted with skill and caution (see, for example, Lass to appear, Minkova to appear). Thus, the OE letter is not a phonological (I-linguistic) object, but fisc, wulf, wulfas, drifan, as examples of preserved E-language from the past, can be used as evidence in the reconstruction of the I-language of the speakers who wrote the Old English texts that survive. In the same way, the comparison of attested forms such as three with an initial fricative (and Gothic nominative neuter þrija, the oldest Germanic form for ‘three’) with the overwhelming majority of other Indo-European languages which have a stop (such as Latin tres and Old Church Slavonic trije), as summarized in (8), can be used as evidence in the reconstruction of a /t/ in the I-language of the PIE speakers (who happened not to leave us any written records). The OE reconstruction of an underlying /f/, which was realized as [f ] and [v], is a reconstruction − it is not anything that is found 4
In fact, the precise laryngeal characteristics of the segments involved are subject to debate, like those of IE, but for dif ferent reasons (see Honeybone 2005). The current article is not the place for that debate, however, and the basic point stands whatever position is taken in laryngeal phonology. Once again, the fact that there is such debate shows that it is important to get the reconstruction right.
History and Historical Linguistics
31
in texts. Past phonological systems need to be reconstructed by (historical) phonologists using evidence from written sources (and from elsewhere, such as the forms found in living descendent dialects, and direct evidence from commentaries on pronunciation, where it is available), and this means that both OE /f/ and PIE /t/ are reconstructed and thus have the same phonological status, meaning that neither (or both) need an asterisk.5 We have dealt with historical phonology in some detail here, but the same conceptual issues arise in the reconstruction of other linguistic levels. I give here much briefer examples of the reconstruction of past syntactic and semantic states. The same range of indirect and direct evidence is available to do such reconstruction (written records, the comparative method, linguistic commentaries from the past), but comparative reconstruction is less successful in syntax because the systems involved are less straightforwardly compared. A syntactic example: OE is reconstructed as a verb-second (V2) language, similar to contemporary German and Dutch. This means that the finite verb is expected to be the second constituent in main clauses (some syntactic processes can ‘hide’ this order, but it is assumed to be fundamental). Unlike in contemporary English, where the subject typically comes directly before the verb, many dif ferent kinds of constituents could directly precede the verb in OE: subjects, objects, adverbials etc. The sentences in (9), from Orosius, and (10), from the Anglo-Saxon Chronicles (both taken from Kroch & Taylor 1997), show the ef fects of V2: as is shown in the aligned gloss below the original OE, (9) has the finite verb (marked as
5
In fact, it can be argued that the reconstructed linguistic entities of past languages have the same status as analyses of contemporary I-languages, too. Neither a past linguistic state nor a present one can be directly observed by a linguist. Rather, we need to use whatever E-linguistic evidence is available to us to figure out what the linguistic system is that we are investigating, and the result of linguistic analysis – the model of linguistic knowledge that we develop – is always constructed by the linguist. There is more and better evidence available to those investigating a contemporary language, such as native speaker intuitions, recordings, controlled experiments and massive corpora, but this construction of analyses of contemporary languages is essentially the same kind of activity as the reconstruction of ideal languages: both are building models of I-languages using the best available evidence.
32
Patrick Honeybone
‘V2’), preceded by its direct object (‘O’) and followed by the subject (‘S’), and (10) has an adverbial (‘A’) where (9) has an object. The reconstruction of the synchronic syntactic patterns which allowed this type of word order need to be figured out by historical syntacticians, and several subtle syntactic subgeneralizations about the precise patterning of V2 constructions in OE have been discovered in a long strand of research on the topic (for example, van Kemenade 1987, Pintzuk 1991, Kroch, Taylor & Ringe 2000). (9) þæt hus
hæfdon Romane
to
ðæm anum tacne
that building had
Romans
with the
O
S
etc.
(10) þær
V2
one
wearþ
se cyning Bagsecg
ofslægen
there
was
the king Bagsecg
slain
A
V2
S
etc.
geworht
feature constructed
A semantic example: as is well known, and as was discussed in section 2, the meaning of words in and from the past needs subtle reconstruction. As an example of this, (11a) shows one sentence from an English letter from around 1740, from a Dr Rutherforth of Cambridge, which requires careful reconstruction of the meanings of most of its content words in order to be properly understood. This is shown in the translation into contemporary English in (11b). Both text and translation are taken from Honeybone & Honeybone (2010, p.c.); the main cases of semantic reconstruction are identified through underlining, emboldening and italics. (11a) I suppose that natural philosophy and polite literature are the branches of science that you chief ly improve (11b) I assume that the fields of knowledge that you work in are chief ly science and philology
The meanings of words exist in the minds of the speakers of the languages that they are part of, so the meanings that have been reconstructed in (11b) are things that existed in the minds of a speaker of mid-18th-century English. This is thus, once again, a case of I-linguistic, cognitive reconstruction.
History and Historical Linguistics
33
In working to understand the linguistic systems and states discussed above, structural historical linguists are (whether they describe it as such or not), aiming to reconstruct past cognitive states. This has led to the writing of largely synchronic historical grammars, such as Campbell (1959), Hogg (1992a) and Braune, Ebbinghaus & Heidermanns (2004). It also expands the database of synchronic linguists, in the quest to understand what is possible or common in language. For example, Dresher & Lahiri (1991) show that the structure of the metrical foot in the phonology of Early Germanic languages was an uneven moraic trochee, a type of foot which had previously been claimed to be unattested (see Kager 1995). In order to make the case that historical linguistics engages in cognitive reconstruction, one final assumption (which has thus far been silently made) needs brief discussion. The ontological status of the entities that historical linguists reconstruct has, in fact, been a subject of some controversy in the literature. Fox (1995: 9), discussing the status of the forms derived through comparative reconstruction, explains the issue thus: Are we entitled to claim for such reconstructions the status of earlier linguistic forms? This controversy can be summed up in a confrontation between two views of reconstruction: the formulist and the realist. The formulist view regards reconstructions merely as formulae which represent the various relationships within the data, while the less cautious realist view assumes that reconstructions can be taken to represent genuine historical forms of a real language, which happen not to have been recorded.
As discussed above, I take all reconstructions – both those done on the basis of the comparative method and those derived from written sources – to have the same status, so at least some of the points considered here are relevant to all aspects of reconstructive historical linguistics (although they have largely been previously considered in relation to the results of comparative reconstruction). Can we really assume that (for example) the phonological system reconstructed for PIE, or OE, or any other past language, should be seen as a ‘real’ I-language? The formulist position sees the reconstructions of historical linguists as useful fictions which are essentially place holders for the comparisons themselves – they should not be taken to be the real forms of any language. This position has inf luential advocates, for example Meillet (e.g. 1903), but it must be wrong. It is only if we assume that
34
Patrick Honeybone
reconstructed languages are real languages, and should thus be given the same status as the models of I-language that we construct for contemporary languages, that uniformitarianism can place a constraint on reconstruction (as we saw above that it must). Linguistic typology and linguistic theory can only be used in arguments about reconstruction (such as the Glottalic Theory, for example), if we assume that reconstructed languages are the same sort of thing as the languages that linguists work with to figure out typological and theoretical generalizations in the first place. If we do not, then we have no control on what we can reconstruct. (The need to reject postmodern doubt that we can ever understand the things going on in the minds of others for these claims to hold will be clear.) The realist position is well supported in historical linguistics. The authors of the two most detailed recent considerations of historical linguistic method (Fox 1995 and Lass 1997) both tend towards realism in reconstruction. Thus Fox (1995: 13) writes that ‘as realists rightly maintain, to be significant the results [of reconstruction] must be interpretable in historical terms’, and Lass (1997: 274) that ‘protolanguages […] are simply languages like any others, if with poorer information sources’. The formulist position seeks to place the things that are reconstructed outside of the minds of speakers, by claiming that they are not languages at all, but this claim is not consistent with the practice of historical linguists, nor is it compelling. The aspects of languages that are reconstructed in historical linguistics must be assumed to represent aspects of past I-languages. We should not claim that languages are always (or ever) fully reconstructable (especially not if we assume that variation is grammatically inherent in language, as it is often dif ficult to reconstruct much of the way in which a language must have varied) but we can reconstruct parts of the linguistic systems involved (such as the underlying phonological contrasts of IE, or finite verb placement in OE), and these parts existed in the minds of speakers. Historical linguistics does, then, in the ways described above, seek to engage in cognitive reconstruction.
History and Historical Linguistics
35
3.2 Does historical linguistics only involve cognitive reconstruction? Any acquaintance with work on structural historical linguistics will show that it does not just involve the reconstruction of past linguistic systems. The synchronic aspect of historical linguistics can be underestimated, but it is subservient to diachronic study: much of historical linguistics aims to understand how and why one linguistic state can change into another. Thus, part of the interest in comparing Gothic þrija with Old Church Slavonic trije certainly is to reconstruct PIE forms, but part of it is to understand the phonological change that occurred in Germanic, by which stops became fricatives – part of the famous Grimm’s Law. And part of the interest in working out that OE had a single series of underlying fricatives is to go on to find out how this changed, so that we now have two series. English has lost V2, at least in the basic type of clauses illustrated in (9) and (10), so that the subject must now always precede the verb, and this has been the major focus of work on V2 in the history of English. Also, of course, the changes that can be seen by comparing (11a) and (11b) have occurred and are of major interest to historical linguists. Synchronic systems are often only reconstructed so that they can be compared with other, temporarily distinct systems and changes can be observed. One central aim of historical linguistics is to work out a theory of linguistic change, addressing such questions as: how does change happen? Can it happen within an individual? Or does it typically happen cross-generationally, when a new generation misacquires the language of their parents? Historical linguistics can try to work out if anything is possible in change, or whether we can formulate a notion of what would be an ‘impossible change’,6 and if so, why they are impossible. These are diachronic questions and particular changes may not be anything to do with I-language – they could be due to population movements, or misperception, or structural ambiguity and
6
Indeed, it seems that we can: a change such as p, t, k > f, θ, x can occur, and did as part of Grimm’s Law, thus explaining the correspondence of non-Germanic /t/ with Germanic /θ/, but the reverse does not ever seem to be attested (see, for example, Kümmel 2007).
36
Patrick Honeybone
the unconscious choice of speakers to adopt a dif ferent linguistic system. While I-linguistic factors may be involved (if change can occur within a grammar), they need not be. Historical linguistics does engage in cognitive reconstruction, but that is not all that it does.
4 What do historians do? Do historians engage in cognitive reconstruction? As was the case in the last section, certain philosophical commitments need to be made in order for this claim to hold. The necessary conceptual position in history is probably proportionately less common than that sketched for historical linguistics above, but it is compelling, is well represented in the literature and is accepted (if maybe tacitly) in certain key intellectual traditions. 4.1 Does history involve cognitive reconstruction? The commitment in terms of the philosophy of history that is needed to make my case is an interpretation of the (at least partly) idealist approach to history, which is associated mostly closely with the historian R.G. Collingwood. Idealism focuses on the mental world, leaning towards the notion that a range of things (which might not seem to do so on a ‘common-sense’ approach) actually only exist in the mind. At its most extreme, idealism can argue that no aspect of reality exists independent of our perceptions, but Collingwood was wary of being identified as a total idealist (D’Oro 2006), and we do not need to subscribe to full-blown philosophical idealism to accept Collingwood’s ideas (in the same what that it is not actually necessary to accept full-blown generativism in order to accept the mentalist understanding of language which fits with Chomsky’s ideas). We can retain the term ‘idealism’ to describe the approach, despite these and Collingwood’s own caveats, however, as is conventional in the philosophy of history. The idealist view of history can be understood as cognitivist in this way, in part
History and Historical Linguistics
37
echoing (3a): (i) history aims to understand past actions; (ii) these actions are due to human causality; (iii) human causality is due to the thoughts that the actors have; and (iv) thoughts exist in the cognitive realm. The Idea of History (1946) is Collingwood’s classic work on the philosophy of history (Collini & Williams 2004). It includes his famous definition of the idea of history: ‘history should be (a) a science, or an answering of questions; (b) concerned with human actions in the past; (c) pursued by interpretation of evidence; and (d) for the sake of human self-knowledge’ (1946: 10–11). The relevant points here are (b), as mentioned above, and (d), which provides the perspective which leads to the ideas in focus here. Collingwood argued that we can understand ourselves through history because we can develop an understanding of psychology through history, and that this is because history is, practically, psychology. Collingwood wrote about a conception of psychology as a type of historical study, whose aim ‘would be to detect types or patterns of activity, repeated over and over again in history itself ’ (1946: 224, see Connelly & Costall 2000), and this requires us to understand the thoughts of those who performed these activities in the past. (The need to reject postmodern doubt that we can understand the minds of others will again be clear for these claims to hold.) The crucial part of Collingwood’s argument for our purposes is his emphasis ‘upon history as the interpretation of purposive action and hence upon the primacy of the need to recover historical agents’ own understanding of their situation’ (Collini & Williams 2004). This is linked to his conception of psychology, but is not tied to it. Like many others, Collini & Williams (2004) cite Collingwood (1946: 215) as his most famous claim, given here in (12). (12) ‘all history is the history of thought … and therefore all history is the re-enactment of past thought in the historian’s own mind’
Collini & Williams (2004) write further that ‘this was intended not as a prescription of the method of historical enquiry, but a description of its success: only insofar as historians can “re-enact” in their own thought what an act meant to its agent can they explain that act.’ This leads us to an understanding of the definition of the object of study in history as actions, as D’Oro (2006) explains:
38
Patrick Honeybone The subject matter of history, understood as a science of the mind, is actions − actions understood not simply as the doings of human beings but of human beings in so far as they are rational. Actions, in the sense in which they constitute the subject matter of historical investigation have an ‘inside’ that events lack. To explain an event all we need to do is to subsume it under a general law that is obtained by inductive generalisation, through the observation of repeated events of type B following events of type A. In order to understand an action, by contrast, we need to render it intelligible by reconstructing the thought processes that inform it.
On this approach, the key goal of history is cognitive re-enactment, as in (12) – the historian should aim to re-enact (that is, to reconstruct) the thought processes that led to the historical actions that history tries to understand. The uniformitarian principle of (7) is important here: we must assume that the mind has not changed in the way that it manipulates or creates thoughts throughout history, for if it had, past thoughts could be unrethinkable. If we take Collingwood at his word (1946: 301), re-enactment should be absolutely literal cognitive reconstruction: the very thought that an agent had in the past can be repeated. […] in its immediacy, as an actual experience of his own, Plato’s argument must undoubtedly have grown up out of a discussion of some sort, though I do not know what it was, and been closely connected to such a discussion. Yet if I not only read his argument but understand it, follow it in my own mind re-enacting it with and for myself, the process of argument which I go through is not a process resembling Plato’s, it is actually Plato’s so far as I understand him correctly.
The claim is not that we need to engage in telepathy. Rather, ‘it is possible in principle to re-enact the thoughts of others because thoughts, unlike physiological processes, are not private items unique to the person who has them, but publicly rethinkable propositional contents’ (D’Oro: 2006). D’Oro further explains that this approach has been seen as counterintuitive: since to say that the thought of the agent and that of the historian are the same appears to presuppose that there is only one rather than two numerically distinct acts of thought, that of the historian and that of the agent. Collingwood’s point, however, is that, since thought proper is conceptually distinct from the physiological process in which it is instantiated, the criterion of numerical identity that is usually applied to physiological processes is not applicable to thought. Thoughts, in other
History and Historical Linguistics
39
words, are to be distinguished on the basis of purely qualitative criteria, and if there are two people entertaining the (qualitatively) same thought, there is (numerically) only one thought since there is only one propositional content.
If a historian is able to fully reconstruct the thoughts7 that led someone from the past to carry out certain actions, they can be said to understand those actions, and will thus be able to write accurate historical works about those actions. In this sense, historians must engage in cognitive reconstruction. Although the act of writing about this is not reconstruction in itself, the same applies to the act of a historical linguist writing about the PIE consonantal system or OE syntax: the reconstruction must be done first, in order for the work which focuses on it to be written. Part of such work might indeed simply be a record of the reconstructions: a table of PIE consonants, or a description of the thinking of a historical agent, but much of it will be ref lection on the result of the reconstruction, and an attempt to place it in its context. The implications of all this are that, in order to understand the actions of those who agitated for national self-determination in Austria-Hungary, for example, or who persecuted non-Catholics in fifteenth-, sixteenth- and seventeenth-century Spain, or who agreed the Acts of Union between England and Scotland, we need to rethink the thoughts that led to the actions involved. Most of the evidence for this is found in texts which are contemporary with (or, maybe, later than) the actions involved, and considerable skill is required to interpret these documents (informed by historical linguistics where necessary, as discussed above, particularly around (11)). The approach sketched out here requires the historian to understand the mindset of the people that they are interested in, to develop a sense of empathy with them (again, this rejects postmodern arguments, such as those of Jenkins 1991). An example of the impact of this approach is considered in M. Honeybone (2008: 16), who is discussing the writing of the history of a well-known witchcraft trial from the 17th century in the Vale of Belvoir in the English Midlands: 7
We may not need to assume that the very same thought exists in the minds of the original actor and the historian, as Collingwood argued – if a thought is simply repeated, we may escape the danger of Collingwood’s mind-body distinction.
40
Patrick Honeybone European scholars have appreciated that the concept of early-modern witchcraft arose as an imaginative response to the widespread myth of the witches’ sabbath, defined as ‘A midnight meeting of demons, sorcerers and witches, presided over by the Devil, supposed in mediæval times to have been held annually as an orgy or festival.’ [SOED] After extensive historical investigation it is now accepted that no actual ‘sabbath’ ever took place. Equally, however, the notion of the sabbath existed absolutely in the minds of sixteenth- and seventeenth-century men and women. This is a perfect justification for Collingwood’s concept of history as the process of rethinking the thoughts of past people: ‘wrong ways of thinking are just as much historical facts as right ones, and, no less than they, determine the situation (always a thought situation) in which the man who shares them is placed’. (Collingwood, 1946: 317)
Reconstructing past thoughts is surely not easy, but it is just as surely the ideal that we should strive for in attempts to understand the past (at least, if we follow Collingwood). As the above point shows, if we are open to all the beliefs of those whose past we consider – ‘false’ beliefs and the general intellectual context, as well as anything we can glean about an individual writing a document – we can hope to come close enough to rethink the thoughts of those whose actions we are interested in. Other historical artefacts, such as buildings and clothing from the past, can also help in the interpretation of past people’s thoughts, and this has led some to reject the idealist position. As Evans (1997: 92) writes, ‘[m]any sources are not written at all. Getting inside the head of someone who buried treasure in a grave in the fourth century, or made a newsreel in the twentieth, is far from easy.’ This is a reasonable point, of course, but it is reminiscent of the mistaken distinction in historical linguistics between reconstructed forms which are standardly given an asterisk and those that are not, as discussed above. Those forms without asterisks are still reconstructed, even though they derive from written sources. Carr (1961: 16) explains that, in the same way, written sources, while privileged among types of evidence, still require a considerable amount of interpretation. He criticizes the nineteenthcentury ‘fetishism of facts and documents’, explaining that: The documents were the Ark of the Covenant in the temple of facts. The reverent historian approached them with bowed head and spoke of them in awed tones. If you find it in the documents, it is so. But what, when we get down to it, do these documents – the decrees, the treaties, the rent-rolls, the blue books, the of ficial correspondence, the private letters and diaries – tell us? No document can tell us more
History and Historical Linguistics
41
than what the author of the document thought – what he thought had happened, what he thought ought to happen or would happen, or perhaps only what he wanted others to think he thought, or even only what he himself thought he thought. None of this means anything until the historian has got to work on it and deciphered it.
Just as linguistic reconstruction can proceed on the basis of texts or comparison, historical reconstruction can proceed on the basis of texts or other historical material, but information found in documents still needs processing to form the basis of reconstruction. It all requires careful interpretation in order for us to be able to do the reconstruction. The comparative method and the use of non-written sources simply need more care and provide less good evidence. Although wary of some of the implications of Collingwood’s ideas, Carr (1961) is generally sympathetic to the idealist approach, because it shows that history is not just a matter of finding dusty facts in documents. Thus, also citing Collingwood (1946), with an echo of (12), Carr (1961: 22) writes: ‘The past which a historian studies is not a dead past, but a past which in some sense is still living in the present.’ But a past act is dead, i.e. meaningless to the historian, unless he can understand the thought that lay behind it. Hence ‘all history is the history of thought,’ and ‘history is the re-enactment in the historian’s mind of the thought whose history he is studying.’ The reconstitution of the past in the historian’s mind is dependent on empirical evidence. But it is not in itself an empirical process, and cannot consist in a mere recital of facts. On the contrary, the process of reconstitution governs the selection and interpretation of the facts: this, indeed, is what makes them historical facts.
Given all of the above, it is not unreasonable to conclude that historians need to reconstruct past cognitive states and actions (although this may not be a historian’s conscious aim) when they work to understand past actions. It may be that this must sometimes remain an unachievable ideal, in that we can only get close to the original thought, because evidence for it is lacking. The ideal remains, though, that the thoughts that provided the context for the actions that we know of need to be reconstructed, and in principle can be rethought in the mind of the historian (who can then aim to invoke them in the mind of the reader of their work). History does, then, in the ways described above, seek to engage in cognitive reconstruction.
42
Patrick Honeybone
4.2 Does history only involve cognitive reconstruction? Just as there are with historical linguistics, there are some aspects of history which are not themselves attempts at cognitive reconstruction. We have already considered that the actual writing of history, just like the writing of historical linguistics, may rely on reconstruction, but also involves much more than this. It is also surely the case that much of the work done by the historian in order to get to the position of being able to re-enact past thoughts is not reconstruction itself: the work that a historian carries out with documents and artefacts, for example, and conducting the background work which is necessary to ‘do the history’. This is rather dif ferent from the conclusion of section 3.2. It may be that history is essentially a synchronic discipline, to the extent that it does not seek to deal with diachronic change to the same extent that historical linguistics does. Rather than largely focusing on processes of change between one system and another, and the possible types of such changes that can occur, there is more emphasis in history than in historical linguistics on understanding particular times and places. While not all of the conduct of history is an engagement in cognitive reconstruction, more of it is than is the case in historical linguistics.
5 Two types of cognitive reconstruction The case has been made above that the two disciplines under discussion here both engage in cognitive reconstruction. We have seen that it is not all that they do, and it should be remembered that I am focusing on structural historical linguistics in my discussion here, but I have argued that cognitive reconstruction is fundamental to the success of both. One final point needs to be considered: while the notion of cognitive reconstruction connects the two disciplines, it also distinguishes them, because the two types of cognitive reconstruction may be conceptually similar, but they are also fundamentally dif ferent.
History and Historical Linguistics
43
There are clear conceptual connections between the two types of reconstruction. For example, both rely on uniformitarianism − that the human mind has been the same throughout history2, since we evolved as a species, which means that the nature of language and of thought have always been the same. This, in turn, means that we stand a chance of being able to engage in cognitive reconstruction as part of our disciplines. Furthermore, both disciplines have to work with sources which are imperfect ref lections of their true object of study: often written sources, which need careful interpretation. There is a substantial dif ference between the two types of reconstruction, however. This derives from a distinction between the objects of study in the two disciplines, set out here in (13). (13a) language is part of the unconscious mind (13b) agentive thoughts are part of the conscious mind
The distinction in (13) is an important one, and it explains many of the dif ferences between the approaches discussed above. While the distinction between conscious and unconscious mental activity is a complex one – far too complex to be properly considered here – we can assume a basic distinction between those aspects of the cognitive world to which we can, in principle, have direct access, and those to which we cannot. We know an extraordinary number of subtle generalizations about the languages that we speak, but we also know that we do not have direct conscious access to this knowledge – we have intuitions about what is ungrammatical, but we cannot easily explain why. Linguistic rules are part of our unconscious knowledge. Contrary to this, we can ref lect on the thoughts which lead us to perform specific actions, and such agentive thoughts are the subject matter of history, on the interpretation developed above. One implication of (13) is that historical linguists do not try to reenact a past linguistic state in their own minds, whereas, if we follow the idealist position, this is what historians should try to do. Of course, the reconstruction that historical linguists engage in is a product of their minds, but there is no expectation that the reconstruction should function in their minds as it did in the minds of the original speakers. In history, there
44
Patrick Honeybone
is some sense in which a reconstruction should function in the same way in the historian as it did in the original thinker. In a sense, then, historical linguistics engages in a type of cognitive reconstruction which is less connected to the researcher than is the case in history: historians use the part of the mind that the object of the reconstruction resided in to do the reconstruction, whereas historical linguists reconstruct unconscious linguistic knowledge with conscious intent. These dif ferences mean that quite distinct methodological tools and approaches are appropriate for the dif ferent types of reconstruction that the two disciplines engage in. Thus, while structural historical linguists and historians can surely learn from each other, we cannot expect to do things in the same way. We can likely better understand both disciplines if we see the dif ferences as just as important as the similarities. To conclude: on the basis of the philosophical positions defended here, historical linguistics and history do indeed both seek to engage in cognitive reconstruction, although the reconstruction involved is of dif ferent types. It may be that we do not always succeed in our attempts at such reconstruction, or that the reconstruction can only be partial, and it is clear that both disciplines do not only engage in such reconstruction, and sometimes only do it so that they can consider something else, but I hope to have shown here (and to have considered a range of non-obvious but connected issues along the way) that historical linguistics and history are both connected and distinguished by the notion of cognitive reconstruction.
References Borsley, R. 2000. ‘A postmodern critique of linguistics’. Ms, University of Essex. accessed 12 August 2011. Braune, W. 1883. Gothic Grammar with selections for reading and a glossary. Trans. G.H. Balg. New York: Westermann. Braune, W., Ebbinghaus, E., and Heidermanns, F. 2004. Gotische Grammatik, 20th edn. Tübingen: Niemeyer.
History and Historical Linguistics
45
Burke, P. 2004. Languages and Communities in Early Modern Europe. Cambridge: Cambridge University Press. Campbell, A. 1959. Old English Grammar. Oxford: Clarendon Press. Carr, E.H. 1961. What is History? Harmondsworth: Penguin. Chomsky, N. 1965. Aspects of the Theory of Syntax. Cambridge, MA: MIT Press. Chomsky, N. 1986. Knowledge of Language: its Nature, Origin, and Use. New York: Praeger. Collingwood, R.G. 1946. The Idea of History. Oxford: Clarendon Press. Collini, S., & Williams, B. 2004. ‘Collingwood, Robin George (1889–1943)’. In: Oxford Dictionary of National Biography. Oxford: Oxford University Press. accessed 12 August 2011. Connelly, J., & Costall, A. 2000. ‘R.G. Collingwood and the Idea of a Historical Psychology.’ Theory & Psychology 10, 147–170. D’Oro, G. 2006. ‘Robin George Collingwood.’ In: Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy, accessed 12 August 2011. de Saussure, Ferdinand. 1916. ‘Cours de linguistique générale’ In: Bally, C., & A. Sechehaye (eds), with A. Riedlinger, Lausanne and Paris: Payot. Trans. W. Baskin, Course in General Linguistics. Glasgow: Fontana/Collins, 1977. Derrida, J. 1967. De la grammatologie. Paris: Editions de Minuit. Dobson, E. 1968. English Pronunciation 1500–1700. 2 vols. 2nd edn. Oxford: Clarendon Press. Dresher, B.E. 1999. ‘Ferdinand, we hardly knew you.’ Glot International 4/6, 9. accessed 12 August 2011. Dresher, B.E., & Lahiri, A. 1991. ‘The Germanic Foot: Metrical coherence in Germanic.’ Linguistic Inquiry 22, 251–286. Evans, R.J. 1997. In Defence of History. London: Granta. Evans, R.J.W. 1998. The Language of History and the History of Language. Inaugural lecture delivered before the University of Oxford. Oxford: Clarendon Press. Fox, A. 1995. Linguistic Reconstruction: An Introduction to Theory and Method. Oxford: Oxford University Press. Gamkrelidze, T., & Ivanov, V. 1973. ‘Sprachtypologie und die Rekonstruktion der geminindogermanischen Verschlüsse.’ Phonetica 27, 150–156. Hart, J. 1569. An Orthographie: Conteyning the Due Order and Reason, howe to Write or Paint Thimage of Mannes Voice, Most Like to the Life or Nature. London: William Seres. Hogg, R. 1992a. A Grammar of Old English. Vol 1: Phonology. Oxford: Blackwell. Hogg, R. 1992b. ‘Phonology and Morphology.’ In: R. Hogg (ed.), The Cambridge History of the English Language, I: The Beginnings to 1066. Cambridge: Cambridge University Press, 67–167.
46
Patrick Honeybone
Honeybone, D., & Honeybone, M. 2010. The Correspondence of the Spalding Gentlemen’s Society 1710–1761. Lincoln: Lincoln Record Society. Honeybone, M. 2008. ‘Practising history: Writing a history book today’. East Midland Historian 16, 12–17. Honeybone, P. 2005. ‘Diachronic evidence in segmental phonology: the case of obstruent laryngeal specifications.’ In: van Oostendorp, M., and van de Weijer, J. (eds), The Internal Organization of Phonological Segments. Berlin: Mouton de Gruyter, 319–354. Jenkins, K. 1991. Re-thinking History. London: Routledge. Job, M. 1989. ‘Sound change typology and the “Ejective Model”.’ In: Vennemann, T. (ed.), The New Sound of Indo-European. Berlin: Mouton de Gruyter. Jones, C. 1989. A History of English Phonology. London: Longman. Kager, R. 1995. ‘The metrical theory of word stress.’ In: Goldsmith. J. (ed.), The Handbook of Phonological Theory. Oxford: Blackwell, 367–402. Kemenade, A. van 1987. Syntactic Case and Morphological Case in the History of English. Dordrecht: Foris. Kroch, A., & Taylor, A. 1997. ‘Verb movement in Old and Middle English: Dialect variation and language contact.’ In: van Kemenade, A., and Vincent, N. (eds), Parameters of Morphosyntactic Change. Cambridge: Cambridge University Press, 297–325. Kroch, A., Taylor, A., & Ringe, D. 2000. ‘The Middle English verb-second constraint: a case study in language contact and language change’. In: Herring, S., van Reenan, P., and Schosler, L. (eds), Textual Parameters in Older Languages. Amsterdam: John Benjamins, 353–391. Kümmel, M. 2007. Konsonantenwandel: Bausteine zu einer Typologie des Lautwandels und ihre Konsequenzen. Wiesbaden: Reichert. Lass, R. 1994. Old English: A Historical Linguistic Companion. Cambridge: Cambridge University Press. Lass, R. 1997. Language Change and Historical Linguistics. Cambridge: Cambridge University Press. Lass, R. forthcoming. ‘Interpreting alphabetic orthographies and orthographic change.’ In: Honeybone, P., and Salmons, J. (eds), The Oxford Handbook of Historical Phonology. Oxford: Oxford University Press. Marwick, A. 2001a. ‘The Fundamentals of History’. History in Focus 2. accessed 12 August 2011. Marwick, A. 2001b. The New Nature of History: Knowledge, Evidence, Language. Basingstoke: Palgrave. Meillet, A. 1903. Introduction à l’étude comparative des langues indo-européennes. Paris: Hachette.
History and Historical Linguistics
47
Minkova, D. forthcoming. ‘Establishing phonemic contrast in written sources: direct and indirect evidence.’ In: Honeybone, P., and Salmons, J. (eds), The Oxford Handbook of Historical Phonology. Oxford: Oxford University Press. Munslow, A. 2001. Author’s response to O’Brien (2001). History in Focus 2. accessed 12 August 2011. Newmeyer, F. 1986. The Politics of Linguistics. Chicago: University of Chicago Press. O’Brien, P. 2001. Book review: ‘An Engagement with Postmodern Foes, Literary Theorists and Friends on the Borders with History.’ History in Focus 2. accessed 12 August 2011. Pei, M. 1965. Invitation to Linguistics: a Basic Introduction to the Science of Language. London: Allen & Unwin. Pintzuk, S. 1991. Phrase structures in competition: Variation and change in Old English word order. PhD thesis, University of Pennsylvania. The Shorter Oxford English Dictionary on Historical Principles (SOED). 1987. 3rd edn. Oxford: Oxford University Press. Törnquist-Plewa, B. 2000. ‘Contrasting ethnic nationalisms: Eastern central Europe’. In: Barbour, S., and Carmichael, C. (eds), Language and Nationalism in Europe. Oxford: Oxford University Press, 183–220. Trudgill, P. 2004. New-Dialect Formation: The Inevitability of Colonial Englishes. Edinburgh: Edinburgh University Press. Warkentin, J. 2009 (1985). ‘Historical Geography.’ The Canadian Encyclopedia. Historical Foundation of Canada. accessed 12 August 2011. Whitney, W. 1867. Language and the Study of Language: Twelve Lectures on the Principles of Linguistic Science. London: Trübner.
Nicholas M. Wolf
History and Linguistics: The Irish Language as a Case Study in an Interdisciplinary Approach to Culture
Abstract The Irish language has been accorded a central place in historical scholarship on modern Ireland as the question of the causes and scope of its decline has become closely intertwined with explanations for the broader cultural transformation that occurred in these same years. Yet prevailing accounts of the relationship between cultural and linguistic change raise a number of uncertainties about current Irish historiography in this area, prompting an opportunity to investigate how a better disciplinary interaction between history and linguistics might improve scholarly approaches to Ireland’s past. This chapter begins by reviewing the scholarship on this subject in an Irish context before stepping back to consider the larger methodological possibilities and challenges of combining the studies of cultural history and language. Finally, this chapter will sketch out new approaches to Ireland’s linguistic history that take a more integrative approach and are intended to prompt the possibility of their application in the study of other historical and geographic contexts as well.
1 Irish historiography and language Historians of Ireland have devoted considerable attention to the island’s linguistic past. Undoubtedly, this willingness to focus on language can be attributed to the heightened public visibility of the movement to revive Irish over the past one hundred years, including ongoing debates today in both Northern Ireland and the Republic over how best to reverse its decline. Yet
50
Nicholas M. Wolf
the attention devoted to language as part of Ireland’s larger history cannot simply be attributed to its continued contemporary relevance. On the contrary, historians have long portrayed the decline of Irish – and particularly the precipitous shift toward English of the nineteenth century in which the proportion of Irish-speakers declined from approximately one-half of the island’s population in 1800 to less than 15 per cent in 1900 – as a key turning point in the emergence of modern Ireland. The importance accorded to language shift in Ireland’s history was already evident in the first full-length consideration of the issue, Daniel Corkery’s The Fortunes of the Irish Language (1954). For Corkery, the history of the Irish language followed the same plotlines as the history of the Irish nation itself. Much as the Irish nation had been usurped by British rule before re-emerging at the end of the nineteenth century with the establishment of a full-f ledged nationalist movement at the popular level, the Irish language had been undermined by a foreign government before being slowly returned to the centre of national life through the ef forts of revivalist organizations like the Gaelic League (founded 1893). In assuming a direct correspondence between nation and language identity, Corkery echoed the formulations of turn-of-the-century language revivalists themselves (e.g. Hyde 1986 [1892]: 153) who saw it as self-evident that a nation – in a typical nineteenth-century Romantic formulation – was the highest manifestation of a people’s culture and that language was the unique marker of that nation’s existence. Since Corkery considered the history of the nation and the history of the language to be synonymous, his account rarely provided any linguistic details. Rather, Corkery saw it as suf ficient to trace the various political movements to resist or end British rule in Ireland between the seventeenth and twentieth centuries. Whether these historical actors ever explicitly voiced support for the Irish language – and he freely admitted that only with the Gaelic League was a body truly committed to the revival of Irish manifested – was immaterial to Corkery, since any antiBritish action would, in his formulation, represent an implied advocacy of the modern Irish nation and therefore of the Irish language. This particular formulation of language and nation ref lected wider concepts in the general public understandings of Ireland’s history that had solidified by the time of publication of Corkery’s book. Developments in
History and Linguistics
51
Irish history seen as oppositional to the re-emergence of an Irish nation on a culturally distinct basis were elevated to centre stage as explanations for the decline of Irish, regardless of whether this made sense from a sociolinguistic standpoint. Daniel O’Connell (1775–1847), for instance, whose political movements for Catholic emancipation and repeal of the Act of Union in the 1820s and 1840s had not acknowledged a need to restore a separate Irish culture, was therefore seen by scholars as committing a particularly dire sin of omission that contributed to language near-death. The Catholic clergy were similarly singled out as primary causes for the decline of Irish insofar as they had not actively sustained its use despite their leadership roles in local parishes. Finally, the national school system established in 1831 – which, tellingly, also had long been a target of the language revival movement because it did not, until the 1870s, include Irish in the curriculum in any way – was identified by historians of Corkery’s generation as the most significant cause for the shift to English. A subtle but powerful reconsideration of these conclusions took place in subsequent years as more archivally intensive, scientifically oriented scholarship emerged in the 1960s. Among this scholarship, the work of Maureen Wall, Brian Ó Cuív, and Seán de Fréine deserves most attention. Wall’s writing took the most revisionist stance of the three. The notion that O’Connell, the national schools, and priests had together prompted language shift had, Wall wrote (1969: 81), been an ‘oversimplification’ perpetuated among both academic and general histories ‘in the manner of a catechism answer’. Although Wall provided a slate of evidence challenging earlier assumptions about shift, her most persuasive involved chronology: estimates showed that Irish speaking had already declined considerably before 1800, long before O’Connell’s leadership of a popular movement to secure Catholic emancipation in the 1820s, the introduction of the state schools in 1831, or the re-emergence of a powerful Irish priesthood in those same years. Such findings were echoed in work by Ó Cuív (1969, 1980), who introduced a measure of careful scholarship and mastery of archival sources in his approach evident, for example, in his use of census material and close reading of early documents in Irish. It was de Fréine’s first English-language version of his earlier writings on the subject, The Great Silence (1965), however, that articulated the most
52
Nicholas M. Wolf
in-depth explanation for the decline of Irish that continues to undergird current historiography. Like Corkery, de Fréine of fered little in the way of sociolinguistic data or analysis. His work instead took as a starting point the investigation of a narrow question: why did Ireland experience a persistent f lood of emigration in the nineteenth and twentieth centuries? The answer, according to de Fréine, could not be found in the usual theories that had been of fered over the years to explain Irish migration, among them economic factors, racially based causes, or religious justifications. De Fréine even cast a critical eye on the role of imperial oppression, arguing that British rule could hardly explain ongoing emigration from postindependence Ireland nor account for discrepancies between Ireland and other nations subject to foreign rule which had not experienced similar mass exodus. Rather, de Fréine proposed that a rupture in Ireland’s culture itself had occurred in the nineteenth century. Culture, in his formulation, is an organic entity made up of communal bonds and shared understandings among all members of a community. Since the ability to pass on these traditions is enabled by language, it was the loss of the Irish language in the nineteenth century that created the abrupt break in Irish cultural continuity and thus the long-lasting climate of national psychological harm and emigration (1965: 79–104). De Fréine’s emphasis on nineteenth-century cultural disjuncture has had a long afterlife in research into other aspects of Irish history in this period. The concept of cultural loss, for instance, appears in Emmet Larkin’s seminal work positing a Catholic ‘devotional revolution’ in the 1850s and 1860s that enshrined continental European religious practices in place of older indigenous devotions. Larkin (1972: 649) argued that a major factor encouraging the adoption of these religious forms had been the decline of Irish-speaking culture in the postfamine period that had made Irish men and women into cultural ‘emigrants’ and therefore susceptible to new Catholic orthodoxies. More recent studies of the decline of folk practices such as healing by fairy doctors and other forms of popular cultural similarly maintain many of the assumptions charted by de Fréine. Angela Bourke’s The Burning of Bridget Cleary (1999) portrays the linguistic frontiers between Irish- and English-speakers in nineteenth-century Ireland as directly mapping the break between a religiously unorthodox, orally
History and Linguistics
53
based, pre-modern culture and an Anglicized modern world that emerged sometime in the late 1800s. Notably, this model has also underpinned the latest research into language shift itself, as in Ní Mhóráin’s (1997: 9–20) characterization of the decline of Irish as a ‘comhlint chultúrtha idir dhá theanga’ (cultural conf lict between two languages) in which one culture was substituted for another. Despite its longevity, this framework deserves closer scrutiny and revision – especially in light of counterevidence emerging in recent research. Some weaknesses were already evident in The Great Silence itself. If the loss of the Irish language caused the loss of culture, what, according to de Fréine, had sparked language shift? Here, his analysis began to show a circularity as he fell back on political and economic explanations to account for the shift that he had rejected earlier in his study. The Great Silence argues that the failure of the Jacobite cause in 1745 – the last attempt to restore the Stuarts to the English crown – had cemented the powers of the Englishspeaking Protestant ascendancy in Ireland. At the same time, the growing strength of the Irish Catholic middle classes in the late eighteenth and early nineteenth centuries prompted a desire to learn English, the language of commerce. In short, Irish declined because English became the language of economic advancement and political dominance, rather than the language of invasion and oppression as it once had been. Yet this embrace of English-speaking society looks very much like the ‘cultural’ rift that de Fréine had earlier proposed as a result (rather than a cause) of language shift, and indeed he admits that it was ‘the social conditions of the second half of the eighteenth century’ – among which one would presumably include a new, Anglicized middle-class sensibility – that ‘fostered the change of language which ensued’ (1965: 141–145). Recent archival investigations have raised further questions. It had been assumed, for example, that the Irish language had become associated with the domestic sphere by the early nineteenth century, allowing the English language to exclusively dominate the new modes of politics emerging by way of widespread newspaper circulation, electioneering, the large outdoor rallies known as ‘monster meetings,’ and other features of the public sphere. This exclusion of the Irish language from politics was supposedly a symptom of its larger incompatibility with modernity (e.g.
54
Nicholas M. Wolf
Ó Loingsigh 1974: 13). But ongoing examinations of the surviving Irishlanguage manuscripts of the nineteenth century by Cornelius Buttimer, for example, have shown a sophisticated awareness among Irish-language commentators of world political events, including the War of American Independence and the Napoleonic Wars (1990a, 1990b, 1992, 1994). The notion that Irish-speakers conceived of a fundamental divide between the Irish-speaker and modern politics looks untenable in light of these findings. Similar discoveries regarding the use (rather than the prohibition) of Irish to expand and entrench new religious sensibilities and orthodoxies have also cast doubts on earlier conclusions by Larkin (Ó Ciosáin 2005: 150–151, and Wolf 2008: 86–138). Most seriously, there are fundamental uncertainties about the way culture itself is conceptualized in this historiography. Often, it is portrayed as an external entity independent of human agency – evident in the concept that culture can be ‘possessed’ or ‘lost’ rather than made through ongoing human interaction and contestation. Furthermore as Niall Ó Ciosáin (2005: 139–142) has pointed out, these studies fail to articulate the nature of the connection between culture and language despite their preeminent reliance on a direct link in order to make their case. De Fréine, for instance, had viewed it as self-evident that language has a fundamental role in perpetuating culture, but never clearly defined the practices, beliefs, or relationships that he believed entailed ‘culture’ or how the Irish language sustained them. At best, he merely insisted (1965: 105–108) that the culture of Irish-speakers not be defined narrowly – that is, by the attributes of the rural communities in which most nineteenth- and twentieth-century speakers could be found. As for how one language could sustain a particular ‘Irish culture’, but not another, the reader of The Great Silence is left without direction. Clearly, a better formulation of the terms of debate in this historiographical investigation is needed.
History and Linguistics
55
2 Approaches to language and culture The problematic treatment of the relationship between culture and language by scholars is hardly confined to de Fréine or to the field of Irish history. The history of this subject continues to be underexplored by historians, although there are recent signs that the field is growing (Burke 1987: 1 and 2004: 3). Even when discussing the issue, historians have often treated language as a transparent phenomenon due to the discipline’s lack of a firm command of developments in the field of linguistics, especially sociolinguistic approaches that aim at studying the phenomenon of language in the wider context of human society. Consequently, when language appears in historical accounts it is usually as an afterthought – that is, as an element encapsulated by the larger concept of ‘culture’ but without considering the fundamental dynamics of language itself. Historians rarely, for example, acknowledge the ways in which language encompasses variety (as in dialect, register, or other aspects of a speech repertoire), complexity in application and operation (e.g. conversational structure, bilingualism, code-switching), or a combination of any number of these elements. Nor is there usually a consideration that speakers themselves were aware of language dif ference, or that language itself might shape the ways speakers conceive of their social relationships. This oversight may be understandable, however, given the unsettled – even conf licting – treatment of the language-culture question in the discipline where an answer might naturally be sought to this question: linguistics. On the one hand, there is a firmly established approach to language that sees its features as indexing larger categories in society such as religion, class, gender, and ethnicity. Such studies seek to understand how dif ferences or changes in language might be explained by the status and features of the communities created by these categories. Thus, for example, a recent publication on overgeneralizations of certain gender forms in Dutch among non-native communities considers the possibility of correlating this tendency with immigrant identity (Cornips 2008: 114–120). Lanza and Svendsen (2007: 277–278, 284–296), while acknowledging the multiplicity of identities and the problems of overly mechanical approaches to ethnicity,
56
Nicholas M. Wolf
nevertheless see language and identity as separate variables whose connection they seek to elucidate for multilingual Filipino language networks in Oslo. Conversely, language may be treated as an independent entity that determines cultural identity. Jones (1998: 129–142) investigates the ways in which the growth of a school-learned, standardized dialect of Breton has in recent decades prompted a wider Breton identity that competes with older allegiances to the local commune and to a broader assimilation to French national identity among the inhabitants of Brittany. Such findings would appear to provide a suf ficient basis for a historical investigation into language by tracing the transformation of these external social categories. However, this picture is complicated by another a welldeveloped body of research that emphasizes the endogenous nature in which language behaves. One need not confine oneself, say, to the debate over Neogrammarian explanations for language change to find an ongoing conversation over the relative weight of internal and external factors in driving variation. Even among those that see social factors at play in linguistic change, there is disagreement about whether those developments require speakers simply to interact – as in concepts of koineization, for instance, in which cultural context only needs to provide details about how and when a speech act takes place – or whether speakers themselves enact change directly based on the ideologies of categories of gender, class, or nation. As result, work such as that recently published by Trudgill (2008: 241–254) can make a persuasive case for dialect accommodation of colonial Englishes that explicitly excludes cultural identity as an explanatory factor, while long-standing research into the fundamentally behavioural nature of language change (e.g. Labov 1994: 13–34) would likewise seem to undermine an oversimplified view of cultural identity as an automatic explanation for any particular linguistic manifestation. My point in citing these studies is not to comment on the merit of their conclusions, which involve nuanced consideration of many other linguistic questions beyond the issue of the relationship between culture and language. Nor is it possible to trace here the entire trajectory of linguistic debate over culture and language which, of course, has long been one of the central features of the discipline throughout a professional history encompassing Hymes, Ferguson, Whorf, and other scholars of the mid- and
History and Linguistics
57
late twentieth century. Rather, my objective here is to suggest that the historian searches in vain among linguistic case studies for a ready model to use in addressing the language-culture problem. Fortunately, developments common to both the disciplines of history and linguistics in the past three decades have better equipped scholars to address this question as well as find common ground. A solution to a historiographical uncertainty such as that of the Irish language in nineteenth-century Ireland may in fact require a hybrid examination of the language-culture problem by scholars. Sketching out these evolving approaches to culture helps make clear the possibilities (and even necessity) of combining historical and linguistic techniques in order to untangle the past.
3 Culture as a process By the early 1980s, a number of studies emerging out of cultural anthropology began to place increasing emphasis on the role of human agency in defining and ordering the relationships that made up society (Ortner 1984: 144–156). In part this was a reversal of longstanding structuralist views in the social sciences of human behaviour as a by-product of broader economic, political, or social realities – an approach that had begun to appear overly deterministic in its treatment of the individual or community. Instead, ‘culture’, or ‘the cultural’, was re-defined as the sum of human actions and discourses (i.e. ‘practices’) which together produced (and gave meaning to) collective understandings of categories (or ‘identities’) such as class, gender, or nationality. This rejection of the notion of innate identity brought along with it a number of linked concepts. It prompted the consideration of community bonds as a function of a ‘process’ or ‘production’ that relied on exclusion (i.e. defining the ‘Other’) as much as inclusion. Furthermore, the treatment of identity as contingent, mutable, and multiple meant a greater emphasis on diachronic study in order to explain change. This contrasted with older anthropological studies that had tended to of fer largely synchronic observations of communities.
58
Nicholas M. Wolf
This shift in anthropology, given this renewed concern with identifying and understanding change, has solidified its impact on the discipline of history in the past twenty years. The result has been a body of scholarship sharing a number of key assumptions regarding the need to historicize cultural practices and to understand how those practices are mediated by inequalities in power and inf luence (Sewell 1999: 52–55). Identities, these historians urge, cannot be treated as transparent and unchanging. Not only are they contingent on time and place, but cultural categories themselves must be treated as historically determined. One of the most active fields emphasizing this need to historicize cultural identity has been the history of nationality and nationalism. Here, scholars such as Hobsbawm and Ranger (1983) and Anderson (2006), among others, have insisted that communities not only constantly negotiate the features of their own national identity, but that the concept of ‘nation’ itself as an organizing principle for a collective group is a product of the past 200 years. The same idea can be applied to class or gender, similar categories that must be understood to have a dependent relationship with the society in which they exist. The integration of this approach with the discipline has been liberating in terms of describing how history works. Armed with a greater awareness of how cultural ‘production’ aims to define identity, historians can trace changing social categories through all levels of society, placing particular emphasis on the capacities and limits of various individuals to control those same categories. This also means that social settings hereto assumed to be nonpolitical – the private sphere, for example, in which certain roles such as gender are defined and enacted – have received significant attention due to an awareness of how they shape history. The need to explain how these identities are negotiated and constructed demanded a closer look at the role of language in driving this process, a development that reinforced previous concerns among structuralists and nonstructuralists alike. Described as a ‘linguistic turn’ in interpreting culture, this historiographical revolution significantly expanded the tools brought to bear by historians on dif ficult archival sources even as it raised awareness of the ways in which language determined the act of writing history itself. The history of events such as the Reformation and the French Revolution, according to scholars like Davis (1975) and
History and Linguistics
59
Ozouf (1988 [1976]), could be reinterpreted by paying closer attention to how human behaviour itself functioned as a linguistic text, whether in the form of the rebellious rituals of adolescent male journeymen printers in the case of Davis, or the quasi-religious French Revolutionary celebrations studied by Ozouf. Others, most notably Hayden White (e.g. 1987: 2–3), drew attention to the ways in which language, through conventions of grammar and syntax, defines dif ferences between narrative and nonnarrative forms of historical reporting – a distinction considered essential to modern western concepts of the past. Ultimately, by focusing on the means by which culture was negotiated (in this case, language as discourse), these approaches made essential contributions to the emerging historiography of the 1980s and later. Yet even as historians have used the concept of language to inform their new, anthropology-based approach to culture, their methods continue to define language narrowly. Language, in many of these studies, remains a metaphorical and a readable ‘text’ in a Saussurian system in which discourse (e.g. expressions printed in newspapers, or phrases used in political speech) acts as a signifier that produces meaning. By contrast, the speech act as a behavioural or ethnographic performance in the sense understood by sociolinguists remains largely unacknowledged by historians, who tend to borrow from literary disciplines far more than linguistics. There are exceptions. A number of historians addressing the history of nationalism and the emergence of the modern state have specifically addressed the role of language standardization and the existence of minority dialects in these developments. Early work on France, for instance, by Eugen Weber (1976) has been supplemented in recent years by Ford (1993) and Bell (1995) who have concentrated on the many linkages between French Republicanism’s concept of universalism, the emergence of French national identity, and the enshrinement of an of ficial dialect of French. The case of the multilingual Habsburg empire has received a similarly detailed overview by Robert Evans (2004), while scholarship on the social history of Welsh (e.g. Jenkins 2000) has been unusually active. These studies have been particularly helpful in directing the attention of scholars to the role of the state in shaping multilingual societies, thereby reminding historians of both the power and limits of governments to direct their population’s linguistic makeup.
60
Nicholas M. Wolf
Fortunately, it can be argued that the present identity-based approach to cultural history invites historians to incorporate a more sociolinguistically informed look at language. Certainly, the role of language as one of many tools in establishing the cultural identity of speakers both consciously and unconsciously – hence the importance of language to the cultural process – should be self-evident. As a matter of fact linguists have long been exploring this area, especially as a result of the same developments in anthropology that had made such a strong impact on historians. There have been, for instance, calls from within the sociolinguistic field for a more critical use of social, economic, or political contexts as an explanation for language features. As Cameron (1990: 81–86) argues, doubt must be cast on the treatment of social structures as independent and determinative inf luences on language; the search by linguists for correlative links between measured variables such as ethnicity, class, or gender and linguistic features assumes, rather than proves, a certain relationship between culture and language. This has prompted linguists to devote more time to understanding how the relationship between language and identity really operates. For Gafaraga (2005: 282–283), extending Cameron’s rejection of language as simply ref lecting essentialized social identities, language itself ‘structures society’ – a clear call to take up the notion of cultural identity as a contingent set of categories dependent in part on language itself. These identities, as Irvine and Gal (2000: 74–76) note, have been increasingly considered by linguists since at least the 1970s to encompass not shared, homogenized bonds based on language, but rather a ‘managed’ linguistic diversity. Cultural identities, in this formulation, are not given entities but are constantly calibrated based on assertions of dif ference with an excluded outside group. Indeed, Irvine and Gal go even further in pointing out that a community’s conceptualization of the features and significance of its own language (as distinct from the language as measured by the linguistic scholar) is a product of the same cultural process of definition and exclusion responsible for any other category of identity. This emphasis on language as defining identity – rather than cultural identity defining language – adds weight to the explanatory powers of psychology- and behaviour-based explanations for events like language change
History and Linguistics
61
or language shift. Consequences of language contact such as creolization, koineization and dialect levelling can be accounted for quite comfortably within a framework that addresses directly the dynamics of human interaction, the context that Mufwene (2008: 15–26) refers to as the ‘ecology’ of speakers. This is not to suggest that the social realities and circumstances of speakers do not help mediate how (and if ) those interactions take place. However, recognizing identity as an unstable and constantly mediated category makes cultural explanations unsuitable for such large-scale phenomena like language variation. Indeed, if it is accepted that identity only acquires meaning through the definitions and understandings of individuals themselves, it is not possible to conceive of culture as a pre-existing condition capable of determining language. Such a conclusion is discernible in the discomfort expressed by Trudgill (2008: 252–253) concerning ‘identity’ as an explanation for the form of colonial Englishes, and more fully developed in a response by Tuten (2008: 259–260) to Trudgill’s conclusion. Speaking of koineization in medieval Spain, Tuten postulates that formation of a ‘frontier identity’ may be a simultaneous feature of language change, a much more plausible paradigm for culture than the theories for colonial dialects that Trudgill ef fectively refutes. The virtues of this approach include the ways it opens up possibilities for scholars to focus on the impact of language variation on identity by liberating culture from serving as a determinant of language. The study of how speakers perceive their own speech acts, the meanings they attribute to them, and how that meaning is deployed to shape worldviews would thus occupy centre stage in this approach. Irvine and Gal’s concepts of erasure and iconization certainly of fer compelling tools for unpacking this process: the role of language change (or language maintenance) in defining identity would be mediated by the meaning that a community itself attributes to linguistic features. One wonders how the history of emerging working-class solidarity in the early industrial era, for instance, might be rewritten based on an exploration of the communal bonds of shared linguistic features like dialect created by diverse workers increasingly interacting in a large-scale factory setting.
62
Nicholas M. Wolf
4 Proposals for ongoing research To return to the Irish case, two ingredients of any future research agenda in this field are suggested by the above attempt to sketch out a disciplinary overlap between linguistics and history. First, there is a need to take a more behaviourally based approach to understanding language shift in Ireland that integrates many of the observational approaches and questions asked by sociolinguists, especially in the areas of language acquisition, language change (certainly a phenomenon such as accommodation), and language maintenance. The justification for this lies partly in the facts already known about Ireland’s language shift. It is accepted that the adoption of English and even the abandonment of Irish had already begun in parts of the country as early as the seventeenth century. This process had commenced even earlier in the region around Dublin subject to English rule, the Pale. Such an extended time frame suggests a gradual and long-term transformation in language use, making attempts to explain language shift in terms of specific events such as the introduction of national schools, the Great Famine, the Act of Union (1801), or any other such single occurrence insuf ficient. Indeed, this chronology had long ago been recognized by Wall (1969), although its ramifications for explaining the decline of the Irish language have yet to become fully entrenched in the scholarly literature. Similarly, the geolinguistic account of Ireland’s language shift that has already been sketched out by FitzGerald (1984 and 2003) suggests not a replacement of Irish by English in an uneven patchwork fashion, but in a clear and receding line running roughly north-south and moving westward over at least two centuries. This, too, argues against any island-wide cultural shift in prompting language change, but rather a gradual and persistent behaviourbased transformation that requires specific sociolinguistic investigation into intergeneration transfer and network ties. An advantage of this approach is that it helps support a second necessary investigative feature that arises out of the initial historiographic question of this paper regarding the connection between Irish cultural identity and Irish speaking in the nineteenth century. By removing the need to force language shift to conform to a timeline of cultural change, the historian
History and Linguistics
63
is free to consider a more complex treatment of Irish culture itself. This approach should incorporate the understanding of cultural identity as a product of language behaviour, with both treated as constructed enterprises based on a number of ingredients: the speech act itself, the speaker’s and attitudes towards that language (including erasure, iconization, etc.), and the inf luence of key institutions and settings (such as religious bodies, the family, the marketplace, or the public sphere) that help define speakers’ views. The result will likely be the discovery of multiple identities constructed by language use, from a broader Irishness defined as the Irishspeaking nation to a local Irishness based on dialect. It is likely, moreover, that plural geographical organizations of cultural identity will not be the only finding when this more open approach is taken. The anti-modern/ modern scheme that scholars often map directly onto the Irish/English linguistic divide should also be considered as only one of several possible views by nineteenth-century speakers of their language. Since such attitudes are not intrinsic, but rather historically contingent, the existence of alternate views of the Irish language as compatible with modernity – however hidden or exceptional – will need to be acknowledged and explored. Obviously, simply adopting this more expansive view of how language ‘worked’ hardly solves the problem of retrieving such detailed sociolinguistic answers from the past. As historical linguists repeatedly acknowledge, the archival record is not always suited to answering such questions. It is far easier for historians to simply argue that a broad economic, political, or social transformation somehow prompted language change than to uncover the behavioural details at the local or individual level that were the actual cause for shift. But this should not discourage scholars from attempting to do so, and in any case the contention of this chapter remains that a shift in approach is needed as much as any return to the archives. Since Corkery first published The Fortunes of the Irish Language in the 1950s, the methods applied by scholars to this issue have not evolved beyond many of the findings of subsequent revisions in the 1960s. This fact in itself should prompt a return to some of the fundamental assumptions behind these studies. Fortunately, the centrality of the Irish language to Irish historiography suggests that this issue will remain salient long enough for scholars to revisit this particular aspect of the history of Ireland.
64
Nicholas M. Wolf
References Anderson, Benedict. 2006. Imagined Communities: Ref lections on the Origin and Spread of Nationalism. London and New York: Verso. Bell, David. 1995. ‘Lingua Populi, Lingua Dei: Language, Religion, and the Origins of the French Revolutionary Nationalism.’ American Historical Review 100, 1403–1437. Bourke, Angela. 1999. The Burning of Bridget Cleary. New York: Penguin. Burke, Peter. 1987. ‘Introduction.’ In: Peter Burke and Roy Porter (eds), The Social History of Language. Cambridge: Cambridge University Press, 1–20. Burke, Peter. 2004. Languages and Communities in Early Modern Europe. Cambridge: Cambridge University Press. Buttimer, Cornelius. 1990a. ‘A Cork Gaelic Text on a Napoleonic Campaign.’ Journal of the Cork Historical and Archaeological Society 95, 107–119. Buttimer, Cornelius. 1990b. ‘An Irish Text on the “War of Jenkins’ Ear”.’ Celtica 21, 75–98. Buttimer, Cornelius. 1992. ‘A Gaelic Reaction to Robert Emmet’s Rebellion.’ Journal of the Cork Historical and Archaeological Society 97, 36–53. Buttimer, Cornelius. 1994. ‘Cogadh Sagsana Nuadh Sonn: Reporting the American Revolution.’ Studia Hibernica 28, 63–101. Cameron, Deborah. 1990. ‘Demythologizing Sociolinguistics: Why Language Does Not Ref lect Society.’ In: John E. Joseph and Talbot J. Taylor (eds), Ideologies of Language. London and New York: Routledge, 79–93. Corkery, Daniel. 1956. The Fortunes of the Irish Language. Repr. edn. Cork: Mercier Press. Cornips, Leonie. 2008. ‘Loosing Grammatical Gender in Dutch: The Result of Bilingual Acquisition and/or an Act of Identity?’ International Journal of Bilingualism 12, 105–124. Davis, Natalie Zemon. 1975. Society and Culture in Early Modern France. Stanford: Stanford University Press. de Fréine, Seán. 1965. The Great Silence. Dublin: Foilseacháin Náisiúnta Teoranta. Evans, R.J.W. 2004. ‘Language and State-Building: The Case of the Habsburg Monarchy.’ Austrian History Yearbook 35, 1–24. FitzGerald, Garret. 1984. ‘Estimates for Baronies of Minimum Level of Irish-Speaking Amongst Successive Decennial Cohorts: 1771–1781 to 1861–1871.’ In: Proceedings of the Royal Irish Academy, Section C, 84, 117–155.
History and Linguistics
65
FitzGerald, Garret. 2003. ‘Irish-Speaking in the Pre-Famine Period: A Study Based on the 1911 Census Data for People Born Before 1851 and Still Alive in 1911.’ In: Proceedings of the Royal Irish Academy, Section C, 103. 191–283. Ford, Caroline. 1993. Creating the Nation in Provincial France. Religion and Political Identity in Brittany. Princeton, NJ: Princeton University Press. Gafaraga, Joseph. 2005. ‘Demythologising Language Alternation Studies: Conversation Structure vs. Social Structure in Bilingual Interaction.’ Journal of Pragmatics 37, 281–300. Hobsbawm, Eric, & Terence Ranger (eds). 1983. The Invention of Tradition. Cambridge: Cambridge University Press. Hyde, Douglas. 1986 [1892]. ‘On the Necessity for De-Anglicising Ireland.’ In: Breandán Ó Conaire (ed.), Language, Lore and Lyrics. Essays and Lectures. Dublin: Irish Academic Press, 153–170. Irvine, Judith T., & Susan Gal. 2000. ‘Language Ideology and Linguistic Dif ferentiation.’ In: Paul V. Kroskity (ed.), Regimes of Language: Ideologies, Polities, and Identities. Santa Fe (NM): School of American Research Press and Oxford: James Currey Ltd, 35–83. Jenkins, Geraint H. (ed.). 2000. The Welsh Language and its Social Domains, 1801–1911. Cardif f : University of Wales Press. Jones, Mari. 1998. ‘Death of a Language, Birth of an Identity: Brittany and the Bretons’. Language Problems and Language Planning 22, 129–142. Labov, William. 1994. Principles of Linguistic Change. Volume 2: Social Factors. Malden, MA: Blackwell Publishers. Lanza, Elizabeth, & Bente Ailin Svendsen. 2007. ‘Tell Me Who Your Friends Are and I Might Be Able to Tell You What Language(s) You Speak: Social Network Analysis, Multilingualism, and Identity’. International Journal of Bilingualism 11, 275–300. Larkin, Emmet. 1972. ‘The Devotional Revolution in Ireland, 1850–1875.’ American Historical Review 77, 625–652. Mufwene, Salikoko S. 2008. Language Evolution. Contact, Competition, and Change. London: Continuum International Publishing Group. Ní Mhóráin, Brighid. 1997. Thiar sa Mhainistir atá an Ghaolainn Bhreá: Meath na Gaeilge in Uíbh Ráthach. An Daingean: An Sagart. Ó Ciosáin, Niall. 2005. ‘Gaelic Culture and Language Shift.’ In: Laurence M. Geary and Margaret Kelleher (eds), Nineteenth-Century Ireland: A Guide to Recent Research. Dublin: University College Dublin Press, 136–152. Ó Cuív, Brian (ed.). 1969. A View of the Irish Language. Dublin: Stationery Of fice. Ó Cuív, Brian. 1980. Irish Dialects and Irish-Speaking Districts: Three Lectures by Brian Ó Cuív. Dublin: Dublin Institute for Advanced Studies.
66
Nicholas M. Wolf
Ó Loingsigh, Padraig. 1974. ‘The Irish Language in the Nineteenth Century.’ In: Oideas 14, 5–21. Ortner, Sherry. 1984. ‘Theory in Anthropology Since the Sixties.’ Comparative Studies in Society and History 26, 126–166. Ozouf, Mona. 1988 [1976]. Festivals and the French Revolution. Trans. Alan Sheridan. Cambridge: Harvard University Press. Sewell, William H. 1999. ‘The Concept(s) of Culture.’ In: Victoria E. Bonnell and Lynn Hunt (eds), Beyond the Cultural Turn. Berkeley: University of California Press. 35–61. Trudgill, Peter. 2008. ‘Colonial Dialect Contact in the History of European Languages: On the Irrelevance of Identity to New-Dialect Formation.’ Language in Society 37, 241–254. Tuten, Donald. 2008. ‘Identity Formation and Accommodation: Sequential and Simultaneous Relations.’ Language in Society 37, 259–262. Wall, Maureen. 1969. ‘The Decline of the Irish Language.’ In: Brian Ó Cuív (ed.), A View of the Irish Language. Dublin: Stationery Of fice, 81–90. Weber, Eugen. 1976. Peasants into Frenchmen. Stanford: Stanford University Press. White, Hayden. 1987. The Content of the Form: Narrative Discourse and Historical Representation. Baltimore: Johns Hopkins Press. Wolf, Nicholas. 2008. Language Change and the Evolution of Religion, Community, and Culture in Ireland, 1800–1900. PhD dissertation. University of WisconsinMadison.
Brian D. Joseph
Historical Linguistics and Sociolinguistics: Strange Bedfellows or Natural Friends?
Abstract Inasmuch as historical linguists and sociolinguists are both interested in dif ferent types of variation – diachronic in the former case and synchronic in the latter case – one might think that they necessarily have much in common. While it is ultimately argued here that such is indeed the case, by way of exploring the relationship between historical linguistics and sociolinguistics, several notions that are stock in trade for most sociolinguists are re-examined from a historical linguist’s perspective. It is suggested, for instance, that ‘change in progress’ is a mirage, and that chain shifts are not as dramatic an event as one might think. At the same time, though, the enterprises are linked by the Uniformitarian Principle, and that construct is extended into general historical investigations through the positing of a ‘social or humanistic uniformitarianism’ that emphasizes what is similar across time in human circumstances and in human reactions to those circumstances.
1 Introduction It should be clear that historical linguistics and sociolinguistics have something to do with one another and presumably something to contribute to one another. Both areas of investigation are interested in variation, diachronic variation in the former case, synchronic variation in the latter case. Moreover, since language is always embedded in some social context, language change necessarily takes place in a given social setting. Such
68
Brian D. Joseph
connections are part of the rationale for a characterization of the study of the social dimensions of language change as ‘sociohistorical linguistics’ and they suggest that both social historians and sociohistorical linguists might well be in a position not only to learn from one another but also to learn from the exploration of the interrelationships of language and history and of linguistics and historiography. At the same time, though, a comparison just of sociolinguistics with historical linguistics would seem to be a useful way of getting at the potential for ‘value-added’ that a consideration of social factors of fers to historical linguistics. Accordingly, in what follows, I aim to see what points of similarity and dif ference there are between these two commonly paired concerns – note the very terms ‘socio-historical linguistics’ / ‘historical sociolinguistics’, after all – with an ultimate goal of determining if the coupling of the two is, as the title suggests, the result of joint membership in a natural class or is instead a forced marriage. In the course of so doing, I re-examine and to some extent debunk, or at least attempt to debunk, a number of concepts that both historical linguistics and sociolinguistics hold dear. In many instances, I pose questions about notions and practices without necessarily of fering answers. In the spirit of Socrates’ adage about the unexamined life,1 my hope is that asking the right questions is helpful even if clear answers are not of fered. Some of what follows may seem obvious and maybe even trivial to the intended audience of sociolinguists, historical linguists, and social historians, but my intent is in part to call attention here to some shortcuts that practicing socio-historical linguists routinely use. In this way, we can be sure that we are aware of what we are doing when we employ them. I see two important reasons for doing this. First, it is sometimes the case that practitioners can be deluded or deceived in the worst case or just even distracted by their own terminology and their own practices, so that raising questions can be a way of heightening awareness. Second, there is always a risk that others outside our subfield might adopt (and then alter or misconstrue) our practices without fully understanding why we do what we do, and being explicit about the practices can thus be a safeguard against that. 1
From Plato’s Apology (38a): ‘The unexamined life is not worth living’.
Historical Linguistics and Sociolinguistics
69
To illustrate this last point, we can note that historical linguists often compress sound changes when talking about them. Thus, the development in Latin in which an original s occurring between vowels ultimately turned into an r, known to linguists as ‘rhotacism’, is generally discussed as being simply s > r /V__V, even though there is good reason to believe that it went through a stage of s > z (and then z > r). For one thing, Proto-IndoEuropean *z independently developed into Latin r, as in mergo ‘dive’ from *mezg- (where Sanskrit majj- ‘to dive’ of fers confirmation for assuming *mezg- as underlying Latin merg-). And, in Oscan, a language very closely related to Latin, z is found corresponding to Latin intervocalic r from earlier *s. Thus we talk about s > r in Latin, while recognizing this to be shorthand for s > z > r (maybe even s > z > ž > ř > r). Nonetheless, in an early generative historical linguistic account of this change (e.g. Kiparsky 1971), it was treated simply as the addition of a rule of s => r /V__V to the grammar, thus taking traditional historical linguists’ practice as defining the issue conclusively. In this way, such a treatment failed to observe Andersen’s (1989) crucial distinction between diachronic correspondence, the static comparison of two possibly widely separated stages of a language, and innovation, the first appearance of a new structure, sound, feature, etc.
2 Some similarities and dif ferences As noted at the outset, a key point of rapprochement between historical linguistics and sociolinguistics is that both disciplines are interested in variation, in the one case in variation across time periods and in the other in variation within a given time period. And relatedly, both are interested in change, but there is a dif ference here: for historical linguists, change is the focus, whereas for sociolinguists, change is just one side of the overall phenomenon of variation. Even so, there is a further, far more crucial dif ference though, in the way each discipline approaches change, and that is in regard to the relevant timescale. Sociolinguistics typically examines what can be construed as
70
Brian D. Joseph
‘change in progress’ (on which see below), typically based on contemporary usage, while historical linguistics typically examines developments that are over and done with, often, but not always, long past over and done with. Yet, even in this dif ference there is a point of contact, of course, in that the tenets of (contemporary) sociolinguistic investigation and interpretation can be applied to past situations, as done by Teodorsson (1976), for instance, with regard to the role of genderlectal variation in Koine Greek vowel changes. This is in keeping with the Uniformitarian Principle that the processes at work in the past are no dif ferent in kind from those evident today.2 Continuing in the point and counterpoint mode, I note that the available data is also a key point of dif ference. It is especially important to keep in mind here Labov’s (1994: 11) telling characterization of historical linguistics as ‘the art of making the most of bad data’. While I would prefer, following the reasoning given in Janda and Joseph (2003: 14), to say ‘imperfect data’ rather than ‘bad data’, because the data available in a historical account typically is simply not as complete as data collected in a contemporary investigation can be, it must be admitted that the sentiment expressed in Labov’s statement is right on target: we do the best we can in a historical investigation within the limitations imposed by the data.
3 Plus ça change … – the more things change the less things change The points just made lead into what might be the single most important question bearing on possible dif ferences between sociolinguistics and historical linguistics: with regard to changes (putatively) observable now, are there dif ferences in nature/type of change or just in amount or degree? 2
On the history of the Uniformitarian Principle, first promulgated for geology, and its subsequent application to other historical sciences, including linguistics and for relevant references, see Janda and Joseph (2003).
Historical Linguistics and Sociolinguistics
71
That is to say, do present circumstances make for fundamentally dif ferent types of change now, and is an answer here tied to issues of the availability of data? Relatedly, if the answer is that any dif ferences are just a matter of degree, we can ask how we might quantify rate of change in a meaningful way so as to make it possible to judge and compare the extent of change. It is important to note here that the ‘long (temporal) view’ of historical linguistics teaches that the answer to these questions about the nature of change in dif ferent eras is that the types of change one observes now, that is the mechanisms and processes of change, are essentially the same now as they were in the past. This, again, is what the Uniformitarian Principle asserts. Still, for many, especially lay, observers, that is not the case; thus, there are some who say that the internet, for example, has created a new type of communication and a new type of human interaction; Paredes & Mário (2007), for instance, claim that ‘the growth of the Internet and its associated technologies did open space for a new type of human interaction: virtual, social interaction environments,’ and Yonkers (2008) confidently states that ‘I see this [i.e. Facebook] as an emerging new method of communication in which the rules are still evolving.’ One has to assume that with these new methods of communication and interaction, these observers would expect to eventually see dif ferent types of change emerging. My answer here is to stand by Uniformitarianism and expect that if any dif ferences are to be observed, they will be dif ferences in degree rather than kind. Presumably, time will tell here, but we can glean some insights into this matter from a few small case studies where developments today and those in earlier years are parallel but dif ferent, and the dif ferences are quantitative and not qualitative. For instance, there is a plethora of acronyms in American English today. Interestingly, a good many of these coinages derive from modern institutions, such as governmental agencies, e.g. IRS (Internal Revenue Service), EPA (Environmental Protection Agency), FAA (Federal Aviation Authority), and OSHA (Occupational Safety and Health Administration); from contemporary professional associations, including sports leagues, and related sponsored activities like conferences, e.g. LSA (Linguistic Society of America), TESOL (Teachers of English to Speakers of Other Languages), CLS ((annual meeting of the) Chicago Linguistic Society),
72
Brian D. Joseph
NCAA (National Collegiate Athletic Association), and BCS ((college football) Bowl Championship Series), to name a few; and from modern technology, e.g. CPU (central processing unit), LOL (laughing out loud (as a text-messaging abbreviation), among others. The obvious recency of these coinages makes it appealing to think of this acronymic coinage as a modern phenomenon. But is it really something new? As it happens, there are documentable instances of acronymic coinage from the past, including ancient times. The most likely source, for instance, of (Latin) elementum is acronymic in nature (see Ernout-Meillet 1939: s.v.), with the word deriving from a fixed sequence ‘L M N’, much like English ‘(learning one’s) ABCs’, a claim that gains some plausibility when one considers that these letters are the first three of the second half of the Canaanite alphabet from which the Latin alphabet ultimately was derived, and thus a likely starting point for some ancient scribal lessons. The ubiquitous English – and now worldwide – word OK is another case in point, if it derives, as Read (1963a, b) has argued,3 from ‘Oll Korrect’, a ‘comical misspelling’ of an initialism ref lecting a jocular pronunciation of ‘all correct’ in east coast American slang of the 1840s that revolved around the use of acronyms and humorous word play. These two historical examples indicate that acronymic coinages have been possible for a long time. It may well be that modernity has provided increased opportunities for this seemingly mundane word-formation process to be realized and to f lourish, but it is hardly a new type of change. Moreover, given that we actually know very little about one-of f kinds of spontaneous colloquial coinages in ancient times or even in earlier stages of modern English, it is hard to feel confident about saying that elementum and OK are unique sorts of examples for their respective eras. The burden of proof rather should rest with anyone saying that these examples do not indicate a possible word-formation strategy for those periods, one with a possibly broad extent of use. 3
I am not completely convinced that Read is right about OK, but it is the etymology that seems to have gained the greatest acceptance. We must always remember that etymology, in the words of Eric Hamp (1998: 14n.2) is a ‘brittle science’, in the sense that etymologies are generally hypotheses about the earlier form of a word and not established fact. This applies of course to the case of elementum, too.
Historical Linguistics and Sociolinguistics
73
A similar sort of argument can be made with regard to blending as a type of compounding. That is, we find numerous examples in modern times of blended forms, not just the famous deliberate creations like Lewis Carroll’s chortle, coined for comic ef fect, or weathermen’s snizzle (for ‘snowy drizzle’), coined to fill a lexical gap, but rather the considerable number of such forms to be found in names for new products or ideas and in the ‘breathless’ sort of celebrity-based journalism that is intended to attract the attention of readers interested in the latest about some well known personality. That is, the demands of these venues, all related to marketing ideas or products or newspapers or television shows, seem to call for something ‘fresh’ and ‘new’, and blends appear to fit the bill. Thus in the US today we find product names such as Snausage and Pup-peroni for doggie treats, the verb decorganize for the commercially based new concept of ‘organizing your home or business with appropriate décor’,4 and labels in the US celebrity media for celebrity couples such as Brangelina (= Brad Pitt + Angelina Jolie) and TomKat (= Tom Cruise + Katie Holmes).5 In this regard one can even cite deliberately humorous coinages such as the following from the Washington Post’s 2009 Mensa Invitational, which asked readers to take any word from the dictionary, alter it by adding, subtracting, or changing one letter, and supply a new definition).
4
5
In a definition that recognizes the commercial side to the word and its use, this verb is defined by The Urban Dictionary accessed 4 May 2010) as: ‘To organize one’s belongings into trendy and obnoxious looking containers. These containers are typically purchased at specialty retail stores and departments such as The Container Store and Target. These containers usually involve the following concepts: large f loral patterns, bright colors, transparent/translucent forms and unique methods of latching and sealing’. There are at least two other relatively recently emerging ‘social’ domains where blending is rampant now: drink names, e.g. appletini, crantini, mangotini, etc., for mixed drinks served in a martini glass and consisting of vodka (or the like) and apple/cranberry/mango (etc.) juice, giving a pattern so prevalent that -tini could be treated as a morpheme (I thank Marivic Lesho for bringing this example to my attention) and names for hybrid dogs, e.g. labradoodle for a mix of Labrador retriever and poodle, cockapoo for a mix of cocker spaniel and poodle, etc.
74
Brian D. Joseph
Here is a sampling of the forms that emerged from the competition, each one ultimately a blend, even if created under a particular set of constraints:6 ignoranus ‘a person who’s both stupid and a jerk’ bozone ‘the substance surrounding stupid people that stops bright ideas from penetrating’ Beelzebug ‘Satan in the form of a mosquito that gets into your bedroom at two or three in the morning and cannot be cast out’.
The social context for these creations – commerce, competition, and such – is certainly interesting, but so too is the apparent folk-linguistic view underlying them that words should be compositional in meaning. To return to the issue at hand, however, again we can ask if such blending is really all that new, or if instead we are just witnessing an increase in the documentation of such forms and an increase in their use; thus again, one has to wonder if this is a qualitative change or just a quantitative one. Here the evidence of an apparent ancient composite blend may point towards the same-in-kind-but-dif ferent-in-extent assessment. In particular, the source of English bring, and more generally Germanic *bringan-, is disputed, as the verb in that form seems to be isolated within Indo-European; one etymology (in the American Heritage Dictionary s.v.) takes it to be a blended form of two roots for ‘carry’, thus *bher-enek- (=> *bhr-enk-). If this account is right, then it would constitute evidence of a very old blending process and thus suggest that the novelty and abundance we see in today’s blends indicates a shift in degree associated with the demands of modernity rather than a new kind of change per se. Admittedly, what we do not, and probably cannot, know is whether this *bhr-enk-, if properly etymologized here, had the same sort of playful or edgy feel to it for speakers of Proto-Indo-European that some of the abovementioned recent blends in American English have; that is, we really have virtually no way
6
Further evidence of the prevalence of these formations comes from an article entitled ‘Lingo-making’ by Mark Leibovich and Grant Barrett in The New York Times’s Week in Review section of Sunday December 20, 2009. In reviewing some of the new words and phrases that constituted ‘Buzzwords 2009’, they noted some presumably deliberately coined blends such as aporkalypse for ‘undue worry in response to swine f lu’ and vook for ‘a digital book that includes some video in its text’.
Historical Linguistics and Sociolinguistics
75
of getting at Proto-Indo-European playful word-formation strategies that would correspond among the PIE speech community to the ‘hot news’ sort of blends we see today. That is, a form like *bhr-enk- may of fer a glimpse at the formal mechanisms at the disposal of Proto-Indo-European speakers, but unfortunately says nothing about its af fective value. It is important to note that if the type of change is not really dif ferent in these cases, then in a sense we are looking at a certain sort of stability. This fact must always be kept in mind when claims about change are made. I say ‘a certain sort of stability’ since there are several works recently that make claims about the relative stability of certain types of structures, where ‘stable’ means ‘less prone to change’,7 but skepticism might be in order here. That is, talking about change/stability in language can be likened to talking about change/stability in the weather in Columbus (and elsewhere): ‘if you don’t like the weather, just wait a few minutes!’ That is, if you are willing to wait long enough, everything in language is subject to change – except for the most foundational aspects of Universal grammar, the elements, like the ‘design features’ (Hockett, 1960), without which we cannot say we have a ‘language’. That is, a statement about some part of structure being ‘stable’ must be taken as relative at best, and can only make sense in social terms, i.e. in terms of what a speech community ‘allows’ to persist in their language. As a related point, there is the very interesting set of observations that Harrington et al. (2000/2006) have made on change in the vowel realizations in Queen Elizabeth’s Christmas speeches throughout her reign. In particular, over her fifty years or so on the throne, her vowels have shifted in these speeches to something more in line with younger speakers of the standard southern British dialect. While these dif ferences would seem to be a clear case of change over the course of an individual’s lifespan, they must be assessed against the fact that we do not really know how the Queen spoke colloquially, in unguarded moments, at other stages in her life. That is, one could argue that what Harrington et al. have measured is changes in the Queen’s linguistic performance in a very circumscribed and 7
Nichols (2003), for instance, while recognizing (p. 284) that ‘nothing in language […] is truly immutable’, argues nonetheless for certain elements showing greater stability, for instance personal pronouns.
76
Brian D. Joseph
controlled set of circumstances, or perhaps even changes in her perception of how a Queen could or should (verbally) behave in a certain context. We are not in a position, absent such data on the Queen’s colloquial usage in earlier years, to tell if there has really been any change in her linguistic competence; she may simply be acting (or perhaps even reverting to) more colloquial realizations due to a recognition that expectations of the Queen’s role may have changed over the years. The same kind of objection can be raised to the fine and revealing study of ‘changes’ with the use of as far as to mark topics, as documented by Rickford et al. (1995). They demonstrate that in the materials they examined over a roughly 200-year span into contemporary English, topic-marking As far as NOUN + ‘VERBAL.CODA’ (e.g. As far as Bristol is concerned, I just love the city) shows a significant decrease in use compared to As far as NOUN (without the verbal coda, e.g. As far as Bristol, I just love the city); there are early examples (two occur in Jane Austen’s Emma, 1816) but the increase in recent years is dramatic. Equally dramatic too, however, is the increase in the availability of colloquial materials in recent decades; that is, it is not clear how much of Jane Austen’s writing, for instance, is suitably colloquial nor do we have a clear idea of what her own colloquial usage was – the couple of examples that are in her works may be the tip of the early nineteenth-century iceberg, the mass of which is hidden from view. The common thread in these last few examples is that because of an absence of crucial data, claims of change are perhaps overstated, and there may actually be more stability in such cases than change.8 Moreover, change in external circumstances may give certain processes a chance to operate but – very much in Uniformitarian style – the processes themselves are familiar ones at all times. 8
A similar view is expressed by Dennis Baron in the 14 November 2009 posting on his ‘Web of Language’ blog () entitled ‘Dirty words you can say on television: WTF as the newest cable channel?’. In commenting on whether swearing is more prevalent now than in the past, he says ‘It’s hard to track whether people are swearing more than they used to, since there’s little swearing in the historical record, and it’s just about impossible to know how much people swore in conversation fifty years ago, or in the eighteenth century, or back in Middle Ages, or when humans first emerged from the primordial mists’.
Historical Linguistics and Sociolinguistics
77
4 Yet another key point – a basic foundational issue: What is change? A concept that has figured prominently in the discussion so far is ‘change’, and indeed much in the conclusions drawn here depends on how one defines ‘change’. In particular, is change to be located in the initial structural innovation or rather in the spread of the structural innovation? Linguists take dif ferent stances on this; moreover, the position taken has consequences for various interrelated issues. I survey here some of those issues that stem from a decision one way or the other on this matter, though, as promised, some are merely posed as questions without there necessarily being answers attached to them. First, if change is defined in terms of the structural innovation itself, then there is less of a rapprochement between historical linguistics and sociolinguistics. This is because historical linguistics examines completed changes (innovations) both as to their actuation (point of origin of the innovation), and their complete or incomplete spread (since that is how dialect/language dif ferentiation occurs), whereas sociolinguistics focuses just on the spread of an innovation. Second, we might add contact linguistics as a subfield interested in change, and as closer to some extent to historical linguistics in terms of focus.9 In particular, while contact linguistics examines change brought on by a particular social setting (contact between peoples), and thus might seem to be more aligned with sociolinguistics, studies on language contact have tended to be focused more on actuation (contact as a cause of innovations) than on spread. Third, if change is defined rather in terms of the spread of innovations, then something akin to observation of change is possible (contrary to the traditional view, as in Bloomfield (1933), whereby change was held to be inherently unobservable); this is what sociolinguists mean by talking about examining ‘change in progress’. Admittedly, in another sense, this notion could relate to the timescale issue
9
I thank my colleague (and contact linguist par excellence!) Don Winford for reminding me of this important point.
78
Brian D. Joseph
mentioned above, with historical linguists looking at completed events at a great time distance and sociolinguists looking at unfolding events in a more finely grained time scale. This leads into a fourth issue or question: The sociolinguistic literature notwithstanding, is there really observable change in progress? Haugen (1958: 777) had an interesting point to make in this regard concerning borrowing (cited in Thomason 1997: 181): ‘Since the actual moment of borrowing is rarely observable, most conclusions about interference are based on inferences’. If change is defined as the innovation, then Haugen’s point could be made too with regard to change; however, even if change is defined in terms of spread, Haugen’s skepticism is justified since it is not the spread itself that is observed but rather only evidence of the aftermath of spread (in the form of certain distributions in the population at large). This means that talking about vowel fronting in terms, say, like ‘Speaker X shows a high degree of fronting of [u]’ may technically be inappropriate, for it seems to imply that the movement itself is being witnessed. Rather, what is in fact being viewed is a static result of (someone having engaged in) the process of fronting; and, that act of fronting need not even have been undertaken by the speaker who is observed, in that such a speaker may well have simply inherited a word (i.e. learned it from a source such as a parent) with a particular vowel in it that happens to be fronter than the realization inherited (learned) by some other speaker. In some ways, this is rather like the situation commented on in Joseph (2011) regarding use of the term ‘grammaticalized’. There I suggest that instead of saying that a speaker shows a grammaticalized feature (or structure, or construction, etc.), we might rather say, more neutrally, that the speaker shows a more grammatical feature (structure, etc.); my reason for this terminological suggestion is that talking in terms of something being ‘grammaticalized’, rather than simply stating the static fact of the speaker showing grammatical status for the feature in question, implies that the history of the feature in question and the directionality of its development are known when oftentimes the history is just a matter of speculation. Fifth, as a matter closely related to the consideration of ‘change-inprogress’, there is yet another way one might interpret this notion that merits some discussion. This other interpretation has to do with the widespread belief that much of what is seen in language change is gradual. That is, a
Historical Linguistics and Sociolinguistics
79
progression of shifts is often given as a basis for talking about change as gradual, so that the Latin rhotacism change alluded to above in section 1 could be called gradual if it in fact proceeded as s > z > r (or even s > z > ž > ř > r). This characterization is a reasonable one, and common to be sure, but it must be recognized that a language could in principle stop after any of these steps. That is, progression to the next step is not a necessary outcome, and indeed, if there was any appreciable time between the shifts, perhaps as little as a generation or even less, then clearly the stage reached by a given shift was a viable language state that speakers could learn and use unaltered. This means that in a certain sense, each shift can be treated as a change in and of itself, each one an innovation. Admittedly, from an etymological standpoint, gradual means ‘step-by-step’, but under the interpretation suggested here, instead of using examples like Latin rhotacism to show ‘gradual change’, it might be better to call it ‘(the result of ) cumulative change’, so as to emphasize the temporal sequencing of separate and distinct innovations. In this view too, it might be said that every change, every shift, every innovation, is abrupt, in that each one represents a break from, a deviation from, an existing norm, even if it is a ‘micro-step’, just a small innovation. Sixth, relevant here too is the claim that ‘chain shifts’ represent a special mechanism of change (so Labov 2007); we can understand ‘chain shift’ in a broad sense, referring to any linked sound changes, including cases where sounds A and B move jointly in the direction of sounds C and D respectively, or in a narrow sense, including cases where sound A moving in the direction of sound B is said to trigger a change in B, moving it in the direction of a dif ferent sound C. With a view of ‘change’ that focuses on innovation rather than spread, this claim needs to be re-examined as to mechanisms and outcome. In particular, if sound A moves in the direction of sound B, then one of two things can happen: either there is merger (of A with B) or there is no merger. And, as for the no-merger situation, while it can be the case that there is no merger because A does not move all the way into the realm of B, another way in which there could be no merger would be B moving somewhere; in such a case, why must that be viewed as having been caused by A’s movement? In such a case, how do we rule out the possibility of these simply being two independent changes, especially since we know that mergers do happen.
80
Brian D. Joseph
To elaborate on this last point somewhat with a real example, we can consider the sound shifts between Proto-Indo-European (PIE) and the Germanic (Gmc.) languages known collectively as ‘Grimm’s Law’ whereby PIE *dh ultimately gives Gmc. d, and PIE *d ultimately gives Gmc. t. It is possible to conceptualize these changes as chain shifts in that when PIE *dh moved in the direction of PIE *d on the way to Gmc., PIE *d moved in the direction of *t. This certainly looks like a chain shift when presented this way, with *dh > d triggering the *d > t change. However some other languages, for instance the Balto-Slavic group, underwent a merger of *dh and *d, so that when *dh moved in the direction of *d for Balto-Slavic, the *d stayed put, and a merger occurred. Thus a causal connection between the two Gmc. changes may be hard to maintain. Seventh, there is the related issue, related since chain shifts would seem to involve whole systems of sounds, or at least subsystems within the overall phonological space, of whether there can be ‘system-based’ motivation for change, e.g. change motivated by such considerations as symmetry and balance in the phonological inventory. Here, I would suggest that it is fair to at least wonder whether the system is a reality, and a concern – subconsciously – for speakers as such, or rather is actually just a nicety that is important for linguists only. In other words, are speakers as taken by ‘symmetry’ as linguists are? Eighth, apparent chain shifts seem to involve one change causing another change, but more generally, it can be asked what it means to talk in terms of one ‘event’ causing another? In some instances, temporal adjacency is all that can be demonstrated or perhaps one event is just one contributing factor to an outcome. To take an analogy from sports, if Bristol City’s Liam Fontaine makes a pass to Nicky Maynard that leads to a goal, he is credited with an assist, and rightly so, but it may be as much Maynard’s ef fort that caused the goal as Fontaine’s pass. Moreover, in a certain sense, everything that precedes X lays the foundation for X, so one could argue that Fontaine’s being born on 7 January 1986 was a causal contributor to the pass he made leading to a Maynard’s goal. And, to continue with a critique of chain shifts, one kind of chain shift – the broad-sense type in which A & B both undergo the same shift, e.g. to C & D – can be reinterpreted to be really just a matter of how we
Historical Linguistics and Sociolinguistics
81
formulate the relevant events. For instance, mid-vowel raising in northern dialects of Modern Greek af fects both e and o, but their shifting in tandem may be a function of generalization over mid-vowels rather than e going first and then ‘pulling along’ with it the o (in what is sometimes called a ‘drag chain’). In that case, rather than seeing the change(s) as two separate events that are linked by a special type of shifting mechanism, one could simply say that the relevant mechanism is ‘rule generalization’. And, once one takes that step, then it would seem that the real mechanism involved is analogical in nature, specifically a ‘phonetic analogy’ (in the sense of Vennemann, 1972) that can be schematized as: e is-to i as o is-to X, with X then ‘solved for’ as u (that is, e : i :: o : X, X => u). And, if that is the case, then the change involved may not be sound change in the strict sense, if it is analogically induced. Alternatively, it may have to be admitted that ana logy plays a far greater role in sound change than we might think. In this regard, the interpretation of fered by Anttila (1977) for the Grimm’s Law shifts is noteworthy, as he suggests that they may have had an analogical basis (p : f :: t : θ :: k : x, etc.); see also Hock (2003), where there is further argumentation about sound change as analogy, and Durian (2009, 2012), who pursues a similar line of reasoning concerning some of the vowel shifts evident in contemporary English of Central Ohio. Ninth, to shift to a dif ferent thorny issue, we can ask how one might go about quantifying change – assuming, that is, that we even can quantify change – in the light of these dif ferent definitions. In particular, if each ‘micro-innovation’ counts as a change, and if each micro-innovation involves a relatively small shift of some sort, then even fairly straightforward changes that can be characterized as the alteration of a single feature, such as k => x (where only manner of articulation changes from stop to fricative) could represent a large number of minute adjustments, each of which could constitute a separate change, e.g. k first becoming aspirated, then various adjustments to the strength and timing of the aspiration occurring leading to an af fricate-like realization (e.g. as kx or even kx), which is then simplified to x. Tenth, if change is to be located in micro-innovations, then it becomes hard, in taking the long temporal view that many historical linguists do, to legitimately talk about the ‘same’ change over long periods of time. For
82
Brian D. Joseph
instance, repeatedly throughout the history of Greek, though sometimes at a time distance of several hundred years, one finds the loss of word-final -n occurring. However, is it the ‘same’ change, activated at dif ferent points in time or a separate and distinct change in each era in which it is found? Without a clear reason for linking the events, such as the recurrence in dif ferent time periods of a highly unusual or particularistic constraint on the change, it would seem prudent to treat such cases as involving temporally and ontologically distinct instances of final n-loss, even at the expense of seeming to miss a generalization (a ‘recurring tendency’ in a given language) over a long time-span; for one thing, once one is talking about time spans of 120 or more years, it must be borne in mind that the speakers, i.e. the members of the af fected speech community, are an entirely dif ferent set of individuals. For such reasons – the absence of any ‘smoking gun’ types of special conditions repeated across time, and the complete turnover of the af fected population with each ‘wave’ – I am inclined to distinguish (at least) three waves of n > Ø /__#, i.e. loss of n in word-final position, in the history of Greek, instead of talking about a persistent final-n-loss change that stretched over a thousand or more years. Eleventh, once the issue of attempting to quantify rate of change is considered, a related question arises. In particular, except perhaps in cases of intense language contact of the sort that can lead to language mixing, one has to wonder if it is really possible to talk (as some sociolinguists and historical linguists do) about ‘massive’ or ‘far-reaching’ changes. It would seem that every language at every stage of its development is an amalgam of elements (sounds, morphemes, words, constructions, etc.) that are carried over from previous stages and elements that have been altered in some innovative way. Thus ‘massiveness’ of change often simply means that the analyst has identified a set of elements that show change, without a concern for the concomitantly conservative elements that show no change. This view is especially compelling when one considers that linguists at any one time typically focus on at best a handful of elements that do change when in fact a language contains thousands and thousands of ‘points’ (sounds, morphemes, words, constructions, meanings, etc.) that could in principle change. Thus, ‘massiveness’ may only be in the eye of the linguist attending to a small subset of changing points in the language overall.
Historical Linguistics and Sociolinguistics
83
Taking this one step further, if this amalgam view is right, then it would seem dif ficult to talk about languages and dialects being ‘conservative’ or ‘innovative’ (in general). It would seem that such an assessment makes sense only as relative to particular features. A useful example to always remember in this regard is the situation with Lithuanian. Lithuanian is often said to be highly ‘conservative’ within Indo-European, but on closer inspection, that is only true with regard to there being a large number of cases within the nominal system (eight, as reconstructed for Proto-Indo-European, and seven in modern Lithuanian, and then only if one ignores the gain of three cases (illative, allative, and adessive) in Old Lithuanian10 and their subsequent loss by the modern era; that is, conservatism is there but only for the noun system and only for (most of ) the eight original cases.
5 Conclusions Even if historical linguistics is focused more on change and less on the spread of change, every practicing historical linguist recognizes that understanding spread is one dimension to a full understanding of the forces of diachrony. Thus overall, historical linguistics and sociolinguistics form a natural class rather than an uneasy forced alliance. Still, it is possible to express some wishes for historical linguists and for sociolinguists in terms of practices and research questions. First, historical linguists could be more interested in stability, even if, as stated above, all we can really talk about is relative stability and not absolute or inherent stability. As noted above in section 3, there is recent literature on this topic, but still for most practicing historical linguists, change is where the action is, so to speak, and not stability. Relatedly, a key question to be addressed is whether change is the default state for language diachronically or whether instead (relative) 10
These cases were secondary formations that came about from the fusing of existing ‘primary’ case forms (genitive, accusative, locative) with postpositions.
84
Brian D. Joseph
stability is the default state. And sociolinguists could be less concerned about ‘change in progress’ since it is often hard to be sure that there really is a change going on; with any variation claimed to betoken change, the question of directionality is an issue (which form is innovative and which is conservative) and so too, as a result, is the question of whether the innovative variant will ultimately take hold or not, as it is not always the case that it does. Admittedly, investigations of change in progress will always be important and revealing and it is not entirely clear what should take the place of interest in the phenomenon; ‘identity sociolinguistics’ is not enough, as there are real issues in determining which states of af fairs constitute generational change and which do not that identity alone does not answer. As for social history and social historians, to return to the theme sounded at the outset, perhaps there is a dif ferent type of uniformitarianism to consider and always keep in mind, another dimension to the adage that ‘plus ça change, plus c’est la même chose’. In particular, the raw material that many historical linguists have to work from are written or orally transmitted texts and thus typically literature or historical accounts from days gone by. Importantly, they generally show humans from thousands of years ago to be much like humans of today in terms of the things that they are interested in and worry about. That is, there is a uniformitarianism with regard to the hopes and aspirations of individuals in the distant past and those we observe and feel today. Similarly, day-to-day concerns about one’s welfare, about being nourished and loved, and so on, seem to be as important to individuals in the past as they are to us today. We can refer to this uniformitarianism as a ‘social or humanistic uniformitarianism’, and, drawing on Joseph (1999), we can note a couple of examples that ref lect this state of af fairs, and show stability, and thus uniformity, in what individuals’ lives and interests and concern were. A first example comes from Wendy O’Flaherty, in her 1981 translation of the Rigveda (the oldest Sanskrit text, composed c.1200 BC but with parts much older). In her preface, she comments on the content of the text as follows, noting that the hymns of the Rigveda show ‘conf lict within the nuclear family and uneasiness about the mystery of birth from male and female parents; the preciousness of animals […]; the wish for knowledge,
Historical Linguistics and Sociolinguistics
85
inspiration, long life, and immortality’. In many ways, these seem to be universal human concerns, and such a characterization of content could be applied to what one sees today in the tabloid press (such as the National Enquirer in the US) or in titles on bestseller lists. A second example is Melchert (1991), writing on the last minutes of life for Hittite king Hattusili (from the second millennium BC). Here are the relevant facts as Melchert lays them out: Hattusili was apparently dictating his last will and testament to a scribe when he suf fered an ultimately fatal or near-fatal episode as he finished the of ficial dictation at the end. He then began ref lecting somewhat incoherently about his impending death, producing ravings which were dutifully copied down and recorded for posterity by the scribe.11 Hattusili ends with an exhortation to a woman he has been calling for: ‘Protect me on your bosom from the earth’, apparently his real last words. Melchert interprets these last words as follows: it is known that the Hittites practiced burial (not cremation) but believed in an afterlife and immortality in divine form for its kings; thus, he writes, ‘Despite […] assurances of happy immortality, however, the dying Hattusili is frightened. He sees only the immediate certainty that he will soon be put down into the cold, dark earth alone, and like many a poor mortal since he finds this a terrifying prospect’. Further, by way of linking modern-day folks with those that preceded them 3,500 years ago, Melchert says, with real eloquence: ‘there seems to be little fundamental dif ference between us and ancient peoples when it comes to facing death. Hattusili’s words speak to us directly across the centuries. His fear is palpable. We not only at once understand but also are moved by his agony and his desperate cry for his loved one’s tender comfort. These emotions are neither Hittite nor Indo-European, neither ancient nor modern, but simply human’. And, finally, to bring this social/humanistic uniformitarianism back to the issue of language change per se, let me mention these notable lines from Chaucer’s Troilus and Criseyde (II.22–28):
11
This is presumably the closest we will ever come to having a tape recording of spontaneous speech from the second millennium BC!
86
Brian D. Joseph Ye knowe ek that in forme of speche is chaunge Withinne a thousand yeer, and wordes tho That hadden pris, now wonder nyce and straunge Us thinketh hem, and yet thei spake hem so, And spedde as wel in love as men now do; Ek for to wynnen love in sondry ages, In sondry londes, sondry ben usages. [You know too that in form of speech there is change within a thousand years, and words as well that had value now wondrously odd and strange they seem to us, and yet they spoke them in this way, and it served as well in love as men do now; also to win love in various ages, in various lands, various were the usages.]
Thus, not only is change itself something that is pervasive, but so too is the recognition of change and an interest in change.
References Andersen, Henning. 1989. ‘Understanding linguistic innovations.’ In: Leiv Egil Breivik and Ernst Håkon Jahr (eds), Language change: Contributions to the study of its causes. Berlin: de Gruyter, 5–27. Anttila, Raimo. 1977. Analogy. The Hague: Mouton de Gruyter. Bloomfield, Leonard. 1933. Language. New York: Holt, Rinehart, Winston. Durian, David. 2009. ‘Purely a chain shift?: An exploration of the “Canadian Shift” in the US Midland’. Paper presented at NWAV 38, Ottawa, Canada. Durian, David. 2012. A new perspective on vowel variation throughout the 20th Century in Columbus, OH. PhD dissertation. Ohio State University. Ernout, A., & Antoine Meillet. 1939. Dictionnaire étymologique de la langue latine: histoire des mots. Paris: Klincksieck. Hamp, Eric P. 1998. ‘Some draft principles for classification’. In: Brian D. Joseph and Joseph C. Salmons (eds), Nostratic. Sifting the evidence. Amsterdam: John Benjamins, 13–15.
Historical Linguistics and Sociolinguistics
87
Harrington, J., S. Palethorpe & C. Watson. 2000a, 21 December. Does the Queen speak the Queen’s English? Nature 408, 927–928. Harrington, J., S. Palethorpe & C. Watson. 2000b. ‘Monophthongal vowel changes in received pronunciations: An acoustic analysis of the Queen’s Christmas broadcasts’. Journal of the International Phonetic Association 30, 63–78. Harrington, Jonathan. 2006. ‘An acoustic analysis of “happy-tensing” in the Queen’s Christmas broadcasts’. Journal of Phonetics 34.4, 439–457. Haugen, Einar. 1958. ‘Language contact’. Proceedings of the International Congress of Linguists 8, 771–785. Hock, Hans Henrich. 2003. ‘Analogical change’. In: Joseph & Janda (eds), 441–460. Hockett, Charles. 1960. ‘The Origin of Speech’. Scientific American 203, 89–97. Janda, Richard D., and Brian D. Joseph. 2003. ‘On language, change, and language change – or, of history, linguistics, and historical linguistics’. In: Joseph & Janda (eds), 2–180. Joseph, Brian D., & Richard D. Janda (eds). 2003. Handbook of historical linguistics. Oxford: Blackwell. Joseph, Brian D. 1999. ‘Linguistics for “Everystudent”’. Studies in the Language Sciences 28.2 (Fall 1998), 123–133. Joseph, Brian D. 2011. ‘A critique of grammaticalization’. In: Bernd Heine and Heiko Narrog (eds), Handbook of grammaticalization. Oxford: Oxford University Press. Kiparsky, Paul. 1971. ‘Historical Linguistics’. In: William O. Dingwall (ed.), A survey of linguistic science. College Park: University of Maryland Linguistics Program, 576–642. Labov, William. 1994. Principles of linguistic change: Internal factors. Oxford: Blackwell. Labov, William. 2007. Transmission and dif fusion. Language 83, 344–387. Melchert, H. Craig. 1991. ‘Death and the Hittite king’. In: Roger Pearson (ed.), Perspectives on Indo-European Language, Culture and Religion. Studies in Honor of Edgar C. Polomé, Vol. I.182–188 (Journal of Indo-European Studies, Monograph Number 7). McLean, VA: Institute for the Study of Man, 182–188. Nichols, Johanna. 2003. ‘Diversity and stability in language’. In: Joseph & Janda (eds), 283–310. O’Flaherty, Wendy. 1981. The Rig Veda. An Anthology. New York: Penguin Books. Paredes, Hugo, and Mário Martins, F. 2007. Social Theatres: A Web-Based Regulated Social Interaction Environment (Lecture Notes in Computer Science). Berlin: Springer. Read, Allen Walker. 1963a. The First Stage in the History of ‘O.K.’. American Speech 38.1, 5–27.
88
Brian D. Joseph
Read, Allen W. 1963b. The Second Stage in the History of ‘O.K.’. American Speech 38.2, 83–102. Rickford, John R., Thomas A. Wasow, Norma Mendoza-Denton & Juli Espinoza. 1995. Syntactic variation and change in progress: Loss of the verbal coda in topic-restricting ‘as far as’ constructions. Language 71.1, 102–131. Teodorsson, Sven-Tage. 1979. Phonological variation in Classical Attic and the development of Koine. Glotta 57, 61–75. Thomason, Sarah G. 1997. On mechanisms of interference. In: Stig Eliasson and Ernst Håkon Jahr (eds), Language and its ecology: Essays in memory of Einar Haugen. Berlin: de Gruyter, 181–207. Vennemann, Theo. 1972. Phonetic analogy and conceptual analogy. In: Theo Vennemann and Terence Wilbur (eds), Schuchardt, the Neogrammarians, and the Transformational Theory of Phonological Change: Four Essays. Frankfurt: Athenäum, 181–204. Yonkers, V. 2008. New type of communication with facebook? Connecting 2 the World 13 September 2008. accessed 12 August 2011.
Nicola McLelland
From Humanist History to Linguistic Theory: The Case of the Germanic Rootword1
Abstract This paper traces how the notion of the monosyllabic rootword – a concept that had its origins in sixteenth-century German humanist historiography – became central to German ref lection on the nature and structure of language. Indexical of German antiquity and originality in the sixteenth and seventeenth centuries, it came to play a part in eighteenth-century debates on language origin, underlay lexicography into the mid-nineteenth century, and also inf luenced comparative linguists examining Semitic and Sanskrit well into the nineteenth century. In J.G. Schottelius’s (1612–1676) hands, it yielded the basis of present-day word-formation theory, and dif fering beliefs about the integrity of the rootword led to dif ferences in spelling between German and Dutch that are still apparent today.
1 Historiography and German linguistics Recently historians’ methods have inf luenced sociolinguists and historical linguists, who have followed their historian colleagues to the archives (Elspaß 2005; Elspaß et al., 2007). But it is not just the methods of history
1
Since this paper was first submitted, Sections 3 and 4 have been integrated in slightly re-arranged and revised form in a longer article on Justus Georgius Schottelius: McLelland (2010).
90
Nicola McLelland
that can inf luence linguistic study. There is no shortage of examples of how historiographers’ readings or (re-)constructions of the past have shaped linguistic ref lection in Germany. One might note the anxious focus, for a while at least, on possible dif ferences between the language of the GDR and FRG, ref lecting anxieties about the recent history of the Germany states (Hellmann and Schröder, 2008), the Marxist view of history that encouraged a particular type of sociolinguistics in the GDR (Stevenson, 2003), or the grand ‘national monument’ of the Deutsches Wörterbuch begun by the Grimm brothers in 1852 (and completed in 1960 (Haß-Zumkehr, 2000)).2 The most notorious examples of how politically motivated readings of history could skew German linguistic study are those of the National Socialist era (Hutton, 1999), tainting figures like Leo Weisgerber (cf. Erlinger 2007 and references there), Jost Trier, and Heinz Kloss (1952) to a greater or lesser degree. Hüllen (2002: 302–332) gives further examples of how the National Socialist reading of the past shaped the work of lesser-known linguistic figures in ways that now seem distasteful. In this article, I consider a case rather more remote in time: sixteenth-century humanist historians’ re-readings of Tacitus and their idealization of a Germanic past. I shall show how the key notion of the rootword in German first emerged from such humanist readings of ancient Germanic history, became imbued with cultural symbolism, and was ultimately incorporated into mainstream linguistic ref lection.
2 Humanist history and heritage-building The slice of German historiography that I want to examine here is the German humanist project of creating a shared German heritage in which to take pride. Central to the project was the rediscovery and exploitation of Tacitus’s account of Germania by humanist historians of the sixteenth 2
The overlaps and mutual fertilization between language history and cultural history have been a focus of interest for some time in German studies, as the volume of papers published a decade ago suggests: Sprachgeschichte als Kulturgeschichte, Gardt et al. (ed.) (1999).
From Humanist History to Linguistic Theory
91
century (Mertens 2004). There has been much interest in German studies in how readings of recreations of the notion of Germania have infused German studies over the centuries (cf. Krebs, 2010; Lee & McLelland, 2011). Mertens (2004: 38) commented that historians and philologists need to work together in examining this process of reinventing the past. This paper is a contribution to that discussion. Tacitus’s Germania itself is a short text – about fifteen pages in translation – and as an ethnographic account of the Germanic tribes, it is highly ambiguous. Passages praising the valour and sexual restraint of the people sit cheek by jowl with accounts of the poor buildings and the lack of agriculture; the potentially touching image of all babes suckled by their own mothers is undercut by the account of the squalor of family life, where those same children grow up ‘naked and nasty’ (nudi ac sordidi, Tac. Germ. 20, transl. Thomas Gordon). Yet Tacitus’ Germania formed the basis for the humanist creation of a proud German past. As a measure of its inf luence, Mertens (2004: 61) calculates that there were twenty editions of it in the period 1472–1519, meaning that about 6,000 copies were in circulation over a time span of fifty years. The first inf luential reading of Tacitus’s Germania was that of Enea Silvio Piccolomini (1405–1464), whose pamphlet Germania (1457–1458) read Tacitus in such a way as to portray the ancient German tribes as nomadic and primitive, barely distinguishable from brutes, so establishing a discontinuity between that era and the more civilized present. In contrast, Giannantonio Campano’s (1429–1477) reading of the same text – in a speech intended to motivate the Regensburg Reichstag of 1471 to take up arms against the Turks (though the speech was never delivered, it was published in the 1480s after Campano’s death; Mertens 2004: 72–75) – emphasized both the valour of the Germans and their unity as a people. Even though the entire second half of Tacitus’ Germania deals with the separate tribes, Campano focused on one statement in Germania 1.2 that the Germans were an unmixed people native to the territories where they lived; hence – Campano reasoned – they must never yet have been conquered by invaders. Equally positive was Campano’s interpretation of the name Germani, which Conrad Celtis (1459–1508) would follow: the Latin word germani (‘brothers’) expressed the sense of brotherhood amongst the Germanic peoples themselves. From a collection of tribes that had come together, Germania was now viewed by Campano as a ‘natürliche
92
Nicola McLelland
Einheit’ (Mertens 2004: 76–77), a people with a common origin, who only later split up into dif ferent tribes. Conrad Celtis (1459–1508) added to the claims of Germanic valour and indigenousness their impressive geographic expansion. He also praised the uncorrupted simplicity of the Germanic peoples, by contrast with the customs of his own time. So where Enea had seen the Germanic past as something primitive but happily overcome, Celtis found in the same Tacitus text a Germanic past that was primitive in the positive sense of unspoilt (Mertens 2004: 82).3 Heinrich Bebel (1472–1518) followed suit, emphasizing the original purity of an unmixed, valorous, undefeated people, and these topoi were soon very firmly established. As evidence of how strong and long-lasting was the tendency to attribute such virtues to the Germanic ancestors, consider Justus-Georgius Schottelius’s (1612–1676) annotation of a German proverb in the list he compiled as part of his Ausführliche Arbeit ‘Comprehensive Work’ on the German language (1663: 1129). Schottelius was a cultural patriot who clearly shared the agenda of his humanist predecessors to portray the Germanic past positively, and like them, he saw proverbs as an expression of German culture and character. So in the proverb Penning ist Pfennings Bruder (already found as no. 71 in the 1529 proverb collection of Johannes Agricola of Eisleben (1494–1566); cf. Mieder 1982), Schottelius found evidence of the sense of equality and fairness amongst the ancient Germans: Hieraus erweiset sichs / wie unsere alten Teutschen Collation / Wolleben und Freude gehalten haben / nemlich / daß ein jeder sein Essen hat mitbracht / und zum Getränk haben sie einen Pfennig neben den andern gelegt zu gleicher Zechen / und ist einer nicht höher beschweret worden dan der ander. (Schottelius 1663: 1129)4 3
4
One particular problem for humanist commentators seeking a more positive view of ancient Germania was Tacitus’ reference to the Germans’ human sacrifices to their deities. Conrad Celtis, in his Germania edition of 1500, suggested an emendation to the text: one should read huius for humanis, hence ‘his [of the God Mercury] sacrifices’, rather than ‘human sacrifices’. Enea’s edition gives the of fending passage in the margin, apparently because of a printer’s error, but many editions following it omitted the problematic passage altogether (Mertens 2004: 74). For fuller discussion of Schottelius’s treatment of proverbs, see McLelland (2011a, chapter 5).
From Humanist History to Linguistic Theory
93
[This shows how our ancient Germans enjoyed eating together, a good life, and joy, namely that each brought his food, and for drinking they each laid a penny next to another in a row, and not one was burdened [financially] more than another.]
Seventeenth-century art history provides evidence of continued ‘bi-polar’ readings of Tacitus in the Netherlands too. Dutch historiographers initially condemned the barbarous Germanic tribes, but, motivated in part by the Dutch Republic’s need for a founding myth after the Twelve Years’ Truce of 1609,5 they later homed in on certain positive characteristics (Weststeijn 2008: 68, n.86, citing Van der Waal 1952: 172–210).6 Conveniently, Tacitus had singled out the Batavians as the bravest tribe of the Rhineland, and he had also praised their noble simplicity (Weststeijn 2008; 50, 58; Cf. Germania 29). Now, in Germany the 26 engravings illustrating the German Philipp Cluverius’s 3-volume Germania antiqua (1616) were the first systematic illustrations of ancient Germania. Cluverius cited and paraphrased Tacitus on nearly every page, and the engravings illustrated the same points: a warriorlike simplicity (cf. Krebs 2010), including for instance the details already mentioned of a mother nursing her baby (fig. 12; cf. Tac. Germ. 20.1: sua nemque mater uberibus alit …), children playing naked on the ground (fig. p. 158; cf. Tac. Germ. 20.1: In omni domo nudi ac sordidi in hos artus, in haec corpora quae miramur, excrescunt), and animal skins as clothing. In contrast, illustrations by Antonio Tempesta for Batavorum cum romanis bellum (by Otto Vaenius, Antwerp 1612; cf. Westteijn 2008: 50) portrayed the Germanic ancestors as civilized, dressed in classical-style clothing. The Anglo-Dutch antiquarian Richard Verstegan’s (c. 1550–1640) portrayal of the ancient peoples of Germania had also already been more fanciful. Verstegan (1605) praised the simplicity, masculinity and virtue of the ancient tribes, but also contrived to portray their pagan worship as virtuous, writing that they ‘lived according to the lawe of nature and reason, wanting nothing 5 6
The truce with the Hapsburg empire achieved nominal independence for the Dutch republic in 1609. Dutch historians apparently rehearsed a commonplace of Roman historiography for describing and appropriating foreign civilizations: the simplicity of barbaric tribes was deemed a positive value which had been lost by the ancients themselves (Van der Waal 1952: 58).
94
Nicola McLelland
but the knowledge of the true God’ (Weststeijn 2008: 51, citing Verstegan 1605: 67). One of Verstegan’s illustrations showed an idol of their god Tuisco (Figure 1, from Weststeijn 2008: 53). While Verstegan’s illustration potrays the statue of Tuisco simply clothed in some sort of cloak (cf. Tacitus, Germania 17) and sporting long hair and a beard (Tacitus, Germania 31), the overall impression is one of classical elegance. Significantly, a band of people is seen approaching in the background from the Tower of Babel. Here, the imagery expresses a tour de force already achieved by Annius de Viterbo (Giovanni Nanni, c. 1432–1502) in his Antiquitatum Variarum of 1498 (often known as Antiquities, a forgery of a ‘lost’ history by Berosus, so often called the Pseudo-Berosus), which contrived to combine the Tacitean mention of a Germanic god called Tuisco with a universal world history that chimed with Biblical accounts of the confusion of tongues. According to this account, Tuisco, an additional son of Noah and the father of the Germanic race, left Babel with his followers to settle the lands now called Germany (cf. Jones 1999, chapter 1). One could pursue the shifting readings of Germania in the sixteenth and seventeenth centuries in far greater detail, but for my purposes the essential points have been made. Ancient Germania was – after some earlier, more negative readings – portrayed positively as pure, uncorrupted, strong, and original, with a shared and single point of origin at Babel.
3 The rootword as guarantor of Germanic purity, antiquity and originality It is in the context of positive ‘heritage-building’ (Considine 2007) for Germany and her Germanic neighbours that a key linguistic notion about the German language emerged, the rootword. Tacitus’s Germania had said almost nothing about the language of the Germanic peoples (though he cited two words, for ‘spear’ and ‘amber’), but Francisus Irenicus
From Humanist History to Linguistic Theory
95
(Franz Friedlieb, 1494 or 1495–1553), in his Germaniae exegesis, described by Mertens (2004: 78) as ‘a kind of handbook about “Germany” that combines ethnography, history, geography and natural history’ (transl. NMcL), wrote of the German language that omnis imperatiuus, ac omnia pene germanica uocabula monosyllaba sunt ut brot, susz, disch, der, dem, lisz, morn etc. et pene quicquid monosyllabam excedit, pere grinum ac non germanicum est … (Irenicus 1518: lib II cap 31 fo 38v, cited Jellinek, 1898 (1985 rpt.): 60) [every imperative and almost every German word are monosyllabic, such as brot, susz, disch, der, dem, lisz, morn, etc. and almost everything that is longer than a monosyllable, is foreign and not German. (Translation NMcL)]
In this short statement, three crucial features that characterize later discourse about rootwords are already present: 1 2 3
They are monosyllabic; They are identical with the (second person singular) imperative form of the verb; They are characteristically German.
This created heritage of the Germanic rootword had a long shadow in later linguistic ref lection, as I shall show below by taking each of the three aspects in turn. 3.1 Monosyllabic rootwords Jellinek (1898: 61) assumed, plausibly, that Irenicus’ belief in monosyllabic rootwords was the result of extrapolating from the southern German varieties of German, characterized by apocope, i.e. sag, not sage. What is important is that this feature of the rootword was not, at this early stage, indexical of any other qualities of German, or indeed of the Germans. Beatus Rhenanus (1485–1547), at the head of the last section of Book II of his 1531 Rerum Germanicarium Libri Tres, a section headed Germanornum lingua ‘the language of the Germans’, repeated Irenicus’ observation: that
96
Nicola McLelland
the German language gaudet vocibus primigeniis monosyllabis ac nonnulla a Prouincia lingua mutuatus est ‘rejoices in originally monosyllabic words and took some from the languages of the provinces’ (cited from the edition by Mundt 2008: 268–69, English translation NMcL).7 The Austrian Johannes Cuspinianus (1473–1529) made the same point, and Georg Henisch (1549–1618) cited both these authorities in the preface to his incomplete German dictionary (Henisch, 1616 [1973], Rössing-Hager, 1984–1985). If it had its origins in a view of the language that was coloured by the local experience of a southern German variety, the notion of the rootword evidently suited north Germanic scholars too, whose languages also exhibited apocoped forms. It soon became a commonplace in linguistic thought in the Low Countries, and in Sweden too, where Georg Stiernhielm and the grammarian Tiällmann (1696) both cited Becanus’s Hermathena (Bk 2, p. 25 of Omnia Antverpiae 1580) (Haapamäki 2002: 55). For it was with the Flemish cultural patriot Goropius Becanus (1518–1572) that the rootword became more than a neutral feature of the German language, and became indexical of the same Germanic qualities that I identified as typical of cultural patriotic discourse at the end of Section 2: pure, uncorrupted, strong, and original, with a shared and single point of origin at Babel. Becanus used the notion of the rootword as evidence of the originality, and therefore antiquity and purity, of the ancestral Germanic language. He believed that the most ancient language on earth would be the simplest language, and that the simplest language would contain the most monosyllabic rootwords. Since the number of short words was (he asserted) higher
7
Beatus Rhenanus marks an important step in the reception of Tacitean notions of Germania. He published both the Germania itself, and a commentary (1519), as well as a later work, Rerum Germanicarium Libri Tres, to which I shall return. He rejected Annius de Viterbo’s Pseudo-Berosus as fiction, and at the same time insisted that Tacitus’s Germania should be read as a document of its time. He ‘thus made a substantial step forward in the development of a historical approach to the sources themselves, a step from the philosophical method of textual reconstruction to historical criticism and to the methodical explication of developments’ (Mertens 2004: 93).
From Humanist History to Linguistic Theory
97
in Flemish than in Latin, Greek or Hebrew, Becanus reasoned that his was the oldest language, from which all others were derived. He attempted to demonstrate this by showing that words in other languages were originally composed of Flemish rootwords. For instance, Becanus (1569) argued that the name of the Germanic tribe of Saxons has at its root sagun, a compound made up of the two roots Sac + Gun, where Sac = ‘Sack’, but meaning here ‘causa’ and Gun should be read as in German Gunst, yielding the meaning ‘favoured by the first cause, i.e. favoured by God’ for the name of the Saxons (Metcalf 1974: 273–274). The German grammarian Schottelius went a step further in using monosyllabicity as an argument for the originality of German. Agreeing with Becanus’s view that simple things have simple words, so the oldest names are short and simple, he adduced a further powerful argument, that the phylogenetic origin of language would have been similar to the ontogenetic process of language acquisition in children. Just as children first learn to utter monosyllables, so the original words of language were short and simple (Schottelius, 1663 [1967]: 61, oration 4, ¶32). Hence German, rich in original ancient rootwords, had evidently remained closer and truer to the origin of language than other languages. Schottelius’s very inf luential grammar of German (Schottelius 1641, 1651, 1663) with its first book of ‘orations’ on the German language acted like a kind of funnel through which the notion of the rootword was transmitted to later generations of language scholars. Long after Johann Daniel Longolius (1677–1740) had rejected the notion of monosyllabicity as a criterion for Primitiva (Longolius 1715: 63; cf. Kaltz 2005: 126), Enlightenment debates about the origin of language – now stripped of any national chauvinism – still remained indebted to the notion of the monosyllabic word as a feature of the first languages. Consider this passage from Herder’s famous prize-winning treatise on the origin of language (Herder 1772; cf. Kief fer, 1978, Neis, 2003). Herder’s broad argument here is that human language is not perfectly logical, and so is not of divine origin. Einem Verteidiger des übernatürlichen Ursprunges [= Johann Peter Süßmilch (1707– 1767), 1766)] ist’s göttliche Ordnung der Sprache, ‘daß die meisten Stammwörter einsilbig, die Verba meistens zweisilbig sind und also die Sprache nach dem Maße des Gedächtnisses eingeteilt sei’.
98
Nicola McLelland [To a defender of the supernatural origin [of language] [= Johann Peter Süßmilch (1707–1767), 1766] it is the divine arrangement of language, ‘that most rootwords are monosyllabic, the verbs mostly bisyllabic, and that the language is thus divided up according to the measure of the mind.’ (Translation NMcL)]
Süßmilch had presumably considered verbs to be more complex than monosyllabic (nominal) roots, and so logically subordinate to them in thought. Language apparently mirrored thought so perfectly that it must be of divine origin. Herder disagreed, though not disputing the assumption that monosyllabicity entailed originality. Rather, he argued that since surviving verbs in the oldest languages are bisyllabic, they are evidence that the first, oldest, monosyllabic word have been lost, just as the babbling of children is no longer present in the speech of adults: In den Resten der für die älteste angenommenen Sprache sind die Wurzeln alle zweisilbige Verba; welches ich nun aus dem vorigen sehr gut erklären kann […]. Diese Verba nämlich sind unmittelbar auf die Laute und Interjektionen der tönenden Natur gebauet, die oft noch in ihnen tönen, hie und da auch noch als Interjektionen aufbehalten sind; meistens aber mußten sie, als halbinartikulierte Töne, verlorengehen, da sich die Sprache formte. In den morgenländischen Sprachen fehlen also diese ersten Versuche der stammelnden Zunge; aber daß sie fehlen und nur ihre regelmäßigen Reste in den Verbis tönen, das eben zeigt von der Ursprünglichkeit und – Menschlichkeit der Sprache. Sind diese Stämme Schätze und Abstraktionen aus dem Verstande Gottes oder die ersten Laute des horchenden Ohrs? die ersten Schälle der stammelnden Zunge? Das Menschengeschlecht in seiner Kindheit hat sich ja eben die Sprache geformet, die ein Unmündiger stammlet: es ist das lallende Wörterbuch der Ammenstube – wo bleibt das im Munde der Erwachsnen? [In the remains of the languages assumed to be oldest, the roots are all bisyllabic verbs; which I can now very well explain […]. For these verbs are built directly on the sounds and interjections of resounding nature, which often still resound in them, and here and there are still preserved as interjections; but mostly they were inevitably lost, as half-inarticulate sounds, when language was formed. In the eastern languages these first attempts of the stammering tongue are lacking; but the very fact that they are lacking and only their relics resound in the verbs, is exactly what shows the originality and – humanity of language. Are these roots treasures and abstractions from the mind of God, or the first sounds of the hearkening ear? the first sounds of the stammering tongue? Humankind formed in its infancy the language that a child stammers: it is the babbling dictionary of the nursery – where does that remain in the mouth of an adult?]
From Humanist History to Linguistic Theory
99
The details of the argument are not my concern here. What matters is that the equation of monosyllabicity with early, primitive language is accepted on both sides of the origin of language debate. 3.2 The verbal root and the imperative of the verb The second feature of the rootword identified by Irenicus was that it is identical with the second person imperative. Jellinek (1898) summarized the discussion for and against this view. Noted by Henisch (1616), it was championed most vigorously by the German grammarian Schottelius and by the popularizer of his ideas, Harsdörf fer (1607–1658). Yet for many Germans, the second person imperative wasn’t monosyllabic at all (an uncertainty – mach or mache! – that still persists; cf. McLelland, 2011b). The heated debate about and ultimate fate of such paragogic or apocoped -e’s is a familiar instance of the interactions between social history and language history, as the non-apocoped forms in the language became identified with northern Protestantism in the eighteenth century (Habermann 1997). At any rate, Jellinek noted that despite much opposition from within the Fruitbearing Society (Fruchtbringende Gesellschaft), the Schottelian view dominated. Beyond Germany too, the Danish grammarian Peder Syv (Syv 1979 [1663]: 132) took the rootword to be the imperative, and so did the Swedish grammarian Tiällmann (1696: 212–213), who listed irregular verbs by this root (cf. McLelland, 2011a, Chapter 7; McLelland 2010). Even once it ceased to play a strong role in German grammar,8 the dogma of the rootword imperative was particularly problematic to later language scholars because it was at odds with what was known about the triliteral and hence bisyllabic verbal roots of the Semitic languages. As Jellinek (1913: 138) had already implied, it has continued to be assumed (Barbarić 1981: 1195; Takada 1998: 25, n. 98), that Schottelius took the idea of the rootword from Semitic, especially Hebrew grammar, but there
8
There remained, however, a tendency to use the notion of the rootword more loosely as a pedagogical tool, as in the case of Martin Aedler’s (1680) German grammar for English learners (cf. Van der Lubbe 2007: 186).
100
Nicola McLelland
is no evidence that he did more that follow his humanist sources. In fact, later European grammarians of Hebrew and the Semitic languages struggled mightily to reconcile the monosyllabic verbal root – which it can be shown that they had inherited in a direct line of descent from Schottelius (cf. McLelland, 2011b, chapter 7) – with what they saw before them in the Semitic languages. Wilhelm von Humboldt concluded that the typically bisyllabic roots of the Semitic languages must have emerged from an earlier monosyllabic stage; only in a second stage of linguistic development did the recognition of a need for grammatical specification lead to the addition of a second syllable for the purposes of inf lection (cf. Rousseau 1984: 300, 320). Franz Bopp (1820) extended the debate about monosyllabic verbal roots to Sanskrit in comparison with the Semitic languages, when he sought to demonstrate that Sanskrit roots were monosyllabic; that they ‘may contain as few letters as are requisite to constitute a monosyllable’ (Bopp 1820: 8, cited Rousseau 1984: 309, 311). Ultimately, then, a humanist observation about a southern variety of German had a significant impact on the European study of Semitic and even Sanksrit. 3.3 Rootwords as typically Germanic building blocks for compounding In his obsession with identifying and interpreting the ‘true’ roots of Germanic proper names like Saxon, Becanus followed the example of Beatus Rhenanus, who had included much discussion of German proper names in the commentary to his edition of Germania in 1519 (Mertens 2004: 92). The Protestant humanist pedagogue Philipp Melanchthon followed suit in his editions of the Germania for school use (1538, 1557). For Melanchthon, the value in etymological study of German proper names lay on the one hand in the reminder of the inconstancy of all human concerns, and on the other in engendering a love for one’s fatherland (patriae amor, cf. Mertens 2004: 94). So identifying rootwords in German names became the task of any cultural patriot. The Dutchman Simon Stevin (1548–1620) went a step further. Better known in the history of mathematics as someone who popularized the use of decimals, Stevin is important in the history of linguistic ref lection
From Humanist History to Linguistic Theory
101
for his Vytspraeck vande Weerdicheyt der duytsche tael, his ‘oration on the worth of the duytsch language’,9 which occupied 24 sides at the front of his work De Weeghdaet (1586), ‘On the art of weighing’. This oration at the front of the book served the purpose of justifying the author’s use of his Dutch vernacular in the work that followed. In his oration, Stevin argued that simple things merit simple (i.e. monosyllabic) words, an idea that he seems to have taken from Becanus, or at the very least shared with him. Crucially for Stevin, such monosyllabic rootwords could then be combined to form new, more complex terms, a facility especially useful for coining technical terms. Stevin demonstrated in a series of tables (1586: 67–79) that Duytsch had 742 eensilbighe woorden inden eersten persoon; daerder de Latinen alleenlick 5 hebben; de Griecken gheen eyghentlicke, maer langhe vercort tot 45 (‘742 monosyllabic words in the first person, while the Latins have only 5 and the Greeks have no monosyllables, but only 45 long words that have been contracted’). The demonstration was developed further in a reworking of this treatise that appeared in 1608 (Stevin 1608: 24–38), where Stevin listed the monosyllabic verb roots as in the 1,586 Uytspraeck, but added 1,428 monosyllabic nouns, adjectives, etc. for Duytsch, 158 for Latin, 220 for Greek, giving totals of 2,170 for Duytsch, 163 for Latin, and 265 for Greek. Since Duytsch had by far the greatest number of these rootwords, Stevin argued that Duytsch was a richer and more capable language than either of the classical languages, and was particularly well-suited to coining new terms in mathematics (of which wiskonde ‘mathematics’ is a prime example; today Dutch is the only European language that does not use some variant of mathematica, from the Greek, via Latin).
9
It is dif ficult here to translate Duytsch. It means for Stevin something that encompasses Dutch and German at least. Evidence of its wider sense is the anecdote Stevin relates of the German humanist Henricus Glareanus switching from Latin to German during an oration on Suetonius because he found the German words to have the greater emotional appeal (Stevin 1586, in Dijksterhuis 1955: 86–87; Schottelius paraphrases Stevin’s account in the AA 64–65, 4: 42). Here, Stevin uses for German the same word, Duytsch, that he uses for his own vernacular.
102
Nicola McLelland
For Stevin, then, rootwords took on the additionally important feature of indicating both the richness and conciseness (brevitas) of German, in which it was superior to Greek, Latin and Hebrew. Its brevity was evident in the abundance of monosyllabic words; its richness in their almost infinite capacity for compounding. Want dit moet ghy weten, dat de sprakens goetheyt niet alleen voorderlick en is om de Consten bequaemlick daer duer te leeren, maer oock den Vinders in haer soucking […] wy segghen van Stof f waerheyt, […] en dierghelijcke, daer t’volghende vol af is; welcke woorden de Griecken soo cort, ende by haren yderman soo verstaenlic, oock so eyghentlick haer grondt beteeckenende, noyt en hebben connen segghen, nu niet en connen, noch, dat kennelick ghenouch is, inder eewicheyt niet connen en sullen. want datter niet in en is en cander niet uytghetrocken worden. [For you must know that the excellence of language is conducive not only to learning the arts well through it, but also to the search of the inventors […] we talk of Stof fwaerheyt [‘specific gravity’], […] and the like, in which the following [text] abounds; which words the Greeks never were, are not now, and never to all eternity will be able to say so shortly and so universally intelligibly to everyone of them, and also describing its nature so aptly, as is suf ficiently obvious. For what is not in it [sc. in a language], cannot be extracted from it’. (Stevin 1586: 5–6, in Dijksterhuis 1955: 384–387)]
Similarly, in a passage lifted wholesale by Schottelius (1663: 14, 1: 42), Stevin took pride in the unique ability of Duytsch to form compounds: want waer wildy spraken halen daermen duer segghen sal, Euestaltwichtich, Recht hefwicht, Scheefdaellini, en dierghelijcke daer de Weeghconst vol af is? sy en sijn der niet, de Natuer heeft daer toe aldereyghentlicxt het Duytsch veroirdent. [for where would you find any languages in which one can say Euestaltwichtich, Recht hefwicht, Scheefdaellini [‘of equal apparent weight’, ‘vertical lifting weight’, ‘oblique lowering line’] and the like, in which the Art of Weighing abounds? They do not exist, Nature has specially designed Duytsch for it. (Dijksterhuis 1955: 86–87)]
In the work of the German Schottelius, the monosyllabic rootword found its cultural-patriotic apotheosis. It was an expression of pure, unmixed, ancient Germanness and a feature of original, early languages. And it was the basic unit of language, for Schottelius developed Stevin’s observations about the combinability of German rootwords into an entire linguistic
From Humanist History to Linguistic Theory
103
theory. That view of German is eloquently illustrated by the frontispiece to Caspar Stieler’s (1691) dictionary, where the German language appears as a banyan tree sustained by its aerial roots that double as trunks (Figure 2; cf. McLelland 2002 for fuller discussion of this imagery, and McLelland 2011a, chapter 2). Stieler’s dictionary, the first full monolingual dictionary of German, listed German words by their root, rather than in strict alphabetical order, as the entry following illustrates: Sample entry for Baum, abridged from Stieler (1691: col.113–117) Baum / der / pl. die Beume / arbor, arbos. Baum etiam quod dicitur quodvis lignum aedificationi aptum. […] Beume pf lanzen / arbustare, arbores serere […] Der Baum trägt ein Jahr üms ander / arbor alternat fructus. Fachbaum […] Obstbeume […] Wilde Beume […] Beumlein / das. dim. arbuscula […] Beumen […] Faulbaumen […] Kürbaumen […] Baumlechtig & beumechtig […] Baumelen […] Baumelung […]
The cultural patriotic belief in the rootword as guarantor of German antiquity and purity, as monosyllabic, and as identical with the imperative all faded in the eighteenth century, but Schottelius’s belief in its centrality as the basic unit of language survived. In the mid-nineteenth century, it was still the structuring principle for the as yet unsurpassed Middle High German dictionary of Benecke (Benecke 1854–1866 [1990]),10 structured just like Stieler’s dictionary by rootword lemmata, and to which a more user-friendly alphabetical index was published by Lexer in the 1870s (Lexer, 1872–1878 [1979]). 10
The first fascicles of the new Middle High German dictionary which will ultimately replace Benecke’s work have been appearing since 2006 (Stuttgart: S. Hirzel, ed. by Kurt Gärtner, Klaus Grubmüller, and Karl Stackmann).
104
Nicola McLelland
Meanwhile, from Steinbach’s (1724) important insight that the root might not be identical with any existing form of the word ( Jellinek 1898: 71) and then the systematic, synchronic comparison of Semitic and Sanskrit roots by Bopp and others, the idea of the root as an abstraction gradually emerged, removed from etymological speculations, and, Rousseau claims, ultimately made ‘thinkable’ the idea of the morpheme (Rousseau 1984: 294). If Rousseau is right, then we must note the role of the sixteenthcentury humanist Irenicus, at several removes, in developing the abstract notion of the morpheme. Furthermore, as any student of German linguistics knows, Schottelius’s rootword-based account of derivation and compounding as the two basic types of word-formation remains current in 21st-century introductions to German morphology. Naumann (2000: 42) makes the distinction in essentially the same terms as Schottelius: Für die Bildung deutscher Wörter […] hat die Wortbildungsforschung vor allem zwei Grundtypen erarbeitet, den Typ der Komposition oder Zusammensetzung ursprünglich voneinander abhängige Wörter zu einem neuen Wort aus mindestens zwei Teilen und den Typ der Derivation oder Ableitung, d.h. Wortbildung mittels Af fixen, also Präfixen, Infixen und Suf fixen. As regards the formation of German words, researchers of morphology have postulated two basic types of word formation: composition or the merging of at least two, previously independent words to form one new word and derivation, i.e. word formation by means of af fixes, i.e. prefixes, infixes, and suf fixes.
4 Conclusion: The rootword in other Germanic languages I have already mentioned the inf luence of the rootword on Danish and Swedish grammarians, to which should be added the striking case of a grammar of Russian that, though never printed, had some inf luence (McLelland 2010; Huterer 2001). One final example from Dutch may suf fice to illustrate the long shadow of the Germanic root, in linguistic practice as well as in theory, and that is the case of systematic dif ferences between Dutch and
105
From Humanist History to Linguistic Theory
German spelling. In German, Schottelius (AA 191–192) adjudged that we should write Pferd because of the inf lected form Pferde, even though the -d in Pferd sounds like a -t because of final devoicing (or final fortition). In Schottelius’s language theory, the integrity of the rootword dictated that the rootword should remain the same throughout all word-forms. The Dutch grammarian Arnold Moonen (1644–1711) was a close adherent of Schottelius’s linguistic theory, and held with him that the rootwords should remain the ‘same’, but he accepted that this ‘sameness’ could include consonant alternations that ref lected actual pronunciation (Moonen, 1706). He was thus the first of a series of eighteenth-century Dutch grammarians to take issue with the forms that the strict analogical principle of Schottelius would require, such as wijv–wijven (‘woman, women’, see (Schaars 1988: 92), preferring instead to mark the dif ference in pronunciation. So Moonen accepted alternations of voiced and voiceless consonants in wordforms like geef–geeven (‘give’, first person sg., and infinitive or pl.). Séwel (1708) would follow Moonen, and the defeat of the analogical principle by the phonetic became established. The result is still evident in the contrast between standard Dutch and standard German spelling, where Dutch alternates consonants, and German does not (Schaars 1988: 94; Table 7.3; see Table 1).11 Table 1 Consonant alternations in Dutch spelling, compared with German Dutch
German
(English)
huis, huizen geef, geven
Haus, Häuser gib!, geben
‘house’ ‘give’
To conclude, this study has shown how the rootword – first extrapolated by an early sixteenth-century humanist from southern German as monosyllabic and identical with the imperative, and then repeated by other humanist historians – grew into a defining characteristic of Germanic language theory. Having become indexical of antiquity and originality, it fed into
11
Compare English wife–wives, hoof–hooves.
106
Nicola McLelland
eighteenth-century debates on language origin. Understood as a verbal root, it inf luenced comparative linguists examining Semitic and Sanskrit well into the nineteenth century. As the building block of derivation and compounding, it inf luenced lexicography into the mid-nineteenth century; it yielded in Schottelius’s work the basis of present-day word-formation theory; and dif fering views as to its integrity led to dif ferences in spelling that are still salient in the Germanic languages today.
References Aedler, Martin [anon.]. 1680. The Hig [sic] Dutch Minerva […]. London: Printed for the author [Facsimile reprint Menston, England: Scolar Press, 1972]. Benecke, Georg Friedrich. 1854–1866 [1990]. Mittelhochdeutsches Wörterbuch. Mit Benutzung des Nachlasses von Georg Friedrich Benecke ausgearbeitet von Wilhelm Müller und Friedrich Zarncke. Nachdruck der Ausgabe Leipzig 1854–1866. 4 vols. and index. Stuttgart: S. Hirzel. Bopp, Franz. 1820 [1974]. ‘Analytical Comparison of the Sanskrit, Greek, Latin and Teutonic Languages, Shewing the Original Identity of Their Grammatical Structure’. Annals of Oriental Literature 1, 1–64. Repr. edn. Amsterdam: E.F.K. Koerner, Benjamins. Considine, John. 2007. Dictionaries in Early Modern Europe. Cambridge: Cambridge University Press. Dijksterhuis, E.J. (ed.). 1955. The Principal Works of Simon Stevin. Vol. I. Amsterdam: C.V. Swets & Zeitlinger. Elspaß, Stephan. 2005. Sprachgeschichte von unten. Untersuchungen zum geschriebenen Alltagsdeutsch im 19. Jahrhundert. Tübingen: Niemeyer. Elspaß, Stephan et al. (eds). 2007. Germanic Language Histories ‘from Below’ (1700– 2000). Berlin: de Gruyter. Erlinger, Hans Dieter. 2007. ‘Zur Geschichte des muttersprachlichen Sprachunterrichts in Deutschland 18.-20. Jahrhundert’. In: Geschichte der Sprachtheorie 6/2. Sprachtheorien der Neuzeit III/2, Tübingen: Narr, 526–558. Gardt, Andreas, Ulrike Haß-Zumkehr & Thomas Roelcke (eds). 1999. Sprachgeschichte als Kulturgeschichte. Berlin: de Gruyter.
From Humanist History to Linguistic Theory
107
Habermann, Mechthild. 1997. ‘Das sogenannte “Lutherische e”. Zum Streit um einen armen Buchstaben’. Sprachwissenschaft 22, 435–477. Haß-Zumkehr, Ulrike. 2000. ‘Das Deutsche Wörterbuch von Jacob Grimm und Wilhelm Grimm als Nationaldenkmal.’ In: Andreas Gardt (ed.), Nation und Sprache. Die Diskussion ihres Verhältnisses in Geschichte und Gegenwart. Berlin: de Gruyter, 247–272. Hellmann, Manfred, & Schröder, Marianne (eds). 2008. Sprache und Kommunikation in Deutschland Ost und West. Ein Reader zu fünfzig Jahren Forschung. Unter Mitarbeit von Ulla Fix. Mit einem Geleitwort von Wolfgang Thierse. Germanistische Linguistik 192–194. Hildesheim: Olms. Henisch, Georg. 1616 [1973]. Teütsche Sprach und Weißheit: Thesaurus linguae et sapientiae Germanicae, A-G. Augsburg 1616. Repr. edn. Hildesheim. Hüllen, Werner. 2002. Collected Papers on the History of Linguistic Ideas. Münster: Nodus. Huterer, Andrea. 2001. Die Wortbildungslehr in der Anweisung zur Erlernung der Slavonisch-Rußischen Sprache (1705–1729) von Johann Werner Paus. Munich: Sagner. Hutton, Chris. 1999. Linguistics and the Third Reich. Mother-tongue fascism, race and the science of language. London, New York: Routledge. Jellinek, Max. 1898 [1985]. Ein Kapitel aus der Geschichte der deutschen Grammatik. In: R. Heinzel and F. Detter (eds), Abhandlungen zur germanischen Philologie. Halle: Niemeyer, 31–111. Jellinek, Max. 1913–1914. Geschichte der neuhochdeutschen Grammatik von den Anfängen bis auf Adelung. Heidelberg: Carl Winter. Jones, William J. 1999. Images of Language. German attitudes to European languages from 1500 to 1800. Amsterdam: Benjamins. Kaltz, Barbara. 2005. ‘Zur Herausbildung der Wortbildungslehre in der deutschen Grammatikographie von den Anfängen bis zum Ende des 19. Jahrhunderts.’ In: Schmitter, Peter (ed.), Geschichte der Sprachtheorie. 6/1 Sprachtheorien der Neuzeit III/1. Tübingen: Narr, 105–161. Kief fer, Bruce. 1978. ‘Herder’s treatment of Süssmilch’s theory of the origin of language in the Abhandlung über den Ursprung der Sprache: a re-evaluation’. The Germanic Review 53, 56–105. Krebs, Christopher. 2010. ‘A Dangerous Book: The Reception of Tacitus’ Germania’. In: The Cambridge Companion to Tacitus, ed. A.J. Woodbridge. Cambridge: CUP, 280–299. Lee, Christina, & Nicola McLelland (eds). 2011. Germania Remembered. Tuscon, AZ: ACMRS (Medieval and Renaissance Texts and Studies).
108
Nicola McLelland
Lexer, Matthias. 1872–1878 [1979]. Mittelhochdeutsches Handwörterbuch: zugleich als Supplement und alphabetischer Index zum Mittelhochdeutschen Wörterbuch von Benecke-Müller-Zarncke. Repr. of the Leipzig edition 1872–1878. Stuttgart: Hirzel. Longolius, Johann Daniel. 1715. Einleitung zur gründtlicher Erkäntniß einer ieden, insonderheit aber der Teutschen Sprache. Budissin: Richter. McLelland, Nicola. 2010. ‘Justus-Georg Schottelius and European linguistic thought’. Historiographia Linguistica 37.1, 1–30. McLelland, Nicola. 2011a, 2011. J.G. Schottelius’s Ausführliche Arbeit von der Teutschen Haubtsprache (1663) and its place in early modern European vernacular language study. Oxford: Blackwell (Publications of the Philological Society). McLelland, Nicola. 2011b. Des guten Gebrauchs Wegzeigere (Schottelius 1663: 10) – du bon usage dans la tradition allemande 1200–2000. In: Wendy Ayres-Bennett (ed.), Le Bon Usage. Paris: ENS. Metcalf, George J. 1974. ‘The Indo-European hypothesis in the sixteenth and seven teenth centuries’. In: Hymes, Dell (ed.), Studies in the History of Linguistics: Traditions and paradigms. London: Bloomington. Mertens, Dieter. 2004. ‘Die Instrumentalisierung der “Germania” des Tacitus durch die deutschen Humanisten. Take-of f Phase der deutschen Tacitus-Rezeption und der Umbau der Geschichtsbilder’. In: Beck, Heinrich, Dieter Geuenich, Heiko Steuer & Dietrich Hakelberg (eds, Zur Geschichte der Gleichsetzung germanischdeutsch, by Johannes Hoops et al. Sprache und Namen, Geschichte und Institutionen. Berlin: de Gruyter, 37–101. Mieder, Wolfgang. 1982. Die Eintstelluing der Grammatiker Schottelius und Gottsched zum Sprichwort. Sprachspiegel 38, 70–75. Moonen, Arnold. 1706. Nederduitsche Spraekkunst. Amsterdam: François Halma. Mundt (ed.). 2008. Beatus Rhenanus. Rerum Germanicarium Libri Tres (1531): Ausgabe, Übersetzung, Studien. Tübingen: Niemeyer. Neis, Cordula. 2003. Anthropologie im Sprachdenken des 18. Jahrhunderts. Die Berliner Preisfrage nach dem Ursprung der Sprache (1771). Berlin: de Gruyter. Rössing-Hager, Monika. 1984–1985. ‘Ansätze zu einer deutschen Sprachgeschichts schreibung vom Humanismus bis ins 18. Jahrhundert.’ In: Besch, Werner, Oskar Reichmann & Stefan Sonderegger (eds), HSK Sprachgeschichte. Ein Handbuch zur Geschichte der deutschen Sprache und ihrer Erforschung. (Art.1144). Berlin: de Gruyter, 1564f f. Rousseau, Jean. 1984. ‘La racine arabe et son traitement par les grammairiens européens (1505–1831).’ Bulletin de la Societé de Linguistique de Paris 79, 285–321. Schaars, Frans. 1988. De Nederduitsche Spraekkunst (1706) van Arnold Moonen (1644– 1711). Wijhe: Quarto.
From Humanist History to Linguistic Theory
109
Schottelius, Justus Georg. 1641 [1651]. Teutsche Sprachkunst […]. Braunschweig: Gruber. Rev. edn. Braunschweig: Zilliger. Schottelius, Justus Georg. 1663 [1967]. Ausführliche Arbeit von der teutschen Haubtsprache. Braunschweig: Zilliger. Repr. edn. Wolfgang Hecht. Tübingen: Niemeyer. Séwel, Willem. 1708. Nederduytsche Spraakkonst. Amsterdam: Assuerus Lansvelt. Steinbach, Christoph Ernst. 1724. Kürtze und gründliche Anweisung zur Deutschen Sprache. […]. Rostochii et Parchimi: Apug Georg Ludw. Fritsch. Stevenson, Patrick. 2003. Language and German Disunity. A sociolinguistic history of East and West in Germany, 1945–2000. Oxford: Oxford University Press. Stevin, Simon. 1586. De Weeghdaet / Beschreven Dver Simon Stevin van Brugghe [contains Vtspraeck vande weerdicheyt der dvytsche tael]. Leiden: Inde Druckerye van Christof fel Plantijn. By François van Raphelinghen. Stieler, Kaspar. 1691 [1968]. Der Teutschen Sprache Stammbaum und Fortwachs […]. Nuremberg: Johann Hof fman. Repr. mit einem Nachwort von Stefan Sonderegger. München: Kösel. Süßmilch, Johann Peter. 1766. Versuch eines Beweises, dass die erste Sprache ihren Ursprung nicht vom Menschen, sondern allein vom Schöpfer erhalten habe (1766). Berlin: Buchladen der Realschule. Takada, Hiroyuki. 1998. Grammatik und Sprachwirklichkeit von 1640–1700. Zur Rolle deutscher Grammatiker im schriftsprachlichen Ausgleichsprozeß: Tübingen: Niemeyer. Van der Lubbe, Fredericka. 2007. Martin Aedler and the High Dutch Minerva. The First German Grammar for the English. Duisburger Arbeiten zur Sprach- und Kulturwissenschaft, 68. Frankfurt: Lang. Van der Waal, H. 1952. Die eeuwen vaderlandsche geschied-uitbeelding, 1500–1800: een iconologische studie. The Hague: Martinus Nijhof f. Verstegan, R. 1605. A Restitution of Decayed Intelligence: In Antiquities. Concerning the Most Noble and Renowmed [sic] English Nation. Antwerp. Weststeijn, Thijs. 2008. ‘The Germanic origins of art: Dutch and English antiquity according to Verstegan, Junius and Van Hoogstraten.’ Dutch Crossing 32, 44–70.
Agnete Nesse
Editorial Practices and Language Choice: ‘Low German Language Monuments’ in Norway
Abstract A great deal of documents from Norway in the late Middle Ages and early modern times were written in Middle Low German. Since the middle of the nineteenth century, many Low German documents have been edited, and editorial practices have varied significantly. Some editors have chosen to publish the texts without comments or explanations; others have used combinations of quoting (in the original language) and referencing (in Norwegian) within the same text. There are also editors who have chosen to translate the documents into Norwegian, only brief ly mentioning that the original was written in another language. The editorial practice often arises from the research questions perceived by the editor as relevant for the document in question. Given that what was considered as important for research questions has changed over time, we see that editions that are strongly linked to one specific research tradition can be dif ficult to use successfully today.
1 Introduction When working with historical texts, we are normally aware ‘that texts are produced and reproduced under specific social and institutional conditions’ (McGann 1991: 21). In other words, a given text is not simply the work of an author, but should also be considered as a ref lection of the time and the society in which it was written. It is therefore important, when working with historical texts, to regard spelling, grammar, style, argument and content
112
Agnete Nesse
as related to the social conditions of the writer – and of the copiers, who could also be important contributors to a given text (Haugen 1988: 72). Editorial practice must be viewed in the same way. In order to understand and successfully use an edition, it is necessary to analyse the potential research questions of the editor. It is important to recognize the relationship between the edition and the theoretical and ideological background of the editor, in order to avoid misinterpretation of historical texts. We should ask ourselves a number of questions concerning the current editions we use and the future editions we want to publish: Why is a given text edited at a specific time and in a specific manner? What aspects of the text ref lect the interests of the time and the society in which it is edited? What choices did the editor make, and why? (McGann 1991: 23). There is an extensive amount of literature on the subject of editing and literary criticism, written by historians, linguists and literary critics (Fidjestøl et al. 1988, McGann 1991, Hunter 2006). The authors’ dif ferent approaches to the subject are not only concerned with their academic background, but also with the text type itself. Working with printed texts entails a dif ferent kind of challenge from working with handwritten manuscripts (as shown by Hunter 2006), and those who work with several manuscripts of the same text follow dif ferent methods to those working with just one manuscript. Regarding some of the editors’ choices, I have chosen to examine texts from Norway written in Low German in the sixteenth and seventeenth centuries. The texts were edited over a specific period of time, from the latter half of the nineteenth to the first half of the twentieth century. None of the texts were originally printed, hence they are all handwritten manuscripts, and with the exception of one text, only one manuscript exists in each case. I will only give a brief analysis of the dif ferent texts here, as my main focus will be the choices the editors made according to the language of the editions – in short, to what extent they translated the texts into Norwegian. One might assume that this is essentially an easy task: either they translated the texts or they did not. Yet there is a lot more to such ‘bilingual editing’, and I will discuss five dif ferent strategies.
Editorial Practices and Language Choice
113
2 Transcription only Obviously, one strategy when it comes to editing German ‘language monuments’ is simply to transcribe the handwritten text and publish it. There are dif ferent views among editors as to how many and what kind of alterations can or should be made to a text within this strategy. For example, with regard to dealing with abbreviations: Should abbreviations be expanded, and if so, how? Punctuation marks and capital letters are also often changed or added to make the text more readable, even within academic editions. The Norwegian collection of mediaeval charters, Diplomatarium Norwegicum, serves as an example of this. The first volume was published in 1847, today twenty-two volumes with around 20,000 texts have been published, both on paper and digitally,1 and the twenty-third volume is on its way. In the preface of the first volume (Lange and Unger 1847: VIII–IX), the editors write that even though they have left all spelling mistakes in the editions, they have expanded abbreviations, because using særskilte dertil stöbte Tegn (‘specially cast symbols’) were expensive and for many readers the use of such symbols made the text dif ficult to read. Furthermore, capital letters are used on all nouns, til Lettelse for Historikeren (‘for the ease of the historian’). The preface states that the edition is for both historians and linguists, but that the two groups should use it somewhat dif ferently. The linguist could achieve a general picture of the linguistic changes in the period (from the eleventh century until 1570), and study dialectal dif ferences. However, for a detailed study of the text, the linguist should, according to the editors, consult the original manuscript (p. IX). In addition, linguists should work only with those texts in the edition that were based on original manuscripts, not with those based on copies. The texts based on copies are meant for historians only (p. X), and are therefore treated dif ferently: mistakes are corrected, the orthography is normalized, and so is the punctuation. The editors’ awareness of the dif ficulties in providing one edition for both linguists and historians is remarkable, in light of this being one of the very first academic editions of charters in Norway. 1
.
114
Agnete Nesse
Hunter (2006: 76–80) of fers an interesting discussion about how abbreviations, changed letter-forms and the like should best be presented in print. In his eagerness to modernize old texts within the framework of academic editing, he strongly supports a view that, at least in the eyes of a linguist, seems rather extreme. His claim (p. 85) that ‘Overall, the aim is to produce an edition which does justice to the content of the manuscript, paying attention to its actual appearance but not fetishizing this’ shows little respect for the linguist’s need to know the actual orthography of the writer, for example when it comes to abbreviations. However, the question of abbreviations is only mildly relevant to the text I have chosen, since the number of abbreviations in the manuscript is very few. Yet, as we can see from the photograph on the next page and from the copy of the edition (Nielsen 1877), the abbreviated forms tym ˜ er and Tym ˜ erholt in the manuscript become tymmer and Tymmerholt in the edition, with no marking of the expansion. The text in question is in the form of a dialogue between a Norwegian, De Norman, and a Hanseatic merchant, De copman, supposedly written in the early sixteenth century in Lübeck. De Norman represents the native inhabitants of Bergen, and he complains about the behaviour of the Hanseatic League, represented by De Copman. The merchant, on the other hand, claims that the accusations are false, and that the Hanseatic merchants are in fact following the law. He adds that if this is not the case, they will punish those who break the law. Nielsen’s first edition is followed by a number of comments, but none of the comments are concerned with the lack of translation. Is that because all his potential readers, Norwegian academics in the late 1870s, could read Middle Low German? Or was it because his potential readers were expected to handle linguistic challenges of this kind, by taking the time needed to find out what the text was all about? From the dialogue De dudesche unde de Norman manuscript2 Nielsen’s edition (1877: 26): 2
The city archive of Lübeck, Bergenfahrerarchiv, no. 1418.
Editorial Practices and Language Choice
115
De Norman. Item dat de copman suluen In den schof f vart of ft er [sic] volck vmme tymmer Berneholt vnde anderholt vp ores sulues kost vnde nicht enkopen van deme Normanne alsze oldinges plach to wesen Dar is der Normanne bergynge mede vorstort de sick hyr mede plegen to Bargende. De copman. Wy enlatet nenerleye Tymmerholt of fte berne holt houwen of fte halen ydt ensy myt der guder lude wyllen of fte orlof f den de schof f tobehort deme gheue wy dar al vul vor Breckt dar wol an dat is vns vnwitlick De do so vele dar vor alse recht is. [The Norwegian: Further that the merchants themselves go into the woods and obtain timber and firewood at cost price and do not buy it from the Norwegians as it has been customary since olden times. This ruins the livelihood of the Norwegians. The Merchant: We do not permit that timber is chopped or fetched, unless it is with the will or permission of the good people. To those to whom the woods belong, we give what they ask. If there is anyone who acts against this, it is unknown to us. In that case, he will get what is right (i.e. punishment).]
What Nielsen does comment on, however, is the fact that he chose not to change the language into a standardized version of Middle Low German. This was an important issue of the time: Scholars busied themselves with what James Milroy (2005: 336) has called ‘retrospective normalisation’, constructing standards like Old Norse, Middle Low German and Middle English.3 Nielsen’s reason for not standardizing was not that he was against it, but that he did not know how to do it:
3
Old Norse was standardized during the 1800s by Peter Andreas Munch and Rudolf Keyser, amongst others. The reason was mainly romantic-puristic in nature: an aim to achieve the perfection that had once existed. But also pedagogical arguments were important: by standardizing the language, it was thought to be easier to read (Mørck 2006: 27).
116
Agnete Nesse Da jeg ikke tidligere har syslet med Udgiverarbeide af ældre nedertyske Sprogmonumenter, har jeg anseet det for den retteste Fremgangsmaade at levere et bogstavret Aftryk, uden at følge den Normalorthografi, som i den senere Tid er anvendt, f. Ex. i Udgaven af hanserecesserne. (Nielsen 1877: 34) [As I do not have previous experience in editing older Low German language monuments, I have seen the most correct procedure to make a letter for letter print, without following the standard orthography, that has been used lately, for example in the editions of the Hanserecesse.]
As a result of this, Nielsen was criticized by the German scholar K. Hegel for having added grammatical mistakes to the text (Nielsen 1892: 1). Yet why did Nielsen choose not to translate this text into Dano-Norwegian? Did he want to distance himself from the Hanseatic League? He interpreted the text as pro-Hanseatic, so editing it at all might have been controversial at a time when the emphasis on the Hanseatic period was negative. However, Nielsen was not one of the most nationally orientated historians; he belonged to a more urban, European-minded school, so his choice of text might have been an attempt to oppose the majority of historians, who concentrated on building a sense of national identity. We could see his editorial practice as a compromise between showing his colleagues that the text existed, and not involving himself with the text by translating it. Was the dialogue pro-Hanseatic? According to Friedrich Bruns, a German historian who worked on the Hanseatic relationship between Bergen and Lübeck from many angles, the text was written by Jacob Dus, the secretary for the Bergen traders in Lübeck from 1512 until 1538 (Bruns 1901: 142). This may well be the case: the secretary for the Bergen traders must have had access to the dif ferent amendments, complaints and other letters concerning the Hanseatic settlement in Bergen. This dialogue text is also clearly a transformation of several political documents dealing with conf licts in Bergen into a didactical text. The most prominent of these is an amendment issued in both Norwegian and Low German by King Christof fer in 1444 (Taranger 1912: 234–243). The part of De Norman in the dialogue follows this amendment quite closely; the part of De Copman is written as a defence, instead of a counter-attack. This is a choice made
Editorial Practices and Language Choice
117
by Dus; there are several documents where the Hanseatic Merchants complain about the Norwegians in Bergen, so in theory, the dialogue could have been written with De Copman as the plaintif f and De Norman as the defendant.
3 The text with comments Our next strategy is chosen by editors who acknowledge that their readers may have some problems with reading the original language, and therefore include word lists in their editions. In this case, we clearly see how important the target audience is. The first quotation is from an edition published in 1897, and the intended readers appear to be Norwegians who did not necessarily understand Middle Low German. The text itself is a Hanseatic journal (Der Gesellen Boeck) written between 1604 and 1671. Notes were taken each year in May and June, recording the apprentices who had participated in the initiation “games” (gespelet), and how this was celebrated – usually by happily drinking beer in the assembly rooms or in the gardens. Since the manuscript in question has many name lists and many words that are repeated several times, the editors chose to make a word list instead of a full translation: Man har dog ment, til veiledning for det publikum, som faar bogen ihænde, at give en oversættelse af de vigtigste og vanskeligere steder, eller av de steder, hvor en stadig gjenkommende vending første gang findes. (Bendixen and Krohn 1897: 16) [We have, for the guidance for the audience of this book, given a translation of the most important and more dif ficult words, or where a frequently repeated phrase comes for the first time.]
The editor makes the judgement for the readers as regards what is important and what is dif ficult within the text: yet, in reality, the readers’ interpretation of the text often contradicts the editor’s own one.
118
Agnete Nesse
As a contrast to Bendixen and Krohn, we may look at an edition from 1946 of another kind of journal,4 written as an appendix to the Norwegian Germanist Olav Brattegard’s doctoral thesis. This is a so-called Maskopbok, where the people who owned shares in parts of the same buildings and piers recorded, once a year, what each owner paid for maintenance. Here too, a word list is included, but it is clear that he has a dif ferent group of readers in mind. Eine Auswahl zur Erläuterung einiger mit den Verhältnissen in der hansischen Niederlassung zu Bergen eng verknüpften Wörter. (Brattegard 1946: 78) [A selection to explain some words that are closely connected to the conditions of the Hanseatic settlement in Bergen.]
Many of the words connected to the Hanseatic settlement were Norwegian loan words that were part of the ‘Bergen-German’ language. Examples from Brattegard’s word list are gollfe (modern Norwegian gulv or golv [f loor]) and the phrase kuss mig j Rouven (modern Norwegian kyss meg i ræven [kiss my ass]), which are perfectly understandable for Norwegian readers, but not for German ones. As a result, Brattegard, who wrote and published his thesis in Bergen, explained words of Norwegian origin to his readers. However, since we know that his aim was to analyse the Low German language in Bergen, it makes sense that he commented only on those words that belonged to this language, i.e. words that were not used in the Low German found in Germany. In other words, he addressed readers who were familiar with Low German of the sixteenth and seventeenth century in Germany. What the editions of these journals have in common is the fact that the editors have not been interested in the authors, or in the authors’ intention concerning their texts, simply because there were so many authors, or rather scribes. Brattegard (1946: 19–41) recognized sixty-two dif ferent
4
These journals, of which several still remain, written during the period 1529–1936, in Low German, High German, Danish and Norwegian, are given a closer presentation in Nesse (forthcoming) ‘The Neighbours’ books of Bryggen’.
Editorial Practices and Language Choice
119
writers in the period from 1577 to 1670 in one journal alone. This, of course, of fers great opportunities for variation studies, and Brattegard does give a description of the orthography of the dif ferent writers, concentrating on the features that shed light on the transmission from Low to High German (pp. 18–41).
4 The text with translation The strategy of editing a text and then translating the whole text is time consuming, but reader-friendly. In this way, those who can read the language of the original do so, and those who do not understand the original language can still enjoy the content of the text. Only one of the many German texts written in Norway has been edited in this way,5 namely a Hanseatic journal from the sixteenth and seventeenth centuries, written partly in Low German and partly in High German, and translated into Dano-Norwegian (Bendixen & Krohn 1895). It is probably correct to say that no other text has been used so much for research on the Hanseatic society in Bergen as this one. It seems that the choice to edit both the original language and translation not only makes the text accessible, but it also gives the text an authenticity that makes its use particularly valuable. Hence, owing to its translation, this one small text has provided a more comprehensible (historical) account of the manuscript, on behalf of a large amount of other texts. Brattegard (1945: 43) is critical of the editorial practice of Bendixen and Krohn, because of the lack of accuracy in the editions. With regard to linguistic research, Brattegard’s own edition of a similar journal (1946) is a much better choice, given that it is both more accurate and its commentary more thorough.
5
In addition, there are four letters that were translated from Norwegian into Low German in the 1550s, where both versions are edited, three in Nesse 2008 and one in Nesse 2009.
120
Agnete Nesse
5 Translation only Including translation in a discussion of dif ferent editorial strategies may be controversial since, after all, the questions that arise when reading a translation are dif ferent to those that arise from looking at an edition. These questions are not concerned with how the text is presented, in terms of orthography, but rather with how the semantic level is retained. Nevertheless, there may be reasons to include this kind of work, as Hunter (2006: 88) does, if one wants to explore the dif ferent ways a historical text can be presented to modern readers, and what implications the editors’ choices have for research possibilities. One of the Norwegian editors who preferred translation to the original editing of texts that were not from the Middle Ages was Nicolay Nicolaysen, a contemporary of Nielsen. However, the two historians belonged to dif ferent schools and this may explain their dif ferent choice of editing. As mentioned earlier, Nielsen belonged to a school interested in urban, European history. Nicolaysen, on the other hand, worked within the national romantic paradigm. Among his many activities was the restoration of old buildings, such as churches – restorations that are today viewed as rather ruthless, given that everything that was added after the time of the original construction was taken away, so that the core building could once again take pride of place (Lidén 2005: 46–47). This can be compared to his ideas of how historical texts from dif ferent periods should be treated: Anything that was changed after the Old Norse language (or Middle Low German for that matter), was not of linguistic interest, and could be ‘taken away’ – translated into modern Dano-Norwegian, whether it was written in German or Norwegian. This had to do with the process of nation building in Norway, a process that focused on the Middle Ages as the period before the Dano-Norwegian unification ‘ruined’ the political, cultural and linguistic independence of the country. Since the written language was inf luenced by Danish after the fourteenth century, the nation builders considered it as impure. Nevertheless, Nicolaysen chose to edit several of the texts from the ‘400-year night’ (= the Dano-Norwegian union), and his three-volume work consisting of texts written in Norway between 1550 and 1760 is of great value, even if his
Editorial Practices and Language Choice
121
editorial practices are considered as questionable by modern linguists and historians. When editing Die Nordische Saw (The Nordic Sow) from the 1580s, a religious and political text that is very critical of both the Hanseatic League and the Dano-Norwegian noblesse, he comments: Dog kan der neppe være tale om nogen anden fremgangsmaade ved udgivelsen, end […] at gjengive bogen i uddrag og paa modersmaalet. Thi dels indeholder den en hel del almindelige og vidtløftige gudelige betragtninger, som ikke ere trykning verd, og dels er sproget forskjelligt i de forskjellige gjenparter, endog i samme afskrift, da det bevæger sig mellem en slags Plattydsk og Højtydsk, hvortil kommer, at retskrivningen er meget unøjagtig. (Nicolaysen 1868: 4) [Still there can hardly be suggestion of any other approach for editing, than […] to reproduce the book in extract and in the mother tongue. Partly, it contains a whole lot of common and high-f lying religious views that are not worth printing, and partly the language is dif ferent in the dif ferent manuscripts, even within the same manuscript, where it moves between a kind of Low German and High German. In addition, the spelling is very inaccurate.]
Front page of one of the manuscripts.6 The text under the drawing reads (cf. figure 4): Lucie am .21. Capitt: Es wirden Zeichen geschein ahn der Sohn, Mohn vnd Stirne, Vnd auf f Erden wirtt leutten Bange Sein/: [There will be signs in the sun, moon and stars. And on the earth, the people will be afraid. (Luke 21:25)]
When Nicolaysen wrote mother tongue, he was referring to his own mother tongue, not the author’s. The term modersmålet was used for the DanoNorwegian written language. Neither Danish nor Norwegian were accurate enough, and the term Dano-Norwegian was considered too long. Hence mother tongue was used as a compromise. 6
The royal library of Copenhagen, Gl.kgl.S.2836, made available to me by Geir Atle Ersland.
122
Agnete Nesse
For researchers today, it might be interesting to investigate the facts that determined Nicolaysen’s choice of editing. For those working with the history of ideas and religion, the combination of Christianity and astronomy that was so typical of the latter half of the sixteenth century would definitely merit further study, since the text is constructed like many religious texts, with metaphors and parables that are then outlined for pedagogical purposes. For those concerned with variation linguistics, the way the author and /or the copiers have mixed Low and High German, and the inaccuracy of spelling, are of particular interest. Ultimately, what did Nicolaysen want to achieve when he edited this text? He wrote that he wanted to show the ‘spirit and tone’ of the text, and include anything of historical interest. In terms of historical interest, this mainly involved the fact that the Danish king was a corrupt ruler, who let the Hanseatic League exploit the Norwegian population, both economically and morally. It is clear, therefore, that editing texts from the sixteenth century in the nineteenth century was very much an act of interpretation – in this case, interpreting the past to serve the political aims of the present, i.e. the independence of Norway. In comparison, a modern edition might emphasize the hatred which the anonymous Lutheran author displays to members of other religious groups. Catholics, Calvinists and Anabaptists are described in much the same way as Muslims are today in many western societies, thereby illustrating that religious dif ferences between the immigrants and the native population need not be great in order for hostility to arise.
6 Combining text and translation on sentence level The last strategy of bilingual editing that I am going to discuss is in many ways the most intriguing, and also the one that is hardest to analyse. Like translation, this is not editing in the strict sense, but it is nonetheless important to mention in an overview such as this. In this case, the editor splits the text into parts that he chooses to quote directly, and parts that he chooses
Editorial Practices and Language Choice
123
to shorten and translate. This was not an uncommon technique, and my interpretation of this strategy has previously been that is was simply a way to save space, whilst ensuring that those passages that were most important to the editor were quoted directly (Nesse 2002: 143–146). When we look at the following, this may be a plausible explanation: Da i 1660 (29 april) der ehrsame Mikkel von Drebben auss eigener authoritet ohne versammling der nachbarn und gesellen i eldhuset havde spenderet Gesellerne 1 Rdl. og derved handlet imot Gaardsrettens § 20, maatte han bøde 1 Tønde Mel. (Nielsen 1892: 13)
The phrase that is a direct quote is set in italics; the translated and shortened passage is in ordinary font. The original version from the manuscript7 reads as follows: Anno: 1660 denn 29 April: hat der Ersahme Mickel von drebben: auß eigener authoritet, ohne versamblung der nachbarn und gesellen in sohlgarten denn gesellen: in elthause einen schlichten R: spendiret: als uns nachbar nicht anders wißent: undt hat damit erwehnter Mickell von drebben: wieder den 20 Artikkel unßer gartenrechte gehandelt: midt himit an den armen verbrochen Lauth selbige vorberuhrten artikell, i thune mehl und damit alles aufgehobenn. [Anno 1660, on April 29th, the honourable Mickel von Drebben treated the young merchants to 1 Reichstaler (worth of beer) without this being agreed on in the assembly of neighbours, and without the neighbours’ knowledge. In this way, the aforementioned Mickel von Drebben has violated article 20 of our association rules, and must pay to the poor people one barrel of f lour, and by that it will all be annulled.]
Some of the direct quotes in German involve individual words of no particular significance concerning the comprehension of the text, such as kersse (candle), malzeit (meal) etc. This means that there must be reasons other than perceived semantic importance regarding the editor’s choice to quote and to translate certain words or phrases. Nielsen may have seen himself as a mediator who could make the text more relevant and interesting to his readers than the original text as a whole would have been. He may have
7
University library of Bergen, Ms 258.
124
Agnete Nesse
thought, like Nicolaysen when editing ‘The Norwegian Sow’, that the whole text from within these journals was not worth editing. In his edition, he chose excerpts from a large number of manuscripts, in both German and Dano-Norwegian, from the sixteenth century up until his own time. He translated a wide range of extracts from the journals to give his readers an impression of how the Hanseatic League was organized. He even quoted some parts of the text, to give the readers a historical f lavour of Hanseatic authenticity. Hence, the direct quotes in German may have less to do with language and semantic importance than with style.
7 The editor as interpreter In conclusion, we see editions of historical texts as voices from the time and the society in which they were made. The question remains, of course, as to whether this way of looking at editions of historical texts has any relevance for us today. Discussions of editing today are mostly discussions of digitalization. The possibilities that digital tools have to of fer seem to be limitless, yet the excitement surrounding methodological possibilities can easily overshadow the more theoretical aspects of editing. This becomes evident if we look at some of the articles and books written in the 1980s on the prospects of how computers could be used in analysing historical texts (Rindal 1988). Since such articles ref lect very dif ferent ideas from the interests of those dealing with digital editing today (for example the mediaeval text archive Menota), we see how important the technical possibilities are when formulating research questions. There can be no doubt that digitalization opens up great possibilities when it comes to bilingual editing. By using parallel screens, with one language per screen, and hyperlinks to explain words and phrases, one can make editions that fulfil the needs of both historians and linguists. Nevertheless, there are still choices to make and research questions to formulate. Consequently, by analysing earlier editions, we will be better prepared to create new ones.
Editorial Practices and Language Choice
125
References Bendixen, B.E., & Krohn, W.D. 1895. ‘Dat Gartenrecht in den Jakobsfjorden vnndt Bellgarden.’ Bergens historiske forenings skrifter 1. Bendixen, B.E., & Krohn, W.D. 1897. ‘Der Gesellen Boeck jm Jakubsforden vnde Belgarden, med indledning.’ Bergens historiske forenings skrifter 3. Brattegard, Olav. 1945. Die Mittelniederdeutsche Geschäftssprache des hansischen Kaufmanns zu Bergen. I. Die Sprache der Blütezeit. Bergen: John Grieg. Brattegard, Olav. 1946. Die Mittelniederdeutsche Geschäftssprache des hansischen Kaufmanns zu Bergen. II Der Ausklang des Niederdeutschen. Bergen: John Grieg. Braunmüller, Kurt, & Diercks, Willy (eds). 1993. Niederdeutsch und die skandinavischen Sprachen I. Heidelberg: C. Winter. Braunmüller, Kurt (ed.). 1995. Niederdeutsch und die skandinavischen Sprachen II. Heidelberg: C. Winter. Bruns, Friedrich. 1901. ‘Norweger und deutsche zu Bergen.’ In: Hansische Geschichts blätter. Leipzig: Verlag von Duncker & Hublot, 142–152. Halldórsson, Ólafur. 1988. ‘Tanker om tekstkritiske udgaver.’ In: Fidjestøl, Bjarne, Haugen, Odd Einar & Rindal, Magnus (eds), Tekstkritisk teori og praksis. Eit nordisk symposium. Godøysund 19.-22. mai 1987. Oslo: Novus, 11–23. Hunter, Michael. 2006. Editing Early Modern Texts. An Introduction to Principles and Practice. New York: Palgrave Macmillan. Lange, Christian, & Carl B. Unger. 1847. Diplomatarium Norwegicum, Første Samling, Første Hefte. Christiania: P.T. Mallings Forlagshandel. Lidén, Hans-Emil. 2005. Nicolay Nicolaysen. Et blad av norsk kulturminneverns historie. Oslo: Abstrakt forlag. McGann, Jerome J. 1991. The Textual Condition. Princeton, NJ: Princeton University Press. Menota hand book. Medieval Nordic Text Archive. accessed 12 August 2011. Milroy, James. 2003. ‘Purist ideologies on Historical Descriptions of English.’ In: Langer, Nils, & Winifried Davies. (eds), Linguistic Purism in the Germanic Languages. Berlin, New York: Walter de Gruyter. 324–342. Mørck, Endre. 2006. Unormaliserte middelaldertekster. Tekstprøver, tekstkommentarer og bakgrunnsstof f. Bergen: Fagbokforlaget. Nesse, Agnete. 2002. Språkkontakt mellom norsk og tysk i hansatidens Bergen. Oslo: Novus.
126
Agnete Nesse
Nesse, Agnete. 2008. ‘Bilingual texts from a bilingual city.’ In: Ersland, Geir Atle & Marco Trebbi (eds), Neue Studien zum Archiv und zur Sprache der Hanseaten. Bergen: Museum Vest. 47–64. Nesse, Agnete. 2009. ‘Flerspråklige kilder fra Bergen.’ In: Folkmålsstudier 47, 109–132. Nesse, Agnete. forthcoming. ‘Four languages, one text type: The neighbours’ books of Bryggen 1529–1936.’ In: Stenroos, Merja, & Martti Mäkinen (eds), Proceedings from Historical Language and Literacy in the North Sea Area, Stavanger 2009. John Benjamins. Nicolaysen, Nicolay. 1868. Norske Magasin. Skrifter og optegnelser angaaende Norge og forfattede efter reformationen. Vol. 2. Christiania: Johan Dahls forlag. Nielsen, Yngvar. 1877. ‘De dudesche Kopman unde de Norman.’ In: Forhandlinger i Videnskabs-Selskabet i Christiania Aar 1876 8. Christiania: Jacob Dybwad. Nielsen, Yngvar. 1892. ‘Vedtægter og Dokumenter fra det hanseatiske Kontor i Bergen og dettes enkelte Gaarde.’ In: Christiania Videnskabs-Selskabs Forhandlinger for 1892 7. Christiania: Jacob Dybwad. Rindal, Magnus. 1988. ‘Bruk av datamaskin i arbeidet med norrøne tekstar.’ In: Fidjestøl, Bjarne, Haugen, Odd Einar & Magnus Rindal (eds), Tekstkritisk teori og praksis. Eit nordisk symposium. Godøysund 19.-22. mai 1987. Oslo: Novus. 46–62. Taranger, Absalon (ed). 1912. Norges gamle love, anden række 1388–1604, første bind 1388–1447, tekst. Christiania: Grøndahl & søn.
Part 2
Standardization and Authenticity
Robert Evans
Of ficial Languages: A Brief Prehistory1
Abstract Historians of politics and administration have largely remained distant from the concerns of historical sociolinguistics, whereas practitioners of the latter have tended either to neglect the politics of language or to use vague and sweeping terms like ‘national’ and above all ‘of ficial’ language. This article examines some cases of how languages came to be formally recognized and regulated by dif ferent states or governments. It argues that the process was much slower than has often been assumed, precisely because languages which were dominant in social and cultural terms long needed no of ficial mandate, especially when mechanisms of authority were comparatively little developed.
Historians always have to do with language; but they have made slow progress towards any real interdisciplinary dialogue with linguists. So far it has been mainly social and literary aspects of their common territory which have attracted attention (Burke, 2004). Nowadays the worlds of (shifting) semantics and discourse are also coming into play more fully – as the present collection will confirm. Yet traditional ‘mainstream’ history, especially of the political-administrative kind, has remained distant. And this distance seems to be mutual, shared by historical sociolinguistics, which either neglects the politics of language or is given to vague and sweeping usage of terms like ‘national’ and above all ‘of ficial’ language (cf. Haarmann 1975: 92–119, a useful compendium for its day, but superficial). That ref lects an underlying incongruity. Languages nowadays represent a major political issue: the formal and prescribed status of one or 1
This text incorporates some material used in Evans (2006) and Evans (2011).
130
Robert Evans
more of them in a given state can be cause for fierce contention. However, they have only recently become so, and particularly in those states with the most powerful languages. Thus France has experienced centuries of constitutions, currently that of the Fifth Republic since 1958; but only on 25 June 1992 was a clause added to assert that ‘la langue de la République est le français’. In the US, English enjoys of ficial status in 34 states and territories, but not in the country as a whole.2 The equivalent situation in the UK is perhaps moot, but the rights of English remain basically unwritten, and the same goes for some other comfortably established European languages too, such as Swedish in Sweden. In the USSR, Russian was a ‘national’ (narodny) and an ‘international’ (mezhdunatsionalny), but not a ‘state’ (gosudarstvenny) tongue. The Italian Costituzione of 1947 (art.3) and the German Grundgesetz of 1949 (art.3, iii) have general statements about the equality of citizens in respect of a list of criteria including language, and about rights for minority languages. In Italy further guarantees for specific minority tongues were introduced in 1999, and eventually in 2007 a measure was proposed to confirm the formal status of Italian itself. In Germany that has not happened – and even the wording of the Grundgesetz largely repeats the relevant paragraph (art. 113) of the Weimar constitution of 1919.3 Is this state of af fairs more than a curiosity, important mainly for current pressure groups, political scientists and the like? Surely modern ‘national’, standard languages have a clear hegemony anyway? And what about the mass of more contingent language regulations over the centuries, designed to serve the same purpose? That, however, is precisely the problem, and the sense of my title for the present essay: the notion of ‘of ficial’ language is de facto readily applied these days in much earlier contexts where it is inappropriate. There are real issues here relevant for 2 3
See accessed 30 May 2011, and links. URLs: ; ; ; ; accessed 15 October 2010.
Of ficial Languages
131
sociolinguistic and other kinds of analysis: precisely why of ficial language – in any proper acceptation – has been a rare, often even a non-existent category until lately. A related distinction needs to be made at the outset between, on the one hand, statute law, and in particular modern constitutions, and, on the other, earlier systems, usually monarchical, of administrational directive: decrees, orders, ordinances, regulations, patents, rescripts, edicts … the list of terms employed is long and (revealingly) confusing. These were products of quite dif ferent and by now unfamiliar forms of executive authority. Not all of them were necessarily even fully public statements of intent. It is correspondingly dangerous to assume that ‘language legislation’ in pre-constitutional systems belonged to any ‘of ficial’ category at all, in our sense. Equally, what we now identify as significant enactments of that sort were not necessarily seen as such at the time, or through the intervening period. In what follows I shall try to do some preliminary mapping, with a concentration on Great Britain and on central Europe, regions with significantly divergent experience in the field. I shall introduce half a dozen brief case studies, some of them familiar. Yet their very familiarity can be perilous! These texts are not straightforward exemplars. And we should note areas where the apparent absence of evidence may be equally telling. Besides, many gaps have still to be filled – for whole vast areas, like Scandinavia, we lack readily accessible information. * It’s a wild but serviceable generalization that we find little sign of ‘of ficial’ action to support language across a swathe of ancient and medieval empires and composite monarchies. Prominent examples would be the largely unenforced spread of Latin through the Roman Empire, and its coexistence with a continuing widespread use of Greek at both popular and elite levels. Then, outside the territories under the cultural domination of Byzantium, medieval Latin took on the role which had previously fallen more to Greek, as the language of civilization (Budinszky 1881, Neumann & Untermann 1980, Richter 1995, and cf. Waquet 1998 for the post-medieval development). Unusually early evidence of suasion seems, by contrast, to come precisely from the precocious statehood of countries in western Europe
132
Robert Evans
whose main languages later didn’t need safeguards. I cite three of the most famous cases, two from the British Isles and one from France, although the caveat just entered will need to be borne in mind. In 1366 the English authorities in an already partially conquered Ireland issued the statutes of Kilkenny. They include decrees whose tenor is clearly dictated by a situation of languages in conf lict. English is set against Gaelic in aggressive fashion, as with condemnation of the fact that many of the local settlers, forsaking the English language [etc.], live and govern themselves according to the language [etc.] of the Irish enemies […] whereby […] the English language [etc.] […] are put in subjection and decayed, and the Irish enemies exalted and raised up, contrary to reason.
It is therefore commanded, among other things, that ‘every Englishman do use the English language’ [etc], and severe punishments are threatened: If any English, or Irish living among the English, use the Irish language among themselves, contrary to this ordinance, and therefore be attainted, his lands and tenements, if he have any, shall be seized into the hands of his immediate lord, [or, if not] his body shall be taken of f by one of the of ficers of our Lord the King, and committed to the next gaol. (Crowley 2000: 14–16)
But we should bear in mind that all this was part of a much longer and broader text, and had an impact only in the Pale of English settlement around Dublin; and that the statute itself was actually issued in French (as were most acts of the Irish parliament till 1472). Moreover, its backdrop lay, as the above extracts imply, in a loss of ground by English locally. That weakness continued in Ireland into and through the sixteenth century. Kilkenny was revoked in its linguistic aspect in 1495, though evidently not because its concerns had been assuaged, since it had to be reimposed a century later. Meanwhile, a series of related enactments were passed to secure or assert English as a tongue requisite for the law, for appropriate forms of schooling, and for grants of urban citizenship (Crowley 2000: 20f f., cf. also Hogan 1927: 15–36 and Rockel 1989: 64–72). Thus the provisions for English in Ireland were still in essence a defensive response.
Of ficial Languages
133
A related instance is the text of the so-called Act of Union, issued in 1536, for the incorporation of Wales into England. Wales, a complex of lordships and jurisdictions in the west of Britain, had in fact developed a more sophisticated vernacular legal tradition in the Middle Ages than England. Yet the English language was already taking over in administration there, alongside Latin, by the fifteenth century (Smith 1997, for the historical context, cf. Williams 1997: 253–278). The 1536 measure ref lected that, particularly in its § 20: Also be it enacted by the authority aforesaid, that all justices, commissioners, [etc., etc.] shall proclaim and keep the sessions, courts [etc.] in the English tongue; and all oaths … and af fidavits [etc.] to be given and one in the English tongue; and also that from henceforth no person or persons that use the Welsh speech or language shall have or enjoy any manner, of fice, or fees within this realm of England, Wales or other the King’s dominion, upon pain of forfeiting the same of fice or fees, unless he or they use and exercise the English speech or language. (Bowen 1908: 87)
We should note that this provision related only to Wales. Although there is mention of England too, it is clearly aimed at Welsh people alone. Remember that England under the Tudors still made much use of law French in its court proceedings, and verdicts even continued to be recorded in Latin till 1732 (Catto 2003). The decision taken in 1536 was connected directly with executive homogeneity, smoothness and ef ficacity, and with judicial reorganization. It confirmed a sociolinguistic process. The ‘Act of Union’ (Deddf Uno), with – inter multa alia – its ‘language clause’ (cymal iaith), as much later nationalists would call it, provoked no significant protest in the country at all, and was actively welcomed by the gentry, attracted by the prospect of equality of status and by the opportunities to be vouchsafed them if they operated in English ( Jenkins and P.R. Roberts in Jenkins 1997: 78f f ). There may have been a culturo-political. agenda. ‘His Highness’ [King Henry VIII], as the Act indicates at another point, ‘minding and intending to reduce [his Welsh subjects] to the perfect order, notice and knowledge of his laws of this his realm’, did undertake ‘utterly to extirp all and singular the sinister Usages and Customs dif fering from the same’ (Bowen 1908: 75). Did such ‘sinister Usages and Customs’ include their Celtic language,
134
Robert Evans
so radically alien to the English eye and ear? If so, however, the agenda was not a consistent one. In 1563, legislation enjoined the production of religious texts in Welsh, building on the claim in the twenty-fourth of the recently promulgated Thirty-nine Elizabethan Articles of faith: ‘It is a thing plainly repugnant to the Word of God, and the custom of the Primitive Church, to have Publick Prayer in the Church, or to minister the Sacraments in a tongue not understanded of the people.’ From that initiative grew the immensely inf luential Welsh Prayer Book and Bible translations, which did so much to sustain the language over the next centuries, for all its subordinate status in worldly af fairs (Williams 1967, Morgan 1988). Exactly contemporary with the legislation for Wales is the best-known of all these directives, the Ordinance of Villers-Cotterêts, issued in 1539. This lengthy document revised procedures of governance within the French monarchy. Towards the end of it, to cover what we would nowadays call matters of ‘transparency’, we find articles 110 and 111: nous voullons et ordonnons que [les arrêts] soient faictz et escriptz si clerement qu’il n’y ayt ne puisse avoir aucune ambiguïté ou incertitude [decrees so clear that there is not and cannot be any ambiguity or incertitude] … Et pour ce que telles choses sont souventes fois advenues sur l’intelligence des motz latins contenuz esd arrestz [understanding of the Latin words contained in the decrees] nous voulons que doresnavant tous arrestz … de nos courtz souveraines [all judgments of high courts] (etc.) soient prononcez, enregistrez et delivrez aux parties en langaige maternel françois et non autrement. (Chaurand 1999: 149)
The earlier part of this justification stresses public ignorance of Latin texts, and hence the scope for their abuse. But the larger issue appears at the end, and has given rise to much debate. Is this ‘langaige maternel françois’ to be equated with the langue du roi, the speech of the Valois court and administration, as long argued – pro and con – by both supporters and detractors of the French state, and by experts like the famous and exhaustive chronicler of French linguistic history, Ferdinand Brunot? Or did it continue to include dialects (previously specified as ‘vulgaires du pays’ or similar) (Brunot 1906: 30–32)?4 4
Cf. also the important qualifications of the traditional view in Fioretti (1950) and Trudeau (1983), as well as the more recent discussion in Babel (2007).
Of ficial Languages
135
Whatever its purpose, Villers-Cotterêts certainly in the event encouraged standard forms of French, which were anyway spreading (Febvre 1924). French still needed its Def fences (of the kind mounted most memorably by Joachim du Bellay); and dialect might still seem a liability for it. Yet clearly it was becoming the language of an increasingly secure culture and of reasonably uniform channels of state power (cf. the useful survey in Schmitt (2000), esp. 682f f.). * In central Europe, later to be the classic linguistic battleground, there were few earlier signs of trouble. Latin and German coexisted for centuries as public media in the Holy Roman Empire, apparently without any serious confusion or friction (Hattenhauer, 1987). Polish underwent a gradual emancipation, from under a shadow cast by both Latin and German; but there, too, clashes remained limited, and mainly took place at a local, municipal level. And the multiethnicity of the Dual Commonwealth of Poland-Lithuania went with a de facto multilingualism which allowed for gradual language shift. The year 1697 saw the final paper confirmation of a sociolinguistic process right across the Rzeczpospolita with the abolition of a residual administrative role for Ruthene in Lithuania and its replacement by Polish.5 Multilingual Hungary and Croatia needed no authoritative assertion of the rights of Latin which for hundreds of years retained, even enhanced, its well-nigh all-embracing functions in public life. The one big exception was Bohemia.6 There already in the high Middle Ages Czech–German discord had a clear linguistic dimension, most famously at the university of Prague, where in 1409 the decree of Kuttenberg, alias Kutná Hora, enforced Czech hegemony. The Hussite wars and the establishment of the Utraquist church then confirmed Czech vis-à-vis both Latin and German as the language of the politically dominant estates and towns, as of non-Catholic religion. All this was progressively inscribed in a series of decrees for the diets, law courts, etc., from the later fifteenth century. But it was also accompanied by a defensive strategy for 5 6
Cf. Evans (2006) for further references. For the details of what follows, see Evans (2011) and references there.
136
Robert Evans
this ostensibly beleaguered Slav culture, increasingly so in the sixteenth century (contrast the comparative self-assurance of Polish-speakers). Hence, alongside much evidence of coexistence in Bohemia, we have some notable private and public statements of linguistic concern, from the religious leader Jan Blahoslav in 1571 to the oppositional commentators Pavel Ješín and Pavel Stránský at the time of the revolt against the Habsburgs which broke out in 1618. Above all we have the evidence of a strange and perhaps unique law of 1615, ‘On the Preservation and Cultivation of the Ancient Czech Language’ [O zachování starožitného jazyka českého a vzdělání jeho]. This bemoans, inter alia, that the extinction of the Czech language would bring with it also the extinction of the Czech nation […] introducing all manner of foreign tongues and nations […] more and more foreigners continually enter the country […] many of them unable to string together three words of Czech […] foreign priests with no Czech are installed and maintained […].
As remedies it enacts that such foreigners should forthwith be obliged to have their children learn Czech from their youth […]. [H]eirs and heiresses […] who can speak the Czech language well shall inherit landed property […] in double measure […]. [N]o foreigner who does not know Czech and cannot intelligibly present his needs in that tongue shall by any means be accepted as a denizen of the country or as citizen of a town […]. [N]o such foreigner newly accepted into the country nor his children to the third generation shall be preferred to any state or municipal or other of fice, or to any courts of law [etc., etc.]. (Skála 1865–1870: 355–358)
These weird and unenforceable stipulations were evidently never implemented – and anyway the Thirty Years War soon intervened. In fact the fears underlying this outburst did come to be realized after the defeat of the Bohemian revolt in 1620. But the decline of Czech was not accomplished by legal means (the revised land ordinance, or constitution of the day, merely placed German alongside it as a language of state). Altogether there appears to be little sign of language ‘legislation’ through the ancien regime Europe of that and the succeeding period. Further examples of active intervention might be found in sixteenth-century
Of ficial Languages
137
Savoy, where promotion of the vernacular ran pari passu with what we have already examined in France, and seventeenth-century Sweden, where the introduction of Swedish into Skåne, once that province was conquered from the Danes, is cited as a precocious campaign of linguistic education (although both standard tongues may have been equally remote from the dialect of the indigenous population). Yet such state-led endeavours were typically ad hoc and localized, at best: witness the decrees of the Bourbon kings requiring newly acquired territories to be administered in French (Babel 2007). Only in the eighteenth century did things begin to accelerate. One early and distinctive case was another act of Bourbon policy: that of the junior branch which after 1700 won control of Spain by military force. Its tough Decretos de Nueva Planta (1716) enforcing the ‘lengua Castellana’ for courts etc. in defeated Catalonia inaugurated an irregular series of measures which buttressed the legal status of language in Spain down to our own day.7 Change was, however, mainly associated with reformist absolutisms post-1750, and it remained piecemeal and inconsistent. Emperor Joseph II, for instance, gave support to vernaculars, for the instruction of the people, alongside the ancillary linguistic aspects to his headlong initiatives for centralizing and homogenizing his realms.8 I shall return to him shortly, in the context of Hungary. * In the history of of ficial languages, as in so much else, the French Revolution played a transformative role, and one which is now better attended to in the literature (cf. Certeau et al. 1975, Renzi 1981, Bell 1995, Bell 2001). It invented the modern written European constitution (alongside the abortive 7
8
Decrees reproduced at [last accessed 30.5.11]. The Spanish constitution of 1931 already prescribed that ‘el castellano es el idioma of ficial de la República. Todo español tiene obligación de saberlo y derecho a usarlo’, a wording substantially repeated in 1978. See Lexikon der roma nistischen Linguistik (368–378). For the vicissitudes of Catalan, see Ferrer i Gironès (1985). I am very grateful to Rodney Sampson (Bristol) for help with Spanish. Much evidence passim in Derek Beales (1987–2009).
138
Robert Evans
Polish equivalent). Some revolutionaries – above all the doctrinaire Abbé Grégoire – pressed for the exclusive claims of the ‘langue nationale’. This was soon abandoned (and we should bear in mind that the Revolution also temporarily discontinued the Académie Française, which had established criteria for that language). But it presaged the immense post-revolutionary and Napoleonic centralizing pressures for standard French and the whole nineteenth-century normalizing experience described according to the ‘Weber thesis’ as turning ‘peasants’ (and others) into (linguistic) ‘Frenchmen’ (Weber 1976). The impact of these developments on ‘of ficial’ language policy was slower and more indirect. The French model of the nation, combined with the enthusiasms of the Romantic age, tended to revive and assert minority as well as majority cultures on an ethnic base. The resultant f lashpoints lay above all in the linguistic congeries of central Europe. In two important cases this yielded explicitly multilingual stipulations. One was Switzerland, where – it is frequently overlooked – until 1798 all the full cantons had used German, so the issue was not formally raised earlier (cf. Arquint et al. 1982). Even the expanded post-Napoleonic Confederation did not mention language in its constitution. Only in 1848 did the next such document, ref lecting the liberal takeover of that year, declare that ‘[t]he three chief languages of Switzerland, German, French and Italian, are the national languages of the Confederation’.9 Still nothing further; and even this article stood as number 109, under ‘miscellaneous provisions’, after long sections on political, military, commercial, religious, legal, and institutional matters. The same form of words would pertain until 1938, when Rhaetoroman was added, while the first three tongues gained formal ‘administrative’ status, as Amtssprachen too. Austria, under its own emperors from 1804, had a similar tradition of German linguistic dominance; but there no language commanded a majority of speakers, and a wide range of other vernaculars existed. The one serious attempt at ‘Germanizing’ proved a short phase during the 1850s, and since the government was then busy promoting a rhetoric of 9
‘Die drei Hauptsprachen der Schweiz, die deutsche, französische und italienische, sind Nationalsprachen des Bundes’: see article 109 in www.verfassungen.de/ch/ verf48-i.htm [last accessed 13.10.2010].
Of ficial Languages
139
multinational harmony, it elicited no clear public statements. Then in the 1860s the Habsburg rulers were forced into granting definitive autonomy to Hungary and concessions to the other nationalities in the rest of the Monarchy. Hence the Austrian constitutional provision of 1867 to the ef fect that [a]ll the peoples [Volksstämme] of the state have equal rights, and every Volksstamm has the inviolable right to preserve and cultivate its nationality and language. The equality of all customary languages in school, of fice and public life is recognized by the state.10
This brief but portentous statement ef fectively came to ensure a multiplicity of languages of state – and it also went with a parallel guarantee of the freedom of all citizens from any obligation to learn a second language. For all the protests of nationalists, neither then nor later did German become any more of ficial than the other vernaculars; though the stenographic practice of the Reichsrat, Austria’s parliament, of not recording any speeches delivered in those vernaculars, pointed up the continuing practical hegemony of German in the public life of the state. The clearest and most detailed example known to me of the making of an of ficial language is the nineteenth-century evolution in the other half of the Habsburg Monarchy: Hungary. That country saw the longest survival anywhere in Europe of historic bipolar (ruler and estates), corporative arrangements of rule, and without a measurable language problem despite a range of spoken tongues as great as in Austria. The much-interpreted maxim of Hungary’s first king, St Stephen, that ‘regnum unius linguae uniusque moris imbecille et fragile est’, was at least ben trovato for this situation, where Latin held its place in administrative functions, and in mediating relations with the largely absentee monarchs, while the vernaculars, some more international and some more local, all operated concomitantly (cf. Evans 2007 for an overview). 10
‘Jeder Volksstamm hat ein unverletzliches Recht auf Wahrung und Pf lege seiner Nationaltät und Sprache. Die Gleichberechtigung aller landesüblichen Sprachen in Schule, Amt und öf fentlichem Leben wird vom Staate anerkannt’: Bernatzik (1906: 370). See Stourzh (1985) for an authoritative commentary and Fischel (1910) on nineteenth-century Austrian language legislation in general.
140
Robert Evans
Everything quickly began to change, for languages too, with the challenge in the 1780s from a radically reformist ruler, Joseph II, who prescribed German henceforth for all administrative purposes. His measure, although short-lived and futile, yielded a constitutional and then a romantic backlash. This provoked a singularly protracted, concentrated and virulent campaign for legislation in favour of Magyar, the best-established of Hungary’s spoken languages, which delivered a stream of enactments initiated in 1792 and completed by 1844.11 It simultaneously fuelled a secondary backlash from the country’s other nationalities: resistance at the diet mainly from Croat deputies (who regularly protested in Latin!); and beyond the diet a pamphlet war, with supplications and complaints about ‘Magyarization’. What limited ef fective Magyarization there was in fact proceeded organically and peacefully; whereas the public and of ficial measures in its favour contributed to widespread unrest and then in 1848–1849 to a civil war which brought temporary nemesis to the Magyar camp. After the hiatus of the 1850s, further extended debates took place, in counterpoint with the accomplishment of Hungarian self-government. They resulted by 1868 in what is described as a ‘nationalities law’ (törvény a nemzetiségekről), in fact the most elaborate language legislation of the whole period. It starts with an important preamble on what constitutes the ‘of ficial’ sphere: Whereas all citizens of Hungary, according to the basic principles of the constitution, form one nation in political terms, the indivisible unitary Hungarian [magyar] nation, of which every citizen of the country, whatever nationality [nemzetiség] he may belong to, is an equal member; and whereas, further, that equality can only be liable to specific regulation in respect of the of ficial usage of the various languages prevailing in the country, and in so far as the unity of the country, the practical scope of government and administration, and the precise dispensation of justice require … [we therefore enact as follows …]12
11 12
Full documentation in Gy. Szekfű (1926) and in a number of brilliant works by Daniel Rapant, of which the first, Rapant (1927–1931) sets the scene in virtuoso fashion. ‘Törvénycikk a nemzetiségi egyenjogúság tárgyában: Minthogy Magyarország összes honpolgárai az alkotmány alapelvei szerint is politikai tekintetben egy nemzetet képeznek, az oszthatlan egységes magyar nemzetet, melynek a hon minden polgára, bármely nemzetiséghez tartozzék is, egyenjogú tagja;
Of ficial Languages
141
Magyar was reasserted as ‘language of state’ (államnyelv); and precise provision made for other vernaculars in (lower levels of ) administration, education, and the judiciary. Minorities qualified for such provision if they formed at least 20 per cent of the population locally; they could draw on translation facilities and interpreters, keep records in their own tongue; etc. Much of this was merely a paper exercise. The regulations were not properly enforced – a consideration we must always bear in mind when assessing ‘of ficial’ sources. But anyway the non-Magyar leaders rejected the law’s assumptions from the beginning and sought to achieve something much closer to the Austrian solution which we have already encountered.13 In the end their attitudes contributed greatly to the breakdown of historic Hungary after the First World War. * Ironically the Hungarian legislation of 1868 served as a model for some of the many new enactments after 1918 across central and eastern Europe which at last did place language firmly in the of ficial sphere. Czech, for instance, was now restored, after exactly 300 years, to pre-eminence in the new Czechoslovakia, even if it shared its place with its less well-versed sister-tongue of Slovak (Kučera, 1999).14 In their turn, the post-war settlements in Europe established a yardstick for the formal rights of languages, at least those with state backing, across more and more of the world. Why did of ficial languages have such a long prehistory? A few conclusions may seem to grow out of the very cursory examination of fered here.
13 14
minthogy továbbá ezen egyenjogúság egyedül az országban divatozó többféle nyelvek hivatalos használatára nézve, és csak annyiban eshetik külön szabályok alá, a men�nyiben ezt az ország egysége, a kormányzat és közigazgatás gyakorlati lehetősége s az igazság pontos kiszolgáltatása szükségessé teszik …’: Magyar törvények, 1865–8 (Pest, 1869), 151 [my translation]. For the debates surrounding this cryptic and unwieldy formulation, see Schlett (2002). For the wider context, cf. Evans (2004). There is no proper study; but many clues in Kemény (1952–1985). The peace treaty spoke of a ‘langue tchèque’; but the constitution identified the language of state as ‘Czechoslovak’, in its two varieties (znění), each of which was de facto of ficial in one of the two halves of the country. For the 1920 laws and their implementation, see Kučera (1999).
142
Robert Evans
It was, of course, the development of administrative organs and the spread of literacy which first generated the requirements for of ficial prescriptiveness about language. So did the rise of formal institutions needing a (single) medium of communication, particularly diets, parliaments, and the like. Yet the processes were slow-moving, and they did not invalidate surviving modes of more casual and unstudied negotiation. Slow too was the emergence of language as a mark of prestige and badge of allegiance. That went with the gradually changing role and priorities of multilingual elites, and their interaction with increasingly prominent, and maybe increasingly mobile, monoglot social groups. Moreover, a key underlying factor in rendering language an of ficial concern has repeatedly been a sense of threat and vulnerability: from the time of the Kilkenny statutes and the Bohemian patriots of 1615, even to recent reactions to the perceived challenge from English – or franglais – in France, or Spanish in the US. So has my theme been ultimately a kind of non-subject, the more so since some of the most celebrated examples of earlier language enactments may (as we have seen) not bear the weight of all that was subsequently attributed to them? No, because the episodes we have observed were significant ones, ripe to be built upon in later rhetoric. Fuller study of them would throw up much evidence of the actual workings of both of ficial and unof ficial channels of linguistic control and direction. In the end they all have to do with power, with its longer-term, subtler manifestations in that public sphere which language always occupies, and thus with political and cultural kinds of interplay between individuals and groups. As such they make a claim on the attention of historians and linguists alike.
References Arquint, Jachen C. et al. 1982. Die viersprachige Schweiz. Zurich: Benziger. Babel, Rainer. 2007. ‘Sprache und Politik im Frankreich der Frühen Neuzeit. Eine Bestandsaufnahme.’ In: Thomas Nicklas & Matthias Schnettger (eds), Politik und Sprache im frühneuzeitlichen Europa. Mainz: von Zabern, 33–50.
Of ficial Languages
143
Beales, Derek. 1987–2009. Joseph II. Vol. I: In the Shadow of Maria Theresa, 1741–80. Vol. II: Against the World, 1780–90. Cambridge: Cambridge University Press. Bell, David A. 1995. ‘Lingua Populi, Lingua Dei: Language, Religion and the Origins of French Revolutionary Nationalism.’ American Historical Review 100, 1403–1437. Bell, David A. 2001. The Cult of the Nation in France: Inventing Nationalism, 1680–1800. Cambridge, MA: Harvard University Press. Bernatzik, Edmund. 1906. Die österreichischen Verfassungsgesetze. Leipzig : Hirschfeld. Bowen, Ivor. 1908. (ed.). The Statutes of Wales. London. Brunot, Ferdinand. 1906. Histoire de la langue française des origines à 1900. Vol. II: Le seizième siècle. Paris: Armand Colin. Budinszky, Alexander. 1881. Die Ausbreitung der lateinischen Sprache über Italien und die Provinzen des römischen Reiches. Berlin: Wilhelm Hertz. Burke, Peter. 2004. Languages and Communities in Early Modern Europe. Cambridge: Cambridge University Press. Catto, Jeremy. 2003. ‘Written English: The Making of the Language, 1370–1400’. Past & Present 179, 24–59. de Certeau, Michel et al. 1975. Une politique de la langue: la Révolution française et les patois. L’enquête de Grégoire. Paris: Gallimard. Chaurand, Jacques (ed.). 1999. Nouvelle histoire de la langue française. Paris: Éditions du Seuil. Crowley, Tony (ed.) 2000. The Politics of Language in Ireland, 1366–1922. A Sourcebook. London: Routledge. Evans, Robert J.W. 2004. ‘Language and State-building: The Case of the Habsburg Monarchy.’ Austrian History Yearbook 35, 1–24. Evans, Robert J.W. 2006. ‘The Politics of Language in Europe, c. 1525–1697.’ Przegląd Historyczny 97, 455–476. Evans, Robert J.W. 2007. ‘The Politics of Language and the Languages of Politics: Latin and the Vernaculars in Eighteenth-Century Hungary.’ In: Hamish Scott & Brendan Simms (eds), Cultures of Power in Europe during the Long Eighteenth Century. Cambridge: Cambridge University Press, 200–224. Evans, Robert J.W. 2011. ‘Language and Politics: Bohemia in International Context, 1409–1627.’ In: Eva Doležalová and Jaroslav Pánek (eds), Confession and Nation in the Era of Reformations. Central Europe in Comparative Perspective. Prague: Institute of History, 155–182. Febvre, Lucien. 1924. ‘Politique royale ou civilisation française? Remarques sur un problème d’histoire linguistique.’ Revue de Synthèse Historique 38, 37–53.
144
Robert Evans
Ferrer i Gironès, Francesc. 1985. La persecució política de la llengua catalana. Barcelona: Edicions 62. Fioretti, P. 1950. ‘L’ordonnance de Villers-Cotterêts.’ Le Français Moderne 18, 277–288. Fischel, Alfred (ed.). 1910. Das österreichische Sprachenrecht: Eine Quellensammlung. Brünn: Irrgang. Haarmann, Harald. 1975. Soziologie und Politik der Sprachen Europas. Munich: dtv. Hattenhauer, Hans. 1987. Zur Geschichte der deutschen Rechts- und Gesetzessprache. Göttingen: Vandenhoeck & Ruprecht. Hogan, Jeremiah J. 1927. The English Language in Ireland. Dublin: The Educational Company of Ireland. Jenkins, Geraint H (ed.). 1997. Y Gymraeg yn ei Disgleirdeb: Yr Iaith Gymraeg cyn y Chwyldro Diwydiannol [The Welsh Language before the Industrial Revolution]. Cardif f : Gwasg Prifysgol Cymru. Kemény, Gábor G. (ed.). 1952–1985. I����������������������������������������������� ratok a nemzetiségi kérdés történetéhez Magyarországon a dualizmus korában. 7 vols. Budapest: Tankönyvkiadó. Lexikon der romanistischen Linguistik. 1988–2005. ed. by Günter Holtus et al. 8 vols. Tübingen: Niemeyer. Kučera, Jaroslav. 1999. Minderheit im Nationalstaat: die Sprachenfrage in den tschechisch–deutschen Beziehungen, 1918–38. Munich: Oldenbourg. Morgan, Prys. 1988. Beibl i Gymru. Gwasg Cambria. Neumann, Günter, & Jürgen Untermann. 1980. Die Sprachen im römischen Reich der Kaiserzeit. Cologne: Rheinland Verlag. Rapant, Daniel. 1927–1931. K počiatkom maďarizácie. 2 vols. Bratislava: Filozofická fakulta Univerzity Komenského. Renzi, Lorenzo. 1981. La politica linguistica della Rivoluzione francese: studio sulle origini e la natura del giacobinismo linguistico. Naples: Ligouri. Richter, Michael. 1995. Studies in Medieval Language and Culture. Blackrock/Dublin: Four Courts Press. Rockel, Martin. 1989. Grundzüge einer Geschichte der irischen Sprache. Vienna: Österreichische Akademie der Wissenschaften. Schlett, István (ed.). 2002. A nemzetiségi törvényjavaslat országgyűlési vitája, 1868 Budapest: Kortárs Kiado. Schmitt, Christian. 2000. ‘Nation und Sprache: das Französische’. In: Andreas Gardt (ed.), Nation und Sprache. Die Diskussion ihres Verhältnisses in Geschichte und Gegenwart. Berlin: de Gruyter. 673–745. Skála ze Zhoře, Pavel. 1865–1870. Historie česká, od r. 1602 do r. 1623, ed. Karel Tieftrunk. 5 vols. Prague: Nakl. Kněhkupectví I.L. Kober. Smith, Llinos Beverley. 1997. ‘Yr Iaith Gymraeg cyn 1536.’ In: Jenkins, 15–44.
Of ficial Languages
145
Stourzh, Gerald. 1985. Die Gleichberechtigung der Nationalitäten in der Verfassung und Verwaltung Österreichs, 1848–1918. Vienna: Österreichische Akademie der Wissenschaften. Szekfű, Gyula (ed.). 1926. Iratok a magyar államnyelv kérdésének történetéhez, 1790– 1848. Budapest: Magyar Történelmi Társulat. Trudeau, Danielle. 1983. ‘L’ordonnance de Villers-Cotterêts et la langue française: histoire ou interpretation.’ Bibliothèque d’Humanisme et Renaissance 45, 461–472. Waquet, Françoise. 1998. Le latin ou l’empire d’un signe, XVIe–XXe siècle. Paris: Albin Michel. Weber, Eugen. 1976. Peasants into Frenchmen: The Modernization of Rural France, 1870–1914. Stanford: Stanford University Press. Williams, Glanmor. 1967. Welsh Reformation Essays. Cardif f : University of Wales. Williams, Glanmor. 1987. Recovery, Reorientation and Reformation: Wales, c.1415–1642. Oxford: Oxford University Press.
Tomasz Kamusella
Classifying the Slavic Languages, or the Politics of Classification1
Abstract At present it is often accepted that the Slavic languages are divided into three subfamilies: Western, Eastern and Southern. However, up to the turn of the twentieth century other classifications had prevailed before this tripartite division became established as the standard model. The taxon of subfamily in the so-called ‘genetic’ classification of languages is generally applied to groups of interrelated languages that are spoken in a contiguous area (dialect continuum) and that are geographically isolated from other ‘kin’ subfamilies by non-related languages. This, however, is not true of the Western and Eastern Slavic languages, which are separated from one another not on any areal or linguistic basis, but for political-cum-cultural reasons (such as distinctions among populations professing Catholicism or Orthodox Christianity, use of the Latin script versus Cyrillic, being part of the Russian Empire or not). Linguists tend to support this genetic classification (perhaps under the inf luence of political considerations), while other social scientists accept the universal applicability of this standard genetic model because of a mistaken belief that it is grounded in the findings of linguistic research.
1
The text gained much from corrections and suggestions of the two anonymous readers, whom I thank for their help. Another word of thanks goes to Michael O’Gorman for reading a version of the text. Obviously, I am responsible for any remaining infelicities.
148
Tomasz Kamusella
1 Introduction In this article I analyse the emergence and configuration of various classificatory schemes of the Slavic languages, which were devised mainly during the nineteenth century. Next, I trace the way in which one of these schemes, the triple one, became the current standard. I propose that the standard’s distinction between the Western and Eastern Slavic languages is based on extralinguistic factors. Both groups of these languages gradually shade from one into another within the North Slavic dialect continuum, meaning that from the linguistic perspective, they could be usefully classified as a single, Northern, branch of the Slavic languages. In premodern and early modern Europe legitimizing histories and myths which gave larger, usually politically significant, groups of people rights to their status, to their territory or to other privileges, were grounded in religion or in the mythical past, usually connected to the Roman Empire, heroic Antiquity or the Bible. The search for the lost tribes of Israel or declaring this or that group descended from Brutus or another Roman aristocrat were accepted instruments of political and scholarly discourse. A change came with the development of historiography as a discipline which should formulate its theses on the basis of facts, substantiated most often by written records. This rationalistic rigor within the domain of historiography has not led to the relegation of politically favored myths and legends within public discourse altogether to the realm of fairy tales. Politically motivated myths pertaining to the past still occupy a large space in the collective imaginations of nations; most notably, they are propagated through schools to future citizens. In the case of Poland, the sixteenth-century myth of the separate ethnic, Sarmatian, origins of the nobility in the Kingdom of Poland is no longer seriously advanced. This wholly unsubstantiated myth may have been shed because it clashed with the newer political needs of the nationalists at the turn of the twentieth century who wished to fashion a Polish nation from all those perceived as the speakers of the Polish language, extending to include the peasantry. The nobles had tended to perceive their serfs as dif ferent not only in status, but also in origin, by claiming that peasants
Classifying the Slavic Languages, or the Politics of Classification
149
descended from the Biblical figure Ham, or in a more rational vein, that they were of Slavic origin, and thus dif ferent from and more vulgar than the nobles who claimed descent from the glories of Sarmatia. The origins of the Poles, the Czechs and the Russians are popularly explained in a late thirteenth-century tale of three brothers, Lech, Czech and Rus, eponyms of the three nations. The legend achieved canonical status in Poland in the late nineteenth century and is still told in Poland to pupils in their early years at elementary school. The tale does not mention any further brothers, representing the Belarusians or the Ukrainians, because when it coalesced, the ethnonym Rus (or rather Rus’), or Ruthenians in English, referred to all of the Slavophone Orthodox population, from which the present-day nations of the Belarusians, the Russians and the Ukrainians later emerged. This story about the three brothers purportedly explains, on the one hand, the common, Slavic origin of the three nations, and on the other, how they became dif ferentiated into three separate nations. The tale’s focus on ethnolinguistic closeness and dif ference is a ref lection of the rise of Pan-Slavism and ethnolinguistic Slavic nationalisms as increasingly potent political ideologies in the second half of the nineteenth century. These were preoccupations of the times, which gradually erased previously dominant ideas of the divine legitimization of temporal power. In Central Europe this was replaced with ethnolinguistic nationalism, and the entailed idea of the ethnolinguistically homogenous (or ‘pure’ in nationalist parlance) nation-state. Ethnolinguistically based nationalism could be presented to the population at large as ‘modern’ or ‘progressive,’ not only because it allowed for changing the old order, but also because beginning in the nineteenth century, European scholars made the ‘genetic’ relations between languages an accepted component of scholarly discourse, which later philologists, often without much substantiation, claimed as ‘scientific’. In 1770 János Sajnovics (1733–1785) noticed a genetic link between Hungarian and Saami (Lapp), and sixteen years later was followed by William Jones (1746–1794) who demonstrated the existence of a similar link connecting Sanskrit to Latin. This led to the rise of historical linguistics, which became a significant tool for both legitimizing and disputing the ethnolinguistic claims of national and pan-national movements. Moreover, Johann Gottfried
150
Tomasz Kamusella
Herder’s (1744–1803) opinions on language, languages, cultures and ethnic groups, culminating in Ideen zur Philosophie der Geschichte der Menschheit (1784–1791, Outlines of a Philosophy of the History of Man, 1800) presented the national movements of the nineteenth century with a ready-made corpus of apparently incontrovertible ideas and ‘facts’ for furthering and legitimizing their pet political projects. From that moment on, especially in Central Europe in the wake of the Napoleonic Wars, it was no longer suf ficient for a political project to rely on the long-lasting existence of a given state. The state’s population, in order to qualify as a nation, had to enjoy a ‘scientifically proven’ common ethnolinguistic origin and ideally this should now be bolstered by the adoption of a single standard form of a common national language. Ernest Gellner (1925–1995) memorably defined this ethnolinguistic variety of nationalism as the drive toward attaining congruence between the cultural unit and the political unit, that is, that an ethnolinguistic nation should be housed in its own nation-state, not shared by any other nation (Gellner 1983: 1). This tendency delegitimized the plethora of polities on the territory of the former Holy Roman Empire and in the Apennine Peninsula; these polities were perceived as being inhabited by German- and Italian-speakers, respectively, and this led to the founding of the Kingdom of Italy (1861) and of the German Empire (1871) as ethnolinguistic nationstates-in-making. Furthermore, the non-ethnolinguistic Austrian Empire, in an ef fort to arrest the fraying of its statehood legitimization which was being increasingly undermined by the rise of ethnolinguistic nationalism, surrendered strategically to this new ideology by transforming itself into Austria-Hungary and facilitating the emergence of variously defined and organized ethnolinguistic autonomies at the sub-state (regional) level.
2 Toward a science of language Philologists in the second half of the nineteenth century aspired to elevate their preoccupation with texts, grammars, translations and classifications to the status of a science of the same rank of exactness as the natural sciences.
Classifying the Slavic Languages, or the Politics of Classification
151
The methods of historical linguists were speculative and only narrowly applicable. An unexpected source of prestigious hard scientific support for their ef forts came from contemporary biology. In chapter fourteen of his On the Origin of Species (1859), Charles Darwin (1809–1882) proposed that his concept of the ‘tree of life,’ intended to illustrate and explicate the process of speciation, could be employed for explaining the rise and dif ferentiation of languages. Darwin suggested that ‘[i]f we possessed a perfect pedigree of mankind, a genealogical arrangement of the races of man would af ford the best classification of the various languages now spoken throughout the world; and if all extinct languages, and all intermediate and slowly changing dialects, were to be included, such an arrangement would be the only possible one’ (cited in Cavalli-Sforza 1994: 381–382). Having been introduced to Darwin’s theory by his friend, the Prussian naturalist Ernst Haeckel (1834–1919), the linguist August Schleicher (1821–1868) from Saxe-Meiningen wrote Die Darwinsche Theorie und die Sprachwissenschaft (1863, Darwinism Tested by the Science of Language, 1869). Schleicher made the image of language as an ‘organism’ into one of the most lasting metaphors in the popular thinking on language. Therefore, most of us now tend unref lectively to believe that languages are ‘born’, ‘grow’ and ‘die’ as though they were species of animals or plants, not phenomena solely dependent on the species of Homo sapiens sapiens. Schleicher also borrowed the Linnaean system of biological classification and used it for classifying these ‘organisms of languages’ (Sprachorganismen) (Maher 1983; Richards 2004: 13; Schleicher 1869: 13). Schleicher developed Darwin’s tentative and diagram-like tree of life concept (with no actual data on the tree), into a full-f ledged ‘family tree’ (Stammbaum) of the Indo-European languages. The British theological writer Frederic William Farrar (1831–1903) extended this device and the biologizing approach to language classification in his Families of Speech (1870). Haeckel in numerous inf luential works adopted as a concrete and uncontested description of reality this initially only heuristically intended device, frequently drawn in the form of a literal tree to classify plants, animals and ‘species [that is, races] of man’ (Alter 1999: 75, 92, 112–121). In this manner, in the twentieth century the image of ‘family tree’ entered textbooks and the popular imagination as the main tool for classifying, understanding and ordering various social and historical phenomena in
152
Tomasz Kamusella
time and space. This, however, presupposes both the discrete nature of such phenomena and their causal linking in accordance with the linear model of time. Despite the widespread dissemination of the idea of the genealogical tree, linguists realize that languages cannot be the unambiguously delineated entities of popular belief. Languages are the creation of human beings using them for communication. Languages interact, appear, disappear and shade into one another, ref lecting the spatial and social mobility of human groups and their interactions with one another. The Thuringian pioneer of creole linguistics, Hugo Schuchardt (1842–1927), in his Der Vokalismus des Vulgärlateins (1868, Vocalism of Vulgar Latin) and the Prussian linguist, Johannes Schmidt (1843–1901), in his Die Verwandtschaftsverhältnsse der indogermanischen Sprachen (1872, The Relationships among the Indo-European Languages) independently proposed the metaphor of the ‘wave’ (Welle) for classifying languages. This was the beginning of areal linguistics, which researches mutual inf luences between languages spoken side by side in neighboring areas or within a single area. This approach to classification allowed for the grouping of languages into non-genealogical, variously named, ‘linguistic areas,’ ‘linguistic leagues,’ or ‘linguistic unions’ (Sprachbünde). But the fact that terminology for referring to these kinds of linguistic relationships still has not stabilized in English indicates that it remains a specialist field, whose insights have not yet extended into popular knowledge (Busmann 1996: 519; Décsy 1973). The image of the ‘genealogical tree’ is more attractive to the average language user, due to its anthropomorphic quality (speaking of ‘parent’ and ‘of fspring’ languages), and the prevalent treatment of languages as discrete entities that parallels popular thinking on nations and states as similarly discrete entities. Nationalism became the first globally accepted ideology of statehood legitimization in the wake of decolonization and the breakup into nation-states of the Soviet Union, the last large and emphatically non-national polity (Kohn 1962). Thus, at the beginning of the twentyfirst century we tend to believe that humankind is ‘naturally’ divided into nations which should be housed separately in their own nation-states. In Central and Eastern Europe where ethnolinguistic nationalism is dominant,
Classifying the Slavic Languages, or the Politics of Classification
153
language has been seen as the defining basis of the nation. This type of nationalism defines the nation as all speakers of a language. This ideological principle reinforces even more the fallacious belief in the discrete character of languages, to the studious disregard of dialects, of multilingualism and of the fact that languages continuously change through time and interact with other languages, following their speakers’ social, economic, political and cultural choices and interactions (Kamusella 2006). Not surprisingly, the classification of languages into linguistic areas, which transcends and blurs political boundaries, and which lets one see one’s speech in various degrees of closeness to the standard of one’s national language as well as to national languages in the adjacent nation-states, even if they are not ‘genetically’ related to one’s language, is an anathema to the national vision of the world. Hence Schleicher’s admonition that ‘[w]hat some call a language, others term a dialect, and vice versa’ (1869: 46) is conveniently forgotten. In our national world of today, there is a preference for Leonard Bloomfield’s postulate that dialects are mutually comprehensible while languages are not (1926: 162). The mutual comprehensibility of such languages as Bosnian, Croatian, Montenegrin or Serbian is conveniently overlooked, as, for similar reasons of political interest, is the mutual incomprehensibility of the dialects of Chinese. Sometimes even professional linguists (especially in Central and Eastern Europe, where their research is expected to uphold and reconfirm the enshrined tenets of ethnolinguistic nationalism) brush aside such discrepancies between the ideologically determined stance and the evident factual position. They fail to or pretend not to see themselves as part of a political game of nationalism and power struggle to which they contribute with their (ostensibly scientific and value-free) pronouncements and theses on language.
154
Tomasz Kamusella
3 Slavic languages: Toward the tyranny of genealogical classification 3.1 A dual classification The study of Slavic languages as a branch of philology emerged at the turn of the nineteenth century, most obviously represented by the works of the Bohemian (Czech) scholar Josef Dobrovský (or, as he himself spelled it in his signature: Dobrowsky, 1753–1829). He wrote in German but spoke Czech and knew other Slavic languages. Between 1806 and 1822 he developed a dual classification of the Slavic languages. The first group comprised Russian, Old Church Slavonic, Illyrian (used in Bulgaria, Serbia, Bosnia and Dalmatia), Croatian and Windisch (Slovenian). In the second group Dobrovský included: Slovak, Czech, Upper Lusatian (Upper Sorbian), Lower Lusatian (Lower Sorbian) and Polish (Ottův slovník naučný 1901: 448). Table 2 Dual classification of Slavic Group 1
Group 2
Russian Old Church Slavonic Illyrian (used in Bulgaria, Serbia, Bosnia and Dalmatia) Croatian Windish (Slovenian)
Slovak Czech Upper Lusatian (Upper Sorbian) Lower Lusatian (Lower Sorbian) Polish
He may have distinguished the former group because the Orthodox, Roman Catholic, Greek Catholic (Uniate), and Muslim speakers of the aforementioned vernaculars either used non-Slavic languages or Church Slavonic for of ficial and written purposes. The standardization of Russian was then already under way,2 but the political will to impose it throughout the Rus-
2
Russian as a standard language began to emerge in the second half of the eighteenth century from the mixture of the Moscow dialect and Church Slavonic, or the ‘middle style’ as proposed by Mikhail Lomonosov (1711–1765) in 1755.
Classifying the Slavic Languages, or the Politics of Classification
155
sian Empire did not appear until after the Polish-Lithuanian nobility’s failed uprising against the Tsar (1830–1831). Russian then replaced Polish as the of ficial language in those Polish-Lithuanian provinces that had been directly incorporated into the empire at the close of the eighteenth century. Furthermore, apart from Russian, all the vernaculars from Dobrovský’s first group were believed to have derived from Old Church Slavonic, while Russian and the languages spoken by Orthodox Slavs functioned in a diglossic symbiosis with Church Slavonic. Dobrovský’s second group of Slavic languages included those which were spoken exclusively by either Roman Catholics or Protestants, and which had been used in writing since the sixteenth century (with the partial exception of Slovak, which became more generally perceived as separate from Czech only in the second half of the nineteenth century). None of these languages originated from Church Slavonic or was symbiotically connected to it. When Dobrovský put forward his classification, of the languages in Group 2, only Polish was used in any of ficial capacity, though merely at a substate level. 3.2 A triple classification The ethnically Slovenian linguist in Austrian service, Bartholomäus ( Jernej) Kopitar (1780–1844), and the Russian Slavicist, Aleksandr Vostokov (1781– 1864), disagreed with Dobrovský’s dual classification. Today’s triple classification of the Slavic languages dates back to Vostokov’s work Rassuzhdene o tserkevno-slavianskom iazyke (1820, An Essay on the Church Slavonic Language). He insisted that Russian did not belong to either of Dobrovský’s two groups of the Slavic languages, but that it constituted a transitional (though in its own right distinctive) bridge between the Western and South-Eastern Slavic languages. He also argued that Russian was closer to the South-Eastern Slavic languages than to the Western Slavic languages. It is likely that the tradition of Church Slavonic literacy and the use of Cyrillic common to the South-Eastern Slavs and to the Russians convinced him of the existence of a special closeness between these two groups of Slavs and their languages (Ottův slovník naučný 1901: 448): Western Slavic languages – Russian [Eastern Slavic lgs?] – South-Eastern Slavic languages
156
Tomasz Kamusella
3.3 An elaborated dual classification The Czech philologist of Slovak origin, Pavel Josef Šafařík (1795–1861, spelled Šáfarik in modern Slovak), in his opus magnum, Slowanské starožitnosti (1837, Slavic Antiquities), supported Dobrovský’s classification. In his scheme Russian, Bulgarian and Illyrian belonged to the SouthEastern branch of the Slavic languages. The Illyrian language was composed from Serbian, Croatian and Slovenian construed as dialects. Lekhian, Czechoslovak and Polabian formed the Western branch. ‘Lekhian’ was the Ruthenian-language3 ethnonym employed for referring to the Poles. It survives to this day in Lithuanian (Lankas), Magyar (Lengyel) and colloquial Ukrainian (Lakh). In the late Middle Ages the Latinized form Lechitae functioned as a synonym for the Poles, and this term was also used in Poland-Lithuania. Through Latin it entered the Polish language as Lechici as a poetic term for the Poles. According to Šafařík, the Lekhian language, or rather sub-branch, included the dialects of Polish, Silesian and Pomeranian. The Czechoslovak sub-branch was composed of the dialects of Czech, Moravian and Slovak. The Polabian sub-branch embraced Sorbian and those Slavic dialects west of the River Oder which had become extinct prior to the nineteenth century. Šafařík’s classification doubled as that of the Slavic peoples (ethnic groups or nations) in line with the ethnolinguistic equation of language with nation and state already adopted by German nationalists in the course of the Napoleonic Wars at the beginning of the nineteenth century (Šafařík 1837: 483).
3
Ruthenian (ruski) was a Cyrillic-based vernacular Slavic language used in an of ficial capacity in the Grand Duchy of Lithuania with its two chancery centers at Vil’na (Vilnius) and Kyiv. It was replaced with Polish in 1699. Today, Belarusian linguists claim the language of the Vil’na Ruthenian languages as ‘Old Belarusian’ and their Ukrainian counterparts see Ruthenian-language documents produced at Kyiv as written in Old Ukrainian (Anichenka 1969).
157
Classifying the Slavic Languages, or the Politics of Classification Table 3 Elaborated dual classification of Slavic 1. South-Eastern Slavic languages Russian
Bulgarian Illyrian – Serbian – Croatian – Slovenian
2. Western Slavic languages Lekhian
Czechoslovak
Polabian
– Polish – Silesian – Pomeranian [Kashubian]
– Czech – Moravian – Slovak
– Sorbian – Polabian (extinct)
3.4 Another triple classification and its variations Frantíšek Palacký (Franz Palacky, 1798–1876) initiated Czech national historiography, writing first in German, and after 1848 also in Czech. Later he became the first leader of the Czech national movement. When still known as Franz Palacky, in 1836 he published the first volume of his Geschichte von Böhmen (History of Bohemia). In this work he divided the Slavs and their languages into three groups. In the first he included Russian and Bulgarian seen as the direct descendents of Old Church Slavonic – the first written Slavic language. The second, or South-Western, group included Serbian, Croatian and Carinthian (Slovenian). In the third, or North-Western, group Palacký placed Lekhian, which could be interpreted as being composed from Czech, Polish, Slovak and Sorbian. Table 4 Revised triple classification of Slavic I 1. [Eastern]
2. South-Western
3. North-Western
Old Church Slavonic
Serbian Croatian Carinthian (Slovenian)
Lekhian
Russian Bulgarian
158
Tomasz Kamusella
At the same time Vostokov’s insistence on the special separate position of Russian among the Slavic languages became popular in Russia. The Russian linguist Mikhail Maksimovich (1804–1873) in his Istoriia drevnei russkoi slovesnosti (1839, History of Early Russian Literature) and Nachatki russkoi filologii (1848, An Introduction to Russian Philology) proposed that Russian constituted a separate branch of the Slavic languages in its own right. The two other branches included the South-Western and North-Western Slavic languages, as in Palacký’s classification. Interestingly, Maksimovich distinguished two sub-branches in the Russian group: the South Russian consisting of Little Russian (or today’s Ukrainian), and the North-Eastern consisting of Great Russian (Russian) and White Russian (Belarusian). At that time Russia’s imperial political ideology was extended into the concept of the Great Russian language that, purportedly, stemmed directly from Old Church Slavonic and comprised Little, or Southern Russian (Ukrainian) and White, or North-Eastern Russian (Belarusian) as its dialects. Table 5 Revised triple classification of Slavic II 1. Russian [Eastern Slavic languages?] South Russian Little Russian (Ukrainian)
2. South-Western North-Eastern Slavic languages Great Russian (Russian) White Russian (Belarusian)
3. North-Western Slavic languages
In 1843 and 1845 the Russian philologist Izmail Ivanovich Sreznevskii (1812–1880) criticized Šafařík’s dual classification of the Slavic languages, and gave his own elaborated version of the triple division. In Sreznevskii’s scheme the Eastern group comprised Great Russian and Little Russian. The South-Western included Church Slavonic, Bulgarian, Serbo-Croatian (or Serbian and Croatian as two literary varieties of a single language), Carinthian (Slovenian). In the North-Western group he placed Polish, Polabian, Lusatian (Sorbian), Czech and Slovak (Ottův slovník naučný 1901: 449).
Classifying the Slavic Languages, or the Politics of Classification
159
Table 6 Revised triple classification of Slavic III 1. Eastern group
2. South-Western group
3. North-Western group
Great Russian Little Russian
Church Slavonic Bulgarian Serbo-Croatian (or Serbian and Croatian) Carinthian (Slovenian)
Polish Polabian Lusatian (Sorbian) Czech Slovak
3.5 Schleicher: A dual classification with a twist In 1858 August Schleicher forcefully entered the ongoing discussion on the classification of the Slavic languages (or dialects) with his article ‘Kurzer Abriß der Geschichte der slawischen Sprache’ (A Brief Outline of the History of the Slavic Language). He criticized the then-emerging triple division of the Slavic languages and argued for the dual one, as proposed by Dobrovský. A qualitative leap occurred in 1863. In his work Die Darwinsche Theorie und die Sprachwissenschaft, Schleicher presented the genealogical tree of the Indo-Germanic (Indo-European) languages. To the typical ranking of languages governed by perceived closeness among them, Schleicher added the dimension of time. The idea of a language as self-contained entity, unambiguously separate from others of the same kind, was projected into the past. The result was a genealogical tree of rapidly branching out languages. In this scheme, at a single moment, from previously common (parent) languages perfectly separate (dif ferent) of fspring languages split in an inexplicable fashion. Darwin treated his highly schematic ‘trees of life’ as a mental prop for illustrating the idea of speciation that occurs in the course of evolution. In a radical departure from Darwin’s limited conceptual and illustrative use of the ‘tree’, Schleicher appended meticulous labels to the branches and twigs of his genealogical tree of the IndoGermanic (that is, Indo-European) languages, and of fered it as an exact picture and explanation of how these languages ‘evolved’ through time. According to him, the pedigree of the Slavic languages commenced with
160
Tomasz Kamusella
the splitting of the Indo-Germanic ur-language into the Slavo-German and Aryan-Greek-Italic-Celtic stock (basis) languages, or Grundsprache. Next the former branched out into the German and Slavo-Latvian (LithuanoSlavic) stock languages. From the Slavo-Latvian, the Baltic (Lithuanian) and Slavic stock languages emerged. Beyond this point Schleicher applied his ‘genetic’ (diachronic) approach to the classification of the Slavic languages. At first he drew on Dobrovský’s classification when he distinguished the two languages of West Slavic and South-Eastern Slavic, which he presented as originating from the common Slavic stock language. In the next phase the West Slavic language spawned Sorbian, Polish, Czech and Polabian, the last-named language being already extinct by the nineteenth century. This branch closely corresponded to the second group of Slavic languages in Dobrovský’s classification, to the Western group in Šafařík’s classification, and to the North-Western group in Palacký’s, Maksimovich’s and Sreznevskii’s triple classifications. Schleicher decided that the South-Eastern Slavic language bifurcated into Russian and the South Slavic language. Thanks to this subdivision he merged the insights of the dual and triple classifications of the Slavic languages. Schleicher also bowed to Vostokov’s and other Russian philologists’ insistence that Russian should be accorded a special place, or, best of all, a separate classificatory category in its own right. Instead of bestowing the same classificatory rank on the remaining Slavic languages, as the proponents of the triple division did, Schleicher split the South Slavic language into the Serbo-Slovenian and Bulgarian languages, which amounted to the tentative introduction of a quadruple division of the Slavic languages. Bulgarian, like Russian, constituted a group on its own, whereas Serbo-Slovenian, predictably, gave rise to Serbian and Slovenian. All the final-position languages were adorned with tiny tufts of hair-like twigs representing the dialects of these languages (see Figure 5).
Classifying the Slavic Languages, or the Politics of Classification
161
Figure 5 A. Schleicher’s genealogical tree of the Indo-Germanic (Indo-European) languages (1873: diagram after p. 33)
Schleicher’s diachronic classification scheme of the Slavic languages gave a priority to the dual division, though at a later (lower) level of classification advocated the triple division, and at an even later (lower) level proposed the quadruple division of these languages. In his 1871 work Laut- und Formenlehre der polabischen Sprache (The Sound and Morphological Structure of the Polabian Language), Schleicher subdivided the West Slavic languages. According to him, the West Slavic language gave rise to the Lekhian and Czech languages. Then the former split into Polish and the extinct Polabian, while the latter split into Czechoslovak and Sorbian. At the lowest classificatory level of Schleicher’s scheme it meant a quintuple division of the Slavic languages (Alter 1999: 75; Ottův slovník naučný 1901: 449; Schleicher 1869: diagram after p. 69).
162
Tomasz Kamusella
3.6 A dual-cum-quadruple classification Schleicher’s genetic tree of the Slavic languages inspired the Czech linguist Jan Gebauer (1838–1907) to propose his own multilevel scheme for classifying the Slavic languages in 1870. However, he rejected Schleicher’s concept of a genealogical tree with the entailed dimension of time (linguistic evolution), and returned to the idea of a synchronic multi-level classificatory scheme. As the basic level he chose the dual division of the Slavic languages, also preferred by Schleicher. This level consisted of the South-Eastern and Western groups. The latter included Czech, Polish, Sorbian and extinct Polabian, which corresponded to Schleicher’s view before he subdivided this group in 1871. In the case of the South-Eastern group, Gebauer reduced Schleicher’s further subdivisions into three subgroups of equal rank: Russian, Bulgarian and Serbocroatoslovenian.4 The first subgroup embraced the Great Russian (Russian) and Little Russian (Ukrainian) languages, alongside the White Russian (Belarusian) dialects. In the second subgroup Old Bulgarian (Old Church Slavonic) and Modern Bulgarian were placed, while in the third he placed SerboCroatian and Slovenian. Unlike Schleicher, Gebauer specified the dialects of all the aforementioned languages with the exception of the category of the ‘White Russian dialects’, which was not subdivided into component dialects. Significantly, Ukrainian’s Carpathian dialect today can be identified with the Rusyn (Lemkian) language; Modern Bulgarian’s Macedonian dialect with the Macedonian language; Czech’s Hungarian-Slovak dialect with the Slovak language; and Polish’s Kashubian and Silesian dialects with the Kashubian language and the Silesian language, respectively. It is also 4
Gebauer’s coinage, ‘Serbocroatoslovenian,’ proved quite inf luential, due to the sustained broad personal and intellectual contacts maintained between the Czechs and the Balkan Slavs within the framework of Austria-Hungary. Bowing to the normative idea of ethnolinguistic nationalism that a nation should speak a single language, which, in turn, would be the sole of ficial language of the nation’s nation-state, the 1921 constitution of the Kingdom of Serbs, Croats, and Slovenes (after 1929, Yugoslavia) announced Serbocroatoslovenian (later, popularly known as Yugoslavian) as the of ficial language of this polity.
163
Classifying the Slavic Languages, or the Politics of Classification
interesting to see that among Slovenian’s dialects, Gebauer distinguished Croato-Slovenian as a transitional dialect between the Slovenian language and Serbo-Croatian. Having skipped one level of subdivision (Russian vs South Slavic) present in Schleicher’s scheme, Gebauer did not incorporate the triple division of the Slavic languages in his scheme, but adopted Schleicher’s quadruple division (Rieger 1870: 645). Table 7 Dual-cum-quadruple classification of Slavic
Western Slavic languages
South-Eastern Slavic languages Russian Great Russian (Russian) Little Russian (Ukrainian) White Russian dialects (Belarussian)
Bulgarian Old Bulgarian (Old Church Slavonic) Modern Bulgarian (Bulgarian)
Serbocroatoslovenian Czech Polish Serbo-Croatian Sorbian Slovenian Polabian (extinct)
4 A Criticism of the Classificatory Schemes The catalogue of dual, triple, quadruple and quintuple classifications of the Slavic languages, of fered by Dobrovský, Šafařík, Palacký, Maksimovich, Sreznevskii, Schleicher and Gebauer, became the basis for further variants ceaselessly worked out by Slavicists. In the late 1870s the Russian Slavicist, Aleksandr Kochubinskii (1845–1907), disputed this method of classifying the Slavic languages in his work, K voprosu o vzaimnykh otnosheniiakh slavianskikh narechii (1877–1878, On the Issue of Mutual Relationships among the Slavic Dialects). He noticed that depending on which linguistic features (be they syntactical, morphological, phonetic, lexical, or a mixture of them) is considered the site of the defining characteristics of languages,
164
Tomasz Kamusella
an endless chain of dif fering classifications of the Slavic languages can be generated. The dependence of the outcome of a project of linguistic research on the original choice of analytical classificatory scheme is thus evidently susceptible either to the internal personal prejudices of the researcher, or to the external demands of political expediency. Even Kochubinskii failed to disentangle himself from the temptations which he criticized. He came to the conclusion that the proper task of Slavicists was to decide which of the Slavic languages is the oldest. Not surprisingly, he accorded this distinction to his native Russian, arguing that it is closest to the first written Slavic language – Old Church Slavonic. It took Johannes Schmidt’s Zur Geschichte der indogermanischen Voca lismus (1875, On the History of Indo-Germanic Vocalism) to undermine the validity of Schleicher’s model of a genealogical tree. First, like Kochubinskii, Schmidt argued that it does not make sense to come up with isoglosses (based on selected linguistic criteria) to separate Slavic languages from one another. This urge for establishing such isoglosses, in my view, amounts to the projection onto the linguistic reality of the perceived or striven-for clear-cut boundaries among ethnically construed nations, and of the existing political borders. But this linguistic reality is continuous not discrete, so that the Slavic languages (construed as dialect chains) shade into one another. This intuition of Schmidt’s became the basis for the later development of the concept of dialect continuum,5 which is bounded by other, often sharply (genetically) dif ferent, dialect continua. Second, having dealt with the synchronic plane, he criticized the diachronic dimension of the genetic tree division of the Slavic languages. He observed that languages were not born from earlier languages that were extinguished in this process. The observable variety of languages arose through the internal dif ferentiation of the Slavic dialect continuum 5
The concept of dialect continuum arose in the wake of the development of creole linguistics. First, in 1914 Schuchardt spoke of ‘language or speech continuum’ (Sprachkontinuum) and the US linguists, John E Reinecke and Aiko Tokimasa, transplanted this term to English as ‘speech continuum’ in 1934. On this basis, in 1971 another US linguist, David DeCamp, coined the actual term ‘dialect continuum’ (DeCamp 1971; Reincke and Tokimasa 1934; Schuchardt 1914).
Classifying the Slavic Languages, or the Politics of Classification
165
due to historical and political processes such as state formation, warfare, population movements, invasions and so on. Schmidt’s and Schuchardt’s insights underlay the areal (wave, Welle) theory of representing relationships between languages. Faced with the immense success of the graphic representations of Schleicher’s genealogical tree (Stammbaum), Schmidt devised a circle diagram to illustrate his ideas on how languages are related to and inf luence one another. The circle was divided into triangular pieces, whose apexes met in the middle of the circle as in a pie. Each piece of such a ‘linguistic pie’ represented a language construed as a segment of a dialect continuum prior the emergence of a standard language. Applied to the Slavic languages this circular diagram placed languages side by side in the order Sorbian – Polish – Russian – Bulgarian – Serbo-Croatian – Slovenian – Czech – Sorbian (returning to the starting point) (Ottův slovník naučný 1901: 449–450). Although the wave classification theory emphasized the gradual change within the dialect continuum, Schmidt’s circle diagram partially defied this logic by introducing boundaries among the ‘dialectal languages’. The Polish linguist Jan Baudouin de Courtenay (1845–1929) in his book, Übersicht der slavischen Sprachenwelt (1884, An Overview of the World of the Slavic Languages), repeated that linguists should not aspire either to establishing or to identifying clearly delineated borders that would unambiguously separate Slavic languages from one another. Dialect territories popularly associated with standard languages are indistinguishable parts of the Slavic dialect continuum and gradually shade from one into another. Like Schmidt, Baudoin de Courtenay also criticized Schleicher’s concept of the common (stock) languages of (Proto-)Slavic, Western Slavic and South-Eastern Slavic from which all the existing Slavic languages were supposed to have branched out. Such languages had not existed. The rise and persistence of the concept of such ur-languages is a function of the entrenched dominance of the genealogical tree model for visualizing the classification of species, languages and human groups (Ottův slovník naučný 1901: 450). The attraction of the genealogical tree method of classification was both graphical and political. By presenting elaborate genealogical trees of ethnic groups, animals, plants and languages as ‘scientific’, the depiction of such images added to the scientific prestige and potential success of books
166
Tomasz Kamusella
and articles in which they were used. The potential to represent certain languages as older than others on such trees could be mobilized for political ends which were not intrinsically linguistic. One of the main tools for justifying the continued existence of an existing state, or the establishment of a new (or renewed) state for a group of people construed as a nation was provided by ancient pedigree. Those nations and states which were successfully presented as ‘older’ acquired a ‘higher right’ to existence than their rivals perceived as ‘younger’. The Württemberg philosopher Georg W.F. Hegel (1770–1831) set this dif ferentiation in stone, terming the older states or nations as ‘nations with history’ and the younger as ‘nations without history.’ Since the 1840s this distinction had became popular equally among both nationalist and Marxian thinkers (Herod 1976). The graphic representation of the Stammbaum perfectly suited the propaganda aim of ‘proving’ that one nation or state was older than another. This device allowed for conf lating and projecting into the distant past a contemporary language, ethnicity and statehood. Given the extraordinary political utility of the genealogical tree method of classification, it was adopted as the dominant paradigm, and the ‘wave’ model of language classification remained secluded as a minority voice within scholarly discourse, despite its explicatory power. There are two clear reasons for the public success of the ‘tree’ model and the relative failure of the ‘wave’ model. First, the proponents of the wave-based areal method of classification did not develop an appealing graphical representation for it. Second, this method of classification went against the nationalist logic of the radical normative isomorphism (tight spatial and ideological overlapping) of state, nation and language. In the age of nationalism that commenced in the nineteenth century and has lasted to this day, this condemned the wave theory to obscurity.
Classifying the Slavic Languages, or the Politics of Classification
167
5 Today’s consensus: Triple classification In the twentieth century the family tree classification and presentation of the Slavic languages was the conventional wisdom, and it is still so in the twenty-first century. The triple classification of these languages also became part of the scholarly norm in the interwar period, and as such it has thrived unchallenged to this day. Two recent English-language academic overviews of the Slavic languages follow this pattern with minor variations. In 1980, Alexander Schenker and Edward Stankiewicz in their The Slavic Literary Languages: Formation and Development included Bulgarian, Macedonian, Serbo-Croatian and Slovenian in the Southern branch. Belarusian, Russian and Ukrainian were placed in the Eastern branch; the Western branch included Czech, Kashubian, Polish, Slovak and Sorbian. Church Slavonic, the survey of which opened the book, was accorded a separate place outside the branches of the triple classification (Schenker and Stankiewicz 1980). Thirteen years later, the editors of The Slavonic Languages decided to add the liturgical language of Old Church Slavonic to the Southern branch and to add the extinct Polabian to the Western branch. They also gave priority to the West Slavic languages over the East Slavic ones presented at the end of this work (Comrie and Corbett 1993). Scholars still produce variations on the genetic triple division of the Slavic languages. For instance, in 1992, the Polish Slavicist, Ewa Siatkowska, presented one in her Rodzina języków zachodniosłowiańskich (The Family of the Western Slavic Languages). According to her the Proto-Slavic language was the precursor of the Proto-Rus’ language, which gave rise to the Eastern group of the Slavic languages. With time the Proto-Ruthenian (Proto-Russian) language emerged from Proto-Rus’, and, later, split into White Ruthenian (Belarusian), Great Ruthenian (Russian) and Little Ruthenian (Ukrainian). Interestingly, Siatkowska proposed that Bulgarian, Serbo-Croatian and Slovenian emerged directly from Proto-Slavic, but form a Southern group of Slavic languages due to the geographical proximity maintained between the users of all three languages. Further, in Siatkowska’s scheme, Bulgarian gave rise to Macedonian, and Old Church
168
Tomasz Kamusella
Slavonic was made into an equidistant relative (though whether predecessor or side-branch is unclear) of Bulgarian and Macedonian. Lastly, the languages of the Western group (Slovak, Czech, Upper Sorbian, Lower Sorbian, Polish and the extinct Polabian) are portrayed as emerging from a common but nameless proto-language. Perhaps she feared that naming such a proto-language would necessitate giving precedence to one of the Western Slavic languages over the others, as in the case of Proto-Russian and Great Russian to the detriment of Belarusian and Ukrainian (Siatkowska 1992: 34). Such academic variations on the theme of the triple genetic division of the Slavic languages are not ref lected in general textbooks or encyclopedias. The standard triple division remains the dominant model. It molds the imagination of new generations of Slavicists and linguists, and has already become the accepted ‘truth’ among intellectuals worldwide. Through endless reiteration and reproduction in mass education and mass communication, the standard triple classification of the Slavic languages overshadows and condemns to oblivion its own ambiguous origin and the rival classifications that had persisted into the three first decades of the twentieth century. It seems that this triple division, having initially been advanced as a tentative consensus reached in the milieu of Russian and Central European Slavophone philologists, is in danger of congealing into an unchallengeable dogma-like consensus. Unfortunately, this state of af fairs discourages further scholarly probing into the issue.
6 Conclusion One can paraphrase Schleicher’s opinion: ‘many glossologists [linguists] speak of the Slavonic dialects, others of the Slavonic languages’ (1869: 46) by proposing that most Slavicists speak now of the triple division of the Slavic languages, while in the past others advanced alternative classification models, especially dual models. Languages and dialects are largely
Classifying the Slavic Languages, or the Politics of Classification
169
subjectively defined entities, existing in the eye of the beholder. However, although the distinction between them is less linguistic but rather imagined, it is nevertheless real enough, enshrined to this end in constitutions and legislation, regulated through of ficially approved grammars and dictionaries, and imparted to the population at large through popular education and mass media. The same is true of the genealogical tree-type classifications of genetically closely related languages. In their field work on dialects, linguists working in this paradigm determine the so-called ‘bundles of isoglosses’ that allow them to buttress one or another ethnolinguistically defined political project with an appropriate ‘linguistic boundary’. Such a boundary between languages or groups of languages lends itself, in turn, to a desired classificatory scheme. In this manner the error of anachronism is committed, as findings of dialectal research are usually transposed onto standardized national languages. These languages are a product of recent times, commonly not older than one or two centuries, while dialects are much older, being the very medium of oral (that is, natural) communication between humans. Thus, the origin of a dialect (obviously, without denying constant change, which any form of spoken language undergoes from generation to generation) can be dated to the time a stable social group employing it had come into being, be it in a village or town in the case of agricultural societies or in extended family clans, when it comes to nomads. Hence, the popular subsuming of dialects under the roof of languages (Dachsprache) is an anachronism. This tendency is widely accepted as appropriately scientific, because it agrees with the political needs of nation states, whose legitimacy is at least to some degree contingent on the idea that language equates to nation, which, in turn, equates to an ethnolinguistically defined nation-state for such a nation. (I term this underlying principle of ethnolinguistic nationalism the ‘normative isomorphism of language, nation and nation state’.) On a broader plane, bundles of dialects ascribed in this way to standard languages are subsumed under the larger classificatory categories of language branches. Some even believe that in the past such branches used to be singular parent (stock) languages from which contemporary (of fspring) languages emerged. However, this model leaves out of account the pervasive phenomenon of unwritten spoken dialects
170
Tomasz Kamusella
which had sustained the bulk of human communication throughout Central Europe until the rise of genuine full literacy in the first two decades after World War II, when language committed to writing began to exert an unprecedented mass inf luence, ef fectively leveling dialectal dif ferences. What are we then to make of the established tradition that distinguishes between the branches of West and East Slavic languages as separate from each other? In the latter case, the Rus’ (‘Old Russian’) language can be proposed as the common progenitor of today’s Belarusian, Russian and Ukrainian. However in the case of the West Slavic languages, linguists are either silent on identifying a parent language of this group or indecisively put it on diagrams without naming it. First, there was in the past no polity which had contained all the territories inhabited by West Slavs. Second, proposing a clearly delineated and unambiguously named progenitor of the West Slavic languages would mean, in ethnolinguistically determined politics, conceding to a single West Slavic nation the right to dominate the others. That is how, since the mid-nineteenth century, Russian tsars and then the Kremlin have repeatedly employed the politicized claim of identifying the Rus’ language with (Great) Russian in order to ‘prove’ that Belarusian and Ukrainian are ‘mere dialects (narechia)’ of the Russian language, and hence, Belarusians and Ukrainians are ‘ethnographic branches’ of the (Great) Russian nation. Another problem that challenges the now already conventional separation of the West Slavic branch of languages from the Eastern branch is the impossibility of distinguishing on a purely linguistic basis which of the ‘border or transitional dialects’ belong to the West Slavic or East Slavic languages. There is no geographical discontinuity in the form of a sea or a vast swath of land inhabited by non-Slavic speakers that would separate West Slavs from East Slavs, unlike in the case of South Slavs who are separated from both the other groups by the territories inhabited by Finno-Ugric-speakers (Hungarians) and Romance-speakers (Romanians and Moldovans). The undeniable linguistic continuity between West and East Slavic languages obliges the consideration of the question of whether the basis of the supposed linguistic separation of the West and East Slavic languages is extralinguistic, or, as best, perilinguistic. The extralinguistic factors may
Classifying the Slavic Languages, or the Politics of Classification
171
be religious, historic or political. An overwhelming majority of West Slavs profess either Roman Catholicism or Protestantism, while most East Slavs profess either Orthodox Christianity or Greek (Uniate) Catholicism. Looking at the West-East Slavic divide from a political vantage, one can readily notice that the lands inhabited by East Slavic-speakers were historically included within the boundaries of Rus’. The Russian Empire, as Muscovy renamed itself in 1721, grew from the policy of ‘gathering the lands of Rus’. Not surprisingly as a result, the perceived linguistic boundary between West and East Slavs largely happens to coincide with nineteenth-century Russia’s western frontier. The minor exceptions to the rule were St Petersburg’s historically non-Rus’an autonomous Congress Kingdom of Poland, and Vienna’s historically Rus’an eastern half of Galicia. This discrepancy disappeared after the westward expansion of the Soviet Union in the wake of World War II. Then all the historically Rus’an lands, thus seen as East Slavic, found themselves in one polity, and today are shared between Belarus, Russia and Ukraine. The perilinguistic division between West and East Slavs is that between the Latin and Cyrillic alphabets.6 All West Slavic languages are written in Latin characters, while East Slavic ones are written in Cyrillic. This ref lects an old religious division, as in Central Europe Western Christianity (Roman Catholicism and later Protestantism) was liturgically and culturally connected to the Latin language with its specific script, while Slavic Eastern Christianity (Orthodoxy, and later Greek Catholicism or Uniatism) had and retains a similar relationship with Cyrillic-based Church Slavonic. However, the now apparently clear-cut writing system division between East and West Slavs is of recent making. Until the 1850s some Little Ruthenian (Ukrainian) publications appeared in Latin characters in Galicia, while Belarusian-language book production was bi-scriptural, in the Latin script and Cyrillic, until after World War I.
6
I refer to writing as ‘perilinguistic,’ due to its close relationship with language. Technically this technology is as ‘extralinguistic’ as political fiat, but scripts map languages at various linguistic levels, and have been in symbiotic relationships with some of them for five millennia.
172
Tomasz Kamusella
Bearing these facts in mind, it is clear that the division between West and East Slavs is religious, historic and political in its nature, not linguistic. Unfortunately, this division, extralinguistic and perilinguistic in its nature, doubles as the linguistic division between the Eastern and Western branches of the Slavic languages. I am sure it would be equally beneficial for linguistics and politics to decouple, or at least demythologize, this spurious complex. Then linguistics would not be demeaned as a ‘scientific argument’ for propping up political projects, and, conversely, unspoken political axioms of national correctness would not have to be observed in pursuing linguistic research. But one can ask if all that means that there is no possibility of classifying East and West Slavic languages on any linguistic ground. That would be going one step too far. A solution to this dilemma is to cease to be dazzled by the deceptive allure of the genealogical tree model of classification and instead to introduce into the standard model some elements of the areal classification model of dialect continuum. Should we agree to adopt this proposal, it emerges clearly that there are two Slavic dialect continua – Northern and Southern (Chambers and Trudgill 1998: 6; Crystal 1987: 25).
References Alter, Stephan G. 1999. Darwinism and the Linguistic Image: Language, Race, and Natural Theology in the Nineteenth Century. Baltimore: The Johns Hopkins University Press. Anichenka, U.V. 1969. Belaruska-ukrainskiia pis’mova-mounyia suviazi. Minsk: Navuka i tekhnika. Bloomfield, Leonard. 1926. ‘A Set of Postulates for the Science of Language’. Language, 153–164. Busmann, Hadumod, ed. 1996. Routledge Dictionary of Language and Linguistics. London: Routledge. Cavalli-Sforza, L. Luca, Menozzi, Paolo, & Piazza, Alberto. 1994. The History and Geography of Human Genes. Princeton, NJ: Princeton University Press.
Classifying the Slavic Languages, or the Politics of Classification
173
Chambers’s Encyclopaedia: A Dictionary of Universal Knowledge (vol 9). 1908. London: William & Robert Chambers and Philadelphia: J.B. Lippincott. Chambers, J.K., & Trudgill, Peter. 1998. Dialectology. Cambridge: Cambridge University Press. Comrie, Bernard, & Corbett, Greville (eds). 1993. The Slavonic Languages. London: Routledge. Crystal, David. 1987. The Cambridge Encyclopedia of Language. Cambridge: Cambridge University Press. DeCamp, David. 1971. ‘Toward a Generative Analysis of a Post-Creole Speech Continuum.’ In: Hymes, Dell (ed.), Pidginization and Creolization of Language. Cambridge: Cambridge University Press, 349–370. Décsy, Gyula. 1973. Die linguistische Struktur Europas. Vergangenheit – Gegenwart – Zukunft. Wiesbaden: Otto Harrassowitz. Gellner, Ernest. 1983. Nations and Nationalism. Oxford: Blackwell. Haardt, W., & Gustawicz, B. 1909. Kozenna Atlas szkolny. Vienna: Edward Hölzl Geographisches Institut. Herod, Charles C. 1976. The Nation in the History of Marxian Thought: The Concept of Nations with History and Nations without History. The Hague: Martinus Nijhof f. Kamusella, Tomasz. 2006. ‘The Isomorphism of Language, Nation, and State: The Case of Central Europe.’ In: Burszta, W. et al. (eds), Nationalisms Across the Globe: An Overview of Nationalisms of State-Endowed and Stateless Nations (Vol 2: The World). Poznań: Wyższa Szkoła Nauk Humanistycznych i Dziennikarstwa. 57–92. Kochubinskii, Aleksandr. 1877–1878. K voprosu o vzaimnykh otnosheniiakh slavianskikh narechii. Odessa: Ulrikh i Shults. Kohn, Hans. 1962. The Age of Nationalism: The First Era of Global History. New York: Harper & Row. Lehr-Spławiński, Tadeusz (ed.). 1949. Chrestomatia słowiańska (Vol I: Teksty południowo-słowiańskie, Vol II: Teksty zachodnio-słowiańskie, Vol III: Teksty wschodnio-słowiańskie). Kraków: Wydawnictwo Studium Słowiańskiego Uniwersytetu Jagiellońskiego and Ossolineum. Maher, J. Peter, & Koerner, Konrad (eds). 1983. Linguistics and Evolutionary Theory: Three Essays by August Schleicher, Ernst Haeckel, and Wilhelm Bleek. Amsterdam: John Benjamins. Meyers Konversations-Lexikon 1889. Vol. 14. Leipzig: Verlag des Bibliographischen Instituts. Ottův slovník naučný. 1901. Vol. 22. Praha: J. Otto.
174
Tomasz Kamusella
Reinecke, John E., & Aiko Tokimasa. 1934. ‘The English Dialect of Hawaii.’ American Speech, 9, 48–58, 122–130. Richards, Robert J. 2004. The Linguistic Creation of Man: Charles Darwin, August Schleicher, Ernst Haeckel, and the Missing Link in Nineteenth-Century Evolution Theory. accessed 2 April 2009. Rieger, František. 1870. Slovník naučný (vol 8). Praha: I.L. Kober. Šafařík, Pavel Josef. 1837. Slowanské starožitnosti, Praha: Tiskem Jana Spurného. Schenker, Alexander M., & Stankiewicz, Edward (eds). 1980. The Slavic Literary Languages: Formation and Development. New Haven, CT: Yale Concilium on International and Area Studies. Schleicher, August. 1858. ‘Kurzer Abriß der Geschichte der slawischen Sprache.’ Beiträge zur vergleichenden Sprachforschung 1, 1–27. Schleicher, August. 1869. Darwinism Tested by the Science of Language. London: John Camden Hotten. Schleicher, August. 1873. Die Darwinsche Theorie und die Sprachwissenschaft. Of fenes Sendschreiben an Herrn Dr. Ernst Häckel. 2nd edn. Weimar: Böhlau. Schuchardt, Hugo. 1914. Die Sprache der Saramakkaneger in Surinam. Amsterdam: Johannes Muller. Siatkowska, Ewa. 1992. Rodzina języków zachodniosłowiańskich. Warszawa: PWN.
José del Valle
Linguistic History and the Development of Normative Regimes: The Royal Spanish Academy’s Disputed Transatlantic Authority
Abstract The Royal Spanish Academy (RAE) was created in 1713 to preserve the purity of the language in Spain and the territories of the Empire. However, once Spain’s American colonies became independent in the early nineteenth century, the Academy’s authority was challenged and declarations of linguistic independence proliferated. The RAE’s 1870 initiative to create associated language academies and thus retain control over the language was only moderately successful: a strong perception remained that the Spaniards would not share linguistic power. In 1950, Mexican president Miguel Alemán invited the RAE and all associated Academies to meet in Mexico and coordinate ef forts; all attended but the Spaniards. Incensed by this rejection, a member of the Mexican delegation called for linguistic independence; subordination to the RAE should end and a truly democratic association of Academies should be created. His proposal was soundly defeated. In this article, I discuss the debates surrounding the RAE’s absence as discursive sites where the nature of the pan-Hispanic community was being worked out.
1 Introduction On 14 June 1950, Mexican president Miguel Alemán (1902–1983) launched an original and ambitious cultural initiative that would end up having a profound impact – through a series of both expected and unexpected
176
José del Valle
turns – on the institutional framework that supports the standardization of Spanish:1 he charged the Mexican Academy of the Spanish Language with organizing a conference whose goal would be to protect the nature and integrity of Spanish. All language academies, from Spain, Latin America, and the Philippines, were invited.2
2 The Royal Spanish Academy The history of these institutions begins, of course, much earlier, in 1713, when the Royal Spanish Academy (Real Academia Española: RAE) was created in Spain under the protection of the first Bourbon King, Felipe V. Modelled after the Académie Française – created in 1635 – and Italy’s Academia della Crusca – created in 1582 – it embraced as its most pressing responsibility the production of a dictionary that, in addition to stabilizing Spanish and safeguarding its purity, would also stand as a proud display
1
2
My understanding of standardization draws from two dif ferent but related categories. I rely, first, on Einar Haugen’s definition of language planning as ‘the activity of preparing a normative orthography, grammar, and dictionary for the guidance of writers and speakers in a non-homogeneous speech community. In this practical application of linguistic knowledge we are proceeding beyond descriptive linguistics into an area where judgment must be exercised in the form of choices among available linguistic forms’ (Haugen 1972: 133). Secondly, I rely on Milroy and Milroy’s (1999) take; in particular on their recognition of standardization both as process and ideology. Thus, I see the history of Spanish not as the development of a highly standardized language (not as a single-minded march towards the standard, as James Milroy has ironically put it) but as a deeply ideological process through which Spanish is associated with various forms and given social meanings at dif ferent points in history. In this understanding, the historicity of language is to be found not only in its deployment along a chronological axis but also – and especially – in the dynamic nature of its relation to cultural, political, and social contexts (see also Blommaert 1999). In this article, ‘Latin America’ refers only to the Spanish-speaking countries in the continent.
Linguistic History and the Development of Normative Regimes
177
of the country’s cultural stature. For the Academy’s founding fathers, the dictionary – its first edition was published in six volumes between 1726 and 1739 – was a matter of national pride (Álvarez de Miranda 1995: 272).3 As the institution, slowly but surely, consolidated its status within Spain – broadening its scope to include orthography (first published in 1741) and grammar (first published in 1771) and strengthening its ties to the State apparatus4 – it also began to concern itself with its wavering reputation overseas, especially after the wars that, during the second decade of the nineteenth century, led to the collapse of the Spanish Empire and the emergence of new Spanish-speaking independent nations.
3 The origins of panhispanism Obviously, the collapse of the Spanish Empire and the independence of the American colonies resulted in broader projects of emancipation aimed at the construction of cultural systems autonomous – at least, relatively autonomous – from that of the former metropolis (Rama 1982). In this context, the intellectual leaders of the independence movement as well as the most prominent figures of the subsequent generations tended to turn their eyes away from Spain, and more often than not, France and the Anglo-Saxon world came to be preferred political and cultural beacons for the emerging Latin American countries. However, in the process of developing new national – properly American – identities, language 3
4
For a lengthy and of ficial history of the Royal Spanish Academy see Zamora Vicente 2000. A good article-length overview of the institution’s history is Álvarez de Miranda 1995. For an in-depth study of the linguistic climate in eighteenth century Spain see Lázaro Carreter 1949. The RAE’s grammar ‘was declared by Carlos III in 1780 the of ficial textbook for teaching Spanish in schools’ (Álvarez de Miranda 1995: 276). In the nineteenth century, the Academy’s orthography became of ficial in schools in 1844 (see Laura Villa’s article in this volume) and its grammar became again mandatory in 1854.
178
José del Valle
was a strategic resource, and the predominantly white creole elites that had led the independence movement were Spanish-speaking. Like North Americans vis-à-vis the British, Latin American elites had to reconcile the embrace and cultivation of the colonial language – which was their own – with the cultural and linguistic autonomy that they had struggled so vigorously to attain. Esteban Echeverría (1805–1851), for example, a prominent Argentinian intellectual, accepted the language, Spanish, without reservations but claimed the legitimate right of Spanish-speaking Americans to control its elaboration on their side of the Atlantic: ‘The only legacy that Americans could accept, and do gladly accept from Spain because it is truly precious, is the language; but they accept it on condition that it be improved, that it be progressively transformed, in other words, on condition that it become emancipated’ (quoted in Alfón 2008: 52).5 Similarly, Juan Bautista Alberdi (1810–1884), of the same generation, linking language and thought and embracing France’s intellectual tradition, predicted divergent linguistic developments on both sides of the Atlantic: ‘If language is just one aspect of thinking, ours demands intimate harmony with our American way of thinking, a thousand times more congenial to the rapid and direct movement of French thought than to the interminable contours of Spanish thought’ (quoted in Alfón 2008: 53). Domingo Faustino Sarmiento (1811–1888), a firm and salient voice among those of his generation (he was Argentina’s president between 1868 and 1874), also af firmed Latin America’s linguistic independence and aimed many of his rhetorical darts at what he felt was a powerless and useless Royal Spanish Academy: ‘To expect a powerless academy – which has no authority even in Spain itself, lacks prestige, and is benumbed by the knowledge of its own impotence – to give us rules that will in the end be useless to us is an humiliation unworthy of nations that have attained the rank of same’ (Sarmiento 1843: 29). Ángel Rosenblat, who studied the linguistic ideas of this Argentinean generation (Rosenblat 1960), summarized their position stating that ‘they were of one mind in their cultural and linguistic anti-Spanish attitude, which at times became 5
Throughout the article, all translations from Spanish are mine.
Linguistic History and the Development of Normative Regimes
179
plain Hispanophobia; in their fervent and neophyte enthusiasm for French literature and thought; […] in their af firmation of [North] America as a source of inspiration; […] in their rejection of any form of academic or academistic tutelage; in their af firmation of the language’s freedom, so that it might go forward with the new ideas’ (Rosenblat 1960: 557). In Spain, the cultural, economic, and political establishment did not remain completely passive after the colonial mishap. Through private and at times public initiatives, a slow but steady mobilization of institutional resources began to unfold in the hope that Spain could somehow recycle to its advantage the transatlantic space left void by the dismantlement of the Empire. These ef forts to engage in post-colonial community-building with Latin America have come to be known variously as hispanismo, hispanoamericanismo, or panhispanismo (Pike 1971, Sepúlveda 2005). While this cultural trend was mostly discursive, it also materialized in the form of a number of cultural initiatives that included congresses and symposia (such as the Ibero-American congresses of 1892 and 1900 organized by the Unión Íbero-Americana) as well as journals (such as La Ilustración Ibérica, La Revista Española de Ambos Mundos, and La Ilustración Española y Americana) (Fogelquist 1968). Panhispanismo was grounded in the belief that a common Spanish culture embodied in the Spanish language existed on both sides of the Atlantic and was the basis for an economically and politically operative entity, for a true hispanofonía. In this use of the term, hispanofonía does not simply refer to a group of nations linked by a shared communicative code; it is rather, following Benedict Anderson’s (1983) notion, an ‘imagined community’ grounded in a common language, itself imagined, that ties together in an emotional bond those who feel they possess it and those who have a sense of loyalty to it (Del Valle 2006 and 2008). It is, therefore, according to Susan Gal and Kathryn Woolard’s definition of the term (2001), a language ideology, a historically situated conception of Spanish as an enactment of a collective order in which Spain performs a central role.
180
José del Valle
4 The creation of associated academies As a result of concerns regarding the possibility of safeguarding linguistic unity and in keeping with broader policies aimed at maintaining Spain’s post-colonial preeminence in Latin America, in 1870, the Academy’s director, Mariano Roca de Togores (1812–1889), Marquis of Molins, appointed a special committee charged with exploring the creation of a series of associated academies throughout the Spanish-speaking world. In a matter of weeks, the committee issued a report stating that, while political ties had been severed by the wars of independence, the common language continued to be the anchor of a single shared intellectual system. They identified the language as the common fatherland and recognized, for the first time, that the majority of Spanish speakers was not to be found in Spain but in the American continent. They further suggested that the active intellectual life of the new nations and their frequent contact with other countries and their languages – more frequent than with Spain itself – should be a source of concern, since it might lead the language to follow dif ferent paths in dif ferent places. In view of these circumstances and following the committee’s recommendation, the decision was made to create a system of associated academies that would secure the preservation of the language and support the Spanish institution’s authority in the former colonies.6 It is crucial to point out that, in designing this new consortium, the Royal Academy retained control over the dictionary, grammar, and orthography; imposed its statutes and regulations; and retained the right to confirm all new members of the associated academies.7
6
7
The first new academies were founded in the 1870s – the first one in Colombia in 1871 – but the full process would turn out to be more cumbersome than expected: many in fact did not take of f until the third decade of the twentieth century or even later. Honduras, for example, did not have an academy until 1949. The report – which includes the statutes that would regulate the relationship between the Royal Academy and the associated academies – was written by Fermín de la Puente Apezechea and published in the Royal Academy’s Memorias (De la Puente Apezechea 1873). The process is also discussed in Zamora Vicente (1999: 347–349).
Linguistic History and the Development of Normative Regimes
181
5 The Academies’ Congress of 1951 Given the Spaniards’ determination to safeguard their authority and selfappointed role as leaders of the standardization process, it seems intriguing, to say the least – certainly fascinating from a geopolitical perspective on language policy – that it was the Mexican and not the Spanish institution that first summoned all academies to the defense of the common language. The fact is that on 14 June 1950, José Rubén Romero (1890–1952), Mexican writer and academician, publicly declared on behalf of president Alemán – whom he served as special advisor – the Mexican government’s commitment to organizing and fully subsidizing this landmark event. He predicted: a beautiful spectacle […] [that will bring together] all academies without exception […] without consideration for relationships among governments […] caring only for the common interest and mutual sympathy that springs from language, the spiritual blood of races. (Comisión Permanente 1952: 11)
An organizing committee was appointed on the 27th and, a few weeks later, dispatched across the Atlantic. Its members landed in Madrid on 13 October and were received at the airport by a generous cohort of Spanish academicians. On the 18th, at a special session of the Royal Academy (broadcast live to Spain and Latin America through Radio Nacional de España), the Mexican delegation issued the of ficial formal invitation and asked the head of the Spanish institution, distinguished philologist Ramón Menéndez Pidal (1869–1968), to of ficiate as chairman of the upcoming conference.8 In his response, Pidal enthusiastically accepted the generous invitation and, with an eloquent speech, reminded his fellow academicians of their immense responsibility: 8
Prior to the trip, the organizing committee had developed a set of rules and regulations. These (article 2) established that the conference’s president would be the director of the Royal Spanish Academy or any Spanish academician that he might wish to appoint as his replacement (Comisión Permanente 1952: 17).
182
José del Valle One does not have to be a prophet to predict and expect with all confidence that this meeting will initiate a crucial period in which the Academies’ mutual relation will be closer and more fruitful in the serious task that they face: maintaining our common language as one of the most powerful bonds created by Humanity. Although this language community may be strained, it is essentially unbreakable […]. Securing the superior, prosperous, and ef fective unity of our language or letting it decay – not into complete fragmentation but through excessive valorization of the dialectal dif ferences that are normal in all languages – is in the hands of those who speak it and in the inspiration of those who write in it; but also – and this appeals to our sense of responsibility – in the conscience and the minds of those of us who belong to these bodies that will now meet in Mexico. (‘Información Académica’ 1950: 480–481)
6 The polemic before the conference By early 1951, everything seemed to be on track and the conference promised to be a celebration of pan-Hispanic harmony and commitment to the joint defense of the common language under Spain’s tutelage and Mexico’s sponsorship. The program had been designed and approved by the Royal Academy.9 But suddenly, on 26 February, the Mexican Academy received a telegram: ‘SURGIDA DIFICULTAD INSUPERABLE / UNSURMOUNTABLE PROBLEM ARISEN’. It was signed by Julio Casares (1877–1964), the Royal Academy’s secretary, and followed by a letter of the same date confirming, with barely a few words more, the Spanish institution’s withdrawal from the conference. The circumstances surrounding the agency’s apparent change of heart would only be aired six weeks later, on 7 April, through a public statement made by Spain’s Education Minister: Upon receiving the invitation from the president of the Republic of Mexico, the Royal Spanish Academy stated that, for patriotic reasons, it demanded, as a necessary moral pre-condition of its participation, that the Mexican government publicly state 9
The program was organized around four major topics: the unity and protection of Spanish, grammar issues, lexical issues, and interacademic cooperation (Comisión Permanente 1952: 21–24).
Linguistic History and the Development of Normative Regimes
183
that it had ended its relations with the red government10 and suspended recognition of the so-called Spanish diplomatic representation in Mexico. Since the Mexican government did not comply with this requirement – which in the present circumstances our national dignity deems indispensible – the Royal Spanish Academy has decided not to attend. (Pagano 1951: 253)
The Spanish government’s attitude triggered, as one might expect, a heated reaction throughout Spanish America and most intensely in Mexico. The daily press filled its pages with articles reporting on the episode and expressing indignation at what was felt to be Spain’s slap in the face to the Mexican people. However, in spite of the commotion, the conference organizers were determined to minimize the impact of the crisis. The opening ceremonies proceeded with apparent normality and with nothing but a mild passing reference to the Spaniards’ absence: We wanted a Hispanism that was fitting and logical, an integral Hispanism led by the Motherland. But since that has been impossible to achieve, the only thing that can be done is what we are trying to do: a self-governed Hispanism […] Provisionally self-governed, it should be understood; because neither the Mexican Academy nor the other academies from this hemisphere nor that of Malaysia have thought for one second of disregarding the authority of the Royal Spanish Academy. (Nemesio García Naranjo, Mexican academician, on April 23rd, while proposing, at the opening ceremonies, a toast to the success of the conference, quoted in Guzmán 1971: 1375)
10
In 1936, during Spain’s Second Republic (1931–1939), the democratically elected leftist government of the Frente Popular came under siege as a result of an attempted military coup. Resistance by forces loyal to the Republic as well as by significant sectors of Spanish society was initially successful and a three-year civil war ensued that ended with the insurgents’ victory in 1939. While Spain’s political regime became a right-wing military dictatorship led by General Francisco Franco, a Republican Spanish government was established in exile. It was referred to by Francoists as ‘the Red government’. Initially Franco’s Spain was excluded from the United Nations, but beginning in late 1950 diplomats succeeded at breaking the isolation and various countries began to formalize relations with Spain. Mexico, which had been the first to recognize the legitimacy of Spain’s Republican government in exile, refused to establish diplomatic relations with Franco’s government (in fact, it would only send an of ficial diplomatic mission to Spain in 1977, two years after Franco’s death).
184
José del Valle
7 The polemic during the conference However, the undercurrent of tension and the symbolic power of the event were too high to be contained. During the conference’s first plenary session, on 27 April, Mexican writer and academician Martín Luis Guzmán (1887–1976), with a passionate but carefully crafted speech, forced his fellow academicians to face the problem and take a stand. The Spanish Academy’s absence, he maintained, was, first, an insult to Mexico and all other Spanishspeaking countries; second, a f lagrant violation of the statutes that, since 1870, had regulated the relationship among the academies;11 and, third, an irresponsible act that jeopardized the Academies’ mission of protecting the nature and integrity of the language. If the problem created by the Spanish Academy were not remedied, he asked, ‘what hope can we have that our indispensible cooperation will prevent the danger that already looms on the horizon: the disintegration of Spanish?’ (Guzmán 1971: 1377). We would risk, he went on, ‘because of Spain and Mexico today or any other two tomorrow, that practical and cooperative unity […] be rendered impossible; that this dangerous fragmentation become chronic’ (Guzmán 1971: 1380). He argued forcefully against the notion that linguistic unity could only be safeguarded through the current hierarchical model in which the Spanish American and Philippine academies were treated as simple subsidiaries. He claimed that, in fact, the opposite was true: real unity could only be achieved through a dignified agreement among equals that imposed on no one a humiliating oath reminiscent of feudal relations and colonial ties: [T]he unity that you are trying to defend does not exist. But the one that we should be advancing has already begun to emerge; the one that would spring from an honorable agreement between equals, between peers; not one on the basis of an oath of service, a humiliating pledge after the existence of feudal lords ended with the end of feudalism and after our status as colonies ended with the end of the Spanish empire. (Guzmán 1971: 1383)
11
Article 11 of the 1870 statutes stated: ‘Since the purpose for which the Associated Academies are created is purely literary, their association with the Spanish Academy must be isolated from any political objective and, consequently, independent from the actions and relations among the respective governments’ (Zamora Vicente 1999: 363).
Linguistic History and the Development of Normative Regimes
185
Guzmán maintained that, through its actions, the Spanish Academy had exposed the weakness of the system and, most importantly, surrendered its credentials to act as the cultural leader of the Spanish-speaking world. He concluded by proposing that the Mexico Conference adopt the following resolution: 1 We recommend that the American and Philippine academies, af filiates of the Royal Spanish Academy, renounce their association with the latter […] and fully assume the autonomy from which they should not abdicate and their full inalienable personality […] 3 We also recommend that […] as soon as they reconstitute themselves as autonomous entities, steps be taken to summon representatives from each and all of them, as well as the Royal Spanish Academy, to meet in Mexico or any other capital city in this continent in order to create, now as equals, a clear, egalitarian, and fruitful association. (Guzmán 1971: 1387)
The discussion that ensued was spirited and, at the end, left two motions on the table: some endorsed the outright rejection of Guzmán’s proposal while others suggested that it be sent to a special committee for further consideration. The Philippines abstained, four delegations – Guatemala, Panama, Paraguay and Uruguay – voted to allow further discussion, and a wide majority of thirteen voted to kill the initiative.12 While the debate triggered by Guzmán’s speech was undoubtedly the biggest controversy of the conference, it was not the only one. New sparks re-ignited the polemic several times as other academicians put forth initiatives that, as in the first debate, were seen by others as threats to linguistic unity. Among those proposals were, for example, the creation, as a joint ef fort by all academies, of a new grand dictionary of Spanish and the production – again, with the consent and participation of all – of a new grammar that would help unify the teaching of the language throughout the continent (Comisión Permanente 1952: 387–990 and 407–416).
12
A detailed description of the debate and vote can be found in Comisión Permanente (1952: 380–383).
186
José del Valle
8 Metalinguistic discourses and linguistic history This necessarily cursory review of the events shows that the Mexico conference ignited an explosion of metalinguistic discourses in which the nature and unity of Spanish as well as the proper organization of the institutions charged with supporting it were central themes. The conference provides us, therefore, with a wealth of precious materials crucial to the reconstruction and interpretation of Spain’s and Spanish America’s recent sociolinguistic history. These materials are, of course, dif ferent in nature from the type of data on which conventional approaches to historical linguistics usually rely. In contrast with the centrality of language structure to the interests of historical pragmatics or socio-historical linguistics – and I select these two because of their explicit concern with context – metalinguistic discourses are central to a dif ferent kind of historical take on language, one that assumes the sociopolitical origins and ef fects of linguistic phenomena and hopes to unwrap the sociopolitical entanglements of the normative regimes in relation to which linguistic practices are deployed.13 By taking this perspective, we search for the ideological character of the linguistic representations that emerged in Mexico. We see them as producing, reproducing, and even challenging a normative regime; as historically situated conceptions of language in general and of Spanish in particular, as enactments of some sort of collective order, and as traces of the cultural, political and social interests involved in disputes over the production and reproduction of a normative regime in which Spanish and its varieties are assigned dif ferential values.14 13
14
With the label ‘normative regime’ I refer to a system of implicit and explicit rules and representations that attach dif ferent social values to dif ferent linguistic forms within a definable, socially structured human community. Context – social conditions and power relations – is crucial in the determination of how a normative regime comes to be as it is and how it changes – or does not change – over time. To a great extent, this approach is consistent with the premises and objectives described in Blommaert 1999. The view of linguistic ideologies that I embrace is presented in detail in Woolard 1998, Kroskrity 2000, and Gal and Woolard 2001.
Linguistic History and the Development of Normative Regimes
187
9 Spanish and the Academies: A disputed normative regime We saw, for example, that the conference was organized on the premise that the nature and integrity of the language needed protection – this was repeatedly stated during the preparatory stages and opening ceremonies as well as in the course of the conference itself. The fear of linguistic deterioration was not by any means a new concern – and it was not, of course, an exclusively Spanish apprehension. It was a historically situated manifestation of one particular form of verbal hygiene (Cameron 1995), a discourse on language closely related to what James and Lesley Milroy have termed the complaint tradition (1999: 24–46). Throughout the nineteenth century, greater political protagonism by the working classes and increased social mobility were considerable sources of anxiety for the bourgeoisie in general and for the lettered elite in particular (González Stephan 1995, Ramos 1993). Similarly, the greater cultural prominence of countries such as France – which, as I indicated above, had become beacons of progress for Spain’s former colonies – was a disconcerting reminder of Spain’s relatively low status in Europe’s cultural hierarchies. These social and cultural concerns made their way to language, which became a discursive and political site where those anxieties were worked out. The undesirable presence of dialectal and foreign forms in textual genres thought to be traditionally safe from such invasions – such as the appearance of dialectal forms and Gallicisms in Spanish and Spanish American literature – was the principal trigger of fears of linguistic deterioration.15 15
The prologue to the Venezuelan Andrés Bello’s 1847 grammar of Spanish for Spanishspeaking Americans of fers a good example of how these fragmentation anxieties played themselves out. Similarly, the controversy between Colombian philologist Rufino José Cuervo and Spanish writer and diplomat Juan Valera – which developed between 1898 and 1905 – was grounded in the Colombian’s prediction that, due to the proliferation of dialectal forms and foreign loans in literature and educated speech, Spanish would eventually split into a multiplicity of dif ferent languages (see Del Valle 2002).
188
José del Valle
In the middle of the twentieth century, as we saw, in the representations of language that emerged around and within the Mexican conference, similar concerns were voiced about the purity of the language. But the context, the circumstances invoked as sources of linguistic deterioration, had taken a new shape: on the one hand, the mass media had now replaced dialectal writers as the main disruption to the internal sociolinguistic hierarchy of Spanish and, on the other, Anglo-Saxon culture – especially the United States’ post-war prowess and cultural sway – had now replaced France as the main threat to the cultural integrity of Hispanic civilization. In keeping with the conference’s objectives, the need to protect the purity of Spanish and the strategies to do so were, of course, preferred objects of discussion – eight out of the thirty-one policy resolutions approved had an explicitly protectionist orientation. But, as we saw from the outset, the discourse of language endangerment went beyond the standard forms of purism. In fact, the need to protect Spanish was linked to fears of fragmentation that haunted academicians just as they had haunted Spanish and Spanish American men of letters at least since the middle of the nineteenth century. In those days, these fragmentation anxieties had been deeply entangled with the crises produced by the fall of the Spanish empire and the nation-building projects undertaken not only by the former colonies but by Spain itself. Still, the focus of the discourses associated with those fears was predominantly linguistic: if dialectal forms were to percolate to the speech of the educated in each Spanish-speaking country – the argument went – the language would soon meet the same destiny as Latin and evolve into a number of related but independent tongues. These arguments, however, had practically no echo in 1951. In fact, one can legitimately conclude that, by this date, although the language of fragmentation was still in play, the threat of actual linguistic divergence was no longer a serious concern – we might recall at this point Menéndez Pidal’s words (‘Although this language community may be strained, it is essentially unbreakable’). What we witness instead is a fractal projection of disintegration anxieties from language itself to language academies. In this new reincarnation of the fragmentation discourse, the nature and unity of the pan-Hispanic linguistic field is threatened not by the possible divergent evolution of linguistic forms but by alternative – and contradictory – conceptualizations of the body politic of the language.
Linguistic History and the Development of Normative Regimes
189
On one hand, Guzmán’s opponents believed the current institutional arrangement to be the adequate framework for defending the nature and unity of the language. In keeping with the spirit of panhispanismo, these Spanish American academicians accepted a language community built under Spain’s tutelage. There was to be no questioning of the Royal Academy’s authority, no challenging of its dictionary and its grammar’s value as the only tools of standardization. Martín Luis Guzmán, however, stood against the traditional model of interacademic relations and the type of panHispanic community that it mirrored. As for all academicians attending the conference, for Guzmán, Spanish was in need of protection as the linguistic, cultural, and political inf luence of Mexico’s powerful northern neighbour loomed in the horizon – a very literal and not at all distant horizon for Mexicans. But a unified institutional approach to this defensive strategy was only possible if the institutions involved renounced a relationship that the Royal Academy’s absence had revealed as tarnished by the imprint of bygone colonial hierarchies. The competent defense of Spanish, claimed Guzmán, had to be grounded in an institutional arrangement in which all Spanish-speaking nations converged as equals.
10 Conclusion To conclude, although the Mexico conference had little immediate impact on the nature of standard Spanish, the debates over the standardization process triggered by Spain’s absence and Guzmán’s proud and lucid voice would in fact end up having a significant ef fect not only on how the Academies relate to each other but also, and most prominently, on how the Academies conceive the norm. We will leave that for future articles, hoping, however, to have convincingly argued for a political and historical approach to language that illuminates, in our case, not only Spanish’s history and present shape but also the complex regimes of linguistic normativity within which it has been inscribed.
190
José del Valle
References Alfón, Fernando. 2008. ‘Los orígenes de las querellas sobre la lengua en Argentina.’ In: González, Horacio (comp.). Beligerancia de los idiomas: un siglo y medio de discusión sobre la lengua latinoamericana. Buenos Aires: Colihue. 43–77. Álvarez de Miranda, Pedro. 1995. ‘La Real Academia Española’. In: Seco, Manuel, & Gregorio Salvador (coords.), La lengua española, hoy. Madrid: Fundación Juan March. 269–279. Anderson, Benedict. 1983. Imagined Communities: Ref lections on the Origin and Spread of Nationalism. London & New York: Verso. Bello, Andrés. 1981. [1847]. Gramática de la lengua castellana destinada al uso de los americanos. In: Obras completas. Vol. 4. Caracas: La Casa de Bello. 1–382. Blommaert, Jan (ed.). 1999. Language ideological debates. Berlin: Mouton de Gruyter. Cameron, Deborah. 1995. Verbal hygiene. London: Routledge. Comisión Permanente del Congreso de Academias de la Lengua Española. 1952. Memoria del Primer Congreso de Academias de la Lengua Española, celebrado en México del 23 de Abril al 6 de Mayo de 1951. México: Academia Mexicana de la Lengua. Cuervo, Rufino J. 1899. ‘Prólogo.’ In: Soto y Calvo, Francisco, Nastasio. Chartres: Durand. vii–x. De la Puente Apezechea, Fermín. 1873. ‘Academias americanas correspondientes á la Española.’ In: Memorias de la Academia Española 4. Madrid: Rivadeneyra, 274–289. Del Valle, José. 2002. ‘Historical Linguistics and Cultural History: The Polemic between Rufino José Cuervo and Juan Valera.’ In: Del Valle, José, & Luis GabrielStheeman (eds), The Battle over Spanish between 1800 and 2000: Language ideologies and Hispanic intellectuals. London: Routledge. 64–77. Del Valle, José. 2006. ‘US Latinos, la hispanofonía and the language ideologies of high modernity’. In: Mar-Molinero, Clare & Miranda Stewart (eds), Globalization and language in the Spanish-speaking world: Macro and micro perspectives. London: Palgrave Macmillan, 27–46. Del Valle, José. 2008. ‘The Pan-Hispanic community and the conceptual structure of linguistic nationalism.’ International Multilingual Research Journal 2, 5–26. Del Valle, José, & Luis Gabriel-Stheeman (eds). 2002. The Battle over Spanish between 1800 and 2000: Language ideologies and Hispanic intellectuals. London: Routledge.
Linguistic History and the Development of Normative Regimes
191
Gal, Susan, & Kathryn A. Woolard. 2001. ‘Constructing languages and publics: Authority and representation.’ In: Gal, Susan, & Kathryn A. Woolard (eds), Languages and publics: The making of authority. Manchester: St Jerome. 1–12. González Stephan, B. 1995. ‘Las disciplinas escriturarias de la patria: constituciones, gramáticas y manuales.’ Estudios 3/5, 19–46. Guzmán, Martín Luis. 1971. Obras completas. México: Compañía General de Editores. Haugen, Einar. 1972. The Ecology of Language. Stanford, CA: Stanford University Press. ‘Información Académica.’ 1950. Boletín de la Real Academia Española XXX, 480–481. Kroskrity, Paul V. 2000. ‘Regimenting Languages: Language Ideological Perspectives.’ In: Kroskrity, Paul V. (ed.), Regimes of language: Ideologies, polities, and identities. Santa Fe, NM: School of American Research Press. 1–34. Lázaro Carreter, Fernando. 1949. Las ideas lingüísticas en España durante el siglo XVIII. Madrid: Espejo. Lázaro Carreter, Fernando. 1980. Estudios de lingüística. Barcelona: Crítica. Menéndez Pidal, Ramón. 1918. ‘La lengua española.’ In: Hispania 1. 1–14. Milroy, James, and Lesley Milroy. 1999. Authority in Language: Investigating Standard English. 3rd edn. London: Routledge. Pagano, José León. 1951. ‘Primer Congreso de Academias de la Lengua Española. Informe del Señor académico don José León Pagano.’ Boletín de la Academia Argentina de Letras XX/76, 249–262. Pike, Fredrik. 1971. Hispanismo, 1898–1936. Notre Dame, IN: University of Notre Dame Press. Rama, Carlos M. 1982. Historia de las relaciones culturales entre España y la América Latina. Siglo xix. México: Fondo de Cultura Económica. Ramos, Julio 1993. ‘El don de la lengua.’ Casa de las Américas 34, 13–25. Rosenblat, Ángel. 1960. ‘Las generaciones argentinas del siglo xix ante el problema de la lengua.’ Revista de la Universidad de Buenos Aires (Quinta época) 4/4, 539–584. Sarmiento, Domingo Faustino. 1949 [1843]. ‘Memoria sobre ortografía americana.’ In: Obras completas, vol. 4. Buenos Aires: Luz del Día, 1–50. Sepúlveda, Isidro. 2005. El sueño de la Madre Patria. Hispanoamericanismo y nacionalismo. Madrid: Marcial Pons. Woolard, Kathryn A. 1998. ‘Introduction: language ideology as a field of inquiry.’ In: Schief felin, Bambi B., Kathryn A. Woolard, & Paul V. Kroskrity (eds), Language ideologies: Practice and theory. New York: Oxford University Press, 3–47. Zamora Vicente, Alonso. 1999. Historia de la Real Academia Española. Madrid: Espasa.
Juan R. Valdez
Colouring Language: Pedro Henríquez Ureña’s Representations of Spanish and Dominican Identity1
Abstract The Latin American linguist and philologist Pedro Henríquez Ureña raised significant and inf luential questions about the origin and development of Spanish in the New World. However, most assessments of his work tend to ignore the fact that these questions were part of fundamental ideological debates. Henríquez Ureña’s work on Dominican Spanish emerged in a political climate in which an instrumentalist approach to language and history became the standard practice for Dominican intellectuals. This practice authorized images of Spanish as representative of the Dominican national character and a specific type of Dominican Spanish that had to be protected from pernicious external forces. Dominican Hispanism strongly reemerged in this context as a political and cultural ideology based on the concept of the Dominican national identity as pure, white and Spanish. In Henríquez Ureña’s studies of Dominican Spanish, linguistic forms become linked to some of these racial and cultural categories. My analysis (inspired by critical linguistic historiography and language-ideological research) takes advantage of several methodological tools available for interdisciplinary research in order to fully understand the sociohistorical meanings and implications of Henríquez Ureña’s linguistic texts.
1
I thank the organizers for the opportunity to participate at the 2009 Bristol conference and my colleagues José Del Valle and Laura Villa for their insightful comments, questions, and feedback. Segments of this analysis appeared in a previous publication (Valdez 2009); it is part of a wider research project whose complete results appear in my book, Tracing Dominican identity: the writings of Pedro Henríquez Ureña (2011).
194
Juan R. Valdez
1 Introduction During defining moments, language has figured prominently as a subject of history in Spain and Latin America. For example, in order to unify the diverse regions of Iberia under one kingdom, King Alfonso X (1221–1284) gave priority to the standardization of written Castilian, an ef fort that led to the first orthography of any Romance language (circa 1270). Antonio de Nebrija (1441–1522), in the prologue to his Castilian Grammar (1492), addressed to Queen Isabel of Spain (1451–1504), advanced the conclusion ‘that language always accompanies empire’, thus persuading the Spanish Crown to pay particular attention to the relationship between language and power. When the Spanish empire had begun to disintegrate a few centuries later, the Venezuelan-born Andrés Bello (1781–1865), concerned with the uncertain future of the newly independent Latin American nations, proposed his Grammar, intended for Latin Americans, as a device that could help check the dynamics of cultural, dialectal, and imperial fragmentation. Finally, after Spain lost its last colonies and entered a cultural crisis, the Spanish intellectual Ramón Menéndez Pidal (1869–1968) produced Orígenes (1926) in which he chronicles the historical conditions that in his view initially allowed the Castilian form of Spanish to dominate. In the Spanish-speaking world, at dif ferent times and in dif ferent contexts, language historians have selected and prioritized only those sociolinguistic facts which give centre stage to some specific images and constructions of the Spanish language. Yet, rarely do they ref lect upon the specific historical complications that had to be overcome or erased in the establishment of a particular linguistic hegemony. It is up to the linguistic historiographer to answer questions such as why and how certain images or visions of the language f lourished while others conspicuously disappeared or withered away. This task can only be accomplished by stepping outside the boundaries of autonomous linguistic research. Pedro Henríquez Ureña (1884–1946), a key figure in Hispanic linguistics and philology, and his work have been the objects of numerous studies. However, to this day his work has not received the critical attention
Colouring Language
195
it deserves (Sarlo 2000). This lacuna is even more obvious in the studies related to his linguistic production. Among Henríquez Ureña’s most important linguistic work, we find first Observaciones sobre el Español en América (1921), El supuesto andalucismo de América (1925) and La gramática castellana (1931), co-authored with Amado Alonso. A linguistic-historiographical approach reveals that his texts abound with some curious ideas that were prevalent in his professional circle and which belonged to what Koerner (1995: 17) calls the ‘intellectual climate’. An intellectual climate is defined by the professional conditions of production, the period’s intellectual currents, and the immediate socioeconomic and political situation. Henríquez Ureña’s seemingly contradictory representations of Latin American Spanish (‘polygenesism’) and Dominican Spanish (‘archaicism’) underscore notions of language, race, and nation that emerged from specific sociopolitical circumstances and struggles. In his classic text, El español en Santo Domingo (1941), Henríquez Ureña advanced the idea that what characterizes the Dominican variety of Spanish is its vocabulary and syntax, the most archaic in the continent. At the same time, the author consistently argued for the absence of African substrate elements in the speech and culture of his country. A language-ideological analysis2 reveals that Henríquez Ureña’s characterization of Dominican Spanish emerges in a specific context, in the debate and struggle for the construction of postcolonial Santo Domingo. Building on the definition of language ideology, of fered by José del Valle (in this volume), as ‘a historically situated conception of Spanish as an enactment of a collective order in which Spain performs a central role’, the purpose of my study is twofold: a) to demonstrate the ways in which these sociopolitical phenomena surface in the construction of Henríquez Ureña’s ‘archaic’ Dominican Spanish; and b) to show how his language representation is intricately connected to historical problems, i.e. the emergence and development of the Dominican state.
2
My study is consistent with this particular line of research defined in Crowley (1996), del Valle and Gabriel-Stheeman (2002), Irvine and Gal (2000), Joseph and Taylor (1990), and Woolard (1998).
196
Juan R. Valdez
2 Language and history in the Dominican Republic Although Spanish is undoubtedly dominant, in the Dominican Republic there are speakers of languages other than Spanish and a significant number of people whose sociolinguistic competence is better characterized as multilingual, multidialectal, or complex repertoire. Contrary to the claims of one analyst (Núñez 2003), it is the complex speech communities and vibrant contact zones which linguistically define this little half-island nation. For example, in the Samaná Province we find speakers of English, descendants of African-American ex-slaves, invited by Haitian president Jean-Pierre Boyer (1776–1850) in 1824, and descendants of English-speaking West Indians who migrated to the Dominican Republic in the late nineteenth and early twentieth centuries. We also find Dominicans who speak a variety of Haitian Creole, locally identified as ‘Patois’, and a large number of Haitian immigrants and their Dominican-born descendants who are bilingual to dif ferent degrees. In sharp contrast to the culture of monoglossia3 and monoethnic ideologies which have historically dominated the development of national identity in the Dominican Republic, this multidialectal and multilingual repertoire has been glossed over, minimized or erased in almost all accounts of Dominican society and history. Manuel Núñez (2003 and 2001), a linguist and member of the Dominican Academy of the Spanish Language, who has become the intellectual spokesperson for ‘anti-Haitianism’, provided a recent example of this monoglossic culture. In justifying his commitment to support and protect the Spanish language, he declared language as the first line of defence in harnessing the Dominican Republic’s identity, 3
According to del Valle (2000: 119–120), the culture of monoglossia, in which the dominant discourse on language is grounded, consists of the principle of focused grammar or the assumption that what linguistically characterizes individuals as well as a community is possession of a well defined and relatively stable grammar and the principle of convergence, or the assumption that people’s linguistic behavior tends to become homogeneous over time through pressure from the dominant norm of the community.
Colouring Language
197
and that the country was not bilingual. He wrote: ‘only the Hispanic component is truly Dominican culture’ (Núñez 2001: 234). Dominicans are not alone in this form of erasure: modern ideologies of the nation advance powerful homogenizing tendencies. However, cultural historian Odalís Pérez (2002) documented the development and adoption by the State of a specifically Dominican nationalist ideology that demands, first, the Hispanization and whitening of Dominican society, second, the adoption of the dominant elites’ cultural behaviour and narratives; and, third, the erasure of cultural features associated with the popular classes. Composed of various ideas and proposals that centred on the historical and cultural bonds between Spain and the Spanish-speaking countries in America, the subject of Dominican Hispanism merits a fuller discussion, but here a summary will suf fice: one of the branches of Hispanism, advanced by both Spanish and Latin American conservative and liberal politicians and intellectuals, defends the formation of a transatlantic Spanish community or identity that justifies Spain’s cultural presence, commercial interests, and foreign policy in Latin America.4 These particular Hispanists proposed a shared (idealized) identity of the Peninsular and American people, based on a cultural community comprised of historical, linguistic, philosophical, and religious elements that underlie the constitutions of the respective states and transcend the political divisions that followed the wars of independence. It is this particular vision of the Pan-Hispanic community that underlies the Dominican brand, amplified to accommodate the specific anti-Haitian ideology that emerged during the foundation of the Dominican state. Elaborated over a long period of social, economic, and political struggles, this of ficial ideology and its related policies were grounded in an af firmation of Dominicanness as the opposite of all black things Haitians symbolized: within this identity framework, Dominicans are, by definition, everything that Haitians are not. Ties with Haiti are denied, especially any tie that binds both parts of the island through an African heritage. The particular circumstances surrounding the birth of the Dominican state 4
See del Valle (in this volume) and Sepúlveda (2005).
198
Juan R. Valdez
shed light on the development and evolution of this ideology. In founding their nation, a group of leading Dominican political figures sought independence from the political jurisdiction of Haiti, which, having gained its independence in 1804, retained control of Santo Domingo until 1844. This dif ficult process of independence from Haiti and the various military attempts of Haitian leaders between 1844 and 1855 to restore their rule in the Spanish-speaking half of the Hispaniola contributed to a nationbuilding ideology that produced a powerful element of self-identification in opposition to Haitians. Already in 1822, years before independence, José Núñez de Cáceres (1772–1846), a local political leader, stated while handing the keys to the city of Santo Domingo over to Jean Pierre Boyer that the Haitian plan for the political unification of the entire island would inevitably fail. According to Rodríguez Demorizi (1975: 17), there and then Núñez de Cáceres declared that the Spanish language alone ‘was a separation wall, as natural and insurmountable as were the Alps or the Pyrenees’.5 Since that early stage and throughout the many phases in the Dominican Republic’s process of nation-building, language and race have been deployed as strategic discursive sites where Dominicanness and Dominicans’ linguistic behaviour become racialized. Historically, many of the authors that contributed to the dominant ethnopolitical discourse have represented the Dominican nation through the lens of race and language.6 Consistently, they envision the Dominican Republic as a homogeneous community without contradictions or adulterations. Along with race, language was and remains the crucial element in the construction of this identity (Bailey 2002 and Toribio 2005). The link between language and national identity is a recurrent theme in nationalist discourses which frequently appears in public discussions of Dominican identity. Time and again, questions concerning language are raised in order to address and, in some cases, ignore larger social issues and problems.
5 6
All translations in this chapter are mine. See San Miguel (2005).
Colouring Language
199
3 Colouring Dominican Spanish Henríquez Ureña’s El Español en Santo Domingo (ESD, from here on), for many years the most comprehensive work on Dominican Spanish, is still a central landmark in any investigation of Dominican and Caribbean dialectology. In this book, Henríquez Ureña advanced the thesis that the vocabulary and syntax of Dominican Spanish is ‘the most archaic’ in the American continent. In chapter three, the author postulated the argument that the archaic nature of the lexicon is unmistakably demonstrated by the widespread use in Dominican speech of expressions that became obsolete in other Spanish varieties. Dominican Spanish, Henríquez Ureña af firmed, remains faithful to its Spanish roots. At the end of chapter four, he posed the following question: are the numerous parallels between the contemporary Dominican lexicon and the Castilian lexicon of the sixteenth and seventeenth centuries suf ficient to support the claim that Dominican Spanish is a derivative of the Castilian Spanish that was spoken in seventeenth-century Salamanca? No, answered Henríquez Ureña (1941: 54). However, this comparison became his point of departure and is implied throughout the book. His description of the lexicon in chapter five contains words such as: aguaitar; bregar; despacharse; dilatarse; guayar; prieto; privar en; tíguere; tostón. The author explained that words like, for example, aguaitar [‘to watch; to look’], were commonly used in the Dominican Republic, after falling out of use in other Spanish-speaking regions where educated people no longer used them. Another example is dizque [‘apparently’] which, according to Henríquez Ureña (1941: 56), was no longer used in most of Spain or Argentina. He claimed that, at the time he was writing ESD, even the Spanish Royal Academy, or rather the lexicographers working on its dictionary, were unfamiliar with many of these Dominican words (1941: 90). For his description of the lexicon he selected a series of words documented in Dominican Spanish, many of them certainly archaic, and he attributed to these words the meaning given to them by the original Spanish colonizers. Henríquez Ureña’s descriptive strategy consisted of connecting typically
200
Juan R. Valdez
Dominican words to their Spanish Golden Age origin. In and of itself, this lexicographic procedure can be perceived as ideologically neutral, yet closer attention and further scrutiny soon shows Henríquez Ureña linking the archaic lexicon to Dominicans’ loyalty to the Hispanic cultural traits and by extension to the purported whiteness of Dominicans. A careful reading of chapters five and six of ESD reveals the discursive construction of a socio-historic image that links the Dominican lexicon to the original Spanish settlers and their linguistic past. In fact, in order to make this colonial connection stronger, Henríquez Ureña advised that ESD should be read along with the supplementary text La cultura y las letras coloniales en Santo Domingo (1936), a text whose cultural and historical facts about Spanish colonialization would help explain and solidify the linguistic ones. We find further evidence of the link between linguistic description and extra-linguistic cultural and racial categories in Henríquez Ureña’s treatment of certain aspects of the phonology of Dominican Spanish. The phonological phenomenon in which both postvocalic /r/ and /l/ are pronounced as a glide, thus, realizing words such as comer ‘to eat’ and porque ‘because’ as [ko.méi] and [pói.ke] has been documented as characteristic of a variety from the northern part of the country known as cibaeño. In describing this peculiar and divergent pronunciation, he highlighted the fact that the phenomenon is also found in rural areas of Puerto Rico where, he emphasized, the people are generally white (Henríquez Ureña 1941: 49). In chapter twelve, the book provides another example of how Henríquez Ureña used linguistic dif ferences to convey and mediate a social and ethnolinguistic identity that is, above all, white. He described a phonological process in which the intervocalic consonant [r] is pronounced as [l]: words such as cerebro ‘brain’ are pronounced as [se.lé.bro]. Earlier scholars thought they had discovered a possible African origin for this feature. Esteban Pichardo, for example, the Dominican-born Cuban lexicographer, had attributed it to the bozales (African-born second language learners of Spanish). Henríquez Ureña, after stating that the phenomenon was generally rare, responded to claims of its African origins by noting that it is never found among blacks that had been born on the island. These, he insisted ‘speak like white natives’ (Henríquez Ureña 1941: 168). Presenting what he thought was insurmountable evidence against the African origin
Colouring Language
201
of this feature, he stated that the phenomenon was also found in Spain where African inf luence was dif ficult, if not impossible, to explain. To summarize, first Henríquez Ureña minimizes the scope of this phonological phenomenon, second he located it within an exclusively Hispanic linguistic tradition, and third he af firmed the Hispanization and whitening of the black Dominican through the acquisition of white speech. This process of strictly associating sociocultural traits with linguistic phenomena is known by linguistic anthropologists as ‘iconization’.7 In Pedro Henríquez Ureña’s linguistic texts as well as in the work of numerous scholars, we find the iconization of Dominican Spanish representing Hispanicness as the only relevant component of Dominican culture.
4 Erasing the black In addition to establishing the iconic relationship between linguistic forms, hispanicity and whiteness among Dominicans, Henríquez Ureña took on the task of contesting the existence of an African cultural and linguistic heritage. In a polemic against the Neogrammarian philologist Wilhelm Meyer-Lübke’s claim of the existence of a ‘black’ dialect of Spanish in the Dominican Republic, Henríquez Ureña (1919: 430) wrote: The black race has never predominated [in the Dominican Republic] and the Castilian language remains pure. There has never been, nor is there now, a black dialect in the Republic. To the contrary, Santo Domingo belongs to that segment of America where the language remains closer to its Castilian roots.
7
Irvine and Gal (2000: 37) defined ‘iconization’ as semiotic and discursive processes by which some aspect of or feature in the linguistic repertoire of a given community is perceived as an iconic representation of its members, as if a linguistic feature could inherently ref lect the nature of a social group. They also defined the concept of ‘erasure’ as the process that ‘renders some people or activities invisible in order to simplify the sociolinguistic field’.
202
Juan R. Valdez
For the observer acquainted with Dominican society and history, this is indeed a surprising minimization if not an all-out erasure of Africanness as a component of Dominican identity. The same minimization emerged in ESD when Henríquez Ureña (1941: 130) took on two of his contemporaries, the Cuban Fernando Ortíz and the Dominican Carlos Larrazábal Blanco, questioning their theories of African inf luence in Cuba and the Dominican Republic. Citing demographic figures from early colonial chroniclers such as Bartolomé de Las Casas (1484–1566) and Gonzalo Fernández de Oviedo (1478–1557), Henríquez Ureña (1941: 132–133) maintained that: In Santo Domingo African inf luence is very scarce […], the fact is that there was no systematic importation of slaves during the first half of the sixteenth century […]. In the nineteenth century, there was even less […] and from the beginning a good number of the slaves did not arrive directly from Africa: It is a fact that they arrived from Spain where they had been bought from the Portuguese; they were already Hispanized. Interestingly enough; the first slaves were not all black; white slaves were also brought to America. Anyway, what characterizes the population of African descent in Santo Domingo is its complete Hispanization.
The image that emerges from this sociodemographic description is of a complete colonization of the eastern part of the Hispaniola by the Spaniards who successfully restricted the arrival of African slaves and prevented their proliferation. In reference to the arrival of migrant workers from Haiti and the English-speaking West Indies in the twentieth century, a phenomenon that was forcing some to question prevailing theories of race in Dominican society, Henríquez Ureña (1941: 133) wrote: This invasion is quickly blackening the country. It is estimated that there are more than 2,000 immigrants; the total population barely reaches a million and a half. Francisco Eugenio Moscoso Puello’s interesting novel, Cañas y bueyes, portrays aspects of this invasion […] Today the recent Caribbean fad of negro poetry that f lourishes in Cuba and Puerto Rico with the works of Luís Pales Matos, Ramón Guirao, José Zacarías Tallet, Alejo Carpentier, Nicolás Guillén, Tomás Blanco, Emilio Ballagas, Marcelino Arozarena, and Vicente Gómez Kemp has reached Santo Domingo. The Negros in [the Dominican] Manuel Cabral’s Doce poemas negros (1935) are primarily Haitians or cocolos from the English-speaking West Indies, because the ones that are native to Santo Domingo have less colorful customs […] Arturo Pellerano Castro’s old and wonderful Criollas are really criollas ‘natives’; they speak of white and light-skinned women [italics in the original].
Colouring Language
203
These passages reveal the insistent presence, in Pedro Henríquez Ureña’s vision of the Dominican Republic, of the dominant ideology of national identity in which blackness is ef fectively erased. Whatever blackness exists in the Spanish-speaking part of the island, as the last quote indicates, is foreign: primarily related to Haitians or cocolos from the English-speaking West Indies.
5 Dominican politics of identity in the twentieth century Language and history inevitably converge in scenarios where people (linguists included) attempt to draw or erase political and cultural boundaries. Nineteenth century nation-building discourse, anti-Haitian rhetoric, and the anti-North American sentiment that followed the US occupation of the country all converged into the Dominican nationalism of the 1920s when Henríquez Ureña was especially intellectually active in the US and Mexico and methodically outlined his ideas on the subject of Dominican Spanish. The brand of Dominican nationalism that emerged during this particular period in the country’s history set forth a series of measures that Dominican leaders attempted to implement during, first, the Horacio Vasquez (1860–1936) presidency (1924–1930). However, the establishment and promotion of anti-Haitian and xenophobic attitudes became more vociferous and violent during the dictatorship of Rafael L. Trujillo (1891–1961). Trujillo and his government launched a complete reevaluation and reconstitution of the Dominican nation. Trujillo, his acolytes, and government of ficials became alarmed by the number of Haitian immigrants in the country, particularly the Haitians that had been long established on the Dominican side of the border. The Trujillo government (1930–1961) initiated projects to ‘Dominicanize’ the border and force out as many Haitians as possible. Not content with the pace of deportation and the failure of some of the policies, in 1937, Trujillo ordered the military to slaughter thousands of Haitians by a river that divides the two countries in the northwest, aptly named El Masacre ‘The Massacre’. In order to distinguish dark-skin
204
Juan R. Valdez
Dominicans from Haitians, a shibboleth test was applied, in which only those that could pronounce the non-lateral simple vibrant [r] in the word ‘perejil’ [parsley] instead of the uvular [R] would be spared from getting shot or hacked to death. This event drew serious international criticism which led many Dominican historians and intellectuals to of fer alternative accounts of this shameful episode and some of the reasons that led to it. This period saw the consolidation of politically conservative forces and what Dominican historian Valentina Peguero (2004: 3) calls ‘the militarization of Dominican culture’. The worst elements in the military, the bourgeois bureaucracy, the Jesuit wing of the Catholic Church, and one of the most powerful and ruthless dictatorships in Latin America’s history converged in the exploitation and domination of the Dominican masses and the legitimization of absolute despotic rule. The era of Trujillo constituted the most reactionary period in the country’s history which Dominicans still struggle to overcome today and strongly conditioned the country’s intellectual and cultural production. It is in this specific political climate that Dominican historians and philologists instituted the instrumentalist and selective approach to language and history as the standard practice. As Dominican author Manuel Matos Moquete (1986: 30) argues, this practice authorized images of the language as representative of lo dominicano ‘Dominicanness’ and ‘pure speech’. It was decreed that Dominicans were responsible for protecting their speech and nationality from Anglicisms, Gallicisms (or rather Haitian creolisms), barbarisms, and ‘deterioration’ at the hands of the lower classes. These sociolinguistic concerns coincided with the active policies of increased European immigration and the eradication of Haitians pursued by the Trujillo government, which invited white Europeans to settle in the Dominican Republic under the most favourable conditions. In this context Dominican Hispanism strongly reemerged as a political and cultural ideology based on the concept of the Dominican national identity as pure, white, Catholic, and Spanish. While not all Dominicans accepted this self-image and neocolonial status, this ideology became, and still is, dominant.
Colouring Language
205
6 Conclusion Pedro Henríquez Ureña’s linguistic analyses of Dominican Spanish go beyond the confines of plain and simple dialectology. When read in light of the intellectual and political context of its emergence, his linguistic description ultimately reveals the moving forces of history. In Henríquez Ureña’s studies of Spanish in Santo Domingo, linguistic forms are iconically mapped onto racial and cultural categories and, in their interaction, an image of the nation emerges, a linguistic ideology that ref lects the white, Hispanic identity that has dominated the discursive construction of the Dominican Republic since its historical emergence as an independent nation. While his knowledge of the linguistic reality of the island was unprecedented and his familiarity with the Spanish language and its history admirable, his texts reveal a complex interaction between the coherent dialectological description of Dominican Spanish and the ideologically driven contemporary discourses of national identity in which whitening the race was a central theme. Ultimately, in order to accomplish some of his objectives in the representation of Dominican Spanish and society, he was unable to disengage from the neocolonial regimes of truth previously established by some of his predecessors and the State apparatus. As Del Valle (2007) and Valdez (2011) have argued, the possibility of analysing such complex semiotic and ideological language phenomena emerges from approaches that go beyond the formal and material dimension of language and examine the culturally and politically situated nature of visions of language, linguistic dif ferences, and communication. Understanding the manner in which linguists’ descriptions and representations of language are related to historically salient ideologies in dif ferent eras and contexts requires an interdisciplinary approach between linguistics and historical studies. A historically situated approach to linguistic phenomena implies crossing the autonomous boundaries that representatives of each discipline sometimes zealously guard.
206
Juan R. Valdez
References Alonso, Amado and Pedro Henríquez Ureña. 1938. Gramática Castellana I & II. Buenos Aires: Losada. Bailey, Benjamin. 2002. Language, race and negotiation of identity: A study of Dominican Americans. New York: LFB Scholarly Publishing. Crowley, Tony. 1996. Language in history: Theories and texts. London: Routledge. Danesi, Marcel 1977. ‘The case for andalucismo re-examined.’ Hispanic Review 45, 181–193. Del Valle, José. 2007. ‘Glotopolítica, ideología y discurso: categorías para el estudio del estatus simbólico del español.’ In: del Valle, José (ed.), La lengua, ¿patria común? Ideas e ideologías del español. Madrid: Iberomamericana, Vervuert. Del Valle, José. 2000. ‘Monoglossic policies for a heteroglossic culture: misinterpreted multilingualism in Modern Galicia.’ Language and Communication 20, 105–132. Del Valle, José, & Luis Gabriel-Stheeman (eds). 2002. The Battle over Spanish between 1800 and 2000: Language ideologies and Hispanic intellectuals. London, New York: Routledge. Henríquez Ureña, Pedro. 1919. ‘La lengua en Santo Domingo (rectificación al romanista Meyer-Lübke).’ In: Obras completas (tomo IV). Santo Domingo: Editora Universal, 49–52. Henríquez Ureña, Pedro. 1975 [1941]. El español en Santo Domingo. Santo Domingo: Taller. Henríquez Ureña, Pedro. 1988 [1936]. ‘La cultura y las letras coloniales en Santo Domingo.’ In: Obra Dominicana. Santo Domingo: Sociedad Dominicana de Bibliófilos Inc. Henríquez Ureña, Pedro. 2003 [1925]. ‘El supuesto andalucismo de América.’ In: Obras completas (tomo IV). Santo Domingo: Editora Universal, 97–104. Henríquez Ureña, Pedro. 2003 [1921]. ‘Observaciones sobre el español en América’. In: Obras completas (tomo IV). Santo Domingo: Editora Universal, 61–85. Irvine, Judith, & Susan Gal. 2000. ‘Language Ideology and Linguistic Dif ferentiation.’ In: Kroskrity, Paul (ed.), Regimes of language: Ideologies, polities and identities. Santa Fe, NM: School of American Research Press, 35–84. Joseph, John, & Talbot Taylor (eds). 1990. Ideologies of language. London: Routledge. Koerner, Konrad. 1995. Professing linguistic historiography. Amsterdam: John Benjamins.
Colouring Language
207
Matos Moquete, Manuel. 1986. La cultura de la lengua. Santo Domingo: Biblioteca Nacional. Menéndez Pidal, Ramón. 1964 [1926]. Orígenes del español: estado lingüístico de la península ibérica hasta el siglo XI. Madrid: Espasa-Calpe. Núñez, Manuel. 2003. ‘La lengua española, compañera de la nación dominicana: discurso de ingreso.’ Boletín de la Academia Dominicana de la Lengua 17, 137–195. Núñez, Manuel. 2001. El ocaso de la nación Dominicana. Santo Domingo: Alfa y Omega. Peguero, Valentina. 2004. The militarization of culture in the Dominican Republic: From captains general to General Trujillo. Lincoln: University of Nebraska Press. Pérez, Odalís. 2002. La Ideología rota: el derrumbe del pensamiento pseudonacionalista dominicano. Santo Domingo: Editora Manatí. San Miguel, Pedro L. 2005. The imagined island: History, identity and utopia in Hispaniola. San Juan, Santo Domingo: Isla Negra, La Trinitaria. Sarlo, Beatriz. 2000. ‘Pedro Henríquez Ureña: lectura de una problemática.’ In: Abellán, José Luis and Ana Maria Barrenechea (eds), Pedro Henríquez Ureña: ensayos. Madrid: Allca XX, 880–889. Sepúlveda, Isidro. 2005. El sueño de la Madre Patria. Hispanoamericanismo y nacionalismo. Madrid: Marcial Pons. Toribio, Jacqueline. 2005. ‘Linguistic displays of identity among Dominicans in national and diasporic settlements.’ In: Davies, C, and J. Brutt-Grif f ler (eds), English and ethnicity. New York: Palgrave, 135–155. Valdez, Juan R. 2011. Tracing Dominican identity: the writings of Pedro Henríquez Ureña. New York: Palgrave. Valdez, Juan R. 2009. ‘The iconization of Dominican Spanish in Pedro Henríquez Ureña’s linguistic texts.’ Spanish in Context 6(2), 176–198. Woolard, Kathryn. 1998. ‘Language ideology as a field of inquiry.’ In: Schief felin, Bambi B., & Kathryn Woolard (eds), Language ideologies: Practice and theory. New York: Oxford University Press, 3–47.
Laura Villa
‘Because When Governments Speak, They Are Not Always Right’: National Construction and Ortho graphic Conf licts in Mid-Nineteenth-Century Spain
Abstract This paper analyses the polemics surrounding the discursive legitimation and political institutionalization of dif ferent spelling systems circulating in mid-nineteenth-century Spain. In 1843, teachers associated with Madrid’s Literary and Scientific Academy of Primary Education developed a simplified orthography and began to implement it in schools. In response to this independent initiative, Queen Isabel II signed a Royal Decree in 1844 that mandated the exclusive use of Royal Spanish Academy’s orthography in Spain’s primary education. The Literary and Scientific Academy contested the imposition and took actions to oppose its implementation, by organizing meetings and publishing essays to defend both the simplified orthography and the legitimacy of the institution. My study examines of ficial documents, textbooks, pamphlets, minutes, etc., which provide us with a record of this linguistic ideological debate that reveals a struggle over authority and power within the linguistic and educational markets.1
1
The reconstruction of this salient episode in the linguistic history of Spain is the result of documentary research carried out in Spain’s National Library and in the Library and Archive of the Royal Spanish Academy thanks to the financial support of the Program for Cultural Cooperation between Spanish Ministry of Culture and United States Universities, managed by the University of Minnesota.
210
Laura Villa
1 Introduction In the mid-nineteenth century, the orthographic and grammatical norms of the Royal Spanish Academy (henceforth RAE)2 were made of ficial. They were to be taught and used in the national school system which, at the time, was being developed in the broader context of the construction of a liberal Spanish nation (Álvarez Junco 2001; Burdiel 1998; Puelles Benítez 1999, 2004). In a larger project I look at the historical conditions that resulted in this of ficialization and the consequent promotion of the linguistic norms proposed by the RAE. The main objectives of this project are, on the one hand, to describe the processes of selection and codification of Spanish as Spain’s national language, and the implementation of the resultant standard variety through the public school system; and on the other, to analyse the increasing relevance of the RAE as the legitimate institution in charge of the standardization of Spanish, as well as the resistant voices that challenged the constitution of the RAE as Spain’s linguistic authority. The study is framed by an ef fort to understand linguistic polemics within the specific historical and cultural context of its production. This entails adopting a linguistic, political and historical approach that looks at the of ficialization of Spanish norms – as well as the discourses surrounding its implementation in the emerging public school system – as part of a broader language planning process deeply involved in tensions over nation-building and the power of the State. These tensions clearly appeared in the debate that surrounded the of ficialization of the RAE’s orthography in 1844 and its imposition as the legitimate spelling norm in Spain’s education system, which is the focus of this paper. The nineteenth century witnessed a number of similar orthographic debates over languages such as Czech,3 2 3
This language institution was created in 1713 after the Académie Française in order to ‘fix the voices and vocabularies of the Castilian language with propriety, elegance, and purity’ (). Czech’s spelling reforms in the nineteenth-century where consistent with a project of linguistic and national revival. The reforms, guided by the Matice česká, a selfappointed linguistic institution in charge of Czech corpus planning, found the opposition of important Czech linguists (Bermel 2007: 95–105).
‘Because When Governments Speak, They Are Not Always Right’
211
Dutch,4 German,5 Spanish,6 and Swedish, Danish and Norwegian,7 to name just a few. These orthographic controversies can be understood as language ideological debates because, one the one hand, in controversies over the spelling system ‘language is central as a topic, a motif, a target’ and ‘language ideologies are being articulated, formed, amended, enforced’ (Blommaert 1999: 1); and on the other, these orthographic debates ‘are part of more general sociolinguistic processes, […] sociopolitical developments, conf licts and struggles’ (2). This model allows us to emphasize the significance of the social and political contexts, as well as the ‘historical horizon of relationships of power, forms of discrimination, social engineering, [and] nation building’ (2), in which these linguistic controversies emerged. The analysis of the polemics surrounding the legitimation of competing spelling systems in mid-nineteenth-century Spain will show that Spain’s orthographic debate ref lected larger socio-political controversies in the educational and national contexts. The standard language, codified 4
5
6
7
Rutten & Vosters in this volume study orthographic practices and discourses on orthography in nineteenth-century Flanders. I would like to thank Rik Vosters (VU Brussels) for a fruitful exchange of ideas and materials on the parallels, dif ferences and theoretical approaches to our objects of study. German’s lack of consensus resulted in the coexistence of several orthographic schools in dif ferent regions until the nineteenth century. The first attempt of orthographic standardization took place in 1876, right after German political unification ( Johnson 2005: 19–22). In addition to the controversial of ficialization in Spain that I analyse, there were simultaneous orthographic debates in other Spanish-speaking countries, such as Chile, Colombia and Mexico, where the cultural separation from the linguistic authority of Spain – the former metropolis – plays a central role. The Chilean orthographic reform – planned by Andrés Bello (1781–1865) and approved, also in 1844, amidst an intense public debate – is central to my wider project for its many connections with Spain’s debate: besides a number of biographical links among the Chilean and Spanish orthographic agents, both spelling disputes shared a crucial tie to intense nation-building processes in which the education system was seen as a central instrument of the State to the promotion of the of ficial orthography (Contreras 1993; Narvaja de Arnoux 2008; Rosenblat 1951). The Scandinavian Orthographic Conference in Stockholm in 1869 recommended some reforms to be implemented in these three languages (Dalen 2005: 1406; Torp 2005: 1427–1431). I would like to thank Olle Josephson (Stockholm) for directing my attention to these Scandinavian orthographic debates.
212
Laura Villa
by and for the school apparatus, constitutes a special linguistic practice in which socio-political antagonisms are produced and reproduced (Balibar & Laporte 1976: 21). These were projected onto the discursive representations constructed by the actors involved in the standardization processes and related educational debates. In consequence, my analysis hopes to unveil the non-linguistic dimensions of the debate over the of ficialization of the RAE’s spelling system, by examining the linguistic ideologies proposed, exploited and reformulated in favor of and in opposition to of ficialization. I rely on the concept of linguistic ideologies as an analytical tool because of its potential, first, to explain ‘the tie of cultural conceptions to social power’ (Woolard 1998: 10), and second, to integrate the domains of language and politics, emphasizing the socio-political and economic interests embedded in linguistic matters (Kroskrity 1999: 2–3). In my analysis I focus on the discursive strategies of formulation, exploitation and reformulation of linguistic representations, that is ‘systems of ideas that articulate notions of language with specific cultural, political and/or social formations’ (Del Valle 2007: 22). These linguistic representations are ‘related to practices and ideas that naturalize and normalize a particular socio-political order and legitimate a certain type of knowledge that supports the exercise of power and authority’ (Del Valle 2007: 22). The of ficialization of the RAE’s spelling system was followed by an intense debate between a group of primary school teachers that proposed and defended a simplified orthography in order to universalize literacy, and governmental agencies that sanctioned and promoted a traditional spelling norm with the aim of centralizing and retaining control over language and education. When read against the broader political and historical context, the discursive confrontation between these two groups reveals the ef forts made by the participants in the debate to naturalize dif ferent conceptualizations of the nation and the national education system through dif ferent strategies of representation of the self and the other.
‘Because When Governments Speak, They Are Not Always Right’
213
2 The orthographic debate in mid-nineteenth-century Spain In 1843, teachers associated with Madrid’s Literary and Scientific Academy of Primary Education – an independent organization – developed a simplified spelling system and began to implement it in schools. In response to this independent initiative, the government, through the Council for Public Instruction – an advisory body created the same year to oversee the public education system – asked Queen Isabel II (1830–1904; reigned 1833–1868) to ban the reformed system from schools and to of ficialize the RAE’s orthographic norms. The Queen agreed and in 1844 signed a Royal Decree that mandated the exclusive use of the RAE’s orthography in Spain’s primary instruction. The Literary and Scientific Academy of Primary Education contested the imposition and took actions to oppose its implementation by moving the debate to the public sphere. They organized meetings and published essays to defend both the simplified orthography and the legitimacy of the institution. For my study I have examined of ficial documents, textbooks, pamphlets, minutes, etc. which provide us with a record of this linguistic ideological debate and reveal a struggle over authority and power within Spain’s linguistic and educational markets. The teachers of the Literary and Scientific Academy defended a ‘radical reform’ – their own words – that would produce a phonological spelling system ‘completely based on the principle that, since writing is the faithful representation of the written word, its signs ought to be fixed and precisely determined’ (Academia 1844: 4a).8 Their orthographic ideal lay in the exact correspondence between sounds and phonemes and, in consequence, they defended the following reforms: first, the simplification of pairs of letters that represented the same sound, such as, , , y , writing cinze instead of quinze, jente instead of gente and
8
All translations are mine. The version respecting the original orthography reads, in Spanish: ‘reforma radical y basada toda sobre el principio de que, siendo la escritura la representacion fiel de la palabra, deben ser fijos sus signos y esactamente determinados’.
214
Laura Villa
Lola i Juan instead of Lola y Juan; second, the substitution of simple letters and for digraphs and and of letter for before and
; third, the replacement of for the combination in intervocalic position and for in pre-consonantal position; fourth, the elimination of silent letters, such as or in the combinations y , writing onbre rather than hombre and giso rather than guiso; finally, the modification of the names of some letters, in order to achieve a regularity in the designation of the sounds, for instance, the letter would be renamed ‘gue’ (pronounced /γe/) instead of the traditional ‘ge’ (/χe/), would be ‘que’ (/ke/) instead of ‘ce’ (/θe/).9 In contrast, the RAE privileged an orthographic system based on the traditional criteria of etymology and usage, since ‘the Orthography of all languages is founded on two principles, namely, the origin of the voices and the alterations introduced in many of them by usage, which is the supreme linguistic arbiter when it becomes general and uniform’ (RAE 1845: ii).10 The disparities of these proposals – a radical phonology-based reform versus an orthography founded on etymology and usage – are rooted in dif ferent traditions that were part of the linguistic culture of the nineteenth century: the ideas of the Enlightenment versus traditional ideas derived from the study of Latin (Mourelle de Lema 2002).11 For the members of the teachers’ Academy, the rationality of a logical form of writing would be the main source of legitimacy for their proposal. And they did in fact claim the following:
9
10 11
Pradel y Alarcón (1845: 7–9) of fers us a contrast between the traditional orthography defended by the RAE and the radical simplification of the primary school teachers. The former would write ‘el hombre que bebe mucho vino no se robustece’, the latter ‘el onbre ce bebe muho bino no se řobusteze’. ‘la Ortografía de todos los idiomas se funda en dos principios; á saber, el orígen de las voces y la alteracion que en muchas de ellas ha introducido el uso, que es el árbitro supremo de las lenguas cuando llega á hacerse general y uniforme’. The discussion on the principles that should drive the spelling system was by no means a new one. The history of the Spanish shows in fact a number of controversies over the spelling system as well as attempts to simplify Spanish orthography from the fifteenth century to the present day (Esteve Serrano 1982).
‘Because When Governments Speak, They Are Not Always Right’
215
where reason is in play there is no resorting to usage, where reason is in play there is no resorting to tradition, where reason is in play there is no other authority, there is no Academy [RAE], no Council for Public Instruction, no government. (Academia 1844: 28a)12
It is necessary to point out the fact that, in spite of their discrepancies, both teachers and government banked on the promotion of one orthographic norm through the education system, as a means of solving the problem posed by the coexistence of dif ferent spelling norms. This alignment with linguistic unity is consistent with a monoglossic ideology that, relying on the romantic correspondence between language and nation, sees linguistic homogeneity as an indispensable requirement for, and the symbolic realization of, national unity.13 In this sense both proposals negatively portrayed the coexistence of several orthographic varieties as a source of disorder and confusion, and discursively represented themselves as the solution to orthographic chaos. For instance, the Royal Decree that imposed the of ficialization of the RAE’s orthography explained the necessity of the intervention because ‘from frequent alterations may result confusion and mistakes’ (Villalaín Benito 1997: 99–100).14 More passionately, a member of the Literary and Scientific Academy, in a pamphlet entitled No more objections and obstacles to the instruction of the people!, stated: ‘So it is clearly seen how unreasonably they label us as orthographic anarchists, when in reality our project hopes to do away precisely with anarchy and not with order’ (Macias 1846: 27).15 12 13
14 15
‘donde hay razon no hay uso, donde hay razon no hay antigüedad, donde hay razon no hay autoridad, no hay Academia, no hay consejo de instruccion pública, no hay gobierno’. Del Valle (2000; 2006) explains that this ideology is grounded on the following assumptions: ‘that what linguistically characterizes an individual as well as a community is possession of a well defined and relatively stable grammar’ (2006: 119) and ‘that people’s linguistic behavior tends to become homogeneous over time through pressure from the dominant norm of the community’ (120). ‘de frecuentes alteraciones puede resultar confusión y equivocaciones’. ‘Véase pues claramente con cuan poca razon se nos acusa de anarquistas ortográficos, cuando en la realidad á extinguir la anarquía y no el buen órden […] es á lo que camina nuestra empresa’.
216
Laura Villa
Both sides of the debate have also in common the discursive representation of their orthography as a good for the nation. Thus, the Royal Decree justified the imposition because ‘all nations always proceed with extreme caution in such a delicate matter, preferring the advantages of a fixed and uniform orthography understood by all’ (Villalaín Benito 1997: 99–100).16 For their part, the teachers’ association publicly defended that the simplified orthography is ‘illustrious, useful and beneficial not only for the good of the children, but for the good of the nation in general’ (Academia 1844: 38b).17 The previous description of the discourses for or against the orthographic simplification leads us to two important conclusions. On the one hand, both teachers and government agreed on the necessity of unifying the orthography ‘for the good of the nation’, relying on nationalist arguments in order to defend the unification of the linguistic market. On the other hand, each actor involved in this controversy represented its proposal as having some value outside the market, presenting the spelling systems as more rational or more natural and, therefore, concealing the extra-linguistic motivation of the debate: a struggle to position themselves as the legitimate authority in a changing educational market. Earlier valuable studies of this important episode in the history of the standardization of Spanish have followed this logic in their analysis: they described the linguistic characteristics of each proposal, linked the of ficialization to the growing power of the central State, or evaluated the benefits of the simplified orthography in terms of the spread of education. Among the most relevant work on the of ficialization of the RAE’s orthography are Esteve Serrano (1982) and Martínez de Sousa (1991), who brief ly describe the episode in their histories of the Spanish orthographic reform, Martínez Alcalde (2010), who also mentions the of ficialization of the RAE’s orthography in her study of the standardization of Spanish spelling, 16 17
‘Todas las naciones proceden siempre con suma circunspección en tan delicado punto, prefiriendo las ventajas de una ortografía fija, uniforme y comprendida por todos’. ‘eminente, útil y provechosa no solo al bien de la niñez, sino de la nacion en general’.
‘Because When Governments Speak, They Are Not Always Right’
217
Rosenblat (1951), a key study to understand the of ficialization as tied to the project of construction of a liberal Spanish nation, and Quilis Merín (2008), who links the orthographic simplification to the universalization of literacy intended by liberal political movements, in her examination of the orthographic ideas in Dominguez’s Dictionario nacional. Despite the great interest of these studies, they focus exclusively on the simplification carried out by the Literary and Scientific Academy before the of ficialization of the RAE’s orthography, thus overlooking the teachers’ reaction to the enforcement of the Royal Decree. These works also failed to take into account the socio-political nature of the debate. An in-depth analysis of primary source and original documentation will reveal an extra-linguistic dimension in the discourses displayed by the participants in this orthographic controversy.
3 The non-linguistic nature of the orthographic debate Interestingly enough, and significantly for the outcome of the debate, a great discursive ef fort was oriented to legitimate not the advisability of the orthography itself, but rather the authority of the institutions that raised their voices in this linguistic-educational matter. The Literary and Scientific Academy put forward a pedagogical argument to defend their proposal: a simplified orthography, they claimed, would be learned faster and ‘written by everyone with more propriety and fewer mistakes’ (Academia 1844: 14b).18 The teachers grounded their legitimacy to intervene in linguistic matters in their experience educating both primary school children and adult workers, both men and women. One member of the educational association explained that the orthographic reform is a response to ‘a thousand and one troubles and inconveniences that only one dedicated to teaching
18
Emphasis in the original Spanish: ‘se escribe por todos con mas propiedad y menos equivocaciones’.
218
Laura Villa
can duly perceive’ (Academia 1844: 12a-b).19 Another teacher denounced the RAE’s orthographic manual imposed as the of ficial textbook in primary schools pointing out that the RAE’s reference book could by no means exceed ‘the many compendia that have been written by experienced teachers that touch and feel the dif ficulties caused by the anomalies in several letters of our alphabet’ (Hernando 1845: 3–4).20 If the Literary and Scientific Academy relied on pedagogical and professional arguments to defend the legitimacy of both their institution and their orthographic proposal, the government, in line with their project of a centralized public education system, used openly political arguments, represented the of ficialization of Spanish orthography as a national necessity and used its power to impose it. The governmental Council for Public Instruction turned to the Crown, the highest institution of political authority and symbolic power, and to the State apparatus, making political leaders and the Committees of Primary Instruction responsible for implementing the Royal Decree. With this of ficialization of the orthography, linguistic matters became ‘a concern of the State’ (Rosenblat 1951: cxxiv): in Spain’s language policy and linguistic discourse, language unity turned into an instrument to maintain and promote the national consensus. After the Royal Decree, the orthography became a matter of national interest because it symbolically represented the nation, and its of ficialization came to be considered ‘an issue of extreme significance due to the serious damage that can be done by the misunderstanding caused by an impure orthography in important documents’ (Villalaín Benito 1997: 99–100).21 The Spanish language was therefore declared a political issue which, due to its centrality to the nation, should be managed by the Crown, the central government and the institutions tied to them, such as the Council for Public Instruction or the RAE. 19
‘mil fatigas é incomodidades, que solo puede apreciar debidamente el que está dedicado á la enseñanza’. 20 ‘lo mucho que hay escrito por profesores práticos en diferentes compendios […] los que tocan y palpan las dificultades que producen en la enseñanza las anomalías que contienen varias letras de nuestro alfabeto’. 21 ‘asunto […] de suma trascendencia por los graves perjuicios que puede acarrear en documentos importantes la equivocada inteligencia de lo escrito por efecto de una ortografía adulterada’.
‘Because When Governments Speak, They Are Not Always Right’
219
The public debate ref lected the political nature of the orthographic controversy and the discourses displayed both in favor of and against the proposed reform were riddled with political representations of the opponent: the teachers were labeled as linguistic anarchists and antipatriotic; they, in turn, represented the central government as absolutist and despotic. For instance, they accused the government of excessive authoritarianism: primary school teachers were forced to comply with the Royal Decree under threat of suspension. In a heated public meeting, organized by the Literary and Scientific Academy to discuss the advantages of the phonological orthography, some teachers called for action against the ‘despotic agents’: ‘Let’s join forces, now is the time, to face our worries and even to confront power, since it rides roughshod over our rights and of fend against our honor’ (Academia 1844: 17a).22 One teacher present in that meeting even made the government responsible for the uprising of the people since ‘when governments speak, they are not always right […] and unfortunately at many times they impose restrictive precepts to good things […] whence their discredit as well as the frequent revolutions that we are going through spring up’ (Academia 1844: 30b).23 The linguistic ideological debate that surrounded the of ficialization of the Spanish orthography confirms that, as Woolard states: ‘[t]he definition of what is and what is not literacy is never a purely technical but always a political matter’ (Woolard 1998: 23). The discursive representation of the self and the other, in this sense, ref lected a broader socio-political struggle. The exchange of accusations, such as the Literary and Scientific Academy’s ardent insistence on the government’s tyranny and despotism, or the government’s fervent proclamation of the teachers’ whim, ignorance and irresponsibility, went beyond the linguistic terrain. What was at stake, I argue, was control over the educational market, that is, the governance of public and private educational spaces, institutions and legal competence.
22 23
‘Unámonos; que aun es hora, para hacer frente á la preocupacion, y aun al poder, cuando holle nuestros derechos ó ultraje nuestro honor’. ‘que cuando los Gobiernos hablan, no siempre tienen razon […] y que muchas veces por desgracia ponen preceptos restrictivos á lo bueno y […] de lo cual nace su descre dito y las frecuentes revoluciones por las cuales vamos atravesando’.
220
Laura Villa
I rely on Bourdieu’s (1991) work and terminology to emphasize the field tensions operating within those educational spaces as well as the struggles over authority within the market. Bourdieu’s model also accounts for the dif ferent value acquired, based on social conditions, by diverse linguistic practices (dif ferent spelling systems, for instance). Moreover I chose the term ‘educational market’ for its relation to the linguistic and labor markets, since the linguistic practices unfavorably valued in the first one could restrict access to the others. For instance, after the of ficialization of the orthography, teachers would be severely evaluated on this subject. Those who failed the exam would lose the authorization to practice their profession. The steps taken by the central government in order to promote a public national education brought about the rearrangement of the educational market, a redistribution of power and, consequently, a struggle to manage the market, intensified by the severe centralization that the public education was undergoing. The Literary and Scientific Academy, an independent association which ‘hoped to gain the monopoly of the Court’s public education’ (Gómez R. de Castro 1983: 50), fought to retain power over the educational space not monopolized by the Catholic Church. The government was also trying to control strategic markets in order to spread the national consensus. They realized that, according to the prominent intellectual Antonio Gil de Zárate (1796–1861), first director of the Council for Public Instruction, also a member of the RAE: ‘education is a matter of power: the one who teaches dominates; given that to teach is to form men, men adapted to the viewpoint of the one who indoctrinates’ (Gil de Zárate 1855: 117).24
24 ‘la cuestión de la enseñanza es una cuestión de poder: el que enseña, domina; puesto que enseñar es formar hombres, y hombres amoldados a las miras del que los adoctrina’.
‘Because When Governments Speak, They Are Not Always Right’
221
4 Linguistic standardization and the nation-building project The of ficialization of the RAE’s linguistic norms, a milestone in the standardization of Spanish, occurred throughout Queen Isabel II’s reign. During those decades the central government engaged in the construction of a liberal monarchic Spanish nation, projected since the Cadiz Constitution (1812) and slowed down during Fernando VII’s (1784–1833) absolutist rule (1814–1820 and 1823–1833).25 The legislative machinery of Queen Isabel II was mobilized at the service of the nation-building project, conceived during her reign from a moderate and traditionalist perspective in keeping with the conservative character of the majority of the governments formed since 1843. As part of the development of State power and institutions, a national centralist school system was developed in the mid-century.26 For example, the education system became fully regulated, through educational plans and legislative initiatives – such as the ‘plan Rivas’ in 1836, the ‘ley Someruelos’ for primary education in 1838 or the ‘plan Pidal’ for secondary and university instruction. The legal constitution of Spain’s instruction culminated in the first comprehensive education law, the ‘ley Moyano’, approved in 1857 – that in its article 88 ratified the special status of the Royal Academy’s linguistic norms. The central government also took steps to create institutions, such as the abovementioned Council for Public Instruction, in charge of controlling public education since 1843. The processes of State formation typically create the conditions for ‘the constitution of a unified linguistic market, dominated by the of ficial language’ which is ‘bound up with the state’ (Bourdieu 1991: 45). 25
I follow Álvarez Junco (2001), Burdiel (1998), and Carr (1969) for a historical contextualization of Queen Isabel II’s reign, as well as for a detailed description of political movements and nation-building processes during her sovereignty. 26 Puelles Benítez’s works (1999, 2004, and 2005) provide a precise account of the significance of the public school system within the politics oriented to the centralization and growth of State power. The author concludes that a conservative turn in the educational project and a lack of investment in public instruction resulted in the failure of the universalization of education in nineteenth-century Spain.
222
Laura Villa
Language seems to be an essential element in the promotion of the national feeling of loyalty and solidarity to the national community that imagines itself as unitary thanks, in part, to the standard language spread through the print, the literature and the press (Anderson 1983). Education appears as a central mechanism in the spread of national consensus and the homogenization of the national community and, therefore, in the standardization process ‘that leads to the construction, legitimation and imposition of an of ficial language, the educational system plays a decisive role’ (Bourdieu 1991: 48). Earlier research on the rise of public school systems in the nineteenth century has shown that the State’s interest in public instruction is necessarily tied to the larger project of national construction, both in its most material dimension and in the formation of collective national ideologies that would legitimate State power (Green 1990: 77). The processes of linguistic standardization in mid-nineteenth-century Spain are also embedded in nation-building processes: the selection, promotion and implementation of the of ficial variety – and all related language policies – contributed to the consolidation of State power in several ways. On the one hand, the standard language made it possible to spread a national school system that would disseminate public instruction and create citizens capable of exercising the rights and fulfilling the duties imposed by the State. On the other, the of ficial variety facilitated the consolidation of the State’s bureaucracy and legislation, by of fering the homogenization required to the viability of a central administration. And finally, the dissemination of (the idea of ) a language shared by all members of the national community promoted a loyalty attachment to the nation, strengthening its symbolic representation. In this fight over the control and management of the education, the RAE began to position itself as the institution of linguistic and cultural reference, a position that this academy would eventually maintain until the present day. The privileged political situation that its most renowned members enjoyed at the particular time covered by my study contributed to the constitution of the RAE as the language authority, as the legitimate prescriptive voice that dictates the norms of linguistic correction. For instance, the director of the RAE at the time of the of ficialization, Francisco Martínez de la Rosa (1787–1862), was also the president of the
‘Because When Governments Speak, They Are Not Always Right’
223
inf luential Ateneo Científico y Literario de Madrid – an exclusive private cultural club where the most significant intellectuals contributed to Spain’s scientific and cultural discussion since the beginning of the nineteenth century – and occupied throughout his life several political and governmental positions, such as representative in the Cortes de Cadiz, president of the Conservative party, Prime Minister, or president of the Congress. The inf luence of the RAE in educational of fices was also crucial: Gil de Zárate, who joined the cultural institution in 1841, held dif ferent of fices within the developing public school system, such as general director of Public Instruction, or director of the Council for Public Instruction. He also played a central role in the redaction of the ley Moyano and his Manual de literatura española, composed between 1842 and 1844, contributed to the early institutionalization of Spain’s literary canon. The Council for Public Instruction, that alerted Queen Isabel II to the abuses of school teachers in orthographic matters, counted among their members with four academicians at the time of its creation: Manuel José de la Quintana (1772–1857), Eugenio de Tapia (1776–1860), Martín Fernández Navarrete (1765–1844) and RAE’s perpetual secretary Juan Nicasio Gallego (1777–1853) (Ceprián Nieto 1991: 437–439). The exploitation of the royal status of the institution and, in particular, of the connections of fered by the political of fices held by its members would determine their ability not only to intervene in the State educational institution but also to access the public spaces of intellectual discussion.
5 Conclusion To conclude, my analysis understands language as a socio-cultural construction and aims at studying linguistic polemics within their specific historical and cultural context. I have presented an example of this historically situated approach to language, analyzing the linguistic ideological debate surrounding the of ficialization of the RAE’s spelling system. The
224
Laura Villa
orthographic debate was certainly one of the most intense linguistic controversies in Spain during the nineteenth century: it reached the public sphere and transcended both the academic realm and the linguistic field. It was not, however, an isolated event. It was in fact embedded in the process of linguistic standardization of Spanish which, in the mid-nineteenth century, played a central role in the representation of and exclusion from the national community. The standard language, likewise, became an essential object, necessarily shared by the members of the national community, and therefore it was promoted through the development of ‘a free, mandatory state education system’ that was felt to be a requirement for building the nation-state, as well as a mechanism, together with the anthem, the f lag, the army and national history, to spread the national consensus (Álvarez Junco 2001: 545). The linguistic ideological debate between governmental agencies and the Literary and Scientific Academy ref lected an intense struggle to gain or retain authority over educational matters in the rearrangement of the market brought about by the centralization of instruction. As previously discussed, the battle over control of the linguistic and educational markets bent to the central government and, in particular, to the RAE, whose orthography and grammar were ratified, by the celebrated ley Moyano, as the of ficial textbooks both for primary and secondary schools. The centrality of the orthographic episode derives from the fact that it shows that the rise of the Spanish orthographic system that is still in use today, responded not to a strictly linguistic process but rather to the complex interaction between linguistic practices and social and political institutions. Our approach aims at bringing to the forefront the sociopolitical dimension of the history of Spanish, by attending to those salient moments that ref lect the dynamic intersection between linguistic, cultural, social and political issues and help us to understand language policies and linguistic authority in contemporary Spain.
‘Because When Governments Speak, They Are Not Always Right’
225
References Academia Literaria i Zientifica de Instruczion Primaria. 1844. Sesion Publica celebrada el dia 3 de octubre de 1844, en el salon del instituto español. Por la Academia de profesores de primera educación, para demostrar las ventajas que ofrece la reforma de ortografía adoptada y publicada por la misma Academia. Dedicada a los profesores y amantes de la educación. Madrid: Imprenta de Dª. Francisca Estevan. Álvarez Junco, José. 2001. Mater dolorosa. La idea de España en el siglo XIX. Madrid: Taurus. Anderson, Benedict. 1983. Imagined Communities. London: Verso. Balibar, Renée, & Dominique Laporte. 1976. Burguesía y lengua. Barcelona: Avance. Bermel, Neil. 2007. Linguistic Authority, Language Ideology, and Metaphor. The Czech Orthographic Wars. Berlin: Mouton de Gruyter. Blommaert, Jan. 1999. Language Ideological Debates. Berlin: Mouton de Gruyter. Bourdieu, Pierre. 1991. Language and Symbolic Power. Cambridge: Harvard Univ. Press. Burdiel, Isabel. 1998. La política en el reinado de Isabel II. Madrid: Marcial Pons. Carr, Raymond. 1969. España 1808–1939. Barcelona: Ariel. Ceprián Nieto, Bernardo. 1991. Del Consejo de Instrucción Pública al Consejo Escolar del Estado: origen y evolución (1836–1986). Madrid: UNED. Contreras, Lidia. 1993. Historia de las ideas ortográficas en Chile. Santiago de Chile: Editorial Universitaria. Dalen, Arnold. 2005. ‘Sources of Written and Oral Language in the 19th Century.’ In: Bandle, Oscar (ed.), The Nordic Languages. An International Handbook of the History of the North Germanic Languages. Berlin: Walter de Gruyter, 1406–1419. Del Valle, José. 2000. ‘Monoglossic Policies for a Heteroglossic Culture: Misinterpreted Multilingualism in Modern Galicia.’ Language and Communication 20, 105–132. Del Valle, José. 2006. ‘US Latinos, la hispanofonía and the Language Ideologies of High Modernity.’ In: Mar-Molinero, Clare, & Miranda Stewart (eds), Globalization and Language in the Spanish-Speaking World: Macro and Micro Perspectives. New York: Palgrave MacMillan, 27–46. Del Valle, José. 2007. ‘Glotopolítica, ideología y discurso: un marco conceptual para el estudio de la difusión del español.’ In: Del Valle, José (ed.), La lengua, ¿patria común? Frankfurt/Madrid: Vervuert/Iberoamericana, 13–29.
226
Laura Villa
Esteve Serrano, Abraham. 1982. Estudios de teoría ortográfica del español. Murcia: Universidad de Murcia. Gil de Zárate, Antonio. 1855. De la Instrucción pública en España. Madrid: Impr. del Colegio de Sordomudos. Gómez R. de Castro, Federico. 1983. ‘La resistencia a las innovaciones. Informe de la Academia de Profesores de Primera Educación.’ Historia de la educación 2, 49–53. Green, Andy. 1990. Education and the State Formation: The Rise of Education Systems in England, France and the USA. Basingstoke: Macmillan Press. Hernando, Victoriano. 1845. Impugnación razonada en contra del prontuario de ortografia castellana que de Real Orden ha compuesto la Academia de la lengua española, con arreglo a su ultimo diccionario, para uso de las escuelas publicas. Madrid: Imprenta de D. Victoriano Hernando. Johnson, Sally. 2005. Spelling Trouble? Language, Ideology and the Reform of German Orthography. Clevedon: Multilingual Matters. Kroskrity, Paul V. 1999. Regimes of Language: Ideologies, Polities, and Identities. Santa Fe, NM: School of American Research Press. Macias, Felipe Antonio. 1846. ¡No mas trabas ni obstáculos ä la instruccion del pueblo! ¡Abajo! (Entre las clases sin pretension de eruditas). Ortografía irracional. Debate lógico, sobre las diferentes anomalias de la ortografía castellana, y sobre la conveniencia ë inconveniencia de su proyectada reforma. Para servir de aditamento crítico, al Manual completo de instruccion primaria elemental y superior de D. Joaquin Avendaño. Bilbao: Imprenta y litografía de Delmas é hijo. Martínez Alcalde, María José. 2010. La fijación ortográfica: norma y argumento historiográfico. Bern: Peter Lang. Martínez de Sousa, José. 1991. Reforma de la ortografía española. Estudio y Pautas. Madrid: Visor. Mourelle de Lema, Manuel. 2002. La teoría lingüística de la España del siglo XIX. Madrid: Grugalma Ediciones. Narvaja de Arnoux, Elvira. 2008. Los discursos sobre la nación y el lenguaje en la formación del Estado (Chile 1842–1862). Buenos Aires: Santiago Arcos. Pradel y Alarcón, Francisco. 1845. Juicio crítico acerca del prontuario de Ortografía de la lengua castellana dispuesto de real órden para el uso de las escuelas públicas por la Real Academia Española, y sobre su impugnacion por los que prestenden la reforma del alfabeto. Madrid: Imprenta del Colegio de Sordo-Mudos y Ciegos. Puelles Benítez, Manuel. 1999. Educación e ideología en la España contemporánea. Madrid: Tecnos. Puelles Benítez, Manuel. 2004. Estado y educación en la España liberal (1809–1857): un sistema educativo nacional frustrado. Barcelona: Pomares-Corredor.
‘Because When Governments Speak, They Are Not Always Right’
227
Puelles Benítez, Manuel. 2005. ‘El Consejo Escolar del Estado: antecedentes, evolución y situación actual.’ In: Revista del Consejo Escolar del Estado: La participación de la comunidad educativa en el debate de la LOE 44, 44–49. accessed 18 August 2011. RAE. 1844. Prontuario de Ortografía de la lengua castellana dispuesto de Real Órden para el uso de las escuelas públicas por la Real Academia Española con arreglo al sistema adoptado en la novena edición de su diccionario. Madrid: Impr. Nacional. RAE. 1999. Ortografía de la lengua española. Madrid: Espasa. . Rosenblat, Ángel. 1951. ‘Las ideas ortográficas de Bello.’ In: Bello, Andrés, Obras completas. V. Caracas: Ministerio de Educación, 9–138. Torp, Arne. 2005. ‘The Nordic Language in the 19th Century I: Phonology and Orthography.’ In: Bandle, Oscar (ed.), The Nordic Languages. An International Handbook of the History of the North Germanic Languages. Berlin: Walter de Gruyter, 1425–1436. Villalaín Benito, José. 1997. Manuales escolares en España. Legislación 1812–1939. Madrid: UNED. Woolard, Kathryn. 1998. ‘Introduction: Language Ideologies as a Field of Inquiry.’ In: Schief felin, Bambi, Kathryn Woolard & Paul V. Kroskrity, Language Ideologies: Practice and Theory. New York: Oxford University Press, 3–50.
Gijsbert Rutten and Rik Vosters
As Many Norms as There Were Scribes? Language History, Norms and Usage in the Southern Netherlands in the Nineteenth Century
Abstract Contrary to the fixed language norms in the Northern part of the Dutch language area, the Southern provinces are commonly said to have known as many orthographical and linguistic norms as there were scribes. According to our analysis of a large number of orthographies and grammars, however, there seems to have been a vivid normative tradition, from which a shared body of linguistic norms can be deduced. In comparing these norms with a corpus of early nineteenth-century manuscripts, we argue against the traditional view of orthographical chaos in the South and even suggest a considerable and increasing dif fusion of the Northern language norms.
1 Introduction The present chapter discusses the linguistic situation in the Southern Netherlands1 in the eighteenth and nineteenth centuries from a historicalsociolinguistic perspective. Whereas a lot of research has been done on 1
Nowadays, roughly speaking, the Dutch-speaking part of Belgium. For stylistic reasons, we use ‘Flanders’ along with ‘the South’, ‘the Southern Netherlands’, etc. We are of course aware of the possible anachronism in using ‘Flanders’ to refer to the Southern Netherlands as a whole, that is, the entire Dutch-speaking part of Belgium.
230
Gijsbert Rutten and Rik Vosters
the history of Dutch linguistics in the Northern Netherlands, the South has gained far less attention, and in so far as historical and/or linguistic research has been done, language ideological myths often seem to have been involved. Two such myths are distinguished in this article: the manynorms myth (section 3.1) and the orthographical-chaos myth (section 3.2). The first refers to the idea that eighteenth and early nineteenth-century grammatical texts from Flanders display a wide variety of norms and do not constitute a coherent normative tradition at all. The latter refers to the idea that actual writing practices in this period are characterized by orthographical chaos. Both myths are reassessed and rejected. First, a vast and coherent Southern normative tradition is identified (section 4.1). Secondly, orthographical practices appear to have been fairly regulated (section 4.2). The latter conclusion is drawn on the basis of a corpus research using judicial and administrative documents from the 1820s. In general, we argue that any interpretation of the historical-linguistic situation should be founded upon empirical studies of that situation. The discussion and demythologizing of the two myths is preceded by an introduction to the historical-sociolinguistic context (section 2.1), which brief ly addresses the political circumstances, and on which the language ideological background of the myths is based (section 2.2).
On the other hand, Willems (1824) already used the term Vlaemsch (‘Flemish’) in this general sense, and whenever we refer to the areas historically belonging to the Duchy of Flanders, we will indicate this by referring to the actual provinces concerned (East Flanders and West Flanders).
As Many Norms as There Were Scribes?
231
2 History and linguistic history 2.1 Historical-sociolinguistic context Following the final defeat of Napoleon Bonaparte in 1815, the Low Countries were united after more than two centuries of separation. The origin of this split between Southern and Northern Netherlands must be dated over two centuries earlier, with the revolt against Spain at the end of the sixteenth century. The North became the independent Republic of the Seven United Provinces and began its so-called Golden Age. The South remained under Spanish (Habsburg) control, and from 1714 onward under Austrian (still Habsburg) rule. After the French invasion in the 1790s, the South was incorporated into France, while the North eventually became a vassal kingdom under Napoleon’s brother Louis Napoleon. After the fall of Napoleon, the European superpowers decided to create an enlarged Dutch buf fer state to the north of France. This led to the United Kingdom of the Netherlands (UKN) in 1815, which united the present-day Dutch, Belgian and Luxembourgish territories under the reign of King William I. The UKN only existed for about fifteen years, when the Belgian Revolution ended it, with the birth of the Kingdom of Belgium in 1830. These fifteen years together, however, are traditionally considered to have been crucial for the future of the Dutch language in Flanders: Without this brief family reunion, Belgium would probably have become a francophone nation. […] The pioneers of the Flemish movement, those who fought for the Dutchification of Flanders after 1830, were all educated during the United Kingdom of the Netherlands. (De Vries et al. 1993: 117; our translation)
From the sixteenth century onwards, language standardization took place in the Dutch language area, especially in the Northern Netherlands in and around the province of Holland and its main city Amsterdam. A preliminary written standard was created in the seventeenth century (macro and micro selection, first codification), maintained in the eighteenth century (further micro selection and codification), and transformed into
232
Gijsbert Rutten and Rik Vosters
nineteenth-century written Standard Dutch (codification and elaboration) (Van der Wal 1995, Van der Sijs 2004, Van der Wal & Van Bree 2008). The nineteenth century soon bore two major results of this standardization: an of ficial orthography (Siegenbeek 1804) and an of ficial grammar (Weiland 1805), endorsed by the state and to be used in the administrative and educational domains. Although this received view of the development of (Standard) Dutch is teleological, too general and unjustifiably disregards language variation (Van der Wal 2006), it does provide us with a clear view of the development of written Dutch in the Northern Netherlands, which eventually led to present-day Standard Dutch. In this view, the strong eighteenth-century normative tradition plays a crucial rule, as it is held responsible for guiding the preliminary seventeenth-century written standard of Holland into the nineteenth century (Rutten 2009a). The reverse of this preoccupation with language standardization in the North is that insight into Southern Dutch of the period is largely still lacking. Linguists have mainly focused on the North, and some even assume that, in a history of Standard Dutch, Southern Dutch has had no role to play after the sixteenth century. Consider Van der Sijs (2004: 53; our translation): Because of the political circumstances, the Southern Low Countries did not contribute to the standard language any more after the fall of Antwerp in 1585. For this reason, this book will essentially not deal with Southern Dutch from the seventeenth century onwards.
Parallel writing traditions and emerging standardization in the South have not received much attention. At the same time, it is acknowledged that some normative linguistic works were published in the Southern Netherlands during the eighteenth century, but these are not considered to constitute a normative tradition; on the contrary, they are criticized for their internal dif ferences in language norms (Smeyers 1959, cf. section 3.1). Furthermore, although we know a fair amount about the status of the language at the time, especially with regard to the opposition of Dutch and French (e.g. Smeyers 1959, De Ridder 1979, De Groof 2004), almost no research has been done into the actual form of Dutch in the Southern Netherlands.
As Many Norms as There Were Scribes?
233
2.2 The image of the eighteenth century in the South For our research into the linguistic situation in the South during the UKN (1815–1830), a look into the Southern Dutch linguistic situation of the eighteenth and early nineteenth centuries is indispensable. The image provided by historians in general of especially the eighteenth century has been quite negative for a long period of time (Hanou 2004, De Vries 2004). Illustrative of this is the image of the Southern Netherlands in the middle of the eighteenth century drawn up by Elias (1963: 106; our translation): The intellectual life in the entire Southern Netherlands […] around 1750 of fers us a view of the most barren landscape one can imagine. There was simply nothing. There was the most complete silence in the deepest intellectual poverty.
Linguistically, this historical image of the eighteenth century is paralleled by the ‘myth of eighteenth-century language decay’ (Van der Horst 2004): the idea that Southern Dutch, as opposed to Northern Dutch, did not show standardization but dialectization, a regression to locally defined varieties. In this vein, Wils (1958: 527–528; our translation) contrasts Northern uniformity with Southern diversity: By the end of the seventeenth century in the North, the colorful diversity in writing slowly yielded to a uniform written language, based on the good usage of the classic authors […] The language in the South had undergone a dif ferent development from the seventeenth century onwards, [and] tended to regress to its purely local character.2
The myth of eighteenth-century language decay can be traced back to contemporary comments on the state of the language. Especially during the early years of the UKN, those advocating a closer connection between Southern Dutch and Northern Dutch, for political reasons, had good cause
2
Although there is no doubt that Northern practice was not as uniform as claimed here – Siegenbeek’s (1804) spelling and Weiland’s (1805) grammar were only obligatory in the administration and education – we will have to leave this additional ‘myth of Northern uniformity’ for another occasion; cf. Vosters et al. (2010).
234
Gijsbert Rutten and Rik Vosters
to uphold the image of eighteenth-century Flanders as an intellectual wasteland (cf. Elias 1963). Consider Jan Frans Willems’ (1793–1846) comments on the eighteenth-century linguistic situation (1819: 34–35): [F]lemish spelling has not been fixed to the level of a general Flemish standard by anyone up to the present. […] [E]ach schoolteacher in the Southern provinces […] considers himself qualified to teach the children whatever language rules his whim might have dictated him. Anarchy is a serious evil, both in spelling and in politics.3
Willems advocated political and linguistic unity with the North. The political context induced commentators such as him to describe the recent past as negatively as possible. A similar ref lex can be witnessed in contemporary histories of the Southern Netherlands and Belgium in which the past functions as a long and dif ficult period of ‘slavery’, that is, of foreign rule, which only ended during the UKN, and especially from 1830 onwards (Peeters 2003). Van der Horst (2004: 73; our translation) also explains the myth of eighteenth-century language decay by referring to its rhetorical function in nineteenth-century linguistic debates: ‘By emphasizing that the South had no tradition of its own, no basis, no language culture, nothing, they strengthened their argument in favor of a closer connection to Northern Dutch’. Building on this myth of eighteenth-century language decay as proposed by Van der Horst (2004), we distinguish two (mis-)conceptions in the traditional literature, two ‘sub-myths’: first, the many-norms myth, as opposed to the vivid normative tradition of the North; and secondly, the orthographical-chaos myth, as opposed to the assumed fixed written language in the Northern Netherlands.
3
‘[D]e Vlaemsche spelling [is], tot heden toe, nog door niemand op vaste gronden van algemeenen Vlaemschen aerd gebracht is. […] [E]lke schoolmeester, in de Zuidelyke Provincien, […] acht zich bevoegd om den kinderen alzulke taelwetten voorteschryven, als hem door het hoofd zyn gewaeid. Anarchie is een erg kwaed, zoowel in de spelling, als in de regering’.
As Many Norms as There Were Scribes?
235
3 The language myths 3.1 The many-norms myth We already mentioned J.F. Willems’s strong claims about the state of Southern Dutch at the end of the eighteenth and start of the nineteenth century. Concerning the body of normative publications in the South, he furthermore explains how he spent many years going through all the books dealing with language and spelling, only to af firm what most already believed to be the case: ‘there are no Flemish orthographies or grammars of any lasting authority’,4 let alone a fully f ledged normative tradition. To further prove his point, Willems discusses several grammars and orthographies which had been available since the early eighteenth century,5 emphasizing inconsistencies and divergent opinions on specific orthographical points. Concluding that Flemish grammarians all adhere to dif ferent norms, Willems’s solution is simple: Flemings should point their gaze northwards, and at least partially adapt to the Northern Dutch orthographical norms (Vosters 2009). Many twentieth-century historians echo Willems’s claims. Sluys (1912: 53; our translation) speaks about ‘the greatest possible confusion’ in normative publications, with every author adhering to a dif ferent spelling system. Concerning the work of Des Roches (1735/1740–1787), no doubt the most authoritative of the eighteenth-century Southern grammarians, he even concludes ‘[n]either his grammar nor his orthography were followed by anyone’ (Sluys 1912: 53; our translation). De Vos (1939: 50–52; our translation) follows suit, using phrases such as ‘mind-numbing drudgery’ to describe most of the eighteenth-century normative works. He also lashes out at Des Roches, who supposedly wanted to promote the Antwerp dialect to the level of a literary language. 4 5
‘Er bestaen […] geene vlaemsche Spel- en Spraekkunsten van doorgaende gezag’. For the second half of the 18th century, he discusses P.B. (1757), the second edition of Verpoorten (1759), Des Roches [1761], Van Belleghem & Waterschoot [1773], Ballieu (21792), a fifth print of Ter Bruggen (11817, 51822), and several other minor works and/or reprints. He also mentions Vaelande [=Van Daele] (1805/1806) and Behaegel (1817), but does not discuss their work in more detail.
236
Gijsbert Rutten and Rik Vosters
Even balanced accounts such as Smeyers (1959: 112), who should certainly be praised for calling attention to the eighteenth-century codifiers and their grammars and orthographies, clearly states that none of the pre1815 grammarians ever strove for a uniform spelling, and that they all had dif ferent linguistic opinions depending on whichever dialect they spoke. After discussing a significant number of normative texts from the South, Smeyers concludes that most grammarians did nothing to contribute to a way out of ‘the maze of orthographical lawlessness’, and that the only thing bringing them together was their obsession with purism and fighting of f loan words (Smeyers 1959: 127–128; our translation). In sum, the idea that the South lacked a proper grammatical tradition and that every grammarian constructed his own idiosyncratic spelling system is defended by Willems (1824), and taken over by historians dealing with the issue in the twentieth century.6 3.2 The orthographical-chaos myth The second myth under discussion is built around the idea of orthographical chaos in Flanders up until the UKN – not just the lack of a normative tradition, but spelling chaos in actual writing practice. ‘Try to read a hundred dif ferent […] books,’ the grammarian P.B. claims in 1757, ‘and you will find a hundred dif ferent spellings’ (P.B. 1757: 3; our translation).7 Half a century later, his much younger colleague Pieter Behaegel (1817: 250; our translation) repeats this lamentation, suggesting that there are ‘almost as many ways of spelling, as there are people who worked on improving the spelling’.8
6
7 8
The fact that Sluys (1912), De Vos (1939) and Smeyers (1959) all base themselves on the claims of J.F. Willems can already be demonstrated by the choice of publications under discussion, which De Vos supplements with the anonymous Inleyding from Dendermonde (1785), and Smeyers completes with Van Boterdael (21776, 31785). ‘[W]ant leést honderd verscheyde schriften, zelfs boeken, gy zult honderd verscheyde spellingen vinden’. ‘Men ziet in onze landstreéken bynae zoo veel wyzen van spellen; als er verscheydene persoónen zyn, die zich op het verbeteren der spelling toegelegd hebben’.
As Many Norms as There Were Scribes?
237
Especially during the period of the UKN, this idea of linguistic decay during the preceding Austrian and French rule becomes commonplace, and the image persists well into the twentieth century. De Vos (1939: 45) mentions that decades of foreign control prevented the Dutch language in the South from manifesting itself as a civilized language of culture, which made it shrivel to the level of patois. Deneckere (1954: 326) takes this still one step further, denying the existence of supraregional written varieties by claiming that even administrative documents were hardly intelligible from one town to another. Also in the 1950s, but repeating these claims more recently,9 Wils describes how, under the French rule, ‘Southern Dutch withered and weakened to such an extent that all contact with Northern Dutch threatened to be lost’ (Wils 1956: 529; our translation). Along the same lines as Deneckere (1954), Wils (1956: 530; our translation) is also very clear in reporting that ‘Flemish dialects and spellings were still being used in school books, in courts of justice and in notary deeds, and in the administration’ during the early years of the UKN. Maybe the best summary of this viewpoint comes from Suf feleers (1979: 19; our translation): ‘As opposed to the relative uniformity in the written language of the North, absolute chaos ruled the South’.
4 Revisiting the norm-myth 4.1 The many-norms myth revisited Do the eighteenth and early nineteenth centuries indeed display normative chaos in the South? Was there no real normative tradition but, at the most, a relatively small collection of grammatical works characterized by normative variation, as even Smeyers (1959) implied? It should be noted beforehand that spelling dif ferences between authors of grammars and orthographies
9
Cf. Wils (2003: 33): ‘Voor een hele generatie werd het onmogelijk een Nederlandstalige cultuur te verwerven, zodat de taal verschrompelde en verarmde’.
238
Gijsbert Rutten and Rik Vosters
in itself, as well as an interest in purist activities, do not prove the absence of a vivid normative tradition. On the contrary, both characteristics also apply to the seventeenth- and eighteenth-century linguistic situation in the North, which is commonly described as heavily prescriptive (Knol 1977: 70–71). A steady stream of publications containing diverging opinions only proves the existence of a linguistic debate, not its absence. Moreover, on closer examination there appears to have been a vivid normative tradition in the South, and the norms proposed are not at all based upon local dialect features. After two early works from the first half of the century,10 it is precisely from 1750 onwards – when Elias (1963) envisioned an intellectual wasteland – that several linguistic publications have come down to us. In the 1750s and 1760s, three Antwerp grammarians laid the foundation of the Southern normative tradition of the later eighteenth and early nineteenth centuries: Jan Domien Verpoorten (1706–1773), P.B. (?-?) and Jan Des Roches. In 1752, the schoolteacher Verpoorten published the first edition of his Woôrden-schat, oft letterkonst, ‘Vocabulary, or grammar’.11 The greater part of the book consists of lists of loan words, mainly of French or Latin origin, with Dutch equivalents. Verpoorten also brief ly discusses some spelling issues. Verpoorten’s ‘new manner of writing’, as he proudly calls it, among other things has to do with getting rid of ‘superf luous’ consonants in consonant clusters representing only one sound. We summarize: [k] which is commonly spelled in auslaut and which should be spelled e.g. ik ‘I’ instead of ick; [γ] which is commonly spelled in anlaut and which should be spelled e.g. geven ‘give’ instead of gheven.
These kind of spelling proposals are not in any way related to the dialect of Antwerp. Instead, these are innovations already put forward in the seventeenth- and eighteenth-century normative tradition in the North, as well
10 11
E.C.P. (1713), see Dibbets (2003); Stéven (1714), see Rutten (2009c). The discussion of Verpoorten (1752, 1759) and Bincken (1757) is based upon Rutten (2009b).
As Many Norms as There Were Scribes?
239
as in the scarce early eighteenth-century grammars from the South (E.C.P. 1713, Stéven 1714). Verpoorten is just linking up with and re-implementing common orthographical innovations. Verpoorten’s few pages on orthography are rather basic and mainly contain examples. Similar orthographical proposals are put forward five years later by P.B. in his Fondamenten ofte grond-regels der Neder-duytsche spel-konst, ‘Foundations or basic rules of Dutch orthography’ (1757). P.B. criticizes schoolteachers who only present examples of correct spellings without explaining the rules, but he does not mention Verpoorten. In 1759, however, Verpoorten publishes the second edition of his Woôrdenschat, in which he extensively elaborates on all kinds of grammatical rules, orthographical as well as morphological, without mentioning P.B., and at the same time constantly referring to the Northern normative tradition. An appeal to Northern norms apparently strengthened one’s proposals. It seems that P.B. and Verpoorten were competitors, linguistically as well as commercially on the schoolbook market. They take part in an implicit yet lively linguistic discussion that rapidly changed from fairly basic orthographical and lexical (purist) matters into a broader linguistic approach. This broader approach is further developed by another Antwerp schoolteacher. In 1761, Des Roches published the Nieuwe Nederduytsche spraek-konst, ‘New Dutch grammar’. Contrary to his predecessors, Des Roches does not limit himself to spelling, loanwords and some morphological issues, but writes a full grammar of Dutch, based on the Northern tradition, texts from the Latin and French grammatical traditions, as well as his Southern colleagues P.B. and Verpoorten, albeit without mentioning them (Rutten 2009d). Des Roches’ grammar is the first Southern grammar for decades and counts as one of the most important contributions to the codification of Dutch in the South throughout the eighteenth century. These three Antwerp grammarians of the 1750s and 1760s were aware of and reacted to each other’s works. They proposed similar grammatical rules and presumably taught these rules in their classes. Furthermore, for our research concerning the period of the UKN, it is important to remark that this Southern normative tradition survived into the later eighteenth and early nineteenth centuries as well. Verpoorten published a third edition in 1767. The books by P.B. and Des Roches were printed over and
240
Gijsbert Rutten and Rik Vosters
over again, well into the nineteenth century, and especially of P.B., many (partial) pirate editions appeared. In the last decades of the eighteenth century and at the beginning of the nineteenth century, the normative tradition was not only continued but intensified as well: dozens of works were published in the whole of the Southern Netherlands, which are always concerned with orthography and pronunciation, but often also with other grammatical features.12 The potential spread of these works was wide, as we know that most primary schools in the later eighteenth century owned a grammar, along with reading matters and a catechism (Put 1990: 202). Besides, nine out of ten elementary schools for boys in the city of Leuven of fered orthography as a separate subject according to a 1795 survey (Put 1990: 208). However, the fact that there was a vivid normative tradition in itself does not imply that it was coherent as well. Therefore, we tried to distill language norms from this vast body of normative works. We selected several recurrent features and made an inventory of the prescribed use in the grammars, a selection of which will be shown below. The choice of these features depended on their importance in the nineteenth-century spelling debates, in which the two most important spelling options for every feature were divided into a typically ‘Southern’ and a typically ‘Northern’ variant (Bormans 1841; see also Vosters 2009). The features are the following: (1) dotted or undotted [ei], e.g. wijn or wyn ‘wine’; (2) the second element in the diphthongs [ei] and [œy], either or , e.g. klein or kleyn ‘small’, and bruin or bruyn, ‘brown’; (3) vowel lengthening, either by adding an or by doubling the original vowel, e.g. zwaard or zwaerd ‘sword’ (with [a:]), zuur or zuer ‘sour’ (with [y]); (4) the form of the definite and indefinite article in the nominative singular masculine form: spelled with or without a final , e.g. de man or den man ‘the man’, een man or eenen man ‘a man’; (5) the so-called superf luous letters: or in anlaut, or in auslaut, e.g. ik or ick ‘I’, and gheven or geven ‘give’.
12
E.g. Van Belleghem & Waterschoot [1773], Janssens (1775), Van Boterdael (1785), Ballieu (21792), Van Aerschot (1807), De Neckere (1815), Henckel (1815), Behaegel (1817), Gyselynck (1819). See also Table 1 and the references.
As Many Norms as There Were Scribes?
241
We examined these features in a selection of normative works from the South. The results can be seen in Table 8. The features are displayed horizontally. Vertically, the names of the authors are given, along with the date of the used edition and its origin. The bottom line gives the of ficial Northern spelling rule of Siegenbeek (1804). When the prescribed use in one of the Southern grammars coincides with this Northern rule, the feature is shown in boldface. Note that and are generally accepted and that every grammarian in North and South rejects and – these features cannot be considered to be either typically Northern or Southern. As can be deduced from the table, most Southern grammars before 1815, including the early eighteenth-century works as well as the Antwerp grammarians of the 1750s and 1760s dif fer from the Northern 1804 rule on nearly all of the selected features. More importantly, it is clear that there is almost complete general agreement in the South. In other words, there was a vast Southern normative tradition in which spelling choices were to a great extent identical. However, in the period of the UKN both writing traditions collide. The Southern tradition still continues for a while, but the Northern norms are gradually being brought to the fore, sometimes only as alternatives (e.g. Cannaert 1823), often as the only norm. In sum, we have argued that a distinct Southern orthographical tradition existed, as can be seen from the implicit debates among grammar writers, and from the coherent normative framework put forward by dif ferent grammars and orthographies published especially in the second half of the eighteenth century.
1757 1761 1792 1807
P.B.
Des Roches
Ballieu
Van Aerschot
Antwerp
1714 1773 1775 1785 1792 1805
Stéven
Van Belleghem & W
Janssens
Anon. [Inleyding]
Van Boterdael
Vaelande
-Flanders
y
y
y
y
y
y
ij
y
y
y
y
y
y/ij
Author
Anon. [Grond-regels]
Origin
Brabant/
1817
Year
y
y/ij
-y
-y/-i
-y
-y
-y
-i
-y
-y
-y
-y
-y
-y
-y
-y
-y/-i
During the United Kingdom of the Netherlands
1713
East/West- E.C.P.
1752
Verpoorten
Brabant/
Year
Author
Origin
Before the United Kingdom of the Netherlands
V+e
V+e/V+V
V+e
V+e
-n
-n/-ø
-n
-n
-n
-n
V+e V+e
-n
-n
-n
-n
-n
-n
-n
-n
-n/-ø
V+e
V+e
V+V
V+e
V+e
V+e
V+e
V+e
V+e/V+V
Table 8 Spelling features and the normative framework
g-
gh-/g-
g-
g-
g-
g-
g-
g-
g-
g-
g-
g-
g-
g-
gh-/g-
-k
-ck/-k
-k
-k
-k
-k
-k
-k
-k
-k
-k
-k
-k
-k
-ck/-k
242 Gijsbert Rutten and Rik Vosters
Ter Bruggen
1822
y
-y
V+e
-n
g-
-k
Zilgens
1824
y
-y
V+e
-n
g-
-k
Willems
1824
y
-i
V+e
-ø
g-
-k
East/West- De Neckere
1815
y
-y
V+e
-n
g-
-k
Flanders
Henckel
1815
ij
-i
V+e
-n
g-
-k
Gyselynck
1819
y
-y
V+e
-n
g-
-k
Cannaert
1823
y/ij
-y-i
V+e/V+V
-n/-ø
g-
-k
Moke
1823
ij
-i
V+V
-ø
g-
-k
Behaegel
1824
y
-y
V+V
-n
g-
-k
De Simpel
1827
ij
-i
V+V
-ø
g-
-k
Limburg
Anon. [Beginselen]
1819
ij
-i
V+V
-ø
g-
-k
In French
Van der Pijl
1815
ij
-i
V+V
-ø
g-
-k
Meijer
1820
ij
-i
V+V
-ø
g-
-k
As Many Norms as There Were Scribes?
Antwerp
Official norm Origin
Author
Year
y/ij
-y/-i
V+e/V+V
-n/-ø
gh-/g-
-ck/-k
North
Siegenbeek
1804
ij
-i
V+V
-ø
g-
-k
243
244
Gijsbert Rutten and Rik Vosters
4.2 The orthographical-chaos myth revisited The final section of this article will assess the ‘orthographical-chaos myth’. Did scribes in dif ferent towns write in their own dialects, and how chaotic were the actual spelling practices? To put the earlier claims to the test, we used a digitized collection of handwritten documents, which are being transcribed at the Centre for Linguistics of the Vrije Universiteit Brussel as part of ongoing research.13 The corpus contains formal and less formal texts from the judicial and administrative domain, including a roughly equal number of: (1) (2) (3)
police reports, drawn up at the local level by police constables, rangers, or other members of the municipal authorities; interrogation reports, written down by district-level scribes and signed by the juge d’instruction in question; indictments, issued by the professional scribes of one of the high courts.
Among these three text types, which already range from the very local to the supraregional level, all five Southern provinces are represented, with an equal amount of material per region coming from a main city and dif ferent peripheral towns or villages. A test version of the full corpus was used, containing a total of 61,912 words (excluding editorial and linguistic markup). The material allows us to compare writing practices in dif ferent regions, and furthermore has a built-in diachronic dimension, with texts from approximately 1823 and 1829. Both years have been chosen because of their sociohistorical importance. In January 1823, language laws came into practice that made the use of Dutch compulsory in most of the government administration and judicature in the Dutch-speaking Southern provinces. For the majority of the departments which were operating in French before the Dutchification policy took place, this means that the documents under investigation are among the first of their kind to be 13
The actual source material was gathered and compiled into a digital image database by Isabel Rotthier and the Royal Academy for Dutch Language and Literature (kantl; see Rotthier 2007). For more background on the research project as a whole, see Vosters & Vandenbussche (2008) and Vosters & Vandenbussche (2009).
As Many Norms as There Were Scribes?
245
written in Dutch since before the French rule of 1794–1814. The manuscripts give us an interesting overview of the Dutch language in Flanders during the early years of the UKN. This allows us to compare them with 1829, at the end of the Kingdom, and to see if any changes occurred after those years of political union with the North. The investigation of this corpus material consisted of electronic searches for the orthographical variables discussed earlier: how often do traditional Southern forms such as or occur, in comparison with the and forms codified by the Siegenbeek norm in the North? It must be emphasized at this point, however, that we refer to orthographical variants as being either Southern or Northern based on the findings of our norms research. ‘Southern’, then, means that a spelling variant coincides with the prescribed usage in the large majority of the pre-1815 Southern normative publications, while ‘Northern’ is used to indicate that a variant corresponds to the prescribed use in the of ficialized orthography of Siegenbeek (1804).14 Clearly, this approach is reductionist in enforcing a strict dichotomy, and the presented results should only be interpreted as a first and tentative indication of the spread and success of both orthographical traditions, rather than capturing all possible variation, at least as far as the selected features and the employed sources go. We do not mean to suggest that there is any direct causality between linguistic forms being prescribed in normative publications and their use in actual written language.
14
Note that the so-called superf luous letters and were also investigated, even though they seem to have disappeared from the Southern normative tradition in the eighteenth century already. The results for these searches are not taken up in the tables below, as they represent values below the 1% mark. These orthographical variants had made their way out of grammars and out of usage by the time of the UKN.
246
Gijsbert Rutten and Rik Vosters Table 9 Southern (left) and Northern (right) spelling variants per feature (distribution per cent) in a digitized corpus of handwritten documents from the Southern Netherlands15 dotting of [ei]
diphthongs
vowel lengthening
y
ij
-y
-i
V+e
V+V
Total
66%
34%
21%
79%
21%
79%
1823
74%
26%
34%
66%
33%
67%
1829
55%
45%
3%
97%
4%
96%
article N Sg M
Total
-n
-ø
Southern
Northern
Total
17%
83%
38%
62%
1823
25%
75%
48%
52%
1829
10%
90%
24%
76%
A first observation while searching through the corpus is that the situation was far from chaotic. As can be deduced from Table 9, scribes seem to have been f lexible in adopting the Northern norm: overall, 62 per cent of the total number of tokens was Northern. For three features (diphthongs with -y, long vowels in -e and masculine den), the Flemish variant is quite rare, being used in less than a quarter of all cases. In the case of the dotting of the [ei], the Southern variant stands stronger, which might be due to the minimal orthographical dif ference between and in handwriting. More surprising is that a similar pattern holds true for the situation in 1823 already. Rather than what might be expected from observations in the literature (i.e. a disparate but Southern preference), slightly over half 15
For the sake of clarity, we only shown the results per cent. The actual number of attestions is high in all cases (a total of 10,295 analyzed tokens, 3,917 of which are Southern and 6,378 are Northern). The full numbers can be sent upon request. Proper names, place names, uncertain transcriptions, and stretches of text in a foreign language were excluded from the present analyses.
247
As Many Norms as There Were Scribes?
of all the attested forms correspond to the Northern Siegenbeek norm (52 per cent). Although the Flemish is still very generally used, the other features occur in a Northern form in over two-thirds of the cases. This is highly remarkable, considering the claims about the spelling chaos and strong dialectal uses in written documents in Flanders, and considering that the investigated documents are among the first of their kind to be written in Dutch for all of the South. Furthermore, when we analyze the data diachronically, a clear movement away from the Southern variants can be observed. Three quarters of all features in 1829 are Northern, and the Southern spellings for diphthongs with a , for long vowels with an , and for den as the masculine form of the article have almost completely disappeared. In general, there is an increase of about 20 to 30 per cent in Northern forms for each of the individual features between 1823 and 1829. Table 10 Spelling variants per province (distribution per cent)16
1823
Total Southern forms
Total Northern forms
1829
Total Southern forms
Total Northern forms
Antwerp
36%
64%
Antwerp
44%
56%
Brabant
71%
29%
Brabant
33%
67%
Limburg
32%
68%
Limburg
21%
79%
East Fl.
68%
32%
East Fl.
30%
70%
West Fl.
62%
38%
West Fl.
33%
67%
The text distribution per province also allows us to look into regional variation, as shown in Table 10. In 1823, we still observe rather strong regional dif ferences, with Brabant, East Flanders and West Flanders showing a clear preference for the Southern forms, and the Northern forms prevailing in 16
The indictment section of the corpus has been excluded from these regional analyses, as these documents all originate from one of two supraregional high court of fices (Brussels or Liège). This still leaves the total number of analyzed tokens at 5,499.
248
Gijsbert Rutten and Rik Vosters
Limburg and Antwerp. However, when we look at the situation for 1829, we cannot only discern the general trend towards the Siegenbeek norm as discussed above, but it is also remarkable that the regionally divergent patterns make way for a more uniform preference for the Northern variants. In 1829, the Northern forms are used in the majority of cases everywhere, and the figures lie much closer together in general. Great shifts can be observed in those provinces which were still mainly using Southern spellings in 1823, with a doubled number of Northern forms in Brabant and East Flanders.17 To come back to the myth of spelling chaos in the Southern Low Countries, our corpus research allows us to draw several conclusions. The orthographical landscape in general is not at all distinctly Southern, and even when the Dutchification policy had just come into practice in 1823, the Siegenbeek variants were already widely spread. Apart from this far from chaotic spelling situation, we might add that none of the documents we transcribed and examined showed any sign of transliterated dialect (cf. Wils 1956). The number of Northern forms furthermore casts doubts on the claim that administrative documents were hardly intelligible from one town to another (cf. Deneckere 1954). The steady spread of the Northern spelling norms in a mere six years between the measuring points furthermore emphasizes the importance of the political union between North and South for the increasing convergence in writing practices during the nineteenth century. These results supplement the earlier findings of Vanhecke (2007), who investigated orthographical shifts in town council reports of seven Flemish municipalities between 1795 and 1900. She concluded that 1823 marked the start of a gradual shift from Southern practice to more Northern Siegenbeek variants, and concludes that the dif ferent orthographical shifts under changing political circumstances, often without a change of hand, are evidence of the remarkable linguistic competence of the scribes. Additionally, her large-scale investigation of language choice (Dutch/French) in 131 Southern chancery administrations shows that there is a uninterrupted 17
We did observe a slight decrease in Northern forms for the province of Antwerp, which calls for further research.
As Many Norms as There Were Scribes?
249
tradition of Dutch usage all throughout the nineteenth century, which in itself already challenges the idea of the language ‘withering and weakening’ under foreign inf luence (cf. Wils 1956).
5 Conclusion In this chapter we discussed the dif ferences between language myths prevailing in the historical and historical-linguistic literature, and the actual linguistic situation in Flanders during the eighteenth and early nineteenth century. Focusing on orthography, we first explored the surprisingly large body of normative publications available for the South in the decades preceding the UKN. There seemed to have been a clear and coherent tradition of Southern writing practices and orthographical prescriptions. The publications analyzed were not based on dialectal uses, and often maintained an awareness of the tradition of good usage in the Northern Netherlands. During the brief union of North and South between 1815 and 1830, grammarians and schoolteachers from Flanders played their own part in the new political scenario. A number of them switched to what had then become the of ficial spelling norm in the North (Siegenbeek 1804), while others continued to propagate Southern writing practices. Building on the importance of the UKN, we then proceeded to investigate actual writing practices in a corpus of handwritten documents from the 1820s. We already witnessed a preference for Northern spellings in the first documents written after the 1823 Dutchification, which led us to believe that the linguistic situation at the start of the UKN could hardly be described as ‘absolute chaos’. Moreover, we observed the firm spread of the Northern variants, which revealed how the way people wrote was directly or indirectly inf luenced by the contemporary socio-historical circumstances. In sum, the period of the UKN has been shown to be an interesting case study at the intersection of both Northern and Southern writing traditions, where the foundations of the later linguistic convergence of Northern and Southern Dutch developed.
250
Gijsbert Rutten and Rik Vosters
References Ballieu, J. 1792 [1772]. Néderduytsche spel- en spraek-konst. Antwerp: J.E. Parys. Behaegel, Pieter. 1817. Nederduytsche Spraekkunst. Eerste boekdeel. Bruges: Wed. De Moor en Zoon. Bormans, Jean Henri. 1841. Verslag over de verhandelingen ingekomen bij het staetsbestuer van België ten gevolge der taelkundige prijsvraeg. Ghent: Annoot-Braeckman. De Groof, Jetje 2004. Taalpolitiek en taalplanning in Vlaanderen in de lange negentiende eeuw. Unpublished doctoral dissertation. Brussels: Vrije Universiteit Brussel. De Neckere, Philips Jaques. 1815. Bewerp van Vlaemsche spelling, gevolgd van eenen oogopslag op de nederlandsche dichtkunst. Ypres: J.B. Smaelen-Moerman. De Ridder, Paul. 1999. ‘The use of languages in Brussels before 1794.’ In: Blockmans, Wim, Marc Boone, & Thérèse de Hemptinne (eds), Secretum Scriptorum. Liber alumnorum Walter Prevenier. Leuven: Garant, 145–164. De Vos, Hendrik. 1939. Moedertaalonderwijs in de Nederlanden. Een historisch-kritisch overzicht van de methoden bij de studie van de moedertaal in het middelbaar onderwijs sedert het begin van de 19e eeuw. Turnhout: Van Mierlo-Proost. De Vries, Jan W., Roland Willemyns & Peter Burger. 1993. Het verhaal van een taal. Negen eeuwen Nederlands. Amsterdam: Prometheus. De Vries, Marleen. 2004. ‘Efemeriden, seniliteiten, republiekjes der letteren. De geheime fantasieën over literaire genootschappen op een rijtje’. Spiegel der Letteren 46, 241–254. Deneckere, Marcel. 1954. Histoire de la langue française dans les Flandres (1770–1823). Ghent: Romanica Gandensia. Des Roches, Jan. [1761] (2007). Nieuwe Nederduytsche Spraek-konst. Derden Druk, oversien en verbetert doór den Autheur. Antwerp: Grangé. Reprint by J.M. van der Horst & Jozef Smeyers. Amsterdam: Stichting Neerlandistiek VU / Nodus. Dibbets, Geert. 2003. ‘Een nieuw spoor van de Port-Royalgrammatica in Nederland’. In: Dibbets, Geert, Taal kundig geregeld. Amsterdam: Stichting Neerlandistiek VU / Nodus, 129–156. E.C.P. [De Witte, Egidius]. 1713. Ontwerp van eene Nederduytsche Spraek-konst. [Meenen]: Theodorus vanden Eynden. Elias, Hendrik Jozef. 1963. Geschiedenis van de Vlaamse gedachte. De grondslagen van de nieuwe tijd (1780–1830). Antwerp: De Nederlandsche Boekhandel. Gyselynck, C.L. 1819. Nieuwe grond-beginselen der Vlaemsche tael, geschikt na de beste Nederduytsche taelmeesters. Ghent: J. Begyn.
As Many Norms as There Were Scribes?
251
[Grond-regels]. 1817. Grond-regels der Nederduytsche spel-konst. Mechelen: P.J. Hanicq. Hanou, André. ‘“De Verlichting brandt.” Een kwart eeuw verlichte letteren in Noord en Zuid (1670–1815)’. Spiegel der Letteren 46, 215–225. Henckel, Frans Lodewijk N. 1815. Nieuwe Vlaemsche spraek-konst. Ghent: P.F. de Goesin-Verhaege. [Inleyding]. 1785. Inleyding tot de grondregels der Vlaemsche spraek- en spelkonste. Dendermonde: Wed. J. Decaju. Janssens, Balduinus. 1775. Verbeterde Vlaemsche spraek- en spel-konste. Bruges: Joseph De Busscher. Knol, Jan. 1977. ‘De grammatica in de achttiende eeuw’. In: Bakker, Dirk Miente & Geert Dibbets (eds), Geschiedenis van de Nederlandse taalkunde. Den Bosch: Malmberg, 65–112. [P.B.]. 1757. Fondamenten ofte Grond-Regels der Neder-Duytsche Spel-Konst. Antwerp: Hubertus Bincken. Peeters, Evert. 2003. Het labyrint van het verleden. Natie, vrijheid en geweld in de Belgische geschiedschrijving (1787–1850). Leuven: Universitaire Pers. Put, Eddy. 1990. De cleijne schoolen. Het volksonderwijs in het hertogdom Brabant tussen Katholieke Reformatie en Verlichting (eind 16de eeuw – 1795). Leuven: Universitaire Pers. Rotthier, Isabel. 2007. ‘“In the picture”. Een bronnencorpus / beeldbank van juridische teksten uit de periode van het Verenigd Koninkrijk der Nederlanden’. In: Handelingen van de Koninklijke Zuid-Nederlandse Maatschappij voor Taal- en Letter kunde en Geschiedenis LX (2006). 131–149. Rutten, Gijsbert. 2009a. ‘Grammar to the people. The Dutch language and the public sphere in the 18th century with special reference to Kornelis van der Palm’. Beiträge zur Geschichte der Sprachwissenschaft 19, 55–86. Rutten, Gijsbert. 2009b. ‘Lezen en schrijven volgens Verpoorten en Bincken. Onderwijsvernieuwingen in Antwerp in de achttiende eeuw’. E-Meesterwerk 2. accessed 18 August 2011. Rutten, Gijsbert. 2009c. ‘Over Andries Stéven, wie het Voorschrift-boek schreef.’ In: Boogaart, Ronny, Josien Lalleman, Marijke Mooijaart & Marijke van der Wal (eds), Woorden wisselen. Voor Ariane van Santen bij haar afscheid van de Leidse universiteit. Leiden: Stichting Neerlandistiek Leiden, 301–312. Rutten, Gijsbert. 2009d. ‘De bronnen van Des Roches. Jan Des Roches’ Nieuwe Nederduytsche spraek-konst (1761) en de taalgeschiedenis van de achttiende eeuw’. Tijdschrift voor Nederlandse Taal- en Letterkunde 125 (4), 362–384.
252
Gijsbert Rutten and Rik Vosters
Siegenbeek, Matthijs. 1804. Verhandeling over de Nederduitsche spelling, ter bevordering van eenparigheid in dezelve. Amsterdam: Allart. Sluys, A.L.I. 1912. Geschiedenis van het onderwijs in de drie graden in België tijdens de Fransche overheersching en onder de regeering van Willem I. Ghent: Koninklijke Vlaamsche Academie voor Taal- en Letterkunde / Sif fer. Smeyers, Jozef. 1959. Vlaams taal- en volksbewustzijn in het Zuidnederlands geestesleven van de 18de eeuw. Ghent: Secretarie der Academie. Stéven, Andries. 1714 [1784]. Nieuwen Néderlandschen Voorschrift-boek. Ypres: Moerman. Suf feleers, Tony. 1979. Taalverzorging in Vlaanderen. Een opiniegeschiedenis. Bruges: Orion / Gottmer. Ter Bruggen, Joannes Abraham. 1822 [1817]. Nederduytsche spraek-konst ten gebruyke der schoólen. Antwerp: J.S. Schoesetters. Vaelande van Ieper [Van Daele, François Donatien]. 1805/1806. Tyd-Verdryf. Ondersoek op de Néder-duytsche spraek-konst. [Ypres]: [De Varver]. Van Aerschot, M. 1807. Nieuwe Nederduytsche spraek- en spel-konst. Turnhout: J.H. Le Tellier. Van Belleghem, P.J. & Daniël Waterschoot. [1773]. Deure oft Ingang tot de Nederduytsche Taele. Bruges: Van Praet. Van Boterdael, L. 1785 [1776]. Gemaklyke wyze om óp korten tyd grooten voortgang te doen in de Nederduytsche spel-konst. Bruges: Joseph Bogaert. Van der Horst, Joop. 2004. ‘Schreef J.B.C. Verlooy echt zo gebrekkig? Het 19de / 20ste-eeuwse beeld van de 18de eeuw getoetst’. Verslagen en Mededelingen van de Koninklijke Academie voor Nederlandse Taal- en Letterkunde 114 (1), 71–82. Van der Sijs, Nicoline. 2004. Taal als mensenwerk. Het ontstaan van het ABN. The Hague: Sdu. Van der Wal, Marijke. 1995. De moedertaal centraal. Standaardisatie-aspecten in de Nederlanden omstreeks 1650. The Hague: Sdu. Van der Wal, Marijke. 2006. Onvoltooid verleden tijd. Witte vlekken in de taalgeschiedenis. Amsterdam: Koninklijke Nederlandse Akademie van Wetenschappen. Van der Wal, Marijke & Cor van Bree. 2008. Geschiedenis van het Nederlands. 5th edn. Houten: Spectrum. Vanhecke, Eline. 2007. Stedelijke kanselarijtaal in Vlaanderen in de negentiende eeuw. Unpublished doctoral dissertation. Brussels: Vrije Universiteit Brussel. Verpoorten, Jan Domien. 1752. Woorden-schat oft letter-konst. Antwerp: A.J. du Caju. Verpoorten, Jan Domien. 1759. Woorden-schat oft letter-konst. 2nd edn. Antwerp: Berbie.
As Many Norms as There Were Scribes?
253
Vosters, Rik. 2009. ‘Integrationisten en particularisten? Taalstrijd in Vlaanderen tijdens het Verenigd Koninkrijk der Nederlanden (1815–1830)’. Handelingen van de Koninklijke Zuid-Nederlandse Maatschappij voor Taal- en Letterkunde en Geschiedenis LXII (2008), 41–58. Vosters, Rik, Gijsbert Rusten, V Marijke van der Wal. 2010. ‘Mythes op de pijubank. Naar een herwaardering van de taalsituatie in de Nederlanden in de achttiende en negtutiende eeuw’. Verslagen en Mededelingen van de Koninklijke Academie voor Nederlandse Taal- en Letterliunde, 120, 93–112. Vosters, Rik, & Wim Vandenbussche. 2008. ‘Wijzer worden over Willem. Taalgebruik in Vlaanderen ten tijde van het Verenigd Koninkrijk der Nederlanden (1814–1830)’. Internationale Neerlandistiek 46 (3), 2–22. Vosters, Rik, & Wim Vandenbussche. 2009. ‘Nieuw onderzoek naar taalbeleid en taalvariatie in de Zuidelijke Nederlanden ten tijde van Willem I’. In: Backus, Ad, Merel Keijzer, Ineke Vedder & Bert Weltens (eds), Artikelen van de Zesde Anéla-conferentie. Delft: Eburon, 389–395. Weiland, Petrus. 1805. Nederduitsche Spraakkunst. Amsterdam: Allart. Willems, Jan Frans. 1824. Over de Hollandsche en Vlaemsche schryfwyzen van het Nederduitsch. Antwerpen: Wed. Schoesetters. Wils, Lode. 1956. ‘Vlaams en Hollands in het Verenigd Koninkrijk’. In: Dietsche Warande en Belfort, 527–536. Wils, Lode. 2003 [2001]. Waarom Vlaanderen Nederlands spreekt. Leuven: Davidsfonds.
Anneleen Vanden Boer
Language and Nation: The Case of the German-Speaking Minority in Belgium
Abstract This paper investigates the use of language as a tool for nation-building, based on a study of the German-speaking minority in Belgium, in particular with regard to the correlation between knowledge of a language and the sense of af filiation to the nationstate for German-speaking Belgians. There are two opposing discourses asserting on the one hand that German-speaking Belgians ought to be ‘Frenchified’ or even ‘degermanized’ to become better Belgians and, more recently, on the other hand that the only ‘loyal Belgians’ are those who are native speakers of German. My findings appear to confirm the ‘stereotypes’ that German-speaking Belgians are royalists and patriots. Also, the national identity of this minority is no longer supported by knowledge of one specific language. A good overall knowledge of all of ficial languages, however, is a good predictor of the strength of Belgian identity.
1 Introduction The focus of this paper concerns the relationship between language command and sense of af filiation to the nation-state, based on a study of the German-speaking minority in Belgium. There are two interesting and opposing points of view on this issue, which can be traced to two dif ferent moments in Belgian history. The earlier of these two discourses asserts that German-speaking Belgians ought to be ‘Frenchified’ or even ‘degermanized’ to become better Belgians. More recent opinion, on the other hand, states
256
Anneleen Vanden Boer
that the only ‘loyal Belgians’ are those who are native speakers of German. Based on empirical evidence collected in 2008, I will investigate the correlation between knowledge of a language and the sense of af filiation to the nation-state for German-speaking Belgians. I will begin by giving an overview of the geographical and historical background of the (of ficially) recognized German-speaking minority in Belgium. This will be followed by a discussion of the theory relating to the notion of language as a tool for nation-building in this specific context. I will then proceed to explain the methodology of my own survey, reporting and interpreting the results before attempting to draw conclusions from my findings.
2 Geographical and historical context Belgium has ten million inhabitants, roughly six million of whom speak Dutch, four million French and seventy thousand German. Dutch, French and German are the of ficial national languages in Belgium. The Germanspeaking Belgians are recognized as a national linguistic minority and enjoy special language rights. The German-speaking part of Belgium consists of the areas surrounding three cities: Eupen, Sankt Vith and Malmedy. This region has been home to the German-speaking minority and belonged to Belgium since 1920. Although not itself the focus of my research, I would also like to draw attention to three additional areas where German is spoken. These are situated around the municipalities of Welkenraedt (north-east of Eupen), Bého (south of Sankt Vith) and Arlon (in the south of Belgium at the border with the Grand Duchy of Luxembourg). All three were initially German-speaking areas that have been part of Belgium since its foundation in 1830. Over the course of time they have become almost completely Frenchified. The German-speaking inhabitants in these areas are not recognized as a linguistic minority.
Language and Nation
257
The territory of the municipalities concentrated around Eupen and Sankt Vith, by contrast, has formally been recognized as a German-speaking area. This means that German is the of ficial administrative language in this region. Politically, the municipalities around Eupen and Sankt Vith constitute part of the German-speaking Community. This community is a legal entity of the Belgian federation with autonomy in personal and cultural af fairs. By contrast, the municipalities Malmedy and Waimes are situated in the French-speaking area, where the of ficial language is French, although the German-speaking minority living in these areas enjoys language facilities.1Malmedy and Waimes are not part of the German speaking community, but of the French community (cf. Vanden Boer, 2009). Subjected to a change in nationality four times since the Congress of Vienna in 1815, each of these areas has borne witness to an eventful past. Before being conquered by French forces at the end of the eighteenth century and annexed by the French empire, Eupen and Sankt Vith belonged to the Austrian Netherlands, to the Duchy of Limburg and the Duchy of Luxembourg respectively. Malmedy formed a sovereign abbey territory together with Stavelot. After the defeat of Napoleon in 1815 both duchies were divided during the Congress of Vienna. No attention was paid to their socio-cultural characteristics, as economic interests prevailed, and the duchies were thus divided between the United Netherlands and Prussia. Eupen, Sankt Vith and Malmedy became part of Prussia. Following the First World War, Belgium was recognized as a victim of war by the Treaty of Versailles in 1920, and was granted permission to annex the municipalities around Eupen, Sankt Vith and Malmedy. Twenty years later, however, the areas were incorporated into Germany again during the course of the Second World War. By the end of World War Two the area was liberated, returning once more to the control of Belgium, and continues to be governed by the country today (cf. Vanden Boer, 2008).
1
Language facilities give a citizen with a mother tongue other than the language of ficially spoken in the language area the right to use his own language in his contacts with of ficial authorities.
258
Anneleen Vanden Boer
3 Theoretical outline The central topic of this paper is the theory of language as a tool for nationbuilding, an issue which has been characterized by two opposing discourses with regard to the position of the German-speaking minority in Belgium. I will provide a brief contextual overview of the foundations for these two distinct viewpoints on the subject, before discussing my own data. View 1: ‘German-speaking Belgians should be Frenchified to become better Belgians.’ Following the annexation of the region in the immediate aftermath of the First World War, the areas surrounding Eupen, Sankt Vith and Malmedy were ruled as a colony by the Belgian government. Governor Baltia, who had previously been posted in the Congo, an African colony also belonging to Belgium, was put in charge of the region and governed from 1920 to 1925. The aim was to integrate the newly acquired areas into Belgium quickly. As most of the elite spoke only French, this lead to a drastic wave of ‘Frenchification’, with the French language being used in Belgium as a tool for nation building during this time. French had been implemented by the Belgian government in its African colony, and consequently this policy was also applied to the German-speaking area. French was the language of good Belgians. For Baltia the Belgification of EupenMalmedy implied Frenchification. ([translation A.V.B.] Wenselaers 2008: 63) It was Baltia’s duty to adapt the former Prussian territory […] to the Belgian structures. An important measure was the introduction of French […]. ([translation A.V.B.] Darquennes 2005: 6)
After twenty years of Belgian sovereignty, the areas around Eupen, Sankt Vith and Malmedy were annexed by Germany during the Second World War. The war had filled the Belgian population with aversion to Germany and the German language, and so when the German-speaking areas were liberated at the end of the Second World War, the cry for ‘Frenchification’ of these areas was immense. The inhabitants of these regions had fought
Language and Nation
259
with the German enemy as they had also done during the First World War. It made no dif ference to the Belgians that this time Germany had forced them to do so. After the war the Belgian government did not know what to do with the German speaking inhabitants. The attempts in the 1920 and 1930 to Belgify the population did not succeed. It was considered to eliminate the German language and culture immediately this time, starting with a radical denazification. ([translation A.V.B.] Wenselaers 2008: 105)
A remarkable trend emerged in the areas around Montzen, Bého and Arlon (see above). These, originally German-speaking areas began to voluntarily ‘Frenchify’ themselves in order to distinguish themselves from the ‘German enemy’. This tendency was already noticeable after the First World War but increased after the Second World War. The Prussian invasion in 1914 evoked a strong belgicist ref lex in the Old-Belgian areas, which was ref lected mostly in the language use. The dif ference between the own German dialect and the standard German was emphasized and French was used more frequently. The use of French became a sign of patriotic unionism. ([translation A.V.B.] Verdoodt 1968: 9)
View 2: ‘German-speaking Belgians are the last Belgians standing.’ The parliamentary year 2007–2008 marked the beginning of a political crisis in Belgium. Since its foundation Belgium has united two big language communities: one Dutch-speaking (Flemings) and one French-speaking (Walloons). The collaboration between them often creates problems. From the 1960s to the year 2001, several state reforms have redefined the Belgian state model. In 1993 this resulted in the creation of a federal state with widespread autonomy for Flanders and Wallonia. The Belgian federal structure still could not reconcile a stable balance between Flemings and Walloons, however, and the reforms continued after that. The tension reached a climax after the federal elections in June 2007 when it became impossible to form a government as a result of the disagreements between Flemings and Walloons over the aims of the new political manifesto, which was to include yet another state reform. The future of Belgium and the option of splitting
260
Anneleen Vanden Boer
up the country was publically debated. The German-speaking community was repeatedly overlooked in this discussion, and as a result, assumed the role of the underdog. For this reason it is often said that German-speaking Belgians are the last true Belgians standing: The Eastern Belgians are unknown to many Belgians. They are viewed as monarchists, Catholics or the best protected minority in the world. Or even as the last Belgians. ([translation A.V.B.] Wenselaers 2008: 165)
4 Method I will use data from a written survey I conducted in September and October 2008 among a cross-section of 977 non-German-speaking Belgians (NGB) and German-speaking Belgians (GB) between 20 and 74 years old, with the aim of revealing the current relationship between citizenship and language command within the German-speaking minority of present-day Belgian society. The survey has been designed specifically to enable the correlation of the scores relating to the variables ‘language command’ and ‘mother tongue’ with the scores relating to the variables ‘royalism’, ‘patriotism’ and ‘separatism’. The results will be analysed using the SPSS software and based on frequency tables and the Kendall tau b test for correlation.2 Where the variable ‘language command’ or ‘mother tongue’ is used, it is important to note that the scores for language knowledge are based on self-assessment by the respondent. Bias as a result of over- or underestimation of linguistic knowledge on the part of the respondent is, of course, possible. In some cases I will use an overall score for ‘language command’. This score is composed of separate scores for command of Dutch and/or 2
Correlations with a rho-value (ρ) higher than 0.05 are not significant. Correlations with a rho-value (ρ) smaller than 0.05 are significant. Correlations with a rho-value (ρ) smaller than 0.01 are highly significant.
Language and Nation
261
French and/or German. For each respondent the mother tongue has been not been considered in this overall score. When I use the label ‘foreign national language 1’ (FNL1) this refers to the foreign national language that is most likely to be the best known in a specific language group (e.g. French is the FNL1 of the GB). The label ‘foreign national language 2’ (FNL2) is the foreign national language that is more likely to be known, only after a certain level of the FNL1 has been reached. The label ‘Feeling of af filiation to Belgium’ refers to an overall score, composed by the individual scores on: – degree of royalism (Do you think it is desirable that Belgium has a king?) – degree of unionism (Do you think Belgium should be split up into two independent states, Flanders and Wallonia?) – degree of patriotism (To what extent do you consider yourself Belgian?)
These individual responses will be evaluated using likert items. This is a bipolar scaling method commonly used in questionnaires to measure the extent to which a respondent agrees or disagrees with a given statement, and to allow a degree of dif ferentiation in the answers.
5 Results 5.1 A loyal Belgian is likely to speak German Let us first take a look at the correlation between the feeling of af filiation with Belgium and the mother tongue of the respondents, and answer the question: Is there a specific language group in Belgium that feels significantly more af filiated to Belgium than the others? These results are shown in Table 11. Highly significant relations are indicated in bold.
262
Anneleen Vanden Boer
Table 11 Correlation between mother tongue and the sense of af filiation to Belgium GB + NGB
Feeling of af filiation to Belgium
French mother tongue
0.1386
Dutch mother tongue
< 0.0001 INVERSE CORRELATION
German mother tongue
< 0.0001
There is no significant relation as far as French-speaking Belgians are concerned. In Dutch and German-speaking Belgians, by contrast, we notice a high significance, but for opposing reasons within each group. If Dutch is the mother tongue there is a high probability that one will feel less af filiated to Belgium. Speaking German by contrast, manifests itself additionally in higher likelihood of feeling more af filiated to Belgium. This result also appears in the frequency tables of my questionnaire (Table 12). These show the results of GB and NGB. I dif ferentiate between GB from Malmedy and Waimes, situated in the French-speaking area (GB MAWA) and the GB from the residents of the German-speaking Community, situated in the German speaking area (GB GC). Table 12 Frequency distribution for the variables indicating the sense of af filiation to Belgium Frequency agrees with the presence of a king in Belgium NGB
GB MAWA
GB GC
Strongly disagree
18.3
8.8
6.9
Disagree
9.2
4.4
4.4
Neutral
17.9
8.8
18.6
Agree
13.3
14.7
10.1
Strongly agree
41.3
63.2
60.1
%
263
Language and Nation Frequency agrees with the separation of Belgium in Flanders and Wallonia NGB
GB MAWA
GB GC
Strongly disagree
54.4
82.1
80.3
Disagree
14.1
6.0
7.1
Neutral
7.1
4.5
4.4
Agree
8.3
1.5
3.9
Strongly agree
16.2
6.0
4.4
NGB
GB MAWA
GB GC
Very low
5.9
4.4
4.8
Low
6.3
1.5
3.5
Quite low
11.7
10.3
11.1
Quite high
25.5
20.6
17.9
High
28.5
20.6
28.1
Very high
22.2
42.6
34.6
%
Frequency degree of patriotism %
In contrast to NGB we notice a distinctly more positive answer pattern for GB. In particular, the GB living in the Malmedy-Waimesarea are shown to be the most loyal Belgians. 5.2 German-speaking Belgians are the most multilingual Before studying the relation between language command and Belgian identity, we should further investigate the level of language command of Belgian citizens (Table 13). In this paper we will only consider the language command of the Belgian national languages Dutch, French and German.
264
Anneleen Vanden Boer Table 13 Frequency distributions for the language command
Frequency degree of knowledge of Dutch NGB % DU FR None at all n.a. 7.2 Very poor n.a. 20.5 Poor n.a. 24.1 Suf ficient n.a. 25.3 Good n.a. 9.6 Very good n.a. 9.6 Native n.a. 3.6 Frequency degree of knowledge of French NGB % DU FR None at all 2.6 n.a. Very poor 3.2 n.a. Poor 9.0 n.a. Suf ficient 32.9 n.a. Good 31.6 n.a. Very good 18.1 n.a. Native 2.6 n.a. Frequency degree of knowledge of German NGB % DU FR None at all 9.9 51.9 Very poor 13.2 14.8 Poor 20.4 16.0 Suf ficient 36.2 11.1 Good 17.8 2.5 Very good 2.0 3.7 Native 0.7 0.0 Note. n.a.: not applicable DU: native Dutch speakers FR: native French speakers
GB MAWA
GB GC
14.1 18.8 15.6 32.8 12.5 4.7 1.6
16.8 10.6 18.7 30.2 16.5 6.7 0.5
GB MAWA
GB GC
0.0 0.0 0.0 4.4 17.6 47.1 30.9
1.4 1.7 4.8 17.3 37.9 32.4 4.6
GB MAWA
GB GC
n.a. n.a. n.a. n.a. n.a. n.a. n.a.
n.a. n.a. n.a. n.a. n.a. n.a. n.a.
265
Language and Nation
The frequency tables concerning the degree of knowledge of Dutch and French among GB clearly show that almost every GB has at least suf ficient linguistic command of French. GB MAWA emerge as having an exceptionally good knowledge of French. One GB out of two has suf ficient command or better of Dutch. For Dutch there is no big dif ference between GB MAWA and GB GC. The knowledge of the German language among NGB in general is quite bad, with French-speaking Belgians in particular scoring very poorly; eight out of ten French-speaking Belgians have no or only poor command of German. The German language level of Dutchspeaking Belgians is comparable to the Dutch language level of GB, which is poor to suf ficient. The language knowledge of French among Dutchspeaking Belgians is not as good as that of GB, but still better than the language level of Dutch among French-speaking Belgians. Overall the GB can be defined as the most multilingual Belgians, whereas French speaking Belgians are the least multilingual. 5.3 Multilingualism supports the Belgian identity of German-speaking Belgians Having successfully analysed knowledge of the national languages, we can now investigate the relationship between this knowledge and the feeling of af filiation to Belgium. I will do this by measuring the significance of the relation between both variables. The significant relations are underlined and high significance is indicated in bold (see Table 14). Table 14 Correlation between language command and the sense of af filiation to Belgium GB
Degree of royalism
Degree of unionism
Degree of patriotism
Sense of af filiation to Belgium
Command of French
0.375
0.718
0.015
0.086
Command of Dutch
0.709
0.462
0.002
0.026
Overall level of language knowledge
0.457
0.856
0.003
0.037
266
Anneleen Vanden Boer
To study this correlation, I will consider both the composite variable of ‘sense of af filiation to Belgium’ (in the last column) and the component parts of that variable. If we consider the composite variable ‘sense of af filiation to Belgium’ and its relation to language knowledge, we observe that, with the exception of the relationship with command of French, which is almost significant, the sense of af filiation to Belgium is likely to increase when the level of language knowledge increases. Next we will also consider the component parts separately. Here we notice that the degree of both royalism and unionism have no significant relation to language knowledge. In contrast, the correlation between patriotism and language knowledge turns out to be highly significant. In order to find out whether this result is characteristic for all Belgians, I conducted the same test for Dutch- and French-speaking Belgians. In both groups no significant relationship between language command and Belgian identity could be found. Therefore we cannot assume that multilingualism is, in itself, a mark of Belgian identity.
6 Conclusion The results of my empirical research appear to confirm some ‘stereotypes’ that appear in the academic literature on the subject: German-speaking Belgians are royalists and patriots. In this investigation I have demonstrated that Belgians with German as their mother tongue are likely to be more loyal with regard to Belgium than other Belgians. After the two World Wars it was impossible to think that the most loyal Belgians would speak German. To create ‘good Belgians’ a radical Frenchification of the German-speaking Belgians seemed the only way. Fifty years later, the social demographic has obviously changed, and the national identity of this minority is no longer supported by knowledge of one specific language. A good overall knowledge of all of ficial languages, however, is a good predictor of the strength of Belgian identity.
Language and Nation
267
Not only ‘Belgian-ness’, but also multilingualism, is a defining part of the identity of German-speaking Belgians, many of whom could even be additionally defined as bilingual French-German. This multilingualism and the strong sense of, particularly patriotic, af filiation to Belgium, seem concurrent in enhancing the mutual intensity of one another. It would have been interesting to discover that multilingualism is an intrinsic part of Belgian identity on a more national level too, but it does not seem to be significant for the Dutch and French linguistic groups. Neither multilingualism, nor mastery of a specific language, can be described as components of a more general Belgian identity.
References Darquennes, Jeroen. 2006. ‘Duits als autochtone taal in België: toekomstige onderzoeksdesiderata.’ In: Handelingen van de Koninklijke Zuid-Nederlandse Maatschappij voor Taal- en Letterkunde. Gent KZM, 93–109. Vanden Boer, Anneleen. 2008. ‘Die Deutschsprachigen in Belgien.’ Der Sprachdienst 6/08, 245–252. Vanden Boer, Anneleen. 2009. ‘Die Position der deutschsprachigen Minderheit im belgischen gemeinschaftlichen Rahmen – erste Befunde.’ Muttersprache, 13–32. Verdoodt, Albert. 1968. Zweisprachige Nachbarn: die deutschen Hochsprach- und Mundartgruppen in Ost-Belgien, dem Elsass, Ost-Lothringen und Luxemburg. Wien: Braumüller. Wenselaers, Selm. 2008. De laatste Belgen, een geschiedenis van de Oostkantons. Antwerp: Meulenhof f/Manteau.
Part 3
Demographics and Social Dynamics
Richard Ingham
The Decline of Bilingual Competence in French in Medieval England: Evidence from the PROME Database
Abstract The maintenance and demise of Anglo-Norman is examined using petitions in the Parliament Rolls of Medieval England (PROME) electronic database. Noun gender is shown to have been largely error-free in contexts involving preposition + article combinations around 1300, but to have become massively error-prone from about 1370 on. The change is argued to have been related to the disappearance of French as a vehicle language for education in first schools attended by clerks, around the time of the Black Death. Conventional accounts of later Anglo-Norman as an instructed foreign language are shown to be inadequate to deal with the phonologically driven patterning of other gender marking errors which, it is concluded, were an ef fect of spoken transmission.
1 Introduction In the later medieval period, vernacular languages in many regions of Europe began to challenge the hegemony of Latin over the written word for purposes of record-keeping. In England the principal beneficiary of the rising use of the vernacular as a written medium between about 1250 and 1370 was not English but French, in the insular variety which will be referred to here as Anglo-Norman, following the convention set by the Anglo-Norman
272
Richard Ingham
Dictionary (Rothwell et al. 2005).1 If standard textbook accounts are to be believed, this is in some ways surprising. To quote Singh (2005: 109): ‘As England moved into the thirteenth century, the tide turned increasingly in favour of English and proficiency in French waned.’ Yet far from declining, the use of French in written documents spread across genres in the later thirteenth and fourteenth centuries, in a way reminiscent of how a standard language is created (Haugen 1966). Indeed it looked to some in the earlier fourteenth century that French was threatening to become a kind of of ficial language.2 The far greater use made of French in charters, wills, chronicles, correspondence and so on at this time is easy to observe, and would certainly have been very noticeable to contemporaries: for a while, at least, French became virtually another language of record alongside Latin, which it had not been in the first two centuries after the Conquest. It is not easy to understand what enabled the survival of linguistic competence in French that permitted this expansion in the functional range of Anglo-Norman. Previous approaches to this question have proposed solutions that for one reason or another are unconvincing. Authors dubbed by Kibbee (1991) ‘French maximisers’, such as Legge (1980), maintained that the use of insular French was widespread and that it was acquired as a spoken vernacular until a late date; ‘French minimisers’, represented by Rothwell (1976, 1978), on the other hand, have claimed that French was taught in England as an instructed second language in medieval schools by this time, and was ‘no more a vernacular than was Latin’. French ‘maximisers’ have been discredited by a number of scholars (see e.g. Berndt 1972, Rothwell 1976), and the critique of them is typically
1
2
An alternative nomenclature would be to reserve the term ‘Anglo-Norman’ for the earlier, largely literary, period lasting until the middle third of the thirteenth century (Pope 1934), and use a term such as ‘Anglo-French’ for the later, mainly non-literary period. For reasons of institutional conventions, as well as in the interest of avoiding the dif ficult issue of what linguistic changes marked the transition from one period to the next, we have not done this, however. A concern which may have lain behind the accusation directed by Edward II at Philip of France at the beginning of Hundred Years’ War, that the French wished to replace English by the French language.
The Decline of Bilingual Competence in French in Medieval England
273
rehearsed in textbooks such as Lass (1987). The weaknesses in the arguments advanced by French ‘minimisers’ are less widely appreciated, but are nevertheless serious, and can be summarized as follows. First, if Anglo-Norman was an instructed language taught in a classroom, it is surprising that its spelling was so irregular. Second, French was still used in English grammar schools as a vehicle language through which to teach Latin in the first half of the fourteenth century, according to contemporary witnesses, the chroniclers Higden and Holcote (Orme 1973). This presupposes that when pupils started Latin they already had some French; teaching one unknown language through another unknown language is unlikely to have been educationally successful. So French cannot have been first encountered in grammar school, which pupils began at the age of seven. Third, pedagogical grammar treatises written around 1250–1300 and discussed by Rothwell (1976) and Hunt (1985) teach Latin tenses through French, not the other way round, so cannot be used as evidence of the instructed nature of later Anglo-Norman. The same applies to word lists put forward by Rothwell (1976) as pedagogical materials for the teaching of French. Fourth, the so-called ‘Treatise’ (a title created and adopted by modern editors of the poem) written by Walter de Bibbesworth in the later thirteenth century is the only item clearly in the category of a text for language learners in the relevant period, but it was intended to support language learning in an aristocratic home setting. It tells us nothing about school contexts, and how less well-of f scholars who went on to become clerks acquired their French. In short, the notion that French had by 1300 become a language of formal instruction in England rests on no empirical foundation. The problem we are left with, however, is to understand how AngloNorman was transmitted across three centuries following the Norman Conquest, the majority of the time with no, or almost no, remaining native speakers.3 If it was neither acquired as a mother tongue in the home nor taught in schools, in what context was it acquired? Present-day societies tell us that there are other options beyond this simple mother tongue vs. instructed foreign language dichotomy. In par3
The royal family is often thought to have been an exception here.
274
Richard Ingham
ticular, a second language may be learnt quasi-naturalistically in childhood, as for instance in immigrant communities in a country where a dif ferent public language from the home language is spoken. In second language acquisition in such contexts, age at initial exposure has repeatedly been shown to be a key determinant of native-like ability (Oyama 1976, Patkowski 1980, Johnson & Newport 1989). Although English would have been the first language acquired by users of later Anglo-Norman, they could in principle have achieved native-like ability in that language variety if they were exposed to it conversationally while young children. In this article the hypothesis will be adopted that those who were professionally involved with creating the texts which have come down to us, and which constitute our evidence base for the study of Anglo-Norman, acquired their knowledge of this variety in early to middle childhood from its use in spoken interaction, not as an instructed classroom subject. The contexts in which this spoken interaction took place were high-status households and the preparatory schools run by the church, known as song schools, to which future clerks were sent before entering grammar school (Orme 1973). If exposure to language took place in such settings, the grammar of Anglo-Norman could have been reliably acquired by young children, even though their pronunciation would have been heavily inf luenced by the ambient English-inf luenced phonology of Anglo-Norman. This scenario is more fully developed by Ingham (2010). In this article we explore one aspect of the linguistic predictions made by the hypothesis just outlined: the marking of noun gender agreement. Noun gender is a domain well suited to the research issue, since numerous studies of second language (henceforth L2) acquisition have established a clear distinction between the essentially target-like attainment of natur alistic child L2 learners and the error-prone performance of instructed adult learners (see e.g. Franceschina 2005 and references therein). Even advanced instructed learners make fairly frequent gender marking errors. Reportedly, they tend to be most accurate on gender marking on articles and somewhat more error-prone on other noun modifiers. These findings, though obtained in a modern context, give some idea of what to expect, had AN been an instructed L2.
The Decline of Bilingual Competence in French in Medieval England
275
Problems with noun gender agreement have indeed been observed in AN. Bolland (1914) commenting adversely on the standard of French in later AN, drew particular attention to gender agreement errors. Tanquerey (ed.) (1916: lv) found gender errors with determiners that were ‘dus a l’ignorance’ as he put it, as well as failures of adjective-noun gender agreement that he took to be either spelling mistakes or due to English inf luence. Kibbee (1996: 10) also saw gender errors as a factor ref lecting English inf luence on AN. However, it is not entirely clear how English inf luence would have operated in this domain. Since Old English noun gender distinctions had largely disappeared by Early Middle English, there was no question of the gender of an English noun af fecting that of an AN noun. Rather, such inf luence would presumably have involved neutralized gender marking,4 either by using a single form to replace the two members of a gender marking opposition, or by using the two forms indiscriminately. Indiscriminate usage is indeed observed, as in the following examples from about 1300 showing gender clashes between modifier and noun: (1)a
en cel prison
Bolland (ed.) (1914: 37)
in this (MASC) prison (FEM)
(1)b
sicum en acun rivere
Nichols (ed.) (1865: 404)
as in some (MASC) river (FEM)
(2)a
de ceste trepas
Bolland (ed.) (1914: 12)
of this (FEM) trespass (MASC)
(2)b
en le haute chemin vers Donestaple
Tanquerey (ed.) (1916: 50)
in the high (FEM) way (MASC) towards Dunstable
However, it is important to recognize that the AN gender agreement errors noted by earlier authors strongly tended to involve a final . A major inf luence on Anglo-Norman phonology was the loss of the final weak ‘e’ 4
As happened in Cappadocian Greek, under the inf luence of Turkish, which lacks grammatical gender (Karatseareas 2009).
276
Richard Ingham
sound in Middle English, already observable by the late twelfth century (Hogg & Denison 2006: 66), among numerous ef fects of English phonology on Anglo-Norman (Duf fell 2005). As noted by Pope (1934: 438), later Anglo-Norman spelling showed the ef fect of this tendency by the omission of final . So apparent gender errors such as (1) – (2) may not have been a matter of imperfect grammar, but rather the erosion of particular phonological means of marking gender distinctions.5 In order to determine whether noun gender in AN was impaired independently of phonology, specifically the loss of weak ‘e’, Ingham (2010) therefore analysed AN texts for gender agreement on pairs of premodifier forms having very distinct phonology, the possessive determiners mon/ma, and son/sa. These forms, in addition to their dif fering vowel qualities, were distinguished by the nasal consonant in the masculine, which is absent in the feminine. Had they too shown gender agreement errors,6 this would have furnished good evidence for imperfect grammar learning, not just phonological inf luence from English. Ingham (in press) searched the petitions in the electronic version of the Parliament Rolls of Medieval England (Given-Wilson et al. 2005), henceforth referred to as ‘PROME’, covering the years 1310–1399. Errors with son and sa were found to be extremely rare events indeed until the last few decades of the fourteenth century, averaging just one in twelve parliamentary sessions. Thereafter, however, they became much more common, with 34 errors occurring in the 29 sessions between 1371 and 1399. Until then grammatical gender properties of nouns must have very reliably known to insular users of French, despite their English-inf luenced phonology. One might suppose that the presence of any errors at all is a sign of non-nativeness in language learning, but this would be over-strong. Continental French texts very occasionally show non-conventional use of noun gender, and in modern studies of noun
5 6
This concept is already familiar in medieval French dialect phonology in the case of Picard, which neutralised the distinction between le/la, me/ma, etc. (Pope 1934: 488). Determined with reference to a standard Old French Dictionary, Greimas (1992).
The Decline of Bilingual Competence in French in Medieval England
277
gender acquisition native speaker errors have also been reported.7 The same would apply to insular French too, until the second half of the fourteenth century. So the results were consistent with imperfect grammar learning by AN users only from the 1370s onwards.
2 The study The claim made by Ingham (2010) that AN users continued to observe noun gender with possessive determiners to a level of accuracy comparable to that of native speakers is acknowledged to be controversial: previous commentators on Anglo-Norman are near-unanimous that its users, by the fourteenth century, were not grammatically proficient. Even in terms of the hypothesis to be adopted here, that of quasi-naturalistic childhood learning of AN, it is open to objection. If only possessive determiners were reliably marked for gender agreement in the language they heard, could naturalistic learners have acquired gender properties of nouns from such contexts, which were relatively uncommon compared with the high frequency of article and premodifier forms? This study therefore aims to identify other contexts, in addition to possessive ones, in which learners of Anglo-Norman could have acquired noun gender. The most frequent opportunities for learners to have heard gender marking agreement were provided by definite articles; indefinite noun phrases in Old French did not yet need an indefinite article. As noted by Tanquerey and others, later AN writers are known to have used le with feminine nouns. An initial examination of PROME indicated that such errors are absent from the petitions presented up to about 1310, suggesting that in this genre at least definite articles resisted gender confusion better
7
A study reported in Franceschina (2005) show Spanish L1 speakers making gender errors on about 1 per cent of occasions (tokens).
278
Richard Ingham
than did weak ‘e’ contexts.8 But from 1315 onwards, clear gender errors with definite articles occur. Between 1315 and 1330, they included the following feminine nouns given le : ley, rente, garde, copie, communalté, tour (‘tower’), feste, discomfiture, journee, and masculine nouns given la: roialme, castel, service, évesché.9 However, the opposition le: la is arguably unstable since the masculine form involves the weak ‘e’ vowel, so there could be doubt over whether such errors are really a grammatical or a phonological phenomenon. If the latter, they would not provide evidence for the loss of noun gender as such. The most fruitful context for investigation seemed to be the observance of gender agreement when a definite article combined with a preceding preposition, especially à and de. As in Modern French, the Old French masculine article form contracted on to the preposition as au and du, alternatively spelt al and del. The feminine counterparts à la and de la also required knowledge of noun class membership, of course, but here it was felt that the growing confusion between le and la observed from 1315 onwards, as mentioned above, would have complicated the issue. Accordingly the data analysis was restricted to the masculine preposition plus article combinations, al/au and del/du, and to see if they occurred only with a masculine nouns. Three exclusionary categories were determined. Since grammatical gender was at issue, animate nouns, where natural gender usually determines grammatical behaviour, were not used. Also excluded were cases where the target variable is followed by an element beginning with a vowel or with . An expression such as del annee might have been written in this form to represent the elision of the /a/ of a feminine definite article, as in Modern French; in this case the use of del preceding a feminine noun 8
9
Le was regularly observed with feminine nouns beginning with a vowel, e.g. le almone, le esglise, le isle, le universeté regularly take le rather than la in the Edward I petitions. This should be seen as a phonologically conditioned sign of the elision of the vowel, or of its reduction to weak ‘e’, rather than a gender error, since feminine nouns beginning with a consonant did not show this. No gender variation is reported for these items in continental Old French by ToblerLommatzsch (1956–).
The Decline of Bilingual Competence in French in Medieval England
279
should not count as a gender agreement error. Note that exclusion was determined on the basis of the word following the article, which was most often the noun, but could be a premodifier of the noun, e.g. avandite in the expression del avaundite dette. Finally, cases were observed where du was written for de e.g. du son doun bien, en fraunchise com hors du fraunchise, or au is written for à. These were not contexts for a definite article, so they were excluded. In sum, therefore, the data to be analysed consisted of cases where a definite article is grammatically expected, preceding an inanimate noun, where the noun or the first premodifier preceding it and following the article had a consonant as its initial phoneme. In scoring frequencies of target items in a corpus, there is a choice to be made between recording the type-frequency or token-frequency of the item or combination targeted. The justification for doing the latter – the approach usually followed in the L2 learning studies cited above – is that in principle each instance is a new opportunity to display correct knowledge or to make an error. However, the often repetitive discourse of the petitions in PROME made this an unsatisfactory procedure: the results would have been f looded with thousands of correct uses of very commonly used nouns, giving a percentage of errors that risked understating the incidence of errors AN users made in more ordinary discourse. Accordingly, we scored noun types used in the target combinations. This gave a much higher profile to any errors that were found. A single error with a noun occurring erroneously in that context would be counted, even though the rest of the time its gender agreement properties were target-like. It was felt that this procedure would pick up the tendency of an instructed L2 learner to make occasional careless mistakes, rather than burying them amongst huge numbers of correct instances. The main data collection concerned the preposition + article data analysis, as described above. However, the performance of AN users in the weak ‘e’ contexts was also assessed, as a comparison. Four frequent modifiers cel, cest, nul, and tout were selected for this purpose, and the gender of noun types co-occurring with the masculine form was scored. Both analyses are reported in the next section. The same database used by Ingham (2010) was searched, except that this time the petitions to parliament in the time of Edward I (1272–1307)
280
Richard Ingham
were included. Unlike other data files, the petitions for this period cannot be generally linked to specific dates, only to the reign as a whole. It was felt that this was not a serious problem, and was outweighed by the need to give as great a time depth as possible to the analysis of gender marking. The study went up to the last parliament in the reign of Richard II (1377–1399). Although there were still petitions in French up to around 1430, by the end of the fourteenth century, grammatical marking involving prepositions and articles in AN was blatantly failing to respect the grammar of French, cf. the uses of a le and du in the following extract from the first parliament of Henry IV: (3) L’evesqe d’Excestre… fist arester le dit Thomas count de Warrewyk, et luy amesner a le tour du Loundres October 1399 ‘The Bishop of Exeter… had the said Th. Earl of Warwick arrested and taken to the Tower of London’
The issue was when this systematic breakdown in the syntax of gender took place.
3 Results Table 15 shows frequencies of masculine and feminine nouns co-occurring with the four preposition + article forms, in each 30-year period of the PROME database between 1280, the presumed date of the earliest Edward I petitions in French, and 1399.10
10
In fact the latest parliament held by Richard II for which petitions are extant was in 1397.
281
The Decline of Bilingual Competence in French in Medieval England Table 15 Frequencies of masculine and feminine nouns used with masculine preposition + article forms in PROME, 1280–1399 del
du
Years
MASC.
FEM.
TOTAL
MASC
FEM
TOTAL
1280–1309
42
1
43
54
1
55
1310–1339
34
5
39
60
4
64
1340–1369
26
2
28
59
7
66
1370–1399
86
94
180
70
54
124
MASC.
FEM.
TOTAL
MASC
FEM
TOTAL
1290–1309
23
1
24
24
2
26
1310–1339
24
0
24
46
4
50
al
au
1340–1369
22
3
25
35
7
42
1370–1399
56
58
114
49
33
82
These figures reveal a very clear change across the time period. The ‘Fem.’ columns, which show the frequencies of errors, display a low level of occurrences in almost every category until 1370–1399. Then errors become approximately as frequent as correct uses. The percentage occurrence of feminine nouns for all four preposition + article types combined is represented in Figure 6, which displays an unmistakeable ‘broken stick’ trend, usually indicative of a change in the status of the language trait under investigation. Were some preposition + article combinations contributing more than others to this trend? It is conceivable, for example, that one preposition changed its distribution more than the other, or that forms ending in were used with more respect for gender than those ending in . These possibilities are examined in Figure 7. Separate graph lines are shown for those cases where the combination ended in and , for cases where the preposition was de, and for those where it was à.
282
Richard Ingham
Figure 6 Percentage occurrence of feminine nouns overall, with del/du and al/au
Figure 7 Percentage of occurrence of feminine nouns with de and à, and with combinations ending in -l and those ending in -u
283
The Decline of Bilingual Competence in French in Medieval England
It can be seen that the trend was virtually identical, regardless of the preposition and regardless of whether the contracted form ended in -l or in -u. With all prep + article combinations, there was a shallow increase to around 10 per cent in the percentage of feminine nouns used with masculine contracted forms between the later thirteenth century and the middle of the fourteenth century. Thereafter, they all sharply increased, to a level of around 50 per cent of cases: in 1370–1399 all contracted forms were thus being used indiscriminately, regardless of noun gender. Whether this means that noun gender as such had now been completely lost in AN is a question to which we return below. In any case, it is clear from these results that the 1370s constituted a major turning point in the observance of noun gender marking: prior to then, errors were rather rare. A way of defending the instructed L2 position might be to take these results to show that at first French was being accurately taught, but that in the last decades of the century the ef fectiveness of instruction declined. Although this is not a hypothesis we find at all plausible, it is worth addressing. If it were correct, we would surely expect a similar trend with gender agreement errors involving weak ‘e’, since they too would surely have been subject to the vigilance of language instructors. A subset of the data was therefore searched to find out if the rate of weak ‘e’ errors increased in the late fourteenth century in line with errors on proposition + article combinations. The Edward I petitions, from around 1300, were compared with petitions to Parliament in the 1370s. Specifically the masculine premodifier forms cel, cest, nul and tut, on which errors had been previously noted, were targeted. Scoring by noun type was done on the same basis as for Table 15, and results are shown in Table 16. Table 16 Frequency of masculine and feminine noun types co-occurring with masculine premodifier forms cel, cest, nul and tut, late thirteenth century/ early fourteenth century and 1370s cel MASC.
cest FEM.
TOTAL
MASC
FEM
TOTAL
1280–1307
20
8
28
6
6
12
1370s
10
6
16
15
10
25
284
Richard Ingham nul MASC.
tout FEM.
TOTAL
MASC
FEM
TOTAL
1290–1307
16
8
24
13
4
17
1370s
17
10
27
18
8
26
Around the turn of the thirteenth century, cel, cest, nul and tout co-occurred with a total of 55 masculine and 28 feminine nouns, the latter counting as errors and constituting 33.7 per cent of instances. In the 1370s the corresponding figure was 60 masculine and 34 feminine nouns (36.2 per cent). It seems no great change took place in the abilities of AN users regarding gender marking with weak ‘e’ forms in the intervening 70 years. Errors where weak ‘e’ was involved were already common at the beginning of the period studied, unlike the very low incidence of errors at that time with preposition + article forms. The constant error rate in the former and the sudden rise in the latter would be very hard to explain in terms of some notional change in the standard of instructed learning. In the remainder of this study, therefore, we shall concentrate on the hypothesis that later AN was acquired naturalistically, at least until around the mid fourteenth century. However, there is still a methodological issue that needs to be addressed. As we saw in Table 15, the number of contexts for appropriate use of del, al etc. for the three decades 1370–1399 is much greater than in the earlier 30-year periods, by a factor of 4 or 5 to 1, in the case of all items except du. This is because the number of petitions in French grew in the later fourteenth century, as did their length in many cases. There may then be a concern that the dif ference in number of contexts biased the outcome by making it more likely to get gender errors in the later period, since more opportunities inevitably arise in a larger body of data. If so, this would invalidate the conclusion that the key factor in the incidence of gender errors was competence in AN in an earlier versus a later time-period. To address the problem, one would ideally take a much larger number of contexts from an earlier period, and see whether gender errors were still negligible. Within the confines of the PROME petitions, this solution is unavailable, so instead it was determined to lump together the Edward I
285
The Decline of Bilingual Competence in French in Medieval England
and Edward II petitions, covering c. 1280–1324, and compare them with those of Richard II (1377–1399). In this way an earlier and a later body of data would be created which would now be considerably closer in terms of sample size, as can be seen from the fact that the form del achieved 662 hits in the Edward I + Edward II petitions, and 1067 hits in the Richard II petitions. The figures still show a numerical advantage for the later period, but it is a far less disproportionate one than the early and late overall contexts in Table 15. Using these two analysis periods, frequencies of masculine and feminine noun types occurring with del were tabulated, as in Table 17. Table 17 Frequencies of del with masculine and feminine nouns, 1280–1324 and 1377–1399 Masculine
Feminine
Edward I and II petitions
50
1
Richard II petitions
81
71
In line with their roughly 60 per cent larger sample size for del, the Richard II petitions produced rather over half as many masculine noun types again as the Edward I and II petitions. But the frequency of feminine nouns rose from 1 to 71, dwarfing the increase in sample size. The greatly increased frequency of late period gender errors on preposition + article combinations in the main data in Table 15 does not seem, therefore, to be merely an artefact of sampling. The much greater error frequencies from the 1370s on seem to have been a genuinely new development.
4 Discussion Two main conclusions seem warranted on the basis of the findings presented above. One is that in the early fourteenth century the writers of these petitions showed a highly accurate knowledge of noun gender.
286
Richard Ingham
The other is that in the last part of the fourteenth century they either lost that knowledge or at least ceased to access it when producing preposition + article combinations. If these claims are to be properly informative, an account is needed of why the observed change took place when it did: what was upholding competence in insular French up to that time, and why did it then give way? The contrast between the frequent agreement errors in weak ‘e’ contexts (Table 16), and their avoidance with preposition + article combinations in the same time-frame (Table 15), makes it dif ficult to appeal to explicit instruction as a way of accounting for the maintenance of Anglo-Norman in the early fourteenth century. If the people who wrote the documents analysed were being taught French as a foreign language, some non-trivial level of gender assignment errors would be expected across various contexts: the gross imbalance discovered between preposition + article contexts and weak ‘e’ contexts would have no reason to arise. If AN was still being learned naturalistically from spoken interaction, however, the outcome observed is readily explained. Child learners would have heard gender marking consistently realized in the preposition + article forms, but not with the weak ‘e’ forms. They would have registered the gender properties of nouns on the basis of their distribution in input – young children’s abilities in this domain are well documented (see e.g. Clark 1985, Granfeldt 2005). AN users’ largely successful performance up to the mid-fourteenth century on possessive determiners, reported by Ingham (2010), had the same ontology. We do not advocate a wholesale reversion to the ‘maximising’ position adopted by Legge (1980), who made two logically distinct claims, one that the use of French was widespread in later medieval English society, and the other that it was a genuine spoken language. The first claim was undoubtedly overstated. Only those with access to a school education, or to life in aristocratic households, would have been able to benefit from childhood exposure to French. Consequently, speakers who acquired a naturalistic competence in French would have made up only a very small proportion of the population as a whole. To this extent the ‘minimising’ perspective taken by Berndt, Rothwell, Lass and others is valid. However, for the people using it that does not mean that AN was reduced to being
The Decline of Bilingual Competence in French in Medieval England
287
a poorly known foreign language. It is our view that spoken transmission in childhood meant that it was maintained within these social circles as a natural language variety, just as Legge envisaged it, until the conditions for its transmission collapsed. The findings on gender agreement in this study and in Ingham (in press) converge on a similar time for when that process took place. By the 1370s, gender marking errors outside weak ‘e’ contexts were becoming common. Why should this have happened just then? To attribute it to English inf luence, as if the latter would have waited until that point to intervene, is unconvincing: AN users had mostly been bilingual for around two centuries (Short 1980). In our view, a crucial point is that at that time the generation of clerks growing up after the Black Death would have been entering the profession. According to Trevisa, around that time the use of French as a medium language in grammar schools was abandoned. Kibbee (1991) has suggested that the Black Death, whose incidence was particularly high in towns and religious communities where AN was used, was related to the demise of AN. We would add that it was the role of priests to look after song schools, so the admission of novices to the clergy having little knowledge of French would have spelt the end for the transmission of AN via that route. Pupils would then have arrived at their grammar school without the basic knowledge of the language gained from spoken interaction that their predecessors would have enjoyed prior to the Black Death. On the hypothesis adopted here, clerks growing to maturity around 1370, and writing the documents surviving from that time, would have been the first generation not to have experienced childhood naturalistic exposure to French in song schools. A dif ference between the outcome of the present enquiry and that of Ingham (2010) must nevertheless be acknowledged. The latter found that from 1310 to 1369 gender agreement errors were negligible. However, as shown in Table 15, the incidence of gender agreement errors with preposition + article forms between 1310 and 1369 stood at over 30 errors (not far short of 10 per cent of the 338 contexts), which was surely too high to consider as negligible. Some loss of the ability to mark noun gender on preposition + noun combinations evidently started to take place before 1370.
288
Richard Ingham
Why should a somewhat higher rate of agreement marking errors have occurred in the preposition + article data than with the possessive determiners analysed by Ingham (in press) during the decades preceding 1370? If we look at the problem from a language learner’s perspective, two sorts of information must be acquired in order to mark gender accurately. The gender of the noun must be known, and specific agreeing forms, such as al, must identified as part of the gender agreement system. These are logically separate, as we see from the existence of the so-called ‘epicene’ (i.e. invariable) adjective forms such as grant, which in Old French did not show gender agreement; the use of grant as a gender-neutral form obviously did not mean the speaker or writer was unaware of noun gender as such. Likewise, the use of al or del with a feminine noun may not have denoted lack of knowledge of gender assignment: the contracted form itself might no longer have been accurately identified as restricted to masculine contexts, rather than behaving as an invariable form. In the long run, of course, as previously gender-marked forms were neutralized, the fewer opportunities there would be for noun gender distinctions to be realized on systematically contrasting forms, and eventually the noun gender system would collapse. This gradual process of losing gender marking context by context seems to have been at work in the later stages of AN: first, by the end of the thirteenth century gender contrasts involving weak ‘e’ on noun modifiers were no longer being reliably made, then as the fourteenth century went on errors arose with preposition + article combinations, and finally in the last decades of the century possessive determiners become error-prone. The process followed this particular sequence because it was originally driven by phonology. Gender agreement with modifiers such as cel and cest was distinguished only by a weak ‘e’ syllable, which was liable to phonological elision, so this was the most vulnerable context. Le and la before a consonant formed a more distinctive contrast as the syllable also contained the initial consonant segment /l/ and so could not simply be elided.11 Accurate agreement marking in PROME was found to have lasted 11
In modern French fast speech, the /l/ segment can assimilate to a vowel at the end of the preceding word, e.g. /tylmeisi/ ‘Tu le mets ici’. However, in medieval French and especially in AN, the retention of word-final consonants would have militated against such a process.
The Decline of Bilingual Competence in French in Medieval England
289
somewhat longer here. Still, the opposition involved weak ‘e’ and to that extent was fragile. Le increasingly became used with feminine nouns, so al and del with a feminine noun could arise if the weak ‘e’ of le preceding a feminine noun was elided, as was prone to happen. This stage could ensue only once the le/la contrast started to erode, which was not yet the case in the late thirteenth century in PROME, but happened in the first half of the fourteenth century; we then see the association of del and al with masculines starting to be undermined. In possessive determiners, on the other hand, which were unaf fected by the attrition of weak ‘e’, gender marking remained highly accurate, only starting to show frequent errors from the 1370s onwards. The observed sequence of gender agreement breakdown thus followed a phonologically motivated series of steps, from the most vulnerable (weak ‘e’ in modifiers), to the least vulnerable (possessives), via preposition + article combinations as an indirectly vulnerable category. It originated as a phonological process but eventually came to af fect gender marking as a grammatical property. In modern advanced instructed L2 learners, a common finding, already noted, is that gender agreement is most successfully achieved with articles, rather than other types of modifier. The PROME data do not show this pattern. Agreement with articles, both per se and in preposition + article combinations, starts to vacillate before agreement with possessive determiners does. The nature and sequencing of the error types observed here show a strong ef fect of phonological conditioning, i.e. a natural process of language transmission at work which was reliant on spoken communication rather than formal instruction.
5 Conclusions In this study an empirical investigation has been carried out of the accuracy of the language used in approximately the final hundred years of the career of AN. The petitions to Parliament in the electronic version of PROME were analysed in order to establish the writers’ level of accuracy on gender
290
Richard Ingham
marking, known to be a good discriminant of native-like ability in a language. The main focus was on the accuracy of gender agreement with preposition + article combinations. For most of the period studied, those who wrote the petitions in PROME showed overwhelmingly correct use of the relevant forms depending on the gender of the noun. The key period in which change in the status of AN became observable was the decade of the 1370s, when errors in preposition + article forms suddenly ballooned. The results of this study are consistent with those of Ingham (2010) on gender agreement with possessive determiners in that the main change took place in the 1370s. Until then, gender errors in Anglo-Norman were largely restricted to contexts where phonological attrition had eroded the means of making gender marking contrasts. Even in the earliest files, from around 1300, they were quite numerous in contexts where phonological neutralization of gender marking in AN is known to have been present. But preposition + article forms, the domain studied here, maintained accurate gender agreement until much later. The sequence of contexts in which gender errors arose seems to bear witness to a transmission scenario up to the mid fourteenth century in which spoken input formed the source of learners’ knowledge of the language, not formal instruction. At this distance in time it is clearly impossible to be certain about the precise nature and extent of the contexts in which this took place. By analysing the textual record, however, it becomes possible to advance some tentative ideas about the way this took place. The PROME database is particularly advantageous here, as the texts it makes available are based on manuscripts roughly contemporaneous with their composition, and are within the same genre, written by the same sort of people on the same sort of topics. Consequently, the changes observed in the series of petitions in the behaviour of the linguistic variables selected can be reliably imputed to the evolution of the language over time, rather than to extraneous complicating and confounding factors which so often bedevil diachronic research. Our findings suggest that Anglo-Norman lost its status as a spoken natural language among educated professionals after the Black Death, rather than following the loss of Normandy to the English crown in 1204, a landmark regularly mentioned by textbook authors seeking politico-historical
The Decline of Bilingual Competence in French in Medieval England
291
explanations for linguistic developments. The expansion of the functions of insular French in the later thirteenth century and fourteenth century was not the work of clerks struggling to write an artificial language they no longer properly controlled, the scenario retailed in many mainstream treatments. It took place because those who were responsible for leaving the documentary records of medieval England were still competent in the living language they had acquired in childhood, which was now achieving the higher status of a second language of record, probably with the active encouragement of the royal government (Lusignan 2004). The extensive use of insular French as a written vernacular at this time is only a paradox, then, if viewed through the fallacious assumption of recent decades that it was no more than a rather poorly learnt written code. Its superficial dif ferences from continental French, though real, have for too long inhibited a proper understanding of how its survival, and also its eventual demise, related to the socio-historical context of the late medieval period in England, when its matrix of transmission was destroyed by the impact of the Black Death.
References Berndt, R. 1972. ‘The period of the final decline of French in Medieval England (14th and early 15th centuries).’ Zeitschrift für Anglistik und Amerikanistik, 20, 341–369. Bolland, W. (ed.). 1914. Select Bills in Eyre, AD 1292–1333. Vol. 30. London: Selden Society. Clanchy, M. 1993. From memory to written record. England 1066–1307. Cambridge, MA: Wiley-Blackwell. Duf fell, M. 2005. ‘Some Phonological Features of Insular French: A Reconstruction’ In: R. Wright & P. Ricketts (eds), Studies on Ibero-Romance Linguistics Dedicated to Ralph Penny. Newark, NJ: Juan de la Cuesta, 103–125. Franceschina, F. 2005. Fossilized Second Language Grammars: The acquisition of grammatical gender. Amsterdam: Benjamins. Granfeldt, J. 2005. ‘The development of gender attribution and gender agreement in French: a comparison of bilingual first and second language learners.’ In:
292
Richard Ingham
J.-M. Dewaele (ed.), French as a foreign language: Multidisciplinary approaches. Clevedon: Multilingual Matters, 164–190. Greimas, A. 1992. Dictionnaire de l’ancien français. Paris: Larousse. Hunt, T. 1985. ‘Anecdota Anglo-Normannica.’ The Yearbook of English Studies 15, 1–17. Ingham, R. 2010. The transmission of later Anglo-Norman: some syntactic evidence. In: R. Ingham (ed.), The Anglo-Norman language and its contexts. Woodbridge: Boydell. Johnson, J. & E. Newport. 1989. ‘Critical period ef fects in second language learning: the ef fect of maturational state on the acquisition of English as a second language.’ Cognitive Psychology 21, 60–99. Kibbee, D. 1991. For to Speke Frenche Trewly: The French Language in England, 1000– 1600. Amsterdam: J. Benjamins. Kibbee, D. 1996. ‘Emigrant languages and acculturation: the case of Anglo-French.’ In: H. Nielsen and L. Schoesler (eds), The origins and development of emigrant languages. RASK supplement No. 6. Denmark: Odense University Press, 1–20. Lass, R. 2006. ‘Phonology and morphology.’ In: R. Hogg & D. Denison (eds), A History of the English Language, 43–108. Legge, M.-D. 1980. ‘Anglo-Norman as a spoken language.’ Anglo-Norman Studies 2, 108–117. Lusignan, S. 2004. La langue du roi. Paris: Presses Universitaires de France. Nichols, F. (ed.). 1865. Britton: The French Text Carefully Revised with an English Translation, Introduction and Notes. 2 vols. Oxford: Clarendon Press. Orme, N. 1973. English Schools in the Middle Ages. London: Macmillan. Oyama, S. 1976. ‘A sensitive period for the acquisition of a nonnative phonological system.’ Journal of Psycholinguistic Research 5:3, 261–283. The Parliament Rolls of Medieval England, 1275–1504 [CD], gen. ed. C. Given-Wilson et al., Scholarly editions, 2005. Patkowski, M. 1980. ‘The sensitive period for the acquisition of syntax in a second language.’ Language Learning 30, 449–472. Pope, M. 1934. From Latin to Modern French, with especial consideration of Anglo-Norman: phonology and morphology. Manchester: Manchester University Press. Rothwell, W. 1976. ‘The role of French in thirteenth-century England.’ Bulletin of the John Rylands University Library of Manchester 58, 445–466. Short, I. 1980. ‘Bilingualism in Anglo-Norman England.’ Romance Philology 33, 467–479. Singh, I. 2005.The History of English: A Student’s. Guide. London: Hodder Arnold. Tanquerey, F. (ed.). 1916. Recueil de lettres anglo-françaises (1265–1399). Paris: Champion. Rothwell, W., Gregory, S. Trotter, D. 2005. Anglo-Norman Dictionary: revised edition, A–C; D–E, 2 vols. London: MHRA. Wilson, R. 1943. ‘English and French in England 1100–1300.’ History 28, 37–60.
Rembert Eufe
Merovingian Coins and Their Inscriptions: A Challenge to Linguists and Historians
Abstract The present article gives a report on a joint project of historians and linguists who are engaged in examining a collection of 478 Merovingian coins conserved in the Bode-Museum in Berlin. The coins minted in the Frankish kingdoms from the late sixth until the early eighth centuries are of a special type: they show a large variety of dif ferent personal and place names. The anthroponyms of the so-called monneyers are of Gallo-Latin and Germanic origin, inviting conclusions on the relations and the mixing between Gallo-Roman and Germanic population groups and their language uses. In contrast, among the toponyms those with Gaulish origin largely prevail. It is considered how both linguists and historians can use this source to gain new insights.
1 Introduction From 2006 to 2009, historians and specialists in German and Romance linguistics from the universities of Berlin, Paderborn and Regensburg carried out a research project focused on a collection of 478 Merovingian coins conserved in the Münzkabinett of the Bode-Museum in Berlin. Their aim has been to compile a printed and an electronic catalogue of this second largest collection of Merovingian coins in the world.1 1
Greule, Jarnut, Kluge & Selig (in preparation) and the digital version IKMK, already available on the museum website.
294
Rembert Eufe
Why are linguists involved in a project like this? Normally, coins display just the name of a king or an emperor, perhaps with a short inscription in his praise or something similar. We do have coins of this type from the Merovingian period, but mostly from the earliest part (see below). From the late sixth until the early eighth centuries, the coins were dif ferent: they came to show a large variety of dif ferent personal and place names, whose number has been estimated at 1,2002 and 600 respectively for the 8,000–10,000 coins alone that we know of at present (Felder 2003: 22). Given these figures, the identification of the names and the explanation of their structure is an ambitious but also very fruitful task.
2 The value of the Merovingian coins as a source It should be recalled that the Merovingian period has left us only very few sources of linguistic information – above all lives of saints and chronicles (posing many unanswerable questions concerning their origin and transmission, as it is typical for texts conveyed by medieval manuscripts). As for the approximately 200 known Merovingian charters, two thirds of them prove to be falsifications, following the recent edition by Kölzer (2001: XII). Only the authenticity of the thirty-eight original charters conserved as originals (not as copies) remains beyond any doubt. These charters contain 12,000 words – whereas one of the most important chronicles, the ten books of histories written by the bishop Gregory of Tours (Buchner 1964) at the end of the sixth century, contain 120,000 words! An advantage of coins is that they are often easier to localize, especially when place names occur on them, as is the case on many of our coins (other wise the find spot and the style and motif of the illustrations are of help). In addition, they can be easily dated, e.g. on the basis of the names of kings mentioned on the coins, where the dates of their reigns are known, but also in the light of the dif ferent coin types, in particular silver or gold coins. 2
Kluge (2007: 83) even gives the number of 2,000 personal names.
Merovingian Coins and Their Inscriptions
295
However, it must be emphasized that not all the coins of fer fairly readable inscriptions. Sometimes the f lan was too small for the die, so that only the lower part of the legend appears. In other cases, the letters are blurred because they were coined with a die worn of f by long previous use. Furthermore, the coins raise considerable philological problems, not only due to abbreviations, but also due to inaccurate engraving (note that the coins are very small, being only the size of a thumbnail), including the commutation of letters. Even the direction of the legends cannot be taken for granted, as is shown in an example mentioned by Felder (2003: 25): according to him, the toponym +COSRVBET on a Merovingian coin is actually a ghost-name. When read in the opposite direction and with a G in the form of an S, it appears as TEBVRGO, to which [STRA]+ can be added which then reads STRATEBURGO C[IVITATE], Strasbourg.3 It has to be underlined that the value of the coins as a linguistic source is very dif ferent for scholars of Romance and Germanic languages. Because the oldest preserved Franconian texts date from the ninth century, the attested names of Germanic origin are important as remnants of older lexical material for Germanic linguists. But the linguistic environment, to which the legends belong as written sources, is primarily a Latin and Pre-Romance one; not surprisingly, the Latin of the inscriptions does not correspond to the standards of Classical Latin, but exhibits the graphophonetic and morphological characteristics of Merovingian Latin which is of particular interest to Romance linguists. An adequate interpretation of the data has to account for the context of use and production in terms of a recontextualization,4 i.e. reconstruction of the sociolinguistic and pragmatico-textual conditions, crucially based on the findings of historians.
3 4
On other coins, the city appears with its Gaulic name ARGENTORATE (see below). Selig / Eufe (in preparation) and the considerations in Greule, Jarnut, Kluge & Selig (in preparation).
296
Rembert Eufe
3 The Berlin collection The 478 Merovingian coins conserved at the Bode-Museum can be assigned to three consecutive periods. The first period, ranging from 500–585, is the period of so-called ‘pseudo-imperial’ coins, emulations of Roman coinages, showing more or less botched or feigned imitations of the model legends. Eighty-five coins belong to this period; however twelve of them cannot be considered as ‘pseudo-imperial’ but ‘royal’ coins with the names of Frankish sovereigns. The second period, running from approx. 585–675, is characterized by ‘typical Merovingian’ gold coinages with frequent mentions of monneyers’ names. In the Bode-Museum collection, a majority of 268 coins belong to this period, including thirty-one royal coins. During the third period from 675–750, the coinage changed in material, switching to silver. One hundred and eighteen coins of the collection stem from this period. To complicate matters, besides two bronze coins even five counterfeits are present, produced mostly in the nineteenth century.
4 Interdisciplinary aspects The central issue of this chapter is the usefulness of collaboration between historians and linguists and their dif ferent perspectives; in this light, research on the coins with their legends consisting mainly of names opens the question on the importance of onomastic research for linguists and historians. In linguistics, onomastics certainly has not been very popular in recent decades. One reason might be that it represents an area of study within linguistics which has the strongest appeal for the amateur researcher, which makes the value of many investigations hard to evaluate. Furthermore, in many respects (above all from a semantic perspective) names cannot be seen as being fully part of the language system. In the perspective of linguistic approaches which insist on explaining language solely in terms of its internal structure, the strong interdisciplinarity of
Merovingian Coins and Their Inscriptions
297
onomastics represents a further disadvantage: for it is usually impossible to focus on a name without also looking for information on its bearer, be it a person (his or her dates of birth and death, origin, relatives etc.), a settlement (its foundation, size, geographical position, population etc.) or something else. Of course, linguists remain primarily interested in names as manifestations of historical languages and should investigate them with regard to other aspects of the languages in question. Moreover, they try to draw conclusions on the linguistic behaviour of those who bear or use the names. For obvious reasons, historians follow the opposite perspective: they focus on the persons and institutions behind the names that is the latter are interesting whenever they provide further information on the former (cf. Geuenich / Hawicks 2006: 82). Some gained researchers (including current project members) working on the coins and their legends can draw on their experience within the context of a very large older project, Nomen et gens. It resulted in a large database of several thousands of name forms, collected in well-known early medieval sources.5 Among other things, this resource should help to answer the question whether or to what extent names of this period allow us to conclude to which ethnic groups (dealing mainly with Germanic groups) their bearers belonged and thus whether typical naming practices for certain ethnic groups existed (Patzold 2008: 139).
5 The anthroponomastic uses of the Merovingian period The Merovingian rule brought about the rearrangement of the ‘anthroponomastic landscape’ of Gaul, which consisted mainly of Latin and Greek names (formed in the classical period, but very often in advanced imperial
5
Cf. the list of already searched texts, mainly of hagiographic type, in Patzold (2006: 148–151). For the problems in finding a system for arranging the onomastic data which meets the standards of both historians and linguists see Geuenich / Hawicks 2008.
298
Rembert Eufe
times, with or without Christian implications)6 but also comprised some Semitic and Celtic ones (be they loans or calques and cover names, often hard to determine with certainty).7 These names were to a very large extent supplanted by the Germanic names of the Franks. As a survey of episcopal names appearing in council records from the provinces in Gaul8 suggests, Germanic names are in the majority from the seventh century onwards.9 It has to be kept in mind that bishops were initially drawn from the ancient Senatorial class, the traditional Roman upper class with privileged access to Latin education. The rate of Latin names is even lower among the names of the persons close to the king mentioned in the Merovingian charters (15 per cent, for both religious and civil dignitaries; Bergmann 1997: 105). Unsurprisingly, the Germanic inf luence was stronger in the North, as is shown by two important sources from the early ninth century, the polyptych of St-Germain des Prés (Paris) with more than 90 per cent Germanic names (Hägermann 1997: 111–112), contrasting with 38.5 per cent of Germanic names in the polyptych of St Victor (Marseille, 814; Bergh 1941: 184). The dif fusion of Germanic names is remarkable because it contrasts with the linguistic Latinization / Romancification of the Germanophone groups in most parts of the Frankish kingdom. In general, their predominance in the Galloromance area lasted until the twelfth century, when saints’ names of Latin and Greek origin started to prevail (such as Jean and Pierre, which became the most frequent male names in the late Middle Ages). 6 7
8 9
For general information on the Latin and Greek name systems see Kajanto (1963, 1965, 1973), Solin (1995, 2003a, 2003b), Solin & Salomies (1994). Coşkun / Zeidler (2003: 10) suspect the name of the fourth-century Bordelaise Ausonius (attested also as a monetary name in the Berlin collection) to be a Gaulish name (perhaps derived from Auso or Ausios) disguised in Latin. In fact, a considerable number of the Latin and Greek names on the coins of the Berlin collection appear concentrated in Gaul, indicating a distinct regional anthroponomastic profile, ascribable at least partly to Gaulish inf luence (cf. Eufe (in preparation)). Note that we refer here just to the provinces of Gaul, comprising the former Roman Germania inferior and Germania superior up to the river Rhine. In fact, no Merovingian coins are known to have been minted east of it. Kajanto (1973, 390–392) counts 21.8 per cent of Germanic names in the period from 511–600, which increased to 59.7 per cent between 600–695. For further statistics concerning the spread of Germanic names in Gaul cf. Bergh (1941: 185).
Merovingian Coins and Their Inscriptions
299
The names of Germanic origin are of three dif ferent types: – bithematic names (as in other Indo-European languages, such as Ancient Greek, but not in Latin), composed of two elements, e.g. Adalbertus, derived from athala ‘noble’ + berhta ‘brilliant’. They represent dif ferent types of compounds, mainly of a determinative morpho-semantic structure (with a first element modifying the meaning of the second, e.g. Theoda-rich / Diet-rich ‘people’ + ‘powerful, rich’ > ‘powerful (second element) among the people’) but also copulative (e.g. Bern-ulf ‘bear’+‘wolf ’) and Bahuvrīhi-composita (e.g. Wolfhelm ‘the one with the helmet decorated with the wolf ’). Note that by this time names without any semantic relation between their two components were created (Greule 1996: 1184), especially by combining elements of the names of the bearer’s parents. – monothematic names, often to be regarded as short forms of bithematic names, e.g. Dodo and Dedo, which our scholars of German have traced back to bithematic names with the first element theud-a ‘people’. These names are supposed to have represented more ‘colloquial’ forms (Greule 1996: 1185).10 Interestingly, in some cases monneyers’ names seem to appear on some coins in a bithematic form, on others in a monothematic one, e. g. Alligisels and Alloni on coins from Angers (Breillat 1935: 24; although we could not find any examples on the Bode-Museum coins). – finally, derivations by means of various suf fixes of Germanic, Latin or mixed origin can occur. In this respect, the Germanic names of our collection do not show great diversity, only some derivations by means of -īna-, -līna- and -linus / -lenus (a typical Merovingian suf fix of mixed origin, see below).
10
Sometimes, the monothematic short names are attested in the sources, as in the case of ‘Berthranno sive Bettone’ (abl. / oblique case in -one), mentioned in the will of a Merovingian bishop (Weidemann 1986: 28).
300
Rembert Eufe
6 The names of the monneyers According to Felder (2003: 26), who examined the names on the 5,000 Merovingian coins conserved in the Cabinet des Médailles at the National Library in Paris, some 75 per cent of the monneyers’ names are of Germanic origin. Our research on the coins of the Berlin collection shows a similar tendency: leaving aside the pseudoimperial coins of the first period, displaying mainly names of kings and Roman emperors and only two monneyers named Maurentius11 (Lyon) and Dedo (Alsheim), we find 109 names of monneyers from the second period12 from 585–675, among them 68 ≈ 62.39 per cent13 of clear Germanic origin, like Adelbertus, Aegulfus, Bertechramnus, Charegisilus, Filaharius, Launomundus, Madelinus and Anglo, Babo, Dedo, Suno and others. To these could be added six other names (≈ 5.5 per cent) that show Germanic bases with an attached diminutive suf fix -lenus / -linus quite common in Merovingian personal names (Abolinus, Agilinus, Audolenus, Gaudelinus, Leobolenus and Theudelenus). Its frequent occurrence whith , abnormal for long -ī- which is in general conserved in Romance (in contrast to short -ĭ-), does not support the explanation as an exclusive outcome of the Germanic diminutive suf fix *-līna-z. It rather suggests a fusion of the latter with the popular element *wini-z ‘friend, beloved, spouse’ (with short -ĭ-) and extraction of the suf fix from names 11
12 13
Due to space constraints, we restrict ourselves to some recapitulary remarks based on the more detailed expositions in Eufe (in preparation) and Greule, Jarnut, Kluge & Selig (in preparation, with comments on the structure and origin of the single names), where the legends are also reproduced in forms more close to what is visible on the coins and more precise datations are given. For simplicity’s sake, we mention the names here only in an adjusted form, which is based to a certain extent on our interpretation of the legends that are sometimes actually quite dif ficult to read and reproduce exactly (on this aspect cf. Eufe in preparation). See Greule, Jarnut, Kluge & Selig for more precise datations of the single coins and the names on them. Percentages have been calculated without regard to unidentifiable names like AOMIO MON (ca. 585–675) or INTE MONETA (ca. 620–640) from the total account.
Merovingian Coins and Their Inscriptions
301
like the aforementioned Madelinus -v-, resulting in forms like CAV[Λ]-LOΠ Γ40 and also + CΛVELONO – VICTVI (three cases against four with , next to three others where the relevant part of the name is abbreviated or illegible).41 Moreover we do have a large number of case endings in comparable syntactic positions. It is important to recall that Old French (as Old Occitan) displays a two-case system.42 The endings of the personal names on the coins testify to the genesis of this system, representing for the second declension (o-declension) a previous stage with a 3-case system. Names like Maurinus or many Germanic names, almost always of a bithematic structure like Sigimundus,43 appear often with -us in the function of a Casus rectus (28 occurrences in the second period).44 In addition, there are some attestations 40 The legend is very and almost illegible, but the can still be clearly discerned. 41 For an overview of phonetical and phonological changes attested on Merovingian coins, cf. Chambon / Greub 2000, who even find evidence for the emerging dia topic subdivision of Gallo-Romance (Chambon / Greub 2000: 167, 171, 174–175). Following Selig / Eufe (in preparation), the sparse occurrences of the features in question should be rediscussed in a broader picture of the formation of GalloRomance. 42 This is actually astonishing because in many other respects, French appears as one of the Romance languages which has developed the most distinctly from Latin – like Romanian, which until the present day has a two-case system (although dif ferent in structure). 43 An exception is monothematic Badus which is regarded as a ref lex of Germ. *baduor *badwa- ‘fight’ (cf. the comment in Greule, Jarnut, Kluge & Selig). 44 In the following, we only indicate the figures that we could ascertain for the second period. For the third period, the data from the Berlin collection are not suf ficient.
308
Rembert Eufe
of the genitive case in place names, like SCI PETRI. But for the majority, these names appear with -o which might represent a Casus obliquus in expansion of use (36 occ.). Unfortunately, Merovingian Latin shows oscillations and blendings of the Latin second and third declensions (use of the types Gallus, gen. Galli and Gallo, gen. Gallonis, with occasional fusion in Gallus, gen. Gallonis and similar), which means that in cases like PRIMANO MONETΛ V and AEGVLFO MCN it cannot be entirely discarded that they are actually intended as nominative forms of Primano, gen. Primanonis (comparable to Petro, gen. Petronis) or Aegulfo, gen. Aegulfonis, respectively.45 The third declension is common with monothematic Germanic names like Betto or Anglo, but also with bithematic names like Chagnomiris, which appear with -o resp. -is / -es as a Casus rectus (9 occ.) and -one / -oni resp. -e / -i (Casus obliquus, 13 occ.). A genitive occurs in the place name IOHANNIS PORTO. And why should historians care about this project? Of course the personal names are indicative of the contact between Gallo-Latin and Germanic-speaking groups. Furthermore, the appearance of the place names on the coins gives precious proofs to historians for the existence of these places during the Merovingian period. In consequence, everyone who writes the medieval history of a certain town or region in the former provinces of Gaul should control if the place(s) he deals with are mentioned on the coins. It must be stressed that for the localization of the place names, a lot of work has still to be done: In most of the older numismatic publications no toponomastic literature has been consulted, the places on the coins are identified on the basis of rather vague phonetic similarities with toponyms on present-day French maps, at most mentioning the existence of remnants from Gallo-Roman antiquity – which actually are very frequent all over former Gaul. Only recently, in Chambon 2001, a first ef fort has been made to examine the localizations current in the numismatic literature on serious toponomastic and philological grounds, according to which only 3/4, 3/5 and 1/5 of the localizations in Limousin, Arvernie and Rouergue, 45 On this question cf. Pitz (2009), who regards these oscillations rather as a continuation of Latin name morphology than as a result of substratum inf luence exerced by Germanic noun inf lection.
Merovingian Coins and Their Inscriptions
309
respectively, can be maintained (Chambon 2001: 348–349). Actually, the identification of the places which the place names in the sources refer to is a very crucial problem for any historical research, relying not only on coins, but also on other sources like manuscripts.46 It is hardly ever accorded the attention it deserves when preparing editions and preparing indices for them, which require a good deal of philological, linguistic and historical background knowledge and meticulous verification of sources, editions and reference guides.47 But how can research on the perception and the division of space or the concept of border in the Middle Ages, for example, arrive at any valid conclusions if the places they rely on were most likely located elsewhere than is assumed?48
9 Continuity of place names, continuity of places? But there is much more for historians about Merovingian coins. They can be discussed in the context of the old question of whether the end of the Roman Empire represents a sudden decline, the ‘crash of civilization’ caused by the barbarian invasions as it has often been regarded as of humanism onwards.49 Or should we better think of it as a ‘surface event’ with scarce impact on a continuous f low of transformations? 46 The importance of toponomastic research has already been stressed by Bloch 1956, 8–14. 47 Cf. the general remarks in Chambon 1997 and 2009. 48 A very striking example is Pitz 2007 on the places names of a mansus indominicatus in Lorraine, in the urbarium of the abbey of Prüm (Rhineland-Palatinate, Germany) as its possession. The author re-examines the identifications of the place names, correcting many of the hitherto proposed identifications and observing that the corresponding places are not dispersed, but situated in fact close to each other, forming a compact and clearly confined manor. 49 On the prerequisites and implications of this question, cf. Fehr (2010: 42–69). Note that Ward-Perkins in his ‘schematic reasoning’ (Fehr 2010: 62) relies on the disappearance of Roman bronze small coins (Ward-Perkins 2005: 110–117).
310
Rembert Eufe
As shown above, the personal names on the coins do testify to the anthroponomastic reorientation towards names of Germanic origin after the Frankish takeover, while this does not hold for the place names. To a great extent, these are of Gaulish and Latin origin, showing an impressive toponomastic continuity which deserves our attention as such, in accordance with the fact that place names represent an important form of collective memory and tradition. Interestingly, place names developed dif ferently in England: The Anglo-Saxon dominance brought about a ‘vast renaming’ (Gelling 1995: 787) of places so that the place names on the Merovingian coins confirm by and large the view of Wickham (2009: 200): The real contrast inside the ex-Roman provinces was not between societies that had been invaded or conquered and the others, but between the Continent and Britain: in the former, the basic Roman political and social structures survived (though they were in most places ramshackle and underfunded).
Although the etymological distribution of the place names appears indicative of a settlement history of late-antiquity Gaul, further investigations are of course needed in order to make clear that the preservation of antique place names corresponded to a continuity of institutions and social life, not just continuity of ruins. In this respect, coins are significant by their very existence, testifying to a certain economic or at least financial activity. It has already been mentioned that a large number of 37 (against 45 other) place names of the Berlin coins refers to main places of former Roman civitates – note that the ‘cellular civitas-structure inherited from the Roman period’ was maintained in ‘perhaps four-fifths of Roman Gaul’, in any case ‘[…] no region of Frankish Gaul exhibits the total breakdown of the civitas-framework which occurred just across the Channel’ (Loseby 2006: 93). But do the old regional capitals turn out to be actual centres of Mero vingian coinage, and if so, did it include all of them? Judging only by the number of monneyers known to have worked there, the following places stand out with ten to twenty-five dif ferent names: Angers, Clermont-
Merovingian Coins and Their Inscriptions
311
Ferrand, Orléans, Bourges, Bordeaux, Limoges, Maastricht, Marsal, Metz,50 Paris, Poitiers, Rennes, Rouen, Rodez, Toul, Toulouse, Trizay, Vienne, Quentovic, exceeded by Chalon-sur-Saône with even thirty names (Pol 2002: 353).51 In addition, other important mints were at work in Reims, Verdun, Dinant, Namur, Treves, Cologne, Autun, Mâcon, Troyes, Lyon, Banassac, Brioude, and, albeit to a lesser extent, also Tournai (Dahmen in preparation). Now, most of these places are indeed well known through archeological and historical research as important administrative, religious and economic capitals of Roman civitates.52 But those who were not definitely deserve a comment: Dinant, Namur, Dorestad, Maastricht and Quentovic, together with Huy, ef fectively emerged as new centres in the heartland of the Franks and north of it. They were situated in or near ancient Roman places, but these had at best the form of small vici or castra. Furthermore, continuity of habitation was ‘fairly rare in the area between the Meuse and the North Sea’ (Verhulst 1999: 5), traceable almost exclusively in Maastricht, Tournai (a former civitas, however), Boulogne and Huy, of which only the latter gives prove of functional continuity in the form of artisanal activities (Verhulst 1999: 21). Especially the medieval emporia of Dorestad and Quentovic indicate the importance of harbours as places of Merovingian coinage, also af firmed by Marseille with its very large production of coins. Furthermore, we have to mention Mâcon, originally a castrum with a port on the Saône. However, in this case Christianity must be named as a decisive factor of transformation: The Merovingian king Childebert I is said to have installed the cult of Saint 50 Halsall (1996: 61) perceives for Metz a civilizational hiatus in the 5th–6th centuries, calling it a ‘town without urban life’ – in any case, if ever traces of urban life subsisted, it seem to have been in civitas main places like this rather than anywhere else. 51 Moreover, coins from Chalon-sur-Saône and Lyon are preserved with two monneyers’ names on one coin. Of course we can judge only from the coins that are known to us, ignoring the original output of coinage. Future large coin finds may of course change the picture. 52 In Carolingian times, then, Pippin again limited coinage to the main places of the civitates, some new centres in the north (Quentovic, Dorestad, Dinant, Namur, Huy and Maastricht) and some large suburban abbeys (Saint-Martin de Tours, Saint-Pierre de Sens, Fixot 1980: 527).
312
Rembert Eufe
Vincent of Saragossa in the cathedral of Mâcon in 543 ( Jeanton 1934: 15). Similarly, Brioude, maybe an old Roman vicus, owes its importance to the veneration of Saint Julian, already beginning in imperial times and later stimulated by Visigothic and Frankish rulers. Of course, there would be a lot to say on the imprint of Christianity and the Catholic Church on the landscape of Gaul;53 as it has certainly been much more important than that left by the Germanic rulers. However, the latter exerted an indirect inf luence inasmuch as they relied on the first, giving their own impact by granting the construction of churches and monasteries or even founding them themselves and promoting the veneration of certain saints, like Remigius of Reims, occurring in S-ANTI REMI VICO. In fact, the coins bear testimony of these general tendencies, with pieces showing legends like ECL + // ID / EPS (maybe from Senlis (Département Oise)), + SCI PETRI // ME[DI]OLΛNO M[O]N (assigned to Saintes (Département Charente-Maritime)), originally Mediolanum Santonum and including since the sixth century a cathedral dedicated to St Peter in its walls) or + SCO FILBER // + GEMEDICO CAV or COL (from Jumièges (Département Seine-Maritime), where Saint Philibert had founded an important monastery in the seventh century).54 The occurrence of ecclesiastical institutions as places of coinage is certainly astonishing and thus calls for further ref lections on the function of the monetary system.
10 Places names and monetary system The large number of place names on the coins leads us to assume that the Frankish kings did not enjoy a monopoly over coinage (Grierson / Blackburn 1986: 97–98), this usually being a very important instrument of power. Among other things, the number of monasteries dif fered strongly between the civitas capitals! In the Viennensis province, e.g. in ten out of fourteen capitals no monasteries have been detected so far, whereas Marseille is likely to have had eight monasteries, Arles five, Vienne assuredly eight, but maybe even thirteen (Biarne 2002: 125). 54 Cf. the comments on these legends in Greule, Jarnut, Kluge & Selig (in preparation). 53
Merovingian Coins and Their Inscriptions
313
Now it is precisely the designations of ecclesiastical institutions, together with other indications like racio fisci, racio domini, scola Regis or in palacio, which raise doubts whether we really deal here with a coin system in the usual sense. They rather point to a special and extremely particularistic tax system, with the monneyers as of ficers responsible for the collection and registration of taxes in public places and their immediate transformation into coins whose purity they had to attest.55 This would not only fit in with the high value of the Merovingian gold tremisses, worth approx. £ 50 (whereas a Carolingian denarius corresponds to only £12), so that they could have been only ‘somewhat clumsy aids to exchange’, better serving as ‘convenient ways of hoarding wealth’ (Wickham 2009: 227). But it can also explain why many sea and river ports appear among the most important Merovingian mints, places where extensive customs revenues must have been charged56 – like in Marsal and some other nearby places in the south of Metz whose importance was related to the extensive production and commerce of salt, also known for antiquity (Stahl 1982, Pol 2001). On these grounds, a study of thirty-four surely identified places mentioned on Merovingian coins from around Limoges has shown that four of them were castra and fourteen are attested as Gallo-Roman or Merovingian vici. In addition, they were all equipped with a parish church and functioned (or still do) as centres of parishes, dedicated to saints popular in Roman and Merovingian times, which further suggests that they were ‘sites of 55
56
This hypothesis is often corroborated by a passage from the Vita of St Eloy (edited in Krusch 1902: 680–681, cited e.g. by Boyer 2007: 152–153), which reports how a monetarius is about to melt and purify the tax revenues collected on a royal manor, not knowing that it has just been conferred to the saint, who makes him stop his work. Cf. the remarks by Strothmann and Dahmen in Greule, Jarnut, Kluge & Selig (in preparation) – note that this episode documents also the exemption of clergymen and landowners from tax allowances, leaving the revenues at their own disposal. In this light, the introduction of silver coins in the second half of the seventh century, probably on the initiative of the mayor Ebroin, appears as a tax reform. Accordingly, Verhulst (1999: 32) supposes that the important numbers of monneyers in Maastricht, Huy, Namur and Dinant ‘could have had a political rather than an economic significance in relation to the position of power of the Austrasian nobility and in particular of the Pippinids in the Meuse Valley and their expansion in a northerly direction along this river to Dorestat […]’.
314
Rembert Eufe
evident public nature’ (Boyer 2007: 150). Of the remaining sixteen place names, one refers to a monastery, all the others are toponyms derived with -acum, indicating old Gallo-Roman rural manors likely to have developed similarly into settlements of public interest. In general, these observations are in line with what we have found out for the place names of the Berlin collection.
11 New places? However, it is worth noting that on the coins there is evidence for Mero vingian institutions which have left some, albeit small, traces on the soil of Gaul: One of the coins from the Bode-Museum is among those minted in palacio ‘in the royal palace’. At least in one case, Palaiseau (Département Essonne), a settlement is known to have grown around a Merovingian palace.57 Moreover, the Paris collection comprises pieces coined in MALLO ARLAVIS, MALLO MATIRIACO and MALLO CAMPIONE (Felder 2003: 472, 466, 473), all localized in Belgica prima and referring to the thingstead where the Germanic government assemblies met. While the first of these places still seems to await localization, the other two are identified as Mairy (Meurthe-et-Moselle, Buchmüller-Pfaf f 1990: 315–316, Haubrichs 1985: 30) and Champ, one of the two hamlets forming present-day Champneuville (Meuse, Haubrichs 1985: 12, 28). Note that even in these cases, the Latin part of the place name is continued, although place names which are traced back to the Frankish designation for the thingstead do exist, such as Les Maus (near Arras, Pas-de-Calais), Le Malet (near Laon, Aisne) and others (Schmidt-Wiegand 1973: 493).
57
This does not call into question that in many towns, on the contrary, Merovingian palaces were established in older Roman public buildings. Pfalzel (Treves) grew out of a fortified Roman villa of the 4th century.
Merovingian Coins and Their Inscriptions
315
Acting on the assumption of the Merovingian coinage as a form of taxation, two places with ‑(i)acum-names yet deserve a comment, because they appear among the places disposing of important mints listed above: At first sight, the role of Banassac (department Lozère) as a centre of coinage from which a great number of coins are preserved comes as a surprise. It is sometimes associated with the existence of a f lourishing pottery industry in Roman times in the third century. In any case, a considerable portion of royal coins in the name of Charibert II, king of Aquitaine (629–632), indicates that it owed its importance to one of the numerous subdivisions of the Merovingian among its various kings. Little is known of Trizay-sur-le-Lay (department Vendée), regarded as the present-day of fspring of TIDIRICIACO. As already said, it is derived from a Germanic name, like only few other places of Merovingian coinage named after the -(i)acum-type. Their exiguous number contradicts earlier views of the monneyers acting on behalf of rich landlords,58 as the latter should have borne predominantly Germanic anthroponyms from which their manors’ names would have been derived.59 TIDIRICIACO as a form of Teuderic-iacum is derived from Theuderich, a name also born by Merovingian kings – as well as TEODEBERCIACO, identified as Thiverzay in the same département and derived from Theudebert. These two names might indicate royal Merovingian manors (especially the first with a large number of monneyers who seem to have been occupied with collecting taxes) in a region where actually few Gallo-Roman settlements are to be found (Gauthier 1999: 146). This would fit partly with the division of Gallo-Romance: On the Atlantic coast, the present-day border between the French and the Occitan area recedes Southwards, across the Loire almost down to the Gironde. But further research would be needed that goes into detail on the settlement history of Pays de la Loire and Poitou-Charentes. 58 59
On these approaches cf. Grierson-Blackburn 1986: 99–100. Buchmüller-Pfaf f (1990: 10) comes across a ‘considerable number’ of -(i)acumtoponyms derived from Germanic anthroponyms in Lorraine which she regards to their majority as ‘(early) Merovingian formations’. Of course the Germanic base of such a toponym does not necessarily indicate a Germanic settlement (BuchmüllerPfaf f 1990: 754). But inversely, Frankish landlords can be expected to bear mostly names of Germanic origin.
316
Rembert Eufe
12 Conclusion As we have seen, Merovingian coins present considerable challenges that are best met in a joint ef fort. The work of numismatists will always be needed with their expertise in assessing datings, comparing iconography, analysing coin finds etc. Such information on preserved Merovingian coins in museums and private collections should be united in data bases which facilitate access for historians and philologists to this type of source. Historians should find out about the Merovingian tax system and the institutions linked with it, not only for getting more precise judgements on the coins, but also for gaining further insights into the economic and political developments between the deposition of the last West-Roman emperor and the Carolingians’ coming into power. Moreover, it turns out fruitful to collaborate with archaeologists who can provide data of the settlement history, including seemingly ‘small’ places like those mentioned on many Merovingian coins but not in other sources. And linguists should carry on with research on place names and re-examine existing identifications with attention to historical phonetics and phonology. In summary, the need for close collaboration between historians and linguists appears evident in our case; their cooperation is particularly recommendable in order to value the coins as a source and open up new perspectives on a period for which only very few other sources of information are extant.
References Belfort, Auguste de. 1892–1895. Description générale des monnaies Mérovingiennes par ordre alphabétique des ateliers. 5 vols. Paris: Société Française de numismatique. Bergh, Åke. 1941. Études d’anthroponymie provençale. Les noms du Polyptyque de Wadalde (a. 814). Göteborg, Elanders Boktryckeri Aktiebolag. Bergmann, Werner. 1997. ‘Personennamen und Gruppenzugehörigkeit nach dem Zeugnis der merowingischen Königsurkunden.’ In: Geuenich et al. (eds), Nomen et gens. Zur historischen Aussagekraft frühmittelalterlicher Personennamen. Berlin: De Gruyter, 94–105.
Merovingian Coins and Their Inscriptions
317
Bloch, Marc (1956): Les caractères originaux de l’histoire rurale française, vol. 2: Supplément établi par R. Dauvergne d’après les travaux de l’auteur. Paris: Colin. Biarne, Jacques. 2002. ‘L’inf luence des monastères sur le paysage urbain en Occident (IVe–VIIe siècles).’ In: Beaujard, Brigitte (ed.), La naissance de la ville chrétienne. Mélanges en hommage à Nancy Gauthier. Tours: Maison des Sciences de l’Homme Villes et Territoires, 123–134. Boyer, Jean François. 2007. ‘À propos des triens mérovingiens: Approche du Système de collecte et de traitement de la recette fiscale en Limousin aux VIe–VIIe siècles.’ Annales du Midi 119, 141–157. Breillat, Pierre. 1935. ‘La langue des inscriptions monétaires mérovingiennes.’ In: Positions des thèses soutenues par les élèves de la promotion de 1935. Paris: Presses Univ. de France, 15–24. Buchmüller-Pfaf f, Monika. 1990. Siedlungsnamen zwischen Spätantike und frühem Mittelalter. Die -(i)acum-Namen der römischen Provinz Belgica Prima. Tübingen: Niemeyer. Buchner, Rudolf (ed.). 1964. Zehn Bücher Geschichten. Darmstadt: Wiss. Buchges. Chambon, Jean-Pierre. 1997. ‘Sur une technique de la linguistique historique: l’identification des noms de lieux, en particulier dans les texts du passé (avec des exemples concernant l’Auvergne et ses marges)’. Lalies. Actes des sessions de linguistique et de literature 17, 55–100. Chambon, Jean-Pierre. 2001. ‘L’identification des noms d’ateliers monétaires méro vingiens (Arvernie et entours): point de vue de linguiste.’ Revue numismatique 157, 47–405. Chambon, Jean-Pierre. 2009. ‘À quelles conditions une onomastique occitane estelle possible?’ In: Latry, Guy (ed.), La voix occitane. Actes du VIIIe congrès de l’Association Internationale d’Études Occitanes. Pessac: Presses Univ. de Bordeaux, 1019–1036. Chambon, Jean-Pierre, & Yves Greub. 2000. ‘Données nouvelles pour la linguistique galloromane: les légendes monétaires mérovingiennes.’ Bulletin de la Société de Linguistique de Paris 95, 147–182. Coşkun, Altay, & Jürgen Zeidler. 2003. ‘“Cover Names” and Nomenclature in Late Roman Gaul. The Evidence of the Bordelaise Poet Ausonius.’ In: Keats-Rohan, Katharine (ed.), Unit for Prosopographical Research (UPR). Oxford: Linacre College. . Dahmen, Karsten. 2011. ‘Zum Münzwesen des Merowingerreiches.’ In: Greule et al. Manuscript in preparation. Deroy, Louis, & Mulon, Marianne. 1993. Dictionnaire de noms de lieux. Paris: Le Robert. Eufe, Rembert. ‘Die Personennamen auf den merowingischen Monetarmünzen als Spiegel der romanisch-germanischen Sprachsynthese im Frankenreich.’ In: Hau
318
Rembert Eufe
brichs, Wolfgang et al. (eds), Kulturelle Integration und Personennamen im frühen und hohen Mittelalter. Manuscript in preparation. Fehr, Hubert. 2010. Germanen und Romanen im Merowingerreich. Frühgeschichtliche Archäologie zwischen Wissenschaft und Zeitgeschehen. Berlin: De Gruyter. Felder, Egon. 2003. Die Personennamen auf den merowingischen Münzen der Bibliothèque Nationale de France. München: Verlag der Bayerischen Akad. der Wiss. Fixot, Michel. 1980. ‘Une image idéale, une réalité dif ficile: les villes du VIIe au IXe siècle.’ In: Duby, Michel (ed.), Histoire de la France urbaine, vol. 1: La ville antique des origines au IXe siècle, 497–563. Gamillscheg, Ernst. 1970. Romania germanica. Sprach- und Siedlungsgeschichte der Germanen auf dem Boden des alten Römerreichs. vol. 1: Zu den ältesten Berührungen zwischen Römern und Germanen. Die Franken. 2nd edn. Berlin: De Gruyter. Gauthier, Pierre. 1999. ‘Les antroponymes dans les toponymes de la Vienne.’ In: Kremer, Dieter (ed.), Onomastik. Akten des 18. Internationalen Kongresses für Namenforschung, vol. IV: Personennamen und Ortsnamen. Tübingen: Niemeyer, 142–153. Gelling, Margareth. 1995. ‘Place-Names in England’. In: Eichler, Ernst (ed.), Namenforschung. Ein internationales Handbuch zur Onomastik. Vol. 3. Berlin, New York: De Gruyter, 786–792. Gendron, Stephane. 2008. L’origine des noms de lieux en France. Essai de toponymie. 2nd edn. Paris: Errance. Geuenich, Dieter, & Hawicks, Heike. 2006. ‘Probleme der Lemmatisierung frühmittel alterlicher Personennamen im interdisziplinären Projekt “Nomen et gens”.’ In: Brylla, Eva & Mats Wahlberg (eds), Proceedings of the 21st International Congress of Onomastic Sciences. Vol. 3. Uppsala: Institutet för sprak och folkminnen, 81–90. Grierson, Philip, & Mark Blackburn. 1986. Medieval European coinage, vol. 1: The early Middle Ages (5th-10th centuries). Cambridge: Cambridge University Press. Greule, Albrecht. 1996. ‘Morphologie und Wortbildung der Vornamen: Germanisch.’ In: Eichler, Ernst (ed.): Namenforschung. Ein internationales Handbuch zur Onomastik, Vol. 2, Berlin: De Gruyter, 1182–1187. Greule, Albrecht, Jörg Jarnut, Bernd Kluge & Maria Selig (eds). 2011. Die mero wingischen Monetarmünzen. Historische, numismatische und philologische Unt ersuchungen. Grundlage des Bestandes im Münzkabinett der Staatlichen Museen zu Berlin. Manuscript in preparation. Hägermann, Dieter. 1997. ‘Die Namengebung in den Unterschichten der Karolinger zeit.’ In: Geuenich, Dieter et al. (eds.): Nomen et gens. Zur historischen Aussagekraft frühmittelalterlicher Personennamen, Berlin: de Gruyter, 106–115.
Merovingian Coins and Their Inscriptions
319
Hägermann, Dieter, Wolfgang Haubrichs & Jörg Jarnut (eds). 2004. Akkulturation. Probleme einer germanisch-romanischen Kultursynthese in Spätantike und frühem Mittelalter. Berlin: De Gruyter. Halsall, Guy. 1996. ‘Towns, societies and ideas: The not-so-strange case of late Roman and early Merovingian Metz.’ In: Neill Christie & Simon T. Loseby (eds), Towns in Transition. Urban Evolution in Late Antiquity and the Early Middle Ages. Aldershot: Scolar, 235–261. Haubrichs, Wolfgang. 1985. ‘Die Urkunde Pippins des Mittleren und Plectruds für St-Vanne in Verdun (702). Toponomastische und besitzgeschichtliche Über legungen zum frühen Besitz der Pippiniden-Arnulfinger und zum Königsgut im Verdunois (avec résumé français).’ Francia 13, 1–46. Haubrichs, Wolfgang. 1997. ‘Galloromanische Kontinuität zwischen unterer Saar und Mosel. Problematik und Chancen einer Auswertung der toponymischen Zeugnisse aus germanistischer Perspektive.’ In: Holtus, Günter et al. (eds), Italica et Romanica. Festschrift für Max Pfister zum 65. Geburtstag. Vol. 3. Tübingen: Niemeyer, 211–237. Haubrichs, Wolfgang. 2004. ‘Romano-germanische Hybridnamen des frühen Mittel alters nördlich der Alpen.’ In: Hägermann, Dieter, et al. (eds), Akkulturation. Probleme einer germanisch-romanischen Kultursynthese in Spätantike und frühem Mittelalter. Berlin: De Gruyter, 179–204. Haubrichs, Wolfgang. 2008. ‘Hybridität und Integration. Vom Siegeszug und Untergang des germanischen Personennamensystems in der Romania.’ In: Dahmen, Wolfgang et al. (eds), Zur Bedeutung der Namenkunde für die Romanistik. Roma nistisches Kolloquium XXII. Tübingen: Narr, 87–140. IKMK. Interaktiver Katalog des Münzkabinetts der Staatlichen Museen zu Berlin. accessed 18 August 2011. Jeanton, Gabriel. 1934. Le vieux Mâcon. Histoire des rues et des quartiers de la ville. Mâcon: Renaudier. Kajanto, Iiro. 1963. Onomastic Studies in the Early Christian Inscriptions of Rome and Carthage. Helsinki: Institutum Romanum Finlandiae. Kajanto, Iiro. 1965. The Latin Cognomina. Helsinki: Soc. Scientiarum Fennica. Kajanto, Iiro. 1973. ‘The Disappearance of Classical Nomenclature in the Merovingian Period.’ In: Due, Otto Steen et al. (eds), Classica et mediaevalia Francisco Blatt septuagenario dedicata. Kopenhagen: Gyldendal, 383–395. Kluge, Bernd. 2007. Numismatik des Mittelalters, vol. 1: Handbuch und Thesaurus Nummorum Medii Aevi. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Kölzer, Theo (ed.). 2001. Die Urkunden der Merowinger. Diplomata regum Francorum e stirpe Merovingica. Hannover: Hahn.
320
Rembert Eufe
Krusch, Bruno (ed). 1902. ‘Vita Eligii Episcopi Noviomagensis.’ In: Krusch, Bruno (ed.), Monumenta Germaniae historica. Scriptores rerum merovingicarum, vol. IV: Passiones Vitaeque Sanctorum aevi Merovingici. Hannover: Hahn, 634–761. Loseby, Simon T. 2006. ‘Decline and Change in the Cities of Late Antique Gaul.’ In: Krause, Jens-Uwe & Christian Witschel (eds), Die Stadt in der Spätantike – Niedergang oder Wandel? Stuttgart: Steiner, 67–104. Patzold, Stef fen. 2008. ‘Probleme der ethnischen Identifikation. Der Beitrag der Geschichtswissenschaft zu einem interdisziplinären Personennamenlexikon am Beispiel der Namen des Frankenreichs (Hagiographie/Historiographie)’. In: Brylla, Eva & Mats Wahlberg (eds), Proceedings of the 21st International Congress of Onomastic Sciences. Uppsala 19–24 August 2002, Vol. 4, Uppsala: Språk- och Folkminnesinstitutet, 135–151. Pitz, Martina. 2007. ‘La représentation de l’espace durant le haut Moyen-Âge: éléments pour une analyse onomastique du Polyptyque de Prüm.’ In: Herbin, Jean-Charles & Michel Tamine (eds), Espace représenté, espace dénommé. Géographie, cartographie, toponymie. Valenciennes: Presses Univ. de Valenciennes, 289–304. Pitz, Martina. 2009. ‘Le type Charles-Charlon ou Eve-Evain en ancient français – vraiment un ref let du superstrat germanique?’ Nouvelle Revue d’Onomastique 51, 175–200. Pohl, Walter. 2004. ‘Vom Nutzen des Germanenbegrif fes zwischen Antike und Mittel alter: eine forschungsgeschichtliche Perspektive’. In: Hägermann et al., 18–34. Pol, Arend. 2001. ‘Marsal.’ In: Hoops, Johannes (ed.): Reallexikon der Germanischen Altertumskunde. Vol. 19. 2nd edn. Berlin: De Gruyter, 345–346. Pol, Arend. 2002. ‘Münzmeister. Merowingerzeit’. In: Hoops, Johannes (ed.), Real lexikon der Germanischen Altertumskunde. Vol. 20. 2nd edn. Berlin: De Gruyter, 353–355. Prou, Maurice. 1910. ‘Notes sur le latin des monnaies mérovingiennes.’ In: Mélanges de philologie romane et d’histoire littéraire of ferts à Maurice Wilmotte. Paris: Champion, 523–540. Prou, Maurice. 1995 [1892]. Les monnaies mérovingiennes. Catalogue des monnaies françaises de la Bibliothèque Nationale. Nîmes: Lacour. Schmidt-Wiegand, Ruth. 1973. ‘Sali. Die Malbergischen Glossen der Lex Salica und die Ausbreitung der Franken.’ In: Petri, Franz (ed.), Siedlung, Sprache und Bevölkerungsstruktur im Frankenreich, 490–530. Selig, Maria, & Rembert Eufe. In preparation ‘Les monnaies mérovingiennes et leur analyse linguistique.’ In: Biville, Frédérique (ed.), Actes du VIIème Colloque International sur le Latin Vulgaire. Solin, Heikki 1995. ‘Namen im Alten Rom.’ In: Eichler, Ernst (ed.), Namenforschung. Ein internationales Handbuch zur Onomastik. Vol. 2. Berlin: De Gruyter, 1041–1048.
Merovingian Coins and Their Inscriptions
321
Solin, Heikki. 2003a. Die griechischen Personennamen in Rom. Ein Namenbuch. 2nd rev. edn. 3 vols. Berlin: De Gruyter. Solin, Heikki. 2003b. ‘����������������������������������������������������������� Le trasformazioni dei nomi personali tra antichità e medioevo.’ In: Rubeis, Flavia De & Walter Pohl (eds), Le scritture dai monasteri. Atti del IIo seminario internazionale di studio ‘I monasteri nell’alto medioevo.’ Rome: Institutum Romanum Finlandiae, 15–45. Solin, Heikki, & Olli Salomies. 1994. Repertorium nominum gentilium et cognominum Latinorum. 2nd edn. Hildesheim: Soc. Scientiarum. Stahl, Alan M. 1982. The Merovingian coinage of the region of Metz, Louvain-La-Neuve: Institut supérieur d’archéologie et d’histoire de l’art. Stotz, Peter. 2000. Handbuch zur lateinischen Sprache des Mittelalters, vol. 2: Bedeutungswandel und Wortbildung. München: Beck. Strothmann, Jürgen. 2011. ‘Merowingischen Monetarmünzen und die Gallia im 7. Jahrhundert. Münzorte und Monetarnamen.’ In: Greule et al. Manuscript in preparation. Verhulst, Adriaan. 1999. The Rise of Cities in North-West Europe. Cambridge: Cambridge University Press. Ward-Perkins, Bryan. 2005. The fall of Rome and the end of civilization. Oxford: Oxford University Press. Weidemann, Margarete. 1986. Das Testament des Bischofs Berthramn von Le Mans vom 27. März 616; Untersuchungen zu Besitz und Geschichte einer fränkischen Familie im 6. und 7. Jahrhundert. Mainz: Verl. d. Röm.-German. Zentralmuseums. Wickham, Chris. 2009. The inheritance of Rome. A history of Europe from 400 to 1000. London: Allen Lane.
Remco Knooihuizen
The Use of Historical Demography for Historical Sociolinguistics: The Case of Dunkirk1
Abstract This chapter combines historical demographic and linguistic research to come to an understanding of language shift in seventeenth-century Northern France. The article details the Francophone immigration to the town of Dunkirk around and after its annexation to France in 1662, focusing in particular on the integration of migrants into the local community through intermarriage. The author then reports on a quantitative dialectological study of the French of Dunkirk in comparison to other French varieties. This study shows similarities in particular with varieties from the migrants’ places of origin, lending credence to the suggestion that Francophone migration played an important role in the language shift. Moreover, this research signals the importance of an integrated social and linguistic perspective in the writing of language histories.
1
This article would not have been possible without the statistical help of Dan Dediu with the demographic data from Section 4, and the support of Paul Heggarty and Warren Maguire in dealing with the linguistic data from Section 5. I would further like to thank April McMahon, Graeme Trousdale, Wilson McLeod, Peter Trudgill, Adam Fox, and the audience at the Bristol conference for their useful comments. This research forms part of a doctoral project supported by funding from the Arts & Humanities Research Council.
324
Remco Knooihuizen
1 Introduction The study of sociolinguistics involves correlating linguistic variation in a community with varying social characteristics of the people in whose linguistic data we found the variation. This means that sociolinguists should want to know who their informants are, and how they behave. Even in modern sociolinguistic research, this is dif ficult enough: we are interested in people’s most natural behaviour, linguistic or otherwise, but by observing and recording that behaviour, we introduce outside factors that compromise the naturalness of the situation – Labov’s Observer’s Paradox. In addition to that, there are so many social and attitudinal factors that could inf luence language use, that it seems almost impossible to obtain a complete record of them. Historical sociolinguistics faces problems of a slightly dif ferent nature. The historical linguistic record may not be inf luenced by the presence of a sociolinguist’s recording device, but we often have no idea what the true circumstances of the linguistic data were either. The speakers, or writers, may even be more abstract entities than the individuals from modern-day sociolinguistics – and either way, our informants are dead. Suf fice it to say that historical sociolinguistics faces a ‘bad data’ problem (Labov 1994: 11), both with regard to the linguistic production itself and with regard to the social background of that production. In order to make the best use of such bad data, we need to apply dif ferent methodologies and draw on data from other fields, such as social history, to complement our research. In this chapter, I explore the links between historical sociolinguistics and historical demography, in order to see how both fields can inform each other. Data for the chapter is drawn from a study of language shift in Early Modern France. I begin the chapter with some introductory remarks about minority languages and language shift in Europe in the Early Modern period (Section 2), before giving a brief introduction of the case study in Section 3. The historical demographic evidence is discussed in Section 4, followed by historical linguistic evidence in Section 5. I finish with some conclusions (Section 6) about both the case study and the research methods used.
The Use of Historical Demography for Historical Sociolinguistics
325
2 Minority languages and demographic evidence Minority language groups are usually defined with reference to a larger group, most commonly the population of a nation-state. By this reasoning, minority languages have only existed since the emergence of nationstates in the Early Modern period (c. 1500–1800). This point is made, for example, in Wright (2004: 24). The Early Modern period is the period of the rise of vernacular languages at the expense of Latin, but also of the rise of certain vernacular languages at the expense of others (see, e.g. Burke 2004, Baggioni 1997). Several minority languages were considerably worse of f with these Early Modern developments. A number of them died out, such as Old Prussian, Orkney and Shetland Norn, and Cornish, while others were restricted in terms of domains of use, geographical area, and number of speakers. Examples of this last category are Welsh, Frisian and Sorbian. Often, restrictive language policies by governments and churches, proscribing the use of minority languages, are blamed for the demise of these languages. There are however a number of problems with such an account. Policies were often aimed predominantly at streamlining government communication, not at changing the language of the minority language speaker’s home. For example, there is great debate about the specific variety of language prescribed in the French Decree of Villers-Cotterêts (1539) (see Martel 2001 for an overview), but the decree’s restriction to the judicial domain is undisputed. Moreover, there is little evidence of the policies being enforced, and there are few ideas about how they could have been enforced. Most problematically, the language policy theory cannot explain why rapid language shift occurred in the absence of such policies (Orkney and Shetland Norn, see Knooihuizen 2005), or why minority languages were maintained to this day despite such restrictive policies, as is the case with Sorbian and most minority languages in France. (See also Evans, this volume, on the propagation of of ficial languages in this period.) Perhaps a better explanation for the language shift in all these cases is the migration of majority-language speakers to the (often recently annexed) minority-language area. A comparative survey of cases across Early Modern
326
Remco Knooihuizen
Europe suggests that such migration may be a necessary and suf ficient condition for language shift; social network theory of fers a possible mechanism for the language shift. A further indication that migration may be a factor in language shift is that some of the anti-minority language (or rather, pro-majority language) legislation is explicitly aimed at causing a language shift through the organized settlement of the minority-language area: this is the case in some areas of Lusatia (the Friderizianische Kolonisation, see Kunze 1999: 9), the Scottish Highlands, and Ulster. In order to investigate how migration may have inf luenced language shift, I will chart the demographic change caused by migration and the interactions between immigrants and locals in Dunkirk, a town in a minoritylanguage area in Northern France.
3 Case study: Dunkirk In the mid-seventeenth century, Dunkirk was a small town of approximately 5,000 inhabitants. The town and the surrounding ‘Westhoek’ area of Flanders were politically part of the Spanish Netherlands. The population were predominantly Dutch speakers, although due to the proximity of the Dutch-French language border, French ‘was never a truly foreign language’ in the area (Coornaert 1970: 136). During the Franco-Spanish war, Dunkirk was taken by French forces, ceded to Cromwell’s England (1659), and finally sold back to France after the Restoration (1662). The surrounding countryside also came under French control (Coornaert 1970: 148–150; Cabantous 1983: 58–59). The annexation of Dunkirk was followed by a migration of French speakers to the town (Lottin & Guignet 2006: 210; Coornaert 1970: 171) and a spectacular growth to approximately 14,000 in 1706 (Lambin 1980: 163–164; Cabantous 1983: 89). These included administrative and military personnel, but also builders involved in the fortification of the town, works that were repeated several times when peace treaties required the fortifications to be taken down (Cabantous 1983: 73; Lottin & Guignet 2006: 210).
The Use of Historical Demography for Historical Sociolinguistics
327
The questions I aim to answer in this article are threefold. Where did the French-speaking immigrants come from? How much did the immigrants interact with the local population? And what were the linguistic ef fects of these interactions?
4 Demographic developments in seventeenth-century Dunkirk 4.1 Data I investigated the interaction between the Dutch-speaking and Frenchspeaking population groups in Dunkirk by analysing marriage registers from the town, as marriage is obvious evidence of interaction. I collected data at ten-year intervals from 1647 to 1697, a period spanning both sides of the annexation. In total, the marriage registers for these six years contain just over 1,000 entries; the uneven distribution over the years shows the growth of the town in this period, especially in the years after the annexation. The entries in the marriage register contain the names of both spouses, of a varying number of witnesses, and of the minister who registered the marriage. By describing the spouses as juvenis ‘youth’ or viduus/vidua ‘widow(er)’, they also contain information on people’s previous marriages. The registers for 1647 and 1657 also mention the spouses’ parish of origin, while the 1697 register contains spouses’ and witnesses’ signatures or a declaration of illiteracy. The data collected from these marriage registers can be seen as fairly representative of Dunkirk society as a whole. There are no indications that the data is biased towards higher classes, for example. As the data is taken from Catholic marriage registers, it does not show the small Protestant community in the town, estimated to be 7 per cent of the population (Coornaert 1970: 117). The Protestants would mostly be Dutch or English speakers; the few English speakers in this Catholic data bear names like Catriona MacDonald and Malachi Donnely.
328
Remco Knooihuizen
4.2 Methods The dif ferent demographic characteristics of the individuals from the marriage registers were correlated with patterns of migration and marriage, looking for significant interactions. Because the demographic aspect of interest, language allegiance, was not directly available from the registers, I used two proxy variables: names and parish of origin. The link between a person’s name and their language of preference is not a direct one, especially in a long-standing contact situation as on the Germanic-Romance linguistic border. There is, however, a tradition of using onomastic evidence for language allegiance, also in this area (e.g. Poulain & Foulon 1981). Despite not being a fully reliable indicator, names are the best indicator available to us. Similarly, a person’s parish of origin is not connected directly to their language allegiance. Societal language preferences do not necessarily say anything about individual language preferences: we may be dealing with a Dutch speaker from a majority French-speaking town. Furthermore, the earliest reliable data on societal language preference dates from the mid-19th century (Pée & Blancquaert 1946), and in the French Flemish situation of a ‘language shift from within’ (Willemyns 1997), two centuries can mean a world of dif ference. However, the correlation between the name-based and the location-based language allegiance of a person is highly significant (p = 5.176 × 10–8), suggesting that these methods of determining language allegiance are reasonably reliable. Further, I used people’s ability to sign (in 1697 only) as a proxy for literacy or education. Again, these are very tenuous links indeed, and this type of evidence of fers only a rough indication of people having had very basic schooling. As problems with this method af fect dif ferent population groups equally (Cressy 1977: 142), however, any patterns emerging from this evidence will still represent actual dif ferences. There is a wide range of f luency in signatures, which was collapsed into a binary variable following the guidelines from Grevet (1991: 44–45). Impressionistically, there appears to be a division between the italic hands of presumed French speakers and the secretary hands of presumed Dutch speakers, but this distinction has not been looked into in more detail.
The Use of Historical Demography for Historical Sociolinguistics
329
4.3 Results 1: Migration patterns Migration patterns could only be investigated for those years when the entries in the marriage register mentioned the spouses’ parish of origin, i.e. 1647 and 1657. Unfortunately, this is before the annexation by France, and no clues to migration patterns immediately post-annexation are available. The pre-annexation data will supply a baseline for migration, and other types of data will have to be used as evidence for continuity or change in these patterns after 1662. (Marriage registers from the late eighteenth century did include this information again, and have been analysed in Cabantous 1983: 93–94; these however have no bearing on the situation in the mid-to-late seventeenth century.) A striking first observation is that the marriage registers contain roughly equal numbers of people from Dunkirk and from elsewhere (Table 18). This high proportion of immigrants is indicative not of the growth of the town – this was minimal before the annexation – but of the fact that mortality in Early Modern urban areas was so high that continued immigration was necessary only to maintain the population size (Sortor 2005: 165). There was no gender imbalance in the immigration in 1647, but in 1657 there were significantly more male than female immigrants due to a drop in the female immigration rate. The reasons for this drop are unknown, but may be found in the ongoing war in the area at the time. The parishes of origin could be identified for 168 immigrants. Of these, 122 (73 per cent) were from parishes where Dutch was the majority language, and 45 (27 per cent) were from majority French-speaking parishes. One immigrant was originally from Lübeck, Germany. An analysis of the distances between Dunkirk and the parishes of origin shows that this was a short-distance migration. The median distance overall was 31.7 km; for obvious reasons – the location of the language border some 30 km south of Dunkirk – the median distance travelled by Dutch-speaking migrants (23.1 km) is smaller than that by French-speaking migrants (57.9 km). As most parishes supplied only one migrant to Dunkirk, there is no evidence of so-called ‘chain migration’. This may however be an ef fect of the small, intermittent sample.
330
Remco Knooihuizen Table 18 Locals and immigrants in the Dunkirk marriage register, by gender and sample year
Men Women
Year
Locals
Immigrants
1647
23 (29.5%)
55 (70.5%)
1657
22 (38.6%)
35 (61.4%)
1647
36 (40.9%)
52 (59.1%)
1657
37 (58.7%)
26 (41.3%)
These findings on migration before the annexation of Dunkirk form a baseline from which to analyse the developments after 1662. The two years for which migration data are available show no dif ference in the areas migrants originated from, but changing patterns in the population after the annexation might give more information about dif ferent migration patterns. 4.4 Results 2: Inter-ethnic marriage patterns Data for all sampled years between 1647 and 1697 was used to look at patterns of ethnic intermarriage in Dunkirk. All individuals in the sample were assigned a language-based ethnicity on the basis of their name. The data comprise just over 1,000 marriages; for both genders, the ratio of Dutchto French-language background is approximately 3:2. The distribution of marriage patterns for all years together is shown in Table 19. Ignoring marriages involving a partner with a language background other than Dutch or French, the pattern shows a clear preference for in-group marriage (p /au/), or, even more important, the wide-spread unrounding of front vowels. Rounding before /l/ occurs only once (außerwöhlten), and so does velarization of MHG /â/ > /ô/ (ohn, a form that has become integrated into modern German). ‘Morphological’ spelling (Voeste 2008: 15–17), an orthographic principle that has shaped the modern standard decisively, is already well represented. ärmlein, Kräntzlein, Wängelein (cf. Arm, Krantz, Wange); but behelt (cf. behalten)
Sixteenth-Century Street Songs and Language History ‘From Below’
379
‘Syllabic’ spelling (Voeste 2008: 13–15), the insertion of , normally as a marker of vowel length, in some cases occurs as in the modern standard; in others it does not. Ehr, sehr, Bůhle, etc. erst
(= NHG); (= NHG);
můht ‘Mut’ war ‘wahr’, wol ‘wohl’
(≠ NHG) (≠ NHG)
The can also be used to enhance the weight of small words (Vergewichtung, Voeste 2005: 12) with lengthened, but originally short, vowels. jhr, jhren (MHG /i/) ‘ihr, ihren’ (= NHG) ~ jr, irem ‘ihr, ihrem’ (≠ NHG)
In this respect the text only partly corresponds with modern German, but one should bear in mind that insertion or omission of remains the most unsystematic feature of German spelling up to this very day. Initial