Language or Dialect? The History of a Conceptual Pair 9780198845713

Citation preview

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

Language or Dialect?

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

Language or Dialect? The History of a Conceptual Pair RAF VAN ROOY

1

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Raf Van Rooy 2020 The moral rights of the author have been asserted First Edition published in 2020 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 0000000000 ISBN 978–0–19–884571–3 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

For Elien [ksin ε ˈɣεrə]

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

Contents Acknowledgements List of Figures List of Tables Conventions

xiii xv xvii xix

1. Introduction

1 I. PREHISTORY, 500 –1500

2. A dive into the prehistory of the conceptual pair 2.1 The Greek dialects between anecdotes and definitions 2.2 Diálektos, a variety of interpretations 2.3 Philology as stimulus 2.4 The Latin West 2.5 Conclusion 3. The exception to the rule: Lingua and idioma in Roger Bacon’s thought 3.1 Bacon at Babel 3.2 Another first for Bacon? 3.3 An outlook on different languages and their dialects: a central precondition 3.4 Thomas Aquinas: a like-minded exegete? 3.5 The Tower of Babel 3.6 Conclusion

15 16 18 21 22 27 28 30 31 35 38 40 42

II. THE ORIGIN OF THE CONCEPTUAL PAIR, 1500–50 4. From dogs and hounds to languages and dialects: The conceptual pair in Conrad Gessner’s work 4.1 Exploring the linguistic world 4.2 Conrad Gessner, certified cataloguer and compiler 4.3 Ad fontes! Gessner reads Clement of Alexandria 4.4 The meanings of dialectus according to Gessner 4.5 Conclusion

47 49 52 55 57 61

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

viii



5. Lingua and dialectus: From synonymy to contrast 5.1 Dialectus as a Latin word 5.2 The key symptom: contrasting lingua to dialectus 5.3 Developing the concept of common language 5.4 Updating dialectus definitions 5.5 The coining of a new phrase: ‘to differ only in dialect’ 5.6 An intricate conceptual web 5.7 Conclusion 6. Hellenism, standardization, and info-lust: The genesis of the conceptual pair in context 6.1 A major pivoting point: the rediscovery of the Greek dialects 6.2 From ancient Greece to Western Europe: standardizing the vernaculars 6.3 Knowledge revolution: information explosion and info-lust 6.4 The countability of language 6.5 A product of appropriation and subconscious adaptation 6.6 Conclusion

62 62 64 67 72 74 76 79 81 81 84 87 88 90 91

III. CONSOLIDATION BY ELABORATION, 1550–1650 7. Space and nation: Greek definitions transformed 7.1 The spatial conception of dialect 7.2 Two topical topoi 7.3 Regional language variation: a universal phenomenon? 7.4 Nation or nations? The ethnic conception of the language/dialect distinction 7.5 Towards a political interpretation of ‘nation’? 7.6 Space and nation: diagnostic criteria? 7.7 Conclusion

95 95 98 99 101 105 107 108

8. Aristotle’s legacy: Substance, accidents, and mutual intelligibility 8.1 Extending the ‘to differ only in dialect’ phrase 8.2 Mithridates: polyglot or not? 8.3 A question of gradation: devising different levels of dialects 8.4 Mutual intelligibility: an early modern criterion 8.5 Johannes Goropius Becanus and immediate mutual intelligibility 8.6 The communicative reach of dialects versus languages 8.7 Conclusion

109 109 113 114 117 119 122 123

9. A subjective touch: Language beats dialect 9.1 Analogical norm or anomalous deviation? 9.2 From common to standard language 9.3 Superior or inferior? 9.4 Conclusion

125 125 127 129 134

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi



ix

10. The conceptual pair and language history: Language generates dialects 10.1 Language mothers and offshoot dialects 10.2 An early critic of the language-historical interpretation 10.3 A discursive strategy for historical classification 10.4 Conclusion

136 136 141 143 146

11. Consolidation by elaboration: Drawing the balance 11.1 The seven main interpretations: a synthesis 11.2 Emancipating the conceptual pair from the Greek heritage 11.3 Conclusion

147 147 151 156

12. The conceptual pair in transition: The case of Georg Stiernhielm 12.1 Georg Stiernhielm, a Swedish all-round scholar 12.2 The conceptual pair according to Stiernhielm 12.3 Conclusion

159 159 163 166

IV. SYSTEMATIZATION AND RATIONALIZATION, 1650–1800 13. Putting the conceptual pair on the scholarly agenda: The orientalist Albert Schultens 13.1 Schultens’s definitions of dialect 13.2 Language, dialect, and degenerate offshoot 13.3 Classes of linguistic variation 13.4 Schultens’s legacy 13.5 Conclusion 14. Lexicostatistics avant la lettre: The historian Johann Christoph Gatterer and the conceptual pair 14.1 Gatterer’s Vorrede 14.2 Characteristic words: Gatterer and basic vocabulary 14.3 Determining the degree of linguistic kinship: Gatterer’s lexicostatistic framework 14.4 The historian Gatterer and the grammarian Adelung 14.5 Conclusion 15. Classes of variation: How do languages and dialects differ? 15.1 Casting off John the Grammarian’s yoke 15.2 From scattered comments to systematization 15.3 Two eighteenth-century outsiders in the quest for linguistic criteria 15.4 Conclusion 16. Between systematization and rationalization: The conceptual pair through the Enlightenment lens 16.1 Towards a dialectological tradition? 16.2 The first sceptical voices 16.3 Conclusion

171 174 176 178 180 181 183 184 186 187 190 191

193 193 195 200 202

204 204 211 216

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

x



V. FROM SILENT ADOPTION TO OUTSPOKEN ABANDONMENT, AFTER 1800 17. From Jones to Gabelentz: Silent adoption and renewed suspicion 17.1 The beginnings of modern linguistics 17.2 Two pioneering dialectologists: Johann Andreas Schmeller and Albert Giese 17.3 Questioning the conceptual pair: August Schleicher and William Dwight Whitney 17.4 The late nineteenth century: between continued usage and increasing scepticism 17.5 Conclusion

219 220

18. Schuchardt the iconoclast 18.1 Tree or waves? 18.2 A pioneering outsider 18.3 Breaking down the walls between languages and dialects 18.4 Linguists, shibbolethists 18.5 Useless abstractions? 18.6 Conclusion

231 231 233 235 238 240 242

19. From Saussure to 1954: Structuralism and the language/dialect distinction 19.1 In Saussure’s class 19.2 Towards a structural dialectology: Uriel Weinreich’s diasystem 19.3 Redefining the conceptual pair: Martinet and Polák 19.4 Conclusion

221 225 228 229

244 245 247 251 254

20. Mutual intelligibility: The number one criterion? 20.1 A criterion making career 20.2 Measuring linguistic distance 20.3 Conclusion

256 258 259 261

21. Between two extremes: Generative and sociolinguistic interpretations 21.1 Generative grammar: no country for old dialects? 21.2 Variables over systems 21.3 Destructive dismissal 21.4 Supplementing the conceptual pair 21.5 By way of conclusion: a quest for alternatives

263 263 266 268 270 273

22. A gentle goodbye? Dialect stripped for parts 22.1 Farewell to the conceptual pair? 22.2 The language/dialect distinction after 1900: the story of a love-hate relationship

275 278 280

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi



23. Language, dialect, and the general public—or how not to popularize knowledge 23.1 The conceptual pair popularized 23.2 The Weinreich witticism in context: the nation-state, language policy, and mass education 23.3 Political activations of the conceptual pair 23.4 Conclusion

xi

284 284 288 289 293

24. Language and dialect between past and future: Terminological success, conceptual failure?

296

References Index

303 333

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

Acknowledgements This book presents ideas which I first began to develop during my PhD fellowship at KU Leuven in the years 2013–17. This research led to a dissertation which I defended in late May 2017 under the supervision of Toon Van Hal and Pierre Swiggers, to both of whom I owe a great deal. It was their fascinating work which excited my interest in the history of linguistics as a field of research. Their feedback on my dissertation has proved invaluable in preparing the present monograph, in which I have tried to decompress my ideas and to bridge the gap between 1800 and the present, left open in my dissertation. I have been able to do so by carrying out extra research during the Winter semester of 2018–19 in Athens. Both my PhD fellowship and my stay in Greece were made possible by the generous support of the Research Foundation—Flanders (FWO). I am likewise much indebted to numerous other scholars, many of whom have selflessly acted as true mentors. To this category belong first and foremost John Considine, Lambert Isebaert, and Han Lamers. They have, moreover, shared with me countless invaluable documents, both primary sources and secondary literature. For the same reason, Josef Eskhult deserves a specific mention here, since he was so generous as to grant me access to his work in progress and to send me scans of source material related to Georg Stiernhielm. I also thank Alexander Maxwell for his helpful and very detailed feedback on Chapters 7–16. I owe a deep debt of gratitude to the staff of libraries in Amsterdam, Athens, Berlin, Copenhagen, Edmonton, Gera, Ghent, Jena, Lille, Leuven, London, Madrid, Munich, Paris, Rostock, and Vienna. I am particularly grateful to the staff of the Forschungsbibliothek in Gotha and the University Library in Leiden, where I have been able to work in perfect peace, generously supported by a Herzog-Ernst-Stipendium of the Fritz Thyssen Stiftung and by a Scaliger Fellowship of the Leiden Scaliger Institute, respectively. I am also grateful to the people behind the many and often freely available digital initiatives, not in the least Google Books and its partnerships with institutions like the Bayerische Staatsbibliothek, without which my research would have been unfeasible. I have, moreover, had privileged access to useful databases such as Early English Books Online and Eighteenth Century Collections Online thanks to, among other institutions, KU Leuven and the University of Alberta. It was, as a result, often unnecessary to spend much time travelling to research libraries across Europe and beyond, even if there are certain dangers tied to the dematerialization of the printed book (Grafton 2009). It may, for instance, lead to a neglect of important material aspects of the books digitized, such as their size, or to a neglect

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

xiv



of individual copies and their unique characteristics, since often only one copy of a book is available in digital format. Additionally, maps and schemes that are printed on pages of larger size are sometimes omitted in online reproductions. Be that as it may, the advantages outweigh the disadvantages, and I am grateful that I have been able to make optimal use of these digital resources. Optical character recognition (OCR) software has been an indispensable tool, too, as this made it possible to perform goal-oriented searches in digitized text files in Roman script, even though it must be added that, in its current state, the tool is not infallible, certainly not for early modern books. It is, however, my hope and expectation that, in the future, the technique will greatly improve and will be extended to non-Roman characters. Last but not least, my wife, Elien, and my parents have as always supported me unconditionally in writing and finalizing this book; for this, and for everything else, I admire and love them.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

List of Figures 3.1. Roger Bacon at his Merton College observatory, oil painting by Ernest Board

29

3.2. The Tower of Babel through the eyes of Pieter Bruegel the Elder, 1563

41

4.1. Dogs in Conrad Gessner’s work

48

4.2. World map by Abraham Ortelius (1527–98), 1570

50

4.3. Conrad Gessner, portrait by Tobias Stimmer (1539–84)

53

4.4. Marble bust of the Pontic king Mithridates VI, first century 

54

5.1. Conceptualization of a language (e.g. Greek) before the early Cinquecento

64

5.2. Conceptualization of a language (e.g. Greek) emerging in the early Cinquecento

64

5.3. Desiderius Erasmus holding a Greek book, by Hans Holbein, 1523

65

5.4. Antesignanus’s perception of the relationship between Greek and French variation

71

6.1. The first separate edition of Adrien Amerot’s On the diverse dialects, 1530

83

6.2. Woodcut showing sixteenth-century printing in action, 1568

86

6.3. Natural history cabinet, 1599

87

7.1. Claude de Saumaise (1588–1653)

104

8.1. Bust of Aristotle, after an original of about 330 

110

8.2. Johannes Goropius Becanus by Philips Galle, 1572

119

10.1. Joseph Justus Scaliger’s influential Diatribe on the languages of the Europeans

137

10.2. Scaliger c. 1607, engraving by Johann Theodor de Bry (1561–1623)

138

10.3. Johann Heinrich Hottinger etched by Georg Meyer, 1664

142

11.1. The dimensional triangle shaping the early modern conceptual pair

149

12.1. Georg Stiernhielm

160

12.2. Impression of Georg Stiernhielm’s tree metaphor

164

13.1. Albert Schultens, 1730

172

13.2. Impression of Albert Schultens’s tripartite conceptual hierarchy

177

14.1. Johann Christoph Gatterer

184

16.1. Caspar Wyss’s Dialectologia sacra, 1650

209

18.1. August Schleicher’s (1863) family tree model

232

18.2. The linguistic continuum as a slanted line according to Schmidt

233

18.3. The linguistic continuum broken up according to Schmidt

233

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

xvi

  

18.4. Hugo Schuchardt in 1922, picture by Hermann Urtel

234

18.5. Schuchardt’s (1928 [1868]: 165) conception of dialect groups visualized

236

20.1. Example of a typical bimodal distribution on a scale

260

21.1. Noam Chomsky in 1977

269

23.1. The Council of Europe’s headquarters in Strasbourg

291

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

List of Tables 4.1. Clement of Alexandria’s Greek compared to Hervet’s and Gessner’s Latin versions 9.1. Transfer of social prejudices to the conceptual pair 11.1. Summary of the main early modern interpretations of the conceptual pair 14.1. Johann Christoph Gatterer’s four categories on the kinship continuum

56 134 148 189

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

Conventions Non-English primary texts, when quoted, are systematically translated into English in the main text. The original passages are given in the footnotes. Translations are mine, unless otherwise indicated. Other relevant passages are occasionally cited in the footnotes in their original form and without a translation. Ancient and medieval Latin texts are quoted from the editions used by Brepols’s Latin databases (Library of Latin Texts A and B as well as Monumenta Germaniae Historica), unless otherwise indicated. I have regularized ancient Latin orthography by reserving for the vowel [u] and for the semivowel [w]. Ancient Greek and Byzantine texts are quoted from the editions used by the Thesaurus Linguae Graecae (TLG) database, unless otherwise indicated. Greek words and short quotations have been transliterated into the Roman alphabet in the main text, with the original Greek quoted between round brackets. For recurring terms, the original Greek is given at the word’s first appearance. Longer quotations are cited in the footnotes and provided only in the original Greek alphabet. Titles of Greek and Latin texts cited in footnotes have normally been taken over from the Brepols and TLG databases. All English Bible quotes are cited according to the English Standard Version. I have quoted early modern and modern texts as I have found them in the original sources. I have resolved abbreviations between square brackets. Errors and misprints have been marked with [sic]. Original bold font or unusual font, for instance Gothic in an otherwise Latin-script text, have been converted to standard Roman script. All italic fonts have been romanized. I have converted the superscript e to the modern umlaut sign in early modern German quotations. When pagination is lacking, I have used the signature markings to refer to the page intended. The capital letters in titles of early modern works have been normalized in the bibliography at the back. Abbreviations in the publisher’s names have not been resolved. Names of Greek, Latin, and early modern authors have been Anglicized whenever this is common in secondary literature. Otherwise, I have opted for one of the most common forms. Life dates are provided in the main text when a source author is first introduced but not in the footnotes. For living persons, I offer the year of birth, if known. I refer to early modern dissertations by mentioning the name of the supervising professor (praeses) as well as the presenting student (respondens), unless I have good reasons to suppose that one of them should be considered the sole author of the dissertation (on this issue see e.g. Considine 2008b). Furthermore, to distinguish between concepts and terms, I rely on italics to denote that the discussion pertains to a term. Finally, I use singular they in order to avoid gendered language.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

1 Introduction When during World War II the linguist Max Weinreich (1894–1969) finished one of his New York lectures on the Yiddish language and its social status, one of his auditors, a young teacher from the Bronx, came up to the front of the room for a follow-up discussion. He boldly addressed Weinreich: ‘What is the difference between a dialect and a language?’—a question frequently asked of linguists in general. As Weinreich tried to give him a satisfactory solution, the teacher interrupted him and said: ‘That I know, but I will give you a better definition: “A language is a dialect with an army and navy”.’ The identity of the auditor is still up for discussion, but his remark was to have a rich and powerful resonance. Weinreich was so impressed by the witty statement that he decided to mention and discuss it at length in one of his publications, a paper, in Yiddish, entitled ‘The YIVO and the problems of our time’, which he had prepared for the annual conference of the YIVO, the Yiddish Scientific Institute, in 1945 (Weinreich 1945). The rest, as they say, is history. Today, as soon as the distinction between the concepts language and dialect is raised in debate, someone will take refuge in the quip recorded by Weinreich to relativize it and demonstrate that it is primarily informed by language-external, sociopolitical circumstances. The question whether one labels a speech form language or dialect depends, in other words, on the history of its speech community, its social status, and its participation in power or lack thereof (see Maxwell 2018). In such debates, the distinction between language and dialect is, almost as a rule, taken for granted. It is, in fact, usually perceived as a kind of ahistorical given, an ever-present and obvious component of our metalinguistic apparatus (cf. Kamusella 2016: 164). The conceptual pairing is, however, not as self-evident as we tend to assume. Like all aspects of human culture, it has a history, which I will sketch in this book. In this history, the early modern era, roughly 1500–1800, was a critical stage. During this period, scholars with various interests and backgrounds began to address, often at great length, questions about linguistic diversity in general and dialectal variation in particular: why are there not only diverse languages, but even differences in every language individually? Is there a litmus test for determining whether a specific form of speech can be considered an actual language or rather a dialect deriving from, and subsumed under, a language? What criteria, causes, and circumstances can be invoked in this debate? Despite the lively interest which early modern scholars took in these issues, historians of linguistics have thus far made no attempt at exploring the origin and Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

2



evolutions of the language/dialect pair in depth. Usually, it is situated in Greek antiquity because of the origin of the term dialect (e.g. Kamusella 2015: 10–14; 2016: 172–5), but I will argue that this view is based on an unfounded assumption. A rare exception is the all-round linguist Mario Alinei (1980: 11–27), who situated the origin of the distinction in the Renaissance in an intriguing paper published four decades ago. Alinei’s compatriot Mirko Tavoni (1986: 221), too, has placed ‘the birth of the modern concept [of dialect] through contact with the Greek sources’ around 1500. In this book, I follow Alinei’s and Tavoni’s lead as I seek to trace the emergence of the conceptual pair, a complex matter on the crossroads of linguistics, intellectual history, and philology. My main aim is to present the poorly known history of the language/dialect distinction to anyone interested, inside and outside academia: linguists, historians, philologists, and non-specialists with a general interest in history and language. Even though the focus is on the early modern era, it is indispensable that I zoom out. In order to fully grasp the history of the conceptual pair in Western linguistic thought, I have to travel back in time to antiquity and the Greek world in particular, with which early modern scholars developed a deep fascination. There, they stumbled across an age-old situation of dialectal diversity that reminded them of their native tongues; this encounter profoundly changed the way they thought about variation within languages. Not only did they adopt the term closely associated with Greek variation, diálektos (διάλεκτος), they also adapted the concept that was intertwined with it to their own here-and-now. The rediscovery of Greek diversity served as an eye-opener for humanists, many of whom assumed that Latin, the language of science, education, administration, and religion, was an invariable tongue that did not suffer from regional differences. I do not only look back to antiquity, as this would result in a very incomplete picture. Understanding the conceptual pairing, its different interpretations, and the suspicions which many present-day scholars harbour about it requires an investigation of modern developments, too. In this period, linguistics emerged as a separate research discipline and was institutionalized. How did this affect the language/dialect distinction? At the same time, the conceptual pair, until then mainly restricted to educated circles, emerged as popular knowledge. In which ways and contexts did this knowledge transfer occur, and why? By answering these questions, it is possible to sketch the history of two intertwined concepts that have been, and still are, frequently used by scholars and laypeople alike in a wide range of different contexts. My main focus is on the Western tradition of grammatical and linguistic scholarship, with a slight West-Germanic tilt for five principal reasons. Firstly, the conceptual pair seems to have been discussed most intensely by authors with a West-Germanic background. Secondly, available scholarship is also concerned with these authors. Thirdly, English texts are currently most easily accessible in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi



3

this digital age. Fourthly, the focus is dictated in part also by the limitations of my linguistic competence. Fifthly, a Germanic tilt also benefits the reader of this book, obviously composed in English. The available source material is extremely varied, especially for the periods before the modern age, when linguistics had not yet emerged as an autonomous branch of study. It ranges from grammars, linguistic handbooks, and treatises over lexica and journal articles to philological and historiographical works of diverging nature. I focus on printed works for the period from the Renaissance onwards, as these generally had a wider circulation than manuscripts and therefore were in a position to achieve a greater impact. In presenting the results of my analysis, I have tried to strike a balance between the general and the specific. I have, on the one hand, taken a bird’s-eye view across twenty-five centuries, tracing the journey of the language/dialect distinction from its prehistory in antiquity to modern language studies. Charting continuities and break-off points has been one of my main concerns. On the other hand, in order to avoid doing injustice to the individuality of scholars and running the risk of losing my grip on the actual sources, I have chosen to draw out telling case studies throughout the book. As such, I have adopted an approach that is more or less in line with David Armitage’s ‘history in ideas’, a genre of intellectual history in which episodes of contestation over meaning form the stepping-stones in a transtemporal narrative constructed over a span of time extending over decades, if not centuries. The ‘ideas’ structuring this history would not be hypostatised entities, making intermittent entries into the mundane world from the idealism’s heavenly spheres, but rather focal points of arguments shaped and debated episodically across time with a conscious—or at least a provable—connection with both earlier and later instances of such struggles. (Armitage 2012: 499)

This longue durée approach, which underlines the variability of concepts and their context-bound constitution, has been further developed by Armitage in a co-publication with Jo Guldi (Guldi and Armitage 2015). I have taken great care to put the different episodes of the history of the conceptual pair in their correct setting, an indispensable step in sound studies in intellectual history, as most scholars agree following Quentin Skinner’s seminal 1969 paper ‘Meaning and Understanding in the History of Ideas’. These contextualization efforts are all the more crucial, since language was usually not studied in and of itself before the nineteenth century. I take contexts as a plural noun, as they differ for each author and each source text individually, and I follow Kristin Asdal and Ingunn Moser’s suggestion that contexts should not be regarded as invariable givens but rather as selective constructions of the historian’s own making (Asdal and Moser 2012: 303).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

4



The main question which I aim to address in this book, then, is the following: how and in which contexts have scholars thus far tried to distinguish a language from a dialect? I have tried to answer this question in a more or less chronological account of twenty-two brief chapters. This choice of presentation will, I hope, make the complex history of the conceptual pair a palpable and well-structured whole for the reader. Further coherence has been created by grouping the chapters into five larger parts, which coincide with the main episodes of the story which I am about to tell. Part I, on Greco-Roman antiquity and the Middle Ages, is the shortest, and with good reason, as I argue that the conceptual pair language/dialect was largely absent from linguistic thought in these periods. This absence also explains why I have called the first part ‘Prehistory’. Chapter 2 starts at the ultimate origin of the term dialect, ancient Greece, unveiling how it was never customary in Greek scholarship, both ancient and medieval, to contrast the term diálektos to a word referring to language. In order to prove this, the chapter treats both passing references to the Greek dialects and influential definitions of the Greek word. It also frames definition attempts in their philological context. Finally, I briefly discuss the Latin tradition up to about 1500, arguing that an obvious opposition of dialect to language cannot be discovered there either. Chapter 3 offers a detailed case study of the main exception to this general tendency, the late medieval scholar and polyglot Roger Bacon, who opposed the terms lingua and idioma in a way that prefigured the later language/dialect distinction. Bacon was able to do so because of his exceptionally broad intellectual and linguistic horizons; he was familiar with English, French, Latin, Greek, and other tongues, as well as the regional variations within some of them. His mastery of Greek in particular was unique in his times. Bacon’s linguistic outlook was a central precondition and a triggering circumstance for his lingua/idioma distinction. He was in good company, since the renowned philosopher Thomas Aquinas presupposed a similar metalinguistic contrast in his exegetical works; yet Thomas was much less explicit about it. Chapter 3 finishes with a brief exposé of biblical exegesis as an overlooked source for ancient and medieval ideas on regional variation. In Part II, I turn to the major argument of this book, endeavouring to reveal the largescale emergence of the distinction between language and dialect in the first half of the sixteenth century. It begins in the same way as Part I ended, with a case study of a prominent scholar. Chapter 4 uncovers the way in which the Swiss humanist Conrad Gessner, an important language scholar, bibliographer, and zoologist, conceived of the Latin term dialectus in opposition to lingua. Renaissance intellectuals were confronted with a major explosion of information, also on the languages of the world, and Gessner was one of the first to try and classify human speech in its great diversity. He did so in his Mithridates of 1555, the first ever language catalogue, in which the term dialectus frequently appeared.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi



5

The word served to bring more nuance into the relationships between speech forms and is, unlike in antiquity and the Middle Ages, clearly taken to be a variety of a language. In this regard, Gessner was inspired not only by ancient sources but also by the works of his contemporaries. Indeed, unlike Roger Bacon, the Swiss humanist was not an isolated pioneer, but rather the exponent of an early sixteenth-century trend. The emergence of this trend is traced in Chapter 5 through its main symptoms. Most tellingly, the first decades of the Cinquecento witnessed an increasing contrasting of the Latin terms dialectus and lingua, after Greek diálektos had been definitively borrowed into Neo-Latin. Another symptom was the creation of the concept of common language, which in its terminological guise as lingua communis was often opposed to dialectus. In addition, dialectus definitions were silently updated by means of determiners suggesting the hierarchical subordination of the term vis-à-vis lingua. In the 1540s, a new Latin phrase moreover appeared, ‘to differ only in dialect(s)’, which implied that related dialects differed from each other only superficially, whereas distinct languages exhibited substantial variation. This collocation enjoyed a rich career. Dialectus and lingua were, however, not contrasted in a very clear fashion from the start, as especially the former term was not yet delimited semantically. Instead, it was part of an intricate conceptual web, including also concepts such as idiom and style. Chapter 6 tries to frame the emergence of the language/dialect distinction in its intellectual and historical context. The rediscovery of the Ancient Greek dialects constituted, I argue, a major pivoting point, as well as the standardization projects that were gathering steam in the early sixteenth century and the information explosion enhanced by the commercialization of the printing press. Attention to the standardization of the vernaculars was especially urgent in Protestant areas, where a uniform language was needed to convey the Word of God to all classes of society, while in Catholic Europe Latin remained the principal language of religion. In the wake of standardization and categorization attempts, linguistic diversity became countable, as boundaries became more fixed than before. Paradoxically, humanists did not realize that they were introducing a new conceptual pair, assuming, instead, that this had been a Greek achievement. Yet they were, in fact, reading the language/dialect distinction into the ancient source texts. As a result, the new contrast was a product of subconscious appropriation and adaptation. Part III, ‘Consolidation by elaboration’, describes how the new conceptual pair became anchored in the metalinguistic apparatus of early modern scholars in the century after 1550 and received various meanings, most of which are still in use today. In Chapter 7, two interpretations with roots in Greek antiquity take centre stage. First, I outline how the spatial conception of dialect established itself after 1550. More circumstantial evidence for the wide dissemination of this geographical interpretation is also briefly treated. Humanists moreover soon recognized the universality of the phenomenon of regional linguistic variation, an intellectual

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

6



achievement that has not yet received due acknowledgement. Even though the geographical interpretation of the language/dialect distinction implied that language covered a larger area than dialect, some early modern scholars believed that good language had its seat only in a state’s capital. Secondly, I treat the emergence of the early modern idea that dialect was particular to a tribe, implying that language was a kind of ethnically overarching phenomenon. There was, however, an unresolved tension with a competing view, associating language with the nation in the political sense. Although widespread in the early modern period, the spatial and ethnic conceptions of dialect as a variety of a language were never used as diagnostic criteria to determine the language/ dialect status of a speech form. Chapter 8 treats two interpretations of the conceptual pair which early modern scholars did consider useful in doing so: the Aristotelian criterion and mutual intelligibility. I argue that in the seventeenth century an interpretation of the language/dialect distinction emerged according to which related dialects showed only accidental differences, whereas distinct languages varied in their substance. This analysis was grounded in two traditional categories of Aristotelian ontology and implied a binary opposition between substantial and accidental variation; numerous humanists realized, however, that linguistic distance comes in degrees. In the remainder of Chapter 8, I contend that the criterion of mutual intelligibility, often bracketed together with the Aristotelian criterion, had its origin in the Renaissance. The eccentric humanist Johannes Goropius Becanus played a key role in its emergence and specified it further as immediate mutual intelligibility. This criterion, too, was conceptualized in predominantly binary terms. Contradictorily, numerous scholars complained about the lack of mutual intelligibility among speakers of related dialects, pointing out the wider communicative reach of a language. From the more or less neutral early modern conceptions of the language/dialect distinction in Chapters 7 and 8, I move to more subjective interpretations in Chapter 9. From the late sixteenth century onwards, dialect was conceptualized as an anomalous deviation from the analogical language under which it resided. I suggest that this interpretation may have had its roots in Greek ideas on the relationship between the Koine and the other dialects. The analogy/anomaly conception was, however, principally grounded in early modern linguistic realities, since the advancing standardization led to a stronger contrast between the prescribed norm and everything deviating from it. Out of this normative interpretation of the language/dialect distinction, the highly subjective idea that language was superior to dialect developed almost naturally in the course of the seventeenth century. Whereas theorizing on dialect had remained neutral in the sixteenth century because of the close link it had with the esteemed Ancient Greek dialects, the dialect concept was detached from them during the seventeenth century. In addition, local elites gradually turned their back on their native dialects,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi



7

embracing the upcoming standard languages. This social evolution likewise enhanced the severe degradation of the dialect concept. Most interpretations launched by early modern scholars were synchronic; they concerned a snapshot in time only. Yet one take on the language/dialect distinction was diachronic, in that language was considered to generate different dialects. The origin and diffusion of this historical conception is the subject of Chapter 10. I argue that even though it had earlier precursors, it was only due to the influence of Joseph Justus Scaliger’s work that this new interpretation became popular. Already in the early seventeenth century, this conception of the language/dialect distinction was moreover framed by Abraham Mylius within a cyclical process of language change. The language-historical interpretation of the conceptual pair, otherwise primarily understood in synchronic terms, soon prompted criticism, voiced most fiercely by the seventeenth-century orientalist Johann Heinrich Hottinger. I also comment briefly on the emergent idea that dialects preserved archaic features, which apparently contradicted the diachronic conception of the language/dialect distinction. Finally, I outline how the conceptual pair came to be used as a handy discursive strategy for historical classifications of language, especially in cases where evidence was lacking. Chapter 11 gives an overview of how different definitions of the conceptual pair emerged and became rooted in the thought of scholars. It does so, first and foremost, by surveying the seven major interpretations of the language/dialect distinction which originated roughly in the century between 1550 and 1650 and the ways in which they interacted with each other. These conceptions were, I argue, shaped by three main related circumstances: the Greek heritage, scholarly interests, and sociolinguistic realities. I moreover develop the idea that in the period 1550–1650 the conceptual pair underwent consolidation not only by elaboration but also as an emancipation from the Greek heritage. This deHellenization resulted in the degradation of dialect as secondary to language on nearly all levels, even though a strictly political interpretation of the distinction was only marginally present in the early modern period. In Chapter 12, the last chapter of Part III, I present the third case study of the book, the Swedish scholar Georg Stiernhielm. Driven by language-historical interests, Stiernhielm defined the conceptual pair primarily in terms of substantial versus accidental differences. This Aristotelian interpretation he made very explicit, tying it to specific linguistic domains such as the lexicon and pronunciation. He moreover invoked mutual intelligibility in his definitions. As he was concerned in the first place with language history, his usage of the terms lingua and dialectus was also steeped in the diachronic interpretation. More in the margin, the analogy/anomaly opposition as well as geography likewise shaped his conception of the distinction. The case of Stiernhielm, who probably did not know much Greek, confirms the emancipatory evolution outlined in the previous chapter. It is, finally, no coincidence that his interest in the conceptual pair surfaced around

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

8



1650, immediately after he had met at the court of Queen Christina two erudite scholars who had also reflected on the subject along similar lines: Christian Ravis and Claude de Saumaise. Stiernhielm’s extensive attention to the conceptual pair and his linking it to linguistic features reveal a tendency towards systematization and rationalization. Indeed, the Swedish philologist may well be regarded as a transitional figure, heralding a new phase in the history of the language/dialect distinction, treated in part IV. This next period coincides roughly with the years 1650–1800, the age of rationalism and the Enlightenment, when the loose reflections of earlier times were replaced by more structured discussions, and language scholars started to involve more strictly linguistic elements in their treatments of the conceptual pair, even though language-external factors were never far away. I first illustrate this tendency towards systematization and rationalization by means of two short case studies, before moving on to more general reflections. Chapter 13 outlines how the eighteenth-century Dutch orientalist Albert Schultens repeatedly defined the term dialectus in a highly systematic fashion. Schultens analysed the conceptual pair principally in Aristotelian terms but tied it also to geographical factors and framed it in a language-historical scheme. He moreover contrasted the analogy of language to the anomaly of dialect. The Dutch orientalist extended the distinction so as to include a third concept, that of degenerate offshoot, which, unlike a dialect, did not preserve the core of the language intact. He also insisted on the linguistic classes in which related dialects allegedly differed from each other. Schultens was a key figure, as he put the conceptualization of dialect on the scholarly agenda, albeit always as a matter of instrumental importance only, and triggered numerous follow-up discussions among his pupils and readers. Chapter 14 presents a short case study from an angle different from that of philology, taking the historian Johann Christoph Gatterer’s ideas on linguistic diversity as its object. It not only serves as another telling example of the tendency towards systematization but also, and especially, represents a climax in eighteenth-century attempts at rationalizing the conceptual pair. Proposing an embryonic lexicostatistic method, Gatterer tried to find an objective way to use linguistic data in writing an encompassing history of tribes and nations, in particular their prehistory. Starting from a basic vocabulary set, Gatterer attempted to quantify linguistic distance. In doing so, he divided the kinship continuum into four sections: unrelated languages, related languages, dialects, and closely related dialects. His innovative methodology, prefiguring modern lexicostatistic approaches, had only limited success, however. Gatterer failed to put it into practice, and the historian was criticized, rather ironically, for his ahistorical method by the grammarian Johann Christoph Adelung. In the last two chapters of Part IV, I adopt a broader perspective again. Chapter 15 outlines the linguistic respects in which related dialects were believed to vary, in opposition to distinct languages, during the early modern era. Initially,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi



9

a Byzantine author was the main source of inspiration to describe dialect-level differences, primarily in Greek handbooks. It was only after 1650 that the levels of variation were treated in a more systematic fashion by non-Hellenists, too. The focus of attention was on the ways in which related dialects varied. The differences were, as most scholars agreed, superficial and mainly situated on the level of pronunciation, letters, and the lexicon. There was, however, no linguistic domain in which related dialects were claimed to never demonstrate variation. Overall, the differences required to speak of distinct languages attracted less attention. Yet many scholars agreed that substantial differences were needed, principally in the roots of words. Sometimes, unusual linguistic criteria were put forward, for instance by Johann Georg Wachter and Ferdinando Galiani in the eighteenth century. Chapter 16 discusses further evidence for the systematization and rationalization of the language/dialect distinction in the period 1650–1800. On the one hand, a kind of dialectological tradition emerged. The study of regional variation became a subfield of philology, albeit never an autonomous one; occasionally, it already was receiving the label of dialectologia, introduced in 1650. For the first time, philologists presented dissertations on dialectal diversity that were no longer exclusively focused on the Greek dialects. On the other hand, scholars adopted more rational attitudes towards the conceptual pair. Some chose to supplement the binary contrast with new concepts. Others advocated to distinguish more clearly among different interpretations of the language/dialect distinction. Confusion persisted, however, throughout the early modern period. The first vocal sceptic of the conceptual pair was Friedrich Carl Fulda, who made painfully clear how arbitrary the distinction actually was. In Part V, I explore the vicissitudes of the conceptual pair in mainstream modern linguistics, with specific attention to their relationship to earlier ideas, both in terms of continuity and breaking points. Chapter 17 surveys the nineteenthcentury fate of the language/dialect distinction, during and after the establishment of linguistics as an autonomous field of study. As at this stage the study of language roughly coincided with historical-comparative grammar, it is no surprise that the language-historical interpretation prevailed. In the 1870s, the spatial conception was foregrounded when dialect geography emerged. Other criteria, too, lived on, especially mutual intelligibility and the Aristotelian interpretation. In the first half of the nineteenth century, only few scholars bothered to discuss the conceptual pair at length, with the exception of the forgotten comparative linguist and Hellenist Albert Giese. In the 1860s, suspicion about the validity of the language/dialect contrast grew, perceptible in the work of August Schleicher and the neogrammarians, but not to such an extent that it was extensively questioned. Perhaps linguists were afraid that this debate would mean the premature end of their young discipline. By the end of the nineteenth century, many prominent linguists had realized the arbitrariness of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

10



the language/dialect distinction, and linguists were forcibly confronted with its problematic nature from within, as Chapter 18 argues. This last case study of the book outlines the ideas of the headstrong German linguist Hugo Schuchardt, who dared to face the full consequences of critically reflecting on the conceptual pair. Considering the nature of linguistic diversity, Schuchardt noticed that there were no such strict divisions as Schleicher’s family tree model presupposed. Instead, he proposed a wave model, thus prefiguring Johannes Schmidt’s image. Language and dialect were merely fictional abstractions from actual linguistic facts and for this reason useless. Schuchardt was not alone in his suspicion. Some of his colleagues likewise acknowledged the abstract nature of the language/dialect distinction, but preferred to keep on using it for practical reasons, whereas others like Jules Gilliéron dismissed its validity and concentrated on linguistic features instead, realizing that dialect boundaries were arbitrary. Chapter 19 surveys the fate of the language/dialect distinction in structuralism. Ferdinand de Saussure’s conception of it is revealed to have been fairly traditional. In Saussure’s wake, mainstream structural linguists usually focused on homogeneous language systems, the langue, rather than the parole, with scant attention to the conceptual pair. In the 1950s, a dialectological turn occurred. The year 1954 in particular was a breaking point, when three structuralist papers devoted to the concept of dialect appeared. Uriel Weinreich suggested the concept of diasystem to capture variation within one language. André Martinet, in turn, tried to redefine dialect scientifically by excluding sociopolitical factors. Václav Polák, finally, argued that substantial morphosyntactic variation was required to speak of distinct languages. Phonological and lexical differences resulted in dialects only. Structuralist discussions of the language/dialect pair remained uncoordinated, however, and had relatively limited impact on subsequent debates, except for Weinreich’s diasystem concept. The brief Chapter 20 treats the success of the criterion of mutual intelligibility since the 1950s, when American linguists interested in Amerindian tongues started to actively test this feature. Pioneers were Carl Voegelin and Zellig Harris, who suggested four methods of answering language/dialect questions, including mutual intelligibility testing. Even though scholars immediately faced numerous problems, the method enjoyed considerable success and is the primary criterion used by language catalogues such as Ethnologue and Glottolog. The criterion was criticized by, among others, Frederick Agard, who proposed nine postulates for determining language/dialect status. Others followed Morris Swadesh’s lexicostatistic lead and tried to quantify the distance between speech forms. According to one of the most recent representatives in this strand, linguistic distance is bimodally distributed, and the language/dialect distinction is, by consequence, backed by the majority of linguistic evidence. Rather artificially, a strict cut-off point is proposed between the language and dialect poles.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi



11

Chapter 21 confronts two opposite perspectives on language and the language/ dialect pair with roots in the 1950s and 1960s. On the one hand, it treats generative approaches to linguistic diversity. In general, generative linguists have assumed that dialect-level variation is produced by minor differences in rules, parameters, or constraints and their ordering or ranking, depending on the generative framework which they follow, whereas distinct languages are characterized by major divergences in the same. On the other hand, sociolinguists have focused on linguistic variables rather than systems. As they correlate linguistic phenomena with language-external attributes, their conceptions of the language/ dialect distinction tend to be rather hybrid, being shaped by linguistic as well as sociopolitical parameters. This externalist approach has been fiercely criticized by Noam Chomsky, the founding father of generativism. Other linguists have adopted more constructive attitudes, either by supplementing the language/dialect distinction or by supplanting it with an entirely new conception of language. Chapter 22 develops the idea that even though many linguists today tend to avoid the term dialect, it has been a major resource for new terminology, especially through -lect derivations. The recent terms doculect and languoid are particularly interesting as they are derived from (dia)lect and language, respectively, and are part of an alternative conception of language. Terminology linked to the language/ dialect distinction is still widely used and has been incorporated into conceptual constellations tailored to specific national contexts. It is moreover striking how noncommittal recent definitions of dialect are, as if linguists want to keep a safe distance from this tarnished term. The final section draws the balance of the modern love-hate relationship with the conceptual pair. Five main attitudes can be distinguished, ranging from tacit preservation through explicit redefinition and supplementation to a quest for alternatives and downright dismissal. The unorganized nature of the debate on the subject is remarkable, as is the persistence of centuries-old interpretations. Even though the book’s focus is on scholarly conceptions of the language/ dialect pair, first within grammar and philology and later within linguistics, it is indispensable to dwell briefly on popular ideas, too, since they have interfered and interacted with the views of linguists in various intriguing ways. Chapter 23 explores the modern popularization of the conceptual distinction, gathering steam in the last decades of the nineteenth century, when the standard language ideologies of nation-states spread to nearly all classes of society, primarily through mass education. This popularization involved a sharp divide between standard language and dialect, which became widely known among the general public. In the process, the conceptual pair was excessively politicized and even politically activated, as demonstrated by some notorious episodes in recent history, including the Oakland Ebonics controversy. The Internet has, moreover, provided laypeople with a forum in which to debate the language/dialect distinction and its applications, which tend to have a strong political colour. It remains to be seen whether

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

12



and how linguists and others, both laypeople and academics, have exchanged ideas on the subject. I tentatively suggest that there indeed has been a conceptual cross-fertilization, a phenomenon requiring further investigation. Chapter 24, finally, surveys by way of conclusion the book’s main arguments. These include especially the emergence of the modern language/dialect distinction during the early sixteenth century and the subsequent formulation of its main interpretations. Above all, however, this chapter emphasizes that the conceptual pair unmistakably has a history, for too long neglected, and that it is not a selfevident given which has always been there. Having established the historicity of the language/dialect distinction, I field the question of whether it has a future, to which I try to offer an answer, both tentative and brief, from my perspective as a historian of language studies. On the one hand, I suggest that a reconceptualization of the distinction can be a viable option. On the other hand, the fact that the conceptual pair has become common knowledge in modern times gives linguists, I believe, not only the opportunity but also, and especially, the responsibility to take on a more prominent societal role in language/dialect disputes.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

I

P R E H I S T O R Y , 5 0 0   – 1500

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

2 A dive into the prehistory of the conceptual pair As the Roman emperor Tiberius (42 – 37) was very fond of Greek rhetoric, he allowed several Greeks to be part of his entourage and live with him. If we are to believe the historian Suetonius (c.  70–after  128), however, they were subject to his capriciousness as much as anyone else. One of Tiberius’s favourite Greeks, going by the name Zeno, for instance, suffered a tragic fate after displeasing the emperor. The Greek had uttered some phrases in an affected fashion, which bothered Tiberius, who asked: ‘What is that so very annoying dialect?’ Zeno answered that it was Doric, which caused the emperor to promptly banish him to the Greek island of Cinaria, present-day Kinaros in the Dodecanese. Zeno had apparently reminded Tiberius of the miserable period in his life when he lived in retirement on the Doric island of Rhodes, before he was crowned emperor.¹ For this reason, he wanted him gone. Suetonius inserted this episode into his biography of Tiberius as he wanted to demonstrate the ruthless character of the Roman emperor. Indirectly, it indicates something else, too. It was widely known in the intellectual circles of antiquity, even in the Latin-speaking western sphere of the Roman Empire, that there were different Greek dialects. Indeed, they absorbed nearly all attention turned to the topic of variation within a language before about 1500, since they were the only dialects relevant for literary and philological study. Greek literature was composed not in one uniform language but in a range of different dialects (see e.g. Colvin 2010). Homer’s masterpieces were sung in a dialectally mixed Kunstsprache. The great orators from ancient Athens drew up their speeches in Attic, the classical form of Greek par excellence. Pastoral poetry was marked by Doric dialect, whereas Sappho chose her native Aeolic speech in her touching lyrical songs. Herodotus, the proclaimed father of historiography, wrote his histories in Ionic.

¹ Suetonius, De vita Caesarum, Tiberius 56.1: ‘Nihilo lenior in convictores Graeculos, quibus vel maxime adquiescebat, Xenonem quendam exquisitius sermocinantem cum interrogasset, quaenam illa tam molesta dialectos esset, et ille respondisset Doridem, relegavit Cinariam, existimans exprobratum sibi veterem secessum, quod Dorice Rhodii loquantur’.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

16

     

2.1 The Greek dialects between anecdotes and definitions Suetonius’s anecdote about Tiberius and Zeno is a fairly banal instance of an ancient author referring to the Greek dialects.² As such, it represents a rather low level of awareness of dialectal variation. Several centuries before Suetonius, the Greek historian Herodotus (c. 485–424 ) was more conscious of Greek diversification, as he was in a position to distinguish four different varieties of Ionic Greek in Asia Minor, part of present-day Turkey. He refers to them as trópous tésseras paragōgéōn, ‘four manners of deviations,’, and kharaktêres glo¯ś sēs tésseres, ‘four distinctions of tongue’.² Herodotus’s wording indicates that he could not yet rely on an established conceptual apparatus and a corresponding metalanguage to talk about dialectal diversity. Instead, he had to resort to words lacking an obvious semantic link with language at that time, such as kharakte¯ŕ (χαρακτήρ), ‘character(istic); distinctive mark; stamp’, trópos (τρόπος), ‘way; manner’, and paragōge¯ ́ (παραγωγή), which for Herodotus apparently meant something like ‘deviation’, ‘twisting’, or ‘seduction’. Paragōge¯ ́ did later become a metalinguistic term meaning ‘derived form’ and ‘inflection’, whereas the root kharak- featured prominently in later Greek definitions of the term diálektos.³ Some generations later, the Athenian general and notoriously difficult historiographer Thucydides (second half of the fifth century ) went a little further still by trying to make sense of the different Greek dialects and to characterize their interrelationships. He reported, for instance, on a case of dialect mixture on Sicily (Historiae 6.5.1). The inhabitants of the city of Himera spoke, Thucydides claimed, a variety that occupied a middle ground between Chalcidian Ionic and Doric, an assertion for which there is, by the way, no historical evidence (Vassallo 2005: 89). The Athenian historian also mentioned that the Greek spoken by the Aetolians was not understood by other Greeks (Historiae 3.94.5), thus apparently proving that not all varieties of Greek were mutually intelligible. Another type of awareness of dialectal variation came to the fore in its dramatic and literary activation. The brilliant playwright Aristophanes (c. 450–c. 385 ), for instance, put specific dialects to comic use at many different occasions, as proved by the masterful analysis of Stephen Colvin (1999). Connected to this was the assumption that a dialect conveyed a specific impression to a listener or reader, which surfaced in other genres, too. For example, the great philosopher Plato (428/7–348/7 ) apparently experienced the Lacedaemonians, Greeks of Doric descent, as brakhúlogos (βραχύλογος), ‘short in speech’ (Leges 641e). A feature frequently attributed to Dorians in general was their alleged ‘broadness’ in speaking

² Historiae 1.142: ‘τρόπους τέσσερας παραγωγέων’ & ‘χαρακτῆρες γλώσσης τέσσερες.’ Cf. Van Rooy (2016d: 247–8). ³ For paragōge¯ ́ see Liddell & Scott (1996: s.v. παραγωγή); Dickey (2007: 251). For the kharak-root see Van Rooy (2016b: 60–1).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    . . .

17

(e.g. Theocritus, Idyllia 15.87–8). These views are among the earliest explicit formulations of the impression which a dialect conveys or, in modern linguistic terminology, of dialect attitudes and perceptions (cf. Edwards 2009: 82–97). The rhetorician and lexicographer Julius Pollux (second half of the second century ) elaborated in more general terms on the variability of language by pointing out that differences in tongue existed from city to city (Onomasticon 2.110). Before him, the geographer Strabo (c. 62 –c.  24) had made a similar suggestion (Geographica 8.1.2). Ancient Greek authors undertook no real attempts at explaining variation in their native language, even though Diogenes Laertius (mid-third century ), author of a work on the lives and thought of various philosophers, suggested colonization as a force behind linguistic change (Vitae philosophorum 1.51). In the past few paragraphs, I have covered the spectrum from vague awareness of dialectal variation to a preliminary search for the causes of the phenomenon. With Diogenes Laertius, I can even go one step further and reach a still higher degree of abstraction. For in his work we find one of the earliest definitions of the key term diálektos, which he quoted from a now lost work by Diogenes of Babylon (c. 240–150 ). This Stoic philosopher must have explained the word as follows: Dialect is speech stamped tribally and in Greek fashion, or speech of a certain region, that is, having a certain quality according to a dialect, as thálatta [θάλαττα, ‘sea’] in the Attic and hēmérē [ἡμέρη, ‘day’] in the Ionic.⁴

Diálektos was conceived by Diogenes as a technical term to designate features of the Greek language that bore the mark of a certain tribe or region. The philosopher expressed this by means of the kharak-root, as Herodotus had done several centuries earlier. The verb kharássō (χαράσσω) meant, among other things, ‘to carve; to engrave; to stamp’, specifically with reference to coins. By using this verb, Diogenes probably wanted to emphasize the distinctive character of a diálektos, and how a tribe and its members could leave their own mark on features of the Greek language. He provided specific examples, too. Thálatta, as opposed to thálassa with double sigma, was a Greek word bearing the stamp of the Attic tribe and the region of Attica, whereas hēmérē, as opposed to hēmérā (ἡμέρα), was typical of the Ionic dialect. Even though Diogenes of Babylon linked diálektos with regional variation, he did not presuppose a clear-cut distinction between dialect and language. Instead, the term was understood as léxis (λέξις), which Diogenes explained very generally

⁴ Diogenes Laertius, Vitae philosophorum 7.56: ‘Διάλεκτος δέ ἐστι λέξις κεχαραγμένη ἐθνικῶς τε καὶ Ἑλληνικῶς, ἢ λέξις ποταπή, τουτέστι ποιὰ κατὰ διάλεκτον, οἷον κατὰ μὲν τὴν Ἀτθίδα θάλαττα, κατὰ δὲ τὴν Ἰάδα ἡμέρη’ (translation adapted from Van Rooy 2016d: 250).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

18

     

as ‘speech that is expressible by means of letters’, phōne¯ ̀ eggrámmatos (ϕωνὴ ἐγγράμματος) in Greek (Diogenes Laertius, Vitae philosophorum 7.56). Accounts such as Diogenes’s, which testify to a larger degree of abstraction than shown by previous observations, will constitute the central thread running through this chapter, as they are indispensable for an adequate understanding of numerous aspects of later theorizing on the language/dialect distinction.

2.2 Diálektos, a variety of interpretations Was Diogenes of Babylon alone in trying to define diálektos? Or was he part of a tradition? The scant evidence available suggests that the latter is true. Indeed, numerous Greek scholars tried to define diálektos as a workable metalinguistic term, albeit in different meanings (Van Rooy 2016b, 2016d). Aristotle’s (384–322 ) early interpretation of diálektos as ‘articulation of the voice by means of the tongue’ is miles away from later definitions focused on the literary varieties of Greek.⁵ Yet even in works devoted to the Ancient Greek dialects, there was considerable disagreement. The term diálektos had an important generic meaning referring to language, ‘manner of speaking; means of communication’, deriving from the Greek middle verb dialégomai (διαλέγομαι), ‘to converse with’. With this generic meaning, the ́ widespread definition idíōma glo¯ssēs or glo¯t́ tēs (ἰδίωμα γλώσσης/γλώττης), found in several philological and grammatical works on Greek, shows affinity. In my view, the definition must have signified ‘particularity of tongue’ rather than ‘particular form of a language’ (pace Consani 2000: 612–13). The difference between the two interpretations lies in the presence or absence of conceptual subsumption. I will make use of the terms subsumption and to subsume in this book to refer to a key development in the history of the conceptual pair language/ dialect: the emergence of the idea that there is a level of linguistic entities, dialects, which fall under the hierarchically higher category of languages. In other words, at a certain point in history, the concept of dialect was subsumed under that of language, an issue which I will address at length in Part II. According to my interpretation, then, the widespread Greek definition idíōma ́ glo¯ssēs meant ‘particularity of tongue’, in which case diálektos was not subsumed under a superordinate language concept. Instead, it was viewed as a language particularity in a very general sense, as an ensemble of language features that did not have a wide application, but were restricted to a certain context of usage. Doric, for instance, was a diálektos because it had long alphas where one would normally expect etas, because it was particular in contracting the vowel sequence

⁵ Historia animalium 535a: ‘διάλεκτος δ’ ἡ τῆς ϕωνῆς ἐστι τῇ γλώττῃ διάρθρωσις’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. ´ ,    

19

epsilon-omicron as epsilon-upsilon , etc. Indeed, nowhere in extant Greek texts are there any indications that the Greeks understood the definition ́ idíōma glo¯ssēs as referring to a ‘particular variety of the Greek language’, and, by consequence, it is impossible to attribute a subsumption of dialect under language to Ancient Greek and Byzantine thought. Nonetheless, the definition idíōma ́ glo¯ssēs was, just like the generic meaning of ‘manner of speaking’, intrinsically linked to linguistic diversity. After all, speaking about a ‘manner of speaking’ or a ‘particularity of tongue’ presupposed the existence of other ‘manners of speaking’ and ‘particularities of tongue’. The definition idíōma glo¯ś sēs is, for example, found in the influential works of John the Grammarian, to be identified with the early Byzantine philologist and theologian John Philoponus (?c.  490–575), and the Byzantine polymath Gregory of Corinth (eleventh / twelfth centuries), also known as Gregorius Pardus. Several ancient and Byzantine scholars further specified the semantic core of the word diálektos in their definitions of the term. What parameters did they invoke? Greek thinkers laid down three principal specifications. First and foremost, a diálektos was interpreted as speech particular to a certain Greek éthnos (ἔθνος), ‘tribe’. This parameter already appeared in the early definition by Diogenes of Babylon, quoted in the previous section. We also encounter it in the work of, among others, the Early Christian author Clement of Alexandria (c.  140/50–before 215/21) and Gregory of Corinth.⁶ These scholars consistently relied on the kharak-root to make clear that a diálektos always bore the ‘stamp’ of a Greek tribe. Their frequent recurrence to this root seems to point to the existence of a continuous tradition, starting with Diogenes at the latest and having a precursor in Herodotus’s usage of the term kharakte¯ ŕ . Clement of Alexandria is an exception in that he did not restrict éthnos to the Greek tribes but extrapolated it to Hebrew, the language of the Old Testament, too (Stromata 6.15.129.2). This broad use suggests once again that diálektos did not figure in a hierarchic constellation in which it was subsumed under a term like glôssa or glôtta (γλῶσσα/γλῶττα), meaning ‘language’. Instead, it could refer to both Hebrew and Ionic, speech forms we would be tempted to call from our modern perspective language and dialect, respectively. Second, a diálektos was speech particular to a certain tópos (τόπος), ‘place’. This specification was indisputably present in ancient definitions but seems to have lost its importance in medieval times. It was, for instance, absent from important Byzantine reference works such as the Suda. What is more, it did not figure in John the Grammarian’s and Gregory of Corinth’s definitions, in which the Greek word tópos, ‘place’, came to be strangely replaced by túpos (τύπος), ‘type, form’.

⁶ Clement of Alexandria, Stromata 1.21.142.3 & 6.15.129.2; Gregory of Corinth, De dialectis 1.1.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

20

     

This change was probably no coincidence and might reflect an inability to tie the Greek dialects to geographical locations, since the dialects had been extinct for quite some time when these authors were writing down, or rather incorrectly copying, their definitions (cf. Dickey 2007: 75; Tribulato 2014: 458–9). Still, in the early Byzantine period, John the Grammarian did regard the geographical dispersion of Greek tribes as a force behind dialectal diversification (Van Rooy 2016d: 264–5). A diálektos, then, was for many Greek scholars ‘speech of a particular form, tied to a specific Greek tribe and place’. Oddly enough, the Greek Koine was usually also classified as one of the main dialects, even though this variety could not be linked to a distinct tribe or region. For the Koine, short for hē koine¯ ̀ diálektos (ἡ κοινὴ διάλεκτος), ‘the common way of speaking’, was a supraregional form of Greek which in the wake of Alexander the Great’s conquests spread across the eastern Mediterranean and beyond, where it served as a lingua franca. This incompatibility remained, however, unresolved in Greek thought, even though there are traces of a debate on the exact position of the Koine vis-à-vis the other dialects (Van Rooy 2016d: 257–8). Thirdly, Greek scholars seem to have regarded a diálektos as written speech particular to an author or a group of authors. They did not take the Greek spoken by the man in the street into consideration, as it was all too susceptible to change and probably difficult to write down in all of its phonic particularities. On the contrary, grammarians concentrated on speech that was expressible by means of letters—Diogenes of Babylon’s phōne¯ ̀ eggrámmatos—and on the canonical authors, who happened to compose in different forms of Greek but at the same time avoided features of too regional a colour. It was this coincidence of Greek literary history that made the Greek dialects relevant to philologists (e.g. Mickey 1981; Morpurgo Davies 1987: 10–11). The literary status of the Greek dialects entailed a close association of the word diálektos with written codification and made it also applicable to other languages with a well-established written tradition, such as Latin, Egyptian, and Hebrew. The playwright Aristophanes and the Byzantine archbishop Michael Choniates (c. 1138–c. 1222) were exceptional in that they used the term diálektos to refer to varieties spoken by lower social classes (Van Rooy 2016d: 258–9, 266–7). Consequently, the word clearly did not have a negative connotation, contrary to many later uses of the English term dialect and its equivalents in other languages. The lexicographer of rare words Hesychius (?fifth/sixth centuries ), for instance, needed the adjective parátonos (παράτονος), ‘ill-sounding’, to add a pejorative meaning to his explanation of barbarismós (βαρβαρισμός), ‘barbarism’, as ‘parátonos diálektos’ (‘παράτονος διάλεκτος’ at Lexicon .210). As a result, the word was often absent from contexts in which we would today intuitively expect it to be present. The Greek geographer Strabo, for instance, described variation in the tongues of Gaul by stating that the Belgians and the Celts ‘do not all speak the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

21

same language, but some of them slightly diverge in their tongues’.⁷ The fact that Strabo was discussing here non-Greek, unwritten tongues might have precluded the use of the word diálektos. Also, the term was not yet interpreted as a variety of a language differing only superficially from other varieties of the same language, as it later was.

2.3 Philology as stimulus Greek scholars proposed various definitions of diálektos, but why did they care to define the term at all? The Greek dialects were relevant only in so far as they were the forms in which the great literary works of Homer and others were composed. It is consequently impossible to identify professional Greek dialectologists. No one devoted his efforts to studying the Greek dialects in and of themselves. Furthermore, the word dialectology, although entirely composed of Greek lexemes, was only coined in the mid-seventeenth century (see Chapter 16, Section 16.1). Still, it remains a fact that treatises devoted to the particularities of the literary dialects, mainly Attic, Ionic, Doric, and Aeolic, and sometimes also including the Koine, were composed at an early stage. In these works, every dialect was discussed separately, and the letter mutations occurring in each of them took centre stage. For instance, to use an example already cited, most treatises pointed out that Attic had double tau where other dialects had double sigma, as in thálassa versus thálatta, ‘sea’. These changes were conceived as ‘modifications of the word’, páthē tês léxeōs (πάθη τῆς λέξεως) in Greek, an approach encompassing several different letter change operations and usually called ‘pathology’ by modern scholars (e.g. Wackernagel 1876; Siebenborn 1976: 150; Ax 1987; Lallot 1995). Early works in this philological tradition by the grammarians Tryphon (second half of the first century ), Apollonius Dyscolus (first half of the second century), and others are lost to the ages. The first entirely extant writing in this tradition might be pseudo-Plutarch’s Life of Homer. In a part of this work, the Greek dialects as they appear in Homer’s epic poems are briefly described. This section came to be extracted from the treatise, most likely in Renaissance Italy, and often received the title On the dialects in Homer. The treatise may date from the Roman period, but this issue is complicated by the fact that it has received later additions (Van Rooy 2018c). Several other but clearly related treatises survive, most of which have been collected by Gottfried Heinrich Schäfer (1764–1840) in the early nineteenth century (Schäfer 1811). One treatise, John the Grammarian’s Tekhniká (Τεχνικά), which can be translated roughly as Grammatical issues, deserves special attention here. Probably

⁷ Geographica 4.1.1: ‘ὁμογλώττους δ᾽ οὐ πάντας, ἀλλ᾽ ἐνίους μικρὸν παραλλάττοντας ταῖς γλώτταις’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

22

     

based on Alexandrine sources and extant in two different redactions, the work exhibits two features of interest as it stands.⁸ On the one hand, the second redaction elaborates on the three linguistic levels on which the Greek dialects allegedly exhibited differences: of entire words, of parts of words, and of word accidents, relating to matters of accent and other diacritical marks. The focus was, however, clearly on the second category of variation, which primarily covered letter mutations. John the Grammarian’s account influenced many early modern discussions of the levels on which related dialects varied, as I will argue in a later chapter (Chapter 15, Section 15.1). On the other hand, the link with philology and the study of literature surfaces very clearly when at the very end of the first redaction, individual authors—exclusively of pagan stock—are connected to specific dialects.⁹ Such associations could be incorrect, as is clear from the fact that the Greek poet Pindar (?522/18–after 446 ) is wrongly said to have written in the Koine, even though the Koine did not yet exist at the time of his writing and his poems have a primarily Doric look (Morpurgo Davies 1987: 18). Philological motives also informed classifications of the Greek dialects (Van Rooy 2016a; 2020a). If Attic and Ionic are today believed to constitute one dialect branch of Greek, ancient and Byzantine scholars separated them based on their distinct literary usages. The position of the Koine vis-à-vis the four literary dialects, moreover, remained a contested issue throughout antiquity and the Byzantine era, as I have mentioned earlier. This struggle with the status of the Koine inhibited the wide acceptance of the idea that the Koine was the normative variety from which the dialects were deviations. As a matter of fact, a standard language concept was absent from Greek theorizing, which does not mean, however, that there did not exist a notion of correct usage among Greek grammarians. There was a tradition of normative linguistic thought separating correct from incorrect forms of Greek (Versteegh 1986; Dickey 2007: 235). The position of the dialects in this correct/incorrect dichotomy nevertheless remained somewhat unclear. Overall, the Greek linguistic ideal of Hellēnismós (Ἑλληνισμός), ‘Greekness’, usually encompassed the canonical literary dialects other than the Koine, too, as James Clackson (2015) has pointed out.

2.4 The Latin West What about the Latin West? How did ancient Roman and Western medieval scholars approach regional linguistic variation? There, the linguistic context was of a very different nature than that in Greece. There was only one language of ⁸ On the work’s sources see Bolognesi (1953: 101–2); Hainsworth (1967: 63–4). ⁹ See John the Grammarian in Manutius et al. (1496: 236-). I refer to the first Renaissance edition because, still, no modern critical edition of this important text is available.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

23

literature and administration: Latin, an Italic tongue that had succeeded in largely wiping out the great linguistic diversity in the western half of the Roman Empire. This situation gave rise to a stricter dichotomy between correct and incorrect language. Ancient authors nonetheless had an eye for regional differences within Latin. To give only one early example, I can mention a passage in one of Plautus’s (c. 250 –after 191 ) comedies. In his Trinummus, the playwright mocked at a phraseological particularity of the city of Praeneste in Latium, present-day Palestrina, some thirty-five kilometres east of Rome. When the slave Stasimus tried to communicate to his master Callicles that his son had betrothed his daughter to a certain Lysiteles, his baffled master asked him when this had happened. Stasimus then replied: ‘just now (tam modo), as the Praenestine says’.¹⁰ The use of tam modo instead of modo bore, in other words, the mark of Praeneste, which Plautus put to comic use. Contrary to what present-day readers might expect, Roman authors did not intuitively compare the Greek literary varieties with regional variation in their native language but linked the Greek dialects with different linguistic registers in Latin instead. The Roman rhetorician Quintilian (c.  35–100) provided a powerful testimony to this association, when discussing so-called Sardism. This fallacy was named after the city of Sardis, situated in modern-day Turkey, which in antiquity was known for its dialectally mixed population. Quintilian explained it as follows: Also, Sardism is the name of a certain speech mixed from a diverging range of tongues, if, for instance, you would confound Doric, Ionic, or even Aeolic words with Attic ones. But we have a similar vice, if someone mixes lofty with lowly words, old with new ones, and poetic with vulgar ones—that is indeed such a monstrosity, as Horace writes in the first part of his book on the art of poetry: “if a painter would want to join a horse’s neck to a human head”—and would place other things of different natures under it.¹¹

Whereas the Greek example referred to regional varieties that had been elevated to literary status, the Latin situation did not pertain to regional linguistic diversity at all but to different stylistic registers, some of which were linked to time-bound and class-based language variation. It can be noted here that the term diálektos did not feature in Quintilian’s comparison, but the orator did use the word when talking

¹⁰ Trinummus 609: ‘ “tam modo” inquit Praenestinus’. ¹¹ Institutio oratoria 8.3.59: ‘Σαρδισμός quoque appellatur quaedam mixta ex varia ratione linguarum oratio, ut si Atticis Dorica, Ionica, Aeolica etiam dicta confundas. Cui simile vitium est apud nos, si quis sublimia humilibus, vetera novis, poetica vulgaribus misceat—id enim tale monstrum, quale Horatius in prima parte libri de arte poetica fingit: Humano capiti cervicem pictor equinam iungere si velit—et cetera ex diversis naturis subiciat’. On the ancient concept of Sardism see Gitner (2018).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

24

     

about the Greek dialects on two different occasions (Institutio oratoria 1.5.29 and 9.4.18). Elsewhere (11.2.50), he referred to the five Greek dialects as Graeci sermonis differentiae, ‘differences of Greek speech’. Quintilian’s choice of words was part of a broader tendency. The Greek word diálektos as well as its Latinized counterpart dialectus were utterly rare in Latin texts before 1500, and where they did appear, they almost always denoted a variety of the Greek language (Van Rooy 2019). What is more, no ancient or medieval Latin author contrasted this term with a word meaning ‘language’. The case of Quintilian suggests that at least some educated ancient Romans were familiar with variation in both Latin and Greek. As I will argue in later chapters (Chapter 3, Section 3.3; Chapter 4, Section 4.1), an awareness of different languages and their internal variation seems to have constituted an important trigger for developing a language/dialect-like distinction. It was, however, by no means a sufficient condition, as ancient Roman authors did not develop a conceptual opposition resembling such a distinction (pace Kamusella 2015: 11–13; 2016: 172–5). Why did this outlook not act as a trigger in the case of the Romans? The answer probably lies in the fact that the Greek dialects were perceived as resembling Latin literary registers rather than Latin regional variation, as Quintilian’s comments on Sardism indicate. There was, moreover, a general lack of interest in the regional and social diversification of Latin, with the bulk of attention being directed to so-called good Latin. What is more, Roman and early medieval authors probably did not view Latin variation as constituting clearly demarcated dialects each with a distinct set of properties, as was the case with the Greek dialects. As a result, they provided only very few generalizations with regard to Latin variation and refrained from offering strict definitions of the metalinguistic terms they used in this context. One might nevertheless note here that in the Greek world, variation other than the one observed among the canonical literary dialects received still less attention than variation in the Latin sphere (Kramer 1989; Müller 2001: 279 n.17). The scholars’ lack of interest in regional language variation did not, of course, mean that they were not familiar with the phenomenon. On the contrary, ancient, medieval, and early Italian Renaissance authors were sufficiently aware that a language could show local particularities, and that two closely related speech forms could be considered to constitute one language. This awareness did not limit itself to Latin. Two well-known examples are the Roman historiographer Tacitus (c.  55–c.  120) and the Early Christian author Jerome (c. 347–419). Tacitus described the speech of the Bretons and Gauls as ‘by no means very different’.¹² Jerome, on the other hand, noted that Galatian, spoken in Anatolia (in present-day Turkey), and Trierisch, spoken in the German city of Trier, were

¹² De vita Iulii Agricolae 11.3: ‘sermo haud multum diversus’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

25

‘almost the same language’, albeit with some corruptions in the former.¹³ Tacitus and Jerome were no doubt right in their observations on these tongues, which all belong to the Celtic family of languages according to modern insights. Gradually, the spectrum of languages in which regional variation was noticed, broadened in the Latin West. In the later Middle Ages, the chronicler Gerald of Wales (1146–1220) was able to remark differences in Welsh and English, and to draw a comparison between the two linguistic contexts. Gerald was, however, still struggling to express in words the two levels of variation he noticed: that of Welsh and English, on the one hand, and that of Welsh and English regional varieties, on the other. The latter he tried to label, rather unsuccessfully, as proprietas idiomatis, a tautological phrase meaning something like ‘property of particularity’ (Gerald of Wales 1868: 177). Much like the ancient Romans, Gerald was aware of different contexts of regional language variation but proved himself unable to develop clear-cut concepts to chart this phenomenon. The difficulty of designing adequate conceptual tools to deal with regional language variation was partly due to the limited reflection on linguistic diversity in the period before 1500. Scholars were more interested in Latin, a wellestablished language with a fixed form. It is telling that only Latin was taken into account by modistic grammarians, an influential school of language philosophers active in thirteenth-century Paris and aiming to uncover the very foundations of grammar.¹⁴ The inability to come up with workable concepts was partly also a consequence of the linguistic realities many medieval writers faced. In a time when the vernacular languages were not yet being upgraded and standardized as full-fledged written tongues, it was often difficult for scholars to perceive clear borders between cognate and neighbouring languages and their varieties. The Germanic tongues in Western Europe, for instance, were believed to constitute one long chain of related tongues. The Italian humanist Enea Silvio Piccolomini (1405–1464), later known as Pope Pius , offered a wonderful example of this point of view. Shortly after 1431, Piccolomini described the wide extension of Germanic tongues in terms of a concatenatio, a linguistic sequence ranging from Austria to Scotland—a dialect chain, a present-day linguist might be tempted to say.¹⁵ Indeed, pre-1500 scholars usually found it far from obvious and often simply irrelevant to distinguish between different types of linguistic diversity. A well-known exception is Dante Alighieri (c. 1265–1321), who composed a treatise On vernacular eloquence around 1305. In this Latin work, the Tuscan poet ¹³ Commentarii in  epistulas Paulinas, Ad Galatas 2.3: ‘lingua eadem paene’. See Denecker (2017: 238, 243–4). ¹⁴ On modistic grammar see e.g. Pinborg (1967); Bursill-Hall (1971); Dahan et al. (1995: 266). ¹⁵ Piccolomini (1960: 350): ‘Scotus vero, quantum ego meis fatis in eam plagam deductus perpendi, non plus ab Anglico quam Australis a Bavaro distat, ut hanc Theutonicam linguam nexu et concatenatione quadam ab Austria usque in Scociam facile productam videas’. See e.g. Borst (1957–63: 969).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

26

     

first distinguished the lingua sì, Italian, from the lingua oïl and the lingua oc, the Romance tongues of northern and southern France, respectively. These tongues were, Dante claimed, initially unitary but had been diversified as an idioma trifarium, a ‘threefold idiom’, by his time (Tavoni 1987: 425–8; 1990; Consani 1991: 159; Trabant 2006: 67–8). He proceeded by dividing the sheer infinite variation in Italian into two main classes, which, in turn, were each characterized by several different linguistic layers. Under these two main classes, there were at least fourteen ‘primary variations’ (variationes primae) across different Italian regions. These fourteen varieties fell into numerous ‘secondary variations’ (variationes secundariae), spoken in smaller regions and cities. Dante went even further by introducing variations that were ‘subsecondary’ (subsecundarius) and by arguing that there were a thousand different tongues within cities. What is more, there were even differences per family (De vulgari eloquentia 1.10.4, 1.10.9, and 1.19.3). Apart from dissecting the phenomenon of linguistic diversity, Dante’s ideas are of interest for another reason, too. They suggest that medieval scholars focused principally on the opposition between Latin, on the one hand, and the chaos of vernacular varieties, on the other, or in Dante’s terminology: grammatica versus vulgaria. The Florentine poet’s interpretation of this opposition was, however, rather atypical, since he valued the natural vulgaria more highly than artificial grammatica. He even promoted an ‘illustrious’ form of the vernacular, a vulgare illustre, as an alternative for Latin grammatica. In Quattrocento Italian humanism, the opposition Latin/vernacular took on a new dimension, as it gave rise to a debate over their precise relationship.¹⁶ There were two main clashing opinions, to present the matter in a somewhat simplified manner. A first group of scholars, including Biondo Flavio (1392–1463) and Poggio Bracciolini (1380–1459), believed that the elite and the populace spoke the same language, albeit with slight, socially determined differences, whereas another group of authors, including Leonardo Bruni (c. 1370–1444) and Lorenzo Valla (c. 1407–1457), claimed that the Roman lower classes employed a language entirely different from the Latin of the higher classes. It can be readily seen how this debate had implications for views on the relationship between Italo-Romance varieties and Latin. The different degrees of linguistic diversity were barely relevant in a model focusing on the opposition Latin/vernacular, if at all. It was largely an either/or question: did these linguistic forms constitute either one single language or entirely distinct languages? In other words, these scholars had no use for an intermediate concept like dialect in the linguistic themes they debated.

¹⁶ See e.g. Tavoni (1982, 1984); Mazzocco (1993: 13–105); Fausel et al. (2002: 200–7); Coseriu and Meisterfeld (2003: 149–71); Celenza (2009); Eskhult (2018).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

27

In sum, it is difficult to maintain on the basis of the available evidence what medievalist Rüdiger Schnell (2008: 568–9) has contended for the later Middle Ages on the basis of German examples, namely that dialects were supposed to vary merely phonetically, whereas languages differed not only as regards their sounds but also lexically, morphologically, and syntactically. Such a generally accepted language/dialect distinction cannot be posited for late medieval times. One could assert, at best, that a number of late medieval authors were, usually very vaguely, aware of the fact that languages differ in regional terms and that linguistic variation comes in different degrees.

2.5 Conclusion Ancient, medieval, and early Renaissance references to, and conceptions of, regional language variation usually had a very incidental character, revealing different levels of awareness of this linguistic phenomenon. For the history of the conceptual pair, Greek interpretations of the key term diálektos are of central importance. At an early stage, this word was tied to notions of ethnicity, locality, and, especially, literariness, since it was first and foremost associated with the study of Greek texts and their diversified linguistic appearance. The philological perspective of ancient and medieval Greek scholars dominated and was bound to leave an indelible mark on later ideas, as I will argue in Part II. For this reason, I have concentrated on the Greek tradition, with some attention to the way Greek conceptions were received in the Latin West and other idiosyncrasies of Western medieval thought. It would be a fruitful avenue for further research to analyse how early approaches to dialectal variation relate to each other cross-culturally, but for my purposes it will suffice to conclude the following on the period before 1500: even though there was widespread awareness of regional language variation, and some authors such as Dante Alighieri even knew that linguistic diversity comes in different degrees, most scholars did not make a clear-cut conceptual distinction between a language-like and a dialect-like entity. There was, however, one conspicuous late medieval exception, which fully deserves a separate treatment.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

3 The exception to the rule Lingua and idioma in Roger Bacon’s thought

Roger Bacon (c. 1214/20–c. 1292) was a man of many firsts.¹ Bacon was, for instance, most likely the first European to describe gunpowder, and a pioneer in optics, a discipline he aspired to introduce into the quadrivium of the new universities in Western Europe, such as the one in Paris, where he lectured for quite some time. In this and other domains of study, Bacon was, moreover, innovative in adopting an inductive approach rather than following the deductive methods of his colleagues (cf. Figure 3.1). He attached great importance to experimental and empirical research in order to reach more general conclusions. It was this very desire to start from concrete reality that led him to become an early advocate of an ad fontes approach in theology, which, in his view, should focus on the Bible text. For this reason, he emphasized the importance of mastering ‘wisdom languages’ (linguae sapientales) other than Latin, too: especially Greek, Hebrew, Arabic, and Aramaic. Indeed, knowledge of languages is the first gate to wisdom, and especially among the Latins, who do not have the text of theology or philosophy, unless from foreign languages; and that is why every man should know languages.²

Or in a more poetical fashion: [The idiomatic individuality of every language] was the main reason why all saints, philosophers, and ancient sages wanted to know other languages, so that they could drink the waters of wisdom more sweetly and more fully in the source itself.³

Bacon even went a step further by strongly recommending the new universities to include the wisdom languages in their curricula and to stop focusing exclusively ¹ Information on Bacon’s life and works is mainly based on Molland (2004) and Hackett (2013). ² Bacon (1859: 102): ‘Notitia linguarum est prima porta sapientiæ, et maxime apud Latinos, qui non habent textum theologiæ, nec philosophiæ, nisi a linguis alienis; et ideo omnis homo deberet scire linguas’. See also Bourgain (1989: 318–19). ³ Bacon (1859: 90): ‘fuit causa principalis quare sancti omnes, et philosophi, et sapientes antiqui, voluerunt scire linguas alias, quatenus in ipso fonte dulcius et plenius biberent aquas sapientiæ’.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

    

29

Figure 3.1 Roger Bacon at his Merton College observatory, oil painting by Ernest Board Source: Wellcome Collection. CC BY 4.0

on Latin. It therefore comes as no surprise that he wanted to contribute to the study of the wisdom languages by composing grammars of Greek and Hebrew, a highly exceptional feat in the thirteenth-century West. Whereas his unique Greek

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

30

    

grammar has been preserved in its entirety, only a fragment of his Hebrew handbook seems to survive, and it is unclear whether he ever finished it (Hirsch in Bacon 1902: 201).

3.1 Bacon at Babel Let me take a closer look at Bacon’s linguistic horizon and his motives to linger at length on linguistic diversity. His perspective was still very much Latin-centred, since Latin remained the most familiar scholarly language for him and his colleagues (Dahan et al. 1995: 266). Bacon pointed out, however, that the language of the Romans was in various respects indebted to Greek and Hebrew, languages which he no doubt sensed to be much older and which he described in contrast to Latin in his pioneering grammars (Hovdhaugen 1990: 117). As can be expected from a thirteenth-century intellectual, Latin was the only language in which he composed his scholarly works. Yet unlike most of his colleagues, he did not blindly accept the monopoly of the language of the ancient Romans on science. He was well aware that there were other languages of interest for scholarship, most importantly Greek and Hebrew. To this end, he had mastered both these languages, making him an exceptional medieval instantiation of Jerome’s ideal of the vir trilinguis, the ‘trilingual man’. Bacon’s linguistic mastery did not stop there, however, as he also had some notions of Biblical Aramaic and, to a lesser extent, Arabic. Bacon’s interest in the three so-called sacred languages must have been part of a broader tendency in Franciscan circles in late medieval times. The Flemish Franciscan scholar Gerard of Huy, for instance, was the author of the work Triglossos, id est liber trium linguarum, basically a glossary, in verses, of Latin, Greek, and Hebrew words in the Bible (Roest 2015: 97). Another example is William de la Mare, an English Franciscan friar active in the 1270s in Paris and influenced by Bacon, who composed a glossary of Hebrew and Greek words in the Bible and who was concerned with the correctness of the Vulgate text (Marenbon 2004). The fact that the Ancient Greek dialects had acquired literary status and thus had become part of the grammatical canon of Greek forced Bacon to comment on this matter in his Greek grammar. He drew up this work, which followed in the tradition of Byzantine grammar and is the first extant Greek grammar written in Latin, around 1268. Still, his grammar was much more concerned with the Greek dialects than the Byzantine elementary handbooks on which he could rely and which were built up in question-and-answer format, and many of Bacon’s remarks were idiosyncratic (see also Vandewalle 1929: 54–5; Consani 1991: 151; RosierCatach 1997: 86). Most of Bacon’s observations on dialectal diversity are found in his Greek grammar, but he also treated the theme of linguistic diversity in his Major work

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     ? 31 of about 1266–7 (cf. Formigari 2004: 84). This work was intended as a lengthy synthesis of natural philosophy, written in response to a mandate of Pope Clement  (ruled 1265–8), of which Bacon learned in July 1266 while in Paris. Bacon also treated this theme in his Third work, most likely conceived as an adapted shorter version of his Major work, which ran the risk of getting lost during its transportation to the Holy See. Finally, in his Compendium of the study of philosophy (c. 1271), a critical description of the intellectual situation in Paris with a defence of his own work, he also elaborated on the issue of linguistic diversity and dialectal variation. Bacon did not restrict himself to noticing variation in Greek, as he showed himself aware of diversity within vernacular tongues as well. Even though the wisdom languages were the focus of his attention, Bacon had an ear for the tongues of his contemporaries, too, with which he came into contact on his travels to different parts of Europe. Born in Somerset, where a South Western variety of Middle English was spoken, he taught at Oxford, where the South Western, West Midland, and East Midland dialect areas came together, and at Paris, to which students came from every area of France. Under these circumstances he must have quickly found out that vernacular languages exhibited regional differences much like those of Greek (cf. Hofmann 1883; Bourgain 1989: 325). All in all, however, Bacon was not very enthusiastic about these vulgar tongues, which in his eyes were useless for learning, since they lacked the necessary philosophical terminology (Dahan et al. 1995: 267). Yet his interest in all things empirical nevertheless seduced him to comment on this phenomenon, undeniably present in everyday life.

3.2 Another first for Bacon? In his attempts at mapping out regional linguistic diversity, Bacon introduced a distinction that helped him make sense of this phenomenon. As he wrote very quickly and very copiously, which resulted in frequent repetitions throughout his writings, we are relatively well-informed about his ideas on this matter. Indeed, by putting together the repeated statements, which differ in detail, we can build up a richer picture than we would get from any one statement (cf. Bourgain 1989: 318; Fausel et al. 2002: 196). When discussing the varieties of the Greek language in his Greek grammar, Bacon proposed to carefully distinguish between two types of linguistic entities, which he called lingua and idioma: For an idiom is a specific property of a language, which one people uses according to its custom. And another people uses another idiom of the same

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

32

     language. For idion in Greek is proprium in Latin, whence idioma, i.e. proprietas.⁴ And I call these diversities idioms and not languages, as many do, because, in reality, they are not different languages but different properties which are idioms of the same language.⁵ An idiom is a property of speaking in a language.⁶

It immediately leaps to the attention that Bacon was operating with linguistic concepts on two different levels. There were different ‘languages’ (linguae), and each of these ‘languages’ was internally diversified in terms of ‘idioms’ (idiomata). This strict division, Bacon suggested in passing, was an innovation of his own. By means of the parenthetic remark ‘as many do’ in the second passage, he complained about the carelessness of other scholars, who used both terms interchangeably. Yet Bacon was not always straightforward himself. His numerous definitions of idioma and his usage of the term reveal a confused concept (cf. Rosier-Catach 1997: 85). An idioma was interpreted as a particularity of speech within one and the same lingua, that is to say, with clear subsumption of idioma under lingua. This relationship most obviously emerges from Bacon’s repeated usage of the genitive phrase idiomata eiusdem linguae, ‘idioms of the same language’. Yet Bacon seems to have applied the term idioma not only to native varieties of a language such as Burgundian French and Northern English, but apparently also to different orthographies and pronunciations of Latin, for instance, by German and Spanish natives. For he stated: Indeed, in the Latin language, which is one, there are many idioms. For the substance of the language itself consists in these things in which all clerics and scholars take part. But there are many idioms according to the multitude of nations using this language. For in many respects the Italians pronounce and write it in one manner, the Spanish in another, the French in yet another, the Germans in still another, and the English, too, in another, and so on. So in this way there was among the Greeks one language according to substance, but there were many properties.⁷

⁴ Bacon (1902: 26): ‘Idioma enim est proprietas lingue determinata, qua vna gens vtitur iuxta suam consuetudinem. Et alia gens eiusdem lingue vtitur alio idiomate. Idion enim grece est proprium latine, a quo idioma hoc est proprietas’. ⁵ Bacon (1902: 27): ‘Et voco has diuersitates idiomata et non linguas vt multi vtuntur, quia in veritate non sunt lingue diuerse sed proprietates diuerse que sunt idiomata eiusdem lingue’. ⁶ Bacon (1902: 75): ‘idioma est proprietas fandi in aliqua lingua’. ⁷ Bacon (1902: 26–7): ‘In lingua enim latina que vna est, sunt multa idiomata. Substancia enim ipsius lingue consistit in hijs in quibus communicant clerici et literati omnes. Idiomata vero sunt multa secundum multitudinem nacionum vtencium hac lingua. Quia aliter in multis pronunciant et scribunt ytalici, et aliter hyspani, et aliter gallici, et aliter teutonici, et aliter anglici et ceteri. Sic igitur fuit apud grecos vna lingua secundum substanciam sed multe proprietates’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     ? 33 There is no doubt that Bacon was indeed referring to the diversified spelling and pronunciation of the Latin language by speakers of different vernacular tongues rather than stating that Italian, Spanish, French, German, and English were daughter languages of Latin.⁸ The former view is corroborated by the fact that Bacon still regarded Latin as the living mother tongue of Western Europe. After all, this idea implies that Latin was subject to internal variation in a way similar to other Western European vernacular tongues and to Greek, another ‘wisdom language’. Also, Bacon was not only insisting on variation in Latin pronunciation but also in writing. He was apparently thinking of the written language par excellence, Latin, and not of the so-called vulgar languages, which lacked a standard as well as an established written tradition. Entirely in line with the Greek tradition, Bacon stressed the connection between an idioma and a natio, ‘tribe’ or ‘nation’: ‘For an idiom is a property of a language with a certain nation’.⁹ For the sake of convenience, the Latin term natio will be translated in this book as ‘nation’, but it should be consistently interpreted in the non-political sense of ‘group of people (allegedly) sharing the same ethnic background and living in the same area’, unless otherwise indicated. Bacon was, moreover, aware of the fact that speech diversifies over space just like everything else that is human, and for this dialect–space correlation he might have been indebted to the Greek tradition, too (e.g. Bacon 1859: 438–9; see Lusignan 1986: 71–2; Bourgain 1989: 318). Bacon did not explicitly define the term lingua but used it in various senses. To the English scholar, it designated not only individual languages such as Latin, but also groups of genealogically related varieties which were considered to constitute one lingua. For instance, he regarded all Gallo-Romance tongues as one lingua Gallicana (Fausel et al. 2002: 197). Bacon nonetheless offered a glimpse of his conception of lingua when discussing the Greek language in his grammar. In his opinion, the lingua Graeca consisted in those elements ‘in which all Greek nations partook, and these elements are called “common” ’.¹⁰ The source for this view of his must have been Byzantine scholarship, in which the idea that the Koine consisted of what was common to the dialects was widespread. For instance, John the Grammarian recalled that some scholars contended that the Koine had no particularities of its own, but was compounded out of the four other dialects (Manutius et al. 1496: 236). Bacon was also convinced that related idioms showed only superficial, accidental variation, for instance, in terms of pronunciation, but constituted a unity in

⁸ On this point I agree with Bourgain (1989: 326), Rizzo (1990: 37–40), and Rosier-Catach (1997: 85), while disagreeing with Consani (1991: 148–56). ⁹ Bacon (1900: 73–4): ‘Idioma enim est proprietas linguae apud aliquam nationem determinatam’. ¹⁰ Bacon (1902: 27): ‘Natura igitur ipsius lingue grece consistit quantum ad ea in quibus omnes naciones grece communicabant et hec vocantur communia’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

34

    

substance.¹¹ This assertion must be viewed against the background of Aristotelian ontology. It is inevitable to quote in this context Bacon’s notorious comment on grammar, which was likewise informed by Aristotelian concepts, but remains difficult to interpret: ‘grammar is one and the same in terms of substance in all languages, even though it can vary accidentally’.¹² Bacon’s phrasing here reminds of the later doctrine of speculative grammar that was only taking a start around his time of writing, but his ideas must be set apart from them, as Even Hovdhaugen (1990: 124–5) has convincingly argued by comparing the terminology of Johannes Dacus’s Grammatical overview (Summa grammatica) of c. 1280 with Bacon’s. Also according to Hovdhaugen (1990: 127), the sentence should be interpreted as follows: ‘The science of analysing and describing language is [in essence] one and the same for all languages although there are accidental variations’. This interpretation seems indeed very plausible, since Bacon probably conceived of grammatica as the categories of grammar that were largely universal, with some accidental particularities. What is more, he explicitly emphasized the utility of comparing Greek and Latin grammar, which did not differ substantially in his eyes. In fact, Latin was derived from Greek grammar, Bacon asserted on the authority of Priscian, a favourite source of his. He did not refer to a specific passage in Priscian’s Latin grammar, but he was surely thinking of comments of this early sixth-century grammarian on the close connection between Latin and the Aeolic Greek dialect. Indeed, Priscian, too, was convinced that Latin derived from Greek and especially from Aeolic—‘Aeolism’ in short, which in antiquity was more a commonplace than a true theory (Stevens 2006–7). Priscian principally related this affinity to the Latin letter and the formally identical Aeolic digamma, a grapheme representing the sound [w] (Keil 1855–80: , 15, 1–6 and , 17, 10–17). The assumption that grammatical categories were universal also seems to emerge from the fact that Bacon (1902: 27, 207) linked the Greek and Hebrew articles with the Latin ‘articular pronoun’ (pronomen articulare), by which he meant demonstrative pronouns such as hic. It is confusing that Bacon (1902: 26–7) used a highly similar expression to describe regional linguistic variation in the vicinity of his notorious dictum on grammar. In particular, he contended that Latin and Greek, although one in substance, were each characterized by many ‘idioms’ (idiomata/proprietates). The precise relationship of this statement to the passage just quoted remains somewhat unclear, but it seems justifiable to argue that they should not be read in close connection, even though they were both shaped by Aristotelian concepts ¹¹ Bacon (1859: 90): ‘Nam lingua Latina est in his omnibus una et eadem, secundum substantiam, sed uariata secundum idiomata diuersa’. ¹² Bacon (1902: 27): ‘grammatica una et eadem est secundum substantiam in omnibus linguis, licet accidentaliter uarietur’. See e.g. Bursill-Hall (1975: 201), Maierù (1983: 743–4), Lusignan (1986: 67), Bourgain (1989: 320–1), Hovdhaugen (1990), Rosier-Catach (1997: 85), and Formigari (2004: 84) for suggestions of interpretation.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

35

and appear in close vicinity. In short, the phrase on grammar concerned the quasi-universal nature of grammatical categories, whereas the assertions on Greek and Latin were, in fact, applications of his lingua/idioma distinction rather than generalizing statements on the nature of grammar. Bacon’s choice for the term lingua is no surprise, but why did he opt for the Greek loanword idioma, from idíōma (ἰδίωμα), to denote a variety of a language and not for dialectus, the Latinized form of a central term in Greek grammatical tradition? As a matter of fact, he did use the Greek term diálektos as well as dialectos, its transcription into the Latin alphabet, but only very occasionally and exclusively to talk about varieties of Greek. Latinized dialectus, however, had not yet been incorporated into the metalinguistic terminology of Bacon’s days (see Chapter 2, Section 2.4). He thus principally relied on other terminological options to express concepts related to linguistic variation. Apart from idioma, he also referred to dialects and dialectal variation in terms of proprietas (loquendi), diversitas, and variare. These words were all well-established Latin expressions. Idioma, for instance, had been in constant use since late antiquity as a word denoting, among other things, linguistic entities, especially in contexts in which their particular properties were stressed.

3.3 An outlook on different languages and their dialects: a central precondition How and why could Bacon develop a conceptual scheme of individual languages each subsuming several different idioms? In his Major work, he explained the opposition lingua/idioma by means of the example of France, where he wrote this work: but it is impossible that the property of one language is preserved in another. For also idioms of the same language are varied among diverse [nations], as is clear from the French language, which is varied by a manifold idiom among the Walloons, the Picards, the Normans, and the Burgundians. And what is properly said in the idiom of the Picards, becomes rough among the Burgundians, indeed even among the Walloons, who live nearer to Picardy; so how much more will this occur among different languages? Therefore, what is well done in one language, cannot be transferred to another according to the property it had from the start.¹³ ¹³ Bacon (1900: 66–7): ‘sed impossibile est quod proprietas unius linguae servetur in alia. Nam et idiomata ejusdem linguae variantur apud diversos, sicut patet de lingua Gallicana, quae apud Gallicos et Picardos et Normannos et Burgundos multiplici variatur idiomate. Et quod proprie dicitur in idiomate Picardorum horrescit apud Burgundos, immo apud Gallicos viciniores: quanto igitur magis accidet hoc apud linguas diversas? Quapropter, quod bene factum est in una lingua, non est possibile ut

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

36

    

Here, Bacon was explaining the idiomatic characteristics of different linguistic forms in the context of translation theory. All idioms of a language had their idiosyncrasies, which applied a fortiori at the level of languages and made literal, word-for-word interpretation impossible. In this passage, we can find a hint as to why Bacon adopted idioma as his term to denote varieties of one and the same language. Even though they showed many similarities, it was their mutual differences and idiosyncratic ‘properties’ (idiomata) that immediately caught the eye. Bacon was, in other words, seemingly jumbling together two concepts modern linguists tend to clearly distinguish: dialect and idiom. The idiomatic properties of each variety of a language may, Bacon observantly remarked, cause mutual laughter. In his Third work, he rephrased his views on this issue and expanded his perspective so as to include almost the entire Romance area: for an idiom is a property of a language distinct from another, such as the Picard, the Walloon, and the Provençal, and all idioms from the borders of Apulia to the borders of Spain. For the Latin language is in all these [regions] one and the same in terms of its substance, but varied according to different idioms. And we manifestly see that something sounds excellent and proper in one idiom and ridiculous in another, as is clear not only from those far removed but also from those very close-by, e.g. from the Picards and the Walloons, for they laugh at each other.¹⁴

It can be noted that Bacon was here clearly thinking of vernacular Romance variation rather than of variation in the pronunciation and orthography of the Latin language. Bacon most likely abstracted the lingua/idioma distinction from the linguistic contexts with which he was most familiar, most notably English and French, and used it to describe and analyse these same as well as other linguistic contexts less familiar to him, most importantly Greek and its dialects as well as the family of languages today known as Semitic. This cognitive process is excellently exemplified by his observation on the relationship between Hebrew and Aramaic:

transferatur in aliam secundum ejus proprietatem quam habuerit in priori’. Gallicus seems to refer to the Walloons, whereas Gallicanus is more generic, denoting Gallo-Romance as a whole. For Bacon’s ideas on French dialectal variation see also Hofmann (1883), Meyer-Lübke (1908), and Lusignan (1986: 67–72). ¹⁴ Bacon (1859: 90): ‘nam idioma est proprietas alicujus linguæ distincta ab alia; ut Picardicum, et Gallicum, et Provinciale, et omnia idiomata a finibus Apuliæ usque ad fines Hispaniæ. Nam lingua Latina est in his omnibus una et eadem, secundum substantiam, sed variata secundum idiomata diversa. Et manifeste videmus quod aliquid optime et proprie sonat in uno idiomate et ridiculose sonat in alio; ut patet non solum de longinquis sed propinquissimis; sicut de Picardis et Gallicis; nam mutuo se derident’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

37

And it is certain that the Hebrews and Chaldeans have the same language but a different idiom, just like the Walloon and the Picard . . . the Hebrew says Eloim for “God” or “Gods”; the Chaldean says Eloa for “sky” or “skies”. For “not”, the Hebrew says lo, the Chaldean says la, and similarly in other words.¹⁵

In his Compendium of the study of philosophy, Bacon made a similar point, but now in more general terms, stating: for Aramaic and Hebrew speech differ as idioms of one language, just like the Picard and the Norman, the Burgundian, the Parisian, and the Walloon; for there is one language among all, namely the French language, but still in several parts it varies accidentally; a diversity that makes idioms, not different languages.¹⁶

Other language families were described in similar terms. In the fourth part of his Major work, devoted to geography, Bacon briefly commented on the Slavic family of tongues, which he apparently regarded as idioms of one and the same language, even though he did not use the term in this context: The Russians are Christians and at the same time schismatic, as they have the rite of the Greeks. Yet they do not use the Greek language, but rather the Slavonic language, which is one of the languages that cover multiple regions. For it comprises Russia, Poland, and Bohemia, and many other nations.¹⁷

It is conceivable that Bacon thought Russian, Polish, and Bohemian to be idioms of one Slavonic language, as he linked them to different Slavic nations, even though Pascale Bourgain (1989: 318) is right in emphasizing that Bacon was not familiar with the Slavic tongues and suggests that he simply believed Slavic to be a unitary tongue. Still, Bacon thought of regional variation as being inherent in language, and even though he did not know any Slavic tongue, it went without saying for him that a language spoken in multiple regions and by multiple nations underwent diversification. Bacon’s acquaintance with several dialect contexts was without doubt an important trigger for his designing a lingua/idioma distinction. As a matter of ¹⁵ Bacon (1900: 73–4): ‘Et certum est quod Hebraei et Chaldaei eandem habent linguam, sed diversum idioma, sicut Gallicus et Picardus . . . Hebraeus dicit Eloim pro Deo vel Diis; Chaldaeus dicit Eloa, pro coelo vel coelis. Pro non, Hebraeus dicit lo, Chaldaeus dicit la, et sic in aliis’. Cf. Demonet (1992: 27); Dahan et al. (1995: 275). ¹⁶ Bacon (1859: 438–9): ‘Chaldæus enim sermo et Hebræus differunt sicut idiomata unius linguæ; ut Picardicum et Normanicum, Burgundicum, Parisiense, et Gallicum; una enim lingua est omnium, scilicet Gallicana, sed tamen in diversis partibus diversificatur accidentaliter; quæ diversitas facit idiomata non linguas diversas’. ¹⁷ Bacon (1900: 360): ‘Rusceni sunt Christiani et sunt schismatici, habentes ritum Graecorum, sed non utuntur lingua Graeca, immo lingua Sclavonica, quae est una de linguis quae plures occupant regiones. Nam Rusciam, Poloniam, et Bohemiam, et multas alias nationes tenet’. Cf. Rosier-Catach (1997: 84).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

38

    

fact, such a conceptual pair made most sense to a scholar engaged, first, in separating different languages from each other and, second, in mapping out their respective internal diversity. As I have contended at length, such an interest clearly emerges from Bacon’s works. I can sum up here once more the different linguistic contexts he mentioned, truly impressive for a thirteenth-century scholar: English, French, Greek, Latin, Romance, and Semitic. Tellingly, Bacon claimed that ‘whoever is happy with their property of speaking, while ignoring the properties of speech of others, is a person of narrow learning (idiota)’.¹⁸ Bacon was adapting here a passage in the Venerable Bede’s commentary on Acts and projecting it onto the level of idioms of a language. Bede’s original reads as follows: For persons of narrow learning (idiotae) were called those who, satisfied with only their own language and with natural science, did not know the study of letters, since the Greeks say “proper” as “ídion”.¹⁹

An outlook on multiple languages and their respective internal diversity, then, seems to have been a necessary precondition for distinguishing between language and dialect-like concepts. That it was not a compelling one has been established by means of the example of the ancient Romans in the previous chapter. Lacking the comparative-contrastive perspective of Roger Bacon, they were blindsided by their assumption that the Greek literary dialects were something completely different from regional variation within Latin.

3.4 Thomas Aquinas: a like-minded exegete? Bacon was in good company in distinguishing two levels of linguistic entities that exhibited different degrees of variation. There is a close parallel to his opposition of lingua to idioma in the exegetical work of the Aristotelian philosopher and Dominican friar Thomas Aquinas (c. 1225–74). In a commentary on the Book of Psalms, Thomas was very careful in distinguishing loquela from sermo, stating that ‘loquelae signify principal languages, but sermones signify varieties of idioms in the same language’ or—put differently—‘loquelae are languages’, whereas ‘sermones are manners of speaking’.²⁰ His distinction between loquela and sermo was triggered by an anacoluthic passage in Psalms 18.4 of the Vulgate version,

¹⁸ Bacon (1902: 26): ‘Et idiota qui est contentus sua proprietate loquendi, nesciens proprietates sermonis aliorum’. Cf. Bacon (1902: 75). ¹⁹ Expositio actuum apostolorum 4.40: ‘Idiotae enim dicebantur qui propria tantum lingua naturalique scientia contenti litterarum studia nesciebant, siquidem Graeci proprium ἴδιον vocant’. ²⁰ In Psalmos reportatio 18.2: ‘Loquelae significant linguas principales; sed sermones significant varietates idiomatum in eadem lingua. Vel loquelae linguae, sermones sunt modi loquendi’. See Lusignan (1986: 61); Eloy (1998: 4).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.  :  - ?

39

which reads: ‘non sunt loquellae neque sermones quorum non audiantur voces eorum’. ‘There are no languages, nor are there speeches of which the voices cannot be heard’. This verse corresponds to Psalms 19.3 in the English Standard Version, which translates the original Hebrew as follows: ‘There is no speech, nor are there words, whose voice is not heard’. Thomas’s terminological contrast resembles to some extent later oppositions of dialect to language, even though he refrained from further elaborating on his distinction, and his definitions, especially of sermo, remained quite vague. He did seem to put the opposition into practice elsewhere in his oeuvre, when commenting on the recognition of Peter as a Galilean by means of his speech in Matthew 26.73: ‘After a little while the bystanders came up and said to Peter, “Certainly you, too, are one of them, for your accent betrays you” ’. In the Latin Vulgate version, the last part reads: ‘loquella tua manifestum te facit’. Thomas Aquinas attempted to explain the passage by rephrasing Jerome’s comments on it and by referring to Gallo-Romance diversification, which was also Roger Bacon’s preferred example: ‘in the same language (lingua), there is often diverse speech (locutio), as is clear in the Île-de-France, Picardy, and Burgundy, and still the language (loquela) is one’.²¹ Again, Thomas used the term loquela here to refer to language, just as in his commentary on Psalms, but its presence here is rather curious. For the Vulgate passage of Matthew 26.73 has the Latin word loquela refer to the Galilean characteristics of his speech, that is to say, to regional variation within a language rather than to differences between distinct languages. Thomas thus must have had the Psalms verse, and his interpretation of it, in mind when commenting on Matthew’s gospel. For Thomas, lingua could apparently serve as a synonym for loquela to refer to language, whereas he refrained from using the term sermo here and opted instead for locutio to denote regional diversity. The example of Thomas indicates the importance of engagement with the Bible text for eliciting observations on dialectal variation. The very same passage in the gospel of Matthew had led Jerome, for instance, to argue that it pointed to variation within Hebrew, as ‘every province and region had its properties and could not avoid the native sound of speaking’—an explanation adopted by the Venerable Bede.²² Jerome even linked Matthew 26.73 to the passage in Judges 12.5–6, which relates how two tribes of Israel, the Gileadites and the Ephraimites, were waging war and the former succeeded in unmasking the latter by means of linguistic evidence. The fleeing Ephraimites were unable to pronounce the word shibbóleth, meaning either ‘stream’ or ‘part of a plant containing the grain’, and ²¹ Super Evangelium Matthaei reportatio 26.7.2296.1: ‘in eadem lingua saepe diversa locutio fit, sicut patet in Francia, et Picardia, et Burgundia, et tamen una loquela est’. ²² Jerome, Commentarii in evangelium Matthaei 1.1452–6: ‘unaquaeque provincia et regio habebat proprietates suas et vernaculum loquendi sonum vitare non possit’. Cf. Bede, In Marci evangelium expositio 4.14.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

40

    

said sibbóleth instead, which led to their bloody execution.²³ The episode in Judges was intensively commented upon, and shibbóleth would eventually lend its name to the concept of shibboleth, designating a particular feature, often of a linguistic nature, setting a specific group of people apart from others. Some decades after Thomas Aquinas, another influential exegete, the Franciscan Nicholas of Lyra (1265–1349), was reminded of Gallo-Romance variation in a manner similar to the great Aristotelian philosopher. Nicholas contrasted Picard to Parisian speech and emphasized that one’s speech variety could be used as a distinctive mark revealing one’s geographical provenance, as a shibboleth: For even though the Hebrew language is one, it was still in some way varied according to the diversity of provinces. And for this reason those from Galilee spoke in some way differently from those who originated from Jerusalem, just as the French tongue is one and yet those who are from Picardy speak differently from those who live in Paris. And by such a difference it can be perceived whence someone originates.²⁴

3.5 The Tower of Babel One of the major linguistic events described in the Bible is the confusion of tongues at the Tower of Babel in Genesis 11.1–9 (cf. Figure 3.2). The text, however, does not offer much information on the precise nature of this confusion. The reader is only told that mutual understanding was made impossible by God, so as to bring the building of the Tower to an end: Now the whole earth had one language and the same words. And as people migrated from the east, they found a plain in the land of Shinar and settled there. And they said to one another, “Come, let us make bricks, and burn them thoroughly.” And they had brick for stone, and bitumen for mortar. Then they said, “Come, let us build ourselves a city and a tower with its top in the heavens, and let us make a name for ourselves, lest we be dispersed over the face of the whole earth.” And the L came down to see the city and the tower, which the children of man had built. And the L said, “Behold, they are one people, and ²³ See e.g. Swiggers (1981) for the Hebrew word, ‫ ִשׂ ֹבֶּלת‬/‫ ִשׁ ֹבֶּלת‬as opposed to ‫ִס ֹבֶּלת‬, in this passage. Cf. Trabant (2006: 24). ²⁴ Nicholas of Lyra (1471: [81–2]): ‘licet enim linguua [sic] Hebrea sit una: tamen aliquo mo[do] uariebat[ur] [sic] secundum diuersitate[m] prouinciarum. & ideo illi de Galylea aliter loquebant[ur] aliquo modo: q[uam] illi qui erant de Hierusalem: sicut lingua Gallica est una: & tamen aliter loquunt [ur] illi qui sunt de Picardia: q[ua]m illi qui habitant Parisius [sic]: & per talem uarietatem potest percipi unde aliquis sit oriundus’. I quote from the printed editio princeps for lack of a modern critical edition. See also Várvaro (2008: 224).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

41

Figure 3.2 The Tower of Babel through the eyes of Pieter Bruegel the Elder, 1563 Held at the Museum Boijmans Van Beuningen, Rotterdam Source: Wikimedia Commons; public domain

they have all one language, and this is only the beginning of what they will do. And nothing that they propose to do will now be impossible for them. Come, let us go down and there confuse their language, so that they may not understand one another’s speech.” So the L dispersed them from there over the face of all the earth, and they left off building the city. Therefore its name was called Babel, because there the L confused the language of all the earth. And from there the L dispersed them over the face of all the earth.

Some modern linguists might conclude that the absence of mutual intelligibility among the workers indicates differences on the level of language rather than dialect, but since the conceptual pair was yet to establish itself, this idea was completely absent from ancient and medieval linguistic thought. Exegetes understood that new languages were spoken, but most of them did not yet have the conceptual apparatus to ask whether it was better to think of them as languages or as dialects. Still, one late medieval author, the Spanish scholar Rodrigo Jiménez de Rada (c. 1170–1247), tried to understand the events at Babel by seemingly working with

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

42

    

a distinction between languages and idioms in the first book of his History of the vicissitudes of Spain, or Gothic history. As a consequence of the building of the Tower, mankind was divided linguistically and, hence, also ethnically. In the aftermath of this double separation, the tribes ‘distinguished the tongues by idioms’.²⁵ Rodrigo did not explain what he exactly meant by these ‘idioms’ and how they related to the various languages that emerged. He seems to have been thinking of an intensification of the differences among the tongues that arose at Babel rather than of an internal diversification within each of the tongues separately. That is, at least, how the use of the active verb may be interpreted, with each tribe, driven by mutual hatred, looking to enhance mutual linguistic differences. Two chapters further on, Rodrigo again juxtaposed lingua and ydioma, this time with reference to Germanic-speaking areas. Germany, Denmark, Norway, Sweden, Flanders, and England were said to ‘have a single language, even though they are distinguished by idioms’.²⁶ The term ydiomata probably needs to be taken here in the sense of ‘particularities’ or ‘properties’ rather than ‘separate linguistic entities subsumed under one and the same language’ (~ dialects). In sum, Rodrigo does not seem to have operated with a distinction between language- and dialectlike entities, even though a superficial reading of his work may convey such an impression.

3.6 Conclusion I can only agree with Gilbert Dahan, Irène Rosier-Catach, and Luisa Valente when they state that Roger Bacon was ‘one of the most subtle and original minds of the Middle Ages, especially regarding the reflection on languages’.²⁷ For this reason, Bacon is often bracketed together with Dante Alighieri in current scholarship. Both late medieval thinkers were intrigued by the phenomenon of regional variation within a language and formulated pioneering ideas about the matter, albeit in clearly distinct contexts.²⁸ Bacon framed it within a broader insistence on the variability of human culture, which he tied first and foremost to the factor of space (Lusignan 1986: 71; Bourgain 1989: 318). In addition, he related regional variation in a language to translation theory. Literal translation conveying all nuances of the source language was impossible, since there existed idiomatic differences even among varieties of individual languages.

²⁵ Historia de rebus Hispanie sive Historia Gothica 1.1: ‘linguas ydiomatibus distinxerunt’. ²⁶ Historia de rebus Hispanie sive Historia Gothica 1.3: ‘unicam habent linguam, licet ydiomatibus dinoscantur’. Cf. Bonfante (1953–4: 681). ²⁷ Dahan et al. (1995: 267): ‘un des esprits les plus subtils et originaux du moyen âge, surtout pour ce qui concerne la réflexion sur les langues’. ²⁸ See Hüllen (2001: 217–18); Fausel et al. (2002: 195). See Chapter 2, Section 2.4, for Dante’s views.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

43

Despite the pioneering character of Bacon’s insights, most modern scholars have only touched sideways on his lingua/idioma distinction. In response to Even Hovdhaugen’s (1990: 117) call, I have endeavoured to cast a clearer light on this issue, which is not as straightforward as is often maintained (cf. e.g. Lusignan 1986: 68). By subsuming idioma under lingua, Roger Bacon was the first to develop and define at some length a language/dialect-like distinction, an innovation of which he was perfectly aware. He mainly related it to Aristotelian notions, with related idioms only showing accidental differences, whereas languages differed substantially. Like diálektos in the Greek tradition, idioma was also closely tied up with the entity of ‘nation’ (natio) in Bacon’s thought. It is moreover noteworthy that the English scholar adopted an objective approach to regional variation within English and French. He did not express any value judgements on individual idioms, even though he was aware that idioms of one language could give cause to mutual ridiculousness among speakers. Bacon’s neutrality in this matter partly resulted from the linguistic context of thirteenthcentury Western Europe. Large-scale vernacular standardization was, after all, still in the future, as were the prescriptive and subjective language attitudes that went with it. At the same time, Bacon’s interpretation of idioma remained somewhat confusing, as he did not link it only to regional variation in vernacular tongues but also to idiomatic properties causing untranslatability and to different ways of writing and pronouncing Latin. Yet since he still conceived of Latin as a mother tongue, this last idea probably seems inconsistent only from our present-day point of view. Still, his interpretation of idioma was far from straightforward, and a similar conclusion can be drawn for his vague usage of lingua, a term he did not explicitly define. Bacon’s case strongly suggests that the design of a language/dialect-like distinction had a peculiar precondition, which could act as a triggering circumstance: a familiarity with regional variation in two or more different languages. To scholars with this specific kind of linguistic outlook, such a conceptual opposition made obvious sense and was a useful descriptive tool for ordering and mapping out the great linguistic diversity they noticed. The fact that Bacon was one of the few medieval scholars who knew Ancient Greek well is surely no coincidence in this regard. His linguistic horizon indeed was exceptionally wide, and most medieval scholars lacked a panoramic view on language, a state of affairs that profoundly changed only in the early modern era. Even though linguistic variation was a well-known phenomenon during the later medieval period, no generalizations on regional variation within a language seem to be extant in texts written before about 1250. Even after this such statements were rare, with the works of Thomas Aquinas and especially Roger Bacon as the main exceptions. Bacon’s oeuvre remained, however, buried in the back rooms of libraries and was only rediscovered and published in modern times.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

44

    

His Greek grammar, for instance, survives in four manuscripts but was published only in 1902 (Nolan and Hirsch in Bacon 1902: lxv–lxxi; Consani 1991: 156; Dahan et al. 1995: 274). Bacon thus remained an isolated early pioneer, whose lingua/idioma opposition did not influence later linguistic thought. Instead, early modern scholars independently reinvented a language/dialect opposition.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

II

THE O RIGIN OF T HE C O N C E P T U A L P A I R , 15 0 0 – 50

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

4 From dogs and hounds to languages and dialects The conceptual pair in Conrad Gessner’s work

Thus far I have treated of dogs those aspects that seemed common to all dogs; now it remains for me to draw up those matters that pertain to diverse dogs separately.¹ The sixteenth-century Swiss polymath Conrad Gessner (1516–65) was greatly invested in cataloguing diverse entities, especially natural ones like animals and plants. In his impressive publications, Gessner usually discussed the different kinds in alphabetical order, even though his entries ‘often included related species under the same heading’ (Ogilvie 2006: 216). So, for instance, he described dogs first in general in his History of animals (cf. Figure 4.1). Then he pointed out that they could be divided into different types, including hounds, the kind hunters used in chasing prey. These types could in their turn be subdivided into different subspecies. The range of natural entities the productive Zürich humanist tried to catalogue and classify was not limited to animals and plants, however. He also tried to impose order on the chaos of existing language forms in his much-read Mithridates, the first language catalogue ever to be published. It was, however, in his preface to Josua Maaler’s (1529–99) German–Latin dictionary of 1561 that he voiced his language-classifying concerns most powerfully: ‘I do not separate Swiss from German but make it a dialect thereof ’.² Indeed, just as a hound was a kind of dog, Swiss was a kind of German; it was a dialect of it. Gessner’s active role in relating these two forms of speech finds a striking expression in his phrasing. It is his deliberate decision—facio, ‘I make’—to classify Swiss as a dialect of German. Gessner’s observation presupposed a contrast between dialectus and lingua, with the former subsumed under the latter, and suggests that, by his time of writing, a distinction similar to Roger Bacon’s lingua/idioma opposition had been introduced, a development of central significance thus far largely neglected by modern scholars; a notable exception is Alinei (1980; cf. also Alinei 1984 [1981]: ¹ Gessner (1551: 235): ‘H de canibus dixi ea quæ canibus communia omnibus uidebantur: reliquu[m] est ut quæ singulatim ad canes diuersos pertinent conscribam’. ² Gessner (1561: *.5): ‘ego Helueticam à Germanica non separo, sed eius dialectum facio’.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

48

       

Figure 4.1 Dogs in Conrad Gessner’s work Source: Wellcome Collection. CC BY 4.0

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    

49

179). How and why did this evolution occur, and what evidence is available? And in what ways did it differ from Bacon’s innovation? Before I attempt to answer these questions both from a close-reading perspective in this chapter and by means of a bird’s-eye view in the next, I should say a brief word on the early modern language horizon and the linguistic profile of humanist scholars, indispensable for an adequate understanding of the early sixteenth-century emergence of the language/dialect pair.

4.1 Exploring the linguistic world How narrow or expanded was the early modern linguistic horizon?³ There can be no doubt whatsoever that it was much broader than it had been in ancient and medieval times. The renowned intellectual historian Peter Burke (2004: 15–42) has rightly spoken of ‘the discovery of language in early modern Europe’, even though it might be more accurate to refer to languages in the plural. The phenomenon of language had been on the scholarly agenda for centuries, but early modern scholars were the first to explore at considerable length the world’s linguistic plurality. This discovery of languages can be framed within the ‘early modern information explosion’ and the concomitant ‘info-lust’ of scholars (Blair 2010: 11–12). They were overwhelmed by all kinds of new information in different branches of learning, a development triggered first and foremost by the printing press and European explorations of the world. With which languages were sixteenth-century scholars primarily involved? To begin with, for the first time in history, it was no longer a rarity to embody Jerome’s ideal of the vir trilinguis, the scholar versed in the three so-called sacred languages: Hebrew, Greek, and Latin, the languages in which the title ‘King of the Jews’ was believed to have been inscribed on Jesus Christ’s cross. In the sixteenth century, this ideal became institutionally embodied in the foundation of trilingual colleges, including the Leuven Collegium Trilingue (1517) and the Paris Collège Royal (1530; see e.g. Papy 2018). In the wake of Hebrew, other so-called Oriental tongues excited increasing attention. Apart from Arabic and Syriac, two other Semitic tongues, the Persian language, of Indo-European descent, also attracted attention; it was even claimed to be closely akin to the Germanic branch of tongues (Van Hal 2018). The explorations of Columbus and others moreover brought new languages within the scope of European scholars, who were often baffled by their entirely different appearance (cf. Figure 4.2). Their interest was aroused mainly from a missionary point of view, with conversion of pagans as the ultimate goal. To this end, many clerics did their best to draw up grammatical

³ For an outstanding overview of the early modern linguistic horizon see Percival (1992).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

50

       

Figure 4.2 World map by Abraham Ortelius (1527–98) showing the expansion of the European horizon, 1570 Source: The Library of Congress. Public domain

codifications of languages such as Zapotec, a tongue spoken in what is now Mexico, an endeavour they undertook with varying degrees of success.⁴ At home, the vernacular tongues, up to then little more than conglomerates of related varieties, were in the process of being standardized. Intellectuals tried to promote one dialect as the norm or to design a common, supraregional variety, and to codify and elaborate it in grammars and handbooks. They aimed to arrive at an efficient and elegant speech form that was widely used and adequate for administrative as well as literary purposes.⁵ This new language could be linked to a centralized state, as in France, where the upgraded vernacular was extensively praised by Joachim du Bellay (c. 1522–60), but that was not necessarily the case, as in Italy, where du Bellay’s source of inspiration, Sperone Speroni (1500–88), made an urgent plea in favour of the vernaculars. Scholars thus put into motion the complex process of standardization, which occurred at different paces for each language. There were, for instance, strong centrifugal tendencies in politically diversified areas such as Germany (Burke 2004: 109). Sometimes there was an active and extensive debate about the variety that should serve as the basis for the common language (see e.g. Giard 1984: 46–7; ⁴ On the history of missionary linguistics see e.g. Hovdhaugen (1996); Zwartjes (2012); Zimmermann and Kellermeier-Rehbein (2015). ⁵ For the four processes of standardization, selection, codification, elaboration, and acceptance, Haugen (1966) is the classical reference. See e.g. also Joseph (1987).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    

51

1992: 206–9). The prototype of such a debate is the Italian language question (questione della lingua), which has been the subject of intensive study (e.g. Tavoni 1984). In other cases, the selection of a specific variety occurred more or less naturally as a result of the prestige of its speakers, as in the cases of South Eastern English and Île-de-France French, or of other historical circumstances such as the Reconquista in the case of Castilian Spanish (Burke 2004: 97). The elaboration of one variety as the norm could, however, entail the admixture of elements from other dialects. Standardization processes were catalysed by the commercialization of the printing press, which almost naturally raised questions about the lack of uniformity of vernacular tongues. The English printer William Caxton’s (c. 1415–92) sighs about the variability of English, for instance, are well-known (Caxton 1490: .i; Harris and Taylor 1997: 87–93; Van Rooy 2018b: 198). Apart from issues of standardization, increased travel activities also fostered curiosity about the regional variation of different languages. Finally, early modern scholars expressed a keen interest in the extinct ancient languages of Europe, such as Gaulish, the ancient Celtic tongue of what is now France (see Van Hal 2013/14), and the then enigmatic tongue of the Gubbio tablets, discovered in 1444 and identified by modern scholars as Umbrian, an Italic tongue related to Latin. Their rediscovery forced intellectuals to reflect extensively on issues of language classification. Especially Gaulish played a key role in the constitution of the dialect concept in the sixteenth century, as I will argue in the next chapter (Section 5.5). This brief sketch has highlighted the multilingual perspective of early modern Europe, and especially of its Republic of Letters (cf. Bots and Waquet 1997), in which being a polyglot was considered an asset. I can even go a step further and assert that a considerable number of scholars had what could be dubbed a multidialectal outlook. For the first time in history, it was not uncommon for an intellectual to be familiar with regional variation in two or more different languages. It seems, by the way, better to call this perspective a ‘multidialectal outlook’ rather than a ‘polydialect competence’ (by analogy to ‘polyglot’), since it was highly unusual for a scholar to truly master different dialects of different languages. The development of this new perspective stood in direct connection with the Renaissance rediscovery of the Ancient Greek language and literature and the subsequent anchoring of Greek studies in Western European scholarship. As a matter of fact, the Greek language provided an additional dialect context next to a scholar’s native one, triggering rudimentary and intuitive comparisons (Van Rooy 2020a: Chapter 8). What is more, the fact that the Greek dialects were the linguistic vehicles transmitting highly valued literary texts left open the possibility that vernacular dialects could be a study object of interest, too. When exactly did the Greek dialects become known to humanist scholars? In the first phase of the Renaissance interest in Greek, the prima grecità, the humanist pioneers Petrarch (1304–74) and Giovanni Boccaccio (1313–75) had tried to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

52

       

master Greek without much success. From the very end of the fourteenth century onwards, Greek studies succeeded in gaining a firm foothold in the north of the Italian peninsula, first in Florence and later also in other city-states. Numerous Byzantine scholars migrated there to teach the elements of Greek to Western students, an influx intensified by the Fall of Constantinople in May 1453. This seconda grecità ended around the turn of the sixteenth century, when Greek studies ripened and Western European scholars gradually took the wheel from their Byzantine masters. The takeover resulted in the terza grecità, with Greek studies spreading and booming in different parts of Western Europe, a development greatly stimulated by the medium of the printing press, as Elizabeth L. Eisenstein (1966: 46–7) has rightly suggested.⁶ Ever since, Western intellectuals have evinced an abiding interest in the Greek heritage. What is more important to my argument, however, is that only in this third phase did the Greek dialects become an inherent part of the curriculum of the student of Greek literature. In conclusion, the largely monolingual, Latin-focused culture of medieval Western Europe gave way to the multilingual and multidialectal perspective of the Renaissance, with humanist scholars enthralled by many different types of tongues, which they categorized roughly as ‘ancient’, ‘Oriental’, ‘exotic’, and ‘vernacular’.

4.2 Conrad Gessner, certified cataloguer and compiler The first scholar to make a serious attempt at systematically exploring the everwidening linguistic horizon was the Zürich humanist, physician, and compiler Conrad Gessner (see Figure 4.3).⁷ Gessner did so by listing alphabetically and describing, in his widely read Mithridates of 1555, a large number of languages and dialects of the then known world, ancient as well as contemporary, while sketching their mutual kinship along the way.⁸ His catalogue, written in Latin and named after the ancient Pontic king Mithridates  (135–63 ), who allegedly mastered more than twenty tongues (Figure 4.4), excellently embodies several aspects of his scholarly activities. Gessner was first and foremost an eager compiler, for which he was partly motivated by financial considerations (cf. Springer and Kinzelbach 2008: v, 39). But it was certainly not all money that drove Gessner. He was at least as much inspired by scientific interests and his passion for collecting, also apparent from his numerous botanical and zoological works. Indeed, the taxonomies used in these branches of biology probably inspired

⁶ See Lamers (2015: 72, 81–2, 166) for the grecità periodization followed here. ⁷ For biographical data see e.g. Fischer (1966); Wellisch (1984: 1–25); Springer and Kinzelbach (2008: 35–45). On the spelling of his name as ‘Gessner’ see Pyle (2000). ⁸ On Gessner’s linguistic work in general see Colombat and Peters (2009: 17–20).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.  ,    

53

Figure 4.3 Conrad Gessner, portrait by Tobias Stimmer (1539–84) Source: Wellcome Collection. CC BY 4.0

Gessner, ‘the Swiss Pliny’, to engage in a similar endeavour for the diversified phenomenon of human language (see Ogilvie 2006: 84; Leu and Opitz 2019). His bio-bibliographical project can likewise be viewed against this background. In his Universal library (1545–9), he tried to provide a complete overview of Latin, Greek, and Hebrew authors, their lives, and their works, again alphabetically organized, in this case by first name. As a result, Gessner has been dubbed both the ‘father of zoology’ and the ‘father of bibliography’ (Springer and Kinzelbach 2008: 27, 28), to which one might add the title of ‘father of language cataloguing’. How did Gessner catalogue and classify the speech forms of the world known to him? And more specifically, what concepts did Gessner employ in order to express different degrees of linguistic kinship? To answer these questions, I have to take a look at the very first pages of his Mithridates, where he introduced one of his key terms: dialectus. The presence of the word in Gessner’s work is not unexpected for two main reasons. On the one hand, it had been naturalized as a Latin term by the mid-sixteenth century and Gessner had been using it at least since 1543 (see Gessner 1543: .6–7). On the other hand, Gessner, trained as a philologist, was ‘especially attached to’ the Greek language and literature, from which the term had

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

54

       

Figure 4.4 Marble bust of the Pontic king Mithridates VI, first century  Source: Eric Gaba. CC BY-SA 2.5

been borrowed (Leu et al. 2008: 16). He had many Greek works in his library and was fluent in the language at age fifteen, in 1531, when he played a part in a Greek performance of a comedy by Aristophanes.⁹ He taught it to fellow students in Strasbourg, Paris, and Bourges to earn some money in the years 1532–3, and he was professor of Greek at Lausanne for three years in 1537–40.¹⁰ He moreover translated many Greek texts into Latin, including works by Porphyry and Johannes Stobaeus, and edited writings of Aristotle and Marcus Aurelius. He composed commentaries on Hippocrates and other Greek authors, and he suggested different allegorical interpretations of Homer’s epic poems (Leu et al. 2008: 16–17). Yet his devotion to Greek studies is perhaps most obvious from an early writing of his, his preface to the 1543 Greek–Latin lexicon he edited, in which he discussed ‘the utility and pre-eminence of the Greek language in every kind of studies’.¹¹ ⁹ Leu et al. (2008: 16–18), containing an incomplete description of the Greek works in Gessner’s library. ¹⁰ On Gessner’s knowledge and teaching of Greek see Springer and Kinzelbach (2008: 37, 38, 40). ¹¹ Gessner (1543: .2): ‘   tia Græcæ linguæ, in omni genere studiorum’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.  !     

55

4.3 Ad fontes! Gessner reads Clement of Alexandria In view of his wide reading in Greek literature, it is no surprise that Gessner’s initial definition of the word dialectus was entirely taken from an Ancient Greek work, the Stromata of the Early Christian Greek author Clement of Alexandria.¹² His Latin rendering of the relevant passage reads as follows in my English translation: A dialect, however, is speech exhibiting a peculiar mark or character of a certain place, or speech showing the proper or common character of a people. In fact, the Greeks distinguish five differences of dialects in their language, Attic, Ionic, Doric, Aeolic, and as a fifth the common.¹³

Gessner’s account seems rather dull at first sight and entirely in agreement with the Greek tradition, a typical instantiation of the humanist call to reach back to ancient sources: Ad fontes! But the opposite is true, as his rendering of Clement’s words is not as faithful as it might seem and, much more importantly, it offers a crucial insight into a key aspect of the metalinguistic apparatus of his Mithridates. What key aspect do I intend here? In order to answer this question, I need to compare Gessner’s Latin with Clement’s original Greek at some length. Table 4.1 puts Clement’s original Greek as it was printed in the Florentine editio princeps of 1550 alongside the first Latin rendering by the productive French translator Gentien Hervet (1499–1584), printed in the same city one year later, and Gessner’s Latin version in his Mithridates of 1555. Looking at the table, it immediately becomes apparent that the Protestant humanist from Zürich did not rely on the Catholic Hervet’s faithful but somewhat inelegant Latin version, nor did he translate Clement’s text verbatim. Instead, he repeatedly provided suggestive renderings and elaborate paraphrases of Clement’s Greek, resulting in a translation that was considerably longer and much less faithful than Hervet’s. According to Gessner, the seventy-two original tongues, hai genikaì diálektoi (αἱ γενικαὶ διάλεκτοι) in Clement’s Greek, should be viewed as languages rather than dialects. This interpretation is suggested by the fact that he translated the phrase as ‘dialecti (linguæ potius) communes’, ‘common dialects (languages rather)’. In what followed, he claimed that there were several dialects subsumed under each common genus (genus commune), that is to say, under each of the

¹² Clement of Alexandria, Stromata 1.21.142.1–4, quoted in Table 4.1. ¹³ Gessner (1555: 1–2): ‘Est autem dialectus dictio peculiarem alicuius loci notam seu characterem prę se ferens: uel dictio quæ propriam communémue gentis charactere[m] ostendit. Græci quide[m] dialectorum suę linguæ differe[n]tias quinq[ue] annotant, Atticam, Ionicam, Dorica[m], Aeolicam, & quintam communem’. For his interpretation of dialectus see e.g. already Metcalf (2013: 71); Peters (1970–1: .165–73; 1972: 260, 264–6); Colombat and Peters (2009: 30–2, 34); Stockhammer (2014: 333); Van Rooy (2016c: 133–4).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

56

       

Table 4.1 Clement of Alexandria’s Greek compared to Hervet’s and Gessner’s Latin versions Clement (1550: 126–7)

Hervet in Clement (1551: 37)

Gessner (1555: 1–2)

Εὔϕορος δὲ καὶ ἄλλοι πολλοὶ τῶν ἱστορικῶν, καὶ ἔθνη καὶ γλῶσσας πέντε καὶ ἑβδομήκοντα λέγουσιν εἶν[αι], ἐπακούσαντες τῆς ϕωνῆς Μωϋσέως λεγούσης. ἦσαν δὲ πᾶσαι αἱ ψυχαὶ ἐξ Ἰακὼβ πέντε καὶ ἑβδομήκοντα, αἱ εἰς αἴγυπτον κατελθοῦσαι. ϕαίνονται δὲ εἶν[αι] καὶ κατὰ τὸν ἀληθῆ λόγον αἱ γενικαὶ διάλεκτοι, δύο καὶ ἑβδομήκοντα ὡς αἱ ἡμέτεραι παραδιδόασι γραϕαί: αἱ δὲ ἄλλαι αἱ πολλαὶ ἐπὶ κοινωνίᾳ διαλέκτων δύο ἢ τριῶν ἢ καὶ πλειόνων γίνονται· διάλεκτος δέ ἐστι λέξις, ἴδιον χαρακτῆρα τόπου ἐμϕαίνουσα: ἢ λέξις ἴδιον ἢ κοινὸν ἔθνους ἐπιϕαίνουσα χαρακτῆρα. ϕασὶ δὲ οἱ Ἕλληνες διαλέκτους εἶν[αι] τὰς παρὰ σϕῖσι, ε ̄ʹ, ἀτθίδα, ἰάδα, δωρίδα, αἰολίδα, καὶ πέμπτην τὴν κοινήν: ἀπεριλήπτους δὲ οὔσας τὰς βαρβάρων ϕωνὰς, μὴ δὲ διαλέκτους, ἀλλὰ γλώσσας λέγεσθαι.

Euporus aute[m] & alii multi Historici, & gentes & linguas dicunt septuaginta quinq[ue] cum audiissent vocem Mosis dice[n]tis. Erant autem omnes animæ ex Iacob septuaginta quinq[ue], quæ descenderunt in Aegyptum. Videntur autem ex vera ratione linguæ seu dialecti, vt vocant, & sermones generales esse septuaginta duæ, ut nostræ tradunt scripturæ. Fiunt autem alię per communionem duarum vel trium vel etiam plurium dialectorum. Est autem dialectus, dictio quę loci proprium ostendit charactere[m], vel, dictio quæ proprium vel communem gentis characterem ostendit. Dicunt autem Gręci esse quinque apud se dialectos, Atticam, Ionicam, Doricam, Aeolicam, & quintam co[m]munem, quæ autem comprehendi nequeunt voces Barbaroru[m], non dici dialectos, sed linguas.

Euphorus & alij multi historicorum, gentes & linguas septuaginta quinq[ue] esse dicunt, inde nimirum impulsi quòd Moses scripsit: Animæ ex Iacob in uniuersum era[n]t septuaginta quinq[ue], quæ in Aegyptum descenderunt. Et sanè uidentur reuera dialecti (linguæ potius) communes duæ & septuaginta, ut in nostrorum etiam monumentis proditum reperitur. Reliquæ uerò multæ sub unum genus commune, quod duas aut tres plurésue dialectos contineat, referendę sunt. Est autem dialectus dictio peculiarem alicuius loci notam seu characterem prę se ferens: uel dictio quæ propriam communémue gentis charactere[m] ostendit. Græci quide[m] dialectorum suę linguæ differe[n]tias quinq[ue] annotant, Atticam, Ionicam, Dorica[m], Aeolicam, & quintam communem. Porrò uoces barbaras (quę scilicet à Græcis usurpantur) cum sint incomprehensibiles, non etiam dialectos, sed glossas uocari aiunt, Clemens Alexandrinus libro 1. Stromatéωn.

Note: Underlining marks the phrases most important for my analysis.

seventy-two original common languages. This idea was lacking in Clement’s Greek original, where, in fact, the emergence of new languages through mixture was discussed (Van Rooy 2013: 39). Gessner moreover rendered the Greek

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.       

57

word form dialéktous (διαλέκτους) by means of the very elaborate paraphrase ‘dialectorum suę linguæ differe[n]tias’, ‘differences of dialects of their language’. The Swiss humanist was clearly struggling with the meaning of the Greek term as it was used by Clement of Alexandria, which differed from his own understanding of dialectus. In fact, Clement interpreted the term diálektos as referring, very generally, to different tongues, presupposing no distinction between language and dialect. Gessner’s adaptations and paraphrases, however, doubtlessly betray a conceptual subsumption of dialect under language; indeed, he reserved the term dialectus principally for forms of speech that resorted under a lingua. Comparing Gessner’s translation to Hervet’s, the only one available at Gessner’s time of writing, the reader is at once struck by the fact that the term dialectus is treated with much less suspicion by the Frenchman. Indeed, unlike Gessner, Hervet did not read a language/dialect distinction into Clement’s work. What is more, Hervet elaborately translated hai genikaì diálektoi as ‘linguæ seu dialecti, vt vocant, & sermones generales’, ‘languages or dialects, as they’—the Greeks—‘call them, and generic tongues’. This phrasing suggests that Hervet interpreted dialectus as a Greek synonym for lingua and sermo, even though the length of his paraphrase indicates that he, too, struggled with Clement’s usage of diálektos. Anyhow, Gessner’s translation was much more interpretive than Hervet’s and influenced many later scholars, few of whom were aware that Gessner was, in fact, quoting Clement and not formulating his own ideas. This error has persisted even in modern scholarship, in spite of the fact that Gessner has clearly marked his source at the end of the long quotation.¹⁴ The modern editors of Gessner’s Mithridates, Bernard Colombat and Manfred Peters, are exceptional in correctly identifying Clement as Gessner’s source (2009: 31).

4.4 The meanings of dialectus according to Gessner Following the passage cited from Clement, Gessner elaborated on other meanings of the word dialectus he had encountered. After observing that the term could also simply mean ‘articulate speech’, as in Aristotle’s work, or ‘conversation’, as in Plato’s oeuvre, Gessner described its grammatical sense in his own words as follows: dialectus was ‘a property of a certain language, either in separate or in several words, a property by which it differs from the common or from other similar or cognate [dialects]’.¹⁵ The polysemy upon which Gessner remarked ¹⁴ See e.g. Mylius (1612: 195); Schottel (1663: 150); Schilter (1728: ); Michaeler (1776: preface); Fischer (1966: 75); Peters (1970–1: .165–6); Ahačič (2008: 96); Simon (2011: 168); Metcalf (2013: 71); Wilkinson (2016: 211–12). ¹⁵ Gessner (1555: 2): ‘Nos dialectum aliâs simpliciter sermonem siue orationem articulatam significare obseruauimus, uel ipsum in pluribus uerbis colloquium: aliâs (apud gram[m]aticos præsertim) linguæ alicuius siue in singulis siue in pluribus uerbis proprietatem, qua à co[m]muni uel

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

58

       

seems to have led to inconsistencies in his own usage of the term dialectus, especially in relation to lingua (cf. Swiggers 1997: 143). As a matter of fact, lingua is frequently used to designate varieties he also dubbed dialecti. The Aeolic dialect of Greek is, for instance, also referred to as Aeolica lingua, ‘Aeolic language’ (Gessner 1555: 5; cf. also fol. 12). Dialectus is moreover found in other, more general senses throughout his work, such as ‘language’ or ‘way of speaking’ (e.g. Gessner 1555: 67, 69; cf. Peters 1970–1: .166). In sum, Gessner, as is typical for a compiler, used the term eclectically, putting the different interpretations he knew into practice as it suited him and as he found them in his sources. From this perspective, it is difficult to maintain the somewhat anachronistic statement of B. Colombat and M. Peters ‘that dialectus very often appears in the Mithridates in the meaning attributed to it today’.¹⁶ The force of this explanation is moreover rather limited, as it is unclear what is meant by the modern interpretation of dialect. Gessner’s definitions were inspired by the Greek tradition, but his varying usage of the term was influenced by his eager reading of the works of contemporary scholars, too. These scholars included, most importantly, theologians such as his teacher Theodore Bibliander (1504/9–64), who occasionally seems to have implied that a dialectus historically derived from a lingua, and philologists such as Johannes Rhellicanus (1477/8–1542), who used the term dialectus to indicate that two related speech forms exhibited superficial linguistic variation only.¹⁷ In other words, Gessner did not encounter the term dialectus in only one isolated context but construed a kind of hybrid concept out of the various usages of the term he came across while reading ancient texts as well as contemporary scholarship. How did Gessner understand dialectus as opposed to lingua? His grammatical definition, quoted above, supplies us with obvious clues. He viewed a dialect as a specific form of a certain language showing differences both within individual words and on the level of syntax. Colombat and Peters (2009: 39) add that Gessner assumed that related dialects also varied slightly on the level of phonetics, phonology, and morphology. This idea might indeed be regarded as implicit in the examples he provided. One could, however, argue that these aspects are all covered by Gessner’s explicit but imprecise statement that dialects showed differences within individual words. For, in making this suggestion, he was no doubt thinking of the permutatio litterarum, ‘change of letters’, a phrase in which the reliquis similibus aut cognatis differt’. On this passage cf. Van Rooy (2016c: 133–4), whence the English translation provided here is adapted. ¹⁶ Colombat and Peters (2009: 32): ‘dialectus apparaît très souvent dans le Mithridate, dans le sens qu’on lui reconnaît aujourd’hui’. Cf. Stockhammer (2014: 333) for a similarly anachronistic analysis. ¹⁷ For Gessner as Bibliander’s pupil see e.g. Gessner (1555: 78); Amirav and Kirn (2011: ). See also Chapter 5.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.       

59

term littera covered both the letter itself and the sound it represented (cf. also Chapter 15, Section 15.2). This ancient Roman framework was designed for etymological purposes by Marcus Terentius Varro (116–27 ) and consisted in four main letter change operations: deletion (detractio), addition (additio), change (commutatio, mutatio, permutatio) of a letter, and a position switch (metathesis) of several letters.¹⁸ Varro’s framework, also outlined by Quintilian, might have been related to Greek pathology, which emerged around the same time but comprised much more letter-change processes (see Chapter 2, Section 2.3). Gessner’s dialect concept remained relatively vague. He nevertheless associated a dialect and its counterpart language with certain parameters. According to Gessner’s grammatical definition, a dialect differed not only from the common language—he no doubt had the Greek Koine in mind—but also from related dialects. It surely implied that this variation should be superficial, without touching the core of the language, and close genealogical kinship was clearly a prerequisite to label related speech forms dialecti. In fact, elsewhere, Gessner (1555: 21v) repeated the humanist commentator Heinrich Glarean’s (1488–1563) statement that it was justified to give two tongues of ancient Gaul that mutually differed only slightly one and the same name. Glarean’s (1538: 19) model was the Greek context, where Aeolic, Attic, Doric, and Ionic were all labelled ‘Greek’ because of their close similarities. Additionally, Gessner claimed that a number of languages now classified as Baltic, including Lithuanian and Old Prussian, ‘vary only by dialects’, a formulation suggesting a strong bond between those tongues.¹⁹ In short, Gessner supposed related dialects to differ from one another only superficially, which may surely be taken to imply that distinct languages displayed more substantial variation. It is unclear whether Gessner regarded a dialect as a regional variety of a language, particular to a certain nation. These two parameters figured in Clement of Alexandria’s definition of diálektos, in which no subsumption of dialect under language was presupposed, but they were not explicitly repeated by Gessner, who usually adopted a very ambiguous attitude towards his sources (Colombat and Peters 2009: 54). Both parameters might be implicit in his statement on the Ancient Greek language that ‘everywhere the populace has something of its own by which it differs from most others’, even though he principally had social variation in mind here.²⁰ The two parameters were, however, no doubt presupposed in the classification of the Ancient Greek dialects into

¹⁸ On Varro’s etymological method see e.g. Pfaffel (1981); Taylor (1996: 7–10, passim). On the permutatio litterarum see e.g. Ax (1987). ¹⁹ See Gessner (1555: 60): ‘ uel Liuorum & Lituanorum & Curorum lingua, eadem est quæ uetus Prussica. dialectis tantum uariant’. Cf. Gessner (1561: *.4, *.5). On this phrase see also Chapter 5, Section 5.5. ²⁰ Gessner (1555: 46): ‘habet enim ubiq[ue] uulgus suum aliquid proprium, quibus à cæteris plerisq[ue] differt’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

60

       

Aeolic, Attic, Ionic, and Doric, excluding the Koine, which, unlike the other four, Gessner did not tie to a Greek tribe or region. Instead, the dialects were subsumed under the Koine in his view.²¹ Mutual intelligibility does not seem to have been a criterion to speak of related dialects for Gessner. Lithuanian, for instance, constituted a unitary language which was divided into four dialects. Speakers of different dialects could not, however, understand one another very well, Gessner (1555: 59) asserted on the authority of the Polish geographer Maciej Miechowita (1457–1523), also known as Matthias de Miechow (Miechowita 1517: E.v). Gessner’s reliance on Matthias’s description of parts of Eastern Europe is also of interest for another reason. From Matthias’s work Gessner knew that the Lithuanian linguistic domain—linguagium, the term is Matthias’s—was ‘quadripartite’ (quadripartitum). Matthias, not yet operating with a clear-cut language/dialect distinction, had described the speech of the Yotvingians, that of the Lithuanians and Samogitians, that of the Prussians, and that of the Latvians as one and the same language, even though there was no mutual intelligibility, except among those having travelled through the different regions (see Stachowski 2013: 312–13). In an attempt at understanding what Matthias meant exactly, Gessner inserted the following parenthetic remark in his quotation from the geographer’s work: Matthias ‘seems to mean that it is one language—he calls it linguagium himself—but distinguished by four dialects’.²² In other words, Gessner relied on the recently emerged conceptual distinction in order to understand the views of a scholar to whom the abstract opposition was apparently not yet known. Gessner repeatedly associated a dialect with lower social classes, which emerges, for instance, from his collocation ‘vulgar dialects’ (dialecti vulgares). Dialectus is contrasted with the lingua communis, ‘common language’, the variety spoken by learned men (Gessner 1555: 46; 1561: *.4). This phrase clearly derived from the usual discourse on the Greek context. It is without doubt a loan translation of the Greek koine¯ ́ (κοινή), which was short for hē koine¯ ̀ diálektos, ‘the common way of speaking’. Gessner, however, found it more adequate to translate this elliptical designation as lingua communis rather than dialectus communis. He used the phrase not only when he spoke of the Greek language, but he also applied it to other languages such as his native German.²³ Gessner also suggested geographical centrality as a characteristic of a common language. This view implied for him that the further a variety was removed from the centre, the more it deviated from

²¹ For Gessner’s views on the Greek dialects see Peters (1970–1: .167–72, .188); Colombat (2008: 78–9); Colombat and Peters (2009: 39–40). ²² Gessner (1555: 59): ‘uidetur sentire linguam unam esse, linguagium ipse uocat, sed quatuor dialectis distinctam’. ²³ See Gessner (1561: *.4). Cf. Peters (1972: 266). For Gessner’s views on Germanic varieties see Metcalf (2013: 77–84). For his views on other dialect contexts see Peters (1970–1: .45–191 [passim]).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

. 

61

the common tongue (Gessner 1555: 58). Finally, in one passage, it seems that Gessner (1561: *.5) conceived of dialects as ‘descendants’ (propagines) of a common ancestral language (lingua principalis), thus adding a language-historical dimension to the conceptual pair, a usage still known today. Gessner did not evaluate dialectal diversity in general terms as a positive or negative phenomenon, even though he associated it from time to time with lower social strata. Yet he attributed certain qualities or deficiencies to individual dialects. This practice was absent from Roger Bacon’s work, and its presence in Gessner’s writings might be seen in relation to the ongoing standardization of vernacular tongues in his days and the competition between dialects that was part of it. Bavarian German was, for instance, labelled ‘utterly gross’ (crassissima), just like Doric Greek. Attic was the most elegant Greek dialect, because, Gessner suggested, it was closest to the common language (1555: 42, 46). However, in an earlier work, his 1543 praise of the Greek language, he had evaluated the Attic, Doric, and Ionic dialects all very positively (Gessner 1543: .6–7). The presence of an imperial court could make a specific dialect elegant, just like the study of literature, Gessner (1555: 39) argued when discussing the case of Brabantian Dutch. In short, the phenomenon of dialectal diversity per se was not cheered or deplored in Gessner’s linguistic work, but individual dialects did receive evaluative labels, both negative and positive.

4.5 Conclusion ‘I make Swiss a dialect of German’. Conrad Gessner actively subsumed certain speech forms as dialects under specific languages, and he did so independently from Roger Bacon, whose linguistic work was largely unknown in the sixteenth century. Gessner’s relating of linguistic entities in this manner should be framed within his broader intellectual programme, as he tried to throw the enormous amounts of information with which he and his colleagues were confronted into sharper relief. In his catalogues, he created rudimentary groupings of natural objects, involving not only the vast variety of animals and plants of the world but also the great diversity of human language. In the latter case, he made use of a distinction between language and dialect to do so. Even though Gessner did not define the conceptual pair very clearly, a close analysis of the definitions in his work and his terminological usage has made it possible to identify the parameters he associated with both concepts. Unlike Bacon, Gessner was not an isolated case. He was part of a wider intellectual trend and may even be seen as an early culmination of it. In order to understand his position in this complex development, I should zoom out in the next few chapters, which trace the sixteenth-century genesis of the language/dialect pair from a bird’s-eye perspective, while framing it in its intellectual context.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

5 Lingua and dialectus From synonymy to contrast

Conrad Gessner was an eager early adopter of the distinction between language and dialect, but he was most certainly not the first humanist to assume its existence and put it to use in his linguistic work. Instead, he followed in the tracks of scholars at their acme in the first half of the sixteenth century. Where exactly do we perceive the first traces of an assumed distinction between the two concepts? And under what circumstances did this occur?

5.1 Dialectus as a Latin word When in the 1470s the Italian humanist Niccolò Perotti (1429/30–80) was drawing up his Cornucopia, an impressive commentary on Martial turned encyclopaedia, the fact that certain languages show regional differences was widely acknowledged. In his posthumously published bestseller, Perotti pointed out that there was ‘sometimes diversity of speech within one language’, which he exemplified, not surprisingly, by means of the Greek tongue. What is remarkable, however, is Perotti’s remark that just as an individual language is, such diversity of tongue ‘is likewise called lingua’, a polysemy he did not problematize.¹ Perotti was, in other words, not yet operating with a concept of dialect clearly distinct from, and subsumed under, language. He did, most likely, not yet have a notion of a linguistic subsystem within a language. His assumption that variation within a language was not a universal phenomenon perhaps kept him from positing such a general distinction. Perotti stated, after all, that there was ‘sometimes diversity of speech’, a view doubtlessly informed by the widespread Renaissance idea that the Latin language was exempt from dialectal diversity (see Van Rooy 2018a on this idea). In the eyes of Perotti, who devoted most of his attention to Latin and accorded a merely auxiliary position to Greek (Charlet 2013), this lack of variation was obviously a positive property. As he was trained as a Hellenist,

¹ Perotti (1489: 85): ‘Quin etiam in vna lingua est aliquando sermonis diuersitas: quæ similiter lingua dicitur: ut in lingua græca: est lingua: quæ co[m]munis vocatur: & Attica: & Dorica: & Ionica: & Edica [sic]’. Cf. Trovato (1984: 208–9).

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     

63

Perotti must have been familiar with the Greek term diálektos. Its Latin form dialectus was for him, however, not an approved word, which is surely why he did not rely on it when describing diversity within one language, even though the Greek language with its dialects was the prototypical example for this phenomenon. So what is the background of Latinized dialectus? And what is its relevance for the genesis of the conceptual pair? The late fifteenth-century Latinization of the word and its subsequent naturalization as a Latin word were largely due to the false assumption of many humanists that the Roman rhetorician Quintilian had already borrowed Greek diálektos into Latin as dialectus in a much-read passage (Van Rooy 2019). For fifteenth- and early sixteenth-century editors of Quintilian’s Institutes of oratory preferred at 1.5.29 the variant reading dialectos, the accusative plural of Latinized dialectus, which is extremely rare in the manuscript tradition, above the more common reading dialectus. The latter form is the transcription of the Greek accusative plural dialéktous (διαλέκτους) into the Latin alphabet and is accepted in most modern editions of Quintilian’s text, because of its prevalence in the manuscript tradition. As a result of the preference of humanist editors, Latinized dialectus was mistakenly taken to be an ancient term by many humanists. Soon it was naturalized as a Neo-Latin word, even though most humanists did not regard it as a new constituent of the Latin lexicon. It quickly spread across the Republic of Letters, from Italy to Switzerland and Germany and thence to the rest of Europe. The borrowing of the term dialectus into Latin, the scientific lingua franca of the Renaissance, and its subsequent naturalization constituted a key intermediary step in the genesis of the conceptual pair. What is more, the fact that early sixteenth-century scholars started to contrast lingua to dialectus is the foremost symptom of the emergence of a language/dialect distinction, as I will argue in the next section. Yet before moving on to the contrastive usage of the terms lingua and dialectus, it may be useful to emphasize that fifteenth- and early sixteenth-century glosses and definitions of the word dialectus and its Greek original diálektos do not reveal a straightforward distinction between language and dialect. It is variously interpreted as idioma loquendi, ‘particularity of speaking’, proprietas linguae or loquendi, ‘property of tongue’ or ‘of speaking’, and genus loquendi, ‘kind of speaking’.² Arguably, one could interpret the phrase proprietas linguae in terms of a subsumption of ‘dialect’ under ‘language’. It seems, however, more plausible ́ that this reflects the traditional Greek definition of diálektos as glo¯ssēs idíōma, and, as I have suggested earlier (Chapter 2, Section 2. 2), this phrase should be understood as ‘property of tongue’, as a particular linguistic form different from others, and not as a linguistic entity intrinsically subsumed under a language.

² See e.g. Crastone ([1478]: s.v. διάλεκτος), Sabellicus (1490: 64), and Calepino (1502: s.v. dialectus), respectively.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

64

  

The very frequent synonymic use of dialectus and lingua in the early Renaissance seems to confirm my analysis. In a late fifteenth-century commentary on Suetonius, for instance, one reads that ‘the Rhodians speak in the Doric dialectus and lingua’.³ It is likely that this synonymy of lingua and dialectus was borrowed from the Greek language, in which glôssa and diálektos were usually used interchangeably, too (Lambert 2009: 21). To conclude, in the late Quattrocento and early Cinquecento, there was, much as it had been in antiquity and the Middle Ages, a direct awareness of the fact that individual languages could manifest themselves in different forms, which was expressed by often elaborate circumlocutions. Aeolic, for instance, was regarded as one manifestation of the Greek language. No early Renaissance scholar operated, however, with more abstract concepts categorizing one variety, a dialect, as clearly subsumed under another, a language. In other words, a language was not yet seen as an entity roofing different dialects, but different varieties were regarded as constituting different manifestations of one language—contrast Figure 5.1 with Figure 5.2 in this regard. In the first half of the Cinquecento, a change occurred, and the terms lingua and dialectus were increasingly contrasted in several different ways, which are all related and which all appear in the thought of both Conrad Gessner and several of his immediate predecessors and contemporaries. manifestation a (e.g. Aeolic) L

manifestation b (e.g. Attic) A

manifestation c (e.g. Koine)

N

G

manifestation d (e.g. Doric)

U

A

manifestation e (e.g. Ionic) G

E

Figure 5.1 Conceptualization of a language (e.g. Greek) before the early Cinquecento

language (e.g. Greek [Koine])

dialect a (e.g. Aeolic)

dialect b (e.g. Attic)

dialect c (e.g. Doric)

dialect d (e.g. Ionic)

Figure 5.2 Conceptualization of a language (e.g. Greek) emerging in the early Cinquecento

5.2 The key symptom: contrasting lingua to dialectus As Conrad Gessner tried to understand Clement of Alexandria’s comment on the seventy-two original tongues of the world, hai genikaì diálektoi in Greek, the Swiss ³ Beroaldo (1493: 138): ‘Rhodii Dorica dialecto et lingua loquuntur.’

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   :    

65

humanist intuitively sensed that, in his view, Clement had not used an adequate word to refer to these languages. Although rendering Clement’s diálektoi initially as dialecti, Gessner felt the need to add between parentheses ‘languages rather’ (‘linguæ potius’). This addition indicates that for the Swiss humanist, a lingua was something different from a dialectus. Other sixteenth-century scholars, too, opposed the terms lingua and dialectus, and even more explicitly and emphatically so. Such contrastive statements were nearly absent from earlier thought, and their large-scale appearance constitutes the single most compelling piece of evidence for the early sixteenth-century emergence of the conceptual pair. Consider two prominent examples. The Rotterdam humanist Desiderius Erasmus (c. 1466/7/9–1536) provided a very early instance when explaining the Greek term diálektos in the 1519 edition of his commentary on Acts (cf. Figure 5.3): To the Greeks, dialect is a property or fashion of tongue. For example, among the Greeks, even though there is one language (lingua), there are nevertheless five

Figure 5.3 Desiderius Erasmus holding a Greek book, by Hans Holbein, 1523 Source: Wikimedia Commons. Public domain

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

66

   dialects (dialecti), so that he who is versed in Greek can directly recognize whether a speaker is Attic or Doric, Ionic or Lacedaemonian.⁴

Erasmus was contrasting lingua to dialectus here, even though one could argue that the Dutch humanist still imagined the five dialects to be manifestations of the Greek language rather than entities subsumed under it. It is nevertheless not unlikely that he equated the Greek language with the Koine, under which he placed the five Greek dialects Aeolic, Attic, Doric, Ionic, and Lacedaemonian. Here, Erasmus did not mention Aeolic, but it emerges from his other works that he was well aware of its existence (e.g. Erasmus 1528: 169–70). There is a clearer example still in the work of the Spanish humanist pedagogue Juan Luis Vives (1492/3–1540), writing in the Southern Low Countries some fifteen years after Erasmus: There are some who introduce characters speaking each not only in their own language (lingua), but also in their own dialect (dialecto), as the sophist Lucian does in his Sale of creeds. This would be easy to do in those languages that have various dialects, as the Greek formerly had, and all vernaculars now per nation. In Latin, this could hardly happen, unless perhaps by means of the distinction in periods.⁵

The Spanish humanist’s contrasting of lingua and dialectus clearly concerned different types of linguistic entities. Different languages were distinct speech forms, whereas cognate dialects resorted under one and the same language, for which Greek and the vernaculars of Western Europe served as examples. Latin, however, was not regionally diversified, even though it could be distinguished into different periods. Here, Vives was thinking of the traditional division of the history of the Latin language into four periods: ancient, Latin, Roman, and mixed (e.g. Denecker 2017: 229–32). The contrastive usage of the terms lingua and dialectus did not emerge overnight. The late fifteenth to the early sixteenth century was the transitory period in which this gradually became a widely accepted given among humanist scholars, who increasingly encountered passages in which the two terms were contrasted or, in some instances, only seemingly so. A case in point is the Bolognese Hellenist Antonio Urceo (1446–1500), also known as Codro, who in 1485/6 briefly remarked of Greek that ‘it is difficult and hard to learn a foreign language with

⁴ Erasmus (1519: 202): ‘Græcis dialectus est linguæ proprietas aut species, uelut apud Græcos cum una sit lingua, quinq[ue] tame[n] sunt dialecti, ut qui Græce calleat mox possit agnoscere, Atticus sit qui loquatur an Doricus, Ionicus an Lacedæmonius’. ⁵ Vives (1533: .vii): ‘Sunt qui personas loquentes introducunt sua que[m]q[ue] non lingua modo, sed etiam dialecto, vt Lucianus sophista in auctione uitarum, quod pro[m]ptum esset facere in ijs linguis, quę dialectos habent varias, vt fuit olim Græca, & nunc vulgares omnes nationatim. In Latina ægre fieret, nisi forte ex distinctione ætatum’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

67

five dialects’.⁶ Here, a contrast between the two terms seems to have been presupposed. Several years later, however, probably in 1491/2, Codro explicitly treated lingua and dialectus as synonyms: ‘for there are five dialects or languages of the Greeks’.⁷ As Kozma Ahačič (2008: 97) has rightly pointed out, such usage could be dubbed a ‘terminological inconsistency’ from a modern point of view, even though it is more fruitful to understand it as a ‘terminological reality’, which must be taken into account in writing a history of the conceptual constructs underlying it. Codro’s case demonstrates that it can be misleading to rely on a single passage to claim that a language/dialect distinction was presupposed by its author, let alone his contemporaries. Even so, the conclusion remains that a number of leading humanists started to use the terms dialectus and lingua contrastively in the first decades of the sixteenth century. They may have been inspired to do so by their reading of assertions such as Codro’s first observation. It might be useful to borrow the concept of bridging context from linguistics to clarify the impact expressions like Codro’s may have had. A bridging context ‘allows for two interpretations of the linguistic item, the original interpretation and the new, pragmatically inferred interpretation’ (Gipper 2014: 87). In other words, a linguistic item x with meaning  can receive a new meaning  in specific contexts. After a while, meaning  can come to overshadow meaning  and even make it obsolete. Applying the concept of bridging context to dialectus, I can say that dialectus at first only had the generic meaning of ‘way of speaking; property of tongue’, but that it was sometimes used in such a way by certain scholars, including Codro, that another interpretive possibility came about: dialectus as ‘variety subsumed under a language’. A number of scholars reading Codro’s work would, by consequence, have been induced by the humanist’s usage of the term to read a contrastive interpretation into it. This originally context-bound reading resulted in a new meaning for dialectus, which did, however, coexist for a long time with the original, more generic meaning of the word. The direct and explicit opposing of lingua to dialectus is the key symptom of the emergence of the conceptual pair, but the subsumption of dialect under language can be recognized in several other ways, too, all related to this contrastive usage.

5.3 Developing the concept of common language In the first half of the sixteenth century, humanists developed the concept of common language, usually termed lingua communis and perceived as roofing ⁶ Codro (1502: .ii): ‘Atqui difficile et arduum est peregrinam linguam quinq[ue] dialectos habe[n]tem ediscere’. ⁷ Codro (1502: .v): ‘Su[n]t enim græcorum dialecti seu linguæ quinq[ue]’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

68

  

related dialects in the way suggested in Figure 5.2. I have illustrated how Conrad Gessner characterized a dialectus as a variety of speech differing from related dialects as well as from the lingua communis, a kind of supraregional variety used by educated speakers to communicate with one another.⁸ Whence did this concept originate? The new knowledge of the Greek language played a key role in this evolution. On the one hand, the Latin phrase lingua communis was inspired by the Greek elliptical designation hē koine¯ ̀ [diálektos], ‘the common [way of speaking]’. In Latin, substantivizing adjectives was more difficult because of its lack of an article; as a result, the noun lingua was added to communis to render the Greek phrase in Latin. At first, the Greek Koine was widely conceived as one of the five main manifestations of the Greek language. Gradually, however, it came to occupy a place clearly distinct from the other dialects in sixteenth-century grammatical works on Greek. The humanist Philipp Melanchthon (1497–1560), for instance, clearly did so in his popular grammar of the language, first published in 1518 (see Melanchthon 1518: a.i). On the other hand, the concept of common language was extracted from the Greek context and soon applied to vernacular tongues, despite the fact that many of them still lacked a widely accepted norm. On the Italian peninsula, for instance, Vincenzo Colli (c. 1460–1508), a poet also known as ‘Calmeta’, proposed as his answer to the questione della lingua, with explicit reference to the Greek context, a mixed common language based on various Romance varieties spoken at the Roman court. He did this in his Della volgar poesia, a book on vernacular poetry now lost to the ages, but we are well-informed about Calmeta’s ideas through Pietro Bembo’s (1470–1547) refutation of them.⁹ Examples from the Italian context could easily be multiplied for the early sixteenth century. In the second half of the Cinquecento, however, the design of such a common Italian variety was judged impossible by scholars such as Benedetto Varchi (1503–65) and Vincenzo Borghini (1515–80).¹⁰ For German-speaking territory, an early example can be found in Melanchthon’s Greek grammar: The speech common to all is called common language, just like, with us, there is a certain fashion of speaking common to the Swabians, Bavarians, and Ubians. Each of them nevertheless has its own particularities.¹¹

⁸ Chapter 4, Section 4.4. On the insufficiently studied history of the common language concept see Regis (2012), with specific reference to the Italian context. ⁹ Bembo (1525: –). On Calmeta’s views see e.g. Pozzi (1978: 92–3); Trovato (1984: 216–17). ¹⁰ See Varchi (1570: 269–71); Borghini (1971 [before 1580]: 338, 341–3). ¹¹ Melanchthon (1518: a.i): ‘Qui sermo comunis omnibus est, lingua comunis dicitur perinde ut apud nos est aliqua ratio loquendi communis Sueuis, Boiis, Vbiis, singulis tamen sui sunt idiotismi’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

69

An intuitive comparison of the Greek and his native German contexts led Melanchthon to oppose lingua communis to dialecti. It is, however, not yet entirely clear whether he interpreted the linguistic particularities of the Swabians, Bavarians, and Ubians as independent entities subsumed under a common language. He suggested rather that these segments of the Germanspeaking population had a common speech that showed an occasional linguistic particularity, without presupposing two separate linguistic levels. This analysis can, however, not be maintained for his views on the Greek context, as he clearly set the lingua communis apart from the Attic, Ionic, Aeolic, and Doric dialects subsumed under it. Melanchthon’s Protestant partner-in-arms Martin Luther (1483–1546) had a more specific linguistic entity in mind when talking about the German lingua communissima he adopted in his writings. In one of his table talks held between 1530 and 1545, Luther used this phrase, marked by the superlative ‘most common’, to characterize the German of the Saxon chancellery. This variety, he clarified, enabled him to communicate with the entire German-speaking area. It was, after all, employed by ‘all princes of Germany’ as the language of administration.¹² The wide usage of this variety was probably exaggerated by the reformer, but Luther’s activities did, in fact, lead to this form of German becoming the main fundament of the later German standard language. The concept of common language was, remarkably enough, first applied to linguistic areas that lacked a centralized government to speak of and where language standardization was not as straightforward as in more centralized states: the Italian peninsula and German-speaking territories (cf. Burke 2004: 109). Precisely this lack of political unity might have stimulated scholars of these regions, some of them not devoid of patriotic sentiments, to conjure up a common language by way of compensation. It no doubt also relates to the fact that Greek studies boomed early in those areas. Soon the concept of common language was extrapolated to vernacular languages other than German and Italian, too, most notably French, Dutch, Spanish, and English (see e.g. Zimont 2020 for French). Furthermore, already in 1531, Juan Luis Vives (1531: 79) promoted, in general terms, a common form of speech as a welcome instrument to overcome communication problems caused by dialectal diversity and to create a societal bond. In summary, one could say that a double evolution occurred. On the one hand, the Greek Koine came to be sensed by a large number of humanists as something fundamentally different from the other dialects of Greek they knew. On the other hand, scholars abstracted the concept of common language from the Greek context in the course of the early sixteenth century and gave it a wider application. ¹² Luther (1913: 639): ‘Ich rede nach der Sechsischen cantzlei, quam mutuantur omnes principes Germaniae’. For lingua communissima see Luther (1913: 640); Burke (2004: 101). Cf. Luther (1916: 78–9); Borst (1957–63: 1067–8).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

70

  

This is, however, too simplified an account, since the relationship between Greek and the Western European vernacular languages was not simply that of a model being applied to new contexts. As a matter of fact, scholars often relied upon their contemporary linguistic context to understand the Greek state of affairs, as, for instance, Petrus Antesignanus (1525–61) did in his commentary on the Greek grammar of the Brabantian humanist Nicolaus Clenardus (c. 1493/5–1542). Antesignanus, a Protestant French Hellenist who was also active as a Hebraist and musician, offered a window on contemporary French dialectal diversity, which he discussed for the student’s benefit: In order that the entire fact of this matter is put before your eyes and you fully understand what the grammarians mean by particularities (idiomata), I will uncover the whole affair for you with few words by means of the example of our French speech.¹³

In what followed, Antesignanus stated that almost all people wrote and spoke French in the area subjected to the French throne. Nevertheless, not everyone talked as neatly as the people at the royal court and in Paris. This fact made it easy to judge from which region a speaker originated, despite their efforts to avoid recognition. There were some exceptions, however. Some succeeded in forgetting their native dialect and in speaking neatly and purely the common French tongue. They even managed to enrich their French with some phrases and sayings derived from the dialects of the famous cities that were praised for their language. Dialects that had received general recognition could, in other words, be used as flowers adorning the common language (Antesignanus 1554: 11–12). The French humanist then proceeded to the Greek context. He claimed rather anachronistically that all Greeks tried to speak and write the Koine, the precepts of which were transmitted through grammatical writings. Everyone nevertheless retained particularities from their native dialect, as did, for example, Hippocrates, whose speech exhibited features from his native Ionic dialect. Antesignanus also mentioned Attic properties as the features par excellence for the enrichment of the Greek common language. Antesignanus thus clearly suggested parallels between French and Greek diversity. What is more, he used his native French context to clarify the Greek state of affairs, which entailed the retro-projection of a contemporary and familiar onto a past and unfamiliar context (cf. Van Rooy 2020a: Chapter 8). At the same time, however, his views on the Greek dialects were likewise partly determining his

¹³ Antesignanus (1554: 11): ‘Vt autem totum huiusce rei negotium tibi ob oculos proponatur, ac quid Grammatici per idiomata intelligant, penitus noscas, exemplo nostri Gallici sermonis tibi paucis rem omnem aperiam’ (translation adapted from Van Rooy 2016c: 131). What follows is mainly based on Van Rooy (2016c: especially 131–3).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

71

views on French diversity. After all, his idiosyncratic insistence on the idea that French had approved dialects that could lexically adorn the common variety appears to have been partly inspired by the existence of literary dialects in Greek. Figure 5.4 offers a schematic visualization of the way in which Antesignanus coordinated the French and Greek contexts with each other. conscious clarification Greek prototype

French diversity

unconscious fashioning

Figure 5.4 Antesignanus’s perception of the relationship between Greek and French variation Source: adapted from Van Rooy (2016c: 132)

Antesignanus’s intuitive comparison of French and Greek diversity should be seen in terms of a process of mutual reinforcement. These two dialect contexts were indeed similar to a certain degree, but Antesignanus suppressed their differences in order to stress their supposed similarities. The alleged basis for comparability was the idea that both Greek and French diversity were characterized by a common language and some generally approved dialects, by which the common variety could be enriched. Some renowned compatriots of Antesignanus, including Pierre de Ronsard (1524–85), made a similar suggestion, even though French dialects were never as prestigious as the canonical literary varieties of Greek (e.g. Ronsard 1565: 4–5). In sum, Antesignanus’s case suggests that the reason why the Greek Koine was increasingly regarded as an entity clearly distinct from the other Greek dialects cannot be exclusively sought in the Renaissance appropriation of Greek scholarship. It was at least as much fostered by the sociolinguistic reality of sixteenthcentury Europe. It seems, in other words, highly likely that even though most scholars explicitly mentioned the Greek Koine as their model common language— Luther being a notable exception—the existence of vernacular supraregional koines and emerging standard languages reinforced the tendency to set the Greek common language apart from the other Greek dialects. It is therefore safe to state that the concept of common language emerged as the result of accommodating the Greek and vernacular linguistic contexts to each other. This process involved two important steps: first, the intuitive comparison of the Ancient Greek language with a more familiar vernacular state of affairs, often with a model status for the Greek Koine; secondly, the abstraction of the concept of common language from the situations compared, often involving a suppression of the differences between them. As a result, it was now possible for scholars to conceptually superpose a common language on its related dialects. The common language

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

72

  

was closely associated and eventually identified with the linguistic norm, from which dialects deviated, and can, in this capacity, be viewed as a forerunner of the later concept of standard language.

5.4 Updating dialectus definitions I have shown that the definition Conrad Gessner (1555: 2) generically attributed to the grammarians explained dialectus as ‘a property of a certain language (proprietas linguæ alicuius), either in separate or in several words, a property by which it differs from the common or from other similar or cognate [dialects]’ (see Chapter 4, Section 4.4). Such further elaborations of dialectus definitions serve as yet another indication of the large-scale emergence of the conceptual pair in the course of the early sixteenth century. The addition of determiners such as alicuius (‘certain’) and unius et eiusdem (‘one and the same’) to widespread definitions à la proprietas linguae suggest that its authors understood dialectus as a ‘variety of a specific language’ rather than as a ‘property of tongue’ in generic terms. An early example is the definition of the humanist rhetorician Petrus Mosellanus (1493–1524), who defined the Greek word diálektos, admittedly still vaguely, as ‘a certain property, preserving the peculiar character of any particular language’ in his commentary on Quintilian’s Institutes of oratory.¹⁴ According to Mosellanus, a diálektos preserved the core characteristics of a language but had some particularity proper to it and distinguishing it from the language under which it resorted. He thus seems to have been reading a subsumption of dialect under language into the definition of diálektos formulated by the late Byzantine scholar Gregory of Corinth, which he was translating rather suggestively and interpretatively (cf. Chapter 2, Section 2.2, for Gregory’s definition). Mosellanus and Gessner were certainly not alone, and many additional instances could be cited, but let me limit myself to revealing examples from sixteenth-century England, where several prominent lexicographers expanded on traditional definitions of dialectus. Sir Thomas Eliot (c. 1490–1546), for instance, was the first Englishman to define Latin dialectus in his successful Latin–English dictionary, first published in 1538. Modelled on Calepino’s Latin lexicon, Eliot’s Dictionary explained the term as ‘a maner of speche, as we wolde saye diuersities in englysshe, as Northerne speche, Southerne, Kentyshe, Deuenishe, and other lyke’ (1538: ). Interestingly, he applied dialectus to his native context, at the same time opposing the level of the English language to that of its varieties. In another Latin–English dictionary, that of Thomas Cooper (1565: s.v. dialectus), a shorter but related explanation was offered: ‘A maner of ¹⁴ Mosellanus (1527: 42): ‘proprietas quædam, peculiare[m] alicuius linguæ particularis charactere[m] serua[n]s’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

73

speache in any language diuers fro[m] other’. Again, we see that the level of maner of speache was opposed to that of language but this time without exemplification. In a number of English–Latin dictionaries, similar definitions can be found. Richard Howlet (fl. 1552) picked up Eliot’s term diversitie and rendered the English phrase ‘Diuersitie of speache wythin one realme’ in Latin as dialectus. Howlet (1552: .iiii.) seems, in other words, to have interpreted dialectus as a regional variety of the language of a realm, much like Eliot before him, to whom he was indebted. It is, however, in a late sixteenth-century English–Latin lexicon that the two conceptual levels came to the fore most obviously. The Churchman John Rider (1562–1632) believed dialectus to be the Latin equivalent of the lengthy English phrase ‘A propertie of speech, diuerse from the rest of the same language’ (1589: 460). A language that was otherwise the same enclosed within itself several varieties that differed from one another. Looking outside the lexicographical tradition, I can remark that dialect was gradually naturalized as a valid English word from the late 1560s onwards. Early adopters of the word often still felt the need to gloss it in their works, suggesting that its acceptance as an approved English word took some time. Richard Stanihurst (1547–1618), for instance, discussed language diversity in his Description of Irelande in correlation to its geography: As the whole realme of Ireland is sundred into foure principall parts, as before is sayd, so eche parcell differeth very much in [the] Irishe tongue, euery country hauing his dialect or peculiar maner, in speaking the language. (Stanihurst 1577: 4)

Here, too, two distinct levels of linguistic entities were presupposed: the Irish language and the different ways in which it was spoken, its dialects. As in Howlet’s dictionary, the level of language was tied to a ‘realm’ by Stanihurst. In parallel with the expanding and updating of definitions, circumlocutions such as dialect(s) of language x began to appear with increasing frequency. These phrases no doubt implied a subsumption of dialect under language. It is, for instance, very obvious in Conrad Gessner’s statement that he ‘make[s] Swiss a dialect of German’, discussed in the previous chapter. Such phrases appeared most often when scholars made reference to the vernacular or Greek dialects, but they occasionally also surfaced in the context of the so-called Oriental tongues. For instance, the great similarity Syriac and Punic exhibited to Hebrew led the Italian humanist Angelo Canini (1521–57) to conclude ‘that they have to be called dialects of it’.¹⁵ For this insight, Canini was very much

¹⁵ Canini (1555: ã.4): ‘vt eius Dialecti nominari debeant’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

74

  

inspired by the Greek language and its dialects, which were the main subject of his book, entitled Hellenism.

5.5 The coining of a new phrase: ‘to differ only in dialect’ When Conrad Gessner commented on the languages now known as Baltic, he stated that the Livonians, Lithuanians, and Curonians had the same language as the ancient Prussians, to which he added that they ‘only vary in dialects’.¹⁶ In doing so, Gessner intended to emphasize the close relationship between these tongues, which were claimed to exhibit only superficial differences, namely on the level of dialect and—it seems implied—not on the level of language. Gessner did not coin this Neo-Latin collocation, which was, however, still very recent at his time of writing. How did it come into being and in what context?¹⁷ The phrase ‘to vary only in dialect(s)’ is likely to owe part of its origins to two widely used and frequently reprinted Renaissance grammars of Greek, composed in the first decades of the sixteenth century by the German-speaking authors Philipp Melanchthon and Jakob Ceporin (c. 1499/1500–25). In their grammars, certain parts of speech were said to ‘vary in dialects’ (variare dialectis). Ceporin, for instance, described in a paragraph entitled ‘Variation of dialects’ ‘how the prepositive and subjunctive articles vary in dialects’.¹⁸ The prepositive article corresponds to the Greek article ho, hē, tó (ὁ, ἡ, τό), whereas the so-called subjunctive article refers to the relative pronoun hós, he¯,́ hó (ὅς, ἥ, ὅ). In the paragraph on these article forms, Ceporin related how ‘the Dorians turn ē into ā also in the articles: hā timā ́ [‘the honour’]’.¹⁹ The interpretation of dialectal differences as accidental and superficial variation was not yet explicitly present in the phrase variare dialectis, even though later scholars may have understood it in this manner in bridging contexts. In other words, certain applications of the phrase variare dialectis, ‘to vary in dialects’, evoked a new, more specific interpretation: ‘to vary only in dialects’. Superficial variation was most certainly implied when in 1536 the SouthGerman Protestant pastor Andreas Althamer (c. 1500–c. 1539) stated that he considered two German words to be ‘varying somewhat in the German dialect’.²⁰ The earliest printed testimony of the phrase ‘to differ only in dialect’ identified thus far, however, dates to 1543, when it appeared in a posthumously published

¹⁶ Gessner (1555: 60): ‘dialectis tantum uariant’. ¹⁷ See Van Hal and Van Rooy (2017) for a more elaborate discussion of the genesis and uses of this phrase. ¹⁸ Ceporin (1522: .v): ‘V ’; ‘Vt articulus præpositiuus & subiu[nctiuus] dialectis uaria[n]t’. Cf. Melanchthon (1518: p.iii). ¹⁹ Ceporin (1522: .v): ‘Dores uertu[n]t α in η [sic pro η in α] etia[m] in articulis. ἁ τιμά’. ²⁰ Althamer (1536: 275): ‘uariante nonnihil germanica dialecto’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      : ‘    ’

75

commentary on Caesar’s Commentaries on the Gallic war, composed by the philologist and Neo-Latin poet Johannes Rhellicanus (1477/8–1542), born in Rellikon near Zürich: Hence, I do not doubt that Polybius, Pliny, Tacitus, and other writers who visited Germany and Gaul would have indicated that they had the same language, if they varied only in dialect one from another.²¹

Briefly, the phrase seems to have been transposed from early modern grammaticographic discourse on Greek to other textual genres such as language catalogues and especially commentaries on ancient historiographers in which the relationship between ancient Gauls and Germans played a prominent role.²² Significantly, this transposition involved the addition of the specification ‘only’ to the phrase ‘to vary in dialect(s)’. This development suggests that the term dialectus was undergoing a semantic specialization towards a more concrete meaning: ‘variety of a language differing only superficially from other varieties of the same language’, instead of the vaguer interpretations of ‘way of speaking’ and ‘property of tongue’. From the 1560s onwards, the Latin phrase appeared with increasing frequency. Its most common terminological guises appear to have been dialecto tantum variare or sola dialecto differre, with only the word dialectus as a fixed element in the expression. The influential historian and political philosopher Jean Bodin (1529/30–96), for instance, propounded in his 1566 Method for the easy comprehension of history that a wide range of Slavic peoples ‘use the same language . . . and differ only in dialect’, thus giving a very wide interpretation to the term lingua, which transcended political borders.²³ The Neo-Latin collocation, which served different purposes, proved to be very popular and was soon borrowed into several European vernaculars. It was adopted remarkably early in English. In 1566, the Jesuit John Rastell (1530/2–1577) had already stated that, in the Slavic tongue, ‘the difference consiste in Dialecte and proprietie only’, in which he may have been inspired by Bodin’s work, printed some months previously, even though he referred only to Gessner’s language catalogue in the margin (see Rastell 1566: 65, misnumbered as 75). A few pages further, several Germanic varieties were said to ‘differ but in Dialect’, even though they were not mutually intelligible (Rastell 1566: 68). Rastell was, however, not only the first author to use the phrase in English, but he also provided the first

²¹ Rhellicanus (1543: 73): ‘Proinde mihi dubium non est, quin Polybius, Plinius, Tacitus, & alij scriptores, qui Germaniam, & Galliam inuiserunt, indicassent eandem utrisq[ue] linguam fuisse: si dialecto tantum inter se uariassent’. ²² See Van Hal (2013/14) for humanist views on the relationship between Celtic and Germanic. ²³ Bodin (1566: 439): ‘sic enim audio Polonos, Bohemos, Rußios, Lithuanos, Moschouitas, Boßinios, Bulgaros, Seruios, Croatios, Dalmatas, Vandalos eade[m] Sclauorum vti lingua, quæ in Sca[n]dia vsurpatur, ac sola dialecto differre.’

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

76

  

printed instance of the English term dialect found thus far (Van Rooy and Considine 2016: 647–51). Rastell’s testimony suggests that the phrase ‘to differ only in dialect(s)’ originally meant something like ‘to differ only in particularity of tongue’. It may in other words have been inspired by a very traditional definition of dialectus, which did not presuppose a subsumption relation between dialect and language: proprietas linguae, ‘particularity of tongue’. The fact, however, that dialectus appeared very often in the vicinity of terms meaning ‘to differ’ and especially ‘only’ seems to have stimulated the contrasting of dialect with language. What is more, as I will argue in Chapter 8, it became customary towards the end of the sixteenth century to extend the phrase by introducing the ‘language’ pole into it, resulting in collocations such as ‘to differ not in the entire language but in dialect only’. In short, the widespread idea that two varieties ‘differed only in dialect’ not only presupposed but also reinforced an opposition of substantially varying languages to related dialects exhibiting only superficial differences. Rather surprisingly, Greek theorizing and the Greek dialect context did not play a fundamental role in the emergence of this phrase, even though it might have been indirectly inspired by early sixteenth-century discourse on Greek grammar and by the traditional definition proprietas linguae, a Latin calque of Greek idíōma glo¯ś sēs. The emergence of the collocation was principally urged by a need for mapping out the close relationships between different peoples, especially those that were lesserknown, such as the inhabitants of ancient Gaul. Tellingly, I have not encountered any instances of authors claiming that the Greek tribes differed only in dialect, most likely because such a claim would have been perceived as stating the obvious.

5.6 An intricate conceptual web The sixteenth-century establishment of the language/dialect opposition was not a straightforward process, since dialect interfered and interacted with other concepts such as idiom and style in an utterly complex and elusive way. In particular, the intricate relationship between dialect and idiom, in the sense of ‘the specific character or individuality of a language, the manner of expression considered natural to or distinctive of a language, a language’s distinctive phraseology’, surfaced in late fifteenth- and sixteenth-century definitions of dialectus (diálektos) and idioma, both terms with several different meanings.²⁴ The former word was frequently explained as idioma loquendi, ‘property of speaking’, while many authors defined the latter as proprietas sermonis, ‘property

²⁴ The definition of idiom is from the online Oxford English Dictionary, s.v. idiom (last accessed 8 April 2019).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

77

of speech’.²⁵ The definition proprietas linguae, ‘property of tongue’, was even used to explain both dialectus and idioma.²⁶ The terms were, in other words, often perceived as synonyms, or at least near-synonyms. One could, in fact, say that dialectus was an alternative terminological means to express the concept of idiom, which was not as clearly distinguished from dialect as it is in present-day linguistics. As I have argued in Chapter 3, Roger Bacon had already confused the two concepts in a similar manner in the thirteenth century. In the course of the sixteenth century, the word dialectus was singled out as the term most adequate to render the concept of dialect, ‘variety of a language’. The reasons behind this development are not entirely clear. This selection process moreover appears to have occurred undeliberately, as I will argue in the next chapter. It might be related to the fact that the term diálektos, unlike idíōma, had been explicitly and repeatedly associated with particular properties such as ‘specific to a certain tribe’ and—especially—‘specific to a certain place’ in Greek scholarship. This connection would have made dialectus and, later on, its vernacular equivalents more apt candidates to express the concept of dialect than idioma and its vernacular forms, which lacked a close association with such specific parameters. I would, however, not go so far as to claim that idioma was never used to convey the concept of dialect; it was just not the preferred option. The ongoing semantic specialization of the term dialectus did not, however, prevent a confusing usage of it. A remarkable case in point is Conrad Gessner’s teacher Theodore Bibliander, who operated with a quite blurred conceptual apparatus. In several cases, Bibliander seems to have been opposing dialectus to lingua. He stated, for instance, that ‘all dialects of the Greeks are called one Greek language’.²⁷ But elsewhere he confused the concepts of dialect and idiom, most conspicuously when he claimed: I see that some ancient grammarians find it pleasing to call the “idiom” of a language the form of speaking which emerges from the structure of words. For instance, what the Greeks say as “hépomai soi” [‘I follow you’: verb + dative], the Latins say as “sequor te” [‘I follow you’: verb + accusative]. And “mákhomai mákhēn”, ‘I fight a fight’, is called Attic diction. But the matter manifests itself more widely, which is both noticed by reasoning and established by authority. For John the Grammarian, who has written on the dialects of the Greeks, places the difference of tongues either in the entire word—for instance, what the Ionians say as “kárta” [‘very; exceedingly’], the other Greeks say as “lían”—or in a part of the word—the Aeolians, for instance, say “bróda” [‘rose’], the other Greeks “rhóda”—or in the affection, or passions, of a word—the Dorians, for instance, ²⁵ See e.g. Crastone ([1478]: s.v. διάλεκτος) and Calepino (1502: s.v. idioma), respectively. ²⁶ See e.g. Sabellicus (1490: 64) and Estienne (1531: s.v. idioma). ²⁷ Bibliander (1542: 58): ‘omnes Græcorum dialecti una lingua gręca dicuntur’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

78

   pronounce “pantôs” [‘at any rate’] with a circumflex accent, which the others pronounce with a grave accent as “pántōs”.²⁸

Likewise problematic for sixteenth-century scholars was the relationship between dialect and style, in the sense of ‘the manner of expression characteristic of a particular writer (hence of an orator), or of a literary group or period’, especially since philological interests came to influence interpretations of the dialect concept.²⁹ I have indicated earlier how Quintilian compared Latin literary registers to the Greek dialects (Chapter 2, Section 2.4). In the sixteenth century, the confusion of both concepts can be perceived in, for instance, the humanist Joachim Camerarius’s (1500–74) proem to his edition of Herodotus’s Histories of 1541. In his paragraph On the Ionic dialect, Camerarius stated that ‘the dialect, i.e. peculiar kind of discourse, of the Ionians has a great conformity with the Attic tongue’, after which he related where and by whom the Ionic dialect was spoken.³⁰ Both his paraphrase of dialectus as ‘peculiar kind of discourse’ and his characterization of Herodotus’s style as ‘elegant’ and ‘exhibiting a certain fluency particular to Ionic’ seem to betray a certain confusion of dialect with style.³¹ The ambiguous place of dialect in this intricate conceptual web had implications for its interpretations. As a matter of fact, the polysemous term dialectus was often used in such an unstraightforward manner that its meaning needed to be guessed from its immediate context, thus leaving ample room for each reader to interpret the word as it suited them. This interpretive freedom seems to have greatly contributed to, among other things, the idea that a dialect derived from, and was posterior to, a language, which became popular mainly after 1600 (see also Chapter 10). Such a diachronic conception of the conceptual pair was, however, already implicit in the usage of Conrad Gessner, who represented ‘the dialects of the Germans’ as ‘descendants of the same original language’.³² Whence did Gessner draw inspiration for this assumption? I argue that his reading of his teacher Theodore Bibliander’s work is likely to have stimulated his at times genealogical usage of the term dialectus, even though this interpretation did not necessarily correspond to Bibliander’s intentions. When discussing linguistic

²⁸ Bibliander (1542: 89): ‘Ιδίωμα linguæ dici loque[n]di formulam, quæ ex uerboru[m] structura nascitur, uideo placere ueteribus quibusdam gra[m]maticis, ut quod Græci dicu[n]t ἕπομαι σοι, Latini sequor te. Et Attica phrasis dicitur μάχομαι μάχην pugno pugnam. Sed latius rem patere tum ratione animaduertit[ur], tum authoritate comprobatur. Nam Ioannes Gra[m]maticus, qui de Græcorum dialectis scripsit, uel in tota collocat dictione linguaru[m] differentiam, ut quòd Iones dicunt κάρτα, reliqui Gręci λίαν, uel in dictionis parte, ut Aeoles βρόδα dicunt, reliqui Græcorum ῥόδα: uel affectione uocis, aut passionibus, ut Dores παντῶς circunflexo tono efferunt, quod alij graui tono πάντως’. ²⁹ See the online Oxford English Dictionary, s.v. style (last accessed 8 April 2019). ³⁰ Camerarius (1541: β.2): ‘Dialectus autem, id est, peculiare orationis genus, Ionum cum Attica lingua magnam habet communitatem’. ³¹ For the common transfer of textual properties to language varieties see Schlieben-Lange (1992). ³² Gessner (1561: *.5): ‘sanè Germanorum dialecti per uniuersum Rhenum à fontibus eius ad Oceanum usque, omnes eiusde[m] linguæ principalis ceu propagines sunt’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

79

monogenesis, for instance, Gessner’s teacher stated that God acted ‘so that out of one common speech several dialects emerged’.³³ Elsewhere, Bibliander offered a more extensive description of his monogenetic approach to language history: Since Hebrew is the primeval language, the others are generated and born out of it. Some of these degenerated farther from the original speech, such as the dialects of Japheth’s sons. Others kept closer to the first language, such as the dialects of Shem’s sons. Still others occupied a position in-between, such as the dialects of Ham’s sons.³⁴

Confronted with such passages, readers such as Gessner might have intuitively understood dialectus as a ‘variety derived from a chronologically primary language’, even though such an interpretation is simply impossible in other passages of Bibliander’s work, for example, where he observed that ‘the books of all Jewish writers are filled with various dialects: Hebrew, Syriac, Arabic, German, Greek, Latin, Italian, French, Spanish, and what not?’³⁵ Its meaning here remains unclear, but dialectus is surely not to be taken in a genealogical sense. Bibliander’s (1548: 107) discussion of the term dialectus does not provide any clues in this regard, as it is rather a state of the art of Greek usage of the term than an explanation of his own conceptions. Several meanings were listed—‘property of tongue’, ‘language’, ‘speech’, ‘rare word’, etc.—but he did not express his preference for any of them. Readers are left to themselves when it comes to determining what Bibliander precisely meant by dialectus in each context and are free to interpret the term as it suits them.

5.7 Conclusion In the years 1500–50, a distinction between language and dialect emerged, introduced primarily by scholars active in German-speaking areas, with a key role for scholars of Swiss origin such as Theodore Bibliander, Conrad Gessner, and Johannes Rhellicanus. They were motivated to do so by a fascination with the phenomenon of human language and its diversity, often connected with an interest in ancient history.

³³ Bibliander (1542: 52): ‘ut ex uno sermone co[m]muni, plures dialecti prodire[n]t’. ³⁴ Bibliander (1548: 142): ‘Siquidem Ebræa est primigenia, reliquæ ex ea propagatæ & genitæ sunt. Quarum aliæ longius degenerarunt à principali sermone, ut dialecti filiorum Iapheth: quæda[m] arctius adhæserunt linguæ primæ, ut dialecti filiorum Sem: quædam medio se modo habent, ut dialecti filiorum Cham’. ³⁵ Bibliander (1542: 82): ‘Omniumq́[ue] Iudaicoru[m] scriptoru[m] libri uarijs dialectis referti sunt, Hebręa, Syra, Arabica, Germanica, Gręca, Latina Italica, Gallica, Hispanica, & qua no[n]?’

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

80

  

Even if the emergence of the conceptual opposition was not reflected in a consistent metalanguage, it cannot be denied that lingua and dialectus were increasingly employed contrastively from the early sixteenth century onwards. The terminological distinction manifested itself in different ways. Apart from a direct and explicit opposition of the two words, sixteenth-century scholars also developed the concept of common language, lingua communis in Latin. This concept presupposed the existence of non-common varieties, and we indeed find that lingua communis was opposed to dialectus by prominent scholars such as Philipp Melanchthon. Other main manifestations of the lingua/dialectus contrast are the addition of determiners to lingua in definitions of dialectus and the coining of phrases such as ‘dialect(s) of language x’ and ‘to differ only in dialect’, which was usually taken to imply ‘and not on the level of language’. The contrastive usage of lingua and dialectus moreover reveals that several criteria came to be associated with the two concepts, albeit only loosely at first. Various scholars suggested, for instance, that related dialects exhibited superficial differences, whereas distinct languages were marked by substantial variation. It is, however, not easy to define a core dialect concept for the early sixteenth century. If, however, I were to formulate one, I could circumscribe dialect as ‘particular speech form subsumed under a certain language’. This definition would have the advantage of emphasizing a sixteenth-century development with major repercussions for the later history of linguistic thought and praxis: the conceptual subsumption of one linguistic entity, a dialect, under another, a language. Such a formulation nevertheless remains undesirable in my opinion, as it does great injustice to the versatility with which sixteenth-century scholars relied on the dialect concept to attain their intellectual and argumentative goals. The definition could moreover convey the misleading impression that dialect was a clearly distinct and uniform concept, but the opposite is true. Its elusive interference with other concepts as well as the polysemy of the term dialectus greatly hampered more concrete and better-defined interpretations of the concept in the first half of the sixteenth century. Scholars were, in other words, unable to develop a manageable concept allowing them to account for the minor differences they noticed within languages that were perceived as otherwise uniform. It is, finally, remarkable that the emergence of the conceptual pair followed the borrowing of Greek diálektos into Latin fairly quickly, which suggests that the borrowing of dialectus can be taken to forebode intellectual innovation and that scholars were looking for the necessary metalanguage to talk about languageinternal variation in the scientific lingua franca of the Renaissance. Lexical borrowing can in this case be viewed as paving the way for conceptual change. One key issue relating to the sixteenth-century genesis of the language/dialect distinction remains to be addressed: why did it occur as it did, and in which context?

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

6 Hellenism, standardization, and info-lust The genesis of the conceptual pair in context

A considerable number of sixteenth-century scholars suddenly began to assume that the Latin terms lingua and dialectus referred to linguistic entities of different orders. Which contextual factors contributed to the increasing contrastive usage of these two terms? And were humanist scholars consciously innovating, as Roger Bacon was in the thirteenth century?

6.1 A major pivoting point: the rediscovery of the Greek dialects The countless references to the Greek heritage in the two preceding chapters are no coincidence; the rediscovery of the Greek language and literature and, especially, of their formal plurality was a major factor in the emergence of the language/dialect distinction.¹ As a matter of fact, contrasting lingua and dialectus became common practice in the first place among scholars trained as Hellenists, specialists of the Ancient Greek language and literature. The term dialectus was by virtue of its origin closely associated with the Greek language and its different forms, at least initially. Indeed, in most available Greek texts, diálektos referred to a variety of the Greek language and also in the rare instances in which it occurred in Latin texts, it almost exclusively referred to forms of Greek. This semantic focus explains why dialectus was initially only part of the vocabularies of Hellenists, nearly always with reference to the Greek language. Where did the sixteenthcentury interest in the Greek dialects come from? And why did it emerge only then and not earlier, say, after the Byzantine diplomat Manuel Chrysoloras (c. 1355–1415) started teaching Greek in Florence from 1397 onwards? Despite the obvious dialectal diversity present in the literary Greek texts that were being rediscovered on the Italian peninsula, and despite the importance of Greek for the emancipation of the dialectally diverse Italian vernacular, Quattrocento Hellenists showed remarkably little interest in the phenomenon of

¹ A more detailed account of the early modern rediscovery of the Greek dialects is in Van Rooy (2020a), on which I draw here.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

82

, ,  -

dialectal variation.² Why was this the case? When Byzantine émigrés started teaching Greek in Italian city-states, they had to focus on the basics of grammar and were not in a position to trouble their students too much with the complicated matter of the dialects. That is not the entire explanation, however. The Byzantine teachers do not seem to have been experts in the issue either. In 1441, the Italian humanist Francesco Filelfo (1398–1481) lamented that even in Constantinople, where he had lived for several years between 1420 and 1427, no Aeolic was taught (Rotolo 1973–4: 88 n.4; cf. Botley 2010: 71–114). What is more, Greek treatises on the dialects seem to have been largely unknown before the end of the fifteenth century. As a result, texts in certain dialects, mainly Aeolic and to a lesser extent Doric, could not be easily studied during the Quattrocento. Does that mean that Greek dialectal features were not treated at all in the didactic texts composed in the fifteenth century by Byzantine migrants for their Western students?³ Some of them did indeed insert some remarks on the particularities of Attic, Ionic, Doric, and Aeolic. Chrysoloras, for instance, probably made limited use of Gregory of Corinth’s treatise on the Greek dialects when ́ ata (Ἐρωτήματα), Questions composing his grammatical handbook Erōte¯m (Botley 2010: 166 n.70). In fact, he mentioned idiosyncrasies of the canonical dialects throughout this work.⁴ Something similar occurred in grammars of later Byzantine refugees, for instance, in Theodore Gaza’s (c. 1398–c. 1475) four books on Greek grammar, composed around 1461–2.⁵ It seems that Byzantine scholars, as a general rule, eclectically referred to what they perceived to be the most notable dialectal properties. An exception to this tendency was Constantine Lascaris (1434–1501), who authored a treatise in which dialectal and poetical variation in the Greek pronoun was discussed more systematically.⁶ It was only in the grammatical work of Western Hellenists that more attention was devoted to the phenomenon of the Greek dialects. The different forms of the language were especially prominent in grammars from the late fifteenth century onwards. The Italian humanist Urbano Bolzanio (1442–1524) is a telling case, as he clearly struggled with the subject of the dialects and its presentation throughout the different editions of his handbook (Botley 2010: 36–40). The dialects occupied an ever more prominent position in publications on the Greek language during the first decades of the sixteenth century, at the exact same time when the conceptual pair emerged. Tellingly, this period witnessed the printing of separate handbooks for the Greek dialects. The first in its kind was a booklet entitled On the

² On the importance of Greek for the emancipation of the Italian vernacular see e.g. Dionisotti (1968: 51). ³ On these handbooks in general see Botley (2010), with further references. ⁴ See e.g. Chrysoloras (1512: 20). I refer to an early modern edition, as a modern critical edition is still lacking. ⁵ See e.g. Gaza (1495: г.v) for a Doric particularity. Pace Botley (2010: 17). ⁶ Lascaris (1502: z.viii). See Botley (2010: 26, 124, 175 n.272).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

83

Figure 6.1 The first separate edition of Adrien Amerot’s On the diverse dialects, 1530 Source: KU Leuven Libraries. Public domain

diverse dialects of Greek declinations, both in verbs and nouns (Figure 6.1; Amerot 1530). The text was printed in 1530 in Paris, at the atelier of Gérard Morrhy, and was an excerpt from a Greek grammar published in Leuven ten years earlier: the Compendium of Greek grammar by Adrien Amerot (c. 1495–1560) from Soissons, later professor of Greek at the Collegium Trilingue in Leuven (Amerot 1520).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

84

, ,  -

Not only did Amerot’s booklet become a bestseller, enjoying countless reprints during the entire early modern period (Hoven 1985: 5–19); it also initiated a tradition of printed manuals for the Greek dialects lasting up to the present day. In sum, the Greek dialects took up an increasingly visible space in the intellectual landscape of the early sixteenth century. Hellenists devoted ever more attention to the subject and started applying the terms they found in Greek grammatical works to other linguistic contexts as well, most often their native vernacular one. This terminological extrapolation first and foremost involved Latin dialectus, but also the Latin rendering of Greek koine¯ ́ as communis lingua, as I have contended in the previous chapter. The learned status of the Greek language was of central significance in this regard. Its variation could impossibly be neglected by grammarians and philologists and often led them to reflect, however rudimentarily, on their native dialect context, a development that would likely have been delayed without the renewed interest in the Greek heritage. In other words, the multidialectal outlook of sixteenth-century Hellenists triggered intuitive comparisons of different dialect contexts, usually the Greek and their native ones, leading them to develop an abstract conceptual distinction between language and dialect. The rehabilitation of the Greek heritage is not the entire explanation, however, and there are additional circumstances that considerably contributed to the emergence of the conceptual pair. What were these? I argue that the standardization of vernacular tongues and the information explosion with which humanist scholars were confronted were of profound significance as well.

6.2 From ancient Greece to Western Europe: standardizing the vernaculars In the Renaissance, the vernacular tongues of Western Europe were on the rise, partly as a result of the changing social order. The merchant middle class in particular prospered, who felt barely any affinity with Latin, the scholarly lingua franca, and used their native tongues to arrange their businesses. The promotion of the vernaculars at the same time entailed strong competition between them. Humanists such as Joachim du Bellay made a point of defending and elevating their native language, French in du Bellay’s case (cf. Burke 2004: 61–88). This linguistic competition should be framed within the humanists’ desire to set their own imagined national community apart from others.⁷ In doing so, humanists tended to put an emphasis on the borders between peoples and countries as

⁷ On the intensifying link between language and national communities in the early modern era see Joseph (2006: 22).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

85

well as on their own superiority in political, cultural, and moral terms. This kind of profiling was a reflection of their political engagement and aspirations. Put differently, emerging national sentiments and pride led humanists to accord higher status to their native vernaculars, contending that they excelled their neighbours also in linguistic terms. The renowned political scientist and historian Benedict Anderson (2006: 37–46) has pointed out in his well-known Imagined Communities the importance of print-capitalism for the emergence of national consciousness to the detriment of ‘the imagined community of Christendom’. Anderson connected this new consciousness not only with the development of humanist Latin and the rise of Protestantism, but also, and most notably, with the promotion and standardization of vernacular tongues. However, this national turn occurred less selfconsciously than in nineteenth-century nation-states, Anderson argued, and linguistic standards were in the Renaissance never imposed on a state’s entire population. Along similar lines, Caspar Hirschi (2012: 13, 41, 109, 126) has traced the origins of nationalism back to humanist discourse and sees an important role for linguistic factors. According to John E. Joseph (2004: 48, 107), ‘national language’ can even be regarded as a ‘dream of the sixteenth century’, with linguistic standardization being primarily motivated by a desire to emancipate the vernacular from Latin. In view of the emerging national consciousness in sixteenth-century Western Europe, it is no coincidence that this period witnessed the first attempts at systematically associating national groups with stereotypical characteristics (Leerssen 2007: 17). What were the implications of the competition among states and their national vernaculars? This linguistic race presupposed a clear demarcation of one’s own language from other vernaculars and went hand in hand with standardization processes. John Considine (2008a: 104) has aptly summarized it: ‘once one variety became recognized as a standard, it could become a symbol of membership of a wider community and of difference from other communities’. The competing vernacular tongues were identified with the selected varieties which were being codified and elaborated and to which all other dialects were subordinated (Joseph 2006: 29–30). The process of linguistic selection and especially the hierarchical subordination inherent to it no doubt catalysed the emergence of the conceptual pair, all the more since ‘standards, by inviting comparison of linguistic forms, enable conceptualization of non-standard varieties’ (Machan 2003: 98). The impact of two historical circumstances deserves to be emphasized when it comes to the rise of the vernaculars in relation to the genesis of the conceptual pair: the printing press (Figure 6.2) and the Reformation. The appearance of the printing press greatly contributed to the codification of vernacular standard languages, as it encouraged fixing their form (cf. Anderson 2006: 45). The printed word enhanced the formal stability of the vernaculars and intensified the contrast between uniform printed language and the great variability of spoken dialects.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

86

, ,  -

Figure 6.2 Woodcut showing sixteenth-century printing in action, 1568 Source: Wikimedia Commons. Public domain

The Reformation, in turn, led to the promotion of the vernacular, especially in Germany, where Luther became a bestselling scholar with, among other things, his German Bible translation (Padley 1988: 246). Coincidentally, Protestant scholars developed a greater interest in dialectal variation than their Catholic colleagues (cf. Van Rooy 2020c).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.  

87

6.3 Knowledge revolution: information explosion and info-lust Sixteenth-century scholars were moreover confronted with an ‘information explosion’ and concomitantly driven by a great ‘info-lust’, to use two phrases of Ann Blair (2010: 11–12). In other words, scholars faced large amounts of data and felt the desire to gather and organize them, for which they developed a wideranging apparatus of categories. The Renaissance collector’s mania was probably embodied best in the emergence of curiosity cabinets, rooms where all kinds of rare objects were gathered and which prefigured later museums (Figure 6.3; see e.g. the studies in Impey and MacGregor 2001). Languages, impossible to capture live in their spoken form during the Renaissance, were collected through the medium of written and printed texts in libraries. Assiduous attention to empirical data went hand in hand with a critical approach to ancient and medieval texts, with more and more humanists questioning the authority of classical authors by looking at what was right in front of their noses. The emergence of the conceptual pair can be seen as a product of the humanist info-lust, or rather as a side effect of it, since it was considered to be a self-evident given and not a newly conceived metalinguistic tool, as I will argue in

Figure 6.3 Natural history cabinet, 1599 Source: Wikimedia Commons. Public domain

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

88

, ,  -

Section 6.5. The distinction between language and dialect helped scholars to map out and make order in the enormous linguistic diversity that became ever more obvious in the sixteenth century. It is certainly no coincidence that the first extensive language catalogue ever composed, Conrad Gessner’s Mithridates of 1555, was published in the same period in which the language/dialect pair emerged and became common as a metalinguistic distinction. Gessner might even have conceptualized his book as a kind of linguistic curiosity cabinet, in which he put the conceptual pair into practice as a welcome means to group and categorize the phenomenon of language diversity (cf. Sergeev 2019).

6.4 The countability of language The historical circumstances sketched above are linked to a general feature of sixteenth-century linguistic thought: the tendency to conceptualize languages as clearly distinct, enumerable entities, as pointed out at length by the Dutch linguist Joop van der Horst (2008: 136–78). This conception surfaces, first, from the idea that different languages and the variation in each of them could be compared. Second, the ongoing standardization of, and competition among, vernacular languages involved the design of a clearly distinct status and profile for each of them. Thirdly, humanists worked on the basic assumption that languages were objects of study that could be categorized. While it is true that, before the early modern period, many scholars supposed on the authority of the Bible that there was a fixed number of languages, usually seventy-two, few of them were able to actually enumerate them all. Furthermore, counting languages was not an easy endeavour, owing to the fact that many languages consisted in continua of varieties without a recognized standard, let alone clear boundaries. This state of affairs changed in the sixteenth century and had a profound impact on the metalinguistic conceptual apparatus. Indeed, the emphasis on the idea that languages were countable and demarcated seems to have been an important trigger for scholars to develop the language/dialect distinction. While establishing ever clearer boundaries between languages, frequently associated with political borders, it became ever easier and ever more obvious to record variation within one language. The example of the Greek language and its five principal dialects stimulated the idea that vernacular dialects, too, were distinct entities and were therefore countable. As a matter of fact, several Hellenists started enumerating vernacular dialects from the early sixteenth century onwards, often assuming the same number for the Greek and the vernacular dialects. The French scholar Geoffroy Tory (1480–c. 1533) provided an early example, when he noted in his Flowery field of 1529, a manifesto for developing a French standard language:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

89

Our language is as easy to regulate and put in good order as the Greek tongue once was, in which there are five diversities of language, which are the Attic, the Doric, the Aeolic, the Ionic, and the Common tongue, which have certain differences among each other in the declensions of nouns, in the conjugations of verbs, in orthography, in accents, and in pronunciation, as a Greek author called John the Grammarian and several others treat and teach very extensively. Exactly that we would well be able to do for the tongue of the court and of Paris, the Picard tongue, the Lyonnais, the Limousine, and the Provençal. I would discuss some of their differences and correspondences, except that I do not want to be too long here and I leave it to those who are more experienced than me to devote themselves to it.⁸

Some humanists even pointed out that Greek had fewer dialects than their native tongue (e.g. Wolf 1578: 595 on German dialects). What is more, from the early seventeenth century onwards, scholars designed classifications of the principal dialects of a vernacular language, often inspired by the Greek example. The English schoolmaster and grammarian Alexander Gill (1565–1635), for instance, described his native context as follows in Chapter six of his English grammar, entitled ‘Dialects, in which the improper diphthongs are also treated’: There are six principal dialects: Common, Northerners’, Southerners’, Easterners’, Westerners’, Poetic. I neither know nor have heard all their particularities. Yet I will describe them as far as I can those I remember.⁹

Here, Gill enumerated the most important English dialects, and several elements suggest influence from early modern grammars of Greek, with which Gill must have been acquainted. Not only was he a distinguished Hellenist—he taught the language to John Milton—but he also presupposed the existence of a common English variety, dubbed dialectus communis, a designation no doubt inspired by the Greek phrase koine¯ ́ [diálektos] (Van Rooy 2020a: 23). The assumed countability of dialects went hand in hand with a terminological development, as dialects were increasingly accorded distinct labels based on ⁸ Tory (1529: –): ‘Nostre langue est aussi facile a reigler et mettre en bon ordre, que fut iadis la langue Grecque, en la quelle ya cinq diuersites de la[n]gage, qui sont la langue Attique, la Dorique, la Aeolique, la Ionique, & la Comune, qui ont certaines differences entre elles en Declinaisons de noms, en Coniugatio[n]s de verbes, en Orthographe, en Accentz & en Pronunciation. Co[m]mme [sic] vng Aurheur [sic] Grec nomme Ioa[n]nes Gra[m]maticus, & plusienrs [sic] autres traictent & enseignent tresamplement. Tout ainsi pourrions nous bien faire, de la langue de Court & Parrhisiene, de la la[n]gue Picarde, de la Lionnoise, de la Lymosine, & de la Prouuensalle. Ien dirois aucunes differences & accordances/se nestoit que ie ne veulx icy estre trop long, et que ie laisse a plus expertz que moy eulx y employer’. ⁹ Gill (1619: 15): ‘Dialecti: vbi etiam de diphthongis improprijs. Dalecti præcipuæ sunt sex: Communis, Borealium, Australium, Orientalium, Occidentalium, Poetica. Omnia earum idiomata nec novi, nec audiui; quæ tame[n] memini, vt potero dicam’. See Kökeritz (1938).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

90

, ,  -

toponyms, usually adjectives referring to a city or region, as the examples of Tory and Gill indicate. This heightened labelling activity might likewise have been partly inspired by the Greek tradition, especially since the Greek dialects were exceptional in that they received a name early on in antiquity (Adams 2007: 11).

6.5 A product of appropriation and subconscious adaptation I have been calling the emerging terminological contrast between lingua and dialectus a ‘symptom’ of the genesis of the conceptual pair language/dialect, which gives a clear hint about the way in which the distinction came into being. As a matter of fact, sixteenth-century humanists introducing the lingua/dialectus distinction were largely unaware of their scholarly achievement, since they seem to have assumed that they were simply using the term dialectus just as the ancient Greeks had used diálektos. They were, however, reading their own interpretation into the term, a process involving an opposition of dialect to language. The conceptual pair, in other words, came about in a subconscious way, almost by accident, as it were. This uncoordinated emergence stands in startling contrast to the case of the late medieval polyglot Roger Bacon, who was clearly aware that his lingua/idioma pair was an innovation. As I have contended, Bacon conceived it as a reaction against the all too liberal usage of the term lingua by his colleagues. There is a striking parallel with the history of the word dialect here. Just like the emergent opposition of lingua to dialectus, the Latinization of the Greek word diálektos was perceived as a realization of ancient, this time Roman, scholarship, even though fifteenth- and sixteenth-century humanists were actually reading the word dialectus into Quintilian’s work, but mistakenly so. Briefly, the appropriation of Greek diálektos went hand in hand with subconscious conceptual adaptation by sixteenth-century scholars. Inspired by Ashley and Plesch (2002), I understand appropriation here as a complex process of adopting and adapting an existing concept, either consciously or subconsciously, to one’s particular historical context and discourse. The subconscious nature of the appropriation of the Greek word makes it impossible to name the prôtos heurete¯ś , the ‘first inventor’, of the distinction between lingua and dialectus. This historical fuzziness reminds in some ways of the mid-fifteenth-century invention of the printing press.¹⁰ Let me, by way of thought experiment, consider three major points of comparison. First, just as printing with movable type was perfected by the combined efforts of Mainz printers, the conceptual pair did not emerge as the result of a one-man endeavour, but was the product of a specific social group: the early sixteenth-century

¹⁰ The information on the invention of the printing press is based on Rice and Grafton (1994: 1–10).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

91

Republic of Letters, and, as I have argued, especially those members that were skilled in the Greek language. It is, however, not impossible to name a number of key figures. For printing, those were Johann Gutenberg (c. 1395–1468), Johann Fust (c. 1400–1465), and Peter Schöffer (c. 1425–1502). For the language/dialect distinction, one could mention, among several others, Philipp Melanchthon, Juan Luis Vives, Petrus Antesignanus, and Conrad Gessner. The humanists pioneering the conceptual pair were, however, not as geographically concentrated as were the Mainz printers, but this is related to the international character of the Republic of Letters and the mobility of its members. Second, both the printing press and the language/dialect distinction display the typical Renaissance blend of old and new. On the one hand, the invention of printing was much indebted to the import of two Chinese ingenuities, block printing and paper, and to existing European techniques such as oil-based ink, which were combined and used in innovative ways. The conceptual pair, on the other hand, was the result of the humanist appropriation of Greek scholarship on the literary dialects and the concomitant adaptation of Greek definitions of diálektos. They subconsciously tailored the term to the linguistic realities of sixteenth-century Western Europe and their outlook on them. Thirdly, Gutenberg and his colleagues ‘considered printing only a new and particular kind of writing’, which reveals ‘[t]heir difficulty in freeing themselves from traditional conceptions’ and indicates that they were not aware of ‘the unique potentialities of their invention’, to put it with the words of Rice and Grafton (1994: 5). The printers’ lack of perspective has an intriguing parallel in the humanists’ unawareness of the fact that, instead of simply furthering Greek tradition, they were actually making a contribution to linguistic thought that was to exert an impact on scholarship lasting up to this day. It is, however, not only points of comparison which link the printing press and the language/dialect distinction. The production and distribution of printed books ‘turned intellectual work as a whole into a cooperative instead of a solitary human activity’, to quote Rice and Grafton (1994: 8) once more. This very fact no doubt accelerated the emergence of the conceptual pair, which cannot be attributed to one single humanist.

6.6 Conclusion The conceptual pair language/dialect emerged as the result of a complex interplay between several aspects of early modern scholarship and society: the rediscovery of the classical Greek heritage, the standardization of vernacular tongues, and the overwhelming amounts of new information with which scholars were confronted. The language/dialect distinction was especially prominent in the works of Italianand German-speaking scholars, who were the first to appropriate Greek studies

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

92

, ,  -

and for whom standardization was an acute problem due to the absence of a firmly established centralized state and, in the case of Germany, due to the Protestant insistence on adequately translating the Bible into the vernacular. The emergence of the conceptual pair was therefore most likely the result of an adjustment to new linguistic realities as well as of a growing interest in linguistic diversity on the part of sixteenth-century scholars, who made ambiguous use of the Greek heritage in doing so. It may also be stressed that the emergence of the conceptual pair was far from a straightforward process. Its most striking feature is that it was created not rectilinearly but almost by accident as it were, with scholars assuming they were just continuing Greek tradition. They simply took the existence of a language/ dialect distinction for granted, failing to recognize its absence from earlier thought. Additionally, conceptions of dialect as subsumed under language remained rather vague in the first half of the sixteenth century. As a result, it is unclear whether sixteenth-century authors conceived of a dialect as a specific subsystem of a language or as a loose collection of regional features within a language. I am inclining towards the latter answer, since dialectal variation was usually approached in a highly punctual manner, that is, principally in terms of a range of inconsistent letter changes. Whatever the case, it is clear that defining what a dialect exactly was seems to have been a frustrating endeavour for many sixteenth-century scholars. It was only after about 1550 that more elaborate interpretations began to appear, as I set out to argue in Part III.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

III

C O N S O L I D A T I O N BY E L A B O R A T I O N , 1 5 5 0 – 1650

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

7 Space and nation Greek definitions transformed

In the first half of the sixteenth century, Hellenist humanists developed a language/ dialect distinction. They did so independently from Roger Bacon and subconsciously, as they assumed they were just following in the footsteps of their great Greek examples. As a result, they did not feel the need to make explicit what they understood by the conceptual contrast, at least at the outset. After about 1550, their attitude began to change, and scholars increasingly formulated explicit interpretations of the distinction and especially of the dialect pole, thus consolidating their innovation by elaboration. What interpretations did they propose? It is this question which I address in the current and following chapters. Even though the focus is on the years 1550–1650, I make occasional digressions into other time frames to highlight certain developments that are of interest to the interpretation in question. I start this chapter with two conceptions with firm roots in Greek thought: dialect defined in terms of space and nation.

7.1 The spatial conception of dialect [A dialect is] a particularity of some one and the same language, which is different for different provinces or cities using that language.¹ This is how the Wittenberg professor of Greek, Hebrew, and mathematics Erasmus Schmidt (1570–1637) defined his usage of the term dialectus in his handbook on the Greek dialects, published in 1604. Apparently, Schmidt conceived of a dialect as a variety of a language that was in restricted use in regional terms, which implied that a language covered a larger area. When did this geographical interpretation of the conceptual pair emerge? It seems to have appeared only in definitions of dialectus formulated after 1550. An early mentioning of geographical restriction in a definition of dialectus in contrast with lingua occurred in Petrus Antesignanus’s widely read and often reprinted commentary on the Greek grammar by Clenardus. In this work of 1554, Antesignanus defined dialectus as ¹ Schmidt (1604: 4): ‘idioma unius & eiusdem alicuius linguæ quod diversis Provinciis vel Civitatibus eâ utentibus, est diversum’.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

96

   a certain property of speaking, by which the variety of speech is distinguished. This usually occurs between different regions of the same nation, because some speak a little differently from others and something proper from the native soil is engrafted in everyone.²

In contrast, Antesignanus assumed a common language to supersede this restrictedness in regional terms (cf. Chapter 5, Section 5.3). The absence of the geographical interpretation from earlier understandings of the conceptual pair probably relates to the fact that Clement of Alexandria’s Stromata were only published in 1550 in Florence. In this text, the Greek word diálektos was broadly defined as speech that bore the mark of a certain locality, tópos in Greek. Clement, however, had not yet subsumed diálektos under a language-like entity. Antesignanus did not mention Clement’s definition, but it is very likely no coincidence that Antesignanus’s geographical interpretation of the conceptual pair roughly coincided with the first printing of Clement’s work. The second author in whose work the spatial conception of dialect appeared did extensively refer to and even quote Clement’s Stromata, and this only one year after Antesignanus’s work was published. I am, of course, referring to Conrad Gessner, whose reliance on Clement I have analysed at greater length in Chapter 4. Not only the availability of Clement’s work in a printed form since 1550, but also especially the success of Antesignanus’s grammatical commentary of 1554 and Conrad Gessner’s Mithridates of 1555 allowed for the rapid diffusion of the geographical interpretation of dialectus as opposed to lingua. Both works were widely read and enjoyed reprints, especially Antesignanus’s work, part of a popular manual for the Greek language. In the sixteenth century alone, there were at least fifty-five editions of the book across Western Europe (Van Rooy 2016c: 134). It cannot, however, be denied that variation in a language was associated with locality even before the 1550s, albeit on a lower level of abstraction. Notably, the French philosopher Charles de Bovelles (c. 1479–after 1566) drew up an impressive account of Gallo-Romance regional diversity in his Book on the difference of vernacular tongues, where he tied the phenomenon to the prevalence of numerous local legal systems (coutumes; Bovelles 1533), a crucial step in recognizing linguistic difference, standardizing French, and centralizing royal power (Grinberg 2006; Kibbee and Keller 2018). Shortly afterwards, during one of his well-attended table talks in the 1530s or early 1540s, Martin Luther remarked that ‘Germany has so many dialects that people within thirty miles do not understand each other’.³ Luther clearly linked dialectal variation with geographical distance; every thirty ² Antesignanus (1554: 12): ‘sermonis quandam proprietatem, qua distinguitur loquelæ varietas, quę semper solet contingere inter diuersos eiusdem nationis tractus: cùm hi paulò aliter loquantur quàm illi, ac cuique proprium quidpiam è natali solo sit insitum’. ³ Luther (1919: 511): ‘Germania tot habet dialectos, ut in triginta miliaribus homines se mutuo non intelligant’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.     

97

miles there was a dialect so different that it could not be understood by a non-native speaker. Luther was, by the way, not alone in recurring to measures of length to underscore the immensity of dialectal diversity. Several other early modern authors expressed regional variation within a language in terms of miles or milestones. One can read another early example in the Greek grammar of the Spanish humanist Francisco de Vergara (†1545), who took great pains to explain Greek diversity by referring to his native Iberian context. In doing so, Vergara remarked that ‘indeed the differences of some letters and syllables occur with us every tenth milestone in whatever direction’.⁴ Still, it was only after 1550 that the spatial conception of dialect as opposed to language became part of theorizing on the term dialectus. Even though the idea was rarely made explicit in the remainder of the sixteenth century, its frequency steeply increased throughout the seventeenth and especially the eighteenth centuries. Indeed, formulations of the geographical interpretation became ever more substantial from the turn of the seventeenth century onwards, in which case Antesignanus’s and Gessner’s works were not infrequently quoted or paraphrased. For instance, Andreas Hoius (1551–1635), a professor of Greek from Bruges, cited part of Antesignanus’s definition in his Short dissertation on the seats and colonies of the dialects of the Greek language of 1620, even though Hoius did not do him the courtesy of mentioning him by name.⁵ The renowned grammarian Justus Georg Schottel (1612–76), in turn, resorted to Gessner’s Mithridates when defining dialectus in his extensive 1663 description of the German language.⁶ Schottel did explicitly refer to Gessner, but he did not notice that the definition he attributed to Gessner was actually the one Gessner himself had quoted, in Latin, from Clement of Alexandria (see Chapter 4, Section 4.3). It may be clear from the above that early modern scholars directed most attention to the dialect pole when discussing the locality of linguistic entities. Their insistence on the restricted geographical range of dialects implied that language covered a vaster space. Not everyone associated language with a large region, however. What is more, several eighteenth-century scholars closely linked it with the capital of a state, thus suggesting that the geographical coverage of a language was about as restricted as that of a dialect. For example, the unidentified author of the Encyclopédie article ‘Patois’, perhaps to be identified with Louis de Jaucourt (1704–79), claimed that ‘one only speaks the language in the capital’.⁷ ⁴ Vergara (1537: 210): ‘Literaru[m] vero & syllabaru[m] quarunda[m] differe[n]tię decimo quoq[ue] miliari quaqua versum nobis occurrunt’. ⁵ Hoius (1620: 95): ‘Sive, ut παχυμερέστερον explanemus, Sermonis quandam proprietatem, qua loquendi varietas distinguitur, quę inter diversos eiusdem nationis tractus plerumque solet existere’. See van der Aa (1867) and Sacré (1994) for biographical information, the latter providing the correct year of Hoius’s death. ⁶ Schottel (1663: 150): ‘Oder wie Gesnerus in Mithrid. sagt: Dialectus est dictio peculiarem alicujus loci notam seu characterem præ se ferens’. ⁷ Anon. (1765: 174): ‘On ne parle la langue que dans la capitale’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

98

  

In eighteenth-century England, one grammarian argued that ‘[a] Language is always esteemed to be spoke the purest in and near the Capital, as London, Rome, Athens, Jerusalem’ (Bayly 1756: 13). In such cases, language was obviously identified with good language as opposed to the corrupt dialects or, in France, the so-called patois. Indeed, there, the concept of patois, closely cognate to dialect, emerged during the seventeenth century. This word first denoted ‘way of speaking/acting, speech’, later ‘natural language, mother tongue’, and eventually, from the seventeenth century onwards, ‘local variety of a language, generally considered corrupt’ (Thomas 1953: 93–9, 105). Apart from carrying a strongly pejorative connotation, the term patois was associated with delimited areas, often provinces, just as dialect was.⁸

7.2 Two topical topoi Remarkably, the definition of the Greek philosopher Diogenes of Babylon, analysing diálektos as a linguistic expression with regional characteristics (Chapter 2, Section 2.1), was nearly absent from early modern theorizing. Gessner’s teacher Theodore Bibliander (1548: 107) was one of the few early modern scholars who quoted it, but he left out, surprisingly from a modern point of view, the part in which regional particularity was alluded to. Bibliander’s work is nevertheless of importance for intuitively linking dialectus, a term he used rather ambiguously, with a restricted geographical area. He stressed, on the one hand, that each city of a nation had something distinct in its speech (Bibliander 1542: 54) and, on the other hand, that individual provinces had their own dialects (Bibliander 1548: 19–20). In doing so, Bibliander provided early examples of two widespread topoi. First, dialects were often associated with the regional-administrative entity generally termed a province, or provincia in Latin, rather than with isogloss lines, which is very much a modern concept. In late antique and medieval writings, regional variation had already been connected with provincia, most notably by the Early Christian author Jerome. He noticed that, although it was the primeval tongue, Hebrew, too, was a diversified language, since ‘every province and region had its properties and could not avoid the native sound of speaking’.⁹ This link became widespread only during the early modern era, however. The French Hellenist Henri Estienne (1528/31–98), for instance, believed ‘French’ dialects such as Picard and Norman to be connected to the provinces of France (1587: 93). Even the Ancient Greek dialects came to be associated with provinces in a remarkably anachronistic fashion. In a Greek

⁸ See e.g. Besnier (1674: 23). For the pejorative connotations see Chapter 9. ⁹ Commentarii in evangelium Matthaei 1.1452–6: ‘unaquaeque provincia et regio habebat proprietates suas et vernaculum loquendi sonum vitare non possit’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.   

99

grammar published in 1752 in Mannheim, for instance, ‘Greek dialect’ was defined as ‘that mode of speaking which is different in the different provinces of Greece, which uses one language’.¹⁰ Yet the Greek dialects never corresponded to such administrative regions, since ancient Greece was rather a collection of citystates than a united area with a centralized government. Second, it was an early modern topos to state that language varied from city to city and even within individual cities. This idea had already been expressed by the Florentine author Dante, who claimed that there was great variation not only across different Italian towns, but even within cities such as Bologna (De vulgari eloquentia 1.19.3). It was repeated by several early modern scholars, most of them, however, independently of the Tuscan poet. Examples are legion. The Italian humanist historian Benedetto Varchi, for instance, remarked that a language could show differences in pronunciation from city to city and even from village to village. Varchi (1570: 89–90) compared this great diversity of speech with the individuality of everyone’s handwriting. Well before Varchi, the great Roman humanist-printer Aldus Manutius (c. 1449/51–1515) had asked himself: ‘What is the reason that every city has a peculiar tongue, that very often also in the same city they speak variously?’¹¹ Some scholars went even further, claiming that a language could show differences from house to house or from family to family (e.g. Trissino 1529: .ii). In sum, geographical entities of different sizes were involved in discussions of dialectal variation. This diverging geographical scaling of language not only suggests that early modern authors were aware of the gradational nature of linguistic diversity, but is also in line with a main characteristic of Renaissance conceptions of space in general: its fluidity. In fact, the apparent intangibility of dialectal variation might have contributed to the early modern sense of the ‘inherent instability’ of space, which was also connected to, among other things, the lack of fixity of state borders (Kümin 2015: 230).

7.3 Regional language variation: a universal phenomenon? And out of experience we see for display that all the languages of the entire world are somewhat different from locality to locality by a distance of no more than two milestones. If you would want to proceed farther, you will find such diverse tongues that they barely understand each other.¹² ¹⁰ Goldhagen (1752: 32): ‘Est autem Dialectus Græca modus ille loquendi in diversis Prouinciis Græciæ unâ linguâ utentis diversus’. ¹¹ Manutius (1496: *.ii): ‘Quid, quod unaquaeq[ue] urbs peculiarem habet lingua[m]· plerunq[ue] etiam in eodem oppido uarie loquuntur’. ¹² Haloinus (1978 [1533]: 55): ‘Experientiaque videmus ad oculum linguas omnes totius orbis de loco ad locum non amplius duorum miliarium distantia aliquantulum diversas esse; si longius procedere velis, tam diversas invenies, ut vix alius alium intelliget [sic]’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

100

  

This observation by the Flemish nobleman and diplomat Georgius Haloinus (c. 1470–1536/37) on linguistic diversity is one of the earliest explicit acknowledgements of the universality of regional language variation, if not the earliest. Haloinus formulated the idea in his Restauration of the Latin language, a work published in 1533 but written already in 1508, in which he expressed his preference for usage above grammar as the point of reference for correct language. His many diplomatic travels for the Habsburg administration brought him into contact with a wide range of different dialects and languages and heightened his awareness of the phenomenon of regional linguistic diversity and its universality (Matheeussen 1974: .15, .18, .172, .290; 1977: 310–11). The idea was absent from ancient and medieval considerations on language, most likely because Latin was widely regarded as a stable, unchangeable tongue. It is, for instance, telling that the Italian humanist Niccolò Perotti still held around 1478 that ‘there is sometimes’—so not always—‘diversity of speech in one tongue’, which he exemplified by referring to the Greek language (see Chapter 5, Section 5.1). Haloinus, however, believed that there was variation within Latin as well; this he motivated by pointing to the Patavinitas of the Roman historian Livy (59 – 17), an alleged regional particularity in Livy’s speech reported by Quintilian and claimed to be due to his Paduan roots (Van Rooy 2018a). The idea that Latin was not as uniform as was generally assumed had its roots in the Quattrocento, during which Italian humanists debated intensively about the possibility of class-based linguistic variation in ancient Rome (Chapter 2, Section 2.4). It was not until the early sixteenth century, however, that scholars put forward the universality of dialectal diversity. This development was surely related to the ongoing emergence, around that time, of the distinction between language and dialect, which was the product of intuitively comparing different contexts of regional diversity. The idea of dialectal diversity as a universal phenomenon was most widespread in Germanic-speaking territories. The renowned Bruges-born mathematician Simon Stevin (1548/9–1620), for instance, emphasized the universality of dialectal variation in his Discourse on the dignity of the Dutch tongue, written in Dutch and prefixed to his Principles of statics: ‘It is common to every language that in one region of the country it is spoken somewhat differently from another’.¹³ In the seventeenth and eighteenth centuries, especially from about 1650 onwards, the universality idea became a kind of refrain and even came to be presented as generally accepted knowledge: ‘It is well known that not only divers nations, but divers provinces or parts of the same nation who use the same language differ in their dialect and manner of pronunciation’ (Poole 1683: ii.1). The universality of

¹³ Stevin (1586: b.1): ‘Tis yder spraeck ghemeen datse inden eenen oirt des landts wat anders uytghesproken wort als op den anderen’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    ? 101 dialectal diversification led some authors to characterize it as an inevitable process, linked to the general variability of human affairs (e.g. Howell 1642: 144).

7.4 Nation or nations? The ethnic conception of the language/dialect distinction ‘Dialect’ . . . is speech peculiar to some people and this in the same language.¹⁴ Georg Pasor (1570–1637), a German Calvinist professor of theology and classical languages and one of the first specialists in New Testament Greek, associated the Greek term diálektos in his work on the nature of Biblical Greek with the entity of ‘people’. This ethnic component in Pasor’s definition comes as no surprise, since there was a close link in Greek scholarship between diálektos and éthnos, ‘nation, tribe’, probably an even stronger one than there was between diálektos and tópos, ‘place’. In line with Greek tradition, Pasor and others connected dialect with a nation, gens, natio, or populus in Latin, in the non-political sense of ‘group of people sharing the same ethnic background and living in the same area’. This coupling of dialect with nation could be taken to imply that language covered several nations, yet such an idea was rarely made explicit. Already in the early sixteenth century, diálektos and dialectus were frequently explained as ‘speech particular to a certain people’. In the majority of these instances, however, it is impossible to posit a subsumption of dialect under language. For example, in a basic Greek–Latin lexicon of 1514, attached to an edition of the Greek and Latin New Testament which appeared in Alcalá de Henares in Spain, diálektos was explained as ‘idiom, tongue, speech peculiar to whatever nation’.¹⁵ The compiler of the word list cannot, however, be claimed to have presupposed a language/dialect contrast, as he took diálektos to refer very generally to nationparticular speech, without opposing it to a language-like entity. French Hellenists active around the mid-sixteenth century seem to have already associated the entity of ‘nation’ with the conceptual pair. Petrus Antesignanus (1554: 11), for instance, emphasized that there were as many dialects as there were Greek nations, all of which used the Greek language with regional differences. Elsewhere, however, Antesignanus stated that dialectal variation ‘usually occurs between different regions of the same nation’, a contradiction he failed to resolve (see Section 7.1 for the full quote). The association of dialect with nation was also implicit in a German grammarian’s classification of the dialects of his native tongue into two principal groups: ¹⁴ Pasor (1632: 1): ‘Διάλεκτος . . . est sermo cuique populo peculiaris, idque in eadem lingua’. ¹⁵ Guillén de Brocar (1514: b.iii–iv): ‘Idioma.Ligua [sic].sermo proprius cuiusq[ue] nationis’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

102

  

just as the Germans are commonly and abundantly divided into two nations today, the High and Low Germans, likewise our language is generally divided into two particular idioms: High German and Low or Saxon German.¹⁶

The most interesting sixteenth-century testimony for the relationship of a language to nation is provided by the Italian humanist Benedetto Varchi.¹⁷ In his dialogue L’Hercolano, Varchi defined lingua as ‘nothing else than the speech of one or more peoples (popoli), which employ, in expressing their concepts, the same words in the same meanings and with the same accidents’.¹⁸ In what followed, he elaborated on each of the elements included in his definition and by consequence also on the phrase ‘of one or more peoples’. On the one hand, he introduced the element ‘of one people’, because the speech of a group of friends could also have particularities of its own. This specificity was not a reason to call it a lingua, however, as it should rather be called gergo, ‘slang, jargon’. On the other hand, it was possible that one and the same lingua was used by several peoples, as in the cases of Latin and Slavic, either as a result of natural developments or by accident. So, in the eyes of Varchi, who explicitly subsumed dialect under language when discussing the Greek dialects elsewhere in his work, a speech form was a language as soon as it was spoken by at least one people. He did not, however, explain what the ethnic coverage of a dialect was, according to him. The generic meaning ‘speech form particular to a certain nation’ of the term dialectus increasingly interfered with interpretations of the language/dialect pair from the end of the sixteenth century onwards. As this sense of the word was inherited from Greek definitions of diálektos, it is not surprising that one encounters the first clear indications of this interference, prefigured in the thought of Antesignanus, in handbooks for the Greek language and its dialects. For instance, the English antiquarian and Hellenist William Camden (1551–1623) stated the following in his 1595 Greek grammar: One generally reckons four dialects of the Greek language or principal forms of speaking according to the number of the principal nations of Greece, which vary from the common language in some respects, i.e. Attic, Ionic, Doric, Aeolic.¹⁹

¹⁶ Albertus (1573: .7): ‘Sicut Germani communiter et largè in duas gentes hodie diuiduntur, Superiores & Inferiores. Ita generaliter diuiditur lingua nostra in duo Idiomata, in Oberländisch/ vnnd Niderlendisch/oder Sächsisch Teutsch‘. ¹⁷ On Varchi’s linguistic thought and classification in general see Marazzini (1993: 267–73). ¹⁸ Varchi (1570: 87): ‘L, ò  ,  ` ,    ’, ò ` ,  , ò  ,     ,      ,  ’  ’. ¹⁹ Camden (1595: .1): ‘Qatuor numerantur Græcæ linguæ Dialecti siue loquendi formæ præcipuæ pro numero præcipuarum Græciæ gentium, quæ à linguâ communi in nonnullis variant, nimirum Attica, Ionica, Dorica, Æolica’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    ? 103 The ethnic interpretation of the conceptual pair soon appeared outside Greek handbooks as well. An intriguing case in point is the Flemish theologian and orientalist Johannes Drusius (1550–1616), who observed that dialectus was a term with different meanings and devoted particular attention to its ethnic conception. Drusius was led to do so by the Hebrew translation of a New Testament phrase containing the Greek term diálektos. The Hebrew word lāšṓn (‫)ָלשׁוֹן‬, ‘tongue, language’, was used to render diálektos, and Drusius took the Hebrew term to be equivalent to Latin lingua, ‘language’. This apparent mismatch between diálektos, ‘dialect’, and lāšṓn, ‘language’, called for some explanation, he must have thought. Drusius pointed out that dialectus had two principal meanings: the ‘wider’ interpretation as ‘speech particular to a people’, which was distinct from its ‘proper’ sense of ‘speech subsumed under a language’. He clarified the latter meaning of dialectus as ‘a peculiar manner of speaking in an otherwise identical language’ and further explained the former sense as follows: ‘distinctiveness of speaking of different peoples, whether they are of the same language or not’.²⁰ In other words, the broader ethnic meaning did not presuppose a subsumption of dialect under language, whereas the stricter interpretation did. Since the word dialectus remained polysemous throughout the early modern era, it should come as no surprise that many scholars employed it in both of these different but related meanings, often in highly equivocal ways. Not infrequently the ethnic meaning was irrefutably present, with a language/ dialect distinction being assumed at most, as, for instance, in the work of Claude de Saumaise (1588–1653), a French philologist and professor at Leiden university (Figure 7.1). Saumaise even overemphasized the ethnic parameter in defining diálektos, since he wanted to refute the existence of a so-called Hellenistic Greek dialect in antiquity. He did so in his fierce controversy with his great rival Daniel Heinsius (1580–1655), a humanist from Ghent who had presupposed the existence of this Hellenistic dialect.²¹ To speak of a Hellenistic dialect, Saumaise claimed, a Hellenistic tribe was needed, since Greek scholars had defined diálektos as ‘speech particular to a certain tribe’. As he followed Greek tradition rather closely, he did not subsume dialect under language in his own definitions. Saumaise (1643a: especially 6–7, 459–61; 1643b: 312) defined diálektos in the first place as a living, nation-specific variety, not necessarily subsumed under a language and not necessarily related to another diálektos genealogically (Considine 2010: 89). Or in his own words: a diálektos was a ‘distinctive ethnic and local mark’

²⁰ Drusius (1616: 76): ‘Est autem dialectus proprie in eadem alioqui lingua peculiaris loquendi modus, veluti apud Græcos cum una sit lingua quinque tamen sunt dialecti. Sed latiore significatione sumitur Act. 2. 8. & alibi. Sic dico, suus & proprius cujusque gentis sermo .i. diversorum populorum discrimen loquendi sive sint ejusdem linguæ sive non, vocatur διάλεκτος’. ²¹ For this much-studied controversy see e.g. de Jonge (1980: 32–4); Muller (1984b: 391–2); Considine (2010).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

104

  

Figure 7.1 Claude de Saumaise (1588–1653) Source: Wikimedia Commons. Public domain

in speech.²² It is, however, hard to deny that Saumaise assumed the existence of a language/dialect distinction elsewhere in his work, as, for instance, in the following statement: ‘The word glôssa seems more general, diálektos more specific about a “derivation” of whatever language’.²³ He moreover employed the expression non nisi dialecto differre, ‘to not differ except in terms of dialect’, which indicates that he associated dialect with accidental variation and language with substantial differences (Saumaise 1643a: 356, 368; Chapter 8, Section 8.1). In short, even though Saumaise doubtlessly presupposed a distinction between language and dialect, the relationship of the ethnic conception of dialect to his understanding of the conceptual pair remained unclear. This ambiguity must be viewed against the background of his scholarly modus operandi: chaotic and repetitive.

²² Saumaise (1643a: 61): ‘χαρακτὴρ ἐθνικὸς καὶ τοπικός’. ²³ Saumaise (1643a: 456): ‘verbum γλῶσσα generalius videtur, διάλεκτος magis speciale de παραγωγῇ cujusque linguæ’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.      ‘’?

105

The insistence on the interpretation of the conceptual pair in terms of ethnic coverage manifested itself differently in the thought of several of Saumaise’s contemporaries. They conceived of a dialect as a variety of a language that was the usual speech form native to a specific nation, whereas one needed to study to master a language. For example, the German grammarian Justus Georg Schottel remarked the following when elaborating on the distinction between everyday linguistic usage, covering also dialectal differences, and the German language: everyday usage is instilled from the cradle and acquired through itself, but the language is learned through nothing else than artificial instruction and the necessary zeal and reflection.²⁴

As a consequence, a common language was sometimes seen as ethnically unmarked. The German diplomat and scholar Caspar Schoppe (1576–1649), for instance, declared: if someone who otherwise recognizes all dialects, would hear someone else speaking in that [sc. the common German language], he would certainly approve of the speech itself, but would not be able to determine to what nation he belongs.²⁵

Schoppe added that the German common language was easily learned in Speyer, a city in the south-east of modern Germany, and at the imperial court, where people from all over Germany flooded together and dialectal elements were avoided in order not to be exposed to ridicule. In other words, dialect was bound to a nation, whereas language was neutral in terms of nation, an interpretation miles away from modern politicized conceptions of the distinction.

7.5 Towards a political interpretation of ‘nation’? From the mid-seventeenth century onwards, language was increasingly associated with the entity of the state and its inhabitants, sometimes termed nation, which began to receive a more politically coloured meaning. Indeed, language became linked more closely to state structures and was increasingly seen as an important symbol of the distinctive character of a particular state. This politicization of

²⁴ Schottel (1663: 168): ‘Weil der altages Gebrauch von wiegen an eingeflösset/und durch sich selbst angenommen; Die Sprache aber/mit nichten anders/als durch kunstmessige Anleitung un[d] erforderten Fleiß und Nachsinnen/erlernet wird’. ²⁵ Schoppe (1636: 51): ‘si quis ea loquentem audiat, qui ceteroqui omnes Dialectos agnoscit, sermonem quidem ipsum probet, nequaquam tamen, cuius ille nationis sit, constituere possit’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

106

  

language seems to have happened first in England and later on in France, too, and is reflected in linguistic works of the time. For instance, in the preface to his dictionary The new world of English words, Edward Phillips (1630–c. 1696) defined dialect as follows: a Dialect is but the self same Language, spoken in several Provinces of the same Nation, with some small difference; as the pronouncing of a vowel either broader or finer, or some little variation of a word or syllable. (Phillips 1658: b.3)

Here, Phillips’s association of language with nation was no doubt inspired by his native English context, both linguistic and political. He referred to the kingdom of England and its provinces, which were tied, respectively, to the English common language and its dialects. In a similar vein, the lemma for nation in the first edition of the dictionary by the Académie française, the French Academy, assumed the existence of one and the same language for a nation: N. . . . All inhabitants of one and the same state, one and the same country, who live under the same laws and employ the same language etc.²⁶

This development no doubt related to the advancing linguistic standardization and the ever-increasing political centralization occurring in Western European sovereign states such as England and France from the sixteenth century onwards (e.g. Trudeau 1992; Rice and Grafton 1994: 111–24). More generally, the association of vernacular languages with nations was a product of the Renaissance, during which political borders became increasingly determined by linguistic factors and humanists started to feel the need to distinguish their own nation clearly from others (Hirschi 2012: especially 104–79; Chapter 6, Section 6.2). However, in the majority of the early modern source texts, the ethnic interpretation of the language/dialect distinction, with a language covering several nations in the non-political sense and a dialect only one, is more prominent than the association of language with nation as inhabitants of a political state and of dialect with the provinces of such a state. One might therefore even assert, albeit cautiously, that the emphasis on the ethnic interpretation of the conceptual pair deconstructed to a certain extent the national identities invented by humanist scholars, especially since speakers of a certain dialect were very often clearly demarcated as a separate linguistic group by early modern scholars.

²⁶ Académie française (1694: .110): ‘N. . . . Tous les habitants d’un mesme Estat, d’un mesme pays, qui vivent sous mesmes loix, & usent de mesme langage &c.’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.   :  ?

107

7.6 Space and nation: diagnostic criteria? The interpretation of the conceptual pair in terms of ethnic coverage was closely associated with the spatial conception of dialect, discussed earlier in this chapter. Indeed, a nation was intrinsically tied to a certain locality in the eyes of many early modern scholars defining dialect. Yet, although present in many definitions, the geographical and ethnic conceptions of the language/dialect distinction were usually not put into practice as diagnostic criteria, that is to say, as litmus tests to determine whether a specific linguistic variety was a dialect or a language. What is more, in cases where these interpretations were employed to determine whether a linguistic variety deserved a certain label, it was not to argue whether a speech form was either a language or a dialect but rather to deny dialect status to it. I have already mentioned how Claude de Saumaise relied on the ethnic meaning of diálektos to refute the existence of a Hellenistic dialect; there was no Hellenistic nation or land, so there cannot have been a Hellenistic dialect of the Greek language. Let me provide an additional example to elucidate my point. Certain early modern Hellenists had posited the existence of a dialect particular to Greek poets. Soon several of their colleagues invoked the locality of dialect or its belonging to a specific tribe to challenge the existence of this Greek poetical dialect. Two eighteenth-century manuals for the Greek language offer excellent examples of this usage. In a handbook for the Greek dialects published in Nuremberg in 1782, one can read the following: Besides, the poetic dialect, usually added to these four dialects, cannot be called a dialect properly speaking, as it is rather a certain kind of speech not particular to a nation, but to a certain class of writers only.²⁷

A similar statement is found in a Greek grammar published two years earlier in Leipzig, which rebutted the existence of the poetical dialect by means of the geographical interpretation of dialectus: ‘Some turn the liberty the poets take in their verses into a dialect and call it poetical dialect. Such a dialect, however, presupposes a poetical city or region’.²⁸ Even though in these cases the spatial and ethnic conceptions of dialect were applied as touchstones in determining whether a linguistic variety was a dialect, this decision occurred without putting the subsumption of dialect under language into practice.

²⁷ Facius (1782: ): ‘Quae praeterea his quatuor Dialectis vulgo additur Poetica, proprie Dialectus dici nequit, cum dictionis potius sit quoddam genus, non genti, sed scriptorum tantum ordini cuidam proprium’. ²⁸ Haas (1780: 67): ‘Manche machen die Freyheit, deren sich die Poeten in ihren Versen bedienen, zu einem Dialekt, und nennen ihn dialectum poët[icam]. Ein solcher Dialekt aber setzet eine poetische Stadt oder Landschaft voraus’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

108

  

7.7 Conclusion Greek definitions of the term diálektos left their mark on early modern conceptions of the language/dialect opposition in two principal ways. First, from about 1550 onwards, dialect was increasingly understood as a variety restricted in terms of space vis-à-vis the language subsuming it. This geographic conception probably resulted from the availability in print of Clement of Alexandria’s Stromata, which contained several important definitions of diálektos. The geographical interpretation of the conceptual pair was, however, not actively used to distinguish a dialect from a language, unlike in modern linguistics (see e.g. Hinskens et al. 2005: 1). Still, it remains a fact that the geographical interpretation of the conceptual pair frequently occurred in definitions of the term dialectus. Second, a dialect was claimed to be particular to a certain tribe or nation, implying that a language covered several different tribes or nations. This ethnic interpretation of the conceptual pair became widespread only in the seventeenth century, well after the geographical meaning, and it is unclear why this delay occurred. Before that, the ethnic interpretation of the term diálektos/dialectus, already present in Greek scholarship, barely interacted with the, at that time still loose, semantics of the conceptual pair. Just like geographical coverage, the ethnic conception was not actively and consciously used as a criterion in determining whether a linguistic variety was a language or a dialect. In short, even though the majority of early modern scholars assumed that dialects had a limited geographical and ethnic coverage as opposed to a language, in practice they relied on other criteria to determine the language or dialect status of a linguistic entity.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

8 Aristotle’s legacy Substance, accidents, and mutual intelligibility

The impact of the Greek tradition is noticeable not only in the spatial and ethnic conceptions of dialect as subsumed under language. More indirectly, it can be perceived in other emergent interpretations of the conceptual pair, too, for instance in the following formulation by the renowned Swedish poet and philologist Georg Stiernhielm (1598–1672), founded on the notions of substance and accidents from Aristotelian philosophy: ‘Languages differ among each other in substance, in their foundation as it were, but dialects differ in accident’.¹ The present chapter investigates the widespread early modern idea that distinct languages exhibited substantial differences, whereas related dialects were only superficially variegated, and its consequences.

8.1 Extending the ‘to differ only in dialect’ phrase The substance/accidents interpretation was already prefigured in the thought of Roger Bacon (see Chapter 3) and was implicit in the phrase ‘to differ only in dialect’, first coined in the 1540s (see Chapter 5, Section 5.5). It was, however, only in the course of the seventeenth century that it was explicitly formulated as a diagnostic criterion. Before looking into this evolution, however, I should briefly consider its philosophical background. There is no doubt that the traditional distinction of Aristotelian ontology between substance in the sense of ‘fundamental essence’ and accidents as ‘non-fundamental properties’ constituted the intellectual backdrop for this interpretation of the language/dialect distinction. This conception of it may therefore well be dubbed the ‘Aristotelian criterion’. It must be understood against the background of the prevailing Aristotelianism in the arts curricula of many universities in the first half of the early modern period (Schmitt 1983; Brockliss 2003). Aristotle (Figure 8.1) was, however, never explicitly mentioned by scholars discussing the criterion, which is surely due to the fact that the substance/accidence opposition had become received knowledge. This silence resembles a process called, in the sociology of science, ‘obliteration

¹ Stiernhielm (1671: c.4): ‘Linguæ inter se substantia, ceu subjecto; Dialecti vero Accidenti differunt’.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

110

 ’  

Figure 8.1 Bust of Aristotle, after an original of about 330  Source: Wikimedia Commons. Public domain

by incorporation’, during which ideas become so established that the author first expressing them is no longer cited (Merton 1968: 27–8, 35). How did the Aristotelian criterion become explicit? To answer this question, I have to take a start in the early seventeenth-century Low Countries, where there was a great upsurge in interest in linguistic diversity (see especially Van Hal 2010a). In 1605, the Leiden history professor Paulus Merula (1558–1607) drew attention to the following research lacuna concerning the ancient tongues of Gaul: What that difference in language was, whether it was in dialects only or in the entire language, so that the Aquitanians, Celts, and Belgians each had a different language, has not yet been distinctly uncovered by anyone, at least as far as I know.²

² Merula (1605: 419): ‘Quod in Lingua discrimen illud fuerit, an in Dialectis solum, an in tota Lingua, ut ea alia fuerit Aquitanis, alia Celtis, alia Belgis, nemo adhuc, quod ego quidem sciam, diffinite aperuit’. See also Chapter 10, Section 10.3.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   ‘    ’ 

111

Here, the phrase ‘to differ only in dialect’ was extended so as to include the conceptual counterpart of language, too. What is more, Merula further determined the term lingua by means of the Latin adjective totus, meaning ‘whole, entire’, clearly contrasting dialect to language. Extending the phrase in this manner was principally a seventeenth-century development, with the Low Countries as its focal point. Some forerunners of this evolution can, however, be found in the late sixteenth century. The Scottish humanist historian George Buchanan (1506–82), for instance, ‘believe[d] that the Scots did not differ from the Britons in the entire speech but rather in dialect’ in ancient times.³ A little further Buchanan expressed his astonishment that even now after so much time, with the languages of neighbouring nations adulterated and changed for a large part by the arrival of so very many peoples, the Brits still differ among each other not so much in the entire speech as in particularity and dialect.⁴

Such expansions of the phrase ‘to differ only in dialect’ are a sign that scholars were moving towards an explicit formulation of the Aristotelian criterion. The Dutch pastor Abraham Mylius (1563–1637), the first early modern scholar to systematically treat the themes of linguistic diversity, change, and kinship, in his book Belgian language, provided an even clearer example of this tendency. Consider, for instance, the following statements of his: Even though indeed the Chaldean language, by the progression of time, has considerably changed in dialect from Hebrew, yet, as far as substance and root are concerned, it has remained the same.⁵ It can happen that that first distinction of languages [sc. the confusion of tongues at Babel] did not so much occur in respect to the substance of words as, for the most part, in respect to dialect and pronunciation.⁶ As a matter of fact, even if the words display some variation of dialect from our tongue, yet the substance is still the same, except in a few words, which time has worn out, as it has happened in the Greek and Latin language, and as it happens in all languages.⁷

³ Buchanan (1582: 19): ‘Sed nec Scotos a Brittonibus toto sermone, sed dialecto potius discrepasse arbitror’. ⁴ Buchanan (1582: 20): ‘nu[n]c etiam tanto post tempore tot gentium aduentu adulteratis, & magna ex parte mutatis vicinarum nationum linguis adhuc Britanni non tam sermone toto, quam proprietate, & dialecto inter se discrepe[n]t’. ⁵ Mylius (1612: 85): ‘Etiamsi vero Chaldæa lingua, progressu temporis, dialecto non parum sit ab Hebræa mutata, tamen, quod ad substantiam & radicem attinet, mansit eadem’. ⁶ Mylius (1612: 85): ‘Fieri posse, ut illa prima discriminatio linguarum non tam circa dictionum substantiam, quam ut plurimum circa dialectum pronuntiationemque facta sit’. ⁷ Mylius (1612: 151): ‘Nam etsi in vocibus dialecti quædam sit varietas a nostra, tamen substantia est adhuc eadem, nisi in paucis, quas ævum detrivit: ut factum fuit in lingua Græca & Latina, & fit in omnibus linguis’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

112

 ’  

For these views, Mylius was inspired by the close kinship among the Ancient Greek dialects (cf. Mylius 1612: 90, 152). Put differently, even though the Aristotelian criterion was an early modern innovation, it was nevertheless inspired by Greek philosophical categories and sometimes corroborated by the prototypical example of Greek linguistic diversity. Aristotelian categories were, however, not the only means used to express this conception of the language/dialect distinction. To refer to the substantial nature of language-level differences, some scholars used the image of ‘foundation’ (fundus), whereas others claimed that the various dialects of one language agreed in the ‘root’ or ‘radix’.⁸ Family relationships were occasionally also used to stress the superficial differences among dialects (e.g. Saumaise 1643a: 19–20). Towards the middle of the seventeenth century, the Aristotelian criterion became part of more general reflections on the conceptual pair. The extravagant Berlin-born orientalist Christian Ravis (1613–1677) offered the following analysis of closely related language varieties in his Sketch of Hebrew orthography and analogy of 1646, published in Amsterdam: Languages that are generally considered different yet connected through all essential elements, even though they apparently are very distinct by some and indeed many various and innumerable accidental elements, are nevertheless dialects to each other, and they constitute an integral language and a perfect body.⁹

In stating this idea, Ravis was one of the earliest scholars, if not the earliest, to formulate the criterion in such explicit terms. After him, it became more common to do so. Not unsurprisingly, it first occurred in the work of scholars active in England, where another major linguistic work of Ravis, his Generall grammer for the ready attaining of the Ebrew, Samaritan, Calde, Syriac, Arabic, and the Ethiopic languages, was printed in 1650. In this book, one could read similar statements (e.g. Ravis 1650: 44–5, 156). In the 1656 dictionary of Thomas Blount (1618–79), for instance, dialect was defined as follows: Dialect (dialectus) is a manner of speech peculiar to some part of a Country or people, and differing from the manner used by other parts or people, yet all using the same Radical Language, for the main or substance of it. (Blount 1656: s.v. dialect)

The Aristotelian criterion soon also appeared outside England in the work of Georg Stiernhielm, who had met Ravis in Sweden not long after the publication of ⁸ For ‘foundation’ see e.g. Schrieckius (1615: a.4). For ‘radix’ see e.g. Shelford (1635: 239). ⁹ Ravis (1646: 42): ‘Linguæ vulgò æstimatæ diversæ, per omnia tamen essentialia junctæ, licet aliquibus & quidem multis varijsque & innumeris accidentalibus apparenter distinctissimæ, sibi invicem tamen dialecti sunt, & linguam aliquam constituunt integram, corpusque perfectum’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. :    ? 113 the latter’s Sketch (see Chapter 12). However, even after the mid-seventeenth century, the criterion was usually taken for granted. Indeed, relatively few authors included it in their definitions. The primarily implicit nature of the criterion is in line with the fact that early modern scholars assumed the conceptual pair to be an obvious given, rather than an innovation of their own, as I have argued in Chapter 6.

8.2 Mithridates: polyglot or not? The Aristotelian criterion was eagerly put into practice by early modern scholars. I have discussed in Chapter 5 (Section 5.5) how the coining of the phrase ‘to differ only in dialect’ was related to the sixteenth-century interest in the poorly documented languages of ancient Gaul and Germany and their interrelationships. Let me consider here another discursive use, or rather abuse, of the criterion. Overcoming extant linguistic diversity was a difficult enterprise for early modern scholars. They resorted to the uniform Latin tongue, the common and widely understood scientific language, as an antidote to this historical reality. Another, later solution consisted in attempts at designing a universal language that would make other languages obsolete (e.g. Wilkins 1668). A more difficult and timeconsuming method was learning different languages. Indeed, a polyglot competence was a source of admiration, whereas a lack of it was often held up to mockery, certainly in philological circles. For instance, the alleged ability of Mithridates , ancient king of Pontus, to converse with the twenty-two tribes he ruled in their native tongues was widely known and applauded.¹⁰ Perhaps the best testimony to this appraisal is the fact that Conrad Gessner named his language catalogue after him. However, the language/dialect distinction and its interpretation in terms of substance and accidents made it possible to detract from a polyglot’s competence by emphasizing the alleged dialectal relationship among the tongues mastered. In fact, the nobleman Georgius Haloinus had already suggested in the early sixteenth century that king Mithridates either only knew some phrases of those twenty-two languages or spoke languages that ‘were neighbouring and quite similar, as experience teaches that this occurs in almost all regions’.¹¹ Similarly, the historian Paulus Merula considered it unlikely that the Pontic king spoke twenty-two ‘original languages’ (linguae primigeniae), implying that he had mastered different dialects of a restricted number of original languages. Merula (1605: 209) argued in this context that speaking different varieties of Germanic or Romance did not grant one the title of polyglot, multi-linguis in

¹⁰ See Pliny the Elder, Naturalis historia 7.88, for the classical source of this information. ¹¹ Haloinus (1978 [1533]: 126): ‘linguae illae vicinae et fere similes erant, ut in omnibus fere regionibus fieri experientia docet’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

114

 ’  

his Latin, either. This idea was picked up by a number of later scholars. The Anglo-Welsh literary author and grammarian James Howell (c. 1594–1666), for instance, even expressed it by means of a variation on the ‘to differ only in dialect’ phrase: There is also true Greek spoken in some parts of the lesser Asia, where there is no place upon the surface of the earth, for the proportion, where so many differing Languages are spoken, yet most of them are but Dialects and subdialects; so that of those two and twenty tongues, which Mithridates is recorded to have understood, above two parts of three, I beleeve, were but dialects. (Howell 1642: 150)

To sum up, the dialect concept and especially its Aristotelian interpretation was put into practice as a discursive strategy to relativize a polyglot’s linguistic competence, in particular that of the ancient king Mithridates.

8.3 A question of gradation: devising different levels of dialects The Aristotelian criterion suggests that early modern scholars conceived of the language/dialect distinction as a binary feature. However, many of them also acknowledged that linguistic variation in general and dialectal variation in particular come in different degrees. As a result, there was a tension between the Aristotelian criterion and the realization that language diversity is a gradational property. Was this mismatch resolved? And if so, how? Even though there was a broad consensus that dialectal variation comes in different degrees, no early modern scholar explicitly suggested that it constitutes a continuum, let alone that he drew conclusions from it that had consequences for the conceptualization of the language/dialect pair; its arbitrary nature was not recognized. Scholars of the period did, however, usually assume that among related dialects the similarities outweighed the differences. For this reason, some of them pointed out that dialects could be designated by a common name, usually that of the language of which they were dialects. The following was, for instance, claimed for the poorly documented linguistic situation of ancient Gaul, for which the Greek language served as a model: ‘nothing keeps us from giving it one and the same common designation, even though it is not entirely the same, but slightly varied, which also happened in the Greek language’.¹² Conversely, Georgius Haloinus (1978[1533]: 126) remarked that even though French was one language, it did have specific designations for its varieties, because they differed in certain

¹² Glarean (1538: 19): ‘Nihil tamen impedit quin communi appellatione eadem ac una nominetur, quanquam non usquequaq[ue] eade[m], sed paululu[m] uariata, Id quod in Græca quoq[ue] lingua accidit’. Cf. Chapter 4, Section 4.4.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

115

respects. To the mind of some authors, dialect-level differences justified the interchangeable usage of names: ‘the Chaldean and Syriac language are not different languages but only different dialects, which is why someone who speaks Chaldean is also said to speak Syriac’.¹³ But let me return to the acknowledgement that dialectal variation is a matter of gradation: how did it manifest itself in early modern linguistic thought? Certain scholars appear to have dissected the phenomenon of language variation in greater detail by devising various levels of dialects. The German diplomat Caspar Schoppe, for instance, distinguished as follows between two kinds of dialects in a 1636 pedagogical work: It seems proper to call some dialects “principal” or “general” and others “subject to the principal dialects” or “specific”, which even though they depart from some principal dialect in many respects are nevertheless related to it in far more respects.¹⁴

Schoppe made the distinction between ‘general’ and ‘specific’ dialects when emphasizing the importance of learning correct German, a study hampered by the existence of the great variety of dialects. In what followed, he exemplified the class of principal dialects by referring to the German of the Saxon city of Meissen, the Greek of Attica, the Italian of Florence, the French of Orléans, and the Spanish of Toledo, all perceived as the selected standard varieties of the respective languages. ‘Specific dialects’ of, for instance, the Meissen variety included Thuringian and Franconian German. Schoppe might have drawn inspiration from the early modern tradition of Greek dialect studies, where the principal literary dialects, mostly Attic, Ionic, Doric, and Aeolic, were distinguished from and superposed on secondary dialects (see Van Rooy 2020a: 19–24). Schoppe’s distinction was, however, not predominantly inspired by literary motivations, as with the Greek dialects, but by pedagogical and normative considerations of learning the correct standard form of a language. A similar differentiation of kinds of dialects can be found in Claude de Saumaise’s 1643 Commentary on the Hellenistic tongue. Saumaise, however, operated from a more strictly philological point of view, since he tried to analyse and classify linguistic diversity by means of a tripartite conceptual hierarchy, with a focus on the history of Ancient Greek. Saumaise (1643a: e.g. 154–6, 458–61; 1643b: 248) repeatedly differentiated between ‘generic dialects’, ‘specific dialects’, and ‘highly specific dialects’, the last two being also labelled ‘local dialects’. In an ¹³ Leusden (1656: 201): ‘Lingua Chaldaica & Syriaca non sunt diversæ linguæ, sed tantummodo diversæ dialecti; ideoque qui loquitur Chaldaicè etiam Syriacè loqui dicitur’. ¹⁴ Schoppe (1636: 46): ‘Dialectos alias vocare Principes placet, siue generales, alias principibus subiectas siue speciales, quæ quamuis in multis à principe aliqua Dialecto recedant, in multo pluribus tamen ad eam referu[n]tur’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

116

 ’  

academic dissertation presented in Wittenberg on 9 February 1709, which like Saumaise concentrated on the historical status of the Ancient Greek dialects, a different three-layered hierarchy of dialects was proposed (Thryllitsch 1709: .2–4; cf. Van Rooy 2016a: 470). The Ancient Greek language was divided into four principal and primary dialects, each spoken by an entire tribe (dialecti primariae, principales, or ethnikaí [ἐθνικαί]): Ionic, Attic, Doric, and Aeolic. Every primary dialect comprised several secondary, regional dialects (dialecti secundariae or egkho¯ŕ ioi [ἐγχώριοι]), which emerged as a result of the dispersion of the four principal tribes. Finally, each secondary dialect comprised several city or local dialects (dialecti urbicae or topikaí [τοπικαί]). In brief, it became customary from the seventeenth century onwards to divide the phenomenon of dialectal variation into two and sometimes even three hierarchical levels. This division occurred primarily out of a desire to create more order in the world’s linguistic diversity, even though pedagogical considerations could also play a role, as in the case of Caspar Schoppe. Such distinctions had their roots in early modern discourse on the Ancient Greek dialects, divided into principal and minor dialects from the early sixteenth century onwards. What is more, often Greek terminology was used in this context, even when it did not concern Greek dialects, or at least not exclusively. In the course of the eighteenth century, the distinction between primary and secondary dialects became ever more detached from the Greek context and was increasingly applied to other languages. It is, for instance, found in descriptions of diversity in Germanic (ten Kate 1723: .243), Basque (Larramendi 1729: 12), and among the Oriental tongues (Kals 1752: 58). There was great terminological proliferation to express different kinds of dialects. For instance, the Dutch orientalist Jan Willem Kals (1700–81) distinguished between ‘closer dialects’ (dialecti propinquiores) and ‘remoter dialects’ (dialecti remotiores; 1752: 58). One eighteenth-century Italian Hellenist even coined a new word to denote lower-level dialects: the diminutive dialettuzzoli (Mazzarella-Farao 1779: .). The distinction between primary and secondary dialects was especially prominent in the thought of eighteenth-century German scholars, who often referred to the former as Hauptdialekte or Hauptmundarten and to the latter as, for instance, Nebendialekte. The concept of secondary dialect generally entailed a value judgement. Indeed, such dialects were often suggested to be somehow inferior and subordinate to a primary dialect. To rid himself of such a connotation, a scholar could rely on the more neutral term subdialect(us), coined around the middle of the seventeenth century to denote a ‘variety subsumed under a dialect’. The concept had clear forerunners in earlier texts, but it was not until the mid-seventeenth century that it was condensed into a more or less demarcated notion, connected to a specific term. The term was invented, apparently independently, by two different writers active in the decades around 1650: the writer James Howell, who coined English subdialect, and the Swedish philologist Georg Stiernhielm, who introduced Latin

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.  

117

subdialectus (Van Rooy 2019: 760). Lacking an intrinsically subjective semantics, subdialect(us) was frequently used to simply denote a variety that was subsumed under a higher-level dialect or genetically derived from a dialect. It thus had a mainly classificatory function and constituted a more objective means to get a grip of the layered and gradational nature of dialectal diversity. James Howell, for instance, pointed out in his 1642 travel guide that someone journeying to France should be aware of the great variation within French. This language had three dialects, including Provençal, ‘whereof the Gascon is a subdialect’ (Howell 1642: 124). Howell’s analysis was clearly informed by a political bias; Walloon and Provençal were subsumed under French, the language of the king and his administration. Still, he did not want to make a prescriptive statement. Instead, he wanted to provide the reader with an adequate description of the linguistic context of France, especially since he intended to demonstrate how a traveller may ‘arrive to the practicall knowledge of the languages’, as he put it in the title of his travel guide. In short, early modern scholars distinguished between different kinds of dialects, thus expanding their conceptual apparatus in order to cope with the complex and multi-layered phenomenon of linguistic variation. Not only classification was a main goal of this conceptual development, however. It often also served to propagate the selected variety of a vernacular language that was being standardized as the correct linguistic form; the selected variety received a quality label, that of ‘principal dialect’. The early modern insistence on the existence of different kinds of dialects reflected linguistic realities, most importantly the fact that linguistic distance comes in degrees as well as the ongoing standardization of the Western European vernaculars.

8.4 Mutual intelligibility: an early modern criterion So we therefore say that the Jerusalemite tongue was distinct from the Galilean in the manner of dialect. The reason is evident, since Galileans could be distinguished from Jerusalemites by their speech, certainly not because the language was different (otherwise they would not have understood each other clearly), but the dialect was.¹⁵ To paraphrase this seventeenth-century observation on Hebrew diversity: speakers of related dialects can understand each other, whereas speakers of

¹⁵ Pfeiffer and Martini (1699: 212): ‘Sic ergo linguam Hierosolymitanam à Galilæa ad modum dialecti distinctam fuisse dicimus. Ratio evidens est, cum ex sermone Galilæi ab Hierosolymitanis dignosci potuerint. Non certe, quod lingua diversa esset, (aliàs se invicem planè non intellexissent) sed dialectus’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

118

 ’  

distinct languages cannot. In modern linguistics, the validity of this criterion, called mutual intelligibility in short, has been often discussed and generally considered doubtful at the very least, even though many modern scholars still rely on it (see Chambers and Trudgill 1998: 3–4; Chapter 20). How did early modern scholars approach mutual intelligibility, or lack thereof, in relation to the conceptual pair? Sixteenth-century authors generally did not employ it as a criterion to determine whether a speech form was a language or a dialect. Mutual (un)intelligibility was nonetheless eagerly used to characterize the relationship between specific language varieties. What is more, it is striking how often the absence of mutual understanding among speakers of related dialects was emphasized. As I have mentioned in Chapter 7 (Section 7.1), Martin Luther already observed in the first half of the sixteenth century how speakers of German, living only thirty miles apart, were not able to understand each other. The renowned Italo-French classical scholar Joseph Justus Scaliger (1540–1609) suggested that mutual intelligibility was endangered mainly when two lower-class people speaking different dialects were forced to converse, for instance a peasant from Liège and an artisan from Paris. Even though they both spoke varieties of French, they did not understand each other and even seem to have spoken entirely different languages (Scaliger 1610b: 123–4). In other cases, different varieties of a specific language were described as mutually comprehensible. The Greek scholar Theodosius Zygomalas (1544–1607), for instance, explained that the existence of different dialects in vernacular Greek did not prevent mutual understanding, seemingly suggesting that his observation was a more general principle (Zygomalas in Crusius 1584: 99). A sixteenth-century German grammarian appears to have employed mutual intelligibility as a criterion to classify the dialects of his native tongue into two different groups, ‘High German’ (‘Oberländisch’) and ‘Low German’ (‘Niderle[n]disch teutsch’); he clarified that the dialects that were part of the former group were ‘mutually intelligible’ or ‘inuice[m] intelligibiles’ in the Latin original (Albertus 1573: .8). The interpretation of the conceptual pair in terms of mutual intelligibility often went hand in hand with other understandings of it, not in the least the Aristotelian criterion. What is more, it was sometimes regarded as a logical consequence of it. Indeed, if two varieties differed in terms of substance, they were not mutually intelligible and, by consequence, distinct languages. If two varieties exhibited only variation in terms of their accidents, they were mutually intelligible and, by consequence, related dialects. Both conceptions appeared in close connection in Georg Stiernhielm’s ideas on the language/dialect distinction, which are analysed in a separate chapter because of their richness and complexity, and because they mark a transition to a subsequent phase in the history of the conceptual pair (see Chapter 12).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    . . .

119

8.5 Johannes Goropius Becanus and immediate mutual intelligibility Even though mutual intelligibility was keenly used in describing the relationship between specific dialects in the sixteenth century, it did not figure in more general discussions of linguistic diversity, with one major exception: the work of the Brabant physician Johannes Goropius Becanus (1519–73; Figure 8.2), infamous for his theory that Dutch was the primeval tongue of Adam and Eve, put forward in an age where most scholars attributed this privilege to the Hebrew tongue.¹⁶ Mutual intelligibility played a role of primary importance in Goropius’s distinction between ‘different languages’ and ‘identical languages’: Now, since it has been shown what dialect means to Aristotle and what it means to us, it should be enquired which dialects or languages are to be considered

Figure 8.2 Johannes Goropius Becanus by Philips Galle, 1572 Source: Rijksmuseum, Amsterdam. Public domain

¹⁶ See Van Hal (2010a: 77–139) and Frederickx and Van Hal (2015: 111–71) for nuanced discussions of Goropius’s linguistic ideas.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

120

 ’  

identical and which different. The Greeks—in their fashion of speaking— subsumed five different dialects under a single Greek language. We—in the popular fashion of speaking we follow here—believe that it should be determined thus: we call different those languages that mutually differ in such a manner that he who understands one language does not immediately understand the other language, too. However, we call identical those languages that, even though they differ somewhat, nevertheless do not deviate from each other to such an extent that their difference makes conversation and communication impossible.¹⁷

Clearly, Goropius regarded the Latin terms lingua and dialectus as synonyms, choosing to base his metalinguistic distinction on such qualifications as ‘different’ and ‘identical’. Indeed, Goropius concentrated on determining whether specific varieties should be understood as constituting either one and the same language or different languages. His main criterion for doing so was immediate mutual intelligibility, immediate in the sense that speakers had no time for reflection or prior knowledge. In what followed, Goropius exemplified his criterion, arguing that on this ground the Brabantian tongue was distinct from Swabian, just as Spanish was from French and Italian. Brabantian and Flemish, on the other hand, were identical tongues. Even though they were somewhat different, speakers of these varieties could communicate with each other without problems. It should be noted that Goropius seems to have attributed a language/dialect distinction to the Greeks, who, he supposed, subsumed several dialects (dialecti) under their language (sermo). He reformulated the allegedly Greek conceptual pair in ‘the popular fashion of speaking’ as a difference between ‘different’ and ‘identical languages’. Goropius thus adopted a slightly different perspective, but in essence his distinction can be said to boil down to a distinction between language and dialect. After all, he suggested himself that it was merely a matter of terminology. What is more, elsewhere in his work, Goropius appears to have employed the term dialectus in the sense of ‘variety subsumed under a language, immediately intelligible for speakers of related varieties’ in alleged agreement with Greek terminology. For instance, he frequently used formulations such as ‘different dialects of the same language’.¹⁸ Such a usage is in direct contradiction to the fact that Goropius had initially defined the word dialectus as ‘the faculty of

¹⁷ Goropius Becanus (1580: Hermathena 3–4): ‘Iam quoniam quid Dialectus Aristoteli, quid nobis sit, ostensum est; quærendum, quæ dialecti siue linguæ eædem, quæ diuersæ sint habendæ. Græci, more suo locuti, quinque diuersas dialectos sub vno sermone Græco complexi sunt: Nos populari loquendi more, quem hîc sequimur, ita statuendum arbitramur; vt eas linguas diuersas dicamus, quæ sic inter se discrepant, vt non statim qui vnam intelligit, alteram quoque intelligat. Easdem verò; quæ licet nonnihil differant, tantum tamen non recedit altera ab altera, vt discrimen earum colloquendi communicandiq́ue auferat facultatem’. ¹⁸ Goropius Becanus (1580: Hispanica 5): ‘Alibi commemoraui nos Man dicere, Allemanos in eadem significatione Mon, & id genus variationes esse in diuersis eiusdem linguæ dialectis quàm frequentissimas’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    . . .

121

speaking in a native way about everyday matters’, putting it on a par with lingua.¹⁹ This friction can be explained by the fact that dialectus had a double sense for Goropius, referring, on the one hand, to a specific historical variety of a language and, on the other, to human language as a means of expression. Goropius’s case thus constitutes an excellent example of the way in which the polysemy of dialectus could impact on a scholar’s linguistic thought and his terminological flexibility. Goropius’s criterion of immediate mutual intelligibility enjoyed some success in the seventeenth century and was adopted by several scholars who never mentioned Goropius’s name in this context, probably because of the controversial nature of his views on Dutch linguistic primacy. The most striking case is that of the Protestant theologian Johann Caspar Myricaeus (†1653), who in his Syriac grammar copied Goropius’s passage on the distinction between different and identical languages verbatim, but did not refer anywhere to Goropius’s work.²⁰ However, not everyone agreed with Goropius that related dialects were directly mutually intelligible. Abraham Mylius (1612: 85), for instance, who was well acquainted with Goropius’s work, remarked that a speaker of Syriac needed training in order to immediately understand Arabic or Hebrew, this in spite of the fact that these tongues differed only in terms of dialect. The criterion of immediate mutual intelligibility appeared with a remarkably low frequency in eighteenth-century thought. An important exception was the Swedish educator Sven Hof ’s (1703–86) Dialectus Vestrogothica, a monograph on the Västergötland dialect of Swedish.²¹ In it, Hof (1772: 11) claimed that a lack of mutual intelligibility ‘without any peculiar instruction’ (‘sine peculiari institutione’) indicated different languages, whereas related dialects were mutually understandable. More generally, the interpretation in terms of mutual intelligibility was relatively rare in eighteenth-century discussions of the language/dialect pair, which may be connected to the increasing focus on specific linguistic features in analyses of the conceptual distinction (see Chapter 15). The emphasis on immediate mutual intelligibility by Goropius and others suggests that a number of early modern scholars did not take it to be a binary feature but rather saw it as a property with at least three possibilities: speech forms can be either mutually unintelligible, or mutually intelligible after instruction or a period of contact, or immediately mutually intelligible. This insight did not, however, lead scholars to problematize the fact that neighbouring dialects are ¹⁹ I have paraphrased the definition from Goropius Becanus (1580: Hermathena 3): ‘ . . . facultas, vel vsu, vel præceptis, vel ambobus acquisita; qua qui præditus est, de quauis re, in populari consuetudine tractari solita, apud nationis cuiuspiam homines emendatè loqui possit’. ²⁰ Myricaeus (1619: }.2–3): ‘Porro lingvas diversas eas esse dicimus, quæ sic inter se discrepant, ut non statim qui unam intelligit, & alteram possit: easdem quæ, licet non nihil differant, tantum tamen una ab altera non recedit, ut discrimen earum colloquendi communicandique auferat facultatem’. ²¹ On Hof ’s place in the history of linguistics see Droixhe (1978: 123–4); Hovdhaugen et al. (2000: 93–4).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

122

 ’  

more likely to be mutually comprehensible than related dialects that are geographically farther removed from each other. What is more, most authors seemingly chose to ignore the fact that mutual intelligibility comes in degrees and took the property in a strictly binary manner. The English bishop Brian Walton (1600–61), for instance, did so in his prolegomena to the polyglot Bible he edited in 1657, which touched on, among other things, issues of language history. Speakers of related dialects could understand each other, but speakers of different languages could not. That was at least what Walton guessed, as he felt compelled to insert the Latin adverb forte, ‘perhaps’, into this claim of his. Still, he explicitly stated that if there was no mutual intelligibility among speakers of two different tongues, these would be improperly called dialects (Walton 1657: 3). Other scholars seem to have been more confident about the binary nature of the criterion. For instance, in his posthumously published remarks on the linguistic diversity of the Holy Roman Empire, the French magistrate Guillaume Ribier (1578–1663) maintained that speakers of different languages could not understand each other, neither in speaking nor in writing (1666: 27). The awareness of early modern authors that dialectal variation comes in degrees, implicit in the insistence on immediate mutual intelligibility, manifested itself more strongly in other ways, most notably in the assumption that there were different kinds and levels of dialects, as I have argued earlier in this chapter.

8.6 The communicative reach of dialects versus languages The emphasis certain authors put on the lack of mutual intelligibility among related dialects suggests that these linguistic varieties were believed by some scholars to have a restricted communicative reach. Conversely, a wide communicative reach was from time to time posited for a language, as, for instance, by an English bishop in a seventeenth-century discourse on the celebration of Mass in an unknown tongue: Where there are these different Dialects, there generally is one way of speaking, which either from the eloquence, or fashionableness of it, so far prevails, as to be the Standard of the Tongue, and to be used in Writing Books, Letters, &c. and is understood by all. (Williams 1685: 5)

The extensive communicative reach of a language was connected to its written status, causing it to be more permanent than the ever fluctuating dialects, most of which were not used in writing. In the eighteenth century, the emphasis on the extensive communicative reach of a language, often specified as the common language, intensified. The Greek émigré Alexander Helladius (1686–after 28 August 1714), for instance, stated the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

123

following on the vernacular Greek and German languages in his 1714 description of the Greek Orthodox Church: Thus, in the modern era, there is a single common ecclesiastical dialect across the whole of Greece, intelligible by all and at the same time not easily imitable by any plebeian. This is not different from that Attic throughout Germany, which is called “the High German par excellence”.²²

Helladius, although using the term dialectus both for the common variety and for the vernacular dialects, contrasted the communicative reach of the common variety to the miscommunication the dialects of the populace could cause. The latter theme was treated at length in Helladius’s book by means of a series of amusing anecdotes drawn from his own experience in Greece and Germany (Helladius 1714: 192; Van Rooy forthcoming).

8.7 Conclusion A key interpretation of the conceptual pair that emerged during the first part of the early modern period was what I have been calling the ‘Aristotelian criterion’. The idea that languages differed from each other substantially, whereas related dialects varied only accidentally, was actively used as a diagnostic criterion to determine the status of a speech form. It was also used for other, highly specific purposes, for instance to denigrate someone’s polyglot competence. The Aristotelian interpretation featured seldom in generalizing discussions, perhaps because it was supposed to be self-evident, much like the language/dialect distinction itself. Its relationship to the Greek tradition was moreover ambiguous. Even though the criterion was undoubtedly an early modern innovation, it was obviously inspired by traditional philosophical categories associated with Aristotle. The friction between the binary nature of the language/dialect pair and the realization that dialectal variation was a gradational property remained unresolved. Scholars posited different hierarchical levels within dialectal diversity but refrained from elaborating on the question of how this multi-layered analysis of linguistic diversity fitted in with the conceptual pair in general and its Aristotelian interpretation in particular. Closely associated to the Aristotelian interpretation was mutual intelligibility. The dialect or language status of a speech form depended on the question whether or not speakers could understand each other. Some early modern scholars relied ²² Helladius (1714: 187): ‘Ita moderno tempore, per universam Græciam unica est dialectus ecclesiastica communis, & ab omnibus intelligibilis, non idem facile à quolibet plebejo imitabilis, non aliter quam per universam Germaniam Attica illa, quæ κατ’ ἐξοχὴν das Hochteutsche vocatur’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

124

 ’  

on this feature as an absolute criterion. Yet there were several authors who, dissatisfied with a binary interpretation of the feature, specified the criterion as immediate mutual intelligibility. It is rather paradoxical that some authors argued that dialects of one and the same language were mutually understandable, whereas others emphasized the limited communicative reach of dialects and the lack of mutual understanding among speakers of closely related dialects. The realization that dialects were not always mutually intelligible was moreover in contradiction to the widespread early modern idea that dialects varied only on an accidental level. An author’s emphasis on mutual unintelligibility among dialects was often related to contextual factors. For instance, Martin Luther’s stress on the lack of mutual understanding among the speakers of German dialects must be viewed in connection with the confessional goals his linguistic programme had to serve. Luther wanted to justify his adoption of the common language of the Saxon chancellery as the linguistic medium for his German Bible translation; pointing out the great linguistic diversity in Germany and the communicational problems that went with it was part of that justification.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

9 A subjective touch Language beats dialect

Early modern scholars at first limited themselves to formulating more or less objective interpretations of the language/dialect pair, in the sense that they did not associate either concept with normative and evaluative properties. In the seventeenth century, more subjective voices appeared on the scene, and many scholars degraded dialect to a status inferior to that of language. How and why did this evolution occur? In the present chapter, I set out to answer these questions by treating two major interpretations of the language/dialect distinction centring around notions of analogy/anomaly and superiority/inferiority.

9.1 Analogical norm or anomalous deviation? The common language is that which follows the more exact analogy or, as usually happens with extinct languages in particular, the norm of grammatical precepts, from which the dialects variously diverge now and then.¹ According to this Wittenberg disputation on the causes of dialects, a common language reflected a regulated and analogical norm, whereas dialects were characterized by anomalies deviating from that norm.² This interpretation of the conceptual pair had its roots in the close of the sixteenth century. For instance, the Protestant philologist Otto Walper (1543–1624), professor of Greek and Hebrew in Marburg, made the idea of an analogical language and its anomalous dialects explicit in his handbook for the Greek dialects. This work was first published in 1589 as an independent treatise; a year later Walper incorporated it into his Greek grammar, which enjoyed numerous reprints for several decades. In the dedicatory letter prefixed to his handbook, Walper pointed out the following: ¹ Schwartz and Helm (1702: .1): ‘communis lingua est, quae exactiorem analogiam, aut, quod in intermortuis linguis maxime fieri solet, praeceptorum Grammaticorum normam sequitur, a quibus dialecti uarie aliquando diuertuntur’. ² On analogy and anomaly in early modern linguistic thought and their ancient roots see Haßler (2009a).

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

126

  

Even though indeed almost all grammarians add the dialects as if they were some anomalies after the analogy of the precepts for the common language, so that it seems to be quite superfluous to prepare a separate treatment of the dialects, still, if we would examine the matter more thoroughly, we will notice that there will be a rather different conclusion in view of the condition and utility for the studious youth.³

Walper apparently attributed the contrast between the anomalous Greek dialects and the analogical Koine to the methods and grammatical descriptions of his predecessors, in whose works such an opposition was, however, only implicit at most. At the same time, he deplored and tried to remedy the idea that dialects were arbitrary deviations from the regulated common language. All the same, such an opposition was still rare in the sixteenth century. What is more, anomaly was not at all considered to be inherent in the concept of dialect. The English theologian Thomas Stapleton (1535–98), for example, was convinced that the Greek dialects ‘might be gathered and collected in to some Orders of Rules’ (1566: 59). The seventeenth century witnessed a profound change in conceptualizing the language/dialect distinction in highly normative terms. The influential German grammarian Justus Georg Schottel revealingly observed that ‘in all dialects, there is something faulty, which cannot occupy the position of rule in the language itself ’.⁴ Elsewhere, too, Schottel (1663: 158) emphasized the impossibility of formulating rules for dialects, which he conceived as intrinsically variable. In other words, language was the regulated norm or point of reference from which its dialects anomalously deviated. Even though Schottel did not mention the Greek dialects at all in this context, this normative interpretation seems to have originated in Byzantine grammatical practice. Certain Byzantine scholars had opposed the Koine to the dialects at various occasions, suggesting at times that the Koine was analogically constructed, whereas the dialects were anomalous variations on the Koine. They did so, however, without working with an abstract language/dialect-like distinction. This approach was silently adopted by Renaissance Hellenists to become explicit only towards the end of the sixteenth century in the work of Walper, who criticized the grammarians’ default attitude. In the course of the seventeenth century, the analogy/anomaly opposition was generalized and detached from the context of Greek studies. As a result, it became applicable to a wider range of languages and absorbed into the conceptual pair. ³ Walper (1589: †.6): ‘Etsi verò ab omnibus penè Grammaticis post analogiam præceptorum linguæ communis, Dialecti tanquam anomaliæ quædam subijciantur [sic], ita vt fermè superuacaneum esse videatur seorsim de Dialectis tractationem instituere: tamen si rem penitius introspiciamus, diuersum potius ex re & vtilitate studiosæ iuuentutis fore animaduertemus’. ⁴ Schottel (1663: 174): ‘Omnibus dialectis aliquid vitiosi inest, quod locum regulæ in Lingua ipsa habere nequit’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.     

127

Of course, language regulation was an evident and central theme for vernacular grammarians even before the seventeenth century, but it was only after 1600 that more sternly normative attitudes came to dominate language disputes. In France, in particular, scholars intensely debated ‘good usage’ (bon usage) in the tracks of François de Malherbe (1555–1628), a tradition in which the genre of the ‘remarks’ came to play a central role (Ayres-Bennett and Seijido 2011). Metalinguistic considerations on normative concepts, including language and dialect, remained relatively rare throughout the early modern period, however (cf. Haßler 2009b: 674). It seems nonetheless justified to conclude that early modern scholarship on the Greek language stimulated to some extent more general observations on the prescriptivist idea that dialects were anomalous deviations from the linguistic norm. The strongly normative focus of early modern vernacular grammarians, by the way, indicates that their main concern normally was with the relationship between dialects and the language from which they were believed to deviate rather than with the interrelationships of related dialects (cf. Haas 1994: ).

9.2 From common to standard language Language was often further specified as common language when an early modern scholar interpreted the conceptual pair in terms of analogy and anomaly. Out of the idea that a common language constituted the regulated norm from which dialects deviated, the concept of standard language seems to have emerged in the second half of the seventeenth century, with a particular position for England. The English bishop John Williams (?1636–1709) was one of the first to employ the term standard in a linguistic sense, while mentioning at the same time some of the basic features of what most modern linguists would call a standard language. Williams did so in the context of a fundamental problem for the celebration of Mass: in what language should it be read? He observed in this regard: Where there are these different Dialects, there generally is one way of speaking, which either from the eloquence, or fashionableness of it, so far prevails, as to be the Standard of the Tongue, and to be used in Writing Books, Letters, &c. and is understood by all. Such I conceive was anciently that which is called the common Dialect in Greek: And of the like kind is that which is spoken in and about the Court, and by Scholars and persons of a liberal education amongst us; and elsewhere. (Williams 1685: 5; cf. Chapter 8, Section 8.6) Which is (to speak charitably) for want of observing, that the Dialects are but several modes of speaking the same Tongue; and that ordinarily there is some common Standard, which (as I have said) over-rules the rest, and is a guide common to all: As here in England, notwithstanding there be several Dialects,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

128

  

and that there is one in Scotland differs much from them all; yet there is but one Translation of the Bible, and one Service for the use of the whole, and that is fully if not equally understood by all. (Williams 1685: 8)

In Williams’s opinion, the standard variety was selected because of its ‘eloquence’ or ‘fashionableness’, as it was the only one appropriate for written communication and it had the advantage of being intelligible to all speakers of the language. To clarify his standard language concept, Williams referred to both the Greek Koine and the high variety of English, spoken at the English court and by the intelligentsia. The standard variety not only ‘roofed’ the other dialects, to use modern linguistic terminology, but was also an example for them—or in Williams’s phrasing: it overruled and guided them. In other words, the dialects were socially, culturally, and linguistically subordinate to the standard, which was employed by the upper classes and which was the ideal the dialects should aim to attain. Normative thinking intensified in the course of the eighteenth century. For instance, an English classical scholar maintained that ‘the distinction of dialects can be only known to a cultivated and, in some degree, settled state of language, as deviations from an acknowledged standard’ (Wood 1775: 238). In the French tradition, a number of scholars distinguished between two kinds of varieties subsumed under a language; they were either legitimate and regulated dialects, occurring in politically diversified areas, or corrupted patois that lacked rules and deviated from the normative variety of the centralized state, the latter obviously applying to the kingdom of France (cf. Chapter 16, Section 16.2). Their conception and evaluation of a regional variety of a language, in other words, depended on the political context. Language was the analogical norm from which dialects were anomalous deviations. Yet this opposition was not set in stone, as some early modern scholars were aware that certain speech forms which they labelled languages were in fact not much more than what one might call ‘upgraded’ dialects. They assumed, in other words, that language status was something that could be achieved by the standardization of a dialect. Modern research has pointed out that this process consists in the selection of a specific variety as the norm, its grammatical codification, its elaboration, and its wide acceptance as the standard (Chapter 4, Section 4.1). It is, however, possible to find traces of awareness of some of these standardization phenomena in various early modern sources. For instance, scholars noticed that most languages had a principal dialect, the dialect selected as the linguistic norm, as, for instance, the Spanish scholar Juan Luis Vives did already in 1533. Vives (1533: .iii) mentioned Attic for Greek, Castilian for Spanish, Tuscan for Italian, and Parisian for French, all of which he presented as regional varieties of their respective tongues enjoying a privileged status. Codification and elaboration were widely considered indispensable for turning a vernacular dialect into a full-fledged cultivated language that could compete with

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    ? 129 Latin and other major vernaculars (cf. Burke 2004: 89). The German poet and lexicographer Nathan Chytraeus (1543–98), for example, sounded rather desperate about the linguistic state of affairs in Germany when he wrote the following in the dedicatory letter prefixed to his pioneering Latin–Low Saxon lexicon of 1582: To be sure, no matter what dialect of no matter what language has certain elegancies and peculiar ornaments of its own. Doubtlessly these need to be carefully studied and collected by someone who, in the German or whatever other nation, would desire to establish a common language, such as there once was in Greece, and to adorn it with characteristic and clear words and manners of speaking, and besides also with figures, as if it were with jewels and flowers.⁵

The concept of common language, understood as a prestigious means of communication among the elites of a nation, played a key role here, since it was viewed as the goal to achieve. Indeed, Chytraeus expressed the hope that German would make up lost ground in this regard by following the example of the ancient Greeks and other nations of his time. In Chytraeus’s view, however, the common language to be designed did not consist in one single dialect but rather in a compilation of the best features of different dialects, which reveals a distinctly positive attitude towards regional variation. Yet dialect was soon to acquire an overly depreciatory interpretation.

9.3 Superior or inferior?  . . . Particular speech of a province, corrupted from the general or principal language of the kingdom.⁶ The lemma dialecte in Antoine Furetière’s (1619–88) influential French dictionary suggests that, to the lexicographer’s mind, language had positive properties, whereas dialect was primarily linked with negative features. A language was considered superior, pure, and adequate for literature, but a dialect was inferior, corrupted, and not at all fit for literary purposes. This interpretation was closely tied up with the opposition analogy/anomaly and followed almost naturally from it. Language was analogical and therefore superior to a dialect, which was ⁵ Chytraeus (1582: .3-): ‘Habet enim quælibet etiam cuiuslibet linguæ dialectus, suas quasdam concinnitates, sua peculiaria ornamenta, quæ sanè accuratè consideranda & colligenda forent ei, qui in natione Germanica, vel alia quacunq[ue], communem aliquam, qualis in Græcia fuit, linguam constituere; eamq́[ue] vocabulis & modis loquendi proprijs & perspicuis; & præterea figuris quoq[ue] tanquam gemmis & flosculis cuperet exornare’. On Chytraeus’s work as a lexicographer see Considine (2017: 116–17). ⁶ Furetière (1690: s.v. dialecte): ‘DIALECTE . . . Langage particulier d’une Province, corrompu de la Langue generale ou principale du Royaume’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

130

  

anomalous and therefore of lower esteem than a language. This very subjective interpretation of the conceptual pair was nearly absent from sixteenth-century thought. The closest we come to it is in a work by the English literary critic George Puttenham (1529–90). In The arte of English poesie, Puttenham (1589: 119–20) opposed the inalterable language of a nation to speach, that was variable in terms of cultivation, literary usage, and regulation (see Hickey 2010: 15). Still, most sixteenth-century authors did not regard dialects as inherently inferior entities, even though individual dialects often received various evaluative rubrics such as ‘elegant’, ‘coarse’, ‘rustic’, ‘smooth’, and so on. A negative interpretation of the dialect pole of the conceptual opposition might have initially been obstructed by several factors. Firstly, dialectus was still very closely connected with the literary varieties of the revered Greek language in the sixteenth century. Secondly, the highly valued selected variety of a language was often still termed dialectus in the first half of the early modern period, precluding a downgrade of the concept of dialect. Thirdly, local varieties were not yet automatically dismissed as inferior entities. Indeed, several authors even stressed the richness dialects could exhibit. The printer-scholar Henri Estienne (1582: *.iii-), for instance, described how certain dialect features could contribute to the splendour of the French language like beauty marks on a pretty face. Even more tellingly, Estienne (1579: 133–4, 147) suggested that dialects related to a language as a rich man’s country seats to his more luxurious main residence in the city (cf. Swiggers 2009: 73). This comparison implied that even though dialects were not inherently corrupt, they were nevertheless somehow inferior to the normative language of a country, closely associated with the capital. Estienne did loosely associate dialect with vice elsewhere in his linguistic work, for instance when he criticized the French grammarian Jean Pillot (1515–92) for mistaking a fault in his dialect for a grammatical rule (Estienne 1582: 202; see Bruña Cuevas 1998: 528). He moreover emphasized that one could only speak of a dialect’s richness insofar as the depravations and corruptions of ‘the insignificant populace’ (le menu peuple) had been cleared away. For instance, Parisian speech served as the basis for the ‘pure and native French language’ (pur et nayf langage françois), but this elegant variety, too, had to be cleansed from vulgar elements (Estienne 1579: 135, 143). More in general, sixteenth-century scholars did not consider a dialect to be intrinsically typical of lower social classes. For in their time, standardization processes were still under way, and most members of the higher classes still spoke in their local dialects, even though they increasingly tried to write in the common language that was being designed. This sociolinguistic situation explains why humanists like Estienne attributed a prominent role to the dialects in enriching the standard; local varieties benefitted from the social prestige of their speakers and were therefore a factor to be reckoned with, but the fate of the dialects was soon to change.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    ? 131 During the seventeenth century, especially in its second half, corruption, degeneration, and lack of cultivation were ever increasingly associated with dialects, whereas language was conceived as a superior entity linked to purity. An early example of this development can be found in The ambassador, a Latin treatise on diplomacy first published in 1598 and authored by the Italo-French diplomat Carlo Pascale (1547–1625). In this work, Pascale answered the question of what language should be used by a diplomatic mission as follows: As far as I am concerned, I think it should be in that language which is native to the nation of the embassy . . . I am talking here about the embassies of those nations that have their own language, as it is offending to call the inelegant, worthless, and obscure idioms of some provinces languages. I prefer to call them servants of languages.⁷

Pascale, well-versed in the Greek language and literature, referred to these inferior linguistic entities as provincial ‘idioms’ (idiomata), using a term that was not as closely connected with the prestigious Greek literary varieties as dialectus was. This terminological choice made it possible for him to operate with a conceptual opposition consisting of an inferior and a superior pole: idioma versus lingua. Pascale was, in other words, actively looking for a way to distinguish two levels of linguistic entities, a superior and an inferior one, suggesting that there was already a need for it but not yet an established metalinguistic apparatus with which to do so. The idea that dialects were faulty varieties of a language, prominent from the second half of the seventeenth century onwards, only made sense if the literary Greek dialects no longer had a privileged connection with the concept of dialect. One of the most telling testimonies to such a view is Antoine Furetière’s definition of dialecte, explained as ‘particular speech of a province, corrupted from the general or principal language of the kingdom’, as quoted above. Furetière interpreted the term patois even more negatively, defining it as ‘corrupted and churlish speech, such as that of the insignificant populace, peasants, and children who are not yet able to pronounce well’.⁸ Dialecte and patois were, in other words, semantically closely related in Furetière’s eyes, even though patois was primarily associated with social differences and dialecte was understood as a corrupted regional variety of a language. In the eighteenth century, the idea that dialects were corrupt entities led some scholars to even promote the abolition of certain dialects. In an anonymous essay ⁷ Pascale (1598: 269–70): ‘Equidem puto illa quæ iis vernacula est, quorum legatio est . . . Hîc loquor de legationibus earum gentium, quibus est propria lingua. Nam incompta, egena, et contorta idiomata quarumdam prouinciarum tædet vocare linguas. Malo linguarum pedissequas’. ⁸ Furetière (1690: s.v. patois): ‘Langage corrompu & grossier, tel que celuy du menu peuple, des paysans, & des enfants qui ne sçavent pas encore bien prononcer’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

132

  

of 1743, for instance, it was argued that Low Saxon should be wiped out in favour of High German and for the greater glory of the empire (Anon. 1743). Others, however, restored the link with the Greek background to a certain extent and came to regard dialects as full-fledged and legitimate varieties of a language, distinguishing them from the corrupted varieties of a language, termed jargon or patois (Chapter 16, Section 16.2). A number of scholars projected the idea of corruption and degeneration on the Ancient Greek dialects, too, as did John Wilkins (1614–72), who wrote at the very beginning of his well-known Essay towards a real character, and a philosophical language of 1668: The Greek was anciently of very great extent, not onely in Europe, but in Asia too, and Afric, where several Colonies of that Nation were planted; by which dispersion and mixture with other people it did degenerate into several Dialects. (Wilkins 1668: 3)

One seventeenth-century Swiss Hellenist even suggested that the Greek dialects were sheer incorrigible deviations of the Greek language (Wyss 1650: 2). It should be noted that the opposition superiority/inferiority was not perceived as a strictly binary one, as statements such as the following reveal: ‘the more cultivated a vulgar form of speaking is, the more closely it approaches the common language’.⁹ What is more, the cultivated language could become tainted by the inferior dialects. According to the German polymath Daniel Georg Morhof (1639–91), ‘in such a vast diversity of dialects, it can easily happen that the more cultivated language receives some dirt’.¹⁰ The corruptive influence of dialects on the high variety could, however, be remedied, Morhof argued, by political centralization, which would eventually result in linguistic normalization as well (Morhof 1685: 146). It is remarkable that one seventeenth-century French author, the French Jesuit Pierre Besnier (1648–1705), observed that the most polished language was often the most corrupted one, whereas the patois were pure (1674: 23–5). Besnier’s opposition centred around the artificial nature of a cultivated language versus the pristine and primitive quality of the patois. Here, we have a rare example of an early modern scholar rightly suggesting that a standard language intrinsically postdates a situation of dialectal diversity. Other authors formulated different, less historically accurate solutions to the question as to how the standard language related to its cognate dialects in terms of time. The grammarian Justus Georg

⁹ Schwartz and Helm (1702: .1): ‘Quo cultior autem aliqua dicendi forma uulgaris est, eo propius ad . . . linguam communem accedit’. ¹⁰ Morhof (1685: 113): ‘Nos tamen inde observabimus, quam facile sit in tantâ diversitate dialectorum cultiorem linguam aliquid sordium adsciscere’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    ? 133 Schottel (1663: 174), for instance, suggested that scholars had restored the High German language from the existing dialectal diversity to its original form. The association of dialect with degeneration, corruption, and lack of cultivation reflected a number of sociolinguistic evolutions occurring in seventeenth-century Western Europe. Most notably, the high and middle classes—royalty, nobility, intelligentsia, bourgeoisie—went out of their way to write and eventually speak the normative variety, identified as the language. In Germany, for instance, ‘erudite men and those imitating them principally use [the common language] in writing, yes indeed also—as far as it can happen—in pronouncing’.¹¹ In other words, speaking a dialect was no longer solely a geographical but increasingly also a social mark. As a matter of fact, when confronted with the high variety, one’s native dialect could even become a source of shame—or to put it with an early modern English saying: ‘She hath been at London to call a strea a straw, and a waw a wall’. ‘This’, the renowned English lexicographer John Ray (1627–1705) explained, ‘the common people use in scorn of those who having been at London are ashamed to speak their own Countrey dialect’ (1678: 75). Dialects were ever more closely associated with the peasantry and were opposed to the cultivated language of high society, as the testimony of an early seventeenthcentury English grammarian reveals: What I say here regarding the dialects, you must realize, refers only to country people, since among persons of genteel character and cultured upbringing, there is but one universal speech, in pronunciation and meaning.¹²

Similarly, the German grammarian Schottel (1663: 174) reserved the actual language for the men of high social classes, who played an active role in its formation and elaboration. The insistence on this idea seems to have been strongest in the French context, which is probably connected to its high degree of political as well as linguistic centralization (e.g. Rousseau 1781 [1755]: 243–4). In short, whereas dialect did not have an intrinsically negative connotation in the Cinquecento, the concept suffered from a devaluation in the seventeenth and eighteenth centuries. It occurred not only in countries with a high degree of political centralization such as France, but also in politically diversified areas such as the states of the Holy Roman Empire. The increasing frequency of the interpretation can probably be viewed as a consequence of the growing emancipation of the dialect concept from the Greek heritage with its valued literary ¹¹ Schwartz and Helm (1702: .1): ‘lingua communis Germanica . . . quam eruditi, et, qui hos imitantur, potissimum in scriptione, imo et, quantum fieri potest, in pronunciatione, usurpant’. ¹² Gill (1619: 18): ‘Et quod hîc de dialectis loquor, ad rusticos tantùm pertinere velim intelligas: nam mitioribus ingenijs et cultiùs enutritis, unus est ubique sermo & sono, & significatu’. The English translation is adapted from Gill (1972: .104). For Gill’s views on the English dialects see Kökeritz (1938).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

134

  

Table 9.1 Transfer of social prejudices to the conceptual pair -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------▸ Social class

Prejudice

Speech

Conceptual pair

Higher Lower

Cultivated Unrefined

Standard and superior Regional and inferior

Language Dialect

The arrow indicates the direction of transfer

dialects as well as of the great rise in esteem of the emergent standard languages. Indeed, the evaluative interpretation of the conceptual pair appeared around the time that the selected standard variety was gradually being detached from specific regions. The norm was ‘delocalized’, to quote Peter Burke (2004: 110), as part of a broader ‘civilizing process’, involving primarily the European elites: the glamour of the court and its influence on provincial elites encouraged the adoption of the new form of language, which became a sign that its users were distinct from and superior to ordinary people. The reform of speech was part of a wider change, the withdrawal from participation in many forms of popular culture on the part of European elites.

The evaluative interpretation of the conceptual pair was no doubt the result of transferring preconceived ideas about specific social classes to the language variety they used and eventually to the conceptual pair, too, a process schematized very roughly in Table 9.1.

9.4 Conclusion The distinction between language and dialect acquired a highly subjective dimension from the end of the sixteenth century onwards. A dialect was increasingly conceived as an anomalous deviation from the analogical norm of the language. This interpretation appears to have been inspired partly by assumptions about language-external, social realities and partly by standardization processes, in which non-selected dialects were usually only marginally involved, if at all. The fact that non-selected dialects did not undergo processes of grammatical codification and regularization no doubt led early modern scholars to associate dialect with deviation from the codified variety and a lack of analogy. This idea is likely to have paved the way for the interpretation in terms of status and was closely intertwined with it. A dialect was anomalous and therefore inferior, seems to have been the inference, whereas a language was analogical and by consequence superior. A subjective turn occurred early in the history of the conceptual pair,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

. 

135

and early modern authors soon made a habit out of speaking in depreciatory terms about dialects. The link between regional speech and lower social class intensified, which no doubt fostered negative conceptions of dialect. Remarkably enough, early modern scholars generally did not put the interpretation of the conceptual pair in terms of analogy and anomaly into a diachronic perspective. In other words, they did not usually make explicit that a codified, regulated language could come only after a situation of dialectal diversity without an acknowledged standard. Diachrony did, however, occupy a central position in the final major interpretation of the conceptual pair that arose in the early modern period.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

10 The conceptual pair and language history Language generates dialects

When humanists created the language/dialect pair in the sixteenth century, they initially conceptualized it uniquely from a synchronic perspective; a dialect resorted under a language at a specific moment in time. Soon a historical interpretation of the distinction between language and dialect was devised, as numerous early modern scholars began to consider a dialect to be a speech form posterior to and genealogically deriving from a language. In this case, the latter was often metaphorically termed mother language, the former daughter dialect. When and why did this innovation with far-reaching consequences occur?

10.1 Language mothers and offshoot dialects We can call language mothers those from which many dialects— offshoots as it were—are derived. Offshoots of one language mother are indeed connected by a certain correspondence, but there is no mutual kinship among the language mothers, either in words or in analogy.¹ With these observations the great classical philologist and professor at Leiden university Joseph Justus Scaliger (Figure 10.2) opened his very short Diatribe on the languages of the Europeans (Figure 10.1), originally a letter written to his colleague Paulus Merula in 1599 and first published by Merula (1605: 271–2), albeit against Scaliger’s will. Five years later, the letter was included in a posthumous collection of Scaliger’s unpublished writings. In his Diatribe, Scaliger controversially argued that there was no kinship whatsoever among the eleven language mothers of Europe, a remarkable idea that soon met with fierce criticism (see Van Hal 2010b). What is of interest to us, however, is what his opening statement reveals about the metalinguistic concepts with which he was operating. To Scaliger’s mind, dialects derived genealogically from an originally unitary ¹ Scaliger (1610a: 119): ‘L Matrices vocare possumus, ex quibus multæ dialecti, tanquam propagines deductæ sunt. Propagines quidem vnius matricis linguæ commercio inter se aliquo coniunctæ sunt: Matricum vero inter se nulla cognatio est, neque in verbis, neque in analogia’.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     

137

Figure 10.1 Joseph Justus Scaliger’s influential Diatribe on the languages of the Europeans Source: Archive.org. Public domain

language and were therefore secondary to it in terms of time. For this idea, Scaliger was most likely inspired by the work of his predecessors, even if it is impossible to name any precise sources. However, his usage of the Latin term propago, ‘offshoot, descendant’, in close connection with dialectus may have been inspired by Conrad Gessner’s observations on the Germanic family of languages. Scaliger might also have been influenced by certain sixteenth-century texts in which the interpretation was only latently present (see Chapter 4, Section 4.4; Chapter 5, Section 5.6). It was, however, only in Scaliger’s wake that the genealogical interpretation of the conceptual pair came into the limelight (cf. also Metcalf 2013: 52; Formigari 2004: 89). His oft-quoted Diatribe indeed seems to have been of central importance for this evolution, and Scaliger may rightly be regarded as playing a pivotal role in the emergence of the new language-historical conception of the language/ dialect distinction, still in vogue today. For instance, in his immediate environment, his colleague Paulus Merula assumed in 1605 that dialects were inherently posterior to a language, as, for instance, when he referred to the problem of how many original languages there were in the world:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

138

     

Figure 10.2 Scaliger c. 1607, engraving by Johann Theodor de Bry (1561–1623) Source: Rijksmuseum through Europeana. Public domain

Let us not elaborate on the number (which I believe to have been lower than Ephorus and others declare). This is certain, that those original languages, after Hebrew, have produced many dialects by means of migrations hither and thither.²

Here, Merula used the term dialectus for varieties genealogically related but geographically dispersed. Scaliger’s influence reached far beyond his intellectual network in Leiden, however. The Anglo-Welsh author and traveller James Howell, for example, likewise conceptualized dialect as a speech form historically derived from a language. Howell was without doubt inspired by Scaliger’s ideas. In fact, he repeatedly referred to Scaliger’s linguistic classification in his popular collection of largely fictional letters published under the title Epistolae Ho-elianae in 1650. ² Merula (1605: 207–8): ‘De Numero (quem minorem, quam Euphorus ceterique edunt, fuisse censeo) non laboremus: hoc certum, primigenias illas post Hebraicam multas Dialectos, Migrationum huc illuc ope, peperisse’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     

139

In an attempt at fine-tuning Scaliger’s account, Howell (1650: 78) at one instance claimed there were fourteen ‘vernacular and independent tongues of Europe’ as opposed to Scaliger’s eleven. Elsewhere, he suggested that the ‘mother toungs of Europe are thirteen, though Scaliger would have but eleven’ (Howell 1650: 91). Although disagreeing with Scaliger—and himself, for that matter—on the precise number of ‘mother toungs’, Howell did adopt quite naturally Scaliger’s historical conception of dialect throughout his writings. In his Instructions for forreine travell of 1642, he argued, for example, that a good knowledge of Latin was ‘behooffull’, adding that it greatly facilitated the study of Italian, Spanish, and French, which ‘are but Dialects or Daughters’ of it (Howell 1642: 21–2). Dialect and daughter were conceived as metalinguistic synonyms, an equivalence that had been largely absent from sixteenth-century usage. An increasing interest in linguistic diversity, classification, and history led early modern scholars to sedulously adopt the language/dialect pair in this new language-historical sense, especially from the early seventeenth century onwards, this in order to map out the historical interrelationships of the countless tongues appearing on their horizon. It comes as no surprise that in such cases dialectus was often accompanied by words meaning ‘offshoot’, ‘descendant’, and ‘daughter’. A typical early modern statement in which the family metaphor appeared was, for instance, the following: ‘the Latin language mother has produced as daughters and dialects Italian, Spanish, French, Rhaetian, Sardinian, and Romanian’.³ The use of botanical and family metaphors quickly gained momentum in language studies, culminating in nineteenth-century historical-comparative linguistics by conceptual cross-fertilization with other sciences, especially biology (cf. Chapter 17, Section 17.3). The view that a language generated dialects apparently contradicted the emergent idea that dialects tended to exhibit archaic features. A prominent proponent of the latter idea was Pierre Besnier, a French Jesuit whom I have introduced in the previous chapter. In his interesting booklet The reunion of languages, or The art of learning them all by means of only one, first published in Paris in 1674, Besnier treated, among other things, the method of comparing tongues. In doing so, he made the very uncommon suggestion that the cultivated varieties of French, Italian, and Spanish were often the most corrupted forms of these languages, and that among other things dialectal data could shed light on the obscured origin of many of their words (Besnier 1674: 23–5). This view of Besnier’s raises an important question: how can the view that dialects were archaic be reconciled with the historical conception of dialect as inherently posterior to language? At first sight, this idea seems to be a paradox, but it is only an apparent one, as language was conceptualized by Besnier from a more or less synchronic perspective as the ³ Kirchmaier & Jäger (1686: 17): ‘Latina lingva mater filias & dialectos peperit Italicam, Hispanicam, Gallicam, Rhæticam, Sardicam, VValachicam’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

140

     

recently designed normative variety from which dialects deviated rather than as the common ancestor of all the dialects. In fact, during the seventeenth century scholars started to increasingly differentiate between these two conceptualizations of language (cf. also Section 10.2), even though this distinction was usually not reflected in their definitions. The idea that a language diversified over time into related dialects naturally invited scholars to frame this process within larger timespans. Key to this zooming out was the assumption that related dialects gradually became so different from one another that they should be viewed as distinct languages rather than as mere dialects. The first scholar to problematize language change and diversity at length, the Dutch humanist Abraham Mylius, made the following observation already in 1612: In fact, the more distance is left from that time and place when and where this first change occurred, the greater also is the change that prevailed. This is why it is so necessary that languages, because of the excessive distance of both elements [sc. time and place], eventually degenerate entirely from different dialects into other languages, which do not retain anything of the former language anymore and barely even the likeness of a word.⁴

Mylius emphasized the vast differences existing between languages that were originally dialects of one and the same language, even up to the point that there were barely any similarities left. This observation did not prevent Mylius from calling these different languages dialects. In other words, speech forms differing on the level of language in Mylius’s time could be termed dialects because of their past status as dialects of one language, even though they now exhibited considerable differences. It is not difficult to see how this metonymic usage of the term dialectus must have often led to a confusion of synchronic interpretations of the word with its new language-historical meaning. Remarkable about Mylius is that he did not refer to Scaliger’s Diatribe, even though he entertained contacts with Leiden philologists and was acquainted with some of Merula’s work (Van Hal 2010a: 210, 231). The early presence of the language-historical interpretation of the term dialectus in Mylius’s Belgian language can therefore be taken as indirect evidence in favour of his familiarity with Scaliger’s work. There is, however, another possible explanation. Perhaps Mylius was influenced by the works of Theodore Bibliander and Conrad Gessner, which he had read (see e.g. Mylius 1612: 213). As I have contended in Chapter 5,

⁴ Mylius (1612: 85–6): ‘Quanto vero plus spatij ab illo tempore, quando, & ab illo loco, ubi prima illa mutatio contigit, est interpositum, tanto quoque mutatio major obtinuit. Id quod ita usus est, ut linguæ rei utriusque nimiâ distantiâ, ex diversis dialectis, tandem prorsus degenerent in alias, prioris nihildum retinentes, & vix quidem vocis alicujus simulachrum’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    . . .

141

Section 5.6, it is easy to read a genealogical interpretation of dialectus into some passages of these authors’ works, even though such a conception does not necessarily chime with the authors’ intentions. Whatever the case, there was a large group of late sixteenth- and early seventeenth-century authors who tried to develop a workable concept to designate a variety, dialect, genealogically deriving from another variety, language, a process that seems to have been strongly catalysed by Scaliger’s widely read work and, more broadly, by the emerging interest in linguistic diversity as a time-bound phenomenon.

10.2 An early critic of the language-historical interpretation The language-historical interpretation of the conceptual pair soon faced criticism, however, because it entailed the adoption of a different time perspective. The other conceptions of the language/dialect distinction were all synchronic, in that they concerned a snapshot in time, whereas the new meaning was the result of putting language and dialect into a historical scheme. An early critic of this language-historical usage of the conceptual pair was the Swiss orientalist and theologian Johann Heinrich Hottinger (1620–67; Figure 10.3). Having studied in his native land as well as in the Low Countries, including Leiden, Hottinger occupied the chairs of theology and Oriental philology in Zürich and Heidelberg. In 1667, he accepted a professorship at Leiden university, but his boat capsized on his way there. In this accident, he drowned together with three of his children. Some years before this tragic event, in 1661, he had published one of his great works, his Oriental etymological dictionary, in which seven languages featured. To this work, published in Frankfurt, Hottinger prefixed a dissertation explaining the goals, usage, and methodology of his lexicon, in which he also touched upon the relationships among the Oriental tongues included in his work. His view on the position of the Lingua Sancta, the ‘holy language’, in this linguistic group deserves to be quoted in full here: After all, from [the comparative study of Oriental languages], it manifests itself very clearly . . . that the holy language [Hebrew] is the mother of all other Oriental languages and that the remaining are daughters, either more or less degenerate. There are some who find pleasure in the nomenclature of dialect here. To Christian Ravis, in his Disputatio de Dudaim, p. 29, “the Hebrew, Chaldean, Syriac, and Arabic dialects appear to be one single language” [= Ravis 1656: 29]. “These dialects, which by the mistake of some people are now unfortunately dispersed, are believed to differ by a huge amount, even though they are connected like the Aeolic, Doric, Attic, and Ionic, the principal dialects among the others, to the common Greek.” Thus on p. 47 [= Ravis 1656: 47]. Yet I would prefer a mother–daughter relationship, on the one hand because we rightly claim

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

142

     

Figure 10.3 Johann Heinrich Hottinger etched by Georg Meyer, 1664 Source: Wikimedia Commons. Public domain

the first place for Hebrew, just like a mother, and on the other also because some deviate farther from Hebrew than that they can be called dialects. Among dialects, there is equality of privilege. With the authority of the mother intact, however, this cannot be given to the daughters. The dialects of the Greeks chiefly consist in the change of letters, but in some aspects of (at least some) Oriental languages there is such a great kinship with Hebrew as there is a greater difference in other aspects. Arabic certainly differs from Hebrew in the number and figure of letters, the number and form of endings, cases, the quantity of roots (similarly Arabic is richer here than all others), the equipment of conjugations, syntax, and formulas of speaking to such an extent that there is no dialect in Greece that departs from the common Greek by so many steps as the number of parasangs by which Arabic departs from Hebrew. It nevertheless does not cease to be a daughter, as it very splendidly expresses in other respects the inborn face of its mother, as becomes sufficiently evident from the lexicon itself.⁵ ⁵ Hottinger (1661: a.3–4): ‘Hinc enim manifestissimè patet, . Linguam Sanctam reliquarum omnium Orientalium esse matrem, cæteras filias, vel magis, vel minùs degeneres. Sunt, qui Dialecti

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    . . .

143

To summarize Hottinger’s somewhat complex account and its odd comparative constructions: Aeolic, Doric, Attic, and Ionic could be seen as dialects of the Koine Greek language, since they differed only by letter changes from it. In contrast, the Oriental tongues were to be considered different daughter languages of one common Hebrew mother, as they varied on many more levels than the letters alone, including morphology, roots, syntax, and phraseology. Hottinger assumed chronological primacy for Hebrew but not for the Greek Koine. In other words, he justifiably separated the language/dialect distinction from the mother/daughter contrast by comparing the Greek with the Oriental context. His subtle critique is especially remarkable, since it stands in startling contrast with his blind adherence to the age-old dogma of Hebrew primacy.⁶ Remarkable though Hottinger’s insights may be, they apparently did not exert any influence of note on the theorizing about the conceptual pair by his contemporaries and immediate successors. The dominant trend was to define dialect as opposed to language in synchronic terms, even though in actual practice many scholars used these concepts in a very confused manner, conflating synchronic meanings with the language-historical usage. Even if a few other early modern criticasters could be cited, it is not until the modern era that one finds more fundamental critiques of the distinction and the different interpretations attributed to it. That is, however, a matter for later chapters (see Part V).

10.3 A discursive strategy for historical classification The Renaissance ushered in a new era in which the linguistic horizon of scholars broadened remarkably (Chapter 4, Section 4.1). In order to cope with the seemingly endless variation languages could exhibit, polyglot philologists started to compare different languages and arrange them into groups, for which they

hic nomenclaturam amant. Christ[iano] Rav[io] Disput[atione] de Dudajim p. 29. Hebræa Chaldæa Syra & Arabica Dialecti apparent uno numero lingua, quæ jam infaustè quorundam vitio discerptæ, immane quantum dissidere creduntur, cum cohæreant, ut communi Græcæ Æolica, Dorica, Attica, Jonica præcipuæ inter reliquas. Ita p. 47. sed malim matris ad filiam σχέσιν, tum quòd Hebrææ, ceu matri, primas jure vendicemus, tum etiam, quòd longiùs ab Hebræa quædam abeant, quàm quæ Dialecti vocari possint. Inter Dialectos ἰσοτιμία. At hæc, salva matris authoritate, filiabus dari non potest. Græcorum dialecti in elementorum potissimum μεταβολῇ consistunt. At Orientalium Linguarum (quarundam saltem) ita magna in quibusdam cum Hebræa est γειτνίασις, ut in reliquis major sit differentia. Arabica certè ab Hebræa differt Elementorum numero & figura, terminationum numero & forma, casibus, radicum quantitate (sic Arabica locupletior hic est, quàm reliquæ omnes) conjugationum apparatu: constructione loquendi formulis, adeò ut nulla sit in Græcia Dialectus, quæ tot passibus à communi recedat Græca, quot quidem parasangis ab Hebræa abit Arabica. Neque tamen filia esse desinit, quòd in aliis nativum matris vultum, ceu manifestum satis ex ipso fit Lexico, luculentissimè exprimat’. ⁶ On the ancient origin of the idea that Hebrew is the primeval language see e.g. Van Rooy (2013: 30–1); Denecker (2017: 63–4).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

144

     

designed various conceptual means and frameworks. This early modern comparative study of languages is today often termed, somewhat teleologically, precomparative linguistics (Considine and Van Hal 2010: 63–5). One of the conceptual means devised was the language/dialect distinction, which allowed scholars to express various degrees of linguistic kinship and to classify language forms into larger groups of different natures. In the latter half of the sixteenth century, the language/dialect pair emerged as one of the principal classificatory devices used by scholars trying to create order in the vast linguistic diversity of the world, with a primary focus on European and socalled Oriental tongues. An early landmark in this development was Conrad Gessner’s Mithridates of 1555 (see Chapter 4). Joseph Justus Scaliger restyled the distinction between language and dialect even more explicitly in historical terms, as I have argued earlier in this chapter. Several other scholars discussed the concept of dialect out of an interest in linguistic kinship and diversity, and even more authors used language and dialect to express the affinity, or lack thereof, between a wide range of languages. For instance, the Swedish philologist Georg Stiernhielm (1671: f.1) applied the conceptual pair to tongues belonging to what is now known as the Uralic family of languages. Sometimes a scholar felt obliged to admit that there was not enough evidence to determine the precise status of a variety: was it a language, a dialect, or even a subdialect? Comparing a number of Prussian with Lithuanian words, the East-Prussian philologist Philipp Ruhig (1675–1749) confessed in 1745 that the lexical sample available to him did not allow him to determine the relationship between both tongues in more precise terms.⁷ The language/dialect pair proved to be a particularly welcome tool for early modern scholars wanting to make sense of the poorly documented Germanic and Celtic tongues anciently spoken in Germany and especially in Gaul. In this context, they often made use of the phrase ‘to differ only in dialect’ (Chapter 5, Section 5.5). For instance, the Reformed historian Caspar Peucer (1525–1602) claimed that the late antique Goths, Franks, Burgundians, Longobards, Alans, and others were all Germanic nations who spoke the same tongue, ‘as I do not distinguish them in language, but in idiom and dialect only’.⁸ Scaliger’s colleague Paulus Merula, in turn, discussed the linguistic situation of ancient Gaul as follows: On the language of the ancient Gauls, in as far as something certain can be stated, it should primarily be noted what Caesar writes in the first book on the Gaulish ⁷ Ruhig (1745: 53): ‘Hierbey wird nicht verdrüßlich seyn, auch einige Preußische mit dem Littauischen übereinkommende Wörter zu lesen, die Prætorius aus Grunovio anführet, welche aber noch nicht ausmachen, ob diese beyde als dialecti, subdialecti, oder mit vernünftigen analogischen Grammatischen Gründen von einander unterschiedene Sprachen seyn’. ⁸ Peucer (1562: 88–9): ‘Non enim lingua, sed idiomate tantum ac dialecto discerno Gottos, Francos, Burgundos, Longobardos, Alanos, Sueuos, Anglos, sicut reipsa linguis non discreparunt’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    . . .

145

war, i.e. that the whole of Gaul is divided into three parts: one of these is inhabited by the Aquitanians, another by those who are called Celts in their own language but Gauls in Latin, and the third by the Belgians. All these differ mutually in terms of language, habits, and laws. What that difference in language was, whether it was in dialects only or in the entire language, so that the Aquitanians, Celts, and Belgians each had a different language, has not yet been definitively disclosed by anybody, at least as far as I know. After all, who could state something sure and ascertained in a matter so obscure and remote from our age?⁹

Although stressing the thorny nature of the issue, Merula (1605: 419–21) nonetheless did propose a conjecture himself. He opted for a difference on the level of language. Yet all the Gauls spoke one language at first, Welsh, which the Aquitanians later exchanged for Basque and the Belgians for a variety of Germanic, with the Celts preserving their original Welsh language. Merula thus presented a nuanced conjecture grounded in historical developments. The manipulative usage of the conceptual pair and the phrase ‘to differ only in dialect’ was not restricted to ancient Europe, since a similar strategy surfaced in the work of scholars treating the linguistic diversity of the New World. It did so most notably in a controversy between two Dutchmen, the geographer Johannes De Laet (1581–1649) and the polymath Hugo Grotius (1583–1645). In this fierce polemic, the issue of language occupied a major position (see e.g. Laes and Van Houdt 2013). Grotius claimed that linguistic evidence proved a Scandinavian descent of the American nations. De Laet, however, countered Grotius’s argument by describing the vast linguistic diversity of New Spain as follows: their idioms ‘do not only vary in dialects, but are entirely different languages’.¹⁰ One language outshone all: Mexican, by which De Laet meant Nahuatl. He described its wide spread and noticed that ‘it has begun to be common, just like the Latin language once was in Europe and the Greek in Asia’.¹¹ De Laet was thus clearly putting the conceptual pair to use so as to describe the linguistic situation in New Spain. Earlier in his work, De Laet had expressed a similar view with regard to the continent of America as a whole: Another thing was the sheer infinite diversity not only of dialects, but even of languages, too, and there was also such a large variation among neighbouring ⁹ Merula (1605: 419): ‘De Veterum Gallorum Lingua, quo certi quid statuatur, notandum in primis quod Cæsar scribit li. . de Bello Gallico, totam Galliam in tris divisam partis; earumque Vnam incolere Aquitanos; Aliam, qui ipsorum lingua Celtæ, Romana Galli appellantur; Tertiam Belgas: eos omnes Lingua, Institutis, Legibus inter se differre. Quod in Lingua discrimen illud fuerit, an in Dialectis solum, an in tota Lingua, ut ea alia fuerit Aquitanis, alia Celtis, alia Belgis, nemo adhuc, quod ego quidem sciam, diffinite aperuit. Nam in re tam obscura tamque ab ævo nostro remota, quis certi quid & explorati statuerit?’ ¹⁰ De Laet (1643: 27): ‘non tantum dialectis variant, sed plane diversæ sunt linguæ’. ¹¹ ‘communis esse cœpit, quemadmodum in Europa olim latina, & in Asia græca.’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

146

     

peoples that they showed almost nothing in common among themselves and differed from each other entirely in terms of customs and practices.¹²

In other words, the existence of many different dialects as well as languages in the New World disproved, in De Laet’s eyes, a possible Scandinavian origin of the American peoples.

10.4 Conclusion During the late Renaissance, the conceptual pair language/dialect acquired a language-historical dimension, apart from its various synchronic interpretations. This new understanding had forerunners in sixteenth-century texts but gained widespread acceptance only in the course of the seventeenth century, partly due to the influence of Joseph Justus Scaliger and scholars in his network. The new genealogical interpretation remained in vogue not only throughout the early modern period, in spite of occasional criticism, but also in modern linguistics, even up to this day (cf. Chapter 17). Early modern scholars eagerly relied on the historical interpretation of the language/dialect distinction and especially on variations on the newly coined phrase ‘to differ only in dialect’ when trying to prove or refute the kinship among two tongues and, by extension, among the peoples speaking them. It was employed especially keenly in cases where there was room for speculation and disagreement, as was the case with the poorly documented tongues of ancient Europe such as Gaulish and with distant contemporary tongues such as the ones in the New World. In other words, the less familiar an early modern scholar was with a particular set of tongues, the more likely it was that he, wanting to give an impression of the relationship among the tongues in question, resorted to the conceptual pair as a welcome tool to describe or manipulate ethnic history in agreement with his scholarly programme and underlying assumptions. The diachronic interpretation must be framed within the historicalcomparative turn in early modern linguistic interests, which itself was, in part, a side effect of the renewed sense of history current among humanist scholars. Or to put it with the well-chosen words of Luce Giard, ‘the study of language participates in the general movement of historization impregnating Renaissance thought’.¹³ Indeed, linguistic diversification was increasingly seen as a cyclical historical process that took place gradually and into which the language/dialect pair also fitted. ¹² De Laet (1643: 4): ‘Accedebat pene infinita non modo dialectorum, sed & linguarum ipsarum diversitas, & inter vicinos quoque populos, tanta varietas, ut nihil fere inter se commune ostenderent; moribusque & institutis toto cœlo inter se discreparent’. ¹³ Giard (1992: 221): ‘la linguistique participe du mouvement général d’historisation qui imprègne la pensée de la Renaissance’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

11 Consolidation by elaboration Drawing the balance

In the previous chapters, I have attempted to unravel how the language/dialect pair became an established part of the metalinguistic apparatus of early modern scholars. I have mainly done so by developing the idea that scholars elaborated seven principal interpretations of dialect opposed to, and subsumed under, language in the years 1550–1650. In this chapter, I want to draw the balance of how scholars managed to consolidate the position of the conceptual pair in linguistic thought by 1650.

11.1 The seven main interpretations: a synthesis By way of summary I have listed in Table 11.1 the seven main interpretations of the conceptual pair devised and established by scholars in the century between 1550 and 1650. They are summed up more or less in the order in which they can be traced in the early modern source texts. It should be kept in mind that my systematized analysis of the matter does not exactly mirror accounts of the time, which tend to be not particularly organized and can be found in various contexts, usually in grammatical, philological, historiographical, and ethno-geographical works. Not all interpretations were put into practice as diagnostic criteria to determine whether a specific speech form was a language or a dialect. Whereas the Aristotelian criterion and mutual intelligibility were often employed to this end, the geographical and ethnic interpretations, rooted in traditional Greek definitions of diálektos, were not. Instead, they served to simply explain what an author took a dialect to be. Regularity, position in time, and status, although prominent in definitions of the conceptual pair, were likewise only rarely put to explicit use as criteria to set a dialect apart from a language. In fact, scholars often had preconceived ideas about the status of a particular linguistic entity, which is why they usually did not ponder at length on the language/dialect status of a speech form. The seven interpretations could be variously combined, all the more since none of them were mutually exclusive. Still, it is possible to group some of them together, as I have done in the preceding chapters. Not unsurprisingly, geographical and ethnic coverage often co-occurred, as they were both inspired by Greek definitions Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

148

  

Table 11.1 Summary of the main early modern interpretations of the conceptual pair Interpretation

Language

Dialect

Nature of linguistic differences (Aristotelian criterion) Geographical coverage Mutual intelligibility

Substantial

Accidental

Extended (Generally) absent among languages Regulated, analogical norm Group of nations Also: state Primary

Restricted (Generally) present among the dialects of a language Anomalous deviation One nation Also: part of the state Secondary

Superior, pure, literary ~ higher social classes

Inferior, corrupted, unliterary ~ lower social classes

Regularity Ethnic(-political) coverage Position in time (language-historical conception) Status

and can be regarded as two sides of the same coin: a dialect was spoken by a restricted group of people living in a narrowly delimited area. Further, mutual intelligibility can be regarded as a consequence of the Aristotelian criterion. These interpretations were indeed mentioned together by several scholars. What is more, numerous authors were aware that both features are gradual rather than strictly binary. Regularity and status, too, were closely linked and often even merged in the seventeenth and eighteenth centuries, even though not all scholars bracketed them together. This association is especially rare in cases where the literary Greek dialects occupied a prominent place. They were usually perceived as deviating from the analogical Koine, without, however, being stigmatized as inferior linguistic entities, as many vernacular dialects were. Somewhat peculiar was the language-historical interpretation of the conceptual pair, especially since this conception was the only one approaching the relation of dialect to language from a diachronic rather than a synchronic perspective. It was nevertheless not uncommon for later scholars to intuitively link it to the interpretations in terms of regularity and status, assuming that the analogical and superior norm, the language, was inherently prior to anomalous and inferior deviations from it, its dialects. The Dutch language scholar Lambert ten Kate (1674–1731), for instance, juxtaposed language to dialect synchronically in normative terms as well as diachronically in a genealogical sense but did not problematize this double usage in his otherwise highly detailed Introduction to the knowledge of the sublime part of the Dutch language. Ten Kate (1723: .57–8) spoke at one instance of the different Dutch dialects, which he distinguished from the common written language in his book. Elsewhere, however, he discussed

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.     . . .

149

pronunciation rules across different dialects, in which case he meant the historical dialects of the Germanic branch of tongues (ten Kate 1723: .165). Such diverging usage is further evidence for my hypothesis that early modern authors perceived the conceptual pair mainly as a flexible metalinguistic tool that need not be questioned. The interpretations of the conceptual pair emerging in the early modern period can be variously related to the Greek tradition, contemporary scholarly interests, and the sociolinguistic realities of Western Europe, a dimensional triangle visualized in Figure 11.1. Some understandings were inspired by traditional Greek definitions of diálektos, such as the interpretations in terms of geographic and ethnic coverage, or indirectly motivated by Renaissance grammars of Greek or Greek philosophical categories, as in the case of the Aristotelian criterion. Most interpretations were, however, primarily the product of early modern scholarly interests, even though they were often presented as ahistorical givens rather than as innovations, much like the conceptual pair itself. The language-historical interpretation in particular should be situated close to this pole, since its emergence was tightly connected to the emerging early modern interest in language history and diversity. Certain meanings of the language/dialect distinction, especially that of status, reflected first and foremost evolutions in the sociolinguistic realities of early modern society and were for a large part shaped by them. These circumstances included most notably standardization processes, the fact that the standard language spread very unevenly across different social classes, and the increasing importance of linguistic borders in the constitution of political states. Put differently, interpretations of the conceptual pair displayed the typical early modern blend of old and new, albeit in this case with a preponderance of new elements, even though scholars usually did not present them as such.

Greek tradition

Figure 11.1 The dimensional triangle shaping the early modern conceptual pair

Sociolinguistic realities

Scholarly interests

One could say that the appropriation of the Greek term diálektos and its interpretations was more or less completed in the course of the seventeenth century, a process not merely consisting in passive reception but displaying a major creative dimension. The word was, on the one hand, borrowed from Greek into Neo-Latin and most Western European standard languages. On the other hand and more importantly, the appropriation consisted in adding new meanings

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

150

  

to it. This semantic revolution primarily happened by the opposition of the term to words meaning ‘language’ and the resultant various interpretations of this metalinguistic distinction. One should, however, differentiate between presentday perspectives and those of the early modern appropriators of the concept. My analysis suggests that early modern scholars made considerable changes to the original meanings of diálektos, gradually disassociating the word from the Greek heritage. Even so, most of them seem to have assumed that they were simply adopting a Greek term without changing its semantics too much, if at all (Chapter 6, Section 6.5). Early sixteenth-century scholars had left implicit how they conceptualized a language as opposed to a dialect, even though it can be inferred from their work that they often equated language with the standard variety being designed. In the late sixteenth and seventeenth centuries, scholars started to be more explicit about the ways in which they understood the language pole. It may suffice to quote what the famed grammarian of German, Justus Georg Schottel, had to say on the language variety he was describing: The High German language, however, which we discuss and at which this book aims, is not a dialect properly speaking, but the German language in its essence, as the learned, wise, and expert men have finally regained it and use it.¹

This conception was projected back onto Greek as well. Tellingly, the Dutch Calvinist theologian and orientalist Johannes Leusden (1624–99) preferred to call the Greek Koine ‘the language in its very essence’ (lingua ipsissima) rather than ‘common dialect’ (dialectus communis; 1670: 86). The idea of a common language did not, however, become the only interpretation of the language concept. On the contrary, it seems wiser to regard the early modern language/ dialect distinction as comprising several conceptual oppositions, most importantly common (standard) language vs. dialect and language mother vs. derived dialect, but more marginally also language as a linguistic conglomerate consisting of several related dialects. It would therefore be misleading to speak of the early modern language/dialect pair, as if it were a fixed and universally accepted opposition, especially since several interpretations could be applied to it. It is for this reason perhaps better to think of the conceptual distinction as a kind of flexible matrix into which diverging and context-dependent meanings could be fitted. Its interpretations depended on various circumstances, such as, for instance, the linguistic context on which a scholar was commenting, his discursive aims, and his intellectual ¹ Schottel (1663: 174): ‘Die Hochteutsche Sprache aber/ davon wir handelen und worauff dieses Buch zielet/ ist nicht ein Dialectus eigentlich/ sondern Lingua ipsa Germanica, sicut viri docti, sapientes & periti eam tandem receperunt & usurpant’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.     . . .

151

outlook and underlying assumptions. What is more, one single scholar could make varying use of different interpretations of the conceptual pair, as the case of Lambert ten Kate cited above served to illustrate. I must note here that the interpretations I have distinguished for the early modern language/dialect pair have been distilled from various works of different genres, written by a wide range of scholars with diverging backgrounds and interests. These texts can nevertheless be categorized into two main groups. Firstly, there were a restricted number of scholars such as Georg Stiernhielm, who had an obvious and engaged interest in the distinction, although always subordinate to a higher, usually grammatical, philological, or historiographic goal. The case of Stiernhielm for this reason deserves a more thorough discussion, which it will be accorded in the next and final chapter of the third part of this book. Secondly, there was a large group of authors who were not actively involved with the interpretation of the conceptual pair, but whose works do contain revealing testimonies offering traces, however faint, of early modern ideas about the language/dialect distinction. Antoine Furetière’s dictionary definition of dialecte may count as an example of the latter group. Furetière did not offer an interpretation of the word because he had any special interest in it, but because he was trying to compile a dictionary of French words, and as it happened, dialecte had become a naturalized French word. In the years 1550–1650 the language/dialect distinction was fleshed out in a more straightforward and explicit manner than in the first half of the sixteenth century. This development continued during the last part of the early modern period, even though no totally new interpretations of the conceptual pair were proposed. As I argue in Part IV, the contribution of the period 1650–1800 lies elsewhere.

11.2 Emancipating the conceptual pair from the Greek heritage I have mentioned that only two interpretations of the conceptual pair can be directly traced back to Greek scholarship on the dialects, those emphasizing the limited ethnic and geographic coverage of a dialect vis-à-vis the language on which it was claimed to depend. This low level of indebtedness to Greek sources might suggest that in the consolidation period the concept of dialect was gradually emancipated from the Greek heritage and the study thereof. Did this emancipation indeed take place? And if so, what other indications of this tendency can be found? The most compelling evidence in favour of this hypothesis seems to be the introduction of two major interpretations of the language/dialect pair: the analysis in terms of status and the language-historical sense.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

152

  

The strongest indication of the emancipation from the Greek heritage is the seventeenth-century view that a dialect was a variety of speech corrupted from a language (cf. Burke 2013: 29–30). I have mentioned that the renowned German language scholar Justus Georg Schottel claimed that ‘in all dialects there is something faulty, which cannot occupy the position of rule in the language itself ’ and how Antoine Furetière defined dialecte as ‘particular tongue of a province, corrupted from the general or principal language of the kingdom’ (Chapter 9, Sections 9.1 and 9.3). Even though Furetière also referred to the Greek dialects, his interpretation of the term was clearly grounded in his native French context. This French anchoring emerges not only from the fact that he considered a dialect to be a corrupted form of speech but also from his mentioning a kingdom, which cannot be anything other than a reference to the form of government of France in his age. The idea that a dialect was a corrupted form of a language was almost entirely absent from earlier thought, no doubt because the traditional Greek dialects were not conceived as despicable speech forms. On the contrary, they were the linguistic media of a revered literary tradition. The emancipatory tendency emerges also from the work of scholars operating with the language-historical sense. A key figure here was, as I have established, the philologist Joseph Justus Scaliger. For his interpretation of dialect as an offshoot of a mother language, Scaliger was inspired neither by Ancient Greek dialectal diversity nor by Greek theorizing. What is more, Scaliger’s one-sentence discussion of the Greek mother language left the reader in doubt as to which episode in the history of the Greek language he was referring to: There are several idioms of the ΘΕΟΣ mother [sc. the Greek language; theós (θεός) means ‘god’ in Greek], and this is not astonishing with so many distant islands, which differ very much in terms of location as well as the interaction of language.²

Was Scaliger alluding here to diversity in the ancient or the contemporary vernacular language? The fact that he did not mention the canonical Ancient Greek dialects indicates that the last option is highly likely. This conjecture is further supported by his statement that he was taking into account eleven ‘language mothers’, since ‘today there are not more of them extant in the whole of Europe’.³ In other words, Scaliger operated with the distinction between a language mother and the dialects derived from it without mentioning the prototypical Ancient Greek context or relying on Greek scholarship. His silence most

² Scaliger (1610a: 121): ‘Matricis ΘΕΟΣ, pluria sunt Idiomata, quod non mirum in tot Insularum interuallis, quæ vt loco, ita linguæ commercio valde dissident’. ³ Scaliger (1610a: 119): ‘Sunto igitur nobis Matrices eæ, quæ per omnia inter se discrepant, cuiusmodi , non amplius hodie supersunt in vniuersa Europa’ [emphasis mine].

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.     . . .

153

certainly cannot be explained by a superficial knowledge of Ancient Greek on Scaliger’s part. In fact, he was an excellent Hellenist—he claims to have learned the language in only twenty-one days by reading Homer (Scaliger 1594: 56)—and evinced interest in the Greek dialects in other parts of his oeuvre (Grafton 1983: 102–4, 155). He was even creative with the language, coining a new Greek term to denote a variety subsumed under a dialectus, which we might call a subdialect: paradiastrophe¯ ́ (παραδιαστροϕή), a word that does not occur in extant Ancient Greek and Byzantine texts (Scaliger 1610b: 123). Scaliger’s case suggests that trained Hellenists, too, started to omit references to the Ancient Greek dialects and Greek scholarship where one would expect them to be present. The emancipation also surfaces from the tendency to provide idiosyncratic definitions of the term dialect, even in handbooks and dissertations on the Greek language and its dialects, which one might suspect to be rather conservative in this regard and which depended on Greek scholarship in many other respects. A revealing case in point is Erasmus Schmidt’s analysis of the word dialectus, part of which I have already discussed in Chapter 7. Schmidt enumerated four different meanings. For the first three, he referred to Greek texts. He mentioned Aristotle’s peculiar definition ‘articulation of the voice by means of the tongue’, the sense of ‘conversation’ present in Plato’s dialogues, and the meaning of ‘language entirely different from other languages’, for which he referred to the New Testament, Acts 2.6 and 2.8 in particular. Schmidt did not, however, use the word in these senses, as he himself indicated. He offered a fourth interpretation instead, for which he did not rely on Greek scholarship. His own definition unmistakably made a distinction between a dialect and a language, which was motivated by his assumptions about vernacular German and which he projected back onto the Ancient Greek language: Lastly, it sometimes denotes, like in the present treatise, a speech particularity of some one and the same language, which is different in the different provinces and cities using this language. For instance, the Saxons, Misnians, Thuringians, Silesians, Franconians, Swabians, etc. all use the German language; and still each of them has their own dialect, i.e. a peculiar way of writing, pronouncing, forming, constructing, or even naming as a whole.⁴

In contrast to Schmidt, the French Protestant scholar Matthieu Cottière (1581–1656) claimed that the original polysemy of the Greek word diálektos had ⁴ Schmidt (1604: 3–4): ‘Aliquando denique, ut in præsenti tractatu, notat idioma unius & eiusdem alicuius linguæ quod diversis Provinciis vel Civitatibus eâ utentibus, est diversum. Exempli gratia: Germanica lingua omnes utuntur, Saxones, Misnici, Thuringi, Silesii, Franci, Sueci, &c: & tame[n] singulis sua est Διάλεκτος, hoc est peculiaris, vel scribendi, vel pronunciandi, vel formandi, vel construendi, vel planè etiam nominandi ratio’. In the early modern period, Sueci usually denoted ‘Swedes’, but Schmidt is more likely to have meant the Swabians here (Suevi/Suebi).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

154

  

disappeared in the early modern period. Whereas the word designated both ‘language’ and ‘a certain affection of a language’ in Greek texts, it currently only conveyed the latter meaning, Cottière (1646: 208–15) contended. He thus undeniably displayed an awareness that there was a disparity between Greek and early modern usage, which makes him rather exceptional, since most scholars silently assumed conceptual continuity. However, like many of his contemporaries, he was reading the subsumption of dialect under language into the Greek sources on which he was relying. Some scholars clung to the Greek tradition and the Ancient Greek linguistic context. For instance, according to certain authors, the term dialect could exclusively refer to a variety of the Greek language. As a matter of fact, the Ancient Greek context remained the prototypical point of reference in discussions of the concept of dialect, even though it was, in its subsumption under language, an early modern innovation. What is more, the tendency to employ Greek as a reference context also holds true for more obviously new concepts such as standard language, which was occasionally projected back, anachronistically, onto the Ancient Greek Koine. A great friction between the Greek heritage and early modern conceptualizations is observable in several cases. A number of scholars even formulated idiosyncratic definitions of diálektos in Ancient Greek. By proceeding thus, they wanted to convey the impression that they were offering a received interpretation taken from Greek sources, even though they were in fact offering definitions of their own, which they had adopted from the Greek tradition and attuned to their own assumptions. A remarkable case in point is the Bruges-born scholar Andreas Hoius, professor of Greek at Douai university. Hoius provided his readers with the following definition, expressed partly in Greek: we call the character of the language with its own form of speaking dialect . . . Or to explain it in rather rough terms, a certain property of speech, by which a variety of speaking is distinguished, which usually and commonly exists between different regions of the same nation.⁵

As I have already argued (Chapter 7, Section 7.1), the second part of Hoius’s definition was, with slight adaptations, silently taken from an earlier Renaissance grammarian, the French Hellenist Petrus Antesignanus. The first part, however, appears to have been a product of Hoius’s own mind. Inspired by traditional Greek views, he formulated a definition presupposing a subsumption of dialect

⁵ Hoius (1620: 95): ‘Διάλεκτον dicimus τὸν τῆς γλώττης χαρακτῆρα, ἴδιον ἔχοντα τοῦ λέγειν σχῆμα, linguæ characterem, qui propriam habeat loquendi formulam. Sive, ut παχυμερέστερον explanemus, Sermonis quandam proprietatem, qua loquendi varietas distinguitur, quę inter diversos eiusdem nationis tractus plerumque solet existere’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.     . . .

155

under language—especially when read in connection with the subsequent quote from Antesignanus’s work. In sum, Hoius offered an idiosyncratic definition of the term diálektos, but he did so in Greek so as to keep up the appearance of traditionalism. I have already established how Antoine Furetière’s conceptualization of dialect as a speech form corrupted from a language was influenced by his native French context. It was unproblematic for Furetière to extrapolate this idea to the Ancient Greek context, even though the traditional Greek dialects were normally not seen as corruptions of the Greek language in the early modern era. This uncommon view might be related to a lack of acquaintance with the Greek context on Furetière’s part and, perhaps, to a scepticism about Greek culture in general. Indeed, the French lexicographer elsewhere stressed the superiority of contemporary over Greek philosophy (Stanton and Wilkin 2010: 30). However, even professional Hellenists struggled with the tension between the Greek tradition and their own conceptualizations and assumptions. For example, in a 1684 dissertation presented in Wittenberg, although stressing the similarities between Ancient Greek and German dialects, two German Hellenists associated the dialects of their native tongue with lower social classes (Kirchmaier and Crusius 1684: .2v–3). This connection contrasted with their assumption that the Greek dialects were esteemed literary varieties. In summary, there was a general tendency towards emancipating the dialect concept from the Greek heritage. Nevertheless, seventeenth-century authors often still associated dialectal variation closely with the Ancient Greek language. Furthermore, scholars frequently interpreted the Greek context by means of the language/dialect distinction. This analysis was, however, a projection of their own ideas onto a context distant in time and space from their own, betraying the assumption that their native vernacular dialectal diversity was comparable to Ancient Greek variation (cf. Van Rooy 2020a: Chapter 8). Additionally, early modern scholars elaborated on definitions from the Greek tradition, into which they silently introduced a distinction between language and dialect themselves. Put differently, even though theorizing on the conceptual pair increasingly stepped out of the shadow of the Greek heritage in the period 1550–1650, this emancipation often occurred in a highly ambiguous manner. Also, the Ancient Greek dialects continued to attract most scholarly attention throughout the seventeenth and eighteenth centuries for philological reasons. To gain better insight into the emancipation from the Greek heritage and the ambiguity of this process, it might be fruitful to resort here again to a process I have introduced in Chapter 8, Robert K. Merton’s (1968: 27–8, 35) obliteration by incorporation. Indeed, the de-Hellenization seems to display two main properties of this process. Firstly, scholars increasingly omitted references to the Greek sources in their definitions of dialect, even though these often

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

156

  

constituted the bases for their own interpretations. This independence contrasts with sixteenth-century practice, when scholars still either cited a specific Greek author or generically referred to ‘the Greeks’ in their definitions. Secondly, obliteration by incorporation is most likely to take place when the original contributors are renowned, which may also be said to hold true for the Ancient Greeks as a people, held in high esteem by Renaissance humanists. Applying Merton’s concept here has disadvantages, too. Most importantly, it presupposes modern methodological practice and therefore does injustice to certain aspects of early modern scholarly practice, in which matters such as originality and transparency in methodological and conceptual terms were not always prime concerns. Also, unlike the sciences with which Merton was concerned, the process of emancipation I have been describing here did not involve an autonomous field of study. Still, it seems useful to tie the deHellenization of the dialect concept to broader tendencies in the history of ideas, so as to arrive at a better understanding of this conspicuous transformation of the language/dialect distinction.

11.3 Conclusion The conceptual pair underwent two main transformations in the hundred years between about 1550 and 1650: a consolidation by elaboration of its interpretations and an emancipation from the Greek heritage. Even though I have opted to investigate these two evolutions somewhat distinctly with a focus on the first development, I should point out here that they often went hand in hand. Their compatibility can be neatly exemplified by repeating here the dialect lemma in Thomas Blount’s Glossographia, an influential English dictionary from 1656: Dialect (dialectus) is a manner of speech peculiar to some part of a Country or people, and differing from the manner used by other parts or people, yet all using the same Radical Language, for the main or substance of it. (Blount 1656: s.v. dialect; cf. already Chapter 8, Section 8.1)

Blount, on the one hand, provided a definition of his own, which combined several interpretations: the Aristotelian criterion as well as geographical and ethnic coverage. The Greek background was, on the other hand, not prominently present. Blount referred to the Ancient Greek dialects only after mentioning variation in English. What is more, he was entirely dependent on the geographer Peter Heylyn (1599–1662) for information on the Greek dialects. Blount’s definition suggests that scholars of Greek language and literature had lost their initial monopoly over the conceptual pair, with non-Hellenists increasingly contributing to its interpretations.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

. 

157

A major consequence of the de-Hellenization was that the seventeenth century brought an ever-increasing emphasis on the secondary nature of a dialect vis-à-vis a language in functional, grammatical, chronological, and sociocultural terms even if the etymological value of dialects was often stressed. In other words, a dialect was more and more regarded as deviating from, as well as inferior and posterior to, a language. Consider, for instance, the view, expressed in an early seventeenth-century dissertation On languages presented in Leipzig, that the Germanic language was ‘a dialect of none’ (‘nullius dialectus’), an assertion taken to imply correctness, purity, and antiquity (Bachmann and Ludovici 1625: .1). A negative interpretation of the dialect concept had initially been blocked by its close association with the admired literary Greek dialects, which is, however, not the entire explanation. The shift was also connected to the rise in prestige of the vernacular languages as a result of intensifying standardization processes. Normative varieties received an advanced grammatical codification and were ever more widely accepted and used, especially in France, where this evolution took place at the quickest pace. Linguistic standardization was encouraged by political centralization in unified states such as England and France in particular, as I have put forward in Chapter 7, Section 7.5. In the process, the conceptual pair sometimes received a political colour in these areas already during the seventeenth century. In more fragmented regions, too, scholars sometimes assigned a political meaning to the language/dialect distinction, albeit less prominently so. The German orientalist Hiob Ludolf (1624–1704), notably, criticized his former friend Andreas Müller (1630–94) for having included specimens not only of languages but also of dialects in his Lord’s Prayer collection of 1680. Interestingly, Ludolf alluded to the way in which one ought to distinguish between the two concepts in his manuscript notes, written down in his interleaved copy of Müller’s book: Whatever tongues only differ as to provincial pronunciation, are not found in any printed book whatsoever, and do not have their own king or lord, these do not deserve to be listed, as, for instance, Guelderish, which does not differ from Dutch, unless in terms of vulgar pronunciation.⁶

Ludolf put his ideas into practice and struck out certain specimens, for example of Swiss German, which he saw not as a language but as a dialect. The German scholar seems to have taken the conceptual pair principally in two senses here: the Aristotelian interpretation and the language-external criterion of cultural and political status. As opposed to languages, dialects differed from each other only ⁶ ‘Quaecumque linguae tantum prouinciali pronuntiatione differunt, nec in ullo aliquo libro impresso reperiuntur, nec suum proprium regem aut herum habent, eae non merentur referri ex. gr. Geldrica, quae non differt a Belgica nisi plebis pronunciatione’ (quoted from Van Hal forthcoming).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

158

  

superficially, were not used in writing, and could not invoke the authority of a political entity. Ludolf ’s suggestion that dialects ‘only differ as to provincial pronunciation’ may moreover imply that he presupposed mutual intelligibility among related dialects but not among languages. In sum, some seventeenth-century scholars politicized the conceptual pair to a certain extent. In doing so, they paved the way for the modern quip, ‘a language is a dialect with an army and navy’. However, I must agree with Peter Burke (2004: 7) that the politicization of both concepts remained a marginal phenomenon in the early modern period. Indeed, as I suggest in Part V (especially Chapter 23), this interpretation rose to prominence only during the nineteenth century. Before treating this major evolution, however, I have to turn my attention to the fate of the conceptual pair in the last 150 years of the early modern period and, to begin with, to the transitional figure of Georg Stiernhielm.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

12 The conceptual pair in transition The case of Georg Stiernhielm

The previous chapters have offered a synthetic account of the consolidation of the conceptual pair in the years 1550–1650. My choice of presentation has to a considerable extent been determined by the available evidence, which itself is rather fragmentary, since before 1650 few scholars reflected extensively on the distinction between language and dialect. After around 1650, certain authors started to provide more systematic treatments of the issue against the backdrop of philology and historiography. An early witness to this change is the Swedish philologist Georg Stiernhielm (Figure 12.1), whom I have briefly mentioned already several times (Chapter 8; Chapter 10, Section 10.3). His views are, however, remarkable enough to deserve a separate discussion in a historical overview of the conceptual pair, all the more since he marks the transition of the consolidation period (Part III) to the systematization period (Part IV).

12.1 Georg Stiernhielm, a Swedish all-round scholar Born as Göran (Georg) Olofsson in a small Swedish village in the Dalarna region on 7 August 1598, Georg Stiernhielm carved out a brilliant career in different branches.¹ After studies in Uppsala, Greifswald, and elsewhere, and after travels across Europe, Stiernhielm put his expertise and knowledge into practice by serving the Swedish state in many different capacities, mainly as a jurist but also as a soldier and archivist. A child of middle-class parents, he was able to acquire the status of nobility in 1631, taking on his new name Stiernhielm, meaning ‘star helmet’, to match this step forward. Although a man of many talents and despite his new status, he faced financial problems for almost his entire life, up until his appointment as director of the College of Antiquity (Antikvitetskollegium) in 1667, five years before his death on 22 April 1672. The saying on his tombstone ‘Vixit, dum vixit, laetus’—‘As long as he lived, he lived happily’—is therefore somewhat surprising. His poverty might not be unrelated

¹ On Stiernhielm’s life almost all writings available are in Swedish. See e.g. Swartling (1909), who discusses Stiernhielm’s scholarly activities at length.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

160

    

Figure 12.1 Georg Stiernhielm Source: Wikimedia Commons. Public domain

to his supreme arrogance and fickle character. He was, for instance, known to have incurred a nasty injury to his right arm after a bar fight in Dorpat, presentday Tartu, Estonia, in 1641. The damage he suffered was lasting and he was forced to write with his left hand for the rest of his life. Moreover, when in 1649 Stiernhielm returned to Stockholm after a military debacle and became part of the intellectual circle around Queen Christina of Sweden (1626–89), he fell out of grace because he insulted some of the queen’s confidants. Stiernhielm’s volatile character has, however, not influenced the reception of his work as a writer, especially his poetical oeuvre, for which he is mainly known today. What is more, Stiernhielm is even hailed as the father of Swedish poetry for his Hercules, a national epic composed in classical hexameters adapted to the stress accent of his native tongue. Stiernhielm, however, made a name for himself as a scholar, too, in the field of, among other things, mathematics and language studies. The wide range of his interests meant, however, that he did not achieve any fundamental contributions in most branches, and for this reason a nineteenthcentury compatriot of his, Erik Gustaf Geijer (1783–1847) even labelled him ‘a great

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.  ,  - 

161

beginner in everything’.² This judgement might be too harsh, since his contribution to philology and language studies seems to go beyond the level of a beginner, even a great one. His originality can, for instance, be easily recognized, when one reads two lengthy passages from his pen, which shed light on his conceptualization of the language/dialect pair: Here, we must first know how a language differs from a dialect. Languages mutually differ in substance, in their foundation as it were, but dialects differ in accident. Languages are distinguished by matter, dialects by form. And, to put it more plainly, every language is defined by its own words and roots of words, which are both unknown and foreign to another language. A dialect is a deflection of one language into a separate idiom of a certain people, which does not differ by the identity of roots and words, but through the formation of accidents, cases, endings, and the addition, detraction, transposition, and change of a letter or syllable in some words. To these features are added the often different composition and usage of the same words, the variation of accent and spiritus, and above all a pronunciation departing from the common one.³ I call dialect a notable difference of an idiom departing from a first language into different directions. And this difference certainly consists in the occasionally diverging usage of substantial and radical words and in their corruption to another meaning, in their composition, phrasing, declension, conjugation, pronunciation, and accent. If these elements, I say, are so varied that they constitute a remarkable difference but are even so understood, albeit in some cases with difficulty, they constitute a new dialect, even though it was not mixed up with any other foreign language. And for this reason, all languages have their peculiar dialects. Italian has Tuscan, Lombard, Venetian, and Neapolitan; French has Parisian, Toulousian, Picard, Norman, and Provençal; do German, Swedish, and so on not have so many dialects as they have provinces? Every language has its varieties. Yet I regard as different those languages that mutually differ by such a distance that the substantial words themselves and the accidental forms seem entirely different, to such a degree that, in general, those mutually conversing cannot understand each other in any way, unless through an interpreter. Dialects become like this over a long span of time, while mutually departing from each

² Geijer (1850: 160): ‘i allt en stor börjare’. ³ Stiernhielm (1671: c.4): ‘Hic, Lingua, quo differat à Dialecto, prænoscendum est. Linguæ inter se substantia, ceu subjecto; Dialecti vero Accidenti differunt. Linguæ Materia, Dialecti Forma distingvuntur. Et ut explicatius dicam: Lingua quæq[ue] vocabulis & ipsis vocabulorum radicibus sibi propriis, & alienæ linguæ incognitis & peregrinis, definitur. Dialectus est unius linguæ deflexus, in singulare Nationis alicuius idioma, radicum & vocabulorum identitate non differens, sed formatione accidentium, casuum, terminationum, litteræ aut syllabæ in quibusdam additione, exemptione, trajectione, mutatione, quibus accedit compositio, & usus earundem vocum sæpe diversus: Accentus ac Spiritus variatio, & super omnia, à communi recedens pronuntiatio’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

162

    

other further and further, a process in which, as with all things natural, the character of natural progression necessarily appears. As a matter of fact, two entities that were one, when deviating from their unity, the further they evolve and the more borders they cross, the greater the interval is by which they differ both from the unity, their origin, and mutually from each other.⁴

Stiernhielm introduced both original definitions of the conceptual pair in the preface to his edition of bishop Ulfilas’s fourth-century Gothic New Testament translation, published in 1671, one year before his death. He had, however, already drafted large parts of this text around the years 1651–3, at which time he conceived it as part of a planned work entitled Runa Suethica, an outline of which appeared in print in 1669.⁵ Stiernhielm’s preface treated the origin of languages and provided an outline of his language-historical programme, in which the conceptual pair played a key auxiliary role. It served as a handy tool for determining the historical position of different language varieties and their degree of kinship. More specifically, Stiernhielm used it to relegate the sacred language Hebrew to the status of dialect within the family of so-called Oriental tongues. As mentioned in Chapter 10 (Section 10.2), Hebrew was widely assumed to be Man’s original tongue before the modern period, which was why Stiernhielm (1671: c.3–d.4) felt compelled to defend his view at great length. Another main thesis of Stiernhielm was that his native Swedish was ‘the least altered representative of the ancient Scythian protolanguage’ (Considine 2008a: 308), revealing a clear patriotic bias on the part of Sweden’s first national poet. Stiernhielm thus showed himself to be a follower of the so-called Scythian hypothesis by suggesting that Swedish was a dialect of Scythian. This theory had received different interpretations in the early modern period, but the core assumption was more or less the same in all of its varieties: Scythian was the now lost protolanguage of most

⁴ Stiernhielm (1671: d.4–e.1): ‘Dialectum voco, idiomatis à prima lingua in diversa abeuntis, insignem differentiam. Quæ quidem differentia consistit in vocum substantialium seu radicalium diverso interdùm usu, & ad aliam significationem depravatione, compositione, phrasi, declinatione, conjugatione, pronunciatione & Accentu. Hæc, inquam, ita variata, ut insignem constituant differentiam, nec tamen secus intelligantur, licet in quibusdam ægrè, novam constituunt dialectum, etiamsi nulli aliæ linguæ exoticæ commixta fuerit. Atq́[ue] ea ratione, linguæ omnes suas habent peculiares dialectos. Italica Thuscanam, Lombardicam, Venetam, Neapolitanam; Gallica Parisiensem, Tholosanam, Picardicam, Nortmannicam, Provincialem; Germanica, Svedica, & quæ non tot habent Dialectos, quod [sic pro quot] habet Provincias? singulæ suas habent varietates. Linguas verò pro diversis habeo eas, quę eo inter se intervallo distant, ut ipsæ voces substantiales, & forma Accidentalis prorsus aliena videantur, adeò ut vulgò inter se colloquentes, nullo modo, nisi per interpretem, mutuo se intelligere queant. Tales evadunt dialecti longâ temporum morâ, longius longiusq́ue à se invicem recedentes, in quo, ut naturalia omnia, necessum est, ut progressionis naturalis indoles elucescat. Duo enim quæ unum fuere, ex unitate sua digredientia, quò longius & pluribus terminis progrediuntur, eò majori intercapedine, & ab unitate, primordio suo, & à se invicem, distant’. Eskhult’s (forthcoming b) English translation loosely inspired my renderings. ⁵ See Eskhult (forthcoming b) for an edition and English translation of the preface, taking into account manuscript testimonies. The Runa Suethica outline is Stiernhielm (1669).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

163

tongues of Europe and some Asian languages. The hypothesis thus anticipated the nineteenth-century concept of Proto-Indo-European in a number of respects.⁶

12.2 The conceptual pair according to Stiernhielm With the position of the language/dialect distinction in Stiernhielm’s languagehistorical investigations in mind, I can now look in more detail at the ways in which he interpreted it.⁷ Reading his two definitions and other parts of his writings, the reader discovers one essential difference from his predecessors and contemporaries that stands out immediately: the length and detail of Stiernhielm’s treatment of the conceptual pair. Furthermore, his main criterion for distinguishing between language and dialect was absent from sixteenth-century definitions. Languages, he argued, varied in substance, on the level of the lexicon and roots, whereas dialects differed accidentally, on the level of accent, pronunciation, letter mutations, morphology, the meaning of words, and phraseology. The latter kind of variation Stiernhielm (1671: e.2) compared to the different looks individual persons can have. I have identified this criterion as the Aristotelian interpretation in Chapter 8, for which Stiernhielm served as the introductory example. It is not this common understanding of the conceptual pair that deserves attention here, but the fact that he explicitly linked both poles to different linguistic features. Most of the features mentioned are probably well-known to modern readers, but one might call for further explanation. In the first passage Stiernhielm argued that related dialects differed by ‘the addition, detraction, transposition, and change of a letter or syllable in some words’. A long history lies behind this remark. Since Roman antiquity language scholars had identified the four main letter-change processes mentioned by Stiernhielm, which were usually known under the Latin phrase permutatio litterarum, ‘permutation of letters’ (Chapter 4, Section 4.4). Letter-change processes were used very flexibly to various ends, usually to formulate speculative etymologies of words or in the early modern period often also to prove the alleged kinship between two tongues. Stiernhielm had, however, a particular use for these letter changes, since he associated them with differences on the level of dialects rather than languages, thus restricting their applicability. In short, it was not Stiernhielm’s interpretation of the conceptual pair in terms of substance and accidents that was exceptional, but the fact that he connected the Aristotelian criterion explicitly and abundantly with specific linguistic features. ⁶ On the Scythian theory, its genesis, and its different interpretations see e.g. Metcalf (2013: 34–9); Swiggers (1984); Villani (2003); Considine (2010); Van Hal (2010a: especially 335–401, 473–5). ⁷ For Stiernhielm’s dialect concept see e.g. already Metcalf (2013: 49–50); Coseriu (1975: 8–9); Law (2003: 261); Burke (2004: 23); Eskhult (2020; forthcoming b). Swartling’s (1909) biography discusses his linguistic interests passim (pp. 35–6, 70–1, 76–84).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

164

    

Stiernhielm’s adherence to the Aristotelian interpretation surfaces in his usage of a variation on the ‘to differ only in dialect’ phrase, too. He argued that Aramaic, Canaanite, Egyptian, Arabic, and Ethiopian ‘not at all differ from Hebrew except in slight dialect’.⁸ The substantial or accidental nature of linguistic differences had consequences for mutual intelligibility, according to Stiernhielm, who linked these two criteria almost naturally (cf. Coseriu 1975: 9; Law 2003: 261; Chapter 8). Speakers of distinct languages wanting to communicate needed an interpreter, whereas speakers of related dialects were able to converse directly, even if they might encounter occasional difficulties. This last specification suggests that Stiernhielm regarded mutual understanding as a relative rather than an absolute binary property. In fact, Stiernhielm admitted in an interesting 1651 letter to his French Protestant colleague David Blondel (1591–1655) that sometimes dialectal diversity was so strong that mutual intelligibility was impossible. He cited Dalecarlian, the peculiar dialect of the Swedes from his native Dalarna region, as a striking example of such a lack of mutual understanding (Stiernhielm 1937–48: 137–8). Stiernhielm moreover viewed dialects as speech forms deviating from a point of reference, either a common language or a shared ancestral language. The latter seems to reveal that for Stiernhielm the ultimate foundation of the distinction was a genealogical one, which would be in keeping with his strong language-historical interests: a dialect was produced by, and therefore inherently posterior to, a language. The image of a language as a tree trunk with dialects for branches was invoked to make this idea more concrete, a metaphor of which I offer an impression in Figure 12.2. In this context, Stiernhielm described the evolution of a language into dialects, which in turn developed into separate languages, as an inevitable natural process, even occurring when no foreign language was interfering. For this continuous change Stiernhielm (1671: d.4, e.1–2) posited two main causes: distance in time and space.

Dialect B

Di al ec Language

ec al Di

tA

Figure 12.2 Impression of Georg Stiernhielm’s tree metaphor ⁸ Stiernhielm (1671: c.3): ‘nil nisi levi dialecto, differre ab Hebræa’.

tC

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

165

Elsewhere in his preface, Stiernhielm (1671: d.4) bracketed dialect together with province, suggesting that he associated the concept with a delimited geographical entity. Finally, he sometimes seems to have implied that a dialect was a faulty speech form, corrupted from a language, even though he did not seem to present it as an inherent property of a dialect. This interpretation can, however, be read into his usage of Latin phrases such as degenerare in dialectum, ‘to degenerate into a dialect’.⁹ In short, Stiernhielm provided his reader with many clues about his conception of the conceptual pair, which he centred around one main diagnostic criterion with language-historical implications: languages mutually differ in substance, which he principally connected to words and roots, whereas related dialects vary only in accidental terms. His account differed significantly from the often short and vague discussions of his sixteenth- and early seventeenth-century predecessors. Initially, the conceptual pair was closely tied up with the Greek language and scholarly tradition. In the course of the seventeenth century, however, theorizing on the conceptual pair was increasingly stripped from its Greek luggage, as I have tried to show in Chapter 11. What was Stiernhielm’s relationship with the Greek language and heritage? It immediately catches the eye that he barely mentioned Ancient Greek dialectal diversity. His acquaintance with the language seems to have been rather slight, to say the least. He limited himself to noting that Greek could be seen as a dialect of the Scythian protolanguage on the same level as Saxon and Gothic, both ideas for which he was relying on contemporary sources.¹⁰ In his explanation of the conceptual pair, Stiernhielm (1671: c.3–d.1, c.4-, d.4) in fact principally referred to the so-called Oriental languages as well as to Western and Northern European vernacular languages such as Italian, French, German, and his native Swedish tongue. It is striking that he did not name a single Greek dialect, whereas he did attribute dialects to the Hebrew language, an idea for which he was inspired by the shibboleth incident recounted in the Book of Judges. The fact that the Ephraimites were unable to pronounce the [ʃ] sound and when asked to say the word shibbóleth said sibbóleth instead was for Stiernhielm (1671: d.2) a letter permutation and therefore evidence of dialectal diversity within Hebrew (see Chapter 3, Section 3.4). This outline may create the somewhat misleading impression that Stiernhielm was highly original in his ideas on the conceptual pair, all the more so since he did not refer to any sources in his two lengthy definitions. This omission must be viewed in relation to the fact that like most early modern scholars he considered

⁹ For this phrase see the first tree diagram in Stiernhielm (1671), inserted after siglum f.4 in the edition I consulted in April–May 2016 (Gotha, Forschungsbibliothek, Theol 4o 00022e/01). ¹⁰ See Stiernhielm (1671: e.4), where both Casaubon (1650: 139) and Junius (1665: *.3) are quoted. Cf. also Eskhult (forthcoming b: 7.3.4) for a passing reference to the Greek dialects.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

166

    

the language/dialect distinction to be an evident datum; it need not be further problematized. It is nonetheless possible to identify a probable source of inspiration for Stiernhielm’s interpretation of the two concepts. His foregrounding of the Aristotelian criterion was likely inspired by the thought of the ambulant intellectual Christian Ravis, who expressed comparable ideas precisely in the years 1646–50, as I have contended in Chapter 8 (Section 8.1). As a matter of fact, Stiernhielm must have met this Berliner orientalist at the court of Queen Christina of Sweden in the early 1650s, right around the time he was drafting the first version of his preface on the origin of languages. He did not, however, mention Ravis by name. Another important source for Stiernhielm was the French classical scholar Claude de Saumaise, whose Commentary on the Hellenistic tongue of 1643 he did quote favourably in his preface when treating the emergence of new languages out of dialects (see Stiernhielm 1671: e.1-, quoting Saumaise 1643a: 18–19). In this work Stiernhielm may have found confirmation for his Aristotelian interpretation of the conceptual pair, as Saumaise (1643a: e.g. 356, 367) repeatedly employed variations on the phrase ‘to differ only in dialect’. Stiernhielm might have discussed the matter orally with Saumaise, who also spent some time at Queen Christina’s court in 1650–1. Whereas Stiernhielm’s relationship to his sources remains somewhat unclear, there is no doubt that his views on the conceptual pair influenced the thought of several later scholars. They were, most notably, picked up in the well-known dissertation On Europe’s oldest language, published in 1686 in Wittenberg and probably co-authored by Georg Caspar Kirchmaier (1635–1700) and Andreas Jäger (c. 1660–1730; see Considine 2008b on the issue of authorship). Kirchmaier, the supervisor, and Jäger, the presenting student, repeatedly referred to Stiernhielm’s preface in this work. In fact, their lengthy answer to the question ‘How does a language differ from a dialect?’ was quoted almost verbatim from Stiernhielm.¹¹

12.3 Conclusion Georg Stiernhielm devoted considerable attention to the conceptual pair, which he defined first and foremost in Aristotelian terms: distinct languages differ from each other in substance, whereas related dialects do so only accidentally. Stiernhielm was remarkable in that he tied both poles to specific linguistic categories—language to roots and words, and dialect to so-called accidental properties such as pronunciation and letter mutations. He noticed that this distinction also had implications for mutual intelligibility among speakers of ¹¹ See Kirchmaier and Jäger (1686: 17): ‘Lingua quomodò à dialecto differat? exposuit in ad Ulphilam præfat[ione] D[omi]n[us] Georg[ius] Stiernhielmus’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

167

distinct languages as opposed to related dialects. His language-historical programme almost naturally led him to conceive of the relationship between language and dialect in diachronic terms, too. More in the margin, he associated dialect with province and with degeneration. In short, he conflated a variety of interpretations in his conceptualization of the language/dialect distinction, but it is safe to say that the Aristotelian interpretation was his most important criterion, since it featured prominently in both of his extensive definitions of the conceptual pair. In outlining his views, the Swedish scholar, who exceptionally defined both dialectus and lingua, went about much more systematically and provided more thorough and elaborate analyses than his Renaissance predecessors. The criteria by means of which language could be distinguished from dialect were discussed more explicitly and were linked closely to linguistic features, a connection largely absent from earlier thought. Stiernhielm may therefore rightly be regarded as a transitional figure. He stood at the end of the consolidation period, during which the conceptual pair was definitively established as a metalinguistic tool through the explicit development of its main interpretations, and heralded a new phase in the history of the language/dialect distinction. Fully emancipated from the Greek heritage, theorizing on the conceptual pair now became much more systematized as well as rationalized. Before I treat both tendencies in more general terms in Chapters 15 and 16, I will elucidate them by means of two revealing case studies of eighteenth-century scholars with a deep concern over the language/dialect distinction, one a productive Dutch orientalist, the other a prominent German historian.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

IV

SYSTEMATIZATION AND R A T I O N A L I Z A T I O N , 1 6 5 0– 1800

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

13 Putting the conceptual pair on the scholarly agenda The orientalist Albert Schultens

On 22 June 1665, a German orientalist by the name of Levinus Warner ( c. 1618/9) died childless in Istanbul. Out of gratitude Warner left his manuscript collection to his alma mater in the Low Countries, Leiden university (see e.g. Vrolijk et al. 2012). His library contained over 900 books, predominantly in Arabic, but also in Greek, Hebrew, and Armenian, which Warner had found during his many years of service in Ottoman territory, first as secretary of a Dutch merchant and later as diplomat for the Dutch Republic. The books arrived in Leiden in different shipments between 1668 and 1674, but a lack of expertise and interest caused them to lie forgotten in a sweat cloth for decades. Fast forward to the summer of 1709: Albert Schultens (1686–1750; Figure 13.1), a 22-year-old scholar from Groningen, obtained his doctorate in theology at the academy of his hometown after studies there, in Leiden, and with the renowned orientalist Adriaan Reland (1676–1718) in Utrecht.¹ Schultens did so with a dissertation on a New Testament verse in the Gospel of Mark, in which he demonstrated his excellent mastery of Latin, Greek, and Hebrew. He was, however, not a classical trilingual man, since he also mastered Arabic, a tongue by which he was fascinated from his early years. A teenage prodigy, he matriculated at Groningen university at the age of fourteen and professed his love for Arabic in a dissertation held there when he was only nineteen years old. Indeed, he must have become convinced at an early stage that Arabic was the key to arriving at a better understanding of Hebrew. Schultens’s enthusiasm for Arabic no doubt also explained his return to Leiden in late 1709, intrigued as he was by the Warner treasure, which he must have seen in that city when he studied there some years earlier. Schultens was allowed to investigate the Warner collection in peace, which cannot but have stimulated his interest in the Arabic tongue. Yet to his regret he had to combine this study, from 1711 onwards, with a job as pastor in Wassenaar near Leiden. To escape this duty he accepted, in 1713, the chair of Hebrew and antiquities at the university of ¹ Biographical information on Schultens is principally drawn from the very useful but positively biased account of Wensinck (1921).

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

172

       

Figure 13.1 Albert Schultens, 1730 Source: Wikimedia Commons. Public domain

Franeker in the north of Frisia, where he further developed his idea that Arabic could elucidate Hebrew and started to put it into practice. Scholarship on Hebrew, Schultens believed, required a new impetus and perspective achievable by studying its lexicon through the lens of Arabic, which, he complained, had thus far been studied solely in isolation from Hebrew. This comparative approach was all the more essential since the Hebrew text corpus was largely limited to the Old Testament. After all, one did not study the Latin lexicon only through the works of Cicero either. Schultens devoted most of his energies to the comparative study of the Semitic lexicon, primarily through Hebrew and Arabic. A first major product of his research was a volume on Hebrew etymology published in Franeker in 1724. At this stage, he still saw Arabic as a daughter language of the sacred Hebrew language. The initial results of Schultens’s studies impressed the Dutch scholarly community, and in 1729 he received a double offer from Leiden university, which he gladly accepted. He became rector of the Collegium theologicum, an institute for impecunious theology students, and, more importantly, he took up the newly created position of interpres legati Warneriani, ‘interpreter of Warner’s bequest’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

       

173

This appointment put him in a prime position to study Warner’s manuscript collection. At the same time, Schultens’s advanced study of Hebrew and Arabic forced him to revisit his ideas on their relationship. He now argued, more correctly from the perspective of modern scholarly insights, that Arabic was a sister of Hebrew rather than a daughter, both descending from a now lost protolanguage. This new insight was, however, controversial, and his colleagues met it with bitter reproach, as most of them still considered Hebrew to be the uncontested primeval language of mankind (cf. Chapter 10, Section 10.2). Schultens remained in Leiden for the rest of his life, becoming professor of Oriental tongues in 1732 and of Hebrew antiquities in 1740. He continued to publish on Hebrew and Arabic, even though he did not manage to have all of his work printed. Some manuscripts of his, including his Theses on the primeval tongue, lie unedited at Leiden University Library.² Apart from more practical applications of his method, including editions and translations of Hebrew and Arabic texts and a grammar of Hebrew, he also produced a second volume on Hebrew etymology and a polemic ‘program of his principles for language research’, both published in 1738.³ Indeed, Schultens’s new method, especially the prominence of Arabic in it, was not appreciated by fellow theologians such as Jacques Gousset (1635–1704) and Anton Driessen (1684–1748) who claimed that for their discipline the study of Hebrew in itself sufficed (Eskhult 2015: 78–81). Schultens took it as his duty to defend his approach in several of his later works, and he did so successfully. Partly thanks to the scholarly dynasty he was able to establish at the Leiden academy through his son Jan Jacob (1716–88) and his grandson Hendrik Albert (1749–93) after his death on 26 January 1750, his method left an indelible mark on Oriental studies in the eighteenth century.⁴ Schultens’s profound interest in determining the precise relationship between Hebrew and Arabic heightened his sense of the concepts with which he was doing so. Especially in his later work, he was concerned over the metalinguistic tools he applied to this end. One of the concepts with which he engaged most intensively was dialect. As a matter of fact, he defined it at several occasions in different publications, offering the modern reader a transparent window on his ideas about the distinction between language and dialect. In this brief case-study chapter, I address two principal questions. Firstly, how did Albert Schultens conceive of the conceptual pair? Secondly, in what ways did his account differ from earlier discussions? To answer these questions, it is convenient to concentrate on his two

² The Theses will, however, soon be available in an edition procured by Josef Eskhult (forthcoming a). ³ See, respectively, Schultens (1738b) and Schultens (1738c). The description of Schultens (1738c) is Eskhult’s (2015: 74). ⁴ On Schultens’s comparative method see especially Eskhult (2015; forthcoming a). On Schultens’s school of Oriental studies see Nat (1929: 37–103).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

174

       

main publications of 1738: his second volume of Hebrew origins and his Ancient and royal way of Hebraizing.⁵

13.1 Schultens’s definitions of dialect In both of his 1738 works, Schultens discussed at length what he considered to be a dialect, which he did in the context of his new comparative approach to the Hebrew tongue. He believed it necessary to elaborate in detail on the matter, since learned men had thus far failed to understand correctly what a dialect was (Schultens 1738b: *.2). Convinced that the Oriental tongues differed from one another in much the same way as the Greek dialects did, he explained his interpretation of the term as follows in the second volume of his Hebrew origins, a work already finished by the end of 1737: . A dialect is without doubt nothing else and cannot be anything else in the view of all those who find pleasure in speaking accurately than a certain external and accidental variation of one language, which does not pertain to its internal substance, but preserves the foundation untouched and unharmed. Certainly, when this is ruined, altered, and corrupted, it will no longer be a dialect, but a corrupted and impure offshoot, such as that of the Italians and other peoples who have more or less corrupted and polluted the Latin language, overthrown by barbarian invasions from its foundations, so to speak. . These external variations, which happen to a true dialect, can be very conveniently referred to three classes. The first revolves around the elements of the letters, sounds, and manner of pronouncing. The second extends itself to the meanings of the words themselves. The third embraces syntax, patterns of speaking, and the entire discursive course.⁶

In his other work of 1738, which outlined his research programme and postdated his Hebrew origins, Schultens repeated his discussion in an even more schematized and systematized fashion, structuring it as a series of eight theses:

⁵ On Schultens’s views see also Eskhult (2015: 86–7; forthcoming a). ⁶ Schultens (1738b: 20): ‘. Dialectus nempe nihil aliud est, nec esse potest apud omnes qui accurate loqui amant, quam Unius Linguæ variatio quædam externa & accidentalis, quæ ad internam ejus substantiam non pertingit, sed fundamentum integrum illibatumque conservat; hoc enim labefactato, alterato, corrupto, non jam amplius Dialectus erit, sed corrupta & adulterina soboles, qualis est Italorum aliorumque populorum, qui Latinam Linguam invasionibus Barbarorum tanquam a fundamentis suis convulsam plus minusve corruperunt atque adulterarunt. . Hæ variationes externæ, quæ in veram Dialectum cadunt, commodissime ad tres classes possunt referri. Prima versatur circa elementa literarum, sonos, ac pronunciandi modum. Altera ad significationes ipsas verborum sese extendit. Tertia complectitur constructionem, loquendi formas, totumque orationis ambitum’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. ’   

175

. A dialect is a variation of one language which is only external and accidental and does not touch its internal substance, let alone that it taints and corrupts it. . If the root and marrow are tainted, they no longer have to be called dialects, but degenerate offshoots which have flowed away to an entirely other nature. . Whoever would call the Latin language, even though it is connected to the Greek by a very strong bond, a dialect of the same language and would add this language as a fifth to Attic, Ionic, Doric, and Aeolic, would gain praise neither of skill nor of judgement. . The same mistake is made by those who are accustomed to call Italian, Spanish, French, and English speech dialects of the Latin language. . One commits an even bigger blunder, when proposing nearly the same relationship of the Italian, Spanish, French, and English speech vis-à-vis the Latin mother, as of the Chaldean, Syriac, and Arabic dialect vis-à-vis Hebrew. . Finally, one makes the greatest crash, when one, together with the famous Gousset and Driessen recently, thence, as it were from an unconquerable stronghold, assails and attacks the use of the Oriental dialects in elucidating Hebrew. . The affinity and indeed blood-relationship of the Chaldean, Syriac, and Arabic dialects with Hebrew is equally close as that of Ionic, Doric, and Aeolic with Attic. . Assuredly, whoever nevertheless in this way denies that the Hebrew language is to be clarified by means of the dialects, denies the principles which are beyond all doubt conveyed, and these they could not or would not want to call into doubt anywhere else.⁷

⁷ Schultens (1738c: 109): . Dialectus est unius Linguæ variatio tantum externa, & accidentalis, quæ internam ejus substantiam non attingit; nedum ut eam vitiet, atque corrumpat. . Si Radix & Medulla vitietur, non amplius Dialecti appellari debent, sed degeneres propagines, quæ in aliam plane indolem defluxerunt. . Qui Latinam Linguam, triplici licet vinculo Græcæ innexam, appellaret ejusdem Dialectum, atque Atticæ, Ionicæ, Doricæ, Æolicæ, accenseret quintam hancce, nec peritiæ, nec judicii laudem auferret. . Idem peccant, qui Italum, Hispanicum, Gallicum, Anglicum Sermonem, Linguæ Latinæ Dialectos appellitant. . Gravius adhuc offendunt, quum Italici, Hispanici, Gallici, Anglici Sermonis eandem fere statuunt relationem erga Matrem Latinam, quam Chaldaicæ, Syriacæ, & Arabicæ Dialecti erga Hebræam. . Gravissime denique impingunt, quando cum Cl. Gussetio, & Driessenio nuper, inde, tanquam ex Arce invicta, usum Dialectorum Orientalium in Hebræa illustranda impetunt, atque oppugnant. . Adfinitas, atque adeo consanguinitas Dialectorum Chaldaicæ, Syriacæ, Arabicæ, cum Hebræa, æque arcta est ac Jonicæ, Doricæ, Æolicæ, cum Attica. . Qui vel sic tamen negat Linguam Hebræam ex Dialectis illustrandam, Principia negat supra omnem aleam subvecta: & quæ nuspiam alibi posset velletve in dubium revocare’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

176

       

The definitions reveal that Schultens was fairly traditional in interpreting the distinction between language and dialect. His views remind of Georg Stiernhielm’s to a certain extent, in that the contrast centred around the Aristotelian criterion (see Chapters 8 and 12). As a matter of fact, Schultens was familiar with the Swede’s work, which he quoted in his manuscript Theses on the primeval tongue (Eskhult forthcoming a: §). Even though the Aristotelian interpretation took centre stage in his definitions, Schultens (1738b: 104; 1748: ) also seems to have associated dialect with limited geographical coverage and used the conceptual pair in language-historical terms in his work. Before him, Stiernhielm, too, had foregrounded the meaning in terms of substance and accidents in his definition, although also assuming that dialect had limited geographical coverage and genealogically derived from a language. Schultens displays several notable differences vis-à-vis Stiernhielm, however. Let me highlight the three most important ones for our purposes. Firstly, Schultens seems to have associated language more closely with analogy, a feature dialects lacked, even though they could help to uncover the original analogy of a language. He did so in the long preface to one of the practical applications of his method, his edition of the Hebrew text of the Book of Proverbs with a Latin translation and extensive commentary (Schultens 1748: –). Secondly, Schultens did not regard mutual intelligibility as an intrinsic characteristic of related dialects. Instead, he suggested that the more time passed, the more it decreased, that is to say, when a language gradually and insensibly—‘sensim sine sensu’, one of his favourite phrases in Latin—developed into several dialects.⁸ A third difference with Stiernhielm was that Schultens seems to have been wellinformed about the Greek dialects, which he compared ad nauseam with Oriental diversity.⁹ He paralleled the four Greek dialects Attic, Ionic, Doric, and Aeolic to Hebrew, Aramaic, Syriac, and Arabic. This Greek–Oriental linking indicates that what I have dubbed the de-Hellenization of the conceptual pair was a gradual and uneven process. Even though Stiernhielm had already proposed a definition without referring to the Greek background, Schultens, a trained Hellenist, still frequently resorted to it (cf. Chapters 11–12).

13.2 Language, dialect, and degenerate offshoot If it is not in his interpretation of the language/dialect pair that Schultens was innovating, where was he? I argue that he was responsible for two major ⁸ See e.g. Schultens (1739: 235; 1748: ; in Eskhult forthcoming a: §, §, §); Eskhult (2015: 87). ⁹ See e.g. Schultens (1738a [1729]: 36; 1732: 4; 1737: 5; 1738b: 21, 104; 1738c: 106–8; 1739: 187, 189, 190–1, 222, 234; 1748: , ; in Eskhult forthcoming a: §, §, §§–). See also Nat (1929: 39).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. , ,   

177

innovations with consequences for the history of the two concepts. Firstly, Schultens’s definitions were part of an attempt at developing a linguistic method to determine whether a variety was a dialect or a degenerate offshoot of a language, taken here in the language-historical sense of ‘common ancestor’ (Chapter 10). In other words, Schultens operated with a tripartite hierarchy with implications for linguistic chronology (Figure 13.2) rather than with a two-way conceptual distinction.

Corruption

Time

Language (e.g. Greek, Latin, primeval language)

Dialect (e.g. Aeolic, Arabic)

Degenerate offshoot (e.g. French)

Figure 13.2 Impression of Albert Schultens’s tripartite conceptual hierarchy

Schultens differentiated a dialect from a degenerate offshoot in terms of purity. The latter was corrupted, whereas the former preserved the foundation of the ancestral language intact. On the one hand, his positive conception of dialect is likely to relate to the long-standing literary traditions of the speech forms he termed dialects: varieties of the Greek and Oriental language families. On the other hand and more importantly so, the distinction was informed by Schultens’s research agenda. After all, the argument that the Oriental tongues were uncorrupted dialects of a now lost primeval language served to justify the fundamental principle of his language studies, a comparative approach to etymology, which in Schultens’s eyes made sense only if the comparison was based on evidence from uncorrupted varieties. In this context, Schultens repeatedly referred to a linguistic situation he claimed did not exemplify dialectal variation, the relationship of the Romance tongues to the Latin language, which he analysed in terms of corruption in a typically early modern manner (see e.g. Neis 2009). His argument implied that the Romance tongues could not be unproblematically used for the study of Latin etymology. Remarkably, Schultens included English among the Romance tongues, which might be explained by the fact that this language has many French and Latin loanwords as well as by the fact that Schultens’s knowledge of it must have been rather limited. Schultens found it apparently important as well to point out that Latin was not a dialect of Greek. Acknowledging their close relationship, he

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

178

       

refused to follow the age-old tradition going back to antiquity of so-called Aeolism, according to which Latin derived from Greek through its Aeolic dialect (Stevens 2006–7).

13.3 Classes of linguistic variation A second major innovation transpires from the specifications Schultens added to his definition of dialect in his Ancient and royal way of Hebraizing, which elaborated on an earlier brief remark made in his second volume of Hebrew origins: Certainly those external variations, which produce a true dialect, happen either to the elements or manner of pronouncing, or to the meanings of words, or to the entire construction, whence they can all be referred very conveniently to three classes. . In the first class, diversities which play around the bark exhibit themselves, 1. by means of a letter change; 2. by means of a vowel change; 3. by means of a slight transposition of both; 4. by means of a contraction of form; 5. by means of a lengthening of the same; 6. by means of a shifting of the accent. . The second class, which pertains to the leaves and foliage, produces even more variation. 1. In some domestic words, which each people is found either to have received as proper to itself or to have coined for itself. A rare kind. 2. Very frequently, there is variation in secondary meanings, which are introduced through metaphors and other tropes. 3. Hence, such a great distinction often exists, that there seems to be a blatant contradiction between the significations of words which in their origin nevertheless agree very pleasingly. 4. One dialect preserves that origin better than the other. Thus, just as in one dialect there are very few [primary meanings], the same can be found very copiously in the other. . The third class of variations is constituted by: 1. syntax and the fashion of construction, which is often very discordant with, and different from, the innermost agreement of meanings; 2. formulas which inevitably arise in each people out of its domestic rites and affairs; 3. the entire phraseology, in such a manner adjusted to the genius of each people that just as an Attic writer is immediately distinguished from an

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

179

Ionic, a Syrian author can be likewise immediately distinguished from an Arab, and both of them from a Hebrew author.¹⁰

To paraphrase, related dialects exhibited only superficial or external variations, which concerned firstly letters and their pronunciation—the bark of a tree— secondly the meaning of words—the foliage of a tree—and rarely the words themselves, and thirdly syntax, which Schultens did not equate with a tree part. The first class seems to be a variation on the age-old permutation of letters framework, comprising pronunciation and covering also such features as accent and prosody (Chapter 4, Section 4.4). For the second class, the variation of meanings of words across related dialects, Schultens may have drawn on early modern handbooks of Ancient Greek and its dialects, in which this type of semantic variation was frequently mentioned, even if no obvious source can be singled out.¹¹ What is more, he seems to have been one of the first scholars to include it in a systematic treatment of the levels of dialectal variation outside such Greek-centred handbooks. To further refine this class of variation, he made a distinction between primary and secondary meanings, which varied across dialects. He assumed that some dialects preserved the original meanings better than others, and that individual dialects tended to introduce different secondary meanings. The third class involved syntax, which Schultens mainly identified with phraseological and idiomatic properties, features he linked to the customs and ¹⁰ Schultens (1738c: 110–11): ‘Nempe variationes illæ externæ, quæ veram Dialectum efficiunt, incidunt vel in Elementa & pronunciandi modum; vel in significationes verborum; vel in constructionem universam: unde ad tres classes cuncta commodissime revocari possunt. . In prima classe offerunt se diversitates circa Corticem ludentes, 1. Ex immutatione literæ. 2. Ex vocali permutata 3. Ex levicula trajectione utriusque. 4. Ex contractione formæ. 5. Ex productione ejusdem. 6. Ex Accentu aliter & aliter posito. . Altera classis ad folia & frondes pertingens, adhuc plus varietatis parit. 1. In vocibus quibusdam domesticis, quas quæque gens sibi vel proprias adscivisse, vel novasse, reperitur. Genus rarum. 2. Frequentissime variatum in significationibus secundariis, quæ per Metaphoras, aliosque Tropos, invectæ. 3. Tantum hinc discrimen sæpe existit, ut directa fronte pugnare videantur potestates verborum, quæ in Origine tamen amicissime conspirant. 4. Istius Originis alia Dialectus præ alia servantior: sic ut in una paucissimæ, in altera copiosissimæ, eædem inveniantur. . Tertiam classem varietatum constituunt: 1. Syntaxis, & ratio constructionis sæpe cum intima significationum concordia valde discors, & dissonans. 2. Formulæ ex ritibus rebusque domesticis in quaque gente oriundæ. 3. Phraseologia universa, cujusque gentis genio ita attemperata, ut quemadmodum Atticus scriptor ob [sic] Ione primo adspectu discernitur, sic Syrus ab Arabe, & uterque ab Hebræo Auctore statim discriminari queat’. ¹¹ Possible sources include Zwinger (1605: .2) and Ursin (1691: 496).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

180

       

genius of the people speaking the dialect. Schultens did not make explicit how distinct languages varied exactly, which is in line with the broader early modern tendency of leaving the overarching concept of language usually undefined. In sum, Schultens was rather exceptional in devoting extensive and systematic attention to the three ‘classes’ of variation in which related dialects allegedly differed from one another.

13.4 Schultens’s legacy As I have mentioned before, Schultens was able to establish a scholarly dynasty at Leiden university, with his offspring continuing his philological programme. However, his influence reached far beyond his own family, both within and outside the Dutch Republic. Schultens was the teacher of numerous students who managed to build successful careers as orientalists. Was he also able to instil into them his ideas about dialect and how it related to language and degenerate offshoot? It most certainly seems so. Two convincing pieces of evidence can be cited: his teaching and the writings of his pupils. Schultens must have professed his views on what constituted a dialect not only in his writings, both published and manuscript, but also in his Hebrew courses at Leiden university. A unique manuscript preserved at the library of this institution contains lecture notes ‘on Hebrew origins and other topics’ taken by the 23-yearold Lodewijk Caspar Valckenaer (1715–85), a Frisian student of Schultens, who later made a name for himself as a Hellenist.¹² On the fifth folium of this manuscript, Valckenaer noted some of Schultens’s ideas about the interpretation of the term dialectus. These notes, taken right before rushing to Friesland on 6 July 1738, as Valckenaer informs us, are very much in line with the views Schultens expressed elsewhere in his published work, especially, nota bene, in his two monographs published in the same year in which Valckenaer wrote down his notes. These student notes suggest that Schultens did not limit himself to discussing the topic in his publications, but he also actively taught it to the audience of his Hebrew courses at Leiden university. Also in their published work several of Schultens’s pupils prove to have been influenced by their teacher. A number of his students, including the Hellenists Tiberius Hemsterhuis (1685–1766) and Lodewijk Caspar Valckenaer, as well as the orientalist and missionary preacher Jan Willem Kals, discussed the concept of dialect and related issues in their work, albeit to various degrees of dependency on their master’s ideas. Whereas Schultens’s influence on the ideas of the Hellenists Hemsterhuis and Valckenaer was at most limited, the orientalist

¹² See Leiden, University Library,  469., fols. 1–23, ‘de originibus Hebraeis et aliis’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

181

Kals’s conception of dialect as opposed to language was very much indebted to Schultens’s views (van der Linde 1987: 98). Defining the Latin terms lingua and dialectus in the spirit of his teacher, Kals (1752: 9–10) claimed that a correct definition of them was only of secondary importance. Instead, he mainly attempted to prove, very much unlike his former professor, that Hebrew in fact was the primeval language (Kals 1752: 27, 61; van der Linde 1987: 101). He also criticized Schultens for propagating two contradictory views, failing to see that Schultens had actually distanced himself from his initial idea that Hebrew was the primeval language (Kals 1752: 62). Faithful to Schultens’s method, Kals (1752: 67) nevertheless admitted that Hebrew had only been partly preserved, stressing that its dialects were necessary to recover its lost part. He also agreed with his master that language change occurred gradually and insensibly, even taking over his favourite phrase verbatim (Kals 1752: 9–11). Kals’s work, published in Oxford, extended Schultens’s influence far beyond the borders of the Dutch Republic. The list of Schultens’s students influenced by his idea of what constituted a dialect can be enlarged further still. For instance, his Swiss theology student Antoine-Noé de Polier de Bottens (1713–83) authored a dissertation, orally defended in Leiden under Schultens’s supervision in 1739, on the purity of Arabic in comparison with Hebrew, which he both saw as descending from the primeval tongue. Polier consciously opted to speak of the Arabic dialect and not language, a contrast requiring explanation to his mind. The definition of dialectus he provided was very obviously influenced by that of his esteemed teacher (Polier de Bottens 1739: 4). The evidence presented here suggests that Schultens actively encouraged his students to reflect on the concept of dialect. He might have discussed his ideas with travelling scholars, too. It is, for instance, not unlikely that he met the Danzig-born orientalist Benjamin Groddeck (1720–76) passing through the Low Countries in the 1740s. At any rate, Groddeck was the supervisor of a dissertation defended in 1747 in Wittenberg, in which the nature of dialects was discussed with special reference to Hebrew and Arabic, and for which Schultens’s conception of the term and his ideas on the Oriental tongues constituted without doubt the primary source of inspiration (compare Schultens 1738b: 20–1 with Groddeck and Treuge 1747: –).

13.5 Conclusion The Dutch orientalist Albert Schultens took exceptional care to procure his reader with a systematized account of the semantics of the language/dialect pair. Schultens did so because he was dissatisfied with existing definitions and usages of the term dialectus. He went beyond a simple binary contrast, however, transforming the conceptual distinction into a tripartite hierarchy consisting of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

182

       

language, dialect, and degenerate offshoot. His interpretation of these concepts revolved around the Aristotelian criterion, with a clear language-historical dimension, and was, as in Stiernhielm’s work, closely linked to linguistic features, ‘classes’ of variation in Schultens’s own words. His tendency to systematize theorizing on the issue is not only apparent from the content of his works but also emerges from the way in which he presented his ideas typographically. Indeed, his discussion was visualized in a highly schematized fashion, as one can gather especially from his Ancient and royal way of Hebraizing (see Schultens 1738c: 109–10 in particular). Schultens can deservedly be regarded as a scholar who put theorizing on dialect on the scholarly agenda, even though it remained a matter of secondary importance only, as the writings of his pupil Kals suggest. The distinction between language and dialect was indeed first and foremost a tool for early modern scholars, who usually did not problematize it. Its auxiliary status is abundantly clear from the work of a late eighteenth-century historian, who tried so hard to offer an objective interpretation of the conceptual pair that he made its arbitrariness painfully obvious.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

14 Lexicostatistics avant la lettre The historian Johann Christoph Gatterer and the conceptual pair

The protagonists in this book have thus far been primarily philologists fascinated with ancient texts and languages, usually Greek and sometimes Semitic. In the eighteenth century, the linguistic horizon broadened still further. Scholars developed an increasing interest in the dialects of other, mainly vernacular tongues, especially in Scandinavia and German-speaking lands but also in Britain, France, and Italy. There, high-quality grammatical and lexicographical descriptions of dialects began to appear. One of the best-known is no doubt the lexicon of Swedish dialects by Johan Ihre (1707–80), published in 1766. This new interest was fuelled partly by curiosity, partly by intensifying antiquarian and historiographical activities. Some intellectuals went to great lengths to uncover ethnic history and looked at all possible types of evidence in their pursuit. A case in point is the Bavarian historian Johann Christoph Gatterer (1727–99; Figure 14.1), professor of history at Göttingen university for the largest part of his career (1759–99) and from December 1766 onwards also director of the Göttingen Historical Institute (see Dohna 1964 for biographical data). Gatterer owed this position to the renown of his informal historical seminar, in which he propagated the writing of comprehensive history. This universal historiography should be soundly founded on the critical study of different kinds of sources, Gatterer believed, and not only on the work of predecessors, as was often the case in the early modern period. He stressed the importance of ancillary sciences such as chronology, heraldry, and geography, thus paving the way for modern historical scholarship (Grafton 2007: 189–91; Iggers 2011: xix; Gierl 2013: 285). One of the types of evidence Gatterer valued highly was linguistic in nature. More specifically, he contended that a critical study of languages and linguistic kinship helped scholars compose an encompassing history of past societies.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

184

   

Figure 14.1 Johann Christoph Gatterer Source: Wikimedia Commons. Public domain

14.1 Gatterer’s Vorrede In March 1770, Gatterer held a discourse at the Göttingen academy in which he emphasized the utility of languages for writing ethnic history.¹ His lecture was published shortly thereafter under the German title Vorrede, von der historischen Benutzung der Sprachen, or Preface on the historical usage of languages in English. The text was dated 6 April 1770 and prefixed to the thirty-fourth volume of the monumental General world history (Allgemeine Welthistorie), a titanic historiographical project in which Gatterer was closely involved. He later integrated sections of his Preface into his Introduction to synchronistic universal history of 1771 as part of a section on human languages, of which he provided an extensive overview per continent (Gatterer 1771: 101–64, esp. 105–11). Gatterer intended his Preface to be a foundational text. Clearly, he thought very highly of language as a historiographical source, in all of its forms: ¹ Cf. Muller (1984a: 41). For Gatterer’s language-related views see Brekle et al. (1992–2005: .216–18).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.  ’  

185

There is, as is known, a large amount of languages. They are all worthy of a philosopher’s investigation, also the languages of savages, man-eaters. Philosophizing on languages is, in a sense, nothing else than philosophizing on reason itself.²

Gatterer (1770: 5) believed in a close connection between language and people, considering the former ‘the reflection’ (die Abspiegelung) of the latter. For this reason, language comparison had great historiographical utility. Not only did a competence in many languages make one a better historian, but there was also a key historiographical axiom regarding languages and peoples ‘which is not as known as it deserves to be’: ‘Peoples speaking one single or closely related languages belong to one and the very same ethnic stock’.³ This axiom was of especial importance when one tried to reveal the history of prehistoric and mythical times. Gatterer viewed himself as elaborating on Gottfried Wilhelm Leibniz’s (1646–1716) views in the latter’s Brief account of reflections on the origins of peoples, primarily drawn from the evidence of languages (Leibniz 1710). Leibniz did not have the necessary instruments to state something substantial on this matter, even though he should be praised for trying and for inciting later scholars to imitate him (Gatterer 1770: 6; see also Trabant 2006: 190). After his introduction, Gatterer briefly commented on specific languages and language families with reference to earlier scholarship. In some cases, the comparison of languages had already occurred and the kinship of peoples had been demonstrated, such as among the so-called Oriental tongues. Gatterer stipulated the existence of an Oriental people on the basis of the close linguistic kinship between Hebrew, Aramaic, Syriac, Arabic, and Ethiopian. In other cases, linguistic comparison was still required, for instance in order to determine the precise relationship between Romans and Greeks. Still, in both cases, definitive answers could only be found when adequate instruments were available, by which he meant comparative grammars and dictionaries. The comparative tables designed by Gatterer, which he mentioned in his Preface, but which were unfortunately not printed together with it, were intended to illustrate the historiographical use of such linguistic tools. In this part of his Preface, Gatterer also resorted to the concept of dialect to describe certain linguistic entities and relationships. For instance, he stressed the conservative nature of German dialects vis-à-vis the book language (Büchersprache), which was as mutable as fashion in clothing. Gatterer even hyperbolically stated that one could hear Otfrid and Williram,

² Gatterer (1770: 4): ‘Es gibt bekannter massen eine Menge Sprachen: sie sind alle der Untersuchung des Philosophen würdig: auch die Sprachen der Wilden, der Menschenfresser. Ueber Sprachen philosophiren heist gewisser massen nichts anders, als über die Vernunft selbst philosophiren’. ³ Gatterer (1770: 6): ‘Ich sehe hier auf eine Art von historischer Benutzung der Sprachen, die nicht so bekannt ist, als sie es zu seyn verdient. Völker, die einerley oder sehr verwandte Sprachen reden, gehören zu einem und eben demselben Völkerstamme’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

186

   

two medieval German authors, speaking among the Swabians, Franks, Bavarians, and Austrians of his day, which he claimed to know from experience. He also interpreted a number of what are now called the Baltic languages in terms of dialect, just as he did with the Germanic tongues and some Celtic varieties (Gatterer 1770: 5–7). Gatterer then moved on to more general considerations touching on his intriguing method of language comparison and the concepts with which he operated.

14.2 Characteristic words: Gatterer and basic vocabulary As he wanted to put historiography on a par with natural sciences in terms of objectivity, Gatterer (1770: 13) spelled out in detail how he believed accurate language comparison should occur. He observed that ‘in language investigations and also elsewhere, even in common speech, the expressions “cognate languages”, “dialects of a language”, “identical words”, and “different words” occur’. He asked himself in this context: What, then, are called languages? What are called dialects? How far does the kinship of languages need to go to be able to rightly say that these or those languages are not merely cognate languages but only dialects? Since I have not found anywhere a thorough philosophical discussion of this topic, I thus want to venture some thoughts on it myself and present them to experts for further investigation.⁴

In other words, Gatterer’s discussion was triggered by a discontent with previous, superficial treatments of the issue. He proceeded by analysing the linguistic entity of the word. It had a meaning, since it referred to a certain concept (Begriff), and a form, materialized by letters (Gatterer 1770: 13–14). These two aspects were the touchstones one ought to employ in order to determine whether a word present in two or more languages was in fact the same or different. One should, however, proceed methodically. Technical, religious, and onomatopoetic words should be avoided, suggested Gatterer, thus proposing an exclusion criterion. Instead, one should resort to what Gatterer called ‘characteristic words’ (characteristische Wörter), the basic vocabulary of a language, to put it with a phrase from modern

⁴ Gatterer (1770: 13): ‘In Sprachuntersuchungen, und sonst auch, selbst im gemeinen Reden, kommen immer die Ausdrücke vor: verwandte Sprachen, Dialecte einer Sprache, einerley Wörter, verschiedene Wörter. Was heissen nun verwandte Sprachen? Was heissen Dialecte? Wie weit muß die Verwandtschaft der Sprachen gehen, bis man mit Rechte sagen kan, diese oder jene Sprachen sind nicht blos verwandte Sprachen, sondern sie sind nur Dialecte? Weil ich nirgends etwas gründliches hierüber philosophirt finde, so will ich selbst hierin einige Gedanken wagen, und Kennern zu weiterer Prüfung vorlegen’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

187

linguistics. This group of words, Gatterer added, included five categories according to the consensus of all language philosophers: numerals; pronouns; the verb ‘to be’; the most essential words denoting body parts, indispensable pieces of clothing, tools, blood-relationships such as ‘father’, and things one can normally see on a daily basis like ‘the sun’; and ‘root words’ (Stammwörter). These characteristic words were about 300 in number (Gatterer 1770: 15). Gatterer was not the first early modern scholar to work with the concept of basic vocabulary, which has its roots in the seventeenth century, but he was highly specific about it and its methodological use.⁵ The identity of words either was visible and therefore immediately clear or needed to be revealed by means of a principle Gatterer called Reduction. By means of this ‘reduction’, words seemingly different in terms of either meaning or form or in both respects were proved to have an identical origin. In case of semantic divergences, the reduction could be done by showing that the meaning of the words in question derived from one basic idea or agreed by a certain kind of analogy. When there were formal differences, reduction entailed showing that letters had been transposed, added, or removed according to certain rules introduced into the language, or were simply pronounced differently. Gatterer thus still relied on the traditional framework of letter permutations, even though he specified that these must be in agreement with certain language-specific rules. He concluded this train of thought by exemplifying his distinction between visible lexical identity and lexical identity in need of reduction (Gatterer 1770: 14–15).

14.3 Determining the degree of linguistic kinship: Gatterer’s lexicostatistic framework The discussion of the ‘characteristic words’ and the principle of ‘reduction’ constituted the fundament for Gatterer’s analysis of the different degrees to which speech forms could display kinship: Starting from this assumption, I believe to be able to provide some rules by means of which one can determine the distinction between languages and dialects, and between related and unrelated languages, and at the same time the degree of kinship. 1. Languages in which only a few characteristic words are the same—whether this identity is visible or reduced—are unrelated languages and even fewer dialects. These few identical words merely indicate that the human race once

⁵ On the early modern concept of basic vocabulary and its history see Muller (1984a: especially 41–2 for Gatterer’s views; 1986: 17–19).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

188

   

spoke only one language. In some words, the identity can also be merely accidental. 2. Languages in which either half or almost half or in any case up to one-third of the characteristic words are identical can be regarded as related languages. One immediately sees how the degrees of kinship can increase and decrease. 3. Languages in which more than half of the characteristic words are identical can be regarded as dialects of one language. 4. Languages in which two-thirds or more than two-thirds of the characteristic words are identical can be called closely related dialects and their kinship increases or decreases, 1) depending on whether the number of characteristic words is over or under two-thirds and 2) depending on whether the identity is more or less visible. The more visible the identity of the characteristic words is in terms of meaning and letter form, the less forced the reduction is, and the more rarely the reduction needs to happen in order to prove the identity, the surer one can be that languages in which all this can be retrieved belong together as dialects; and the more it happens, the more related the dialects themselves are.⁶

The word Dialect, explained Gatterer, was taken from Sprachphilosophie, ‘language philosophy’. Sprachphilosophie referred in this context to general reflections on language, which in the eighteenth century could comprise historicalcomparative approaches, too (cf. Van Hal 2015: 57–8). Gatterer (1770: 6) used this terminology from language philosophy to elaborate his own linguistic ideas,

⁶ Gatterer (1770: 15): ‘Dieß vorausgesetzt, glaube ich im Stande zu seyn, einige Regeln zu geben, wodurch man den Unterschied zwischen Sprachen und Dialecten, und zwischen verwandten und nicht verwandten Sprachen, und zugleich die Grade der Verwandtschaft bestimmen kan. 1. Sprachen, in denen nur wenige characteristische Wörter eine Identität, sie sey nun sichtbar, oder reducirt, haben, sind nicht verwandte Sprachen, noch weniger Dialecte. Diese wenigen identischen Wörter zeigen nur an, daß einmal das menschliche Geschlecht nur Eine Sprache geredet hat: in einigen kan auch die Identität blos zufällig seyn. 2. Sprachen, in denen die characteristischen Wörter entweder zur Hälfte, oder nahe gegen die Hälfte, oder allenfalls bis zum dritten Theile identisch sind, können für verwandte Sprachen gehalten werden. Man sieht zugleich, wie die Grade der Verwandtschaft steigen und fallen können. 3. Sprachen, in denen die characteristischen Wörter über die Hälfte identisch sind, können für Dialecte Einer Sprache gehalten werden. 4. Sprachen, in denen zwey Drittheile, oder mehr als zwey Drittheile der characteristischen Wörter identisch sind, können nahe verwandte Dialecte heissen, und ihre Verwandtschaft steigt oder fällt, 1) je nachdem die Menge der characteristischen Wörter über oder unter zwey Drittheilen ist, und 2) je nachdem die Identität mehr oder weniger sichtbar ist. Je sichtbarer die Identität der characteristischen Wörter in Bedeutung und buchstäblicher Gestalt ist; je ungezwungener die Reduction ist, und je seltener die Reduction geschehen darf, um die Identität darzuthun: desto gewisser kan man seyn, daß Sprachen, in denen alles dieses anzutreffen ist, als Dialecte zusammengehören, und je mehr es statt findet, desto verwandter sind selbst die Dialecte’. On Gatterer’s definitions see also Muller (1984b: 394).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      

189

which he formulated as one approach within a more encompassing historiographical programme. It is surprising that Gatterer, even though his discussion was triggered by historiographical concerns, offered a mainly synchronic, ahistorical analysis of linguistic interrelationships. A consequence of this focus was that he did not seem to consider a dialect to be posterior to a language like many of his predecessors and contemporaries. The main criterion on which Gatterer relied to distinguish between language and dialect was a linguistic one, and more specifically the degree to which the ‘characteristic words’ of a speech form, which Gatterer generically called Sprache, differed from those of other varieties. The historian Gatterer relied more narrowly on lexical data than the orientalist Albert Schultens and put the framework of letter permutations to use as an etymological principle rather than as a class of variation, as Schultens had done. The quadripartite division into unrelated languages, related languages, dialects, and closely related dialects indicates that Gatterer was looking for objective, linguistic, and quantifiable criteria to determine the degree of linguistic kinship between two speech forms, with a remarkably low language/dialect threshold at 50%. In modern lexicostatistic research, this cut-off point is usually much higher (cf. Chapter 20). Gatterer’s distinctions still remained arbitrary, however. As a matter of fact, he was simply dividing the continuum from ‘no kinship’ to ‘identity’ into four arbitrary categories, which I have visualized in Table 14.1. There was, in other words, a certain tension between the need for laying down conceptual categories in a quantifiable way and the implicit recognition of their relativity and arbitrariness. Gatterer’s pioneering attempt at quantifying linguistic distance by lexical means fitted in with the ambitious programme of the Göttingen school of history and its members, who frequently resorted to statistics in order to analyse historical data (Cheng 2008: 130). Unfortunately, Gatterer does not seem to have actively and systematically applied in his published work the embryonic lexicostatistic method he outlined, but a closer study of his extensive oeuvre might reveal whether it shines through in his preserved scholarly output. Although trained in classical as well as Oriental languages (Dohna 1964: 89; Debus 1998: 10), Gatterer refrained from referring to the Greek dialects in his Table 14.1 Johann Christoph Gatterer’s four categories on the kinship continuum no kinship 0%

identity 33,3%

unrelated languages

50%

related languages

66,6% dialects

100%

closely related dialects

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

190

   

conceptual and methodological outline. As is apparent from his other work, he knew that the extinct ancient language had four dialects: Aeolic, Doric, Ionic, and Attic (Gatterer 1771: 119). He moreover tried to assess the relationship of this language to others and considered it closely cognate to Latin, Germanic, Celtic, and Slavic (Gatterer 1770: 7). It is nonetheless clear that, unlike Schultens but very much in line with the general tendency towards emancipation from the Greek heritage, Gatterer developed his thoughts independently from the Greek context. This autonomy is not very surprising. Gatterer in fact pointed out that he was elaborating a new conceptual framework, motivated by a lack of satisfaction with existing discussions of the matter. His lamentation that there were barely any dialect lexicons and grammars (Gatterer 1770: 9) should no doubt be seen in connection with it and may betray a lack of acquaintance with the early modern dialectological tradition in the philological study of Ancient Greek. Other scholars in Gatterer’s Göttingen circle who engaged in language study likewise refrained from referring to the Greek language in discussing the dialects of other tongues. For instance, the orientalist Johann Gottfried Eichhorn (1752–1827) was an alumnus of the Göttingen academy and member of its school of history, whose mainly historiographical dissertation on Arabic dialects made no reference whatsoever to the Greek language (Eichhorn 1779).

14.4 The historian Gatterer and the grammarian Adelung Gatterer’s ideas on linguistic kinship left their mark on the thought of the influential German grammarian Johann Christoph Adelung (1723–1806). In his 1781 German grammar, Adelung suggested a threefold rather than quadripartite conceptual division, splitting the kinship continuum into different languages, related languages, and dialects. In doing so, Adelung partly adopted the terminology of Gatterer, but he linked it to other linguistic features than the historian had done: If two languages mutually agree in their root words and their flexion and derivation syllables entirely, i.e. with a few exceptions, and the difference merely exists in vowels and cognate consonants, then they are only dialects of each other. If the deviation also covers consonants other than cognate ones and there are noticeable differences in derivations and flexions, then they are related languages, a relationship which has, in its turn, its manifold degrees. Entirely different kinds of derivation and flexion and greater or smaller difference in roots and their meaning result in more or less different languages.⁷

⁷ Adelung (1781: 71): ‘Wenn zwey Sprachen in ihren Wurzelwörtern, Biegungs- und Ableitungssylben im Ganzen, d. i. biß auf einzele Ausnahmen, mit einander übereinstimmen, und der Unterschied bloß in den Hülfslauten und verwandten Hauptlauten bestehet, so sind sie bloße

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

191

Instead of operating solely with Gatterer’s characteristic words, Adelung opted to include in his analysis not only ‘root words’ but also other linguistic features. He considered morphological differences in particular to be an important factor in the constitution of distinct languages, whereas many previous scholars had included flexional variation primarily under the traditional framework of letter changes on the level of dialect. In his Detailed doctrine of the German language, published in 1782, Adelung repeated the statement from his German grammar quoted above. Here, however, it was preceded by an extensive quote from Gatterer’s work. More specifically, he cited Gatterer’s (1771: 106–11) entire explanation of his method for determining linguistic kinship (Adelung 1782: 239–44). The way Adelung introduced the quote suggests that he admired Gatterer’s innovative propositions. Yet Adelung did formulate a reservation, since he was not entirely convinced by the ahistorical use of linguistic data propounded by the historian Gatterer: Despite all of this, it is still not possible to draw from the mere kinship of languages, without the contribution of history, any certain conclusion about a common descent, however much Mister Gatterer (1771: 105) considers this sentence an axiom not in need of any further proof. It is, after all, possible that one people is forced to adopt the language of another, and history learns that this has happened very often. How can we deduce in this case a single origin from a single language?⁸

In other words, Adelung objected that linguistic interrelationships could not be the only criterion to stipulate a common origin for different peoples and emphasized the importance of history itself. Paradoxically, it was the grammarian Adelung making this point in reaction to the historian Gatterer.

14.5 Conclusion Even though historians and ethnographers generally took the conceptual pair for granted when describing nations, regions, and past societies, a small number of Mundarten von einander. Erstreckt sich die Abweichung auch auf andere als verwandte Hauptlaute, und finden sich in den Ableitungen und Biegungen merkliche Unterschiede, so sind es verwandte Sprachen, welche Verwandtschaft denn wieder ihre mannigfaltigen Stufen hat. Ganz verschiedene Arten der Ableitung und Biegung, und größerer oder geringerer Unterschied in den Wurzeln und ihrer Bedeutung geben mehr oder weniger verschiedene Sprachen’. ⁸ Adelung (1782: 243–4): ‘Bey dem allen läßt sich doch aus der bloßen Verwandtschaft der Sprachen ohne Mitwirkung der Geschichte noch kein sicherer Schluß auf eine gemeinschaftliche Abstammung machen, so sehr auch Herr Gatterer S. 105 des angeführten Werkes diesen Satz für ein Axiom hält, welches keines weitern Beweises bedürfe. Es kann ja einem Volke die Sprache eines andern aufgedrungen seyn, und die Geschichte lehret, daß solches sehr häufig geschehen. Wie läßt sich in diesem Falle von einerley Sprache auf einerley Ursprung schließen?’

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

192

   

them did engage in theoretical considerations of variable length on the question as to how the language/dialect distinction could be put to use in their discipline. The most fine-grained historiographical approach of the early modern era was developed by Johann Christoph Gatterer around 1770. The main principle on which Gatterer relied was a well-known one: the close connection between a people and its language and, more specifically, the way in which language could help one reconstruct early human history. A fundamental assumption here was that peoples speaking tongues that were claimed to vary only as dialects rather than as entirely different languages had a common origin. Gatterer’s innovativeness lies in the fact that he tried to make this idea more workable and concrete, while grounding it in an objective mathematical approach in connection with the concept of basic vocabulary. He thus developed a lexicostatistic method avant la lettre, which became popular only much later, in the 1950s (see Chapter 20). His ahistorical perspective on linguistic data, however, attracted criticism from Adelung. Gatterer’s stress on linguistic criteria in determining the status of a speech form should no doubt be seen in connection with the rationalist trends of the eighteenth century. In fact, his wide-ranging historiographical programme was inspired by Enlightenment principles (Gierl 2012). Conceived as an ancillary science for the study of history, Gatterer’s Sprachphilosophie, too, was informed by this intellectual movement. His development of a statistical method for language study likely constitutes the best proof of this. Framing language and dialect in an elaborate method for language comparison, the Göttingen historian provided the most rationalized early modern account of the conceptual pair.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

15 Classes of variation How do languages and dialects differ?

Albert Schultens and Johann Christoph Gatterer, although operating in different contexts and disagreeing in several respects, exhibit three remarkable similarities in their treatment of the conceptual pair. First, both scholars insisted on linguistic properties in determining what distinguished a language from a dialect. This focus on actual language betrays a desire to arrive at more objective, linguistically motivated criteria. Second, they both attempted to systematize their discussions, which also emerges from the schematized typographical setups of their texts. Third, Schultens as well as Gatterer framed the language/dialect distinction within a broader conceptual scheme—Schultens within a tripartite hierarchy and Gatterer within a quadripartite division of the degree of kinship continuum. Are these three similarities between Schultens’s and Gatterer’s ideas coincidental? Or do they exemplify more general tendencies of the later stages of the early modern era? I argue that they are indeed part of a broader trend, an idea which I develop further in this chapter and the next. I first focus on the emergent interest in what Schultens called the ‘classes of variation’, the ways in which related dialects were believed to differ from one another, whether or not in opposition to distinct languages.

15.1 Casting off John the Grammarian’s yoke While it is true that eighteenth-century scholars such as Schultens and Gatterer evinced a keen interest in the classes of variation, it does not mean that their predecessors did not elaborate at all on them. What is more, in order to understand eighteenth-century developments, I should briefly go back in time. The respects in which the Ancient Greek dialects differed from one another had already been systematized to a certain extent in the early Byzantine period. In Chapter 2 (Section 2.3), I have indicated how John the Grammarian distinguished three classes: the entire word, part of a word, and an accident of a word. Put differently, the Greek dialects were claimed to differ in lexical terms, through letter permutations, and through differences in accent and other diacritic marks. Since John the Grammarian became one of the three canonical sources for information on the Greek dialects, his framework was naturally adopted by Renaissance Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

194

  

Hellenists, often literally and unquestioningly, which happened, for instance, in a booklet on the Greek dialects dating to 1569 and published in France (Vuidius 1569: 138). Some philologists extended the framework so as to include other classes of variation. The German professor of Greek Otto Walper (1589: 4–5), for instance, added syntax to John’s three classes in his manual for the dialects aimed at the students of Marburg university and first published in 1589. Theodore Bibliander (1542: 52–3) likewise offered a general account of how dialects varied, according to him on the level of an entire word, of part of a word, and of idiom. His discussion was mainly inspired by John the Grammarian, but Bibliander expressed his views when reflecting on the confusion of tongues at the Tower of Babel rather than on the Greek dialects. As I have argued earlier (Chapter 5, Section 5.6), however, his conception of the Latin term dialectus was highly confused, and it is difficult to grasp how he related it to lingua. During the sixteenth and seventeenth centuries, substantial systematic accounts of the classes of variation typical of related dialects were principally restricted to handbooks for the Ancient Greek language. From the end of the sixteenth century onwards, however, scholars started to insert into their Greek manuals more general accounts, which appear to have had a wider application and were no longer dependent on John the Grammarian’s work. The German Hellenist Jakob Gretser (1562–1625), for instance, who wrote the standard Greek grammar for Jesuit colleges, defined dialectus as follows: A dialect is nothing else than the characteristic and difference of one and the same language, not employed by all using that language but only by some. Hence, dialect is usually called “idiom” or “property”, whether this difference consists in letters of the same word and affections of letters only, or in pronunciation only, or in entire words, or in syntax and construction. The entire variation of dialects indeed seems to be contained in these four parts.¹

Gretser distinguished in generic terms four different classes of variation among related dialects: letters and their affections, pronunciation, entire words, and syntax. John the Grammarian’s tripartite framework was, in other words, gradually supplemented and eventually superseded in Renaissance Greek handbooks. From about 1650 onwards, the classes of variation were extensively discussed outside manuals on the Greek language and its dialects, too, which may be taken as yet another symptom of the gradual de-Hellenization of the dialect concept. This development made a start in works focusing on language origin in general ¹ Gretser (1593: 19): ‘D nihil est aliud, qua[m] vnius & eiusdem linguæ nota & differentia, non omnibus qui illa vtuntur, sed aliquibus tantum vsitata; quorum ratione dialectus idioma seu proprietas dici solet; siue differentia hæc consistat in solis literis eiusdem vocabuli, literarumq́[ue] affectionibus, siue in sola pronuntiatione, siue in verbis integris, siue in syntaxi & constructione. His enim quatuor partibus contineri videtur tota dialectorum varietas’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     

195

and the so-called Oriental tongues in particular. As an example of the former may serve the definition of Georg Stiernhielm, who opposed dialect to language in terms of accidental and substantial variation in his preface on the origin of languages. Stiernhielm linked his definition to linguistic features, too, as my analysis in Chapter 12 suggests. In particular, Stiernhielm (1671: c.3v, d.4v) associated language-level variation with differences in the lexicon and the roots of words, whereas he believed related dialects to vary in terms of letters, morphology, phraseology, lexical meaning, accent, and pronunciation. Stiernhielm’s acquaintance Christian Ravis had expressed similar views in an earlier grammatical publication on different Semitic tongues (see Chapter 12, Section 12.2). According to Ravis (1650: 8, 44–5), related dialects varied mainly in terms of accent, pronunciation, and accidents, whereas languages showed substantial differences on the level of roots, the lexicon, and the meaning of words, which could be designated with a modern phrase as lexical semantics. Like Stiernhielm, Ravis linked this distinction to the Aristotelian interpretation of the conceptual pair, which was no coincidence. As a matter of fact, the attention to classes of linguistic variation intensified when scholars started to formulate the Aristotelian interpretation in explicit terms (see Chapter 8, Section 8.1). The focus on linguistic criteria may seem paradoxical in the light of the increasing prominence of language-external interpretations of the conceptual pair, most importantly in terms of status (see Chapter 9). It is, however, not inconceivable that this linguistic turn in views on the conceptual pair was an intuitive reaction against such subjective views on the language/dialect distinction. Another plausible explanation is that scholars were developing an ever-increasing interest in the specificities of language change and linguistic kinship. Philologists decided, often on the basis of a superficial and intuitive analysis of select linguistic forms and features, whether two speech varieties were related dialects or distinct languages, a decision process with which language-external factors—social, political, and ideological—frequently interfered. The precise mechanisms of such language/dialect decisions deserve a closer analysis, which, however, lies outside of the scope of this book.

15.2 From scattered comments to systematization In the sixteenth and early seventeenth centuries, ideas about the ways in which related dialects varied were primarily expressed in Greek handbooks. However, they can also be gathered from scattered comments in other contexts. Let me treat the three most important types of evidence. Firstly, some humanists discussed how the dialects of a language other than Greek were different. The French printer-scholar Henri Estienne, for instance, elaborated at length on French variation in the outline of his book on the excellence of the French language.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

196

  

Estienne (1579: 138–46) assumed that French dialects, among which he included for instance Savoyard, could differ from each other in terms of pronunciation, lexical meaning, lexicon, and proverbial expressions. Secondly, passing comments reveal a scholar’s assumptions on the classes of variation associated either with language or with dialect. For example, the English Jesuit Robert Persons (1546–1610) commented upon the events at the Tower of Babel, following the authority of a number of pagan testimonies, that if there had not bene some such miracle in the diuision of tongues; no doubt, but that all tongues, being deriued of one, (as all me[n] are of one father,) the same tongues vvould haue retained the self same rootes and principles, as in all dialectes or deriuations of tongues we see that it commeth to passe. (Persons 1585: 102)

This passage suggests that Persons believed languages to differ from one another in terms of lexical roots and ‘principles’, which he assumed to be identical or highly similar for related dialects. It is hard to catch the exact meaning of the term principles here. Persons may have been thinking of primary forms, which would make his phrase ‘rootes and principles’ tautological, or perhaps of grammatical or syntactic rules. Thirdly, some early seventeenth-century works contain utterly concise remarks on languages and linguistic diversity, from which ideas on the classes of variation can be inferred. Joseph Justus Scaliger, for instance, fleshed out his conceptual distinction between language mother and offshoot dialect as follows: ‘the same words effect that it appears as one language, but the transposition, change, and modification of the same words create different offshoots’.² Scaliger thus suggested that languages differed on the level of the lexicon, whereas related dialects exhibited certain variations in individual words. He was no doubt thinking of the age-old framework of letter permutations. What exactly happened after around 1650 and especially in the eighteenth century? Basically, language scholars systematized their views on the classes of variation. The cases of Stiernhielm, Ravis, Schultens, Gatterer, and Adelung, treated in the previous chapters, strongly suggest this. Numerous additional examples could be cited from different parts of Europe and in different contexts, usually in connection with the Greek, vernacular, or Oriental tongues. In such discussions, the ways in which related dialects differed from one another usually took centre stage, and the classes of variation among distinct languages were treated only in the margin, if at all.

² Scaliger (1610a: 119): ‘eadem verba faciunt vnam linquam [sic pro linguam] videri: sed eorundem verborum traiectio, immutatio, inflexio, aliam, atque aliam propaginem facit’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     

197

Reading the many scattered remarks of early modern scholars together with the more systematic eighteenth-century testimonies, one notices that there was no or hardly any level on which related dialects were not believed to exhibit differences. Instead of offering a detailed overview of the classes of variation associated with related dialects and, less often, distinct languages, which runs the risk of being confusing and uninformative, I limit myself here to some general observations. A question that naturally poses itself is the following: why did early modern scholars focus on classes of dialectal variation and not of language-level differences? Since they were unaware of this misbalance, one cannot but conjecture about the reasons behind it. It might, for instance, be connected to the basic assumption that linguistic variation was more or less unrestrained, which would make it rather pointless to posit specific classes of variation on the level of languages. Another pressing question concerns the vast number of classes of dialectal variation scholars presupposed. How could two dialects be assumed to differ in so many different ways and still be closely related? The answer must be sought in the Aristotelian interpretation of the conceptual pair. Even though related dialects differed from one another only on the surface, this did not mean that they did not exhibit variation on many different levels. Despite the great number of classes of variation suggested, which ranged from pronunciation, accent, and letters through morphology, the lexicon, and semantics to syntax and even idiom and style, dialectal differences were nonetheless prototypically associated with only a restricted number of linguistic features, usually pronunciation and so-called letter permutations. The superficiality of this type of variation was often emphasized, but some scholars pointed out that there should be more than a few letter differences in order to speak of a distinct dialect. The Dictionnaire de Trévoux, for instance, an early eighteenth-century French dictionary which enjoyed tremendous success, asserted that the presence of one type of vowel contraction alone was not a reason to regard a speech form as a separate dialect (see Dictionnaire de Trévoux 1721: .760). The strong emphasis on letter variation was no doubt partly inherited from Byzantine treatises on the Greek dialects, in which word modification processes constituted the primary descriptive mechanism. This Greek framework is now usually referred to as ‘pathology’ and was conflated with a similar Roman framework known as the permutation of letters, as pointed out earlier (Chapter 2, Section 2.3; Chapter 4, Section 4.4). It should be noted that the framework of letter mutations was not exclusively used to describe differences on the level of dialects. It was, in fact, of key importance for early modern approaches to etymology as well as linguistic variation and change as a whole, both on the level of language and dialect (see e.g. Cram 1999). Variation which modern linguists would label ‘morphological’ was usually also accounted for by means of such letter-change processes. As a result, early modern scholars tended to associate differences on the level of morphology with superficial linguistic variation,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

198

  

which accounts for the fact that before 1800 only a few intellectuals took differences in morphology to point to different languages. The English orientalist Thomas Hyde (1636–1703), for instance, was exceptional in taking substantial variation in the ‘grammatical part’ (pars grammaticalis) of a language, which for him largely coincided with morphology, as a criterion to distinguish between language and dialect. Hyde advanced this view in his Latin oration on the antiquity of the Arabic tongue, made in 1691, when arguing in favour of a close relationship between this language and Hebrew: Out of said dispersion the Arabic language emerged, but its difference from Hebrew was only dialectal, as is manifest also today both from the grammatical part, in which the main criterion of languages is put, and from the very many words which correspond to Hebrew words, as well as from the characters which are derived from the Hebrew alphabet.³

The emphasis on letters rather than sounds can be explained by the close link many scholars observed between a language and the writing system used to codify it, an inheritance from antiquity. Before the modern period, the term littera was understood in a double manner, designating both the letter itself and the sound it represented. That is why scholars like Thomas Hyde involved writing systems in determining the status of a variety. Scholars were, however, increasingly aware that there was a mismatch between the spoken and the written word, an awareness heightened by their engagement with dialectal variation. The Hellenist Erasmus Schmidt, for instance, reminded his German audience in the early seventeenth century that we are now not talking about those dialects that are expressed by means of different letters, but about those that are written with the same letters, but have a different pronunciation. For instance, some Germans say ‘father’ as Vãter, with a slightly longer a; others Văter with a slightly shorter a; still others Voter with an o. And yet all write it with the same letter a, even though at the same time the pronunciation is realized differently.⁴

Schmidt made this point in a discourse on the pronunciation of Ancient Greek, held in favour of the traditional Byzantine pronunciation and against the new ³ See Hyde (1767 [1691]: 454–5): ‘Ex dictâ dispersione orta est Lingua Arabica, cujus quidem differentia ab Hebraicâ erat tantum dialectica, uti & hodiè constat tam ex Grammaticali parte in quâ præcipuum Linguarum Criterion ponitur, quàm ex plurimis vocibus quæ cum Hebraicis consonant; & à Characteribus ab Hebraïco Alphabeto deflexis’. ⁴ Schmidt (1615: 239): ‘Non est nunc sermo de ijs dialectis, qvæ diversis literis exprimuntur; sed de illis qvæ iisdem literis scriptæ, diversimodam pronunciationem habent. v. g. Patrem aliqvi Germani dicunt Vãter, a longiuscula: aliqvi Văter, a breviuscula: aliqvi Voter per o: & tamen omnes scribunt eandem literam a, qvum interim pronunciatio fiat diversa’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     

199

ideas of Erasmus and others. It seems therefore that the debate on this topic raised his awareness of the fact that letters could be pronounced differently and that there could be an uneven relationship between speaking and writing, especially in dialects. Apart from pronunciation and letter mutations, the lexicon, too, was often perceived as a major level on which related dialects differed from one another. The lexicon as a level of dialectal variation was not only obvious from daily experience but also had clear predecessors in Ancient and Byzantine Greek scholarship. Some Greek philologists had already compiled collections of dialect words, either as a tool for those wanting to write pure Attic or out of an antiquarian and philological interest. As I have mentioned earlier in this chapter, John the Grammarian had moreover included lexical variation among the three planes on which the Ancient Greek literary dialects differed. Since I have tried to offer a brief synthetic account of early modern ideas on the classes of variation, I might have given the impression that these were largely homogeneous. There was sometimes disagreement on this matter, however. Christian Ravis, for instance, countered the widespread opinion that related dialects differed from each other mainly in terms of pronunciation, since he seems to have believed that variation on this level did not suffice to speak of different dialects. That is at least what a passing statement of his suggests, featuring in a discourse he held on the topic of the ‘Easterne Tongues’, in which he discussed the easiness and wide spread of Hebrew: ‘I confesse there are many different pronunciations, but never a Dialect, or Idiome in Africa’ (Ravis 1650: 71). Ravis thus implicitly proposed a taxonomy with three layers: language – dialect – pronunciation. Direct debate among scholars about the classes of variation was, however, almost entirely lacking. Yet scholars did react against general practices. The Leipzig theologian and orientalist Bartholomaeus Mayer (1598–1631), for instance, made the following objection to a group of unnamed scholars who believed that the Hebrews, Aramaeans, and Syrians had one language: On the contrary, I judge that the Chaldean and Syriac languages are not merely different dialects of one and the same language, but that in the Babylonian “speech confusion”, or certainly a little later, a peculiar language emerged, I mean Aramaic, or Syro-Chaldean. This language, even though it approximates more closely the nature of the mother than the other languages, is nevertheless separated from it by such a distance that since it differs from Hebrew in terms of letters, mode of inflection as well as very many essential primary forms, it should be deservedly regarded as another and different language.⁵ ⁵ Mayer (1629: 16–17): ‘E diverso statuo, linguam Chaldæam ac Syram non esse unius & ejusdem linguæ diversas tantum dialectos, sed in γλωσσοσυγχύσει Babylonica, vel certè paulò post,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

200

  

From this observation it is clear that Mayer employed linguistic distance as his main criterion to distinguish a dialect from a language in a discussion triggered by predominant views on the interrelationships of certain Oriental tongues. Mayer’s account is remarkable for two reasons. On the one hand, letter variation was usually closely associated with dialect-level differences and not with distinct languages, as in Mayer’s account. Morphological properties such as flexion were, on the other hand, usually not distinguished from letter variation, but they were by Mayer.

15.3 Two eighteenth-century outsiders in the quest for linguistic criteria Slight differences in terms of pronunciation, letters, and also the lexicon were typically associated, often in very general terms, with dialect-level variation. Towards the end of the early modern period, these categories frequently received a more concrete interpretation, as the cases of Schultens and especially Gatterer suggest. Two eighteenth-century scholars had the language/dialect distinction even depend on one single linguistic feature. In the first and earliest case, the criterion chosen is the most baffling, due to its high specificity as well as its arbitrariness. In 1727, the German language scholar Johann Georg Wachter (1663–1757) saw his German etymological dictionary through the press in Leipzig. To his glossary Wachter added a preface on the original Germanic language, in which he discussed among other things the origin of language and linguistic diversity in general terms. Relying on the Bible, he assumed like most early modern scholars that there had been one primeval language, here unidentified, from which the others descended. He believed these new languages to originally have been dialects of the original tongue which gradually developed into distinct languages: And so out of one single and primitive language (to which the sacred writings and the agreement of nations testify) first various dialects were produced, and almost as many variations as there were families of men or blood-relationships among families. When these were subsequently separated, the dialects were gradually transformed into languages. Later, out of these languages were formed new languages, and dialects of languages, and languages of dialects, not suddenly

peculiarem extitisse linguam, Aramæam puto, sive SyroChaldæam, quæ licet propius accedat ad matris indolem, quam reliquæ, ab ea tamen tanto intervallo distat, ut quia & literis, & inflexionis ratione, & essentialibus insuper thematibus quamplurimis ab Ebræa discrepat, meritò pro lingua alia ac diversa haberi debeat’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.  -  . . .

201

and instantly, but gradually and step by step, with the help of time, space, migrations, and colonies.⁶

In a footnote accompanying this passage, Wachter drily remarked: ‘I distinguish languages from dialects in such a manner that difference of languages arises from consonants, whereas difference of dialects emerges from vowels’.⁷ Even though Wachter formulated his highly specific criterion seemingly independently, his idea was foreshadowed in the thought of the Flemish orientalist Franciscus Raphelengius (1539–97). In a letter to the esteemed philologist Justus Lipsius (1547–1606), dated 18 May 1584, Raphelengius commented on the close relationship he saw between Persian and Dutch, observing: ‘I also find [in Persian] dare for “door” (for deure [Dutch for “door”]), for I do not have a high regard of vowels, since these can only form different dialects’.⁸ This highly specific criterion might have been grounded in the idea that consonants made up the framework of a word, into which vowels were implemented. Even though this consonant/vowel criterion could strike certain modern readers as absurd, it is less so than it at first sight appears. In fact, modern research has confirmed that consonant substitutions often have a greater impact on the degree of intelligibility among speakers of different language varieties than vowel changes (Gooskens in Boberg et al. 2018: 210). Raphelengius and Wachter did not, however, base their criterion on thorough linguistic research but assumed its validity intuitively, without any concern whatsoever about its workability. The second case deserves more extensive attention, since it frames within a debate over the status of a low variety vis-à-vis its roofing literary language. It involves the Enlightenment thinker Ferdinando Galiani (1728–87), born in Chieti in the then kingdom of Naples and today mainly known for his work as an economist and humorist. Yet Galiani also authored a book on the Neapolitan dialect (Del dialetto napoletano), first published in 1779. The booklet was highly popular and incited ‘vivacious reactions’ (Brevini 1999: 1905–6), for instance by his fellow citizen Luigi Sèrio (1744–99). Sèrio reacted against Galiani’s ennobling of their native dialect and propagated a more plebeian variety of Neapolitan in a work published a year later (Sèrio 1780). Copies of Galiani’s book sold out quickly, and he made plans to reedit it with corrections and additions as early as 1780. ⁶ Wachter (1727: a.4–5): ‘Itaque ex una & primitiva Lingua (quod sacræ Literæ & Gentium consensus testantur) suscitatæ sunt primo variæ , & totidem pene variationes, quot hominum familiæ, aut familiarum cognationes(i), quibus deinde separatis, Dialecti paulatim abierunt in  (ii). Ex his Linguis postea formatæ sunt novæ Linguæ & Linguarum Dialecti, & Dialectorum Linguæ, non subito & repente, sed gradatim & pedetentim, temporum, locorum, migrationum, & coloniarum adjumento’. ⁷ Wachter (1727: a.4–5): ‘(ii) Linguas a Dialectis sic distinguo, ut differentia Ling[uarum] sit a consonantibus, Dialectorum a vocalibus’. On Wachter’s etymologicon see Jones (2001: 1106–7) and Considine (2014: 135). ⁸ Raphelengius (1983: 123): ‘Invenio etiam “Dare” pro ostio (pro “deure”) nam non magni facio vocales, cum harum sit tantum formare varios [sic] dialectos’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

202

  

By his death, this work had been finished, but the augmented second edition appeared only posthumously in 1789 (see Cortelazzo 1980: 112–13; Sansone 1999: 201–34; Pennisi 2009). In his discussion of the Neapolitan dialect, Galiani proposed syntax as the main touchstone for determining whether two varieties were ‘different languages’ (lingue diverse) or ‘dialects’ (dialetti). If the syntax was different, it concerned distinct languages. If the syntax was the same or at least very similar, it concerned different dialects of one language, despite any variation occurring on the level of pronunciation, orthography, morphology, and the lexicon (see Galiani 1779: esp. 33–4, cf. also 15, 27, 39). It is, however, not entirely clear what Galiani meant by ‘syntax’, but his outline indicates that he was principally thinking of phraseology and stylistics rather than word order or verb constructions. His focus on syntax led him to regard Neapolitan as a dialect of common Italian, which was based on the archaic literary Tuscan of Petrarch, Boccaccio, and Dante. In doing so, Galiani wanted ‘to lessen the conflict between Tuscan, the prestige vernacular, and local dialects such as those spoken by the educated classes of Naples and southern Italy, by downplaying the phonetic and morphological differences and emphasizing the common syntactic base’. In this context, he ‘proposed a moderate standardization of southern dialects (starting with spelling) and a historically founded description of Neapolitan which gave importance not so much to its literary or folkloric uses as to its administrative uses’ (Pennisi 2009: 508). In other words, Galiani saw Neapolitan as a dialect of the Italian language because he did not want his native speech to supersede the Tuscan-based norm, at least not in all genres of writing. Even though Galiani explicitly took syntax as his main linguistic criterion to distinguish between a language and a dialect, he nevertheless appears to have conceived of dialect as erratic, too, a view inspired partly by his reading of Dante’s oeuvre and partly by his preconceptions about the inferior position of Neapolitan vis-à-vis the Tuscan norm (Galiani 1779: 35, 55). Galiani’s conception of the language/dialect pair was clearly not only grounded in the linguistic feature of syntax but indirectly also in normative ideas about Italian and other Italo-Romance varieties.

15.4 Conclusion By 1650, scholars had teased out the most important interpretations of the conceptual pair. The focus then shifted to rationalizing and systematizing existing ideas. In fact, as I have tried to argue in this chapter, an objective turn occurred after the mid-seventeenth century, which climaxed in the eighteenth century. This objective turn manifested itself as a greater emphasis on linguistic criteria and on the ways in which related dialects differed from one another as opposed to distinct languages. In general, scholars seem to have believed that related dialects varied

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

203

primarily in terms of pronunciation, letters, and the lexicon, even though a wide range of other linguistic levels recurred in discussions of the nature of dialectal diversity. In this context, the superficiality of the variation was frequently emphasized. Language-level variation, however, was closely associated with radical differences in a twofold sense. On the one hand, scholars assumed that distinct languages could vary on all possible levels, and these differences were usually said to be of a substantial or fundamental nature. On the other hand, language-level variation was claimed to manifest itself mainly in the roots of words. The objective turn was partly motivated by rationalist Enlightenment trends, for which the ideas of Johan Christoph Gatterer and Ferdinando Galiani can be cited as prime examples. Paradoxically, this rational attitude did not necessarily mean that the conceptual pair was recognized as potentially problematic; instead, most scholars still perceived it as a self-evident common-sense distinction. Early modern discussions of the classes of variation confirm the trend towards de-Hellenization. Before about 1650, the nature of dialect-level variation was mostly treated in manuals for the Ancient Greek dialects, which were for a large part indebted to the three-class framework of the early Byzantine author John the Grammarian. Afterwards, non-Hellenists took over the wheel, a change involving a different attitude towards linguistic features in relation to the conceptual pair. Numerous scholars started to look more, but far from exclusively, at languageinternal specifics rather than at language-external factors when distinguishing a language from a dialect. Their proposals are sometimes surprising, as for instance Galiani’s insistence on syntax, and occasionally even utterly particularistic, like Wachter’s consonant/vowel criterion.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

16 Between systematization and rationalization The conceptual pair through the Enlightenment lens

The period 1650–1800 was an age of reason, which took off in the wake of René Descartes’s (1596–1650) work and culminated in the Enlightenment movement during the eighteenth century. Rationalist trends went hand in hand with a longing for the systematization of knowledge, evidenced by the popularity of the genre of the encyclopaedia. It is indeed no coincidence that Diderot and d’Alembert’s Encyclopédie is usually considered to be the ‘epitome of Enlightenment literature’ (Mokyr 2002: 68). As I have tried to outline in the previous chapters, this rational turn was reflected in ideas about the language/ dialect pair, too. Scholars such as Albert Schultens and Johann Christoph Gatterer did their best to provide rational and systematic treatments of the matter in their philological and historiographical programmes. They tried to offer objective criteria, grounded in linguistic features. The symptoms of rationalization and systematization go, however, much further. In what other ways do they manifest themselves in seventeenth- and eighteenth-century understandings of the conceptual pair?

16.1 Towards a dialectological tradition? On 22 July 1702, the Saxon philologist Christian Gottlieb Schwartz (1675–1751) and the even more obscure German scholar Abraham Helm presented to their audience at Wittenberg university a Disputation on the causes of dialects, specifically the Greek (see Van Rooy 2020c for a more detailed analysis). The text of their dissertation, most likely authored primarily by Schwartz, the supervisor, was printed in the same year. It offered the early eighteenth-century reader a wellstructured analysis of dialectal diversity as a general phenomenon. First of all, the term dialectus and its numerous existing definitions were discussed. They were, however, all rejected: We leave the opinions of others behind us and add the following definition of our own: a dialect is a peculiar change of a certain common language, especially in

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

205

terms of individual words, and that change is rather slight and different according to the times and peoples using that same language.¹

Schwartz and Helm thus combined several interpretations of the conceptual pair. Firstly, related dialects showed ‘slight’ differences as opposed to distinct languages. Secondly, it was implied that dialects deviated from a common language. Thirdly, dialect was associated with the entity of ‘people’, suggesting that language transcended this level and encompassed different ‘peoples’. Fourthly, the reference to time as a factor in linguistic diversification likely indicates that Schwartz and Helm also took dialect in the language-historical sense of ‘descendant of a language’, an interpretation which was to feature prominently in later chapters of the dissertation. In what followed, Schwartz and Helm (1702: .2–.2v) offered a kind of synthesis based on earlier literature, elaborating extensively on the different elements included in their idiosyncratic definition. The relationship of a dialect to a common language, under which it was subsumed, was first treated. Then, the authors advanced the idea that dialectal variation mainly manifested itself as ‘individual words that are altered in one way or another’, no doubt referring to the traditional letter mutation framework, to which they added that variation on the syntactic level occasionally also occurred.² In this context, Schwartz and Helm (1702: .3–) took great pains to distinguish between the concepts of style and dialect, frequently conflated by scholars because of the literary usage of the Greek dialects (see also Chapter 5, Section 5.6). Next, they discussed the implications of their claim that related dialects exhibited slight differences. Schwartz and Helm continued by treating the different degrees of variation dialects could display, which led them to posit a hierarchy of types of dialects; there were generic dialects under which specific or local dialects resorted. Schwartz and Helm (1702: .3–.1) then provided an overview of the causes behind linguistic change in general and dialectal diversification in particular, stressing its gradualness. In the few remaining pages of the dissertation, the authors moved from the general to the specific and discussed the origin and peculiarities of the Greek dialects, as announced by the title. The disputation held by Schwartz and Helm may thus be considered the earliest work entirely devoted to the problem of dialectal diversification per se, even though they still had the reflex to focus strongly on the Greek context in the final paragraphs. Yet they also alluded, albeit sparsely, to dialect-like variation in German, Romance, and Hebrew. Their text was clearly an attempt at providing

¹ Schwartz and Helm (1702: .2): ‘Nos, missis aliorum sententiis, hanc nostram subiicimus definitionem: Dialectus est Linguae alicuius communis, inprimis secundum singulas uoces, immutatio peculiaris, eaque leuior, et, pro diuersis temporibus, ac populis illa eadem lingua utentibus, diuersa’. ² Schwartz and Helm (1702: .1): ‘uocibus singulis aliter atque aliter inflexis’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

206

   

the reader with a comprehensive and systematic overview of the issue of dialectal diversity. To this end, it treated key aspects related to the phenomenon, including the criteria that should be used to determine the dialect status of a variety, the classes of dialectal variation, the causes of linguistic diversification, and the prototypical Ancient Greek context, knowledge of which was considered an important philological skill in the early modern period, indispensable for anyone desiring direct access to the great Greek authors of the past. It is no coincidence that Schwartz and Helm’s disputation was presented at the Wittenberg academy. This institution had a scholarly tradition of books and dissertations on issues related to dialectal diversity. In the seventeenth century, several works on the Greek dialects were published there (e.g. Schmidt 1604), as were disputations on Oriental tongues (e.g. Pfeiffer and Martini 1663), often conceived in terms of dialect. This tradition continued well into the eighteenth century. Two other dissertations were presented in Wittenberg in the year 1709, one on the Greek dialects (Thryllitsch 1709), and the other on the Greek Koine (Kirchmaier and Thryllitsch 1709). These works, too, were systematizing treatments, even though they were more narrowly focused on the Ancient Greek context. Both disputations had the young Hellenist Georg Friedrich Thryllitsch (1688–1715) as their presenter, and of at least one Thryllitsch was the sole author. Towards the middle of the century, Wittenberg witnessed the publication of yet another dissertation on the topic of dialectal variation, this time applied to the context of the Oriental tongues and depending to a large extent on Albert Schultens’s work (Groddeck and Treuge 1747; see Chapter 13, Section 13.4). In 1782, this time in Berlin, the influential school reformer and late Enlightenment scholar Friedrich Gedike (1754–1803) published an article entitled On dialects, especially the Greek, which appeared in the first volume of the Berlin magazine of sciences and arts. Much like Schwartz and Helm in 1702, Gedike (1782) offered his readers a synthetic discussion of the phenomenon of dialects with specific attention to Greek variation, while also referring to German, Latin, and English diversity. However, very unlike Schwartz and Helm, he barely cited any early modern sources and largely developed his ideas on his own. The topic of dialects kept fascinating Gedike in later years, too. In 1794, he held a lecture on German dialects, which he introduced in a manner similar to his 1782 contribution and in which he referred also to the Greek dialects (Gedike 1794). In fact, he saw a close parallel between Greek and German dialects, comparing Ionic to Low German, both smooth and delicate dialects in his perception. Doric was put next to Upper German, both of which he claimed to be solemn and splendid. Between these two dialects were the esteemed Attic and High German dialects, which approached more closely the solemn Doric and Upper German dialects than Ionic and Low German, respectively. Gedike repeatedly revisited this idea since his first formulation of it in a 1779 work on linguistic purity (see Gedike 1779: –). This way, Gedike could tie German dialects to their highly valued Greek counterparts. In fact, the main

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

207

goal Gedike tried to achieve with his 1794 lecture on the German dialects was to promote the study of Low German (Plattdeutsch), not because it should replace High German as the norm, but because it could be a source of richness and even correction of High German. He even proposed to organize competitions for the collection of dialect words (Gedike 1794: 332). Most systematizing works were published in German-speaking areas. This geographic concentration is not surprising, since Greek studies, which provided the original incentive to study dialectal variation, never ceased to flourish there, unlike in many other regions. These areas moreover witnessed an increasing philological interest in the relationships among Semitic tongues, which most scholars agreed were so closely related that they might well be dubbed ‘dialects’. Last but not least, Protestantism caused a major concern over spreading the Word of God in a unified and widely intelligible form of the German vernacular, which seems to have made Protestant scholars more conscious of the vast variation in their native speech. Yet systematizing treatises were not only produced in Germany. There are examples of encompassing works from other regions as well. For instance, the Swedish teacher and language scholar Sven Hof published in 1772 a pioneering work on the dialect of Västergötland, an area in modern-day Sweden (see Chapter 8, Section 8.5). His book consisted of a thorough grammatical introduction and a long dictionary in which he compared Västergötland dialect words to their Swedish equivalents. The intellectual circumstances in which Hof produced his contribution resembled those of the German works mentioned earlier. Hof, too, was well-versed in the Greek language and its dialects; in fact, scholarship on this tongue was a welcome model for Hof ’s description of the Västergötland dialect (see Hof 1772: 12–13, 23). Additionally, Hof contributed to establishing a norm for the Swedish language, as he wanted to arrive at a better Swedish translation of the Bible and was concerned over Swedish orthography (see Eriksson 1971; Droixhe 1978: 123–4; Hovdhaugen et al. 2000: 93–4). Like Gedike after him, Hof considered dialects to be a source of good words that could or even should replace foreign words. He moreover emphasized that knowledge of the dialects was necessary in order to fully understand one’s native language, its genius, and its orthography. This important idea indicates that Hof did not believe dialects to be intrinsically corrupt. Indeed, they provided evidence for the nature, essence, and history of the language to which they belonged. He nevertheless suggested that a dialect should not be used in writing, claiming that the Swedish variety he described actually should be unlearned rather than learned. What is more, he wanted to teach his compatriots correct Swedish and claimed by way of captatio benevolentiae that there was no glory in writing on vernacular dialects (Hof 1772: 8–10, 28). Hof did, however, observe that dialects differed from each other in terms of esteem: ‘all nations have some dialects they approve most, but others they despise and condemn as if they were absurd’. Interestingly, such

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

208

   

judgements were formed by the ‘imagination and usage of the ears’ rather than by a rational decision.³ It should be clear that the phenomenon of dialectal diversity was gradually becoming a topic in its own right after 1650, culminating in the eighteenth century, when scholars such as Gedike and Hof insisted on the richness and utility of regional dialects (cf. Droixhe 1978: 334–45). It did, however, remain a kind of reflex to concentrate on the prototypical Greek dialects after a more general discussion. Another symptom of the tendency towards systematization and rationalization consists in the fact that it became somewhat more customary to explicitly define language as opposed to dialect from the second half of the seventeenth century onwards. This increased attention to the language pole did not occur only in Georg Stiernhielm’s work and that of some of his readers but also in several writings by followers of Albert Schultens (see Chapters 12–13). It would nevertheless be premature to speak of a full-fledged dialectological tradition in the early modern period, since the research field did not enjoy any disciplinary autonomy, institutionalization, or other scientific organization, for instance through specialized journals, as there was in the later nineteenth century. The degree of dialogue between scholars was moreover limited. Be that as it may, the term dialectology, more specifically in its Latin guise dialectologia, ‘study of dialects’, was coined in this period as a composition of the Greek words diálektos and lógos (λόγος), mirroring compositions like geologia, ‘study of the earth’. In fact, dialectologia featured quite prominently on the title page of a work on the Greek dialects and the way they allegedly figured in the Greek New Testament. The work, entitled Sacred dialectology (Dialectologia sacra; Figure 16.1), was published in Zürich in 1650 and compiled by the local professor of Greek Caspar Wyss (1604/5–59). The book was conceived as a manual to help readers understand the peculiarities of the original language of the New Testament. However absurd and clumsy this might seem to modern readers, the idea that features of all literary Greek dialects could be found in the New Testament turned out to be a persistent one, cultivated by many philologists in the early modern period and first systematized by the Calvinist professor Georg Pasor in the first half of the seventeenth century (see Van Rooy 2020a: Chapter 4 for a more detailed discussion). Pasor’s Swiss colleague and follower Wyss coined the term dialectologia to strictly refer to this subfield of Greek dialect studies. It was, however, soon applied to the study of the Greek dialects in general, apparently first in a 1684 handbook for Greek prosody by the poorly known Stuttgart teacher Johannes Bregius. The

³ Hof (1772: 14): ‘Etenim omnes gentes nonnullas habent dialectos, qvas maxime probant, alias autem, qvas tamqvam insulsas fastidiunt & damnant; qva in re non tam rationis judicium, qvam imaginationem & consvetudinem aurium seqvuntur’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

Figure 16.1 Caspar Wyss’s Dialectologia sacra, 1650 Source: e-rara. Public domain

209

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

210

   

booklet was published in 1684 in Tübingen and had the title Greek prosody, with dialectology, or Prosodia Græca, cum Dialectologiâ in the original Latin. This designation caught on, and several German eighteenth-century Hellenists took it over in this broader sense (see e.g. the title of Nibbe 1725). From here, it was only a small step to apply it to vernacular dialects as well. Indeed, the Swedish scholar Erik Benzelius the Younger (1675–1743) planned to produce a work entitled Swedish dialectology (Dialectologia Suecica), containing word lists of Swedish dialects, which never materialized, however (Considine 2017: 166, 173, 175). In 1755, the term reappeared in the title of a brief description of the Hamburg dialect, Hamburg dialectology or, in Latin, Dialectologia Hamburgensis; this treatise was attached to the second edition of an extensive lexicon of the same dialect. Both the description and the lexicon were produced by the Hamburg scholar and poet Michael Richey (1678–1761), primarily out of a feeling of patriotism.⁴ An additional motivation of Richey’s was dialect preservation, since he believed dialectal features to be a source of linguistic richness (Haas 1994: ). For this reason, he deplored the decline he noticed in the usage of his native dialect. What is more, at his time of writing, even farmers were using High German words in an attempt at passing themselves off as distinguished men (Richey 1755: –, ). It comes as no surprise that Richey studied in Wittenberg around the turn of the eighteenth century, at which time the topic of dialectal diversity was a hot topic in Luther’s city. Even more unsurprisingly, Richey had himself specialized in the Greek language and literature, too; he became professor of Greek and history at the academy of his home city in 1717, holding both chairs until his demise (see Waldberg 1889 for biographical information). Richey (1755: ) made explicit, moreover, in the preface to the second edition of his Hamburg dictionary, that lexica and descriptions of the Greek dialects had served as a source of inspiration for his work. It is not unlikely that he adopted the term dialectologia from works in this tradition, even though he did not cite any specific source in his lexicon. Richey’s pioneering work was eagerly imitated across German-speaking Europe, especially his lexicon but occasionally also his description of the Hamburg dialect, his dialectologia. In fact, shortly after the appearance of the second edition of Richey’s work, Johann Christoph Strodtmann (1717–56) designed a brief Osnabrück dialectology (Dialectologia Osnabrugensis) and also Germanized the term (Strodtmann 1756: , 3–8, 240). It seems that from about this time the study of dialects had a name, dialectologia in Latin or Dialectologie in its German form. The term remained in use to refer to descriptions of dialects and their features, primarily in German-speaking areas, before becoming the designa-

⁴ Richey (1755). The first edition of the lexicon is Richey (1743). See Brekle et al. (1992–2005: .172–5); Considine (2017: 117–19).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

211

tion of a subfield of institutionalized linguistics in the late nineteenth century (cf. Budzinski 2012). The surprising result is that this linguistic subfield received a label before the rubric of linguistics was coined. This fact, too, suggests that a dialectological tradition was emerging after 1650, which flourished in the eighteenth century, particularly in the Holy Roman Empire. Despite the emancipation of the conceptual pair from the Greek heritage, the systematization nevertheless started in the context of Greek philology. One could, in fact, regard Schwartz and Helm’s 1702 disputation as having evolved out of early modern manuals for the Ancient Greek dialects. The difference was, however, that the general discussion at the outset was not only given a wider application, so as to concern the phenomenon of dialectal diversity as a whole, but also constituted the dissertation’s main theme. This structure was then extrapolated to philological studies of the so-called Oriental tongues. Systematization thus mainly occurred within a philological context, with a primary focus on Greek and Semitic diversity, even though book-length treatments of vernacular dialects began to appear on a large scale, too. In these works, however, general conceptual issues were usually not considered to be of direct concern, and authors mostly limited themselves to ad hoc issues involving the description of a particular dialect.

16.2 The first sceptical voices The rationalist attitudes adopted towards the conceptual pair in the eighteenth century are not only noticeable in the linguistic turn occurring in definitions of dialect versus language and in the emergence of an embryonic dialectological tradition. Rationalist reflection also led a number of scholars to field questions about certain usages of the language/dialect distinction and even about its tenability tout court. Scholars were increasingly aware of the fact that the conceptual pair actually covered different semantic oppositions, which was considered undesirable. For this reason, some authors proposed to supplement the binary language/dialect distinction with other concepts and to fit it into larger schemes. I have argued in Chapter 13 that the Dutch orientalist Albert Schultens operated with a tripartite constellation consisting of language, dialect, and degenerate offshoot, thus adding a third element to the conceptual pair. Schultens was certainly not unique in doing so. For example, some years before him, the in his time highly popular French classical scholar and pedagogue Charles Rollin (1661–1741), responsible for a revival of Greek studies in Paris, had posited a tripartite scheme similar to Schultens’s: These idioms or dialects of the Greek language are not like the different jargons dominating in the different provinces of our France, which are an uncouth and corrupted manner of speaking and which do not deserve to be called a language.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

212

   

Each dialect was a perfect language in its kind, which was in vogue with certain peoples and which had its own rules and its own particular beauties.⁵

Rollin made a clear distinction between the cultivated varieties of Greek, termed dialectes or idiomes, which enjoyed immense literary prestige, and the corrupted patois of French, termed jargons, which he associated with the different provinces of France. The French grammarian Nicolas Beauzée (1717–89) designed a similar conceptual scheme in his Encyclopédie article ‘Language’ (‘Langue’). In it, Beauzée (1765: 249) linked language to nation, which embraced several peoples (peuples), whereas he took dialect in the sense of ‘legitimate variety of a language, particular to a people’, as opposed to patois, which he viewed as a degenerate variety of a language spoken by the provincial populace of a unified state. Examples of dialects could be found in German, Italian, and Greek. For patois, one should look at the languages of the Roman Empire and the kingdom of France. The scheme of Schultens, on the one hand, and those of Rollin and Beauzée, on the other, are not entirely the same. In Schultens’s ideas, the historical dimension and the process of language corruption played an essential role, whereas languageexternal factors were of central significance for Rollin and Beauzée, who connected their concepts of dialect and patois principally to cultural and political circumstances. The very fact, however, that they felt the need to design such tripartite divisions seems to indicate that they were no longer satisfied with a simple binary conceptual opposition and that they wanted to arrive at a more nuanced account of the omnipresent phenomenon of linguistic diversity. Scepticism towards the conceptual pair and its polysemy manifested itself in other ways, too. Some scholars insisted on differentiating between several interpretations of the language/dialect distinction, since jumbling them together inevitably led to confusion. For instance, the Berlin-born orientalist Paul Ernst Jablonski (1693–1757) published in 1714 a treatise on the mysterious Lycaonian language, anciently spoken in Asia Minor and mentioned once in the New Testament (Acts 14.11; on Lycaonian see e.g. Bruce 1992). The identity of this language was much debated among early modern scholars. There were two main camps, one arguing that it was a dialect of Greek, the other that it was a distinct language. It was apparently first problematized in sixteenth-century England, where the discussion was part of a larger disagreement between Catholic and Protestant theologians, known as the Challenge controversy (Jenkins 2006: 115–54). In this debate, the practices of the Early Christian Church were the ⁵ Rollin (1726: 117): ‘Il n’en est pas de ces idiomes ou dialectes de la langue grecque, comme des différens jargons qui regnent en différentes provinces de notre France, qui sont une maniere de parler grossiére & corrompue, & qui ne méritent pas d’être appellés un langage. Chaque dialecte étoit un langage parfait dans son genre, qui avoit cours chez certains peuples, qui avoit ses régles & ses beautés particulieres’. On Rollin see the rich biography by Ferté (1902: esp. 16–17).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

213

main topic of discussion, including the language used in Mass, where the issue of the identity of Lycaonian came in. Later, the matter became an autonomous philological topic discussed by Paul Ernst Jablonski. In his Inquiry into the Lycaonian tongue, Jablonski discussed its identity, as well as that of other obscure languages of Asia Minor such as Phrygian. In this context, he reacted against the claim advanced by the Dutch historian Theodorus Rijcke (1640–90) that Phrygian was a dialect of Greek, accusing his predecessor of conceptual malpractice: I am surprised that the most erudite man here takes two things, even though they differ from each other, to be one. An offshoot of a language is one thing, a dialect another. As a matter of fact, a language can depart from another language, of which it is an offshoot, so far that it can no longer be held a dialect of it.⁶

Jablonski exemplified this last statement by referring to the case of the Latin language. The general consensus was that it was an ‘offshoot’ (propago) of Greek, but it could not be regarded as a dialect of it, since it had changed too considerably from it (Jablonski 1714: 16–17). Jablonski thus clearly harboured suspicions with regard to using the conceptual pair in a genealogical sense, proposing the distinction mother/offshoot as an alternative (cf. Chapter 10, Section 10.1). As I have argued earlier (Chapter 10, Section 10.2), a similar concern had already been voiced before Jablonski by another orientalist, the Swiss scholar Johann Heinrich Hottinger. Not all eighteenth-century scholars replaced the language/dialect distinction with a multi-layer model. What is more, the fact that the conceptual pair had different interpretations could still be a source of confusion and debate towards the end of the eighteenth century. For instance, in this period, two English scholars had an argument about the two concepts when discussing the language of the Greek poet Homer. The classical scholar, traveller, and politician Robert Wood (1717–71) stated the following in his Essay on the original genius and writings of Homer, while refuting the views of ‘professed scholars and critics in the Greek tongue’: They compliment [Homer] for having enriched his language with the different dialects of Greece; though the distinction of dialects can be only known to a cultivated, and, in some degree, settled state of language, as deviations from an

⁶ Jablonski (1714: 16), with reference to Rijcke (1684: 465): ‘In quo illud miror Virum erudissimum [sic] pro una re habere, quæ tamen inter se differunt. Aliud propago est linguæ, aliud Dialectus. Potest namque lingua ab alia, cujus est propago, tam procul abire, ut pro ejus dialecto amplius haberi non possit’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

214

   

acknowledged standardb. — They point out his Poetical Licences; forgetting that, in his time, there were no compositions in Prose.⁷

The historian Thomas Howes (1729–1814) responded to this observation of Wood’s in the first volume of Critical observations on books, antient and modern: Certainly Mr. Wood is mistaken in the first place, to assert that a distinction of dialects can be only known to a cultivated state of language; for as dialects are the derivatives and different modifications of one common original, so both these derivations and the common original itself may be all equally barbarous and uncultivated. ([Howes] 1776: 57)

To exemplify his genealogical conception of the distinction, Howes referred to the Gothic language, from which he believed English, French, Dutch, German, and Danish to have descended as dialects. He put this on a par with the Greek context; out of the Pelasgian-Ionic common original language, several Greek dialects developed. The synchronic perspective of Wood, in which standard language was opposed to dialect, clashed with Howes’s language-historical interpretation of related dialects as descending from one common ancestral language. Put differently, Wood stressed the synchronic deviation of dialects from a language, whereas Howes emphasized that the relationship between them is one of genealogical derivation. Their disagreement was, in other words, shaped by the different ways in which they interpreted the conceptual pair. By 1650 the main interpretations of the conceptual pair had been established. The borders between them were, however, not always sharply drawn, as the Wood/Howes case suggests. Some scholars tried to remedy this lack of precision by clearly distinguishing between different understandings of the two concepts, usually involving the language-historical interpretation, or by introducing a new concept into their schemes. One author, the evangelical pastor Friedrich Carl Fulda (1724–88), went even further and adopted an utterly sceptical attitude towards the language/dialect pair, bringing its polysemy to a head. After studies in theology, history, and statistics in Tübingen and Göttingen, Fulda developed a keen interest in the German language. A late bloomer, he started his research on this topic only around 1760. He wrote down his first results in an essay On the two principal dialects of the German language, which won him the award of the Royal society of sciences in Göttingen (Königliche Societät der Wissenschaften zu

⁷ Wood (1775: 238–9). Note b reads: ‘Nor would it be judicious to employ them indifferently. The Bergamasc, Neapolitan, and Venetian dialects, do well on the Italian stage in the mouths of Harlequino, Pulcinello, and Pantalone; but a Tuscan would never think of enriching his language by using them promiscuously in an Epic Poem’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

215

Göttingen) on 9 November 1771.⁸ The essay provided a historical discussion of the Germanic languages, their lexical roots, and the similarities in their grammatical systems (Vogt 1974: 49). It was published two years later in Leipzig and included at the back a table of Germanic roots. The publication marked the beginning of an industrious period of philological scholarship, during which he produced several lengthy works focusing on German linguistic history as well as on its diversity. This activity lasted until his death in 1788, in which year Fulda brought out his Attempt at a general collection of German dialect words. It is, however, his very first publication to which I should return, as it was in this work already that Fulda expressed his doubts about the value of the language/dialect distinction. In his essay, Fulda shrewdly observed: Dialect and language are difficult to ascertain. Different ways in which one and the same people speak a single language in different provinces in such a manner that they nevertheless understand each other—that is very vague . . . If one says that dialect is a slight distinction in a single word, language a complete distinction of the word itself, then doubts are raised by the different, often opposite meanings of the same word in a single or even different regions and the daily rising and falling in value of the same word as well as, on the contrary, the great resemblance among the words of apparently very diverse languages. They are like blood relationship, which alienates itself in both lines, without ceding to be common blood. Are only sisters dialects? Are mother and daughter, aunt and niece no longer? One should call each language by its own name, and kinship with a common name, and one should give up the word play.⁹

Although devoting only one paragraph to the matter, Fulda was one of the first, if not the first, to cast doubt on the utility of the conceptual pair in such explicit terms and to transparently point out its imprecise nature. He specifically mentioned the criterion of mutual intelligibility, which he dismissed as vague, and the Aristotelian interpretation, which were often bracketed together by earlier scholars, too (see Chapter 8). His suspicion was fuelled by the intrinsic variability ⁸ For biographical notes on Fulda see e.g. Franck (1878) and Püschel (2009). On Fulda’s essay see e.g. Vogt (1974: esp. 49–52, 77–8, 84–6) and Brekle et al. (1992–2005: .175–8). ⁹ Fulda (1773: 36): ‘Dialect und Sprache sind schwer zu bestimmen. Verschiedene Arten eines und eben desselben Volks, einerlei Sprache in verschiedenen Provinzen so zu reden, daß sie sich doch gegenseitig verstehen, ist sehr unbestimmt . . . Sagt man, der Dialect sei ein geringer Unterschied von einerlei Wort, die Sprache ein völliger Unterschied des Worts selbsten: so machen die verschiedenen, oft entgegengesezten Bedeutungen desselbigen Worts in einerlei, auch verschiedenen Gegenden, und der täglich steigende und fallende Werth desselben; wie im Gegenteil die grose Uebereinstimmung der Wörter in offenbar höchstverschiedenen Sprachen, Zweifel. Sie sind wie die Blutsfreundschaft, welche sich in beiden Linien entfernt, ohne daß sie aufhöre, ein gemeinsames Blut zu sein. Sind blos Geschwister Dialecte? Sind Mutter und Tochter, Muhme und Nichte keine mehr? Man nenne jede Sprache mit ihrem eigenen, und die Verwandtschaft mit einem gemeinschaftlichen Namen, und gebe das Wortspiel auf ’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

216

   

of human language in time, an intuition doubtless shaped by his languagehistorical interests. The essence of language and linguistic variation, Fulda believed, could not be captured in clear-cut and static concepts. In spite of his reservations, he continued to use terms such as Dialect (later German: Dialekt), no doubt for the sake of convention. In fact, a few pages further, Fulda (1773: 38) put forward in a kind of Aristotelian reflex that, strictly speaking, variation should be limited in order to speak of ‘dialects’, which differed from one another primarily in terms of pronunciation. Fulda did not cast doubt only on the conceptual pair but also on the widespread metaphor of family relationship, which he took to be closely connected with the language/dialect distinction. As an alternative, he proposed the use of specific names to denote individual speech forms and a common label for the kinship among different speech forms; he meant that it would be better to speak of Germanic, which covered, among other languages, German, English, and Dutch, rather than to associate these entities with abstract and metaphorical concepts such as language and dialect, mother and daughter. For Fulda, in conclusion, the distinction between these concepts was suspicious, since it was an all too simplistic abstraction of the complex phenomenon of the variation present in human language, which is always in flux. His scepticism did not, however, keep him from extensively using the terms Sprache and Dialect throughout his linguistic oeuvre.

16.3 Conclusion In the eighteenth century, general treatments of the phenomenon of dialectal diversity appeared on the scholarly scene; these were concentrated mainly in German-speaking Europe and seem to reveal a desire to systematize available knowledge on the topic, very much in the spirit of the Enlightenment. This systematization went hand in hand with the emergence of an embryonic dialectological tradition, best evidenced by the publication of numerous monographs devoted to the dialects of various languages, mainly Greek, vernacular, and Semitic. In addition, the design of more nuanced conceptual schemes and scepticism about the genealogical usage of the conceptual pair may be regarded as symptoms of a wider eighteenth-century rationalist tendency. This reached a culmination in Friedrich Carl Fulda’s work, where the inaccuracy of the conceptual distinction was explicitly problematized for the first time. However, the question of the adequacy of the language/dialect pair came to be addressed at much greater length only in the course of the nineteenth century.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

V

FROM SILENT ADOPTION TO OUTSPOKEN ABANDONMENT, A F T E R 18 0 0

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

17 From Jones to Gabelentz Silent adoption and renewed suspicion

I do not believe that the Greek, Latin, and other European languages are to be considered as derived from the Sanskrit in the state in which we find it in Indian books; I feel rather inclined to consider them altogether as subsequent variations of one original tongue, which, however, the Sanskrit has preserved more perfect than its kindred dialects. (Bopp 1820: 3) This is how the Bavarian linguist Franz Bopp (1791–1867) imagined the relationships between the ancient European languages and Sanskrit. They were dialects of a now lost original tongue, best preserved in Sanskrit. Bopp put forward this statement in 1820 at the publication of his thoroughly revised English edition of his famous 1816 study of the Sanskrit conjugation system. This work rightly earned Bopp the title of founding father of modern historical-comparative linguistics, and shortly after, in 1821, he obtained the chair of comparative grammar, the first of its kind, at the recently founded Friedrich Wilhelm University in Berlin.¹ He was one of the first scholars to thoroughly investigate Sanskrit from a linguistic point of view and compare it meticulously with related tongues from Europe and Asia. His comparative efforts laid the basis for an effective and sound methodology to prove the kinship among the languages now known as ‘Indo-European’, a hypothesis first formulated in the seventeenth century, when the protolanguage was usually referred to as ‘Scythic’, and made popular by Sir William Jones’s (1746–94) enthusiastic discourse to his Asiatic Society in 1786. Jones, Bopp, and others such as Rasmus Rask (1787–1832) stood on the threshold of a new era, in which linguistics finally emerged as a separate discipline, emancipated from other fields of study, in the first place philology. One might expect theoretical and methodological questions to be foregrounded in this period, but was this indeed the case? And more specifically, did theorizing on the conceptual pair benefit from the development of institutionalized linguistics in the nineteenth century? Bopp’s remark quoted above suggests that the genealogical

¹ See Swiggers (2017: 177) for an assessment of Bopp’s principal merits and defects.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

220

   

interpretation of the language/dialect distinction rose to prominence in the context of historical-comparative grammar. Can this observation be generalized?

17.1 The beginnings of modern linguistics Even though Sir William Jones’s contribution to linguistics tends to be overemphasized by intellectual historians and linguists, it is impossible to overlook him, if only because his work contributed to the emergence of the field as a separate scholarly discipline. It did so by ‘put[ting] Sanskrit at the forefront of European interest’ (Rocher 2001: 1160–1). Still, Jones was not yet a rigorous linguistic scholar but rather a historian interested in fitting Asian ethnology into the traditional framework of the Bible (Trautmann 1998: 106–7; Campbell 2006: 247). That he was not a linguist pur sang can be noticed in his varied usage of the term dialect, among other things. He did not define the word, probably because he considered its meaning common knowledge, like most scholars before him. Jones employed it not only as a synonym of language and in the senses of ‘manner of speaking’ and ‘spoken language’ but also to refer to dialect as stemming from a language, much as it had been employed from about 1600 onwards.² The latter usage comes as no surprise, as Jones’s interest in linguistic kinship was shaped primarily by historiographical concerns. This focus was most famously embodied in his hypothesis about a common ancestor for a wide range of tongues, which included not only languages which linguists today consider Indo-European, such as Sanskrit, but also tongues currently classified in other families, such as Egyptian (Poser and Campbell 1992: 230; Blench 2006: 54). It is, however, remarkable that the term dialect did not figure in the famous and oft-quoted passage of his Third discourse in which he expressed this hypothesis (Jones 1799: 26). In the early nineteenth century, too, the conceptual pair was taken for granted, much as it had been in the early modern period. Franz Bopp, for instance, did not see anything wrong with using the terms Mundart and Dialekt and the English word dialect to denote a variety that was part of a group of genealogically related varieties descending from a common ancestral language (see e.g. Bopp 1816: 10, 89, 97, 105, 116). Bopp’s Danish contemporary Rasmus Rask likewise used the language/dialect pair throughout his published work without questioning it. What is more, Rask cultivated certain usages of the concept of dialect very typical of early modern scholarship. For example, he used a variation on the popular phrase ‘to differ only in dialect’ (Rask 2013 [1818]: 201). The German linguist and fairytale collector Jacob Grimm (1785–1863) also used the pair in its languagehistorical sense throughout his influential German grammar (Grimm 1819). ² For the former two meanings see e.g. Jones (1799: 80, 110, 138, 179, 201, 357). For the latter meaning see Jones (1799: 60, 80, 121).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

221

Put differently, Bopp, Rask, and Grimm relied on the conceptual distinction to express, much like Jones, genealogical relationships among different tongues. There were no theoretical discussions of the distinction in mainstream works by these early linguists, no doubt because they initially focused on the actual languages, their interrelationships, and the Indo-European protolanguage rather than on the metalinguistic concepts used to indicate the relationships among them (cf. Morpurgo Davies 1998: 131–3 on Bopp). The conceptual pair did, however, figure in Rasmus Rask’s ‘six degrees’ of linguistic classification, a framework he developed only rudimentarily in two texts of limited circulation. The first was a letter of 29 January 1819, the second a manuscript treatise of unknown date. Rask presented his classification method as his ‘new system for the grouping of all the languages on earth’, suggesting that he created it ex nihilo. The ‘six degrees’ were, in the original Danish, Race, Klasse, Stamme, Gren, Sprog, and Sprogart, ‘race—class—stock—branch—language— dialect’. Rask considered this scheme necessary in order to avoid ‘get[ting] completely lost in the infinite multitude of languages and dialects’.³ Rask did not, however, formulate definitions of the different ‘degrees’ and focused in actual practice on distinguishing between races.

17.2 Two pioneering dialectologists: Johann Andreas Schmeller and Albert Giese The early nineteenth century did, however, witness two scholars with an outspoken interest in dialectal diversity and the concept of dialect in particular. The first, Johann Andreas Schmeller (1785–1852), is widely known among historiographers and linguists.⁴ A pioneer of German dialect studies, Schmeller authored the monograph The dialects of Bavaria (1821), which included a very early dialect map, and composed a four-volume Bavarian dictionary (1827–37). The German philologist treated Bavarian dialects principally out of an antiquarian interest and assumed that dialectal diversification was caused in the first place by languageexternal circumstances. Even though Schmeller (1821) did not explicitly define dialect, readers of his work will notice that he believed dialect-level variation to affect primarily pronunciation, morphology, and the lexicon. More interesting still for early nineteenth-century conceptions of dialect is the almost entirely forgotten Wittenberg-born linguist Albert Giese (1803–34), a student of the German philologist August Boeckh (1785–1867) and Franz Bopp. Giese discussed the conceptual pair extensively in his monograph on the Ancient

³ See Benediktsson (1980: 22–3), whose English translation I quote. See also Morpurgo Davies (1998: 92). ⁴ See e.g. Brunner (1971); Scheuerer (1995); Niebaum and Macha (2006: 55–7).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

222

   

Greek dialects, which focused on Aeolic in particular and on the many myths surrounding this obscure dialect. The work was published only posthumously in 1837, three years after Giese’s untimely death. His treatment of the language/ dialect distinction featured quite prominently in the book. Its first chapter, entitled ‘Language and dialect’, was almost entirely devoted to it, even if it started out as a programmatic text about the aims and restrictions of historical-comparative linguistics.⁵ Giese’s work has been described by a nineteenth-century biographer as ‘one of the first and most inspiring attempts at applying the results and methods of comparative linguistic research to Greek dialectology’.⁶ Not all nineteenth-century readers were, however, convinced of the quality of Giese’s contribution. His fellow student August Friedrich Pott (1802–87), for instance, argued that Giese’s work exhibited certain unscientific characteristics. In particular, Pott (1974 [1884–90]: 92) took offence at Giese’s emphasis on Attic as the main norm of the Ancient Greek language. Another, more striking criticism was formulated by the presumed pioneer of modern Ancient Greek dialectology: Heinrich Ludolf Ahrens (1809–81). Tellingly for the early nineteenth-century lack of attention for metalinguistic concepts, Ahrens (1839: ) criticized his predecessor Giese precisely for his lengthy introduction. Giese used the term Mundart primarily in the genealogical, language-historical sense. Greek, Latin, and Sanskrit were related dialects, each of which could help the linguist complete the forms lacking in the others. Giese was aware that this genealogical interpretation was extrapolated from the micro-level of a language having different dialects to the macro-level of a language family and its members. He moreover framed the two concepts within a ‘forever lasting change’ (immerwährende Bewegung), which, in his view, constituted one of the main challenges for linguistic research (Giese 1837: 6). Several pages later, Giese elaborated on a thorny issue: how could one distinguish between dialects of a language and related languages? He assumed that dialects had a ‘closer connection’ (innigere Verbindung) than related languages, but at the same time he observed that dialects of the same language could be highly divergent, up to the point that a dialect resembled more a related language than a cognate dialect. This realization led him to make critical observations on the genealogical use of the term Dialekt and to elaborate on the ways in which the concepts dialect and language should be distinguished in historical-comparative linguistics: However much the concepts ‘dialect’ and ‘language’ seem to flow into each other in our prehistorically related language family, one can nevertheless offer certain

⁵ See Giese (1837: 3–22): ‘Sprache und Mundart’. The programmatic part coincides with pp. 3–7. ⁶ Leskien (1879: 151): ‘einer der ersten und anregendsten Versuche, die Resultate und die Methode der vergleichenden Sprachforschung auf die griechische Dialectologie anzuwenden’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

223

points strictly distinguishing between these concepts. These are, on the one hand, the original properties of a language and, on the other, certain more general sound laws which arose within the autonomous development of a language, for we find that both elements are preserved and observed in the dialects of a language, but they can lack in a dialect of the related language.⁷

Giese summed up this idea by means of the following maxim: ‘that which connects the dialects, keeps the prehistorically related languages apart’.⁸ He was thus one of the first modern linguists, if not the first, to consciously adapt the conceptual pair to the methodology of the new discipline of historical-comparative linguistics. He was, however, aware that his analysis was tied to a specific time window, since related languages such as Sanskrit, Greek, and Latin were originally themselves dialects of one and the same language. Giese (1837: 12) proceeded by applying his interpretation of the language/ dialect distinction to refute the age-old idea that Latin was a dialect of Greek or a ‘mixed language’ (Mischsprache) composed of the Aeolic dialect and a number of Italic languages. He pointed to, among other things, the Greek augment, not present in Latin, the latter allegedly preserving the most ancient situation. In what followed, Giese teased out the problem of the interrelationships of different dialects and their relationship to the language of which they were part. He put forward the following question in this context, which he did not manage to answer: did dialects derive from an originally unitary language or were there already dialectal differences in the beginning? Giese (1837: 13–14) did, however, suggest that the latter scenario was possible, and that one should judge every dialect relationship separately. When discussing the place of related dialects vis-à-vis the language under which they were subsumed, Giese characterized dialect as a relative concept which had to be defined with respect to the overarching concept of language. Interestingly enough, he used a traditional definition from Greek scholarship to do so, clearly reading his own interpretation into it: The following general bond is in the dialects: which they are not, as the language is abstracted from the dialects as such, and the dialects are nothing else than the ́ language in its difference (idíōma glo¯ssēs [“particularity of tongue”]). Yet the dialects which have appeared in historical times are not the parts of the language

⁷ Giese (1837: 12): ‘So sehr auch bei unsrer urverwandten Sprachfamilie die Begriffe „Mundart“ und „Sprache“ in einander zu fließen scheinen, so lassen sich dennoch bestimmte Punkte angeben, die jene Begriffe streng auseinander halten; diese sind einerseits die ursprünglichen Eigenthümlichkeiten einer Sprache, andrerseits bestimmte allgemeinere Lautgesetze, welche sich innerhalb der selbstständigen Entwickelung einer Sprache bildeten; denn beide finden wir in den Mundarten einer Sprache bewahrt und beobachtet, in einer der verwandten Sprache aber können sie fehlen’. ⁸ Giese (1837: 12): ‘was die Mundarten verbindet, hält die urverwandten Sprachen aus einander’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

224

   

as of the whole, as far as we conceive this as the whole that existed ever since the origin of the language, but they are the varieties of a specific language situation, which, mainly determined by certain general sound laws, has emerged out of another, older language situation . . . And the abovementioned general bond is itself a variable one, as far as a language disperses more or less into dialects in different periods.⁹

Giese tried to reconcile a synchronic interpretation of the conceptual pair with its language-historical usage. This attempt was innovative, since early modern scholars often made confused use of synchronic and diachronic conceptions of language and dialect or, more rarely, argued to distinguish clearly between them. There are other important differences between Giese’s views and early modern thought, too. For instance, his insistence on a comparative approach to the dialects of a language and on the difficulty of distinguishing between related dialects and related languages seems to have no precedents. Even though the latter idea had already been suggested by Fulda in the second half of the eighteenth century, early modern scholars were usually not suspicious about the distinction between language and dialect. As Giese probably had not read Fulda’s paragraph, it is safe to conclude that he arrived at this conclusion on his own. In what followed, Giese offered an alternative definition of dialect in which the ethnic parameter was emphasized, once again clearly inspired by traditional Greek definitions. Rather untraditionally, he linked it to the derivative nature of dialects in relation to the language from which they stemmed (Giese 1837: 14, 17, 21). It is remarkable that Giese, although operating with centuries-old interpretations of the conceptual pair, did not refer to early modern works at all. This lack of acknowledgement of earlier work was typical of nineteenth-century scholarship in the humanities, which sought to break with the unscientific past, thus ignoring early modern achievements.¹⁰ Outside Germany, there was also increasing attention to the study of dialects. In France, in particular, the patois were mapped already in the first decades of the nineteenth century by the efforts of members of the Coquebert de Montbret family, especially Charles-Étienne and his son Eugène, both of whom worked at the French Bureau de la statistique from 1806 onwards (McCain 2018: esp. 41,

⁹ Giese (1837: 14): ‘Jenes allgemeine Band ist in den Dialekten das, was sie nicht sind, denn es ist die Sprache abstrahirt von den Dialekten als solchen, und die Mundarten sind nichts anderes als die Sprache in ihrem Unterschiede (ἰδίωμα γλώσσης). Aber die in der geschichtlichen Zeit zum Vorschein gekommenen Dialekte sind nicht die Glieder der Sprache als des Ganzen, insofern wir diese als das Ganze, welches vom Ursprung der Sprache her bestand, auffassen, sondern sie sind die Arten eines bestimmten Sprachzustandes, welcher, hauptsächlich durch gewisse allgemeine Lautgesetze bedingt, aus einem andern, ältern Sprachzustande entsprungen ist . . . Und das genannte allgemeine Band ist selbst ein veränderliches, je nachdem eine Sprache in den verschiedenen Zeiten mehr oder minder in Dialekte auseinander geht’. Cf. Chapter 2, Section 2.2. ¹⁰ Cf. Lamers et al. (2020) and the other papers in the same journal issue.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

225

78–82). As early as September 1807, Charles-Étienne developed the plan to conduct enquiries into local speech varieties—mainly of Romance but also of Germanic and other stock—which resulted in a thorough geographical study with statistical components and the collection of texts in these regional varieties, including numerous versions of the Parable of the Prodigal Son (cf. [Coquebert de Montbret] 1831). They did not directly engage with the language/dialect distinction, but apparently regarded the patois spoken in Napoleon’s empire as vulgar, impure, and grammatically incorrect (McCain 2018: 78–9). In the spirit of the recent French Revolution, French was propagated as the pure language of the nation (see Chapter 23, Section 23.2). In conclusion, remarkable though the early nineteenth century may be for the history of linguistics in many respects, it is rather uninteresting as it comes to theorizing about the language/dialect distinction. Active comparison of the IndoEuropean languages and their specificities attracted most of the linguists’ attention, which left barely any space for more general reflections on methodological and conceptual issues. The scholarly focus broadened in the last decades of the nineteenth century, when dialectology emerged as a subdiscipline of linguistics and the fixation with the ancient Indo-European tongues gradually ceded to a more encompassing study of language. Several eminent linguists developed an interest in the conceptual pair, which they approached with increasing scepticism. Among them were August Schleicher and William Dwight Whitney.

17.3 Questioning the conceptual pair: August Schleicher and William Dwight Whitney At the outset of his 1861 comparative grammar of Indo-European languages, the Thuringian linguist August Schleicher (1821–68) pointed out that ‘the distinction between mundart, dialect, sprache cannot be determined in general terms’.¹¹ This reservation did not, however, keep Schleicher from using these words throughout his work. Rather unsurprisingly, he took them primarily in the long-established language-historical sense. After all, Schleicher was a historical-comparative linguist who tried to arrive at a better understanding of the Indo-European language family, thus consolidating the young discipline of linguistics. The method he adopted for doing so was as much unprecedented as it was daring; Schleicher relied on a meticulous comparison of the Indo-European dialects to fully reconstruct their common ancestral language, now usually known as ‘Proto-Indo-European’. He even composed a short fable in his version of the reconstructed protolanguage. ¹¹ Schleicher (1861: 4): ‘der unterschid von mundart, dialect, sprache ist im algemeinen nicht fest zu stellen’. On Schleicher’s linguistic views and his place in the history of linguistics see e.g. Robins (1967: 179–80), Koerner (1989); Morpurgo Davies (1998: 167–74, 196–201).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

226

   

The language-historical interpretation of the relationship between language and dialect, moreover, fitted in well with Schleicher’s strict conception of languages as living organisms. In fact, it is not unimaginable that the family metaphors often used in connection with language families and language/dialect relationships encouraged Schleicher in developing such a view on human language, even though it was surely not as constitutive as his study of biology. Schleicher did, however, assign other meanings to mundart and dialect as well, associating it in particular with superficial linguistic variation. For instance, the sound variation ‘:r = d’ in Umbrian, an Italic language, Schleicher (1861: 221) characterized as ‘only dialectal’ (‘nur dialectisch’). Two years later, in his essay Darwinian theory and linguistics, Schleicher (1863: 12–13, 19) consciously reflected on the language-historical interpretation of the conceptual pair, transforming it into four layers of variation: languages, dialects, subdialects, and idiolects. These he compared to concepts from his secondfavourite branch of learning: biology. Unlike biological diversity, however, linguistic variation was of a gradual nature, Schleicher realized, which is why he added that it was impossible for linguists to clearly distinguish between languages, dialects, and subdialects. Furthermore, scholars greatly varied in their usage of the terms Sprache and Mundart: ‘what some call languages, others call dialects and vice versa’.¹² In conclusion, Schleicher was well aware of the arbitrariness of the conceptual pair and the difficulties that presented themselves in defining and applying it. This realization did not, however, keep him from resorting to the conceptual pair in his language-historical work, doubtlessly for lack of any satisfying alternatives. A contemporary of Schleicher, the American linguist and Sanskrit specialist William Dwight Whitney (1827–94) even devoted a lengthy article to the conceptual pair. Whitney was the author of a contribution entitled ‘Languages and Dialects’, which appeared in 1867 in The North American Review.¹³ The article focused on ‘the phenomena of dialects, their rise and their overthrow,—the diversification and assimilation of languages’ (Whitney 1867: 32). Working with the concept of language community, Whitney proved himself aware not only of regional variation in a language but also of the fact that individuals, families, and professional groups exhibit differences of speech. In this context, he expressed a view highly similar to Schleicher’s: Now all these differences, limited as their range may be, are in their essential nature dialectic; the distinction between such idioms, as we may properly call

¹² Schleicher (1863: 19): ‘So war denn begreiflicher Weise noch kein Sprachforscher im Stande eine genügende Definition von Sprache im Gegensatze zu Dialekt u. s. f. zu geben. Was die Einen Sprachen nennen, das nennen die Andern Dialekte und umgekehrt’. ¹³ See e.g. Silverstein (1971) and Alter (2005) on Whitney’s linguistic views.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

227

them, and well-marked dialects, or related but independent languages, is one of degree only, not of kind.¹⁴ It will be noticed that we have used the terms “dialect” and “language” interchangeably, in speaking of any given tongue; and it will also, we trust, have been clearly seen how vain would be the attempt to establish a definite and essential distinction between them, or to give precision to any of the other names which indicate the different degrees of diversity among related tongues. No form of speech, living or dead, of which we have any knowledge, was not or is not a dialect, in the sense of being the idiom of a limited community, among other communities of kindred but somewhat discordant idiom; none is not truly a language, in the sense of being the means of mutual intercourse of a distinct portion of mankind, adapted to their capacity and supplying their needs. (Whitney 1867: 54)

Labels such as idiom, patois, dialect, and language were determined by ‘external circumstances’ and, Whitney implied, not backed by linguistic evidence. Much like Schleicher, Whitney drew a parallel with ‘natural history’ and its concepts of variety and species. Despite claiming to use the terms language and dialect interchangeably, it is clear that Whitney employed the term dialect to express the genealogical kinship among English, German, and Swedish as well as among Russian, Persian, and Hindi. What is more, he relied on the conceptual pair to describe natural linguistic cycles; local varieties of a language developed into dialects, which eventually diversified ‘into distinct, and, finally, widely dissimilar languages’ (Whitney 1867: 38). This idea also implied that related dialects exhibited only little variation, whereas distinct languages differed substantially from each other. Whitney regarded dialectal diversity as a universal phenomenon, ‘inseparable from the being of any language at any stage of its history’ (Whitney 1867: 62). By arguing so, he was reacting against two contemporary scholars, Ernest Renan (1823–92) and Max Müller (1823–1900), who ‘affirm that the natural tendency of language is from diversity to uniformity; that dialects are, in the regular order of things, antecedent to language; that human speech began its history in a state of infinite dialectic division, which has been ever since undergoing coalescence and reduction’ (Whitney 1867: 57, see also pp. 60–1). In other words, despite his claims that both terms could be used interchangeably, Whitney was clearly differentiating language from dialect. Indeed, his actual usage suggests that he was limiting neither the term dialect to his social definition nor the term language to his functional definition but also regarded them as two poles of a conceptual bipartition with genealogical implications. Whitney’s

¹⁴ Whitney (1867: 36). Cf. Alter (2005: 140, 260–1) on Whitney’s relationship to Schleicher’s thought.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

228

   

variable reliance on the conceptual pair lays bare the polysemy of the language/ dialect pair and the flexibility with which linguists could use it. In Whitney’s case, this variability is all the more remarkable since it occurred in one single text entirely devoted to the issue of ‘Languages and Dialects’.

17.4 The late nineteenth century: between continued usage and increasing scepticism Around the time when the validity of the conceptual pair began to be questioned, dialectology emerged as a separate subfield of linguistics, for which Georg Wenker’s (1852–1911) dialect maps are usually cited as a foundational monument today (see e.g. Chambers & Trudgill 1998: 15–16; Schrambke 2010). The results of this new dialectological research seemed to confirm the arbitrary nature of the distinction between language and dialect and led some linguists to argue that both concepts were abstractions made from concrete linguistic facts and did not correspond to any entities in actual reality (see Morpurgo Davies 1998: 238, 289; Chapter 18). Still, in the last decades of the nineteenth century, most linguists kept on employing these abstract concepts in their work because it allowed them to reach ‘empirically valid results’ (Morpurgo Davies 1998: 330). It was especially the so-called neogrammarians, Junggrammatiker in German, who made eager use of the dialect concept. This school of linguists which originated at the University of Leipzig reacted against the narrow focus on the ancient Indo-European tongues and emphasized the importance of dialect studies in historical-comparative research, which, in their view, should be focused on formulating infallible sound laws (ausnahmslose Lautgesetze) and turn the attention to actual living language (see Osthoff and Brugmann 1878: –; Murray 2010). Dialects were mined for their archaisms and, more importantly, served as pure specimens of natural language development; the primary access gates to this living lab of dialects usually were what linguists have later called ‘NORMs’, ‘NonMobile Older Rural Males’ (e.g. Kristiansen in Boberg et al. 2018: 111). The validity of this focus on NORMs later came to be criticized on various grounds, including its gender bias and its assumption of lack of mobility and contact, especially as transportation and communications technologies have since profoundly changed human interactions. In the second half of the nineteenth century, then, contemporary vernacular dialects were, for the first time in history, considered an object worthy of serious and systematic research in and of itself, the primacy of the written word finally being superseded. Early dialect geographers were, however, not much concerned over the distinction between language and dialect, even though their maps foregrounded again the spatial conception of dialect. Neither did many neogrammarians, who strongly disagreed with the dialect geographers on other points

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

229

(Jankowsky 1972: 133), discuss the problematic nature of the conceptual pair at length. The neogrammarian Berthold Delbrück (1842–1922), for instance, still took it for granted when drawing up his Introduction to language study (1880) and interpreted it primarily in the usual language-historical sense (Delbrück 1880: 120–7). His colleague Hermann Paul (1846–1921), in turn, compared language families, languages, and dialects to biological classes, much as Schleicher had done, and emphasized the gradualness of linguistic variation (Paul 1880: 231–2). This observation did not lead him to problematize the conceptual pair, however. Towards the end of the nineteenth century, concern over the validity of the concept pair increased. For instance, when discussing the precise number of the world’s languages, the linguist August Friedrich Pott (1974 [1884–90]: 251), a student of Franz Bopp’s, pointed out that it was difficult to observe a clear distinction between language and dialect. A more interesting example still is the German sinologist Georg von der Gabelentz (1840–93), who reflected on the criteria used to distinguish between language and dialect in his work Language science, its tasks, methods, and current results, which was first published in 1891 and appeared in an enlarged edition posthumously in 1901. Even though Gabelentz’s work was relatively uninfluential, it is worthwhile to take a closer look at his interesting comments on the conceptual pair, since they illustrate very well how it was increasingly questioned in the late nineteenth century.¹⁵ According to Gabelentz, the only correct criterion whereby to distinguish a language from a dialect was mutual intelligibility, which, however, had not yet been defined in an objective fashion. It was this flaw which he tried to remedy. Gabelentz endeavoured to do so by insisting on the importance of the speech of uneducated people. Although attempting to design a workable criterion, he was nonetheless aware that it was impossible to satisfactorily solve its arbitrariness. For there was no freezing or boiling point in mutual intelligibility, as there was with temperature. As a consequence, the language/dialect distinction and the criterion he proposed could only be used for convenience. Still, such distinctions did have scientific value, claimed Gabelentz, even though he refrained from explaining in what this value exactly consisted.

17.5 Conclusion The conceptual pair was silently adopted by linguists when the study of language emerged as an autonomous discipline in the early nineteenth century, as if it were an ever-present and unquestionable given. The language/dialect distinction was not recognized for what it was: a product of the reflection and linguistic experience ¹⁵ Gabelentz (1901: 54–8). On the limited impact of Gabelentz see Elffers (2008), but compare this with Albrecht (2007: 22–3, 178).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

230

   

of Renaissance scholars. The language-historical interpretation was foregrounded at an early stage, obviously connected to the focus of institutionalized language studies in the nineteenth century, largely coinciding with historical-comparative Indo-European linguistics. The early comparative philologist Albert Giese was exceptional among his colleagues in that he consciously attempted to restyle the conceptual pair to the principles of the young discipline. Tellingly, Giese did not elaborate on early modern views but tried instead to connect his ideas to traditional Greek thought. Interpretations other than the language-historical sense lived on, too, principally mutual intelligibility and the Aristotelian criterion. The spatial conception of the distinction was likewise rehabilitated when dialect geography emerged in the late 1870s and became the primary dialectological paradigm for decades to come. In the second half of the nineteenth century, the conceptual pair was increasingly problematized by prominent linguists such as Schleicher, Whitney, and Gabelentz. It comes as no surprise that they were not aware of Fulda’s scepticism, as there seems to have been a general neglect of early modern work. It is, however, remarkable that they refrained from referring to one another’s ideas when discussing the conceptual pair. This lack of debate stands in stark contrast with their treatments of other issues, in which they displayed detailed familiarity with the work of their colleagues. It appears that in their historical-comparative enthusiasm they had no use for theoretical debates on key concepts such as language and dialect, which they initially perceived as obvious auxiliary concepts. Or perhaps they wanted to keep themselves far from this thorny problem, which could shake the very foundations of linguistic study. After all, if it was unclear what constituted a language or a dialect, the very object of this new discipline would be in danger. Yet linguists could not look away for ever.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

18 Schuchardt the iconoclast In her magisterial overview of nineteenth-century linguistics, the late Anna Morpurgo Davies devoted a section to ‘Hugo Schuchardt, language and dialect’.¹ She did so with good reason, as the German linguist Hugo Schuchardt (1842–1927) radically reacted against the common practice among his fellow linguists to take the concepts of language and dialect for granted. What led Schuchardt to formulate a smashing critique of his colleagues and their fundamental assumptions? To answer this question, I have to retrace my steps to the nineteenth-century debate on the modality of linguistic diversification, centring around the question: is it a branching out or a diffusion in waves (cf. Goebl 1983)?

18.1 Tree or waves? Even though August Schleicher on several occasions expressed his discontent with the arbitrary language/dialect distinction, he had done so only in passing and in the end kept on working with it, most likely for practical considerations. In fact, Schleicher needed it to develop his influential Stammbaum theory. According to this organicist model, the original Proto-Indo-European language ramified into distinct families of languages. These families in turn split up into different languages, which finally branched out into different dialects (Figure 18.1). Schleicher’s tree model, in other words, presupposed strict boundaries between linguistic branches, languages, and dialects. It is well known that his student Johannes Schmidt (1843–1901) reacted against the rigidity of Schleicher’s divisions with his wave theory (Wellentheorie). In his model, Schmidt tried to stress the connection between the different Indo-European languages rather than their division. They were initially a continuum of dialects in which some gained the upper hand, thus abolishing adjoining varieties and creating larger gaps in the continuum. The Attic dialect, Schmidt (1872: 27–8) pointed out, eclipsed all other Greek dialects, becoming the main variety of Greek. To clarify his hypothesis, Schmidt employed the image of a slanted line to describe the initial situation, which became indented when gaps were created in the linguistic continuum as certain tongues came to prevail—contrast Figures 18.2 and 18.3.

¹ Morpurgo Davies (1998: 287–90), an important source of inspiration for the present chapter.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

232

  

Figure 18.1 August Schleicher’s (1863) family tree model Source: Archive.org. Public domain

Wave theory, however, has taken its name from another powerful image of Schmidt’s, developed only very briefly a few sentences earlier as an alternative to Schleicher’s Stammbaumtheorie. Linguistic differentiation was like a wave, ‘extending itself by means of concentric circles which become ever weaker as they are further removed from the middle’.² Schmidt tried to develop a model that better matched the complex reality of language diversity and change than the artificially strict divisions inherent to his teacher’s tree model. The alternative models divided linguists into two camps, even though some, like August Leskien (1840–1916), insightfully argued at an early stage that the theories complemented rather than contradicted each other (Leskien 1876). The debate shook linguistics to its foundations—finally, one might even say, as for a long time theoretical issues had been blatantly neglected. One of the major problems that now compelled the attention of linguists was the language/dialect distinction. The basic assumption that ‘languages or dialects were self-contained ² Schmidt (1872: 27): ‘Ich möchte an seine stelle das bild der welle setzen, welche sich in concentrischen mit der entfernung vom mittelpunkte immer schwächer werdenden ringen ausbreitet’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

233

Sanskrit

Celtic

Figure 18.2 The linguistic continuum as a slanted line according to Schmidt Source: Schmidt 1872: 27–8

Sanskrit

Celtic

Figure 18.3 The linguistic continuum broken up according to Schmidt Source: Schmidt 1872: 27–8

units easily identifiable’ was increasingly contested, not in the least by Hugo Schuchardt, who even before Schmidt published his book in 1872 had pointed out the impossibility of adequately classifying languages and dialects and went much further than his colleague, who had still believed in an originally unitary Proto-Indo-European tongue (Morpurgo Davies 1998: 285–7).

18.2 A pioneering outsider Who was this Hugo Schuchardt (Figure 18.4), who adopted an unprecedentedly radical attitude in studying human language?³ Schuchardt was born on 4 February 1842 in the Thuringian town of Gotha, where he spent most of his youth. After passing his final exams at the local gymnasium in 1859, he moved to Jena to study law, but he soon changed to philology, attending the classes of, among others, August Schleicher, the man whose Stammbaumtheorie he later criticized. In 1861 he switched to the University of Bonn, where he obtained a PhD in 1864 with a dissertation on the vowel system of Vulgar Latin, written in scientific Latin. In 1866–8 he published a three-volume German version of this work, which ³ Biographical information is drawn from the detailed and useful website of the Hugo Schuchardt Archiv, consultable at http://schuchardt.uni-graz.at/ (last accessed 14 January 2020). For Schuchardt as outsider see e.g. Swiggers (2000).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

234

  

Figure 18.4 Hugo Schuchardt in 1922, picture by Hermann Urtel Source: Hugo Schuchardt Archiv, Graz. Reproduced with permission

remained his most extensive publication. In 1870, Schuchardt obtained his habilitation at the University of Leipzig with a study of Romansh sound changes, accompanied by a notable trial lecture, ‘On the classification of the Romance dialects’. By this point, he had already developed several of his main fields of interest: Romance philology and contact linguistics, both areas in which he broke new ground. In 1873, Schuchardt was appointed professor of Romance philology at the University of Halle, a neogrammarian stronghold. Three years later he managed to obtain the chair in the same discipline at Graz University on the recommendation of Johannes Schmidt. From the second half of the 1870s onwards, Schuchardt undertook several research trips with various destinations across Europe. His travels to Wales and Spain drew his attention to new languages, and he started working on Celtic in the 1870s and Basque in the late 1880s. Also in the 1880s, he quasi-founded a new linguistic subfield, that of creole linguistics, which for Schuchardt involved studying the mechanisms behind what he typically called ‘language mixture’, or Sprachmischung in German.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     . . .

235

In 1885, he wrote a radical critique of the neogrammarians, reproaching their too systemic and mechanistic approach to language change. By this time, Schuchardt had become a well-known linguist, who corresponded with countless researchers across Europe.⁴ His fame and his extensive scholarly network he owed to his many travels and his pioneering contributions to linguistics. In the 1890s, Schuchardt developed an interest in still other languages, including Caucasian tongues and Iberian, an extinct language anciently spoken on the Iberian peninsula, which he believed to be closely related to Basque. Schuchardt tried to prove this controversial hypothesis rather unsuccessfully. In 1900, Schuchardt retired, but this did not mark the end of his scholarly career. He kept on working on language-specific themes, especially regarding Basque, while broadening his perspective by touching on more general linguistic issues, too. In 1922, his Austrian admirer Leo Spitzer (1887–1960) edited the Hugo Schuchardt-Brevier: Ein Vademecum der allgemeinen Sprachwissenschaft, an anthology of Schuchardt’s work compiled in celebration of his 80th birthday. Copies of this book sold out quickly, and Schuchardt collaborated with Spitzer to produce a new edition, corrected and enlarged by the author himself. He did not witness its eventual publication in 1928, however, as he had passed away on 21 April 1927.

18.3 Breaking down the walls between languages and dialects A student of Schleicher’s, Schuchardt was thoroughly acquainted with his professor’s work. He could even appreciate it, and in the first volume of his book on the Vulgar Latin vowel system, published in 1866, he applauded Schleicher’s (1863: 19) observation that ‘what some call languages, others call dialects and vice versa’ (see Chapter 17, Section 17.3). Schuchardt believed, however, that this needed a further qualification, all too long neglected by fellow linguists: This lies in the essence of language diversity, which is a gradually growing one. Untermundart, Mundart, Dialekt, Sprache are relative concepts. If we wanted to argue that related dialects were promoted to the rank of languages on condition that mutual understanding ceased, then nothing would be gained by that; for the ability of understanding is different in different persons and understanding itself is endlessly gradual.⁵ ⁴ See the online Hugo Schuchardt Archiv for an overview and edition of his vast correspondence. ⁵ Schuchardt (1928: 164): ‘Es liegt dies im Wesen der Sprachdifferenz, die eine allmählich wachsende ist. Untermundart, Mundart, Dialekt, Sprache sind relative Begriffe. Wollten wir behaupten, verwandte Dialekte stiegen dann zur Würde von Sprachen empor, wenn das gegenseitige Verständnis aufhörte, so würde dadurch nichts gewonnen sein; denn die Fähigkeit des Verstehens ist bei verschiedenen verschieden und das Verstehen selbst unendlich abgestuft’,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

  

236

Schuchardt was, in other words, the first to indisputably recognize the complex and gradual nature of linguistic variation, which, he believed, arbitrary concepts such as language, dialect, and subdialect could never adequately convey. For this reason he also rejected mutual intelligibility as a workable criterion. Whereas his illustrious professor had not fully realized the implications of his observations on these key concepts, Schuchardt did not hesitate to problematize them even at the outset of his career—he was only twenty-four when his first volume on the Vulgar Latin vowel system appeared. In the third part of the same work, published in 1868, he further reflected on the gradual nature of linguistic variation, observing that ‘neighbouring Dialekte, Mundarten, Untermundarten etc. are not abruptly delimited from each other but approximate each other, flow into each other’.⁶ Schuchardt considered it impossible to draw strict borders between speech forms, which he believed to be a consequence of contact phenomena. Different language regions were connected because speakers of different varieties interacted with members of other speech communities. The intensity of this contact differed depending on the degree of civilization. This contact was, in other words, a gradational phenomenon itself. Schuchardt tried to visualize his conception of dialectal diversity by means of a telling, although rather complicated scheme (Figure 18.5), which he explained as follows: If in the accompanying figure the four large triangles represent main dialects and the sixteen small ones represent subdialects, then e.g. a, A, α, a are not to the same degree different from the dialects having B as their centre, but a to the greatest degree, A to a lesser degree, α, a to the smallest degree. What is more, the general similarity between α and b is possibly greater than that between α and a.⁷

a

α

A

δ

b

a

D

d

B

b c

d γ

C

β c

Figure 18.5 Schuchardt’s (1928 [1868]: 165) conception of dialect groups visualized

⁶ Schuchardt (1928: 164): ‘so finden wir fast überall, dass benachbarte Dialekte, Mundarten, Untermundarten usw. nicht schroff gegeneinander abgrenzen, sondern sich aneinander annähern, ineinander überfliessen’. ⁷ Schuchardt (1928: 164–5): ‘Wenn in der nebenstehenden Figur die vier grossen Dreiecke Hauptdialekte, die sechzehn kleinen Unterdialekte darstellen, so sind z. B. a, A, α, a nicht im gleichen Grade von den Diall. mit dem Zentrum B verschieden, sondern a am meisten, A weniger, α, a am meisten [sic pro wenigsten]. Ja, die allgemeine Ähnlichkeit zwischen α und b ist möglicherweise grösser, als die zwischen α und a’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.     . . .

237

Yet the schedule was still too rigid in its division of speech forms, Schuchardt believed, which was why he resorted to another image, one which sounds strikingly familiar: Let us imagine language in its entirety as a body of water with an even surface; this water is put in motion as a result of wave centres that are formed at different locations in it, the systems of which, depending on whether the intensity of the driving force is of a greater or smaller extent, traverse each other.⁸

Schuchardt wrote this in 1868, four years before Schmidt developed this image in his booklet on the kinship relationships among Indo-European tongues. Apparently, Schmidt did so independently from his colleague, since on 14 April 1874 he asked Schuchardt in a letter to provide him with the reference to the work in which Schuchardt had arrived at insights similar to his own.⁹ It seems that Schuchardt himself had drawn Schmidt’s attention to their parallel conclusions after a presentation by Schmidt at a philology conference in Leipzig in late May 1872, where they probably met for the first time. Even though Schuchardt referred to his 1870 trial lecture on Romance dialects rather than to his magnum opus on Vulgar Latin, the fact that Schmidt had to make an enquiry about Schuchardt’s writings suggests that he was not well-acquainted with his colleague’s work. It is not very surprising that Schuchardt noticed a parallel between Schmidt’s 1872 treatise and his own lecture on the Romance dialects. In these works, both linguists emphasized the difficulty and even absurdity of drawing exact borders between languages and dialects; both linguists criticized Schleicher’s Stammbaumtheorie, pointing out the unevenness of linguistic change and the asymmetric cohesion between speech forms. Schmidt’s request to have Schuchardt’s lecture in a printed form was in vain, however, since Schuchardt only published it in 1900. He eventually decided to do so because he felt embarrassed, as he had cited this unpublished lecture so very often that he could no longer leave it unprinted. He was moreover aware of the fact that, in spite of its blemishes, its content was pioneering. In this lecture, Schuchardt had argued against the classificability of dialects in general; he spoke in German of ‘die Nichtklassifizirbarkeit der Mundarten’ (Schuchardt 1900: 1*). In other words, he had pointed out the inadequacy of concepts such as dialect because of contact and mixture phenomena to which all speech forms were susceptible, albeit

⁸ Schuchardt (1928: 165): ‘Denken wir uns die Sprache in ihrer Einheit als ein Gewässer mit glattem Spiegel; in Bewegung gesetzt wird dasselbe dadurch, dass an verschiedenen Stellen desselben sich Wellencentra bilden, deren Systeme, je nach der Intensivität der treibenden Kraft von grösserem oder geringerem Umfange, sich durchkreuzen’. Cf. Goebl (1983: 33); Mayrhofer (1983: 135–6); Pirazzini (2013: 29–30). ⁹ See letter 01-10093 at the Hugo Schuchardt Archiv.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

238

  

to different degrees.¹⁰ By publishing this old text, Schuchardt wanted to demonstrate his own originality on this point. The fact that he still found it relevant to divulge his lecture so many decades after he had written and read it may suggest that in 1900 the discussion about key metalinguistic concepts such as language and dialect was intensifying.

18.4 Linguists, shibbolethists Schuchardt adopted a more radical attitude than Schmidt, as he argued that names for languages and dialects were abstract collective terms taken from geography rather than identifiable linguistic entities.¹¹ In fact, there were even differences within cities such as Venice and Rome, he claimed, thus consciously reiterating a centuries-old topos (Schuchardt 1900: 12–14; cf. Chapter 7, Section 7.2). In later publications, Schuchardt dwelt in greater detail on the problematic nature of concepts such as language and dialect. Especially in a treatise on learning the Basque language in its Sara variety, published in 1922 and not included in the Brevier, he developed interesting ideas on the term Mundart, even if it seems that he did not want to lose himself too much in theoretical considerations. When discussing the different accentual systems within Basque and their dialectal mapping, Schuchardt felt compelled to broach ‘the general problem of dialects’ and to touch on the troublesome word Mundart, despite his strong aversion to terminology—nomina sunt odiosa, he complained.¹² Mundart was not a scientific but a popular concept, argued Schuchardt, which linguists had adopted but were unable to adapt to their scientific goals. He then revealed its problematic nature in a surprisingly accurate and concise fashion: On the one hand, it is something relative, with dialect next to itself and language above itself; on the other hand, it is something complex, lacking a required coherence of its parts and lacking a fixed delimitation, so not an individual object, no organism, as it has so often been conceived.¹³

The gist of what followed boils down to this: no clear borders between dialects and languages could be drawn, and linguists who had done so relied on only a selection of mainly phonetic features. And if there were any borders, these were ¹⁰ See the accompanying preface in Schuchardt (1900: 1*) as well as Schuchardt (1928: 193). On the contact phenomena see in particular Schuchardt (1900: 10–11). ¹¹ Schuchardt (1900: 12): ‘nur geographische Kollektivausdrücke’. Cf. Schuchardt (1928: 119): ‘die Mundarten sind uns kaum mehr als geographische Begriffe’. ¹² Schuchardt (1922: 10): ‘das allgemeine Mundartenproblem’. ¹³ Schuchardt (1922: 10): ‘Einerseits etwas Relatives, mit Mundart neben sich und Sprache über sich; anderseits etwas Komplexes, ohne notwendigen Zusammenhang der Teile und ohne feste Umgrenzung, also kein Individuum, kein Organismus, wie man es so oft aufgefaßt hat’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. , 

239

arbitrary and due to a long-term interruption of contact between different speech communities. In deploying this argument, Schuchardt once again referred to his 1870 lecture on Romance dialects. Linguists, he noticed, seemingly could not do without dialects, much as geographers could not do without their precious system of coordinates. And they proceeded wrongly, since they first selected a number of features out of all linguistic facts and then claimed that these were responsible for constituting a dialect. On this limited basis, they construed the linguistic entity of dialect. As a result, it was no more than a ‘mental object’ (Gedankending), to which they had a priori assigned a number of essential properties. They started, in other words, from a fictitious concept and a limited set of features rather than from linguistic facts when studying language variation. Linguists were therefore, Schuchardt ingenuously judged, mere shibbolethists (Schibbolethiker). Just as the biblical Gileadites had used the Hebrew word shibbólet to identify and kill the enemy Ephraimites, who allegedly pronounced it as sibbólet, linguists relied on a few features only to identify distinct dialects, thus paralyzing research into language diversity. Looking at other features, Schuchardt implied, might have led them to different, equally fallacious conclusions. If there was any secure ground at all to introduce dividing lines into the continuum of language, it should be sought in the rhythm and musicality of a variety, which reflected a tribe’s individuality much better than phonetic features did.¹⁴ Yet this was more a cautious proposal by Schuchardt than an incontestable maxim, and he felt forced to conclude that no clear borders could be drawn between dialects, only between groups of linguistic facts. The latter should therefore be the main ingredient of language atlases, for individual features reflected the natural dynamism of language and its inherent mixedness much better than fictitiously static entities such as dialects and languages. After all, language was part of nature and fluctuated with it—or as Schuchardt put it with the famous saying of the Ancient Greek philosopher Heraclitus (fl. c. 500 ): in language, too, ‘everything flows’.¹⁵ For this reason, Schuchardt argued in his review of Ferdinand de Saussure’s Cours de linguistique générale (1916) that it was pointless to make a distinction between general and language-specific linguistics, since ‘there is only one, which at one time explores here, at another time there, which scrutizines at one time this matter and at another time that’.¹⁶ ‘Language’, Schuchardt continued, ‘constitutes a whole, a continuum’.¹⁷ At least, that was how it should be conceived, since there were transitions rather than fixed dividing walls in it.

¹⁴ Schuchardt (1922: 11), speaking of ‘die Eigenart des Stammes’. ¹⁵ Schuchardt (1922: 15). The original Greek saying is ‘pánta rheî’ (‘πάντα ῥεῖ’). ¹⁶ Schuchardt (1928: 318): ‘es gibt eine einzige, die bald hier, bald dort schürft, bald diesen, bald jenen Stoff durchleuchtet’. ¹⁷ ‘Die Sprache bildet eine Einheit, ein Kontinuum.’

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

240

  

18.5 Useless abstractions? In Schuchardt, the modern scepticism of the conceptual pair culminated. It had been ever-growing in the latter half of the nineteenth century, but Schuchardt carried the matter to an extreme and believed he had revealed the true nature of the concepts language and dialect. They were useless abstractions, and linguists should follow his lead and focus on the facts rather than on mental objects grounded in a priori assumptions. Schuchardt, in other words, was questioning one of the most fundamental premises of all language studies up to then: the idea that ‘languages or dialects were self-contained units easily identifiable’ (Morpurgo Davies 1998: 285). Even if linguists such as August Friedrich Pott (1856: 220) had expressed suspicions about the conceptual pair some decades before Schuchardt, they never considered the full consequences of the problem and never adopted an attitude as radical as Schuchardt’s. Several of Schuchardt’s contemporaries, too, realized the problematic character of the language/dialect distinction, but most drew less extreme conclusions. The French comparative linguist Victor Henry (1850–1907), for instance, treated the matter in thesis/antithesis format in his Linguistic antinomies of 1896, an intriguing work with relatively poor distribution.¹⁸ Henry put forward the following thesis: ‘The categories of language, languages and dialects, even of simple words, if examined closely, are only abstractions with no external reality’.¹⁹ The accompanying antithesis read: ‘There exists a science of language, which takes as its object of study the phenomena of the life of language, that is, the life of languages and the life of words’. In other words, linguistics studied language, but language was an abstract concept which did not correspond to anything in actual reality. This mismatch was problematic, Henry granted, especially since a ‘science does not usually debut by declaring itself without an object’ (Henry in Joseph 1996: 125; cf. Morpurgo Davies 1998: 290). Although repeatedly emphasizing that concepts such as language and dialect were mere abstractions deprived of reality and suggesting that perhaps ‘the objective notion of the infinity of speakers’ should replace ‘the empty abstraction’ that was language, Henry felt forced to comply with tradition based on practical considerations: But, after all, what matters are not words, but clear ideas. And if words can clarify ideas, it is on the condition that they not shock established traditions too much. So let us continue—we have no choice—to speak of language, even though we know that there is no language, but only people who speak. Let us continue to

¹⁸ See Joseph (1996); Morpurgo Davies (1998: 288). Noordegraaf (2014: 525) nuances the poor distribution of the work. ¹⁹ I quote from Joseph’s (1996: 124–37, here 124) translation of the first chapter of Henry’s work.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.  

241

speak of language families, of particular languages, of dialects, sub-dialects, patois, correct or incorrect pronunciation, provided that we understand in every case, beneath each of these words, a single latent meaning, the same for everyone: namely, in each of the irreducible units of language that science has provisionally established, an indefinite series of variations which get attenuated imperceptibly as one descends from the race to the nation, the province, the canton, the family and finally the individual—or rather which are parts of individual and initially imperceptible nuances of pronunciation and expression, and which are aggravated by circumstances until they end up at dialectal scission or even at linguistic isolation. (Henry in Joseph 1996: 128)

In short, to Henry it seemed that the concepts maintained their practical use as analytical linguistic tools, a conclusion with which Schuchardt no doubt would have disagreed. Henry’s views on the concepts of language and especially dialect were, however, not shared by a number of his compatriots who pioneered Gallo-Romance dialectology. In the French dialectological tradition, a true Schuchardtian spirit came to dominate, personified by linguists of the calibre of Gaston Paris (1839–1903), Paul Meyer (1840–1917), and Jules Gilliéron (1854–1926). These scholars did not believe in the existence of dialects, only linguistic features, a bold claim at the time, later embodied by Gilliéron’s monumental Linguistic atlas of France (1902–10).²⁰ The evidence did not show any obvious dialect divisions but rather fluctuating isoglosses, these French dialectologists claimed, as they lacked a holistic view on linguistic diversity. Notably, Paris and Meyer, both of whom were familiar with German methods of language study, were involved in a long-running controversy over dialect lines between French (langue d’oïl) and Occitan (langue d’oc) with the Montpellier school of dialectologists, who did work with a concept of self-contained dialects, most notably Octavien Bringuier (1820–75) and Charles de Tourtoulon (1836–1913).²¹ The mixed nature of dialectal diversity was acknowledged also by linguists outside the dialectological tradition, including Michel Bréal (1832–1915), famous for his foundational essay on semantics, in which he criticized the illusion of pure and unmixed dialects (Bréal 1897: 3–4). The French scepticism constituted a reaction against the neogrammarian mechanistic conception of language, which presupposed regular development and discrete entities (cf. Kretzschmar in Boberg et al. 2018: 61). Hans Goebl (1982; 2003) has fittingly labelled the French approach typophobia, a fear of assuming abstract linguistic entities with clear boundaries, as ²⁰ See in particular Goebl (1982; 2003) and Goebl in Boberg et al. (2018: 124–5). Cf. also Desmet (1996: 459); Engler (2000); Desmet et al. (2002); Joseph (2012: 462). ²¹ For an excellent synthesis of the frictions between both groups of French dialectologists as ‘communities of practice’ see most recently Klippi (2020: esp. 114–16 on the dialect concept). See also Hoyt (2006).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

242

  

opposed to the typophilia prevailing in German scholarship and with Graziadio Isaia Ascoli (1829–1907) also in Italy. Ascoli (1876) actively reacted against the views of Paul Meyer and Gaston Paris, with whom he had a dispute over FrancoProvençal dialect lines. In Ascoli’s (1876: 387) view, there were, in fact, separate dialect entities, marked by the ‘simultaneous presence’ or ‘particular combination’ of certain linguistic features.²² He thus adopted a more holistic perspective on regional language diversity than the Parisian school of dialectologists. The opposition between French (Parisian) and German–Italian dialectology also shaped later dialectological traditions. The former tended to focus on words and individual linguistic features, involving social factors in their analysis, whereas the latter tried to discover dialect areas in a more strictly regional sense. Scholars could change their minds as to which approach was preferable. A notable case is the Swiss linguist Louis Gauchat (1866–1942), who, trained in the Parisian dialectological tradition, was typophobic at first, before siding with the typophilic camp, developing the concept of core landscape (Kernlandschaft) to chart dialectal diversity in centres and peripheries (e.g. Lang 1982: 206–7; Desmet et al. 2002: 33–6). Reconciliatory approaches also existed, such as the one adopted by the Austrian-American dialectologist Hans Kurath (1891–1992; see Kretzschmar in Boberg et al. 2018: 63).

18.6 Conclusion Even though there previously had been suspicion of the language/dialect distinction, most nineteenth-century and early twentieth-century linguists kept on using it as a self-evident device in a very flexible way, priming one interpretation or another as it suited them. It was Hugo Schuchardt who fully exploited its problematic nature for the first time by tearing down the walls built between languages and the fences erected within them between dialects ever since the sixteenth century. Schuchardt did so in incisive comments scattered throughout his diversified oeuvre, as he did not produce a grand theoretical work. He richly deserves Anna Morpurgo Davies’s (1998: 288) well-chosen label of ‘iconoclast’, since he radically questioned a number of basic assumptions in linguistics, including the conceptual pair. Schuchardt was, however, most certainly not an isolated thinker but the chief exponent of a more general tendency, as the trends in French dialectology suggest. Linguists increasingly questioned fundamental concepts such as language and dialect, which led to the definitive abandonment of the influential organicistic conception of language, advanced and popularized by August Schleicher. The ²² ‘ma il distintivo necessario del determinato tipo sta appunto nella simultanea presenza o nella particolar combinazione di quei caratteri.’

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

243

question of ‘What is (a) language?’ was now wide open, yet many linguists chose to avoid and even abandon abstract concepts such as language and dialect and to focus instead on concrete linguistic facts and features—very often words— especially as the results of the new subfield of dialect geography confirmed the problematic nature of such abstract concepts (Morpurgo Davies 1998: 289–90; Boberg et al. 2018: 7–8). The linguistic facts, many linguists realized, did not allow anybody to presuppose clearly distinct and self-contained units such as languages and dialects. Language, then, was no longer countable. In mainstream linguistics, however, the conclusions of Schuchardt and the French dialectologists came to be overshadowed by the grand theories of the twentieth century.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

19 From Saussure to 1954 Structuralism and the language/dialect distinction

The closer I come to the present, the more difficult it becomes to provide a satisfying historical account, not only because of the enormous wealth of source material and the vast diversity of ideas and increasingly complicated approaches encountered in it, but also because the distance from the sources shrinks and secondary literature becomes much more limited.¹ For this reason, it is simply impossible to offer a detailed outline of the history of the conceptual pair in post1900 linguistics, which lies outside the scope of the present overview and deserves a study in its own right. I will limit myself to distinguishing a number of notable tendencies in the last 120 years within mainstream linguistics and the ways in which they relate to the previous history of the language/dialect distinction and are anchored in it. I will do so on the basis of the works of key linguists like Ferdinand de Saussure, as well as of recent handbooks in the field of linguistics and, especially, dialectology and language variation. The latter provide useful overviews of the diverging approaches to dialectal variation common in the last century and a half and thus constitute fine embodiments of the various attitudes modern linguists have adopted vis-à-vis the conceptual pair (see e.g. Boberg et al. 2018). A study of these text types will be my basis for providing a first rough sketch of the latest episode in the history of the language/dialect distinction, from which later studies can hopefully benefit. What was the place of the conceptual pair in the main currents of post-1900 linguistics, including most notably structural, generative, and sociolinguistic approaches to the phenomenon of language? In an overview that is partly chronological and partly thematic, I will argue that this question can be answered in terms of a love-hate relationship, since linguists’ attitudes have oscillated between two extremes: a deep-seated belief in the use of the language/dialect distinction and its downright dismissal.

¹ For an indication of the wealth of source material see e.g. the anthology in Göschel et al. (1976) and the papers in Göschel et al. (1980) as well as the chapters and references in Boberg et al. (2018). Cf. e.g. also Shapiro and Schiffman (1981: Chapter 2).

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   ’  

245

19.1 In Saussure’s class The beginning of the twentieth century witnessed a major turn in language studies, personified by no one better than the Swiss linguist Ferdinand de Saussure (1857–1913; see Joseph 2012 for an extensive biography). Whereas during his lifetime Saussure published mainly on classical topics from IndoEuropean historical-comparative linguistics, including a ground-breaking dissertation on the Proto-Indo-European vowel system, he adopted a very different perspective on language in some of his courses at the University of Geneva, published only posthumously in 1916 by his students Charles Bally and Albert Sechehaye (see Saussure 1995 [1916]). Saussure famously introduced a distinction between langue, language as a socially shared and structured system, shaped by the relationships between signs, and parole, actual usage of language in context. Trained as a philologist, he reflected extensively on both of these aspects of language. Over time, Saussure became the symbol of the shift in mainstream linguistics from traditional historical-comparative linguistics to structural linguistics, despite his interest in both approaches. This shift involved a major change in perspective from diachrony to synchrony, even if traditional forms of linguistics, historicalcomparative grammar as well as dialectology, persisted, albeit now relegated to the margin of mainstream research. Language as a system abstracted from time and place became the main object of linguists for about half a century, first in structuralism, then in generative grammar as developed by Noam Chomsky from the 1950s onwards. Since linguists operating within these strands were primarily interested in language as a selfcontained system, linguistic diversity was not high on their research agenda (see e.g. Bierbach 1978: 75). Instead, the starting point was the ideal speaker, often coinciding with the linguist himself, and this ideal speaker did not usually speak dialect. It is telling that no full-fledged structural and generative dialectologies have developed, even though there have been some attempts in this direction, with varying success (see Boberg et al. 2018: Chapters 4–5; the next section, 19.2; Chapter 21, Section 21.1). The comparative philologist’s books and the dialectologist’s questionnaires, one might say, ceded to the armchair, from where the linguist could comfortably think about the underlying system of language. What were the consequences of this turn in mainstream linguistics for ideas about the conceptual pair? The consequences seem to have been limited, at least initially. A pivotal figure, Saussure took a rather traditional, although not unambiguous, view on the language/dialect distinction. Saussure agreed with his French colleagues Gaston Paris and Paul Meyer that it was impossible to draw ‘clear boundaries’ between dialects, especially since ‘languages are in endless flux’. In his course on

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

246

   

geographical linguistics, he even actively taught Meyer’s pointed observation that there were no such things as dialects but only dialect features.² The fuzziness of linguistic borders forced him to think about the conceptual pair and its validity, a topic he broached in his third general linguistics course, taught in the winter semester of 1910–11. One of the students attending this course was Emile Constantin, whose extensive notes have been published in 2005 and which I take as the basis of my discussion. Constantin’s notes suggest that Saussure treated the language/dialect distinction, and especially the concept of dialect, at some length in his classes. The occasion for the Swiss professor to do so was a passing usage of the French term dialecte, which he mentioned while pointing out that linguistic difference was a matter of gradation. Latin and Greek were more closely related to each other than they were to Sanskrit, Saussure remarked. Even closer kinship existed among dialects, at which point he observed: But in pronouncing this word, we immediately add that one should not attach any absolute idea to the term dialect in relation to language. There is no precise point where the designation dialect is reached instead of the designation language.³

Some had proposed mutual intelligibility as a criterion, Saussure was well aware, but he does not seem to have been convinced of its utility. This phenomenon, too, was a question of gradation, and cutting the continuum in half would be an arbitrary act (Saussure 1995 [1916]: 278–9). There is an underlying contradiction here. Saussure on the one hand refrained from granting existence to dialects— there were only dialectal features—but he supposed on the other hand that abstract entities such as languages and dialects did, in fact, exist, even if it was difficult to say when a dialect started being a language. The Swiss linguist was aware of this discrepancy and believed that ‘in practice, one should preserve the term dialect’, yet that this should occur under certain conditions.⁴ Linguists should either ‘agree that one single feature suffices to characterize a dialect’ or if they considered all features, they should ‘limit themselves to a single point on the map and speak of the dialect of some town’.⁵ And even then, dialects should be

² Joseph (2012: 379, 462–3). See Constantin and Saussure (2005: 117–26). Cf. Chapter 18, Section 18.5. ³ Constantin and Saussure (2005: 100): ‘Mais en prononçant ce mot, nous ajoutons tout de suite qu’il ne faut attacher aucune idée absolue au terme de « dialecte » par rapport à celui de langue. Il n’y a aucun point précis où intervienne le nom de dialecte au lieu du nom de langue’. ⁴ Constantin and Saussure (2005: 123): ‘Dans la pratique, il faut conserver le terme de dialecte’. ⁵ Constantin and Saussure (2005: 121): ‘Pour qu’il y ait un dialecte : ou bien il faut 1 ) convenir qu’un seul caractère suffit pour caractériser un dialecte, 2 ) ou bien, si l’on prend tous les caractères, il faut s’enfermer sur un seul point de la carte et parler du dialecte de tel village’. Cf. Saussure (1995 [1916]: 276).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

247

conceived as open rather than closed entities, which took part in broader waves of linguistic innovation, Saussure claimed with clear reference to the wave model (Constantin and Saussure 2005: 125–6; see Chapter 18, Section 18.1). With these conditions in mind, one could continue speaking of dialects; otherwise one was forced to conclude that there were no such things (see also Kretzschmar 2009: 50–3, 55–8; Joseph 2012: 571, 731 n.32). Saussure attributed other values to the language/dialect distinction, however. Whereas he basically rejected the conceptual pair as an adequate means to map out regional linguistic diversity because of its gradational and uneven nature, he maintained it in two other traditional senses: the language-historical interpretation of language mother/daughter dialect and in the opposition standard language/dialect.⁶ With regard to the latter sense, he observed that ‘in a natural language, there are only dialects; a language left to itself is destined to indefinite fractioning’.⁷ From these dialects, that of the most civilized and powerful province was selected to become the literary and general language and was usually influenced by other provinces in due course (Constantin and Saussure 2005: 102–3; cf. Saussure 1995 [1916]: 267–8). Saussure thus recognized that it was impossible to find a purely linguistic criterion to distinguish languages from dialects—if there were any dialects at all—and at the same time that the conceptual distinction was shaped by extralinguistic factors, most notably culture and politics. In conclusion, the Genevan professor was fairly unoriginal when it came to his conception of the language/ dialect pair. He followed trends in French dialectology, with which he was wellacquainted, since he had spent several years in Paris as a lecturer at the École des Hautes Études (Joseph 2012: Part III).

19.2 Towards a structural dialectology: Uriel Weinreich’s diasystem In Saussure’s wake, the focus was for a long time on langue, the language system, and usually not on linguistic diversity, relegated to the parole domain (cf. Gordon in Boberg et al. 2018: 75). His colleagues and students limited themselves to noticing the gradual nature of linguistic variation (e.g. Sechehaye 1908: 251), if they touched upon the matter at all. Indeed, structural linguists of various strands seem to initially have had the tendency to insist on the relative nature of the conceptual pair, while at the same time continuing its usage, most often without

⁶ For an example of a reference to an Indo-European dialect see Constantin and Saussure (2005: 170). ⁷ Constantin and Saussure (2005: 102): ‘Dans une langue naturelle, il n’y a que des dialectes : une langue laissée à elle-même est vouée au fractionnement indéfini’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

248

   

clearly defining it. An exception was the American structuralist Leonard Bloomfield (1887–1949), who proposed in Language, his 1933 linguistics textbook, to expand the conceptual pair in order to overcome its binary nature. Bloomfield (1933: 52, 54) did so by integrating it into a five-layer division of literary standard, colloquial standard, provincial standard, substandard, and local dialect. This cutting up of the variation continuum was most likely as arbitrary as the distinction between language and dialect. Although more nuanced, it was based on a mixture of exclusively language-external criteria, principally geographic and sociocultural parameters. Ambiguous attitudes towards the language/dialect opposition were common among linguists active in the first decades of the twentieth century in general, also outside the sphere of structural linguistics strictly speaking. For instance, dialect was an ‘intangible notion’ according to the French linguist Antoine Meillet (1866–1936), Saussure’s successor in Paris.⁸ By consequence, the concept was not very useful to a comparative linguist, who should look for a more clearly defined research object. However, Meillet’s suspicion did not keep him from eagerly using the term dialecte himself throughout his rich oeuvre, mainly in its language-historical sense, in phrases such as ‘the Indo-European dialects’ (cf. Meillet 1908: 4–5; 1925: 53–9). A similarly ambiguous attitude was adopted by the German-born scholar Edward Sapir (1884–1939) in his textbook Language (Sapir 1921: 159–64). Inspired by his teacher Franz Boas (1858–1942), Sapir approached language principally from an anthropological point of view in the United States and paved the way for the later American structuralist tradition, closely associated with the figure of Leonard Bloomfield. Sapir did, however, contribute to the discussion about the concept of dialect by taking a sociological view on it, in which he did not presuppose any strict language/dialect opposition. In his entry for ‘dialect’ in the Encyclopaedia of the Social Sciences, he remarked that a ‘group of dialects is merely the socialized form of the universal tendency to individual variation in speech’ (Sapir 1931: 123), thus linking dialect to idiolect rather than language (cf. Göschel et al. 1976: –). Given the ambiguity of the attitudes which early twentieth-century linguists adopted towards the conceptual pair, the reader often has to infer their conceptions from their terminological usage. To stick with the example of Sapir, it is clear from his manual Language that he entertained traditional conceptions of the language/dialect distinction. Although noticing that the distinction is relative and arbitrary and nothing more than a snapshot in time, Sapir (1921: 163–4) seems to have assumed that two related dialects became different languages as soon as they were no longer mutually intelligible, even though he did not explicitly

⁸ See Meillet (1924: 15): ‘Cette notion est fuyante’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

249

posit this criterion. Central to his conception of the distinction was the traditional language-historical interpretation. A language developed into dialects, which in turn became languages again in due time. In a footnote, Sapir (1921: 164 n.6) dismissed the interpretation of dialect ‘in contrast to an accepted literary norm’ as irrelevant to his linguistic research. Linguists continued to harbour only superficial suspicions about the conceptual pair well into the 1950s. Symptomatic of this state of affairs is that they did not engage in any direct debate. This lack of interaction was, as I have argued in previous chapters, a widespread phenomenon in the history of the language/ dialect opposition, in which there were only a few flashes of animated discussion, usually exhibiting the same pattern: one or a few scholars reacted against common practices, like Johann Heinrich Hottinger in the seventeenth century (Chapter 10, Section 10.2) and Hugo Schuchardt in the decades around 1900 (Chapter 18). In the 1950s, structural linguists were challenged from within to involve regional variation in their study of language, thus sparking off a more direct and lively debate. The possibility of a structural dialectology was finally examined, an endeavour spurred not in the least by the Polish-American linguist Uriel Weinreich (1926–67; see Malkiel 1967), son of Max, the Yiddish specialist who, it will be remembered, first recorded the quip ‘A language is a dialect with an army and navy’ (see Chapter 1). Weinreich the younger studied linguistics at Columbia University with André Martinet (see the next section, 19.3). For his PhD dissertation (1951) and the resulting foundation-laying 1953 book Languages in Contact, he conducted fieldwork in linguistically diverse Switzerland, where he also came into contact with the work of local dialectologists, which he valued highly (cf. Weinreich 1954: 397 n.18). His structuralist formation as well as his interest in language-contact phenomena and linguistic diversity led him to ask himself: ‘Is a structural dialectology possible?’ In an influential and foundational paper, bearing this question as its title and published in Word, a journal of which he was co-editor, Weinreich (1954) answered it favourably by providing some schematized examples; he took them mainly from phonology, the domain where structuralism achieved most of its successes. He did, however, emphasize that this linguistic approach should occur in conjunction with an externalist form of dialectology, which sought to explain linguistic differences by correlating them with cultural areas and borders (Sprachlandschaft). This form of dialectology was promoted especially by Swiss linguists in reaction to the idolatry of isoglosses in earlier scholarship. Weinreich wanted, in sum, to bridge the gap between structuralism and dialectology. What was the place of the conceptual pair in his attempts at reconciliation? Apparently, it did not have one, since he renounced the common usage of the term dialect and its many connotations, because it was ‘non-technical’ and defined by language-external, non-structural attributes of time, space, and prestige (Weinreich 1954: 399). He, too, in other words, probably believed in the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

250

   

validity of the witticism associated with his father. Weinreich (1954: 397 n.17) also found it confusing that the term dialect was used in phrases such as ‘IndoEuropean dialects’ and ‘literary dialects’. The eternal problem of dividing dialect continua into separate dialects likewise fostered his suspicion, since it could not be solved by simply unveiling isogloss bundles on a map. Weinreich (1954: 397) concluded: It is evident that no unambiguous concept of dialect could emerge even from this optimistic methodology any more than a society can be exhaustively and uniquely divided into “groups”.

Yet if structuralists and dialectologists were to join forces, Weinreich (1954: 399) did not rule out ‘the formulation of a technical concept of “dialect” as a variety or diasystem with certain explicit defining features’. As long as the term dialect did not receive a technical interpretation, Weinreich preferred to operate with the concept of diasystem in his proposed structural dialectology. In Saussure’s wake, he explained, linguists have thus far concentrated on language as a uniform system, an illusion upheld by their usual object of study: standardized rather than folk language. To integrate the latter into structuralist theory, Weinreich suggested conceiving of a diasystem as follows. For structuralists, language is a ‘unique and closed system whose members are defined by opposition to each other and by their functions with respect to each other, not by anything outside of the system’.⁹ Structuralists should now proceed to constructing systems of a higher level out of the discrete and homogeneous systems that are derived from description and that represent each a unique formal organization of the substance of expression and content . . . A “diasystem” can be constructed by the linguistic analyst out of any two systems which have partial similarities (it is these similarities which make it something different from the mere sum of two systems). But this does not mean that it is always a scientist’s construction only: a “diasystem” is experienced in a very real way by bilingual (including “bidialectal”) speakers and corresponds to what students of language contact have called “merged system”.

Dialectal variation could, in other words, be regarded as an ‘aggregate of systems’ with different layers, which Weinreich preferred to term varieties rather than dialects. Apparently, variety was in his eyes a more neutral term than dialect with its dubious connotations. His proposal has proved to be very successful, since variety is still widely used as a passe-partout term by linguists studying variation,

⁹ All quotes in this paragraph are from Weinreich (1954: 388–90).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

251

even though some scholars have criticized its excessively noncommittal nature (see e.g. Macaulay 2010: 62–3). For Uriel Weinreich, in short, dialectology was the ‘study of diasystems’ and the varieties that constituted those diasystems, apparently now presupposing that they were easy to distinguish from each other. The term dialect, he judged, should be abandoned, at least for now, until a scientific definition was reached. Old habits die hard, however, and traces of the age-old language/dialect distinction remained latently present in his work, including his article on structural dialectology. Weinreich referred, for instance, to bilingual and bidialectal speakers, people mastering two languages or two dialects; this terminology suggests how deeply rooted the conceptual pair had become in the minds of linguists, even those renouncing the term dialect as non-scientific.

19.3 Redefining the conceptual pair: Martinet and Polák Around the same time as Weinreich was developing his thoughts on structural dialectology, some of his colleagues attempted to remedy the non-technicality of dialect and to provide a sound basis for distinguishing language from dialect. These included, most notably, the French linguist André Martinet (1908–99) and the Czech structuralist Václav Polák (1912–81), both of whom Weinreich cited in his structural dialectology paper, although admitting that the latter’s article had arrived too late to be fully acknowledged. Weinreich’s teacher Martinet wrote an article entitled ‘Dialect’, published in the journal Romance Philology in 1954, in which he problematized the concept named in its title, because he had noticed how loosely it was being used by contemporary dialectologists. He complained that Sever Pop (1901–61) in his monumental La dialectologie of 1950 had not touched upon so central a concept as dialect, and that dialectologists from the French school, like Gilliéron, had not addressed this issue either, because they did not believe in the existence of dialects and patois, only in isoglosses. This overall neglect compelled him to reflect on the matter himself. Martinet set out by deploring the proliferation of poorly defined terms to talk about various forms of language, including the terms language and dialect. The latter was, for instance, used to refer to allegedly deviant forms of a standard language as well as to legitimate forms of a language. Linguists had, moreover, confounded speech forms derived from a common ancestor with variations on a later developed standard. Martinet’s solution was to formulate more objective interpretations. Genetic kinship, he maintained, should be the primary criterion in deciding whether a speech form was a language or a dialect: As a rule, genetic kinship would seem to be the decisive criterion: Corsican rates as an Italian dialect because it is genetically close to Tuscan, the dialect on which

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

252

   

the Italian “koinê” is based. Corsican is not a French dialect although its speakers have recourse to French as a standard. (Martinet 1954: 7)

The existence of a koine, believed Martinet, could be a secondary criterion for using the term language. He was, however, fully aware that sociopolitical factors often interfered with this objective criterion of genetic kinship, even among linguists. Basque, for instance, was usually regarded as a distinct language because of the genetic criterion, whereas Catalan was frequently claimed to be a Spanish dialect for political reasons: A form of speech of indefinite status will be denied the title of “language” if it can somehow be labeled a dialect of a national language that enjoys the sanction of a special color on the map.

This mixing up of criteria, Martinet (1954: 8) realized, made the linguist look bad, who ‘tends to be just as hesitant as the layman, because actually both use the same terms, and the linguist has simply never taken the trouble to redefine them scientifically’. The linguist should accept responsibility for the terminology of his discipline, according to Martinet, who challenged his colleagues to think about the language/ dialect distinction rather than simply dismiss or relativize it. This redefining of existing terms was, he claimed, also more economical than inventing new ones. In the case of language and dialect, redefinition was all the more appropriate, since the concepts were already in the beloved format of structuralists, the opposition of linguistic units, even though he did grant that language variation was not a simple binary matter, but a gradational phenomenon. For this reason, he added an extra layer to the conceptual pair, that of patois, referring to the speech form of a small community, typically a village, that had lost its overarching provincial dialect and in consequence greatly differed from the standard language of the country. He further proposed the more neutral term vernacular as a near-synonym of patois, which carried negative connotations. Martinet’s tripartite division was not as new as he wanted his readers to believe. His master Antoine Meillet, for instance, had already suggested a similar hierarchy in a work on the comparative study of Indo-European tongues (Meillet 1903: 31). In the early modern period, too, linguistic diversity had already been charted in terms of language, dialect, and subdialect. Martinet was, however, innovative in teasing out the historical conditions in which dialects and patois emerged. He reacted against the exclusive focus on dialect divergence, central to conceptions of regional variation since the sixteenth century: from a language into dialects, which became in time separate languages, which itself developed dialects again. This supposed standard evolution was too simple a picture of sociolinguistic reality, since there were cases of dialect convergence, too, Martinet pointed out.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

253

Although genuinely concerned over the language/dialect distinction and how to redefine it scientifically, Martinet proposed a rather poor conceptualization himself, since he did not further discuss how his ‘decisive criterion’ of genetic kinship should work exactly. It does seem, however, that he had different degrees of kinship in mind, thus perhaps presuming a variation on the Aristotelian criterion (see Chapter 8, Section 8.1). Yet Martinet did not suggest any cut-off point or other method to convincingly differentiate a dialect from a language. Also in 1954, the Czech structuralist Václav Polák, active in the Prague linguistic circle, published an article wholly devoted to the distinction between language and dialect (Polák 1954). In this contribution, which appeared in the Leuvenbased journal Orbis, founded by Sever Pop, Polák attempted to establish a purely linguistic foundation for the conceptual pair, as if he had anticipated Martinet’s call to responsibility. He, too, considered it indispensable that linguists defined their core concepts; he, too, emphasized the importance of linguistic convergence, especially for dialect studies. Finding fault with earlier linguists such as Graziadio Ascoli, Hermann Paul, Hugo Schuchardt, and Antoine Meillet, whom he criticized for their usage of extra-linguistic criteria in defining the conceptual pair, Polák set out to design a strictly linguistic approach to the matter. In typically structuralist fashion, he interpreted dialect as part of the system of a language—a kind of subsystem—which was why it should be understood in linguistic terms. For Polák, dialects resulted from the confrontation of the linguistic system with neighbouring systems, framing it within the concept of Sprachbund, or union des langues in his own French. Egyptian and Coptic, Polák’s beloved example, belonged to the same linguistic system, but through contact with other Hamitic and Bantu tongues Coptic had acquired certain structural characteristics which Egyptian did not have, but which did not cause language-level differences. The structural frame of a language was, in other words, timeless, Polák believed; he used the terms panchronique and achronique himself (Polák 1954: 96). It was only if one perceived distinct morphosyntactic systems that one could speak of different languages. Variation on the level of phonetics and lexico-semantics only generated different dialects or stylistic variants. Polák, in other words, refashioned the conceptual pair in distinctly structural terms in a way vaguely reminding of the early modern Aristotelian criterion. Related dialects exhibited superficial differences belonging to the phonetic and lexical level, whereas distinct languages had separate morphosyntactic systems, the dominant, most essential, and most stable layer of a language. Polák’s (1954: 93) conscious choice to consider diversity from an exclusively languageinternal perspective implied excluding any social factors whatsoever. It was such isolationist approaches to language that the social turn in language studies aimed to counter in the following decade (see Chapter 21, Section 21.2). Polák did, however, make a distinction between horizontal and vertical varieties of a language system. The former were the dialects, coinciding with regional subsystems

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

254

   

of a language, the latter were stylistic variations on the language system such as jargon, slang, and literary language, situated mainly on the level of the lexicon. A purely linguistic, system-based interpretation of the conceptual pair, argued Polák, had practical use, since it was an objective means whereby to distinguish independent languages from each other, especially in cases where there were not yet boundaries linked to literary languages, as in the Caucasus, or where linguists had been using the terms language and dialect sloppily, as in the case of the Slavic family of tongues. Polák believed that the latter—with the exception of Bulgarian—should be conceived of as dialects of one and the same language because they shared in their core the same morphosyntactic system. Linguists, however, usually considered them to be distinct languages, but this view was shaped by extra-linguistic criteria related to politics, culture, and civilization. They were, one might paraphrase, exalting these Slavic dialects to language status because they had an army, navy, and literature. In his exceptional attempt at tying the language/dialect distinction to specific linguistic properties and accommodating it to structuralist conceptions of language, Polák failed to address the eternal problem of how to distinguish between dialects. His suggestion that they were ‘different stages of the continuous variation of a specific linguistic structure’ indicates, however, that he considered it possible to establish some kind of borders between dialects.¹⁰

19.4 Conclusion The year 1954 was a key turning point in the history of the conceptual pair, since structural linguists from different backgrounds suddenly tried to define it in more objective terms. They did so independently from each other, proposing diverging solutions, and they were motivated to do so by the loose and confused usage of the terms language and especially dialect current in their discipline and among the general public. Especially Weinreich’s paper on structural dialectology met with considerable success. His concept of diasystem, which one might regard as a structuralist reinterpretation of the language/dialect distinction, as well as his annotation method for dialect features became widely known and applied but also criticized (see e.g. Gordon in Boberg et al. 2018: 83–4). Some structural linguists, however, did not consider the flexible usage of the terms language and dialect a problem, but an asset. Most notably, according to the American scholar Charles F. Hockett (1916–2000) in his 1958 handbook A Course in Modern Linguistics, ‘[t]he relative looseness of the two terms is a merit, not a defect, for one can add as many precisely delimited technical terms as one needs,

¹⁰ Polák (1954: 98): ‘étapes diverses de la variation continue de la structure de langue en question’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

255

based on various criteria of similarity between idiolects’.¹¹ Typically for the American structuralist tradition, Hockett stipulated the idiolect, the speech form particular to an individual, as the lowest level of linguistic diversity, which constituted the basis for research and from which higher-level abstractions were to be made. Both language and dialect were collections of similar idiolects, he suggested, with the difference that ‘the degree of similarity of the idiolects in a single dialect is presumed to be greater than that of all the idiolects in the language’. Despite claiming that the looseness of the terms language and dialect was an asset and fostered a flexible approach to linguistic diversity, Hockett did propose a specific criterion ‘stem[ming] from an everyday assumption about language’: mutual intelligibility.

¹¹ All citations in this paragraph are from Hockett (1958: 322).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

20 Mutual intelligibility The number one criterion?

Charles Hockett believed that the intuition among laypeople that speakers of related dialects could understand one another, but those of distinct languages could not, required a more objective formulation. Aware that mutual intelligibility was not a black-and-white matter, he saw two options. Firstly, the linguist could impose a yes-or-no answer, which Hockett (1958: 323) believed to be less artificial than it sounded, because ‘most pairs of idiolects, chosen at random, yield results near one end of the scale or the other’. Secondly, the linguist could quantify the matter and look for statistical solutions, a method he discussed at length, motivated by recent empirical studies in the field. In fact, in the early 1950s, American structuralists had started to devise mutual intelligibility tests and conduct them in actual practice, which Uriel Weinreich (1954: 397–8), too, had already appreciated because they yielded insight into linguistic relationships, even though dividing points resulting from such tests still remained arbitrary. A pioneering contribution was the 1951 paper ‘Methods for determining intelligibility among dialects of natural languages’ by the American anthropological linguist Carl F. Voegelin (1906–86) and the well-known structuralist Zellig Harris (1909–92). Voegelin and Harris (1951: 322) recognized the possibilities of a ‘combination of eliciting and text recording’, which allowed to take old problems in linguistic theory and interpretation and address these to new languages, especially to American Indian languages which have the practical advantages of being numerous, diverse, and available. One of these problems concerns the old and difficult question of dialect versus separate language.

They identified four distinct methods used to answer this ‘old question’. Firstly, anthropologists often simply asked informants whether two varieties were so similar that they could be viewed as dialects, thus leaving the matter to the subjective perception of laypeople. Secondly, linguists could quantify the degree of similarity, ‘count sameness’ in short. The results of these two methods often coincided, they noticed, for instance in the case of the Salish languages, spoken in the Pacific Northwest, where independent studies following these distinct methods resulted in similar conclusions.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

 

257

The second method was pioneered by Morris Swadesh (1909–67), a student of Edward Sapir’s, who wanted to measure linguistic nearness and distance rather than solve the language/dialect puzzle (Swadesh 1950). Swadesh did so by a method he later dubbed ‘lexicostatistics’ (Swadesh 1952; 1955; cf. Chapter 14), which started from a list of 100 words for each language compared. The conceptual pair was, however, already defined in his 1950 paper on Salish tongues, but he distanced himself from the interpretation proposed by attributing it to ‘scholars’ in general: Scholars commonly use dialect to refer to a local variety of a language. Language in turn is used for a continuum of dialects such that each is mutually intelligible with its neighbors. (Swadesh 1950: 163)

In what followed, Swadesh rejected the criterion of mutual intelligibility in particular as too subjective. Swadesh had focused on words, but one might as well include phonetic and other features, Voegelin and Harris (1951: 326) argued; they did so prior to Polák, who independently made a similar suggestion (Chapter 19, Section 19.3). This broader perspective brought them to the third method of ‘structural status’, which implied ascribing hierarchical values to the differences and similarities between language varieties. Yet this structural criterion was not enough, and a fourth method should be introduced: that of mutual intelligibility, testing whether one informant understood another. This method was all the more necessary since speakers of highly similar varieties could experience great difficulties in understanding each other because certain near-identical formal elements could receive very different meanings and functions. In conclusion, Voegelin and Harris were looking for objective criteria grounded in linguistic data, on the one hand, and in intelligibility tests, on the other. By way of check, the results of these methods could be compared to speaker perceptions. Soon linguists from Voegelin and Harris’s network and others put this intelligibility testing to practice in studying Amerindian tongues and other understudied language families, for instance those in New Guinea.¹ Scholars at last started to empirically investigate a criterion which for almost four centuries had been loosely suggested and intuitively assessed, ever since the work of Johannes Goropius Becanus in the second half of the sixteenth century (see Chapter 8, Sections 8.4–5).

¹ See e.g. Hickerson et al. (1952); Pierce (1952). Cf. Casad’s (1974) handbook for mutual intelligibility testing. For New Guinea see Wurm and Laycock (1961).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

258

 

20.1 A criterion making career The criteria launched by anthropologists and structuralists in the early 1950s were highly influential. Mutual intelligibility testing, in particular, was to remain a major guiding principle among linguists who continued to use the conceptual pair, even up to this day, with quantitative methods being increasingly adopted (e.g. Hammarström 2008; Gooskens in Boberg et al. 2018: 206). Most notably, it is the main criterion of Ethnologue, the widely known online language catalogue purporting to list every recognized language—7,111 at the time of writing—in which it follows the ISO 939-3 standard for language names, pioneered by Ethnologue.² The use of the mutual intelligibility criterion has been defended by claiming that the resulting classification usually agrees with the ideas of specialists (Hammarström 2005). Rather infelicitously, the criterion is mixed by Ethnologue with such language-external factors as ‘the existence of a common literature or of a common ethnolinguistic identity’ and ‘the existence of well-established distinct ethnolinguistic identities’. The result is a hybrid framework, which, although flexible, is not as objective as it might be. Another well-known catalogue of the world’s known languages, Glottolog, is more strict, since it takes mutual intelligibility as its sole criterion in distinguishing languages, which leads to the figure of 7,596 languages at the time of writing.³ Problems with the criterion of mutual intelligibility were recognized at an early stage (cf. Hammarström 2008; Gooskens in Boberg et al. 2018: Chapter 11). It is, for instance, possible that a speaker of one variety understands a speaker of another better than vice versa, which linguists call asymmetry. Intelligibility can moreover be incomplete and is, as Charles Hockett already pointed out, not a simple yes-or-no question. Previous knowledge on the part of a speaker or listener can moreover blur the results of intelligibility tests, as can (a lack of) good will and other preconceptions (e.g. Rubin 1992). Acknowledgement of such and other problems with mutual intelligibility does not necessarily imply rejecting the validity of the conceptual pair, however, as is abundantly clear from the recent Handbook of dialectology, in which Charlotte Gooskens states: Reflecting on the objections to intelligibility as the most important criterion for distinguishing between dialects and languages (the Abstand criterion), we see that it is not unproblematic to use this criterion. More research is needed before we will be able to establish when two varieties are so different that they are no longer mutually intelligible, and which linguistic factors play a role. Since there is no universally accepted criterion for distinguishing a language from a dialect, the

² See https://www.ethnologue.com/about; https://iso639-3.sil.org/about/scope#Individual%20lan guages, both last accessed 10 January 2020. ³ See https://glottolog.org/glottolog/glottologinformation, last accessed 10 January 2020.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

259

examples provided in the rest of this chapter are from language varieties that are traditionally referred to as dialects as well as closely related varieties that are mostly referred to as languages. (Gooskens in Boberg et al. 2018: 208)

Despite Gooskens’s optimism that a mutual intelligibility threshold might be discovered, it is hard to see how this will work, especially considering the numerous problems involving the criterion, not in the least because it is often a question of gradation and the countless tests conducted ever since the 1950s have not yet revealed such a threshold. What is more, the fact that mutual intelligibility assessments can be distorted owing to language-external factors, such as a speaker’s reluctance to understand the speech of a hostile community, makes its validity even more questionable. Finally, on a more fundamental level, is mutual intelligibility not an effect of underlying similarities in language itself and lack thereof not a result of differences? Would it not be more opportune to study the phenomenon of linguistic diversity at its core, in language itself?

20.2 Measuring linguistic distance Such questions have led some linguists to continue rather on Morris Swadesh’s path and measure the degree of differences between speech forms (‘count sameness’). They consider this method more objective as it starts from language itself and not from a secondary effect, even though the results of both methods, measuring linguistic distance and testing mutual intelligibility, often coincide, studies find (e.g. Biggs 1957: 61). The secondary nature of mutual intelligibility has been argued by the American linguist Frederick B. Agard (1907–93), who formulated in 1971 nine postulates for distinguishing a language from a dialect. In an account that is principally structuralist but assumes the generative split between deep and surface structure, Agard (1971: 7) rejected the mutual intelligibility criterion as ‘a matter of probability’. Instead, he attached great importance to lexical correspondences and, in second place, phonemic inventory and grammatical phenomena. Agard’s main argument was that structural criteria should be sought to distinguish between language and dialect, that is to say, in speech itself and not, for instance, in its speakers. His concern with the conceptual pair was fuelled by the loose usage of laypeople, sociologists, and sociolinguists and ‘their impressionistic criteria’ (Agard 1971: 6). Language-external circumstances, motivated by political, literary, or social factors, should be of no importance for the descriptive linguist. In ‘a scientific distinction’, the focus should be on actual language and—note the structuralist terminology—its diasystems, not on secondary circumstances, which included mutual intelligibility:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

 

260

To sum up: Mutual intelligibility either across languages or within languages is of itself unmeasurable but can nevertheless be a function of measurable phonological, grammatical and lexical factors. (Agard 1971: 20)

Agard’s postulates involved a hierarchical ordering of linguistic criteria to differentiate language from dialect. It was indeed more common among structuralist linguists to propose weighing the different criteria in determining the precise status of a speech form. Usually, however, phonemic similarities or differences were then invoked as the prime criterion (Gordon in Boberg et al. 2018: 83–4). Agard did not propose measuring linguistic distance, as Swadesh had done, but shared with him the desire to find stricter linguistic criteria. Perhaps Agard considered it an impossible or undesirable task to quantify his language/dialect framework. In recent times, however, the availability of advanced software has made it easier to apply statistical methods to linguistic issues in general and to the problematic conceptual pair in particular. One of the latest and highly technical contributions on the language/dialect opposition takes Swadesh’s approach of measuring linguistic distance as its starting point. The paper, published in 2018, relies on a statistical analysis of basic vocabulary items from twenty-two language groups.⁴ Its presumed main author, Søren Wichmann, starting from the belief that the conceptual pair is still useful and should be redefined scientifically, wants to show that lexical, and hence linguistic, distance tends to be ‘bimodally distributed’; this means that there are usually two peaks around which the results tend to situate themselves, which might suggest that there is something like a language cluster and a dialect cluster (see Figure 20.1). t

g/t

Figure 20.1 Example of a typical bimodal distribution on a scale

Results in-between are rare and, hence, constitute interesting problems for scholars. Wichmann (2016: 5) feels confident enough to advance a ‘universal cut-off ’ point which may be used to distinguish pairs of dialects from pairs of languages. This distinguishing criterion is easily applied and it was arrived at by an entirely ⁴ Ran and Wichmann (2018), the published version in Mandarin. Wichmann (2016) is the English draft version I quote here.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

. 

261

objective procedure which can be both replicated and revised on an evolving dataset.

Based on his study of twenty-two language groups, the cut-off point coincides according to Wichmann with a so-called Levenshtein distance normalized (LDN) value of 0.49. Levenshtein distance, also known as edit distance, refers in linguistics to the number of changes needed to arrive from a word in one language to the word denoting the same concept in another. The qualification ‘normalized’ designates that the Levenshtein distance for a word pair has been ‘divided by the length of the longer of the two words’ (Wichmann 2016: 2). Establishing the cut-off point allows, Wichmann proceeds, the determination of whether the language count in catalogues such as Ethnologue is correct. His conclusion is that ‘Ethnologue tends to overdifferentiate, so the number of languages counted in this catalogue would be too high’ (Wichmann 2016: 6). Wichmann leaves it to the reader to conclude on the basis of his proposed cutoff point that, for instance, Indonesian–Malaysian, Bosnian–Croatian, Hindi–Urdu, and Standard German–Swiss German are dialect pairs, whereas Danish–Swedish, Dongshan Chinese–Fuzhou Chinese, Cairo Arabic–Moroccan Arabic, and Catalan–Spanish are pairs of related languages (see Wichmann 2016: 6–7, esp. Table 2). The discussion is rounded off by an interesting suggestion which reintegrates mutual intelligibility into the debate on the language/dialect distinction as a key contextual factor rather than a criterion: what is it about the way that speakers interact which allows us to distinguish languages and dialects? A possible explanation is that there is a threshold of mutual intelligibility where language varieties will be influencing one another if they are below the threshold but will cease to influence one another if they are above it. If mutual intelligibility between variety A and B is impeded completely speakers may take recourse to just the more prestigious of the two, if not some third language, leaving A and B to drift apart more rapidly than would the case if both A and B were used for communication between the two groups. (Wichmann 2016: 7)

By also including the values of a variant of the LDN, Wichmann likewise calculates the period of time required to form new languages, which, he argues, coincides with a window of roughly 934 to 1195 years.

20.3 Conclusion In conclusion, there might be, after all, an empirical measure to distinguish between languages and dialects, even though it seems that further exploration

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

262

 

and fine-tuning of this statistical method is required, for instance, by incorporating non-lexical evidence, too, and by according relative weights to individual features, as Carl Voegelin and Zellig Harris already suggested over sixty years ago and Frederick Agard likewise proposed in his postulates. If in such more comprehensive studies the data still tend to be bimodally distributed, it will be safe to conclude that the language/dialect pair as a metalinguistic abstraction is backed by the majority of the linguistic evidence. Still, there will always exist less numerous in-between situations, which, however, should perhaps not be viewed as invalidating the conceptual distinction but should be studied as intriguing exceptions in a transition zone between the language and the dialect cluster.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

21 Between two extremes Generative and sociolinguistic interpretations

The heyday of structuralism was already coming to an end as Uriel Weinreich and his colleagues were attempting to initiate a dialectological turn in mainstream language studies and narrow the gap with traditional dialectology. In 1957, the paradigm-changing Syntactic Structures by one of Zellig Harris’s students, Noam Chomsky, was published, initiating a tradition of linguistic research that usually left little room for research on dialectal variation. Still, the problem was not entirely neglected, and some generativists tried to frame it within their new theory. Was there any room for the language/dialect distinction? And if so, how was it incorporated into generative theory?

21.1 Generative grammar: no country for old dialects? Early generative linguists conceived of language as consisting of two layers: the underlying set of rules, for which man had an innate capacity, and the way in which they were transformed on the surface into spoken language.¹ This division between deep and surface structure constituted the core of generativetransformational approaches to language at the outset. As generativists focused on discovering the principles of universal grammar, they tried to reduce language and its enormous variation to universally applicable ground rules. Overcoming variation was key, and language differences, especially small-scale ones such as those among dialects, were principally treated as minor variations. Revealing is what Brian Newton (1972: 5) wrote at the outset of his monograph, The generative interpretation of dialect: A study of Modern Greek phonology, published in 1972: Because dialects arise from a more or less uniform language it is possible to show that they can, for the most part, be described in terms of a common set of underlying forms; variation is introduced by the phonological processes on these forms. ¹ On the history of generative grammar see e.g. Freidin (2012). See also Hinskens in Boberg et al. (2018: Chapter 5) for generativism in relation to dialectology.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

264

  

According to Newton (1972: 3), this dialectal variation could be due either to a difference in the order of some underlying rules in two given dialects, or in the failure of a dialect to undergo a certain rule, which he linked to historical developments. In doing so, Newton was following the generativist interpretation of dialects current since Morris Halle’s (1923–2018) foundational 1962 article ‘Phonology in generative grammar’ (Halle 1962; cf. Campbell 1972). Like Polák’s structuralist conception of the language/dialect distinction, generativist interpretations remind somewhat of what I have called the Aristotelian criterion and might well be seen as a more sophisticated variation on this theme. Languages differed from each other considerably in grammatical structures, whereas related dialects showed only limited differences (cf. Löffler 1974: 4). This idea was expressed very transparently by the generative linguist Sol Saporta (1925–2008) in 1965: the grammatical description of a given dialect may be converted into an adequate description of a related dialect by the addition, deletion, or reordering of a relatively small number of rules. Indeed, it is tempting to propose that the degree of difference between dialects is nothing more than a function of the number and type of such changes. (Saporta 1965: 219; for its exemplarity see Campbell 1972: 291)

This grammar and its variations in the rules between underlying and surface forms were, however, usually devoid of any psychological reality, Lyle Campbell (1972) argued in his critical assessment of generative dialectology, which he entitled in Weinreichian fashion ‘Is a generative dialectology possible?’ Apart from the view that dialects were variations on one single grammar, there was one other major strand in early generative dialectology. This alternative approach assumed that grammars of individual dialects should be composed and subsequently compared, which according to Campbell (1972: 294) ‘fails in its overemphasis of the differences at the expense of the similarities, an expense which means that this approach fails to resolve the conflict between structure and variation’. There was no longer a basis of comparison for the two dialects, when two separate grammars were composed which from the point of view of deep structure were as dissimilar as those of Chinese and Quechua; in both cases, the grammars ‘will have different underlying forms with different rules and different surface forms’ (Campbell 1972: 293). The focus of generative linguists has usually not been on the question of how one might be able to distinguish languages from dialects but on how to integrate this type of variation into their theory and reconcile it with the transformational framework. This integration has been an issue in subsequent updates and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      ? 265 revisions of the theory, in most of which Chomsky has played a prominent role.² In the 1980s, the Principles and Parameters theory was formulated, which conceived of universal grammar as a system of very general abstract principles. Different choices of largely syntactic parameters allowed a limited amount of linguistic variation. If these choices concerned major parameters, dubbed macroparametric variation, this resulted in language-level differences, whereas micro-parametric variation was responsible for differences across related dialects. In the early 1990s, the Minimalist programme was proposed as the new major path of generative linguistic study, in which the strictly syntactic approach to variation was abandoned and an insistence on rule and derivation supplanted the traditional division between deep and surface structure. Generativists started to consider now other non-syntactic factors in integrating variation into their theory, including cognition, physiology, and society. In particular, linguists such as Sjef Barbiers have tried to narrow the gap between generative theory and sociolinguistics, much as Weinreich tried to do in the 1950s for structuralist theory and dialectology. In the later 1990s, Optimality Theory abandoned the rule and derivation framework, working instead with universal constraints that work on languages and seek to produce optimal output forms. As the importance of these constraints is different for each language, and as they can be superseded by rivalling constraints, there is variation. Constraint ranking accounts for differences between discrete linguistic systems—i.e. language-level variation—whereas minor variations on this ranking are usually held responsible for dialect-level differences. In conclusion, generative grammar has greatly struggled to incorporate lowerlevel variation into their various theoretical frameworks. The focus has been on the ideal speaker and on the transition from underlying presentations to output forms; language is seen as a closed system, most likely because well-defined standard languages and not dialects have been the central research object— especially English. It would be, however, too rash to claim that the language/ dialect distinction was not a concern for generativists at all. Still, the opposition usually does not figure prominently in their accounts, and one has to look for traces of it in their models of language. It seems that dialect-level variation has by and large been presented in generative approaches as the result of minor changes in rules, parameters, or constraints; language-level differences, in turn, are the product of major divergences in them.

² See Hinskens in Boberg et al. (2018: Chapter 5, especially 89–91), which guides my discussion here. Cf. also Shapiro and Schiffman (1981: 50–1).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

266

  

21.2 Variables over systems The shying away from superficial forms of language variation common in structuralist and generativist approaches set off a strong reaction in the 1960s. Language is a social phenomenon, and bears indelible traces of this identity, many linguists increasingly realized. One scholar in particular contributed greatly to promoting the study of language in its societal context: Uriel Weinreich’s student William Labov. After a brief career as an industrial chemist, Labov switched to linguistics in the 1960s, graduating at Columbia University. His MA thesis on diphthong change in the English spoken on the island of Martha’s Vineyard of 1963 became the first landmark in what is known as sociolinguistics, soon followed by his warehouse study, in which he charted social variation in the pronunciation of [r] among New Yorkers (see Labov 1963; 1972). Labov succeeded in initiating and establishing sociolinguistics as a subfield of institutionalized language studies and bringing it closer to general linguistics. The great appeal of his approach consists in the fact that it makes the correlation between language and society very palpable; it focuses on language in actual usage and tries to discover how linguistic variables are motivated by social circumstances. Even though sociolinguistics has greatly increased our understanding of language variation, scholars active in this strand have generally passed over theoretical questions about what a dialect is and what distinguishes it from a language. Linguistic variables have taken centre stage (cf. Macaulay 2010: 64). As a consequence, sociolinguists tend to follow Hockett’s judgement that the flexibility of terms such as language and dialect is an asset rather than a defect. It allows linguists to speak about language in its many appearances, ranging from distinct languages such as English, national varieties such as American English, varieties of regions as for instance Pacific Northwest English, or varieties of a town or city, including Pittsburgh English. The Scottish sociolinguist Ronald K. S. Macaulay (2010: 65–6), for instance, does not consider this flexible usage problematic and even finds it preferable that such labels match the ideas of lay speakers: Does it make sense to consider such diverse phenomena under the single label of “dialects”? To my mind, yes. If the term dialect is to have any coherent interpretation it must refer to a form of speech based on geographical location, preferably one that the speakers themselves would most likely admit to using . . . The goal of dialectology should be to tabulate those features of speech that characterize a particular locality regardless of its size.

Not unsurprisingly, sociolinguists mostly focus on variation within a language and are therefore less concerned with the question of how to distinguish between language and dialect. In particular, variation within cities has been a prominent

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.   

267

research topic among sociolinguists, since this kind of diversity had been largely neglected by traditional dialectologists, whose attention was monopolized by rural dialects (see e.g. Chambers and Trudgill 1998: Chapter 4). Sociolinguists’ interest in the correlation of linguistic variables with social factors has moreover led them to foreground extra-linguistic factors when speaking about dialect in the sense of ‘regional form of speech’. Various sociocultural, political, and ethnic circumstances are invoked to identify a dialect and a dialect area. Macaulay (2010: 66), for instance, agrees with the Austrian-American linguist Hans Kurath (1972: 76) that ‘[w]hat identifies a locality as a dialect area will depend . . . on “a combination of socio-cultural factors” such as political domains or settlement areas’. Yet no strict boundaries between dialect areas can be observed, as Macaulay (2010: 66 n.2), Labov (2010: 165–6), and many other sociolinguists acknowledge. The term dialect is frequently also used, however, to refer to varieties not bound to a specific locality, as, for instance, the ethnically defined African American Vernacular English (Macaulay 2010: 66–7). Still, sociolinguists consider variation in a language from a linguistic point of view, too. Labov (1973), for instance, has expressed the idea that dialects of one language share the same grammar or, at least, should be encompassed within one single grammar. As such, he accepts assumptions about the existence of a so-called panlectal or polylectal grammar advanced by generativists in the 1960s and early 1970s. This idea has subsequently been doubted by sociolinguists such as Peter Trudgill (1983: Chapter 1). In summary, it seems that among many sociolinguists a rather flexible usage of the term dialect is current, tied primarily to extra-linguistic factors. This externalist perspective is related to the main method of this branch of linguistics: the relating of language variation to language-external attributes. As a result, the language/dialect distinction is usually not problematized extensively by sociolinguists. J. K. Chambers and Peter Trudgill, for instance, treat the matter only very briefly at the outset of their handbook on dialectology, of which the following paragraph is the most telling: It seems, then, that while the criterion of mutual intelligibility may have some relevance, it is not especially useful in helping us to decide what is and is not a language. In fact, our discussion of the Scandinavian languages and German suggests that (unless we want to change radically our everyday assumptions about what a language is) we have to recognise that, paradoxically enough, a “language” is not a particularly linguistic notion at all. Linguistic features obviously come into it, but it is clear that we consider Norwegian, Swedish, Danish and German to be single languages for reasons that are as much political, geographical, historical, sociological and cultural as linguistic. It is of course relevant that all three Scandinavian languages have distinct, codified, standardised forms, with their own orthographies, grammar books, and literatures; that they correspond to three separate nation states; and that their speakers consider

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

268

  

that they speak different languages. The term “language”, then, is from a linguistic point of view a relatively nontechnical term. (Chambers and Trudgill 1998: 4–5)

For this reason, Chambers and Trudgill (1998) abandon the term language and choose to work with the term variety, because it is more neutral and has the advantage of being flexible and widely applicable. It can be ‘appl[ied] to any particular kind of language which we wish, for some purpose, to consider as a single entity’. The term dialect is retained to refer to ‘varieties which are grammatically (and perhaps lexically) as well as phonologically different from other varieties’, whereas accent designates ‘the way in which a speaker pronounces, and therefore refers to a variety which is phonetically and/or phonologically different from other varieties’ (Chambers and Trudgill 1998: 5). Different classes of variation constitute here the vague and hesitatingly proposed criterion to distinguish a dialect from a mere accent.

21.3 Destructive dismissal Chambers and Trudgill (1998) thus propose to abandon the conceptual pair as a technical distinction and to use a more neutral term instead, variety, in combination with dialect and accent. Their dismissal of the language/dialect opposition is still gentle, but it is a dismissal nonetheless. Other linguists have more pronounced condemnations in store, as can be easily gathered from a recent paper by Alexander Maxwell (2018). Untangling the often baffling uses and misuses of the Weinreich witticism in modern linguistics, Maxwell (2018: 272) indicates how some scholars have ‘exploit[ed] the witticism’s humor to denigrate the languagedialect dichotomy and any who invoke it’. Among those cited by Maxwell, Noam Chomsky’s (Figure 21.1) voice is particularly powerful. Where I have had to search for traces of the language/dialect distinction in mainstream generative theory earlier in this chapter, it now becomes clear why. To Chomsky, apparently convinced that it is easy to distinguish languages from each other and from dialects, the conceptual pair is non-scientific. It therefore has no relevance for his theory of language. In his conversations with the French poetess and linguist Mitsou Ronat (1946–84) published in 1977, he explained that ‘in my view, the concept of language is not a linguistic concept’.³ It is worthwhile to quote here at length his objection to the conceptual pair and, more generally, to the externalist approach current in the then young branch of sociolinguistics: ³ Chomsky (1977: 195): ‘A mon avis, la notion de langue n’est pas une notion linguistique’. I quote from Ronat’s French version of Chomsky’s original ‘American’, as the English rendering of the French translation (Chomsky 1979) departs from it considerably.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.  

269

Figure 21.1 Noam Chomsky in 1977 Source: Wikimedia Commons / Nationaal Archief. CC BY-SA 3.0

What is Chinese? Something yellow on a map. The languages which are called Chinese are as diverse as the Romance languages. A Chinese from the north does not understand a Chinese from the south, etc. It is political reasons which define Chinese. Theoretically, nothing allows one to assert that Chinese is one language and that the Romance languages are not another. And still, no one says that Italian and French are one single language. No linguistic argument would support that. I would say more: what makes French a language? I suppose that fifty years ago neighbouring villages spoke dialects of French that were sufficiently different to render intelligibility poor indeed. What, then, is a language? It is said jokingly that a language is something that has an army and a navy. It is not a linguistic concept, nor a linguistic definition. As far as other sociolinguistic questions are concerned, I think that it is justified to ask oneself whether they have been posed in a way which permits answers of some depth . . . You see, questions about languages are always tied to those of power . . . Every one of us speaks a certain number of [idealized linguistic] systems, mixing them in an amusing way. Since our experience is different, our system mixtures are different.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

270

  

Yet I do not believe that outside the reality of these systems there exists a dialect or language reality.⁴

Chomsky continued to repeat this harsh judgement of sociolinguistics in other writings of his in the 1980s and 1990s (see e.g. Chomsky 1986: 15; 1992: 216; 1997: 125). In fact, he has used the Weinreich quip so very often in a disparaging manner that several subsequent scholars have wrongly attributed the saying to the generativist (Maxwell 2018: 267). Misusers of the witticism not only want to show the involvement of nonlinguistic aspects in language studies in a bad light, but they also dismiss the language/dialect distinction as having no value whatsoever. Maxwell has identified another trend among users of the witticism apart from the dismissive one. Some linguists recognize that their field should engage with factors outside language in studying this phenomenon but employ the witticism to keep themselves from doing so. The phrase serves as a disclaimer: ‘there are political factors, but these do not concern me here’. It is striking how often linguists present the conceptual pair as an unworkable instrument because of its political connotations but nonetheless persist in using it, sometimes even only a few sentences apart (Maxwell 2018: 272–5, with striking examples). This unease with the conceptual pair results, Maxwell (2018: 276–8) argues, from the fact that linguists are not trained to conduct political analysis but do feel the need to clearly categorize linguistic diversity.

21.4 Supplementing the conceptual pair Numerous sociolinguists, too, have questioned the validity of the conceptual pair, but instead of simply dismissing the problem they have come up with more constructive solutions. If the language/dialect distinction did not work in and of itself, supplementing it should do the trick. One of the most famous and ⁴ Chomsky (1977: 195–7): ‘Qu’est-ce que le chinois? Quelque chose de jaune sur une carte. Les langues qu’on appelle le chinois sont aussi diverses que les langues romanes. Un chinois du nord ne comprend pas un chinois du sud, etc. Ce sont des raisons politiques qui définissent le chinois. Théoriquement, rien ne permet d’affirmer que le chinois est une langue et que les langues romanes n’en sont pas une autre. Et pourtant, personne ne dit que l’italien et le français sont une même langue. Aucun argument linguistique ne soutiendrait cela. Je dirais plus : qu’est-ce qui fait du français une langue? Je suppose qu’il y a cinquante ans, des villages voisins parlaient des dialectes du français suffisamment différents pour que l’intelligibilité ait été tout à fait faible. Donc, qu’est-ce qu’une langue? On dit par plaisanterie qu’une langue, c’est ce qui a une armée et une marine. Ce n’est pas un concept linguistique, ni une définition linguistique. Quant aux autres questions de sociolinguistique, je crois qu’il est juste de se demander si elles ont été posées d’une manière permettant des réponses de profondeur quelconque . . . Voyez-vous, les questions de langues sont toujours liées à celles du pouvoir . . . Chacun de nous parle un certain nombre de ces systèmes, en les mélangeant de manière amusante. Parce que notre expérience est différente, nos mélanges de systèmes sont différents. Mais je ne crois pas qu’en dehors de la réalité de ces systèmes, il existe une réalité dialecte, ou langue’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

271

influential supplementations of the traditional pair has been proposed by Heinz Kloss (1904–87), a German specialist in minority languages who in 1952 devised a framework separating linguistic from external factors, into which he integrated the conceptual pair.⁵ Kloss distinguished between language by distance (Abstandsprache) and language by elaboration (Ausbausprache). The former referred to speech forms which were so distant from one another in linguistic terms that they could not be considered related dialects. Kloss (1967: 30) was aware that this linguistic distance criterion was a tricky affair but ‘for the moment . . . assume[d] that linguists are in a position to apply final, reliable, and uniform criteria’, since he was himself more interested in the non-linguistic, sociological side of things. He did, however, venture some thoughts on the matter (Kloss 1978: 63–6). Substantial differences in phonology and grammar were required to brand two speech forms separate languages, but the principal defining characteristic was the lexicon, according to Kloss. Related speech forms were to be viewed as dialects if they shared basic vocabulary and if, as a result, speakers of these varieties were able to understand one another. Kloss realized, however, that ever since Swadesh had started computing basic vocabulary likeness (see Chapter 20), linguists disagreed about the cut-off point between language and dialect. Swadesh suggested on the basis of a diachronic study of English vocabulary that, from a synchronic perspective, 81% of the basic vocabulary should be the same in order to speak of related dialects, whereas later scholars proposed a lower percentage. Measuring mutual intelligibility was likewise not without problems, and Kloss cited the traditional objections to it, while pointing out that it was a secondary effect of linguistic similarities, principally in the lexicon but also in grammar. A language by elaboration, an ausbau language, on the other hand, ‘refers to languages which have deliberately been reshaped so as to become vehicles of variegated literary expression’ (Kloss 1967: 30). Kloss’s main interest was glottopolitical, and his work thus concerned these deliberately planned ausbau languages and their relationship to spoken language within society, usually dialects or closely related tongues. Ausbau languages are often superposed on related dialects or closely related tongues; they ‘roof ’ them, in Kloss’s now widely accepted terminology. As a consequence, speakers of a language roofed by very dissimilar ausbau languages might accept dialect status for their native tongue because of the prestige attached to the roofing ausbau language. Some speech forms such as Corsican, in turn, lack a roofing ausbau language closely related to it—Corsica belongs to France politically but its speech is more closely related to Italian—and can therefore be viewed as roofless dialects (Kloss 1978: 60–3).

⁵ Kloss (1952; 1978). I quote from the updated and enlarged 1978 edition as well as from Kloss (1967), an English résumé article.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

272

  

Kloss’s separation of abstand from ausbau was triggered by his observation that linguists, sociologists, and laypeople espoused different ideas about how to distinguish between language and dialect. He found this discrepancy obvious and believed that linguists should not necessarily agree with laypeople’s views, which were inspired by language-external circumstances (Kloss 1978: 24, 27). Even though he focused on the ausbau aspect, he did find it useful to include the abstand perspective in devising a typology of sociolinguistic situations involving ausbau languages. His findings even led him to suggest that not only mutual intelligibility should be involved in determining the relationship between speech forms but also ‘mutual recognizability’. Speakers may not be able to understand a related language but may feel a close connection with it because they recognize features of their own speech in it (Kloss 1967: 36–7 and the interesting n.8). What is more, his conceptual innovation was also intended to facilitate language counting, where, however, he seems to have intermingled the originally structuralist concept of diasystem with language-external factors such as the possession of a literature and usage in schools (Kloss 1978: 31–3). Kloss remained realistic, however: We may not, however, be mistaken about the fact that even with the help of ever so many criteria a clear decision will not always be possible about the question of whether we are dealing with a dialect, an autonomous language, or a variety (style) of an autonomous language.⁶ It is almost impossible to draw the boundaries between language and dialect according to uniform viewpoints. The researchers’ measures are divergent and can change. What is more, non-scientific, purely historical circumstances cannot be entirely overlooked. For instance, the boundaries between language and dialect are traditionally drawn more broadly within the German language space than in the Slavic area.⁷

The conceptual pair was, however, not goalless, Kloss maintained, and one should conceptualize language and dialect not as two entities divided by one line but as separated by a broader transition zone which Kloss (1978: 35) termed ‘border hem’ (Grenzsaum).

⁶ Kloss (1978: 33): ‘Wir dürfen uns freilich nicht darüber täuschen, das auch mit Hilfe noch so vieler Bestimmungsmerkmale eine klare Entscheidung darüber, ob wir es mit einer Mundart zu tun haben oder mit einer selbständigen Sprache oder mit einer Spielart (Schreibart) einer selbständigen Sprache, nicht immer möglich sein wird’. ⁷ Kloss (1978: 34): ‘ . . . es fast unmöglich ist, die Grenze zwischen Sprache und Dialekt nach einheitlichen Gesichtspunkten zu ziehen. Die Maßstäbe der Forscher sind verschieden und können sich ändern. Auch können außerwissenschaftliche, rein historische Umstände nicht ganz außer acht gelassen werden. Z.B. wird innerhalb des deutschen Sprachraumes die Grenze zwischen Sprache und Mundart traditionell weiter gezogen als im slawischen Bereich’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

273

By supplementing the language/dialect distinction with sociological concepts, Kloss had a major impact on sociolinguistics. Yet he struggled with the conceptual pair himself, since he intermingled linguistic distance with language development and planning in assessing the language/dialect status of a speech form, an incoherence of which he was perfectly aware. He moreover acknowledged that historical circumstances could interfere with conceptualizations of language and dialect. Kloss’s contribution was, in conclusion, situated in acknowledging the importance of non-linguistic factors in the debate on the language/dialect pair and in introducing new concepts based on the age-old distinction to capture the sociological side of language; he did not want to provide a final answer to whether or how languages and dialects were different in linguistic terms, even though he evinced a genuine interest in issues of linguistic distance, too.

21.5 By way of conclusion: a quest for alternatives Whereas Kloss supplemented the language/dialect pair, other scholars have recently dismissed it altogether and supplanted it with an entirely different conception of language. William A. Kretzschmar, Jr, for instance, has made a strong case in favour of abandoning the conceptual distinction (see especially Kretzschmar 2009). Language and dialect are ‘observational artifacts’, according to Kretzschmar, who ties his views to those of earlier dialectologists and Saussure; these linguists maintained that empirical study did not allow supposition of the existence of autonomous and clearly delimited languages and dialects (see Chapter 18, especially Section 18.5). The conceptual pair is a construct of what he dubs ‘the linguistics of linguistic structure’, for which he proposes ‘the linguistics of speech’ as the most viable alternative: Taken together, the basic elements of speech correspond to what has been called a “complex system” in sciences ranging from ecology and economics to physics. Order emerges from such systems by means of self-organization, but the order that arises from speech is not the same as what linguists study under the rubric of linguistic structure. (Kretzschmar 2009: 4)

To Kretzschmar, language and dialect are remnants of a persisting tradition of language studies starting from closed systems with their own internal structure. Reasoning often occurs circularly in this tradition, since languages and dialects are first presupposed; then, their features are studied, which results in confirming the languages and dialects already presupposed (cf. Kretzschmar in Boberg et al. 2018: 59–60). Kretzschmar’s alternative has no place for the centuries-old conceptual pair. Overall, the balance among present-day linguists seems to be shifting towards

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

274

  

dismissing it as a useful tool to chart language diversity for various reasons. Kretzschmar, like Hugo Schuchardt, Gaston Paris, Jules Gilliéron, and Paul Meyer, finds fault with the closed system approach and regards the conceptual pair as an inadequate abstraction of linguistic realities. Other linguists, like Jacques Van Keymeulen, dismiss it as a ‘sociological notion’ (see Boberg et al. 2018: 39). Are we moving towards a final abandonment of the conceptual pair in mainstream linguistics? The demise of the language/dialect distinction might well be imminent in academia, but it will definitely not part without leaving a trace.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

22 A gentle goodbye? Dialect stripped for parts

Discontented with, and embarrassed by, the undesired externalist connotations of the conceptual pair, many modern linguists harbour suspicions about it, if they have not chosen to abandon it altogether, and the term dialect with it. Yet many compositions and derivations based on dialect are still in vogue. Some of these, such as subdialect and dialectology, are centuries old (see Chapter 8, Section 8.3; Chapter 16, Section 16.1), but most were coined only in recent history, including adjectives such as dialectal. More interestingly, the dia- element has made a career in terms such as diatopic, diastratic, and diaphasic, a derivation technique popularized by the work of the extravagant Romanist Eugenio Coseriu (1921–2002), who wanted to capture the different dimensions of linguistic diversity with this terminology: geography, society, and register (see e.g. Coseriu 1998). The element -lect, in turn, has proved even more popular to refer to speech varieties and linguistic features of divergent kinds. Like variety, lect per se has been used as a neutral passe-partout term for any homogeneously conceived form of speech (e.g. Fasold 1975: 29; Kloss 1978: 23; Haugen in Alinei 1980: 40). In order to emphasize the different dimensions of linguistic diversity, lect has received various specifiers, such as socio- and regio- to refer to a variety of a certain social class and a certain region, respectively. Especially sociolect has been common since its appearance in the 1960s.¹ Regiolect was introduced to avoid the negative connotations of dialect and to highlight the geographical dimension of variation. A somewhat older but today lesser-known coinage in this strand is communalect, designating the speech of a specific community (e.g. Hickerson et al. 1952: 1). The most successful derivation, however, has been idiolect, referring to the variety of speech of an individual; this term became popular in the late 1940s, owing to its central position in American structuralism, where it was the basic unit of linguistic analysis. In about 1970, scholars introduced ethnolect and technolect, stressing the racial and professional dimensions to variation, respectively. Technolect thus served as an alternative to jargon. Apart from ethnolects, there are also multiethnolects, urban youth languages which have their origin in different countries (e.g. Auer in Boberg et al. 2018: 173). A couple of years after ethnolect and technolect, genderlect ¹ Frequency estimates are based on searches in English works by means of the Google Books Ngram Viewer https://books.google.com/ngrams, last accessed 13 January 2020.

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

276

   ?    

was coined as a label for linguistic features related to different sexes or genders. It will be noticed that the appearance of terms like ethnolect and genderlect nicely dovetails with the emergence of postcolonial and gender studies, no doubt the result of a cross-fertilization with linguistics. More recent years have seen the appearance of natiolect, which emphasizes the link between speech variation and the dimension of the nation-state. From most of these -lect formations, adjectives in -lectal have been derived, mirroring the pair dialect and dialectal. Suffice to say, -lect has been a productive lexeme in recent decades in referring to all kinds of language varieties, very often in the closed-system sense William Kreztschmar has criticized. In the emergent field of lectometry, lect is more and more interpreted not as a clearly delimited linguistic entity but as a ‘collection of linguistic features that can vary along any extra-linguistic contextual dimension in the broadest sense possible’, covering regional, stylistic, social, and perceptual variables.² This field has come into being as a result of the broadening of the scope of dialectometry, popularized by Jean Séguy and Hans Goebl from the 1970s onwards and aimed at quantifying dialect similarities and their areal mapping (see e.g. Séguy 1973 and Goebl 1982 for foundational studies). Lectometry is currently especially popular among sociolinguists and scholars active in the cognitive strand of language studies, which investigates the ways in which language interacts with the human mind. Key cognitive concepts such as salience and prototypicality are often put into practice in lectometric studies. It has, for instance, been argued that ‘lects are not internally homogeneous, but are rather characterized by centrality effects’; as a result, certain features of a lect are more salient than others for language users (Geeraerts and Kristiansen 2015: 376). The language/dialect distinction, however, is usually not problematized within this relatively young field of enquiry, in which the term dialect is present only at the margins. This lack of interest is most likely due to the feature-focused approach to language prominent in such research. Other derivations from dialect have also proved popular, but they are too numerous to discuss in detail here. Compositions with dialect are likewise widespread and include dialect area, dialect chain, dialect continuum, dialect convergence and divergence, dialect levelling, dialect loss, dialect shift, and dialect term or word. These are frequently also used by scholars renouncing dialect as an adequate metalinguistic term. A striking example is James N. Adams, a specialist in the Latin language and its diversification, who stated at the outset of his 2007 monograph The regional diversification of Latin 200 – 600: “Dialect” in popular usage implies a distinct type of speech tied to a precise locality, whereas the reality may be far more complex, with the boundaries, ² The quote is taken from the description of a recent panel on lectometry at www.ling.arts.kuleuven. be/qlvl/lectometrypanel/, last accessed 13 January 2020.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

   ?    

277

regional and linguistic, far more blurred. There is something to be said for terms vaguer than “dialect”, such as “variations, variety, diversity”. In using such terms throughout this book I am acknowledging that we could never from the Latin record determine the full range of local usages of any precisely demarcated region, even if precisely demarcated linguistic regions ever existed. The point will also be made below . . . that there is a difference between “regionalised standard language”, an entity that may be identified in Latin, and a “primary dialect”, something impossible to find in Latin. The lack of evidence for the latter is a good reason for avoiding the term “dialect”. I am, however, happy with the expression “dialect term”. (Adams 2007: 12)

Adams, on the one hand, has general objections to the popular term dialect, while, on the other hand, renouncing its applicability to the context of ancient Latin. This scepticism contrasts with his retaining the composition dialect term; a tension which remains unresolved. Not only dialect has been a productive source of new terminology; language, too, has been stripped for parts in order to express new conceptualizations of the human capacity for speech and its great diversity. In one interesting case, both terms served to convey new concepts: doculect and languoid, coined with the precise purpose of cancelling out any discussion about the language/dialect status of a speech form. They have been proposed as part of ‘a triumvirate of concepts’ next to the term glossonym (Cysouw 2014). This tripartition has been described most extensively in a paper of 2013, published after years of informal discussion at the Max Planck Institute for Evolutionary Anthropology in Leipzig between Michael Cysouw and Jeff Good, who have benefited from the input of many other linguists, including the typologist Martin Haspelmath. Cysouw and Good have introduced their three terms in order to formalize the language concept, a concern triggered by the largescale language cataloguing projects of Ethnologue and Glottolog and the problems involved in this endeavour, not least due to the lack of rigorous definitions of language and dialect (see Cysouw and Good 2013: 331). Languoid is defined as a ‘language-like object’ and ‘refer[s] to an entity used to designate any (possibly hierarchical) grouping of doculects, in principle ranging from a set of idiolects to a high-level language family’ (Cysouw and Good 2013: 347). Doculect, in turn, was coined at Haspelmath’s suggestion to denote ‘a named linguistic variety as attested in a specific resource’, including documents written in a specific language, studies of such a language, or simply vague references to it in, for instance, travel diaries (Cysouw and Good 2013: 356). A specific languoid can receive a label, the glossonym, which linguists and laypeople alike have used to refer to it. Cysouw and Good’s terminology, circulating since 2006, was almost immediately popular and now features not only in many linguistic publications, especially by typologists, but also in online language catalogues, including WALS

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

278

   ?    

Online and Glottolog.³ Their paper of 2013 was intended to present in more explicit terms their meta-model of the world’s linguistic diversity; it articulates their justified belief in the necessity of more accurate terminology in order to document the great variety of languages across the earth. This search for an improved jargon implies for them, like for many other contemporary linguists, the abandonment of the language/dialect pair.

22.1 Farewell to the conceptual pair? It is, however, unlikely that the conceptual pair will disappear altogether from linguistics in the foreseeable future. Not only are its traces indelibly left in the terminology of the discipline, but many linguists also persist in using it, either out of habit or by conscious choice. The very existence of the conceptual pair keeps on fuelling debate and has in recent decades even triggered entire books on the status of a language variety. For instance, Jean-Michel Eloy addressed at length in his 1997 monograph on Picard the question whether this Gallo-Romance speech form is a language or a dialect (see Eloy 1997). Eloy did so in direct response to the 1992 European Charter for Regional or Minority Languages in order to gain the rights protection offered by it (see Chapter 23, Section 23.3), even though France has never ratified this text (cf. also Bichurina 2016 on Franco-Provençal/ Arpitan). After providing a selective history of modern definitions of language and dialect, with a focus on the French tradition, Eloy presented the history of Picard, the theoretical framework of his analysis, and his morphological description of Picard. The evidence he collected led him to formulate three criteria for attributing a separate status to a minority speech form, only two of which Picard fulfilled. Firstly, speakers should be aware of the diglossia, which held true for Picard. Secondly, this feeling of diglossia should be backed by linguistic evidence, a criterion also met in the case of Picard. Thirdly, speakers should be willing to elaborate the language to standard status, a willingness lacking among Picard speakers. Eloy thus combined criteria of various origins in determining the language/dialect status of a form of speech, as has so often occurred in the history of language studies. Unlike his predecessors, however, he did so in a very conscious manner, yet without considering any problems that might be associated with such a hybrid approach. Not only in case studies such as Jean-Michel Eloy’s does the conceptual pair play a prominent role, but even more so in several recent reference works on language and linguistic diversity, not in the least in the Handbook of dialectology ³ See https://wals.info/ and https://glottolog.org/, last accessed 13 January 2020. Cf. Cysouw and Good (2013: 332–3).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.      ? 279 of 2018, edited by, among others, Charles Boberg, a student of William Labov’s. Its chapters are composed by specialists from various backgrounds and as such represent an entire gamut of different attitudes towards the conceptual pair. Several dialectologists contributing to the volume renounce the validity of the distinction, either because it presupposes an underlying linguistic structure as a closed system or because of its non-linguistic connotations.⁴ Others choose to continue its usage. This choice may be implicit, for instance in presenting concrete linguistic examples in terms of the opposition ‘standard language’/‘dialect’ (e.g. Hinskens in Boberg et al. 2018: 95). Explicit definitions of the term dialect are no rarity either, yet it is striking how diverse and noncommittal they are. The following examples may suffice to prove this: The term “dialect” is understood today to refer to a geographically delimited form of language. (Hickey in Boberg et al. 2018: 23) Dialect will be defined here as a geographically determined language variety (local or regional), which is essentially only orally transmitted and is relatively isolated from the roofing standard variety (if any). (Van Keymeulen in Boberg et al. 2018: 39) The focus on geographical space should be taken to imply that dialect refers to regional varieties, as per traditional terminology. (Kristiansen in Boberg et al. 2018: 107) Firstly, it is of course important to clarify what we mean by dialect. In continental European dialectology and sociolinguistics, the term is always used in opposition to a standard language/variety. Hence, the standard is not considered a dialect . . . The relational definition of dialect is, by definition, linked to a standard variety. The terminology suggests—correctly—that once the particular relationship between standard and dialect is established, it will result in the two varieties impacting on each other, usually with the standard influencing the dialect more than the reverse. (Auer in Boberg et al. 2018: 160) The term “dialect” will be used rather loosely to refer to accent (i.e., phonetics/ phonology only) as well as in the strict sense, that is, a language subvariety defined solely by grammatical and lexical properties. (Watt in Boberg et al. 2018: 226)

Do these authors want to distance themselves from the conventionalist interpretations they are proposing because they realize the problematic nature of the conceptual pair? It appears so; the result is an awkward clinging to traditional terminology for lack of any satisfying alternatives. This practice can be downright

⁴ See, respectively, Kretzschmar and Van Keymeulen in Boberg et al. (2018).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

280

   ?    

inconsistent when scholars explicitly dismiss the linguistic validity of the conceptual pair on the very same page (e.g. Van Keymeulen in Boberg et al. 2018: 39). My selective overview in the past four chapters has made it clear that the conceptual pair has been variously understood and assessed in mainstream linguistics during the last century. For reasons of space, I cannot treat in this book the various national traditions, in which the language/dialect distinction has been tailored to a country’s specific context. This local refashioning has often involved the integration of a third hierarchical layer into the presentation of a nation’s linguistic diversity, consisting either in varieties resulting from convergence between dialect and standard or in varieties historically not deriving from the standard but from the same protolanguage as the standard. For instance, in Great Britain, scholars and laypeople alike have supposed the existence of roughly three forms of speech: the English standard language in its received pronunciation, the standard language in a local accent, and regional dialects (Boberg et al. 2018: 4). Many more examples of such accommodations of the conceptual pair to national contexts can be cited. For instance, in modern Greece, a division into standard Modern Greek, regional accents (idiómata; ιδιώματα), and dialects deriving from the Ancient Greek Koine such as Cretan, Cypriot, and Cappadocian is current (see e.g. Trudgill 2003: 49), whereas in Dutch-speaking territory a variety between the standard and the dialects has been distinguished, known as ‘interlanguage’ (tussentaal; e.g. Absillis et al. 2012). National traditions do not always take over the terminology associated with the conceptual pair, however. In Italy, for instance, the term dialetto is often avoided with reference to Italo-Romance tongues such as Sicilian and Venetian, since this label might wrongfully suggest that they derive from the Italian standard language, while they in fact descend from Latin very much as does Tuscan, on which the standard is based (Auer in Boberg et al. 2018: 160).

22.2 The language/dialect distinction after 1900: the story of a love-hate relationship The language/dialect debate in linguistics has raged most vigorously since 1900, especially from the 1950s, when linguists began to devote entire papers to the question. The year 1954 was pivotal in this regard, when publications by structuralists from different backgrounds appeared. The conceptual pair was, however, usually only treated in the margin of mainstream linguistic studies, since the grand theories up until the 1960s focused on homogeneous language systems and ideal speakers. Perhaps structural and generative linguists felt uncomfortable treating the issue, as most of them refrained from investigating the phenomenon of linguistic diversity in great detail. This lack of interest might explain why the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

281

language/dialect distinction before the 1950s was treated at some length principally in introductory manuals to the field, in which the basic concepts of linguistics could not be left implicit. Whereas structuralists and generativists tended to shy away from matters of language variation and focused on idealized systems, the phenomenon of diversity became re-established as an object of study in the 1960s thanks to the work of William Labov, who initiated a true social turn. Yet extralinguistic variables and their correlation with specific linguistic features took and still take centre stage in this field of enquiry, and sociolinguists, too, have refrained from extensively reflecting on the language/dialect distinction. They are very much data-oriented, like traditional dialectologists (cf. Boberg et al. 2018: 12–13). A similar conclusion can be drawn for recent lectometric studies. This tilt towards actual linguistic data may be related to the long-standing problem of delimiting speech forms, first raised at length in the 1860s, as I have argued (Chapter 18). Focusing on data has the advantage of avoiding the seemingly unsolvable conceptual puzzle. In recent years, however, linguists like William Kretzschmar have proposed alternative models of language in order to overcome this problem. Scholars engaged in charting the linguistic diversity of the world are especially drawn to the problem of linguistic delimitation and classification; this holds a fortiori for those focusing on non-European tongues. It is, for instance, no coincidence that statistical measuring of linguistic distance was introduced by Morris Swadesh, a specialist in the indigenous languages of North and Central America, precisely for purposes of classification. Debates about the language/dialect distinction have often been of a cursory nature in that the conceptual pair is rarely broached as the main problem of a discussion. Indeed, although the principal topic of several major journal papers, the issue has been the primary subject of only few conferences, let alone monographs, and this only from the 1970s onwards, when a kind of dialect renaissance occurred (Mattheier 1983: 135). One of the most important exceptions is the international symposium ‘On the theory of dialect’ (‘Zur Theorie des Dialekts’), organized by Joachim Göschel at the Philipps-Universität Marburg from 5 to 10 September 1977. Many leading sociolinguists and dialectologists participated in this event, including the Romanist of Romanian descent Eugenio Coseriu, the American linguist Einar Haugen (1906–94), and Heinz Kloss. The proceedings of the conference have been published in 1980 and include not only the papers themselves but also useful summaries and, more interestingly, a written account of the lengthy oral discussions following many of the presentations (Göschel et al. 1980). Anyone reading this fascinating volume will recognize in the papers and especially the discussions afterwards most of the main tendencies in the modern history of the conceptual pair which I have presented in the past few chapters. Let me concisely recapitulate them here.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

282

   ?    

I maintain that five principal tendencies can be distinguished. Firstly, many scholars do not question the conceptual pair at all and preserve this Renaissance relic, which can be either defined in explicit terms or simply presupposed and used without further definition. This continued usage ranges from monograph-length studies like Jean-Michel Eloy’s over an insistence on mutual intelligibility to a mere passing reference to speech form x as a dialect of language y. Secondly, scholars have tried to redefine the conceptual pair, for instance to integrate it into the theory advocated, as some generative linguists have attempted, or by using new computational means, as in Wichmann’s lexicostatistic analysis.⁵ Thirdly, linguists have supplemented the conceptual pair, because they feel that criteria of a diverging nature have been intermingled in defining it. Heinz Kloss, for instance, has influentially distinguished between linguistic distance (Abstand) and the sociological aspect of elaboration (Ausbau) in treating language/dialect constellations, whereas André Martinet opted for a simpler scheme with language, dialect, and patois. More complicated constellations have been proposed by, among others, Eugenio Coseriu (1980). Fourthly, linguists have proposed entirely new models to capture linguistic diversity, most notably Kretzschmar’s (2009) conception of language as a complex system in which variation is inherent as well as Cysouw and Good’s (2013) ‘triumvirate’ of glossonym, doculect, and languoid, devised for language classification purposes. Fifthly and finally, numerous linguists have chosen a less constructive path and simply dismiss the conceptual pair as irrelevant for language studies, for which they often invoke the Weinreich witticism. Language/dialect disputes are not linguistic but sociopolitical problems, linguists in this category argue. Yet downright dismissal does not necessarily entail avoidance of the terms language and dialect, which can be found in traditional meanings even in the work of linguists suspicious about the conceptual pair. Even though modern scholars often approach the language/dialect distinction with great scepticism, they have found in the controversial word dialect a prolific source of new terminology, to which the many derivations and compositions testify. Age-old interpretations of the conceptual pair live on in many modern ideas, even though scholars do not recognize this continuity. The geographical element, most notably, is still one of the principal parameters in definitions of the term dialect. The criterion of mutual intelligibility has risen to prominence since the first tests in the early 1950s, even though it has proved to be almost as problematic as the language/dialect pair itself. The association of dialect with small-scale variation remains popular, too, for instance in generative approaches. Yet linguistic differences are no longer interpreted in a binary fashion as in the traditional Aristotelian criterion with its opposition substance/accidents, but in the wake of

⁵ See e.g. also Tillinger (2013) for a recent redefinition tailored to the linguistic context of France.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

.    

283

Morris Swadesh’s lexicostatistics a rich tradition of actively measuring linguistic distance has established itself. Intuitive analyses have given way to empirical studies. A crucial question has been that of the cut-off point: where do speech forms become so different that they are distinct languages rather than related dialects? A recent study in this vein has suggested that linguistic diversity is bimodally distributed, which means that there might be something like a dialect pole and a language pole around which speech varieties flock together. This bimodal distribution does, however, remain a tendency with a transition zone rather than an absolute rule without exceptions. The reader may have noticed that in the past few chapters I have for the first time made repeated reference to lay usages of the term dialect in opposition to language. Scholars such as André Martinet, Uriel Weinreich, and Heinz Kloss, to name but a few, expressed the need to arrive at a more technical understanding of the conceptual pair, which, they noticed, non-specialists interpreted in ambiguous ways. Where did these lay conceptions of the language/dialect distinction originate? This question is all the more pressing, since I have argued in Part II that the pair of concepts emerged as a technical metalinguistic instrument in the work of Renaissance Hellenists in the early sixteenth century and not at all as a popular idea. I will briefly look into the vulgarization of the language/dialect opposition in the final chapter of this book, since it will help the reader understand why many modern linguists have had the anxious reflex to distance themselves from the conceptual distinction and, especially, its sociopolitical connotations.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

23 Language, dialect, and the general public— or how not to popularize knowledge The taxonomy of linguistic description—that is, the identification and enumeration of languages—is greatly hampered by the ambiguities and obscurities attaching to the terms “language” and “dialect”. Laymen naturally assume that these terms, which are both popular and scientific in their use, refer to actual entities that are clearly distinguishable and therefore enumerable. A typical question asked of the linguist is: “How many languages are there in the world?” Or: “How many dialects are there in this country?” (Haugen 1966: 922) Einar Haugen’s analysis is spot-on. The terms language and especially dialect have ‘carried the full burden of both scientific and popular usage’ (Haugen 1966: 924). As I have contended in the previous chapters, numerous modern linguists are aware of the awkward fact that they share the language/dialect distinction with the general public. This realization has raised suspicion against the conceptual pair, up to the point where many linguists have barred it from scientific usage, assuming that it had its origin outside scholarship. Yet I have argued in Part II that the conceptual pair emerged for the first time on the large scale in specialist philological works in the early sixteenth century. How did it evolve from a technical tool to a piece of common knowledge?

23.1 The conceptual pair popularized RUn away on Thursday April 12 about two of the Clock in the Morning, from the Right Honourable the Lady Lovelaces House in Little Queen street, John Howel Footman to the said Lady, in a new black Liver, and brown Perriwig, about 28 years old, low of Stature, and speaking a Welch Dialect, having stole several things, and the next night six Silver Forks, and six Silver Spoons, with the Arms of the Family, and other Plate and things were suspected to be stole by the said John out of the said House: whosoever shall bring him, or give

Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    

285

notice of him to Mr. Lion in Holborn Grocer, or discover the said Plate shall be well rewarded for their pains. (Anon. 1672) This ‘advertisement’ was published in The London Gazette edition of 22 April 1672. Such wanted notices were far from uncommon in early modern newspapers. What is surprising, however, is that the word dialect appeared in it as part of the physical description of the wanted criminal. Traditionally included in seventeenth-century English dictionaries of difficult words, dialect remained a technical term for most of the early modern era. Indeed, it seems to have occurred only sporadically outside learned discourse before the nineteenth century. Another intriguing exception stems from late eighteenth-century Scotland. In 1783, the redaction of The Edinburgh Weekly Magazine received a query from one of its readers: ‘What are the causes of a difference in dialect among people who speak the same language?’ In the 11 September edition, the interested reader could find the following answer: VARIETY of dialect may originally and unavoidably arise, in some instances, from difference in the conformation of the organs of speech. The pronunciation of several consonants is affected by this circumstance. But the following causes, which it is unnecessary to illustrate, are sufficient to account for that diversity of speech which is observable in every nation:—Inattention to the pronunciation of others, and the retaining of mistaken pronunciation; the licence of poets; the affectation of novelty in curtailing, lengthening, and compounding of words; borrowing from foreign languages, with a view to adorn or improve; the arbitrary invention of words by those who are unacquainted with many already in use; the rejection of vulgar, and the re-admission of obsolete terms. The effect of these causes becomes remarkable in proportion to the want of acquaintance between the inhabitants of distant provinces. Though, at the building of Babel, diversity of languages was the cause, it is now the effect of separation among mankind. The language of North America will, in a few centuries, become unintelligible to an inhabitant of Great Britain. (Ruddiman 1783)

In the eighteenth century, the market for knowledge opened up, and many newly created magazines and almanacs popularized scientific information.¹ These publications found new readership in educated bourgeois circles; their members cultivated broad interests and inundated the magazines with the most diverging questions, to which the editors tried to formulate sound answers. Regional language diversity, too, drew the attention of some educated laypeople, who

¹ See e.g. Lynn (2006: especially Chapter 2), focusing on eighteenth-century France.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

286

, ,    

increasingly conversed in the standard languages, both in writing and orally. It is in this light not surprising that in his answer to the above query the editor made disparaging comments about the phenomenon of dialect, linking it to incorrectness. Yet he had a very broad conception of it, since he took it to cover also poetical properties and, above all, various kinds of lexical peculiarities. Isolation and lack of contact enhanced these formal differences, and on this basis he predicted the linguistic separation of American English, which was bound to become unintelligible to the British. This prediction might imply that the editor replying assumed that speakers of related dialects could understand one another, whereas speakers of distinct languages could not. Yet he did not offer an explicit criterion by which to differentiate language from dialect. Outside Great Britain, the conceptual pair had also made its way to nonspecialists. The historian Johann Christoph Gatterer, for instance, who developed an embryonic lexicostatistic method in the late eighteenth century (Chapter 14), noticed that phrases such as ‘dialects of a language’ occurred ‘even in common speech’.² Non-specialist familiarity with dialect in opposition to language increased during the first half of the nineteenth century in the wake of Romanticism, which reacted against the extreme rationalism of the Enlightenment and put emotion at the forefront. Pristine and rustic dialects were put to use by Romantic writers for various literary purposes, which included evoking pastoral scenes and character marking.³ Historic dialects were set off from the frigidly formalized standard language that was gaining wide acceptance. It has been maintained that early dialect geographers such as Wenker were still very much working in this Romantic spirit, since they, too, focused on the age-old and allegedly pure dialects of the countryside, which came under pressure from the steadily advancing standard languages. From the end of the nineteenth century, modern linguists realized that laypeople likewise had ideas about what constituted a dialect and what a language. Georg von der Gabelentz reported extensively on popular interpretations of the conceptual pair in the revised edition of his linguistics handbook, posthumously published in 1902: A language is considered the common property of a people, a Dialekt or a Mundart the common property of a region—this may more or less correspond to the ordinary conception. Schriftdeutsch is written and read in the entire fatherland, preached by the chancellors, and learned in the schools: that is why it is language. Bavarian, Swabian, Palatine, etc., to the contrary, are dialects. It will be difficult to understand, in accordance with this viewpoint, that Plattdeutsch is

² Gatterer (1770: 13): ‘In Sprachuntersuchungen, und sonst auch, selbst im gemeinen Reden, kommen immer die Ausdrücke vor: verwandte Sprachen, Dialecte einer Sprache’. ³ See e.g. Hodson (2018) for dialect and Romanticism in Britain.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.    

287

not simply a dialect, too. Newspapers, public bodies, clerics, and teachers, in many cities most citizens indeed speak High German in Low Germany as well. Dialect is spoken by the man in the overall. Educated men, from waiters and up, take pains to speak dialect-free. In the interior, one concedes a particular language only to the people whose speech one does not understand at all, as, for instance, to our Sorbs of Lausitz. This way of judging things is external, superficial, and necessarily leads to inconsistencies. The history of our days has, however, proved that such views can also lead to fairly vexing consequences. Since language is the common property of the people, language community thus justifies the right to political unification, language diversity the right to break away—in this manner the modern nationality principle judges according to the motto: “as far as tongue x resounds”.⁴

Gabelentz thus rejected political parameters in determining the language/ dialect status of a speech form. More fruitful, he maintained, was another popular criterion: mutual intelligibility. In fact, to Gabelentz’s mind this feature could be the only workable touchstone by which to make language/dialect decisions. Yet the criterion was, as it was used by laypeople, not objective and explicit enough. He proposed to take the speech of uneducated people as the foundation of intelligibility analysis, admitting, however, that the criterion remained somewhat arbitrary. The conceptual pair could only be used for convenience, since precise definitions were impossible. A contemporary of Gabelentz tellingly wrote: No scientific or adequate definition can be given. For all practical purposes this will suffice. A language is a big dialect, and a dialect is a little language. (Meiklejohn’s 1891 Book of the English language as quoted in Crowley 2003: 81)

⁴ Gabelentz (1901: 55): ‘Unter einer Sprache denkt man sich das Gemeingut eines Volkes, unter einem Dialekte oder einer Mundart das Gemeingut einer Landschaft—dies dürfte so etwa der Allerweltauffassung entsprechen. Schriftdeutsch wird im ganzen Vaterlande geschrieben und gelesen, von den Kanzeln gepredigt, in den Schulen gelehrt: mithin ist es Sprache. Bairisch, Schwäbisch, Pfälzisch u. s. w. dagegen sind Dialekte. Diesem Standpunkte wird es schwer zu begreifen, dass Plattdeutsch nicht auch blos ein Dialekt ist; Zeitungen, Behörden, Geistliche und Lehrer, in vielen Städten die meisten Bürger reden ja auch in Niederdeutschland hochdeutsch. Dialekt spricht der Mann im Kittel; die Gebildeten, vom Kellner aufwärts, bemühen sich ‚dialektfrei‘ zu sprechen. Im Inlande gesteht man nur den Leuten, deren Rede man gar nicht versteht, eine besondere Sprache zu, so z. B. unsern Lausitzer Wenden. Diese Art die Dinge zu beurtheilen ist äusserlich, oberflächlich und muss zu Inconsequenzen führen. Die Geschichte unserer Tage hat aber bewiesen, dass solche Ansichten auch zu recht ärgerlichen Consequenzen führen können: weil die Sprache Gemeingut des Volkes ist, so begründet Sprachgemeinschaft das Recht zur politischen Vereinigung, Sprachverschiedenheit das Recht zur Losreissung—so urtheilt das moderne Nationalitätsprinzip nach dem Wahlspruche: „soweit die x’sche Zunge klingt“.’ On Gabelentz’s ideas on the conceptual pair see also Chapter 17, Section 17.4.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

288

, ,    

23.2 The Weinreich witticism in context: the nation-state, language policy, and mass education Gabelentz procured a surprisingly thorough and insightful analysis of popular conceptions of the language/dialect distinction. He deplored that the opposition had been misused to claim or deny the right to political autonomy. Nationalist sentiment had penetrated deeply into the conceptual pair. Already in the eighteenth century, the phenomenon of dialect had captured attention outside the confines of academia among educated laypeople. In this stage, however, the lay conception of the language/dialect distinction had been principally shaped by ideas about cultivation and education, or a lack thereof. Yet in the modern period, the conceptual pair became knowledge available to roughly all classes of society. The channels through which the popularization of the conceptual pair occurred demand more systematic research, but the distinctly political dimension which entered the scene in the nineteenth century suggests in which direction to look. Gabelentz, in fact, had already provided part of the answer in his analysis of the linguistic side of the ‘modern nationality principle’. It was the language ideology of modern nation-states which was responsible for politicizing the conceptual pair. Standard languages became an instrument for nation-building. The members of a nation should learn to speak the uniform language imposed by the state and diffused through mass education and media. If one aspired to attain power, mastery of the standard language was indispensable. Language was the statebacked means of communication, dialect a regionally limited form of speech current among citizens lacking both power and good education. Or to put it with the words of Weinreich’s auditor: ‘A language is a dialect with an army and navy’. To make the nation more homogeneous, dialects were sometimes actively suppressed with a view to their eradication. They were useless in the nation-state’s standard language ideology. For reasons of space and clarity, I am generalizing. It is desirable to investigate the matter for each nation-state separately, since there are differences between them all, depending on such factors as the education system designed and applied, the sociopolitical and pedagogical philosophy behind it, and the tenacity with which the standard language was imposed. It occurred very early and strongly in France, where in the aftermath of the French Revolution Henri Grégoire (1750–1831), better known as the Abbé Grégoire, heralded a new era in national attitudes towards dialectal variation. In his well-known report of 1794, presented to the French National Convention, Grégoire discussed ‘the necessity and means for annihilating the patois and universalizing the usage of the French language’.⁵ ⁵ See the report’s title: Rapport sur la nécessité et les moyens d’anéantir le patois, et d’universaliser l’usage de la langue française. Cf. Pop (1950: ., .6–7); Burke (2004: 10). On the place of this report in the history of French nationalism see Bell (2003: 169–97). See Certeau et al. (2002) for the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.      

289

Grégoire did not think it fit for a nation like France, mother of liberty, to remain at the Tower of Babel language-wise. The patois were remnants of feudal society, whereas a uniform national language spoken by all would serve to perpetuate the hard-won liberty of the French Revolution. Interestingly, Grégoire adduced American English as the main linguistic model for French. The annihilation of the patois became a prominent item on the political agenda as an efficient means for freeing the nation from feudality. This radical new start involved a severe degradation of dialects vis-à-vis the early modern period, during which the phenomenon was not as emphatically interpreted in negative terms. State-guided mass education, then, was most likely a factor of prime importance in spreading to nearly all citizens of the state the idea that there was something like a language/dialect distinction. This school revolution gathered steam especially from the last decades of the nineteenth century onwards (see Boli et al. 1985: 146, 155; Meyer et al. 1992: 145; Ferguson 2006: 23). It was probably more decisive in the diffusion of the conceptual pair than, for instance, mass media. Often organized by the nation-state, mass education served to spread its ideology, a ‘package’ of which standard language ideology was a core part (May 2011: 153–4; cf. Joseph 2004: 52). Children of different social classes and backgrounds were confronted on a large scale with the uniform linguistic norm of the nation-state, which they were expected to learn to read, write, and eventually also speak. They were guided in doing so by teachers who were required to draw a sharp dividing line between the ideal of the standard language and the various native dialects of their students (cf. Van der Horst 2008: 189). The conceptual pair, in sum, most likely reached the masses in an educational context shaped by the political and ideological concerns of the nation-state’s leadership. This development deserves much closer attention, especially since it involved the political appropriation of technical terminology coined by philologists during the Renaissance in the first place to describe and categorize the ubiquitous phenomenon of linguistic diversity.

23.3 Political activations of the conceptual pair The vulgarization of the conceptual pair in modern times had enormous consequences for its interpretations, since it led to the foregrounding of languageexternal and often subjective parameters. As a product of state-backed popularization through mass education, the language/dialect distinction was redefined in predominantly political and sociocultural terms. Policy dictated a sharp divide between the uniform standard language and the diversified dialects, and the result relationship between the French Revolution and the patois, and the many follow-up discussions of this issue (cf. Burke 2004: 4 n.11).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

290

, ,    

was a highly prescriptivist attitude towards language. Language was the precinct of power, progression, and cultural superiority, dialect that of subordination, the status quo, and lack of refinement. As such, dialect suffered from a severe social stigma. In popular usage, the conceptual pair came to be defined by external criteria alone, which makes it a very unsuccessful case of knowledge vulgarization indeed, since most early modern scholars and modern linguists had opted either for a mix of language-internal and external criteria or exclusively for languageinternal criteria. The conceptual pair’s new political guise has had far-reaching consequences. At various occasions during the last decades, the language/dialect distinction has been the centrepiece of numerous debates, being activated in policymaking. An already classic example is the so-called Oakland Ebonics controversy raging from late 1996 to early 1997, with substantial follow-up discussion (see e.g. Wolfram 1998; Edwards 2009: 79–82). In a resolution of 18 December 1996, the Oakland School Board (California, US) recognized varieties of Ebonics, or African American Vernacular English, as ‘not merely dialects of English’ and suggested the use of Ebonics in teaching Standard English to Afro-Americans. The result was a fierce debate, because the resolution, using a negative variant on the ‘to differ only in dialect’ phrase, was widely interpreted as claiming that Ebonics constituted a separate language rather than an English dialect. It even led to Senate hearings of linguists and non-specialists alike, including William Labov, who were asked to present their view on the precise status of Ebonics. Tellingly, it was more popular language ideology than linguistic study that determined the stances of laypeople. The American linguist Walt Wolfram, for instance, described how he was often asked whether he ‘believed’ in Ebonics, as if a religious sect were at stake rather than a form of language (Wolfram 1998: esp. 110). Suffice to say, linguistic diversity can be a delicate subject, especially if it amounts to language/dialect disputes. Perhaps more remarkable still is the European Charter for Regional or Minority Languages, which was signed on 5 November 1992 but came into force only on 1 March 1998. The treaty was prepared by the Council of Europe (Figure 23.1), an international organization not to be confused with the European Union, to safeguard the languages spoken by smaller communities of its member states. In order to clarify what was meant by ‘regional or minority languages’, the charter offered in article 1 a definition of this phrase: “regional or minority languages” means languages that are: i traditionally used within a given territory of a State by nationals of that State who form a group numerically smaller than the rest of the State’s population; and ii different from

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

.      

291

the official language(s) of that State; it does not include either dialects of the official language(s) of the State or the languages of migrants.⁶

The charter’s explicit exclusion of dialects forced participating member states to make on a large-scale language/dialect decisions for speech varieties in their territories, often with major implications. Certain minority tongues could not enjoy privileged treatment because governments decided that they could not. For instance, the language of the Rusyns (Ruthenians) was not recognized by Ukraine in the framework of the charter, thus denying it regional language status solely on the basis of the government’s political and ethnic preconceptions about the Rusyns, who are considered to be Ukrainians speaking a Ukrainian dialect and not a separate people with its own language. Contradictorily, Rusyn, spoken not only in south-western Ukraine but also in surrounding countries, did obtain regional language status in other states which have signed and ratified the charter, including Croatia, Romania, and Slovakia. As such, the document embodies more than anything else the problematic politicization of the conceptual pair. It therefore does not seem to be a pity that this charter, noble in spirit but questionable in

Figure 23.1 The Council of Europe’s headquarters in Strasbourg Source: Wikimedia Commons. CC BY 3.0

⁶ Council of Europe (1992: 1–2). Cf. also Gorter and Cenoz (2012: 263); Van Keymeulen in Boberg et al. (2018: 39); Gooskens in Boberg et al. (2018: 205).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

292

, ,    

execution, can be denounced at any given time by the participating countries (Council of Europe 1992: 13, Article 22). In the last decades, the conceptual pair has played its role in non-specialist debates also outside of politics. Discussions on the topic are often strongly coloured ideologically and show a remarkably high degree of vagueness. A number of such recent debates are reflected in online publications. For instance, a blogpost of 16 February 2014, published on the website of The Economist, a London-based weekly news magazine, explained ‘[h]ow a dialect differs from a language’ (R. L. G. 2014). It did so in the aftermath of a debate in Hong Kong about the precise status of Cantonese. The education department had claimed that, contrary to English and Standard Chinese, Cantonese was not an official language of Hong Kong. This assertion led to a debate about the status of Cantonese: was it a separate language or merely a dialect of Chinese? The author of the blogpost, American journalist Robert Lane Greene, tried to offer a definitive solution to the issue. Before dismissing the popular political criterion with reference to the Weinreich quip, he elaborated extensively on mutual intelligibility. The latter was attributed to professional linguists and claimed to be the most objective criterion, ‘the linguist’s common-sense definition’. The author was aware that the criterion was not absolute, but this was no problem for its objectivity: Linguists have a different criterion: if two related kinds of speech are so close that speakers can have a conversation and understand each other, they are dialects of a single language. If comprehension is difficult to impossible, they are distinct languages. Of course, comprehensibility is not either-or, but a continuum—and it may even be asymmetrical. Nonetheless, mutual comprehensibility is the most objective basis for saying whether two kinds of speech are languages or dialects. (R. L. G. 2014)

The conclusion was that Cantonese must be a distinct language, related to other Chinese languages such as Mandarin and Shanghainese. The author added that this ‘objective’ mutual intelligibility criterion ‘can annoy nationalists’ in China as well as elsewhere. To further support the argument, the typical example of Danish and Norwegian was invoked to illustrate that two mutually intelligible varieties could be the standard languages of different nation-states. This political circumstance made speakers of these tongues reluctant to regard their respective standard languages as ‘only’ dialects of one and the same language. In conclusion, the author wanted to create the impression that the interpretation of the conceptual pair was undisputedly clear among linguists and tried to validate it as a means to provide an allegedly objective answer to a politically thorny issue. In recent times, the conceptual pair has been politically activated to make various claims. In such cases, it tends to arouse strong feelings. For reasons of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

. 

293

convenience, I have focused on two notorious recent episodes in language policy and an English-language blogpost, but the uses and appropriations of the conceptual pair outside academic linguistics deserve a closer investigation, taking into account various governing bodies and different languages and national contexts. It might, moreover, be fruitful to study popular attitudes towards the language/ dialect distinction through experiments, questionnaires, or social media. The last resource, in particular, offers many research opportunities not only to dialectologists and variationist linguists but also to scholars investigating language attitudes and popular interpretations of widely known linguistic concepts. Especially Twitter, the largest part of which is accessible to anyone with an Internet connection, provides the researcher with many interesting and easily analysable data.

23.4 Conclusion The conceptual pair gained some ground outside scholarly circles as early as the seventeenth century and especially in educated lay circles during the eighteenth century. It was, however, only with the advent of modernity that the language/ dialect distinction reached nearly all classes of society in the wake of mass communication and especially general education instituted by the nation-state. Language is the official means of communication of the nation-state, the refined standard language, the language of promotion, dialect the diversified and repulsive speech of the lower populace and the status quo. Such has been the major political interpretation of the conceptual pair in modern times, even though in recent decades a countermovement has done much to value dialect in more positive terms. This popular take on the two concepts is motivated exclusively by subjective, language-external parameters. As such, the language/dialect distinction has fallen victim to the language policy and planning, the glottopolitics, of modern nation-states, as well as to the sociolinguistic prejudices of lay speakers. It has become a prime example of how not to popularize scholarly terminology. The discourse of laypeople and linguists alike suggests that there has been, and still is, a strict distinction between lay and specialist ideas, and the latter often condemn the views of the former. Yet it is certainly not the case that ideas from both sides are formulated in complete isolation from each other. Linguists have, for instance, been consulted by policymakers—recall that William Labov partook in the Oakland Ebonics Senate hearings—and have written blogposts and recorded podcasts on the language/dialect distinction, often pointing out its meaninglessness and political inspiration (e.g. McWhorter 2016; 2019). The blogger for The Economist, too, confronted lay with specialist conceptions. This dialectic interaction between laypeople, policymakers, and linguists deserves further study, especially since it will allow linguists to trace currents of influence from linguist to layperson as well as from layperson to linguist.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

294

, ,    

Laypeople looking for an answer have, for instance, turned to linguists to find more objective criteria, such as mutual intelligibility. Linguists opposing the pure and correct norm to the deviant dialects, in turn, have been influenced by subjective conceptions and the standard language ideology of the nation-state in their prescriptivist attitudes, even though it must be granted that they are becoming an increasingly rare breed. Still, Calvet (1974: 46–8) has pointed out that sociopolitical factors underlie numerous definitions of the language/dialect distinction formulated by linguists in the twentieth century. The strict distinction claimed to exist between lay and specialist conceptions adumbrates the fact that there is a great diversity of ideas within each group itself. I have illustrated this abundance of opinions at length over the past chapters for scholarly interpretations of the language/dialect distinction. There is, however, no uniform popular conception of it either, and this variation, too, is a matter requiring further research. Peter Auer has recently suggested that there are two main popular interpretations of dialect, both championing in vagueness (see Boberg et al. 2018: 160). On the one hand, it is viewed as a pure and pristine relic from the past, enjoying prestige because of its antiquity; this is the traditional take on dialect cultivated also by most dialectologists up until the 1960s. Less appreciative is, on the other hand, the conception of dialect as an inferior deviation from the standard language. Further research can no doubt offer a finer-grained picture of the wealth of popular ideas about the language/dialect distinction. I have not insisted on the ways in which the conceptual pair has served, or still serves, in academic branches other than linguistics, which would have led me too far. Yet the ideas of sociologists, anthropologists, historians, and the like also deserve closer study as well as their interactions, both with lay ideas and with views in other academic branches. That these interactions are very real, is proved by, for instance, Louis-Jean Calvet’s Linguistics and colonialism: A short treatise on glottophagy of 1974. In this pioneering and polemical book at the crossroads of linguistics and postcolonial studies, Calvet analysed the ways in which Western linguistic concepts have been carriers of dominant imperialist ideologies and have given shape to ideas about ‘the other’ and their speech. His postcolonial perspective led him to denounce vehemently the conceptual pair not only because of the vagueness with which it has been defined but especially because of its languageexternal, political interpretations (Calvet 1974: 40–54). Imperialist discourse has resulted in the denigration and degradation of certain speech forms to dialect status in favour of the language of the establishment solely on the basis of power relations and preconceptions about Western cultural superiority. To Calvet, ‘a dialect is always nothing more than a defeated language, and the language is a

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

. 

295

dialect which has succeeded politically’, which is why he proposed to abandon this terminology, since the burden of its connotation was too heavy to carry.⁷ Calvet’s book, in short, suggests that cross-fertilization between different academic branches has indeed occurred; it is, for instance, not inconceivable that postcolonialism has contributed much to exposing the political connotations of the language/dialect distinction. Needless to say, Calvet is only one eminent example among many, all of which deserve a more in-depth historiographic study.

⁷ Calvet (1974: 54): ‘le dialecte n’est jamais qu’une langue battue, et . . . la langue est un dialecte qui a réussi politiquement’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

24 Language and dialect between past and future Terminological success, conceptual failure?

I have tried to guide the reader on the age-long journey of the distinction between language and dialect. Its history is, however, not as long as may be suspected. The conceptual pair does not hark back to ancient Greece, as one might guess from the etymology of the word dialect, but rather to the sixteenth century, the period of the Renaissance, when humanists rediscovered and appropriated Greek scholarship. Yet its history, although shorter than often assumed, is still a fascinating one. In fact, it emerged under peculiar conditions in the early sixteenth century. There was a renewed interest in ancient Greek literature, which unlike its Latin counterpart was composed in a gamut of different dialects. Also, reinforced by the invention of the printing press and embryonic national sentiments, processes of language standardization were gathering steam; these involved differentiating the selected variety, the language, from other speech forms, the dialects, even though this contrast did initially not occur in predominantly depreciatory terms for the dialect pole. The linguistic horizon of scholars moreover opened up. For the first time, they had what I have dubbed a ‘multidialectal outlook’, as it was no longer exceptional to be acquainted with the dialects of different languages, typically those of Ancient Greek and one’s native tongue. Humanists needed concepts to bring some structure into the enormous linguistic diversity they observed, and language was as much as anything else the object of classification attempts in the Renaissance, an era which witnessed a true explosion of information. The language/dialect distinction proved a welcome classificatory tool, since it made languages as well as dialects countable entities, an evolution embodied by nothing better than Conrad Gessner’s 1555 language catalogue. The conceptual pair was, in other words, a construct of its age. What Brian Ogilvie (2006: 1) has stated about the Renaissance ‘invention of natural history’ may therefore be taken to apply to the language/dialect distinction, too: ‘it could be the product only of a community’, in particular a community of scholars. As opposed to Roger Bacon’s ephemeral advocating of a similar conceptual opposition in the thirteenth century, humanists were more united in their endeavour, and their innovation met with overwhelming success. Ironically, unlike Bacon, Language or Dialect? The History of a Conceptual Pair. Raf Van Rooy, Oxford University Press (2020). © Raf Van Rooy. DOI: 10.1093/oso/9780198845713.001.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

      

297

they were unaware of the novelty of their distinction, since they assumed that they were simply borrowing it from the greatly admired ancient Greeks, without realizing that they were enforcing their own interpretation on the works of their illustrious predecessors. It is telling that I have had to diagnose, so to speak, the emergence of the language/dialect opposition by means of symptoms arising, roughly, in the years 1500–50. As such, the conceptual pair was a blend of ancient ideas and new insights, a typical product of Renaissance scholarship (cf. Rice and Grafton 1994: xiii, 83). Humanists intuitively transformed the impressionistic concept of ancient Greek and Byzantine scholarship, diálektos as ‘particularity of tongue’, which could acquire secondary parameters such as ‘typical of a certain tribe/region’, into a reified and countable entity. Dialect came to be subsumed under an overarching language concept, a process in which the initially secondary parameters gradually gained a more prominent position. Yet because of its peculiar emergence conditions, the conceptual pair only received explicit interpretations after several decades, from about 1550 onwards. Still, much was left to the reader to guess, and early modern authors could greatly vary in their conceptions of the language/dialect contrast. Daniel Baggioni has aptly summarized this: It would be very difficult to find a beginning of an agreement on the terms language and dialect, which vary considerably in their meaning from one author to the other and from one period to the other.¹

I have, however, tried to illustrate that it is nonetheless possible to observe broader tendencies in early modern ideas. I have identified seven major interpretations of the conceptual pair, which I need not repeat here. The making explicit of how a dialect differed from a language went hand in hand with the emancipation of the distinction from Greek scholarship. From the seventeenth century onwards, it was no longer Hellenists alone who made substantial contributions to conceptions of the language/dialect contrast. After about 1650, once the major early modern interpretations had been established, the focus shifted to the ways in which related dialects differed linguistically from one another in contrast with distinct languages, as I have argued mainly through the work of Georg Stiernhielm and Albert Schultens. This Enlightenment rationalization of the conceptual pair culminated in an embryonic lexicostatistic approach in the work of the eighteenth-century Göttingen historian Johann Christoph Gatterer. At this stage, the first critical voices about the conceptual pair and some of its interpretations were also heard. ¹ Baggioni (1997: 56): ‘On serait bien en peine . . . de trouver un début d’accord sur les termes langue et dialecte qui, d’un auteur à l’autre, d’une époque à l’autre, varient considérablement dans leur signification’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

298

      

Suffice to say, the conceptual pair had already come a long way when it was adopted as a kind of self-evident given by scholars in the early nineteenth century, at which time linguistics emerged as a separate scholarly discipline. Owing to the initial focus of linguists on the interrelationships of Indo-European tongues and their common origin, the language-historical interpretation was foregrounded. Dialects descended from a language, like daughters from a mother. The geographical aspect rose to prominence during the 1870s in the wake of Georg Wenker’s dialect studies and his pioneering linguistic atlas. At the same time, the results of this new dialectological research made scholars such as Hugo Schuchardt realize that dialects were not as countable as was commonly believed ever since the Renaissance and that the distinction between language and dialect was above all arbitrary. For the first time, the conceptual pair came to be questioned on a large scale, a scepticism heard most loudly in France, especially in Paris, during the decades around 1900. While historical-comparative linguistics and dialectology continued to be practised throughout the twentieth century, they were superseded quite rapidly by the grand theories of structuralism and later generative grammar, in which there was usually little space for small-scale language variation and hence also for the conceptual pair. It was only in the 1950s that a change was initiated, when Uriel Weinreich dared to ask the question whether a structural dialectology was possible. Only from this moment onwards did the debate about the language/dialect distinction start to rage in full vigour, to which the diverse journal articles entirely devoted to the matter testify. It is telling that since the 1950s many linguists have tried to come up with objective criteria to separate languages from dialects, very likely reflecting an urge to counter the rise of language-external, often political and subjective interpretations from the late nineteenth century onwards. The booming of the debate was also tied, like in the Renaissance, to an increased concern over counting the languages of the world, embodied in the various online language catalogues created in recent decades. Seen from this point of view, it is not surprising that many contributions to the debate have been fuelled by attempts at classifying the many indigenous languages outside Europe, where strong preconceived ideas about the language/dialect status of individual speech forms were lacking. Even though the debate has raged most intensely in the last seventy years, it has taken place in a rather unstructured fashion, with many intervals and limited direct discussion. In sum, the conceptual pair, although usually taken as an obvious and ahistorical given, is a relic from a not-so-distant past. As yet, few historians of linguistics, let alone linguists themselves, have acknowledged this; a rare exception is the Italian scholar Mario Alinei (1980), who deserves full credit for his analysis of the Renaissance as a pivotal period, even if his interpretation seems to be more socially coloured than the evidence allows. The misleading timelessness of the language/ dialect opposition is probably enhanced by the fact that nobody seems to want to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

      

299

accept responsibility for it; the conceptual pair has, by consequence, become something of an age-old orphan. Created by humanists in the sixteenth century but attributed to the ancient Greeks, the language/dialect distinction rapidly emancipated itself from that Greek patronage. As soon as this emancipation had occurred, it was out there, free for everyone to use and interpret, for better or for worse. Symptomatically, citations of previous definitions of dialect as opposed to language are of a remarkably low frequency, which suggests that there is a reluctance to take part in a direct debate on the interpretation of the conceptual pair. It is moreover telling that criteria have often been attributed to groups of people as large as ‘laypeople’ and ‘linguists’, or vaguer still, to ‘some’ and ‘others’. Mutual intelligibility, for instance, was hailed by Georg von der Gabelentz at the end of the nineteenth century as a criterion of popular origin which linguists should upgrade and make workable. In Greene’s blogpost on the website of The Economist, exactly the opposite occurred; there, mutual intelligibility was promoted as the number one objective criterion to distinguish dialects from languages, which laypeople should take over from linguists (see Chapter 23). Overall, there has been terminological rather than conceptual continuity throughout the ages, from Greek antiquity and Byzantium through the early modern period to modern institutionalized linguistics and lay ideas. It seems, therefore, that in the case of dialect we have a big term rather than a big idea, since it was not the term’s semantics but its formal outlook which has enjoyed enormous success over more than two millennia. Indeed, the term dialect has remained, but numerous interpretations have come and gone. Even if in recent decades the term dialect has aroused suspicion among linguists, mainly owing to the failure to develop an objective metalinguistic concept, its terminological success continues through various derivations and compositions as well as for the sake of tradition—or as the adage goes: old habits die hard. Through my study of the conceptual pair’s history, I hope to have contributed to bringing the long history of language studies back within the horizons of modern linguists, who often assume that the history of their discipline is tantamount to the history of institutionalized linguistics. This book has argued that the deep foundations of a key conceptual distinction were laid several centuries before the nineteenth century. Only by taking earlier developments into account can we grasp more fully the often long and sometimes bumpy road which language studies have taken. The conceptual pair has had a rich and turbulent history, but does it have a future? As elsewhere, here, too, accurate prediction is sheer impossible. Yet the fact that it is out there, known to laypeople and linguists alike, suggests that its usage will not decrease soon. Especially among laypeople and policymakers, the language/dialect opposition may be expected to remain commonly employed, with all the problems

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

300

      

attached to it, while among linguists the general suspicion of it may further develop into an ever-stronger aversion. But it is utterly difficult to make a confident prediction about its fate in academic circles. Even there, the conceptual pair will probably be not abandoned altogether if only because it is so ingrained in many linguists’ mental maps and because it has left indelible traces on linguistic terminology. The fact remains, however, that mainstream linguists today tend to avoid using the term dialect in contrast with language. The main concerns are, on the one hand, the realization that language and dialect appear to be clumsy abstractions of actual linguistic phenomena and, on the other hand, the many undesirable sociopolitical and ideological connotations attached to the conceptual contrast. The latter problem can, I believe, be easily superseded if linguists are more explicit about their own usage and interpretation by clearly defining the terms if they employ them and by abandoning bad citation practices, which are all too common with reference to the conceptual pair (see particularly Maxwell 2018). Explicitness and clarity are all the more indispensable, since there are, as I have asserted, many different interpretations and connotations attached to the terms language and dialect, which bear the load of an age-long and complex history. Trickier is the former problem: is the gap between the reality of language diversity and the concepts of language and dialect too wide to bridge? Since I am first and foremost a historian of linguistics, this is not a question to which I can hope to offer a satisfying answer. To several modern linguists, in any case, this distance has seemed too vast, which is why they avoid the conceptual pair almost like the plague. Only few constructive alternatives have been offered, for instance by William Kretzschmar, who has put forward the appealing proposal of conceptualizing language as a self-organizing complex system. This view might indeed be closer to the truth but remains, of course, an abstraction itself. The question then amounts to asking whether it will ever be possible to reach a satisfying theoretical frame to grasp the inherent diversity of human language. One of the major problems with the conceptual pair is borders in a twofold sense. Firstly, when does a variety become so different from a related variety that it is a different language rather than a related dialect? Secondly, is it at all possible to distinguish between dialects, which often flow into one another, as linguists and dialect geographers have argued in great numbers ever since the 1860s? The delimitation problems suggest that a strictly binary distinction of languages and dialects is impossible to maintain. Does, then, such a division have no basis at all in linguistic evidence? If we would accept this conclusion, we might be faced with more disturbing ponderings. For instance, are we to discard all previous work in which the conceptual pair is put into practice in some way or another? This conclusion is not reasonable, since in the past many linguists assuming the language/dialect contrast have reached valid research results. What is more, recent statistical studies focusing on lexical variation have suggested that there are indeed two major degrees to which speech forms tend to vary, that there might be

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

      

301

something like a dialect and a language pole, and that linguistic diversity tends to organize itself on two major planes (Ran and Wichmann 2018). It is my expectation that linguists in the very near future will be able to fine-tune this statistical approach in determining degrees of linguistic kinship and involve in it different aspects of language, not only the lexicon but also phonology, morphology, grammar, and semantics (cf. already Voegelin and Harris 1951: 327). If such studies would confirm the existence of a dialect and a language pole, perhaps there might be a future for the conceptual pair. In this case, linguistic diversity would have to be reconceptualized as being in three zones flowing into one another: a dialect and a language cluster with an intriguing transition zone in-between. These results may be connected to an assessment of mutual intelligibility and its role: around which point of linguistic distance do two speech forms start being mutually unintelligible? Is mutual intelligibility a conservative force, tying two related speech forms to the dialect cluster? Does gradual loss thereof, for instance through the disappearance of mutual contact, contribute to a shift of these two speech forms through the transition zone to the language cluster? Such analyses of linguistic distance would, of course, remain snapshots in time but reproducible and testable ones. Conducting them over extended periods of time would moreover allow us to chart the nature, degree, and pace of linguistic change. The proposed reconceptualization would also involve taking individual speech forms not as closed systems, as the grand theorists of the twentieth century have usually presupposed, but rather in the complex system sense William Kretzschmar has suggested. Such a perspective would have the advantage of diverting the attention of the seemingly unsolvable puzzle of linguistic borders to the different types of language variation and their causes, properties, and functions. Anyhow, it remains desirable to keep looking for ever more accurate models of language that acknowledge its inherent variability. This search is all the more exigent, since it has become somewhat overshadowed by a great enthusiasm for the specific rather than the general in current linguistic research. Whatever happens with the conceptual pair within linguistics and whether scholars consider it an adequate or relevant abstraction or not, it remains a fact that it is out there for everyone to use and misuse. For this reason alone, I agree with what André Martinet (1954) and other scholars have pointed out in the past century: linguists cannot simply ignore the language/dialect distinction even if they no longer see a future for it in their own discipline. The fact of the matter is that it originated within Renaissance philology and was popularized on a large scale owing to developments in modern history, in particular the rise of the nation-state and general education. As a result, it has become integrated into the wide range of concepts non-specialists can rely on to make sense of their surroundings in everyday life. This everyday usage is why linguists should assume responsibility for the language/dialect pair and its meaning rather than shy away from it as if it were not theirs to engage with. In an age in which science

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 26/8/2020, SPi

302

      

communication is ever more important, this call to responsibility should be welcomed with open arms as an opportunity rather than disowned as a burden. If not, there is a risk that Max Weinreich’s auditor from the Bronx will be right, that dialect will always be language’s poor relation, a language without an army and navy.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

References In early modern books I have retained the original place and publisher names for reasons of findability, but I have added between square brackets the English place names where doubt may arise. van der Aa, A. J. (1867). ‘Hoius (Andreas)’, in Biographisch woordenboek der Nederlanden . . . , continued by K. J. R. van Harderwijk and Dr. G. D. J. Schotel. Haarlem: J. J. van Brederode, 8/2: 975–7. Absillis, Kevin, Jürgen Jaspers, and Sarah Van Hoof (2012). De manke usurpator: Over Verkavelingsvlaams. Ghent: Academia Press. Académie française (1694). Le dictionnaire de l’Académie françoise. Two volumes. A Paris: chez la Veuve de Jean Baptiste Coignard . . . et chez Jean Baptiste Coignard. Adams, James Noel (2007). The Regional Diversification of Latin 200 – 600. Cambridge: Cambridge University Press. Adelung, Johann Christoph (1781). Deutsche Sprachlehre. Berlin: bey Christian Friedrich Voß und Sohn. Adelung, Johann Christoph (1782). Umständliches Lehrgebäude der deutschen Sprache zur Erläuterung der deutschen Sprachlehre für Schulen. Leipzig: verlegts Johann Gottlob Immanuel Breitkopf. Agard, Frederick B. (1971). ‘Language and Dialect: Some Tentative Postulates’, Linguistics 65: 5–24. Ahačič, Kozma (2008). ‘ “Omnis lingua confitebitur Deo”: Writing the First Slovenian Grammar (1584)’, Histoire Épistémologie Langage 30 (2): 93–112. Ahrens, Heinrich Ludolf (1839). De Graecae linguae dialectis. Volume 1. Gottingae [Göttingen]: apud Vandenhoeck et Ruprecht. Albertus, Laurentius (1573). Teutsch Grammatick, oder Sprach-Kunst . . . Augustæ Vindelicorum [Augsburg]: excudebat Michaël Manger. Albrecht, Jörn (2007). Europäischer Strukturalismus: Ein forschungsgeschichtlicher Überblick. 3., erweiterte Auflage. Tübingen: Gunter Narr. Alinei, Mario (1980). ‘Dialect: A Dialectical Approach’, in Göschel et al. (1980: 11–42). Alinei, Mario (1984). ‘«Dialetto»: Un concetto rinascimentale fiorentino’, in Mario Alinei, Lingua e dialetti: Struttura, storia e geografia. Studi linguistici e semiologici. Bologna: Il Mulino, 169–99. Alter, Stephen G. (2005). William Dwight Whitney and the Science of Language. Baltimore, MD & London: The Johns Hopkins University Press. Althamer, Andreas (1536). Commentaria Germaniae in P. Cornelij Taciti Equitis Rom[ani] libellum de situ, moribus, & populis Germanorum. Norimbergæ [Nuremberg]: apud Ioh. Petreium. Amerot, Adrien (1520). Compendiu[m] Græcę gra[m]matices, perspicua breuitate co[m]plectens, quicq[ui]d est octo partium orationis. Louanii [Leuven]: apud Theodoricum Martinum. Amerot, Adrien (1530). De dialectis diuersis declinationum Græcanicarum tam in uerbis quàm nominibus . . . Parisiis [Paris]: ex officina Gerardi Morrhii.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

304



Amirav, Hagit and Hans-Martin Kirn (2011). ‘Introduction’, in Hagit Amirav and HansMartin Kirn (eds), Theodorus Bibliander: De ratione communi omnium linguarum et literarum commentarius. Travaux d’Humanisme et Renaissance 475. Geneva: Droz, –. Anderson, Benedict (2006). Imagined Communities: Reflections on the Origin and Spread of Nationalism. Revised edition. London & New York: Verso. Anon. (1672). ‘Advertisement’, The London Gazette, 22 April 1672, edition 671. Anon. (1743). ‘Daß es nützlich und möglich sey, die niedersächsische Sprache allmählig gar abzuschaffen’, Der deutschen Gesellschaft in Leipzig Nachrichten und Anmerkungen welche die Sprache, Beredsamkeit und Dichtkunst der Deutschen betreffen 3: 383–99. Anon. (1765). ‘Patois’, in Denis Diderot and Jean le Rond d’Alembert (eds), Encyclopédie, ou dictionnaire raisonné des sciences, des arts et des métiers . . . A Neufchastel [Neufchâtel]: chez Samuel Faulche, 12: 174. Antesignanus, Petrus (1554). ‘De dialectis appendix’, in Institutiones linguæ Græcæ, N. Cleonardo autore, cum scholijs P. Antesignani Rapistagnensis. Lugduni [Lyon]: apud Matthiam Bonhomme, 11–16. Armitage, David (2012). ‘What’s the Big Idea? Intellectual History and the Longue Durée’, History of European Ideas 38 (4): 493–507. Ascoli, Graziadio Isaia (1876). ‘P. Meyer e il franco-provenzale’, Archivio glottologico italiano 2: 385–95. Asdal, Kristin and Ingunn Moser (2012). ‘Experiments in Context and Contexting’, Science, Technology & Human Values 37 (4): 291–306. Ashley, Kathleen M. and Véronique Plesch (2002). ‘The Cultural Processes of “Appropriation’’, Journal of Medieval and Early Modern Studies 32 (1): 1–15. Auer, Peter and Jürgen Erich Schmidt, eds (2010). Language and Space: An International Handbook of Linguistic Variation. Handbooks of Linguistics and Communication Science 30.1. Berlin & New York: De Gruyter Mouton. Auroux, Sylvain, ed. (1992). Histoire des idées linguistiques. Volume 2. Philosophie et langage. Liège: Mardaga. Auroux, Sylvain et al., eds (2001). History of the Language Sciences: An International Handbook on the Evolution of the Study of Language from the Beginnings to the Present. Volume 2. Handbücher zur Sprach- und Kommunikationswissenschaft 18. Berlin & New York: Walter de Gruyter. Ax, Wolfram (1987). ‘Quadripertita ratio: Bemerkungen zur Geschichte eines aktuellen Kategoriensystems (Adiectio – Detractio – Transmutatio – Immutatio)’, in Daniel J. Taylor (ed.), The History of Linguistics in the Classical Period. Studies in the History of the Language Sciences 46. Amsterdam & Philadelphia, PA: John Benjamins, 17–40. Ayres-Bennett, Wendy and Magali Seijido (2011). Remarques et observations sur la langue française: Histoire et évolution d’un genre. Histoire et évolution du français 1. Paris: Éditions Classiques Garnier. Bachmann, Andreas [praeses] and Michael Ludovici [respondens] (1625). Περὶ τῶν διαλέκτων διάλεξις ϕιλολογικὴ sive de linguis . . . Lipsiæ [Leipzig]: excudebat Gregorius Ritzsch. Bacon, Roger (1859). Opera quædam hactenus inedita, ed. J. S. Brewer. Volume 1. Rerum Britannicarum medii aevi scriptores. London: Longman, Green, Longman, and Roberts. Bacon, Roger (1900). The ‘Opus majus’ of Roger Bacon, ed. John Henry Bridges. Volume 1. London, Edinburgh & Oxford: Williams and Norgate. Bacon, Roger (1902). The Greek Grammar of Roger Bacon and a Fragment of His Hebrew Grammar, ed. Edmond Nolan and S. A. Hirsch. Cambridge: at the University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



305

Baggioni, Daniel (1997). Langues et nations en Europe. Paris: Payot. Bayly, Anselm (1756). An introduction to languages literary and philosophical . . . Volume 1. London: printed for John and James Rivington. Beauzée, Nicolas (1765). ‘Langue’, in Denis Diderot and Jean le Rond d’Alembert (eds.), Encyclopédie, ou dictionnaire raisonné des sciences, des arts et des métiers . . . A Neufchastel [Neufchâtel]: chez Samuel Faulche, 9: 249–66. Bell, David A. (2003). The Cult of the Nation in France: Inventing Nationalism, 1680–1800. Cambridge, MA & London: Harvard University Press. Bembo, Pietro (1525). Prose . . . nelle quali si ragiona della volgar lingua . . . Vinegia [Venice]: per Giovan Tacuino. Benediktsson, Hreinn (1980). ‘Discussion’, in Torben Thrane et al. (eds), Typology and Genetics of Language: Proceedings of the Rask-Hjelmslev Symposium, Held at the University of Copenhagen 3rd–5th September, 1979. Travaux du Cercle linguistique de Copenhague 20. Copenhagen: The Linguistic Circle of Copenhagen, 17–28. Beroaldo, Filippo (1493). Commentationes . . . in Suetonium Tranquillum. Bononiae [Bologna]: impressit Benedictus Hectoris Bononiensis. Besnier, Pierre (1674). La reunion des langues ou l’art de les apprendre toutes par une seule. A Paris: chez Sebastien Mabre-Cramoisy. Bibliander, Theodore (1542). De optimo genere grammaticorum Hebraicorum commentarius. Basileæ [Basel]: per Hieronymum Curionem. Bibliander, Theodore (1548). De ratione communi omnium linguarum & literaru[m] commentarius. Tiguri [Zürich]: apud Christoph. Frosch. Bichurina, Natalia (2016). Trans-Border Communities in Europe and the Emergence of ‘New’ Languages: From ‘Francoprovençal patois’ to ‘Arpitan’ and ‘Arpitania’. Perpignan: unpublished PhD dissertation (University of Perpignan via Domitia). Bierbach, Christine (1978). Sprache als »Fait social«: Die linguistische Theorie F. de Saussure’s und ihr Verhältnis zu den positivistischen Sozialwissenschaften. Linguistische Arbeiten 59. Tübingen: Max Niemeyer. Biggs, Bruce (1957). ‘Testing Intelligibility among Yuman Languages’, International Journal of American Linguistics 23 (2): 57–62. Blair, Ann M. (2010). Too Much to Know: Managing Scholarly Information before the Modern Age. New Haven, CT & London: Yale University Press. Blench, Roger (2006). ‘Archaeology and Language: Methods and Issues’, in John Bintliff (ed.), A Companion to Archaeology. Malden, MA, Oxford & Carlton: Blackwell, 52–74. Bloomfield, Leonard (1933). Language. New York: Henry Holt. Blount, Thomas (1656). Glossographia . . . London: printed by Tho[mas] Newcomb, and are to be sold by Humphrey Moseley . . . and George Sawbridge. Boberg, Charles, John Nerbonne, and Dominic Watt, eds (2018). The Handbook of Dialectology. Blackwell Handbooks in Linguistics. Hoboken, NJ: Wiley Blackwell. Bodin, Jean (1566). Methodus, ad facilem historiarum cognitionem. Parisiis [Paris]: apud Martinum Iuuenem. Boli, John, Francisco O. Ramirez, and John W. Meyer (1985). ‘Explaining the Origins and Expansion of Mass Education’, Comparative Education Review 29 (2): 145–70. Bolognesi, Giancarlo C. (1953). ‘Sul Περὶ διαλέκτων di Gregorio di Corinto’, Aevum 27 (2): 97–120. Bonfante, Giuliano (1953–4). ‘Ideas on the Kinship of the European Languages from 1200 to 1800’, Journal of World History 1: 679–99. Bopp, Franz (1816). Über das Conjugationssystem der Sanskritsprache in Vergleichung mit jenem der griechischen, lateinischen, persischen und germanischen Sprache. Frankfurt am Main: in der Andreäischen Buchhandlung.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

306



Bopp, Franz (1820). ‘Analytical comparison of the Sanskrit, Greek, Latin, and Teutonic languages, shewing the original identity of their grammatical structure’, Annals of Oriental literature 1: 1–65. Borghini, Vincenzio (1971). ‘Se la diversità della lingua greca è la medesima come la italiana’, in John Robert Woodhouse (ed.), Scritti inediti o rari sulla lingua. Collezione di opere inedite o rare 132. Bologna: commissione per i testi di lingua, 335–47. Borst, Arno (1957–63). Der Turmbau von Babel: Geschichte der Meinungen über Ursprung und Vielfalt der Sprachen und Völker. Four volumes. Stuttgart: Hiersemann. Botley, Paul (2010). Learning Greek in Western Europe, 1396–1529: Grammars, Lexica, and Classroom Texts. Transactions of the American Philosophical Society 100.2. Philadelphia, PA: American Philosophical Society. Bots, Hans and Françoise Waquet (1997). La république des lettres. Paris: Belin. Bourgain, Pascale (1989). ‘Le sens de la langue et des langues chez Roger Bacon’, in Geneviève Contamine (ed.), Traduction et traducteurs au Moyen Âge: Actes du colloque international du CNRS organisé à Paris, Institut de recherche et d’histoire des textes, les 26–28 mai 1986. Documents, études et répertoires [54]. Paris: Centre National de la Recherche Scientifique, 317–31. Bovelles, Charles de (1533). Liber de differentia vulgariu[m] linguaru[m], & Gallici sermonis varietate. Quæ voces apud Gallos sint factitiæ & arbitrariæ, vel barbariæ: quæ item ab origine Latina manarint. De hallucinatione Gallicanorum nominum. Parisiis [Paris]: Ex officina Roberti Stephani. Bréal, Michel (1897). Essai de sémantique (Science des significations). Paris: Hachette. Brekle, Herbert E., Edeltraud Dobnig-Jülch, Hans Jürgen Höller, and Helmut Weiß, eds (1992–2005). Bio-bibliographisches Handbuch zur Sprachwissenschaft des 18. Jahrhunderts: Die Grammatiker, Lexikographen und Sprachtheoretiker des deutschsprachigen Raums mit Beschreibungen ihrer Werke. Eight volumes. Tübingen: Max Niemeyer. Brevini, Franco, ed. (1999). La poesia in dialetto: Storia e testi dalle origini al Novecento. Volume 2. I meridiani. Milan: Arnoldo Mondadori. Brockliss, Laurence (2003). ‘Curricula’, in Hilde de Ridder-Symoens (ed.), A History of the University in Europe. Cambridge: Cambridge University Press, 2: 563–620. Bruce, F. F. (1992). ‘Lycaonia’, in David Noel Freedman (ed.), The Anchor Bible Dictionary. New York: Doubleday, 4: 420–2. Bruña Cuevas, Manuel (1998). ‘L’enseignement de l’r français aux Espagnols (XVIe–XIXe siècles)’, in Teresa García-Sabell Tormo et al. (eds), Les chemins du texte: VI Coloquio da APFFUE (Santiago, 19, 20 e 21 de febreiro de 1997). Cursos e Congresos da Universidade de Santiago de Compostela 110. Santiago de Compostela: Universidade de Santiago de Compostela & A.P.F.F.U.E., 2: 527–39. Brunner, Richard J. (1971). Johann Andreas Schmeller: Sprachwissenschaftler und Philologe. Innsbrucker Beiträge zur Sprachwissenschaft 4. Innsbruck: Institut für Vergleichende Sprachwissenschaft der Universität Innsbruck. Buchanan, George (1582). Rerum Scoticarum historia. Edimburgi [Edinburgh]: apud Alexandrum Arbuthnetum. Budzinski, Laure (2012). ‘Reconsidération étymologique et historique du vocable dialectologie’, Romanica Cracoviensia 12 (1): 11–19. Burke, Peter (2004). Languages and Communities in Early Modern Europe. Cambridge: Cambridge University Press. Burke, Peter (2013). ‘Nationalisms and Vernaculars, 1500–1800’, in John Breuilly (ed.), The Oxford Handbook of the History of Nationalism. Oxford: Oxford University Press, 21–35.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



307

Bursill-Hall, G. L. (1971). Speculative Grammars of the Middle Ages: The Doctrine of partes orationis of the Modistae. Approaches to Semiotics 11. The Hague & Paris: Mouton. Bursill-Hall, G. L. (1975). ‘The Middle Ages’, in Thomas A. Sebeok (ed.), Historiography of Linguistics. Current Trends in Linguistics 13. The Hague & Paris: Mouton, 1: 179–230. Calepino, Ambrogio (1502). Calepinus. Rhegii Lingobardiæ [Reggio Emilia]: i[n]dustria presbyteri Dionysii Berthochi imp[re]ssoris. Calvet, Louis-Jean (1974). Linguistique et colonialisme: Petit traité de glottophagie. Bibliothèque scientifique. Paris: Payot. Camden, William (1595). Institutio Græcæ grammatices compendiaria . . . Londini [London]: excudebat Edm. Bollifant pro Simone Waterson. Camerarius, Joachim (1541). ‘Prooemium in Historias Herodoti . . . ’, in Ἡροδότου λόγοι ἐννέα, οἵπερ ἐπικαλοῦνται μοῦσαι . . . Basileæ [Basel]: in officina Heruagiana, α.2–β.4. Campbell, Lyle (1972). ‘Is a Generative Dialectology Possible?’, Orbis: Bulletin international de documentation linguistique 21 (2): 289–98. Campbell, Lyle (2006). ‘Why Sir William Jones Got It All Wrong, or Jones’ Role in How to Establish Language Families’, Anuario del Seminario de Filología Vasca Julio de Urquijo: International Journal of Basque Linguistics and Philology 40 (1/2): 245–64. Canini, Angelo (1555). Ἑλληνισμός . . . Parisiis [Paris]: apud Guil. Morelium. Casad, Eugene H. (1974). Dialect Intelligibility Testing. Summer Institute of Linguistic Publications in Linguistics and Related Fields 38. Norman, OK: Summer Institute of Linguistics of the University of Oklahoma. Casaubon, Méric (1650). De quatuor linguis commentationis, pars prior: quæ de lingua Hebraica: et, de lingua Saxonica. Londini [London]: typis J. Flesher, sumptibus Ric. Mynne. Caxton, William (1490). ‘[Preface]’, in Vergil, The boke yf Eneydos. [Westminster]: [William Caxton], .i–ii. Celenza, Christopher S. (2009). ‘End Game: Humanist Latin in the Late Fifteenth Century’, in Yanick Maes et al. (eds), Latinitas Perennis. Leiden & Boston, MA: Brill, 2: 201–44. Ceporin, Jakob (1522). Compendium Græcæ grammaticæ. Basileæ [Basel]: apud Valentinu[m] Curionem. Certeau, Michel de, Dominique Julia, and Jacques Revel (2002). Une politique de la langue. La Révolution française et les patois: L’enquête de Grégoire. Nouvelle édition augmentée. Folio: Histoire 117. Paris: Gallimard. Chambers, J. K. and Peter Trudgill (1998). Dialectology. Second edition. Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press. Charlet, Jean-Louis (2013). ‘La culture grecque de Niccolò Perotti’, Cahiers de Recherches Médiévales et Humanistes 25: 259–80. Cheng, Eileen Ka-May (2008). The Plain and Noble Garb of Truth: Nationalism & Impartiality in American Historical Writing, 1784–1860. Athens, GA & London: University of Georgia Press. Chomsky, Noam (1977). Dialogues avec Mitsou Ronat. Traduit de l’américain et présenté par Mitsou Ronat. Dialogues. Paris: Flammarion. Chomsky, Noam (1979). Language and Responsibility, transl. John Viertel. Based on conversations with Mitsou Ronat. New York: Pantheon. Chomsky, Noam (1986). Knowledge of Language: Its Nature, Origin, and Use. Convergence. Westport, CT & London: Praeger. Chomsky, Noam (1992). ‘Explaining Language Use’, Philosophical Topics 20 (1): 205–31. Chomsky, Noam (1997). ‘Language from an Internalist Perspective’, in David Martel Johnson and Christina E. Erneling (eds), The Future of the Cognitive Revolution. New York & Oxford: Oxford University Press, 118–35.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

308



Chrysoloras, Manuel (1512). ‘Ἐρωτήματα τοῦ Χρυσολωρᾶ’, in Ἐρωτήματα τοῦ Χρυσολωρᾶ· Περὶ ἀνωμάλων ῥημάτων . . . Venetijs [Venice]: in ædib[us] Aldi, 3–115. Chytraeus, Nathan (1582). Nomenclator Latinosaxonicus. Rostochii [Rostock]: typis Stephani Myliandri. Clackson, James (2015). ‘Latinitas, Ἑλληνισμός and Standard Languages’, Studi e Saggi Linguistici 53 (2): 309–30. Clement of Alexandria (1550). Κλήμεντος Ἀλεξανδρέως τὰ εὑρισκόμενα ἅπαντα, ed. Pietro Vettori. Florentiæ [Florence]: cudebat . . . Laurentius Torrentinus. Clement of Alexandria (1551). Clementis Alexandrini omnia quae quidem extant opera . . . , transl. Gentien Hervet. Florentiæ [Florence]: Laur. Torrentinus . . . excudebat. Codro, Antonio Urceo (1502). In hoc Codri uolumine haec continentur: Orationes. seu sermones ut ipse appellabat. Epistolæ. Siluæ. Satyræ. Eglogæ. Epigrammata. Bononiæ [Bologna]: eme[n]date ac/ curateq[ue] impressum . . . per Ioannem Antoniu[m] Platonidem. Colombat, Bernard (2008). ‘L’accès aux langues pérégrines dans le Mithridate de Conrad Gessner (1555)’, Histoire Épistémologie Langage 30 (2): 71–92. Colombat, Bernard and Manfred Peters (2009). ‘Introduction’, in Conrad Gessner, Mithridate. Mithridates (1555), ed. Bernard Colombat and Manfred Peters. Travaux d’Humanisme et Renaissance 452. Geneva: Droz, 11–90. Colvin, Stephen (1999). Dialect in Aristophanes and the Politics of Language in Ancient Greek Literature. Oxford: Clarendon Press. Colvin, Stephen (2010). ‘Greek Dialects in the Archaic and Classical Ages’, in Egbert J. Bakker (ed.), A Companion to the Ancient Greek Language. Blackwell Companions to the Ancient World. Malden, MA, Oxford & Chichester: Wiley-Blackwell, 200–12. Consani, Carlo (1991). ΔΙΑΛΕΚΤΟΣ: Contributo alla storia del concetto di ‘dialetto’. Testi linguistici 18. Pisa: Giardini. Consani, Carlo (2000). ‘La nozione di «lingua comune»/«varietà dialettale» nei grammatici tardo-antichi’, in Cristina Vallini (ed.), Le parole per le parole: I logonimi nelle lingue e nel metalinguaggio. Atti del Convegno Napoli, Istituto Universitario Orientale 18–20 dicembre 1997. Rome: Il Calamo, 605–18. Considine, John (2008a). Dictionaries in Early Modern Europe: Lexicography and the Making of Heritage. Cambridge: Cambridge University Press. Considine, John (2008b). ‘Did Andreas Jäger or Georg Caspar Kirchmaier Write the Dissertation De lingua vetustissima Europae (1686)?’, Historiographia Linguistica 35 (1–2): 13–22. Considine, John (2010). ‘Why Was Claude de Saumaise Interested in the Scythian Hypothesis?’, Language & History 53 (2): 81–96. Considine, John (2014). Academy Dictionaries 1600–1800. Cambridge: Cambridge University Press. Considine, John (2017). Small Dictionaries and Curiosity: Lexicography and Fieldwork in Post-Medieval Europe. Oxford: Oxford University Press. Considine, John and Toon Van Hal (2010). ‘Introduction: Classifying and Comparing Languages in Post-Renaissance Europe (1600–1800)’, Language & History 53 (2): 63–9. Constantin, Emile and Ferdinand de Saussure (2005). ‘Linguistique Générale (Cours de M’ Le Professeur de Saussure) Semestre d’hiver 1910–1911’, Cahiers Ferdinand de Saussure 58: 82–290. Cooper, Thomas (1565). Thesaurus linguæ Romanæ & Britannicæ . . . Londini [London]: excusum . . . in ædibus quondam Bertheleti . . . per Henricum VVykes.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



309

[Coquebert de Montbret], [Charles-Étienne or Eugène] (1831). ‘Essai d’un travail sur la géographie de la langue française’, in Mélanges sur les langues, dialectes et patois; renfermant, entre autres, une collection de versions de la parabole de l’enfant prodigue en cent idiomes ou patois différens, presque tous de France; précédés d’un essai d’un travail sur la géographie de la langue française. Paris: au Bureau de l’Almanach du Commerce, 5–29. Cortelazzo, Manlio (1980). I dialetti e la dialettologia in Italia (fino al 1800). Ars linguistica: Commentationes analyticae et criticae 4. Tübingen: Gunter Narr. Coseriu, Eugenio (1975). ‘Stiernhielm, die rumänische Sprache und das merkwürdige Schicksal eines Vaterunsers’, Románica 8: 7–24. Coseriu, Eugenio (1980). ‘“Historische Sprache” und “Dialekt”’, in Göschel (1980: 106–22). Coseriu, Eugenio (1998). ‘Le double problème des unités “dia-s” ’, Les Cahiers δια: Études sur la diachronie et la variation linguistique 1: 9–16. Coseriu, Eugenio and Reinhard Meisterfeld (2003). Geschichte der romanischen Sprachwissenschaft. Volume 1. Tübingen: Gunter Narr. Cottière, Matthieu (1646). De Hellenistis, & lingua Hellenistica, exercitationes secundariæ. [Strasbourg]: typis Rihelianis. Council of Europe (1992). ‘European Charter for Regional or Minority Languages’. European Treaty Series 148. Strasbourg: Council of Europe. Cram, David (1999). ‘Edward Lhuyd and the Doctrine of the Permutation of Letters’, in Gerda Haßler & Peter Schmitter (eds), Sprachdiskussion und Beschreibung von Sprachen im 17. und 18. Jahrhundert. Studium Sprachwissenschaft 32. Münster: Nodus, 317–35. Crastone, Giovanni (1478). Λεξικὸν κατὰ στοιχείων. [Milan]: [Bono Accursio]. Crowley, Tony (2003). Standard English and the Politics of Language. Second edition. Basingstoke & New York: Palgrave Macmillan. Crusius, Martin (1584). Turcogræciæ libri octo. Basileæ [Basel]: per Leonardum Ostenium, Sebastiani Henricpetri impensa. Cysouw, Michael (2014). ‘Languoid, Doculect and Glossonym: Formalizing the Notion “Language”’, Billet. Diversity Linguistics Comment (blog). Online. Available at https:// dlc.hypotheses.org/623. Last accessed 25 July 2020. Cysouw, Michael and Jeff Good (2013). ‘Languoid, Doculect, and Glossonym: Formalizing the Notion “Language”’, Language Documentation & Conservation 7 (December). Available at http://hdl.handle.net/10125/4606. Last accessed 25 July 2020. Dahan, Gilbert, Irène Rosier, and Luisa Valente (1995). ‘L’arabe, le grec, l’hébreu et les vernaculaires’, in Sten Ebbesen (ed.), Sprachtheorien in Spätantike und Mittelalter. Geschichte der Sprachtheorie 3. Tübingen: Gunter Narr, 265–321. Debus, Karl Heinz (1998). Der Gatterer-Apparat. Patrimonia 119. Speyer: Landesarchiv. Delbrück, Berthold (1880). Einleitung in das Sprachstudium: Ein Beitrag zur Geschichte und Methodik der vergleichenden Sprachforschung. Bibliothek indogermanischer Grammatiken 4. Leipzig: Breitkopf und Härtel. Demonet, Marie-Luce (1992). Les Voix du signe: Nature et origine du langage à la Renaissance (1480–1580). Bibliothèque littéraire de la Renaissance: Série 3, 29. Paris & Geneva: Champion & Slatkine. Denecker, Tim (2017). Ideas on Language in Early Latin Christianity: From Tertullian to Isidore of Seville. Supplements to Vigiliae Christianae 142. Leiden & Boston, MA: Brill. Desmet, Piet (1996). La linguistique naturaliste en France (1867–1922): Nature, origine et évolution du langage. Orbis/Supplementa 6. Leuven & Paris: Peeters.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

310



Desmet, Piet, Peter Lauwers, and Pierre Swiggers (2002). ‘Le développement de la dialectologie française avant et après Gilliéron’, in Peter Lauwers, Marie-Rose SimoniAurembou, and Pierre Swiggers (eds), Géographie linguistique et biologie du langage: Autour de Jules Gilliéron. Orbis/Supplementa 20. Leuven & Paris: Peeters, 17–64. De Laet, Johannes (1643). Notæ ad dissertationem Hugonis Grotii de origine gentium Americanarum: et observationes aliquot ad meliorem indaginem difficillimæ illius quæstionis. Amstelodami [Amsterdam]: apud Ludovicum Elzivirium. Dickey, Eleanor (2007). Ancient Greek Scholarship: A Guide to Finding, Reading, and Understanding Scholia, Commentaries, Lexica, and Grammatical Treatises, from Their Beginnings to the Byzantine Period. American Philological Association Classical Resources Series 7. Oxford & New York: Oxford University Press. Dictionnaire de Trévoux (1721). Dictionnaire universel François et Latin . . . Imprimé à Trevoux et se vend à Paris: chez Florentin Delaulne . . . Dionisotti, Carlo (1968). Gli umanisti e il volgare fra Quattro e Cinquecento. Bibliotechina del Saggiatore 29. Florence: Felice Le Monnier. Dohna, Lothar Graf zu (1964). ‘Gatterer, Johann Christoph’, in Neue Deutsche Biographie. Berlin: Duncker & Humblot, 6: 89–91. Droixhe, Daniel (1978). La linguistique et l’appel de l’histoire (1600–1800): Rationalisme et révolutions positivistes. Langue et cultures 10. Geneva: Droz. Drusius, Johannes (1616). Ad uoces Ebraicas Noui Testamenti commentarius duplex . . . , ed. Abel Curiander. Franekeræ Frisiorum [Franeker]: excudebat Fredericus Heynsius. Edwards, John (2009). Language and Identity: An Introduction. Key Topics in Sociolinguistics. Cambridge: Cambridge University Press. Eichhorn, Johann Gottfried (1779). ‘Ueber die verschiedenen Mundarten der Arabischen Sprache’, in John Richardson, Abhandlung über Sprachen, Litteratur und Gebräuche morgenländischer Völker, transl. Friedrich Federau. Leipzig: in der Weygandschen Buchhandlung, 3–72. Eisenstein, Elizabeth L. (1966). ‘Clio and Chronos: An Essay on the Making and Breaking of History-Book Time’, History and Theory 6: 36–64. Elffers, Els (2008). ‘Georg von der Gabelentz and the Rise of General Linguistics’, in Lo van Driel and Theo Janssen (eds), Ontheven aan de tijd: Linguïstisch-historische studies voor Jan Noordegraaf bij zijn zestigste verjaardag. Amsterdam & Münster: Stichting Neerlandistiek & Nodus, 191–200. Eliot, Thomas (1538). The dictionary of syr Thomas Eliot knyght. Londini [London]: in ædibus Thomæ Bertheleti typis impress[um]. Eloy, Jean-Michel (1997). La constitution du picard: Une approche de la notion de langue. Bibliothèque des Cahiers de l’Institut de Linguistique de Louvain 90. Louvain-la-Neuve: Peeters. Eloy, Jean-Michel (1998). ‘Sur l’inventaire des variétés d’oïl de France: Un processus de naissance de langues?’, Travaux de linguistique 9: 3–17. Engler, Rudolf (2000). ‘La géographie linguistique’, in Sylvain Auroux (ed.), Histoire des idées linguistiques. Bruxelles: Mardaga, 3: 239–52. Erasmus, Desiderius (1519). In Nouum Testamentum . . . denuo recognitum, annotationes, ingenti nuper accessione per autorem locupletatæ. Basileæ [Basel]: apud Ioannem Frobenium. Erasmus, Desiderius (1528). De recta Latini Græciq́[ue] sermonis pronu[n]tiatione . . . dialogus . . . Apud inclytam Basilæam [Basel]: in officina Frobeniana. Eriksson, Manne (1971). ‘Sven Hof ’, in Svenskt biografiskt lexikon. Online. Available at https://sok.riksarkivet.se/Sbl/Presentation.aspx?id=13681. Last accessed 10 August 2020. Eskhult, Josef (2015). ‘Albert Schultens (1686–1750) and Primeval Language: The Crisis of a Tradition and the Turning Point of a Discourse’, in Gerda Haßler (ed.), Metasprachliche

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



311

Reflexion und Diskontinuität: Wendepunkte—Krisenzeiten—Umbrüche. Münster: Nodus, 72–94. Eskhult, Josef (2018). ‘Vulgar Latin as an Emergent Concept in the Italian Renaissance (1435–1601): Its Ancient and Medieval Prehistory and Its Emergence and Development in Renaissance Linguistic Thought’, Journal of Latin Linguistics 17 (2): 191–230. Eskhult, Josef (2020). ‘The Concept of Dialect in the Works of Three Seventeenth-Century Linguistic Theorists, Georg Stiernhielm, Claudius Salmasius and Christian Ravius: An Exploratory Survey of Their Intellectual Contexts, Defining Criteria, and Classificatory Aims’, in Van Rooy (2020b). Eskhult, Josef (forthcoming a). Albert Schultens as Teacher and Innovator of Comparative Hebrew Language Studies: Exploratory Survey of Scholarly Authorship and Textual Edition of Unpublished Manuscripts with Introduction, Translation and Reception Studies. Eskhult, Josef (forthcoming b). Georg Stiernhielm (1598–1672) on the Structure, Function and History of Language: Textual Edition of His Theoretical Linguistic Writings with Introduction and Reception Studies. Studia Latina Upsaliensia. Uppsala: Uppsala Universitet. Estienne, Henri (1579). Proiect du livre intitulé De la precellence du langage françois. A Paris: par Mamert Patisson. Estienne, Henri (1582). Hypomneses de Gall[ica] lingua, peregrinis eam discentibus necessariæ . . . [Geneva]: [Henri Estienne]. Estienne, Henri (1587). Dialogus, de bene instituendis Græcæ linguæ studiis. [Geneva]: [Henri Estienne]. Estienne, Robert (1531). Dictionarium, seu Latinæ linguæ thesaurus . . . Parisiis [Paris]: ex officina Roberti Stephani. Facius, Johann Friedrich (1782). Compendium dialectorum Graecarum . . . Norimbergae [Nuremberg]: sumtibus E. C. Grattenaueri. Fasold, Ralph W. (1975). ‘The Bailey Wave Model: A Dynamic Quantitative Paradigm’, in Ralph W. Fasold and Roger W. Shuy (eds), Analyzing Variation in Language: Papers from the Second Colloquium on New Ways of Analyzing Variation. Washington, DC: Georgetown University Press, 27–58. Fausel, Andrea, Jochen Hafner, and Franziska Küenzlen (2002). ‘Sprechen über Sprache und Geschichte: Mittelalterliche und humanistische Entwürfe zur Sprachgeschichtsschreibung’, in Cora Dietl and Dörte Helschinger (eds), Ars und Scientia im Mittelalter und in der Frühen Neuzeit: Ergebnisse interdisziplinärer Forschung. Georg Wieland zum 65. Geburtstag. Tübingen & Basel: A. Francke, 193–214. Ferguson, Gibson (2006). Language Planning and Education. Edinburgh Textbooks in Applied Linguistics. Edinburgh: Edinburgh University Press. Ferté, H. (1902). Rollin: Sa vie, ses œuvres et l’université de son temps. Paris: Hachette. Fischer, Hans (1966). Conrad Gessner (26. März 1516–13. Dezember 1565): Leben und Werk. Neujahrsblatt 168. Zürich: Naturforschenden Gesellschaft in Zürich. Formigari, Lia (2004). A History of Language Philosophies. Studies in the History of the Language Sciences 105. Amsterdam & Philadelphia, PA: John Benjamins. Franck, Jakob (1878). ‘Fulda, Friedrich Karl’, in Allgemeine Deutsche Biographie. Munich & Leipzig: Duncker & Humblot, 8: 192. Frederickx, Eddy and Toon Van Hal (2015). Johannes Goropius Becanus (1519–1573): Brabants arts en taalfanaat. Hilversum: Verloren. Freidin, Robert (2012). ‘A Brief History of Generative Grammar’, in Gillian Russell and Delia Graff Fara (eds), The Routledge Companion to Philosophy of Language. New York & London: Routledge, 895–916.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

312



Fulda, Friedrich Carl (1773). Ueber die beiden Hauptdialecte der Teutschen Sprachen: Eine Preisschrift . . . Leipzig: bey Bernhard Christoph Breitkopf und Sohn. Furetière, Antoine (1690). Dictionaire universel, contenant generalement tous les mots françois . . . Three volumes. A la Haye, et à Rotterdam [The Hague & Rotterdam]: chez Arnout & Reinier Leers. Gabelentz, Georg von der (1901). Die Sprachwissenschaft, ihre Aufgaben, Methoden und bisherigen Ergebnisse, ed. Albrecht Graf von der Schulenburg. Zweite, vermehrte und verbesserte Auflage. Leipzig: Chr. Herm. Tauchnitz. Galiani, Ferdinando (1779). Del dialetto napoletano. Napoli [Naples]: per Vincenzo Mazzola-Vocola. Gatterer, Johann Christoph (1770). ‘Vorrede, von der historischen Benutzung der Sprachen’, in Fortsetzung der Algemeinen Welthistorie durch eine Geselschaft von Gelehrten in Teutschland und Engeland ausgefertiget. Halle: bey Johann Justinus Gebauer, 34: 3–16. Gatterer, Johann Christoph (1771). Einleitung in die synchronistische Universalhistorie zur Erläuterung seiner synchronistischen Tabellen. Göttingen: im Verlag der Wittwe Vandenhoek. Gaza, Theodore (1495). In hoc uolumine haec insunt. Theodori Introductiuæ gra[m]matices libri quatuor . . . Venetiis [Venice]: impressum . . . in ædibus Aldi Romani. Gedike, Friedrich (1779). Gedanken über Purismus und Sprachbereicherung. Berlin: gedrukt bei George Jakob Dekker. Gedike, Friedrich (1782). ‘Ueber Dialekte, besonders die griechischen’, Berlinsches Magazin der Wissenschaften und Künste 1 (2): 1–26. Gedike, Friedrich (1794). ‘Über deutsche Dialekte: Erste Vorlesung’, Beiträge zur deutschen Sprachkunde: Vorgelesen in der Königlichen Akademie der Wissenschaften zu Berlin 1: 292–332. Geeraerts, Dirk and Gitte Kristiansen (2015). ‘Variationist Linguistics’, in Ewa Dabrowska and Dagmar Divjak (eds), Handbook of Cognitive Linguistics. Handbücher zur Sprachund Kommunikationswissenschaft 39. Berlin & Boston, MA: De Gruyter Mouton, 366–89. Geijer, Erik Gustaf (1850). ‘Äreminne öfver Georg Stjernhjelm . . . : Uppläst i Svenska Akademien d. 20 Dec. 1840’, in Svenska akademiens handlingar ifrån år 1796. Stockholm: P. A. Norstedt & Söner, 23: 151–71. Gerald of Wales (1868). Itinerarium Kambriæ, et descriptio Kambriæ, ed. James F. Dimock. Giraldi Cambrensis Opera 6. London: Longmans, Green, Reader, and Dyer. Gessner, Conrad (1543). ‘De utilitate ac præstantia Græcæ linguæ, in omni genere studiorum, ad candidos lectores præfatio’, in Lexicon Græcolatinum. Basileæ [Basel]: per Hieronymum Curionem, .2–7. Gessner, Conrad (1551). Historiæ animalium lib. I de quadrupedibus uiuiparis. Tiguri [Zürich]: apud Christ. Froschouerum. Gessner, Conrad (1555). Mithridates: De differentiis linguarum tum ueterum tum quæ hodie apud diuersas nationes in toto orbe terraru[m] in usu sunt . . . obseruationes. Tiguri [Zürich]: excudebat Froschouerus. Gessner, Conrad (1561). ‘Ad lectorem præfatio super hoc dictionario Germanicolatino . . . ’, in Josua Maaler, Die Teütsch spraach . . . Dictionarium Germanicolatinum nouum. Tiguri [Zürich]: excudebat Christophorus Froschouerus, *.3–7. Giard, Luce (1984). ‘Du latin médiéval au pluriel des langues, le tournant de la Renaissance’, Histoire Épistémologie Langage 6 (1): 35–55. Giard, Luce (1992). ‘L’entrée en lice des vernaculaires’, in Auroux (1992: 206–25).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



313

Gierl, Martin (2012). Geschichte als präzisierte Wissenschaft: Johann Christoph Gatterer und die Historiographie des 18. Jahrhunderts im ganzen Umfang. Fundamenta Historica 4. Stuttgart & Bad Cannstatt: frommann-holzboog. Gierl, Martin (2013). ‘Change of Paradigm as a Squabble between Institutions: The Institute of Historical Sciences, the Society of Sciences, and the Separation of Cultural and Natural Sciences in Göttingen in the Second Half of the Eighteenth Century’, in André Holenstein et al. (eds), Scholars in Action: The Practice of Knowledge and the Figure of the Savant in the 18th Century. Leiden & Boston, MA: Brill, 1: 267–87. Giese, Albert (1837). Ueber den Aeolischen Dialekt: Zwei Bücher. Berlin: in der G. Finckeschen Buchhandlung. Gill, Alexander (1619). Logonomia Anglica. Qua gentis sermo facilius addiscitur. Londini [London]: excudit Iohannes Beale. Gill, Alexander (1972). Logonomia Anglica (1619), ed. Bror Danielsson and Arvid Gabrielson, transl. Robin C. Alston. Two volumes. Acta Universitatis Stockholmiensis: Stockholm Studies in English 26–7. Stockholm: Almqvist & Wiksell. Gipper, Sonja (2014). ‘From Inferential to Mirative: An Interaction-Based Account of an Emerging Semantic Extension’, in Evie Coussé and Ferdinand von Mengden (eds), Usage-Based Approaches to Language Change. Studies in Functional and Structural Linguistics 69. Amsterdam & Philadelphia, PA: John Benjamins, 83–116. Gitner, Adam (2018). ‘Sardismos: A Rhetorical Term for Bilingual or Plurilingual Interaction?’, The Classical Quarterly 68 (2): 689–704. Glarean, Heinrich (1538). In C. Iulii Caesaris clariss. Roman. imperatoris Commentarios de bello Gallico, ac ciuili . . . annotationes. Lugduni [Lyon]: apud Seb. Gryphium. Goebl, Hans (1982). Dialektometrie: Prinzipien und Methoden des Einsatzes der Numerischen Taxonomie im Bereich der Dialektgeographie. Österreichische Akademie der Wissenschaften, Philosophisch-historische Klasse, Denkschriften 157. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Goebl, Hans (1983). ‘ “Stammbaum” und “Welle”: Vergleichende Betrachtungen aus numerisch-taxonomischer Sicht’, Zeitschrift für Sprachwissenschaft 2 (1): 3–44. Goebl, Hans (2003). ‘Graziadio Isaia Ascoli, Carlo Battisti e il ladino: Breve controstoria di una pietra dello scandalo della linguistica a cavallo tra Otto- e Novecento’, in Antonio Trampus and Ulrike Kindl (eds), I linguaggi e la storia. Bologna: Il Mulino, 273–98. Goldhagen, Hermann (1752). ‘Appendix I. De dialectis’, in Institutiones linguae Latinae et Graecae . . . Mannhemii [Mannheim]: ex typographejo electorali aulico apud Nicolaum Pierron, 32–4 [second sequence of pagination]. Goropius Becanus, Johannes (1580). Opera . . . hactenus in lucem non edita . . . Antuerpiæ [Antwerp]: excudebat Christophorus Plantinus. Gorter, Durk and Jasone Cenoz (2012). ‘Legal Rights of Linguistic Minorities in the European Union’, in Peter M. Tiersma and Lawrence M. Solan (eds), The Oxford Handbook of Language and Law. Oxford: Oxford University Press, 261–71. Göschel, Joachim, Pavle Ivić, and Kurt Kehr, eds (1980). Dialekt und Dialektologie: Ergebnisse des internationalen Symposions ‘Zur Theorie des Dialekts’ Marburg/Lahn, 5.–10. September 1977. Zeitschrift für Dialektologie und Linguistik: Beihefte. Neue Folge 26. Wiesbaden: Franz Steiner. Göschel, Joachim, Norbert Nail, and Gaston Van der Elst, eds (1976). Zur Theorie des Dialekts: Aufsätze aus 100 Jahren Forschung mit biographischen Anmerkungen zu den Autoren. Zeitschrift für Dialektologie und Linguistik: Beihefte. Neue Folge 16. Wiesbaden: Franz Steiner.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

314



Grafton, Anthony (1983). Joseph Scaliger: A Study in the History of Classical Scholarship. Volume 1. Oxford-Warburg Studies. Oxford: Clarendon Press. Grafton, Anthony (2007). What Was History? The Art of History in Early Modern Europe. Cambridge: Cambridge University Press. Grafton, Anthony (2009). ‘Codex in Crisis: The Book Dematerializes’, in Anthony Grafton, Worlds Made by Words: Scholarship and Community in the Modern West. Cambridge, MA & London: Harvard University Press, 288–324. Gretser, Jakob (1593). Institutionum de octo partibus orationis, syntaxi et prosodia Græcorum, libri tres. Ingolstadii [Ingolstadt]: excudebat Dauid Sartorius. Grimm, Jacob (1819). Deutsche Grammatik. Volume 1. Göttingen: in der Dieterichschen Buchhandlung. Grinberg, Martine (2006). Écrire les coutumes. Paris: Presses Universitaires de France. Groddeck, Benjamin [praeses] and Michael Gottlieb Treuge [respondens] (1747). Commentatio academica de natura dialectorum ad linguam Hebraicam et Arabicam applicata . . . Vitembergae [Wittenberg]: typis Schlomachianis. Guillén de Brocar, Arnao (1514). ‘Introductio qua[m] breuissima ad Grecas litteras’, in Arnao Guillén de Brocar (ed.), Nouum Testamentum Grece & Latine in Academia Complutensi nouiter impressum. [Alcalá de Henares]: [Arnao Guillén de Brocar], a.i–g.iii. Guldi, Jo and David Armitage (2015). The History Manifesto. Cambridge: Cambridge University Press. Haas, Johann Gottfried (1780). Grichische Grammatik kurz und deutlich insonderheit den Anfängern zum Besten abgefasset. Leipzig: im Schwickertschen Verlage. Haas, Walter (1994). Provinzialwörter: Deutsche Idiotismensammlungen des 18. Jahrhunderts. Historische Wortforschung: Untersuchungen zur Sprach- und Kulturgeschichte des Deutschen in seinen europäischen Bezügen 3. Berlin & New York: Walter de Gruyter. Hackett, Jeremiah (2013). ‘Roger Bacon’, in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy. Online. Available at https://plato.stanford.edu/archives/ spr2015/entries/roger-bacon/. Last accessed 14 January 2020. Hainsworth, J. B. (1967). ‘Greek Views of Greek Dialectology’, Transactions of the Philological Society 65: 62–76. Halle, Morris (1962). ‘Phonology in Generative Grammar’, Word 18 (1–3): 54–72. Haloinus, Georgius (1978 [1533]). De restauratione linguae Latinae libri III, ed. Constantinus Matheeussen. Bibliotheca scriptorum Graecorum et Romanorum Teubneriana. Leipzig: B. G. Teubner. Hammarström, Harald (2005). ‘LINGUIST List 16.2637: General Linguistics: Gordon (2005)’, The LINGUIST List. 12 September 2005. Online. Available at https:// linguistlist.org/issues/16/16-2637.html. Last accessed 14 January 2020. Hammarström, Harald (2008). ‘Counting Languages in Dialect Continua Using the Criterion of Mutual Intelligibility’, Journal of Quantitative Linguistics 15 (1): 34–45. Harris, Roy and Talbot J. Taylor (1997). Landmarks in Linguistic Thought I: The Western Tradition from Socrates to Saussure. Second edition. Routledge History of Linguistic Thought Series. London & New York: Routledge. Haßler, Gerda (2009a). ‘Analogie’, in Haßler and Neis (2009: .658–74). Haßler, Gerda (2009b). ‘Normierung’, in Haßler and Neis (2009: .674–718). Haßler, Gerda and Cordula Neis, eds (2009). Lexikon sprachtheoretischer Grundbegriffe des 17. und 18. Jahrhunderts. Two volumes. Berlin & New York: Walter de Gruyter. Haugen, Einar (1966). ‘Dialect, Language, Nation’, American Anthropologist 68 (4): 922–35. Helladius, Alexander (1714). Status praesens ecclesiae Graecae . . . [Altdorf]: [s. n.].

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



315

Hickerson, Harold, Glen D. Turner, and Nancy P. Hickerson (1952). ‘Testing Procedures for Estimating Transfer of Information among Iroquois Dialects and Languages’, International Journal of American Linguistics 18 (1): 1–8. Hickey, Raymond (2010). ‘Attitudes and Concerns in Eighteenth-Century English’, in Raymond Hickey (ed.), Eighteenth-Century English: Ideology and Change. Studies in English Language. Cambridge: Cambridge University Press, 1–20. Hinskens, Frans, Peter Auer, and Paul Kerswill (2005). ‘The Study of Dialect Convergence and Divergence: Conceptual and Methodological Considerations’, in Peter Auer et al. (eds), Dialect Change: Convergence and Divergence in European Languages. Cambridge: Cambridge University Press, 1–48. Hirschi, Caspar (2012). The Origins of Nationalism: An Alternative History from Ancient Rome to Early Modern Germany. Cambridge: Cambridge University Press. Hockett, Charles Francis (1958). A Course in Modern Linguistics. New York: Macmillan. Hodson, Jane (2018). ‘Literary Uses of Dialect’, in David Duff (ed.), The Oxford Handbook of British Romanticism. Oxford: Oxford University Press, 513–28. Hof, Sven (1772). Dialectus Vestrogothica, ad illustrationem aliquam linguae Suecanae, veteris et hodiernae, dissertatione philologica et vocabulorum Vestrogothicorum indice explanata. Stockholmiae [Stockholm]: typis & impensis Joh. Arv. Carlbohm. Hofmann, Konrad (1883). ‘Roger Bacon über die französischen und englischen Mundarten’, Romanische Forschungen 1 (3): 427–8. Hoius, Andreas (1620). ‘Dissertatiuncula, de linguæ Græcæ dialectorum sedibus et coloniis’, in Andreas Hoius, Orthoepeia, siue de Germana ac recta linguæ Græcæ, et obiter Latinæ, pronuntiatione. Duaci [Douai]: ex officina Ioannis Bogardi, 95–104. van der Horst, Joop (2008). Het einde van de standaardtaal: Een wisseling van Europese taalcultuur. Amsterdam: J. M. Meulenhoff. Hottinger, Johann Heinrich (1661). Etymologicum Orientale; sive Lexicon harmonicum ἑπτάγλωττον . . . Francofurti [Frankfurt am Main]: sumptibus Joh. Wilhelmi Ammonij, & Wilhelmi Serlini. Hovdhaugen, Even (1990). ‘Una et Eadem: Some Observations on Roger Bacon’s Greek Grammar’, in G. L. Bursill-Hall et al. (eds), De ortu grammaticae: Studies in Medieval Grammar and Linguistic Theory in Memory of Jan Pinborg. Studies in the History of the Language Sciences 43. Amsterdam & Philadelphia, PA: John Benjamins, 117–31. Hovdhaugen, Even, ed. (1996). And the Word Was God: Missionary Linguistics and Missionary Grammar. Münster: Nodus. Hovdhaugen, Even, Fred Karlsson, Carol Henriksen, and Bengt Sigurd (2000). The History of Linguistics in the Nordic Countries. Helsinki: Societas Scientiarum Fennica. Hoven, René (1985). Bibliographie de trois auteurs de grammaires grecques contemporains de Nicolas Clénard: Adrien Amerot, Arnold Oridryus, Jean Varennius. Livre – Idées – Société 7. Aubel: P. M. Gason. Howell, James (1642). Instructions for forreine travel . . . London: printed by T. B. for Humphrey Mosley. Howell, James (1650). Epistolæ Ho-Elianæ: Familiar letters domestic and forren . . . The second edition . . . London: printed by W. H. for Humphrey Moseley. [Howes], [Thomas] (1776). Critical observations on books, antient and modern. Volume 1. London: printed by W. Richardson. Howlet, Richard (1552). Abcedarium Anglicolatinum. Londini [London]: ex officina Gulielmi Riddel. Hoyt, David L. (2006). ‘Dialects of Modernization in France and Italy, 1865–1900’, in David L. Hoyt and Karen Oslund (eds), The Study of Language and the Politics of Community in Global Context. Lanham et al.: Lexington Books, 85–118.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

316



Hugo Schuchardt Archiv. Online. Available at http://schuchardt.uni-graz.at/. Last accessed 14 January 2020. Hüllen, Werner (2001). ‘Reflections on Language in the Renaissance’, in Martin Haspelmath et al. (eds), Language Typology and Language Universals . . . Handbooks of Linguistics and Communication Science 20. Berlin & New York: Walter de Gruyter, 1: 210–22. Hyde, Thomas (1767 [1691]). ‘Oratio de linguæ Arabicæ antiquitate, præstantia et utilitate’, in Gregory Sharpe (ed.), Syntagma dissertationum quas olim auctor doctissimus Thomas Hyde . . . separatim edidit . . . Oxonii [Oxford]: e Typographeo Clarendoniano, 2: 451–9. Iggers, Georg G. (2011). ‘Introduction’, in Leopold von Ranke, The Theory and Practice of History, ed. Georg G. Iggers, transl. Wilma A. Iggers. London & New York: Routledge, xi–xlv. Impey, Oliver and Arthur MacGregor, eds (2001). The Origins of Museums: The Cabinet of Curiosities in Sixteenth and Seventeenth Century Europe. London: House of Stratus. Jablonski, Paul Ernst (1714). Disquisitio de lingua Lycaonica . . . Berolini [Berlin]: literis Wesselianis. Jankowsky, Kurt R. (1972). The Neogrammarians: A Re-Evaluation of Their Place in the Development of Linguistic Science. Janua Linguarum: Series Minor 116. The Hague & Paris: Mouton. Jenkins, Gary W. (2006). John Jewel and the English National Church: The Dilemmas of an Erastian Reformer. Aldershot & Burlington: Ashgate. Jones, William (1799). The works of Sir William Jones. London: printed for G. G. and J. Robinson . . . and R. H. Evans . . . . Jones, William Jervis (2001). ‘Early Dialectology, Etymology and Language History in German-Speaking Countries’, in Auroux et al. (2001: 1105–15). de Jonge, Henk J. (1980). De bestudering van het Nieuwe Testament aan de Noordnederlandse universiteiten en het Remonstrants Seminarie van 1575 tot 1700. Verhandelingen der Koninklijke Nederlandse Akademie van Wetenschappen, Afdeling Letterkunde, Nieuwe Reeks 106. Amsterdam, Oxford & New York: Noord-Hollandse Uitgevers Maatschappij. Joseph, John E. (1987). Eloquence and Power: The Rise of Language Standards and Standard Languages. Open Linguistics Series. London: Frances Pinter. Joseph, John E. (1996). ‘ “Undoubtedly a Powerful Influence”: Victor Henry’s Antinomies Linguistiques (1896), with an Annotated Translation of the First Chapter’, Language & Communication 16 (2): 117–44. Joseph, John E. (2004). Language and Identity: National, Ethnic, Religious. Basingstoke & New York: Palgrave Macmillan. Joseph, John E. (2006). Language and Politics. Edinburgh Textbooks in Applied Linguistics. Edinburgh: Edinburgh University Press. Joseph, John E. (2012). Saussure. Oxford & New York: Oxford University Press. Junius, Franciscus (1665). Quatuor D. N. Jesu Christi euangeliorum versiones perantiquæ duæ, Gothica scil[icet] et Anglo-Saxonica . . . Dordrechti [Dordrecht]: typis & sumptibus Junianis. Excudebant Henricus & Joannes Essæi. Kals, Jan Willem (1752). Dissertatio philologica de linguæ Hebrææ natalibus punctisque vocalibus ad eam docendam discendamque summe necessariis. Volume 1. Oxonii [Oxford]: e Theatro Sheldoniano, impensis Joannis Barrett, et prostant apud J. Baldwin . . . & apud Tho. Merrill. Kamusella, Tomasz (2015). Creating Languages in Central Europe during the Last Millennium. Basingstoke & New York: Palgrave Macmillan.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



317

Kamusella, Tomasz (2016). ‘The History of the Normative Opposition of “Language versus Dialect”: From Its Graeco-Latin Origin to Central Europe’s Ethnolinguistic NationStates’, Colloquia Humanistica 5: 164–88. ten Kate, Lambert (1723). Aenleiding tot de kennisse van het verhevene deel der Nederduitsche sprake. Two volumes. Tot Amsterdam: by Rudolph en Gerard Wetstein. Keil, Heinrich, ed. (1855–80). Grammatici Latini. Seven volumes. Lipsiae [Leipzig]: in aedibus B. G. Teubneri. Kibbee, Douglas and Marcus Keller (2018). ‘Introduction’, in Scipion Dupleix, Liberté de la langue françoise dans sa pureté, ed. Douglas Kibbee and Marcus Keller. Remarques et observations sur la langue française 2. Paris: Classiques Garnier, 7–167. Kirchmaier, Georg Caspar [praeses] and Johannes Crusius [respondens] (1684). De originibus & caussis literaturæ Græcæ, ejusdemq́[ue] dialectis . . . Wittenbergæ [Wittenberg]: typis Christiani Schrödteri. Kirchmaier, Georg Caspar [praeses] and Andreas Jäger [respondens] (1686). De lingva vetustissima Europæ, Scytho-Celtica et Gothica . . . Wittenbergæ [Wittenberg]: typis Christiani Schrödteri. Kirchmaier, Georg Wilhelm [praeses] and Georg Friedrich Thryllitsch [respondens] (1709). Eruditorum de dialecto Graecorum communi sententias . . . excutiendas proponet . . . Thryllitius. VVitembergae [Wittenberg]: literis Christiani Schroedteri. Klippi, Carita (2020). ‘Dialect and Its Technologies: Communities of Practice in GalloRomance Dialectology’, in Van Rooy (2020b: 107–35). Kloss, Heinz (1952). Die Entwicklung neuer germanischer Kultursprachen von 1800 bis 1950. Munich: Pohl. Kloss, Heinz (1967). ‘ “Abstand Languages” and “Ausbau Languages” ’, Anthropological Linguistics 9 (7): 29–41. Kloss, Heinz (1978). Die Entwicklung neuer germanischer Kultursprachen seit 1800. 2., erweiterte Auflage. Sprache der Gegenwart 37. Düsseldorf: Pädagogischer Verlag Schwann. Koerner, E. F. K. (1989). ‘August Schleicher and Linguistic Science in the Second Half of the 19th Century’, in E. F. K. Koerner (ed.), Practicing Linguistic Historiography. Studies in the History of the Language Sciences 50. Amsterdam & Philadelphia, PA: John Benjamins, 324–75. Kökeritz, Helge (1938). ‘Alexander Gill (1621) on the Dialects of South and East England’, Studia Neophilologica 11 (1–2): 277–88. Kramer, Johannes (1989). ‘Klassische Sprache und Substandard in der Geschichte des Griechischen’, in Günter Holtus and Edgar Radtke (eds), Sprachlicher Substandard II: Standard und Substandard in der Sprachgeschichte und in der Grammatik. Konzepte der Sprach- und Literaturwissenschaft 44. Tübingen: Max Niemeyer, 55–82. Kretzschmar, Jr, William A. (2009). The Linguistics of Speech. Cambridge & New York: Cambridge University Press. Kümin, Beat (2015). ‘The Uses of Space in Early Modern History—An Afterword’, in Paul Stock (ed.), The Uses of Space in Early Modern History. Palgrave Studies in Cultural and Intellectual History. New York: Palgrave Macmillan, 227–34. Kurath, Hans (1972). Studies in Area Linguistics. Bloomington, IN & London: Indiana University Press. Labov, William (1963). ‘The Social Motivation of a Sound Change’, WORD 19 (3): 273–309. Labov, William (1972). ‘The Social Stratification of (r) in New York City Department Stores’, in William Labov, Sociolinguistic Patterns. Philadelphia, PA: University of Pennsylvania Press, 43–54.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

318



Labov, William (1973). ‘Where Do Grammars Stop?’, in R. W. Shuy (ed.), Report on the Twenty-Third Annual Round Table Meeting on Linguistics and Language Studies. Georgetown Monograph Series on Language and Linguistics 25. Washington, DC: Georgetown University Press, 43–88. Labov, William (2010). Principles of Linguistic Change. Volume 3. Language in Society 39. Malden, MA, Oxford & Chichester: Wiley-Blackwell. Laes, Christian and Toon Van Houdt (2013). ‘Taalkunde met een ranzig randje: Wetenschap en ideologie in het debat tussen Grotius en De Laet over de oorsprong en de status van de Indianen (1641–1644)’, in Toon Van Hal et al. (eds), De tuin der talen: Taalstudie en taalcultuur in de Lage Landen, 1450–1750. Orbis Linguarum: Studies op het gebied van de taalbeschrijving, de taalfilosofie, en de taal- en cultuurgeschiedenis 3. Leuven, Paris & Walpole, MA: Peeters, 125–52. Lallot, Jean (1995). ‘Analogie et pathologie dans la grammaire alexandrine’, Lalies 15: 109–23. Lambert, Frédéric (2009). ‘Les noms des langues chez les Grecs’, Histoire Épistémologie Langage 31 (2): 15–27. Lamers, Han (2015). Greece Reinvented: Transformations of Byzantine Hellenism in Renaissance Italy. Brill’s Studies in Intellectual History 247. Leiden: Brill. Lamers, Han, Toon Van Hal, and Sebastiaan G. Clercx (2020). ‘How to Deal with Scholarly Forgetting in the History of the Humanities: Starting Points for Discussion’, History of Humanities 5 (1): 5–29. Lang, Jürgen (1982). Sprache im Raum: Zu den theoretischen Grundlagen der Mundartforschung. Unter Berücksichtigung des Rätoromanischen und Leonesischen. Beihefte zur Zeitschrift für romanische Philologie 185. Tübingen: Max Niemeyer. Larramendi, Manuel de (1729). El impossible vencido: Arte de la lengua bascongada. En Salamanca: por Antonio Joseph Villargordo Alcaráz. Lascaris, Constantine (1502). De octo partibus orationis liber primus . . . [Venice]: [Aldus Manutius]. Law, Vivien (2003). The History of Linguistics in Europe: From Plato to 1600. Cambridge: Cambridge University Press. Leerssen, Joep (2007). ‘Imagology: History and Method’, in Manfred Beller and Joep Leerssen (eds), Imagology. The Cultural Construction and Literary Representation of National Characters: A Critical Survey. Amsterdam & New York: Rodopi, 17–32. Leibniz, Gottfried Wilhelm (1710). ‘Brevis designatio meditationum de originibus gentium, ductis potissimum ex indicio linguarum’, Miscellanea Berolinensia . . . 1: 1–16. Leskien, August (1876). Die Declination im Slavisch-litauischen und Germanischen. Preisschriften gekrönt und herausgegeben von der fürstlich jablonowski’schen Gesellschaft zu Leipzig. Nr. XI der historisch-nationalökonomischen Section 19. Leipzig: Bei S. Hirzel. Leskien, August (1879). ‘Giese, Albert’, in Allgemeine Deutsche Biographie. Leipzig: Duncker & Humblot, 9: 151. Leu, Urs B. and Peter Opitz, eds (2019). Conrad Gessner (1516–1565): Die Renaissance der Wissenschaften/The Renaissance of Learning. Berlin & Boston, MA: De Gruyter Oldenbourg. Leu, Urs B., Raffael Keller, and Sandra Weidmann (2008). Conrad Gessner’s Private Library. History of Science and Medicine Library 5. Leiden & Boston, MA: Brill. Leusden, Johannes (1656). Philologus Hebræus . . . Ultrajecti [Utrecht]: apud Meinardum à Dreunen.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



319

Leusden, Johannes (1670). ‘Dissertatio undecima, de dialectis Novi T. in genere & in specie’, in Philologus Hebræo-Græcus generalis . . . Ultrajecti [Utrecht]: ex officinâ Anthonii Smytegelt, 83–91. Liddell, Henry George and Robert Scott (1996). A Greek–English Lexicon. Revised and augmented by Sir Henry Stuart Jones with the assistance of Roderick McKenzie and with the cooperation of many scholars. Oxford: Clarendon Press. van der Linde, J. M. (1987). Jan Willem Kals: Leraar der hervormden; advocaat van indiaan en neger. Kampen: J. H. Kok. Löffler, Heinrich (1974). Probleme der Dialektologie: Eine Einführung. Darmstadt: Wissenschaftliche Buchgesellschaft. Lusignan, Serge (1986). Parler vulgairement: Les intellectuels et la langue française aux XIIIe et XIVe siècles. Études médiévales. Paris & Montréal: Librairie philosophique J. Vrin & Les Presses de l’Université de Montréal. Luther, Martin (1913). Werke: Tischreden, 1531–1546, ed. Ernst Kroker. Volume 2. D. Martin Luthers Werke: Kritische Gesamtausgabe 2. Weimar: Hermann Böhlaus Nachfolger. Luther, Martin (1916). Werke: Tischreden, 1531–1546, ed. Ernst Kroker. Volume 4. D. Martin Luthers Werke: Kritische Gesamtausgabe 2. Weimar: Hermann Böhlaus Nachfolger. Luther, Martin (1919). Tischreden 1531–46, ed. Ernst Kroker. Volume 5. D. Martin Luthers Werke: Kritische Gesamtausgabe 2. Weimar: Hermann Böhlaus Nachfolger. Lynn, Michael R. (2006). Popular Science and Public Opinion in Eighteenth-Century France. Studies in Early Modern European History. Manchester: Manchester University Press. Macaulay, Ronald (2010). ‘Dialect’, in Mirjam Fried et al. (eds), Variation and Change: Pragmatic Perspectives. Handbook of Pragmatics Highlights 6. Amsterdam & Philadelphia, PA: John Benjamins, 61–72. Machan, Tim William (2003). English in the Middle Ages. Oxford: Oxford University Press. Maierù, Alfonso (1983). ‘Dante al crocevia?’, Studi medievali 24: 735–48. Malkiel, Yakov (1967). ‘Uriel Weinreich (1926–1967)’, Language 43 (2): 605–7. Manutius, Aldus (1496). ‘Aldus Manucius Basianâs Romanus studiosis omnibus .S .P. D.’, in Aldus Manutius et al. (1496: *.ii–iii). Manutius, Aldus et al., eds (1496). Θησαυρός. Κέρας ἀμαλθείας, καὶ κῆποι Ἀδώνιδος. Thesaurus Cornu copiæ. & Horti Adonidis. Venetiis [Venice]: in domo Aldi Romani. Marazzini, Claudio (1993). ‘Le teorie’, in Luca Serianni and Pietro Trifone (eds), Storia della lingua italiana. Torino: Giulio Einaudi, 1: 231–329. Marenbon, John (2004). ‘Mare, William de La [William de Mara] (Fl. 1272–1279), Franciscan Friar and Theologian’, in Oxford Dictionary of National Biography. Online. Available at www.oxforddnb.com/view/10.1093/ref:odnb/9780198614128.001.0001/odnb9780198614128-e-18025. Last accessed 14 January 2020. Martinet, André (1954). ‘Dialect’, Romance Philology 8: 1–11. Matheeussen, Constant (1974). Een zestiende-eeuws pleidooi tegen de grammatica: Georgius Haloinus en zijn Restauratio linguae Latinae (1533). Leuven: unpublished PhD dissertation (KU Leuven). Matheeussen, Constant (1977). ‘Haloinus (Halewijn, Halewyn, ook Cominius), Georgius’, in Nationaal Biografisch Woordenboek. Brussel: Paleis der Academiën, 7: 310–17. Mattheier, Klaus, ed. (1983). Aspekte der Dialekttheorie. Tübingen: Max Niemeyer. Maxwell, Alexander (2018). ‘When Theory Is a Joke: The Weinreich Witticism in Linguistics’, Beiträge zur Geschichte der Sprachwissenschaft 28 (2): 263–92.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

320



May, Stephen (2011). ‘Language Policy’, in Michael Grenfell, Bourdieu, Language and Linguistics. London & New York: Continuum, 147–69. Mayer, Bartholomaeus (1629). Philologiæ sacræ pars prima . . . Lipsiæ [Leipzig]: sumtibus Gothofredi Grosii . . . Excudebat Johann-Albertus Minzelius. Mayrhofer, Manfred (1983). Sanskrit und die Sprachen Alteuropas: Zwei Jahrhunderte des Widerspiels von Entdeckungen und Irrtümern. Göttingen: Vandenhoeck & Ruprecht. Mazzarella-Farao, Francesco (1779). La Neoellenopedia o sia il nuovo metodo per erudire la gioventù nel greco linguaggio. Volume 1. Napoli [Naples]: nella stamperia porsiliana. Mazzocco, Angelo (1993). Linguistic Theories in Dante and the Humanists: Studies of Language and Intellectual History in Late Medieval and Early Renaissance Italy. Brill’s Studies in Intellectual History 38. Leiden, New York & Cologne: Brill. McCain, Stewart (2018). The Language Question under Napoleon. War, Culture and Society, 1750–1850. Cham: Palgrave Macmillan. McWhorter, John (2016). ‘What’s a Language, Anyway?’, The Atlantic, 19 January 2016. Online. Available at www.theatlantic.com/international/archive/2016/01/differencebetween-language-dialect/424704/. Last accessed 14 January 2020. McWhorter, John (2019). ‘Language vs. Dialect’, Lexicon Valley, 6 August 2019. Online. Available at https://slate.com/podcasts/lexicon-valley/2019/08/john-mcwhorter-onlanguage-versus-dialect>. Last accessed 14 January 2020. Meillet, Antoine (1903). Introduction à l’étude comparative des langues indo-européennes. Paris: Hachette. Meillet, Antoine (1908). Les dialectes indo-européens. Collection linguistique publiée par la Société de linguistique de Paris 1. Paris: Honoré Champion. Meillet, Antoine (1924). ‘Introduction’, in Antoine Meillet and Marcel Cohen (eds), Les langues du monde. Collection linguistique publiée par la Société de linguistique de Paris 16. Paris: Édouard Champion, 1–18. Meillet, Antoine (1925). La méthode comparative en linguistique historique. Oslo, Leipzig, Paris, London & Cambridge, MA: H. Aschehoug and Co. (W. Nygaard), Otto Harrassowitz, Honoré Champion, Williams & Norgate & Harvard University Press. Melanchthon, Philipp (1518). Institutiones Graecae grammaticae. Tubingæ & Hagnoæ [Tübingen & Haguenau]: ex Academia Anshelmiana. Merton, Robert K. (1968). Social Theory and Social Structure. Enlarged edition. New York: The Free Press. Merula, Paulus (1605). Cosmographiæ generalis libri tres . . . [Leiden]: ex Officina Plantiniana Raphelengi F. Metcalf, George J. (2013). On Language Diversity and Relationship from Bibliander to Adelung, ed. Toon Van Hal and Raf Van Rooy. Studies in the History of the Language Sciences 120. Amsterdam & Philadelphia, PA: John Benjamins. Meyer, John W., Francisco O. Ramirez, and Yasemin Nuholu Soysal (1992). ‘World Expansion of Mass Education, 1870–1980’, Sociology of Education 65 (2): 128–49. Meyer-Lübke, Wilhelm (1908). ‘Roger Bacon über die französischen Mundarten’, Zeitschrift für französische Sprache und Literatur 32: 175. Michaeler, Karl Joseph (1776). Tabulae parallelae antiquissimarum Teutonicae linguae dialectorum . . . Oeniponte [Innsbruck]: formis Academicis Wagnerianis. Mickey, Katherine (1981). ‘Dialect Consciousness and Literary Language: An Example from Ancient Greek’, Transactions of the Philological Society 79: 35–66. Miechowita, Maciej (1517). Tractatus de duabus Sarmatijs Asiana: & Europiana & de contentis in eis. Cracouię [Cracow]: impressum . . . opera & impensis prouidi viri d[omi]ni Ioannis Haller.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



321

Mokyr, Joel (2002). The Gifts of Athena: Historical Origins of the Knowledge Economy. Princeton, NJ: Princeton University Press. Molland, George (2004). ‘Bacon, Roger (c.1214–1292?)’, in Oxford Dictionary of National Biography. Online. Available at www.oxforddnb.com/view/article/1008. Last accessed 10 August 2020. Moos, Peter von, ed. (2008). Zwischen Babel und Pfingsten: Sprachdifferenzen und Gesprächsverständigung in der Vormoderne (8.–16. Jahrhundert) . . . Zürich & Berlin: LIT. Morhof, Daniel Georg (1685). De Patavinitate Liviana liber: Ubi de urbanitate & peregrinitate sermonis Latini universè agitur. Kiloni [Kiel]: literis ac sumptibus Joachimi Reumanni. Morpurgo Davies, Anna (1987). ‘The Greek Notion of Dialect’, Verbum 10: 7–28. Morpurgo Davies, Anna (1998). History of Linguistics, ed. Giulio C. Lepschy. Volume 4. Longman Linguistics Library. London & New York: Longman. Mosellanus, Petrus (1527). In M[arci] Fab[ii] Quintiliani Rhetoricas institutiones annotationes. Basileæ [Basel]: apud Adamum Petrum. Muller, Jean-Claude (1984a). ‘Quelques repères pour l’histoire de la notion de vocabulaire de base dans le précomparatisme’, Histoire Épistémologie Langage 6 (2): 37–43. Muller, Jean-Claude (1984b). ‘Saumaise, Monboddo, Adelung: Vers la grammaire comparée’, in Sylvain Auroux et al. (eds), Matériaux pour une histoire des théories linguistiques . . . Travaux et recherches. Lille: Université de Lille III, diffusion P.U.L., 389–96. Muller, Jean-Claude (1986). ‘Early Stages of Language Comparison from Sassetti to Sir William Jones (1786)’, Kratylos 31: 1–31. Müller, Roman (2001). Sprachbewusstsein und Sprachvariation im lateinischen Schrifttum der Antike. Munich: C. H. Beck. Murray, Robert W. (2010). ‘Language and Space: The Neogrammarian Tradition’, in Auer and Schmidt (2010: .70–87). Mylius, Abraham (1612). Lingua Belgica . . . Lugduni Batavorum [Leiden]: pro Bibliopolio Commeliniano, excudebant . . . Vlricus Cornelij & G. Abrahami. Myricaeus, Johann Caspar (1619). Grammaticæ Syro-Chaldææ libri duo . . . [Geneva]: ex typographεῖα Syra Petri de la Rouiere. Nat, Jan (1929). De studie van de Oostersche talen in Nederland in de 18e en de 19e eeuw. Purmerend: J. Muusses. Neis, Cordula (2009). ‘Korruption’, in Haßler and Neis (2009: .567–82). Newton, Brian (1972). The Generative Interpretation of Dialect: A Study of Modern Greek Phonology. Cambridge Studies in Linguistics 8. Cambridge: at the University Press. Nibbe, Johann Barthold (1725). Dialectologia paradigmatica oder Gründliche Anweisung zu denen 4. Haupt-dialectis der Griechischen Sprache . . . Neu-Brandenburg: verlegts Joh. George Schönborn. Nicholas of Lyra (1471). [Postilla litteralis super Matthaeum et Ioannem]. [Rome]: [Konrad Sweinheim & Arnold Pannartz]. Niebaum, Hermann and Jürgen Macha (2006). Einführung in die Dialektologie des Deutschen. 2., neubearbeitete Auflage. Germanistische Arbeitshefte 37. Tübingen: Max Niemeyer. Noordegraaf, Jan (2014). ‘Verre verwanten: D. C. Hesseling (1859–1941) over taal en taalverandering’, in Freek Van de Velde et al. (eds), Patroon en argument: Een dubbelfeestbundel bij het emeritaat van William Van Belle en Joop van der Horst. Leuven: Universitaire Pers, 523–35.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

322



Ogilvie, Brian W. (2006). The Science of Describing: Natural History in Renaissance Europe. Chicago, IL & London: The University of Chicago Press. Osthoff, Hermann and Karl Brugmann (1878). ‘Vorwort’, Morphologische Untersuchungen auf dem Gebiete der indogermanischen Sprachen 1: –. Oxford English Dictionary. Online. Available at www.oed.com/. Last accessed 10 August 2020. Padley, G. A. (1988). Grammatical Theory in Western Europe 1500–1700: Trends in Vernacular Grammar II. Cambridge: Cambridge University Press. Papy, Jan, ed. (2018). The Leuven Collegium Trilingue 1517–1797: Erasmus, Humanist Educational Practice and the New Language Institute Latin – Greek – Hebrew. Leuven, Paris & Bristol, CT: Peeters. Pascale, Carlo (1598). Legatus. Rothomagi [Rouen]: apud Raphaelem Parvivallium. Pasor, Georg (1632). Syllabus brevis Grecarum Novi Testamenti dialectorum. Franekeræ [Franeker]: excudebat Fredericus Heynsius. Paul, Hermann (1880). Principien der Sprachgeschichte. Halle: Max Niemeyer. Pennisi, Antonino (2009). ‘Galiani, Ferdinando’, in Stammerjohann (2009: .508). Percival, W. Keith (1992). ‘La connaissance des langues du monde’, in Auroux (1992: 226–38). Perotti, Niccolò (1489). Cornucopiæ seu come[n]tariorum [sic] linguæ Latinæ. Venetiis [Venice]: per magistrum Paganinum de Paganinis Brixiensem. Persons, Robert (1585). A Christian directorie guiding men to their salvation . . . [S. l.]: [s. n.]. Peters, Manfred (1970–1). Conrad Gessner als Germanist und Linguist. Four volumes. Ghent: unpublished PhD dissertation (Rijksuniversiteit Gent). Peters, Manfred (1972). ‘Sprachwandel und Sprachnorm in Conrad Gessners Mithridates und in seiner Vorrede zu Josua Maalers Teütsch spraach’, Archiv für das Studium der neueren Sprachen und Literaturen 208 (4–6): 256–66. Peucer, Caspar (1562). Tertia pars Chronici Carionis, a Carolo Magno, ubi Philippus Melanthon desijt, usq[ue] ad Fridericum Secundum. VVitebergae [Wittenberg]: excudebat Petrus Seitz. Pfaffel, Wilhelm (1981). Quartus gradus etymologiae: Untersuchungen zur Etymologie Varros in ‘De lingua Latina’. Beiträge zur klassischen Philologie 131. Königstein/Ts. [Königstein im Taunus]: Anton Hain. Pfeiffer, August [praeses] and Johann Georg Martini [respondens] (1663). Diatribe philologica de lingua Galilæa, per quam D. Petrus agnitus fuisse legitur . . . Wittebergæ [Wittenberg]: typis Johannis Haken. Pfeiffer, August [praeses] and Johann Georg Martini [respondens] (1699). ‘Exercitatio decima de dialecto Galilæa, per quam D. Petrus agnitus fuisse legitur . . . ’, in August Pfeiffer, Dubia vexata Script[urae] Sacræ . . . Editio quarta. Lipsiæ [Leipzig]: apud Joh. Frider. Gleditschium, 206–16 [second sequence of pagination]. Phillips, Edward (1658). The new world of English words: or a general dictionary . . . London: printed by E. Tyler, for Nath. Brooke. Piccolomini, Enea Silvio (1960). Ausgewählte Texte aus seinen Schriften . . . , ed./transl. Berthe Widmer. Basel & Stuttgart: Benno Schwabe. Pierce, Joe E. (1952). ‘Dialect Distance Testing in Algonquian’, International Journal of American Linguistics 18 (4): 203–10. Pinborg, Jan (1967). Die Entwicklung der Sprachtheorie im Mittelalter. Münster: Aschendorff. Pirazzini, Daniela (2013). Theorien und Methoden der romanischen Sprachwissenschaft. Romanistische Arbeitshefte 59. Berlin & Boston, MA: Walter de Gruyter.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



323

Polák, Václav (1954). ‘Contribution à l’étude de la notion de langue et de dialecte’, Orbis: Bulletin international de documentation linguistique 3 (1): 89–98. Polier de Bottens, Antoine-Noé [auctor et defendens] (1739). Dissertatio philologica qua disquiritur de puritate dialecti Arabicæ, comparate cum puritate dialecti Hebrææ, in relatione ad antediluvianam linguam . . . Lugduni in Batavis [Leiden]: apud Johannem Luzac. Poole, Matthew (1683). Annotations upon the Holy Bible . . . Volume 1. London: printed by John Richardson, for Thomas Parkhurst . . . Pop, Sever (1950). La dialectologie: Aperçu historique et méthodes d’enquêtes linguistiques. Two volumes. Université de Louvain: Recueil de travaux d’histoire et de philologie 3e série 38. Louvain & Gembloux: chez l’auteur & J. Duculot. Poser, William J. and Lyle Campbell (1992). ‘Indo-European Practice and Historical Methodology’, in Laura A. Buszard-Welcher et al. (eds), Proceedings of the Eighteenth Annual Meeting of the Berkeley Linguistics Society (February 14–17, 1992): General Session and Parasession on the Place of Morphology in a Grammar. Berkeley, CA: Berkeley Linguistic Society, 214–36. Pott, August Friedrich (1856). Die Ungleichheit menschlicher Rassen hauptsächlich vom sprachwissenschaftlichen Standpunkte, unter besonderer Berücksichtigung von des Grafen von Gobineau gleichnamigem Werke . . . Lemgo & Detmold: Meyer’sche Hofbuchhandlung. Pott, August Friedrich (1974 [1884–90]). Einleitung in die Allgemeine Sprachwissenschaft (1884–1890) together with Zur Literatur der Sprachenkunde Europas (Leipzig, 1887), ed. E. F. K. Koerner. Amsterdam Classics in Linguistics, 1800–1925 10. Amsterdam: John Benjamins. Pozzi, Mario (1978). Trattatisti del Cinquecento. Milan: Riccardo Ricciardi. Püschel, Ulrich (2009). ‘Fulda, Friedrich Carl’, in Stammerjohann (2009: .498). Puttenham, George (1589). The arte of English poesie . . . At London: printed by Richard Field. Pyle, Cynthia M. (2000). ‘Conrad Gessner on the Spelling of His Name’, Archives of Natural History 27 (2): 175–86. R. L. G. (2014). ‘How a Dialect Differs from a Language: The Economist Explains’. The Economist (Online), 16 February 2014. Online. Available at www.economist.com/theeconomist-explains/2014/02/16/how-a-dialect-differs-from-a-language. Last accessed 10 August 2020. Ran, Qibin and Søren Wichmann (2018). ‘Zenyang qufen yuyan yu fangyan: Jiyu hexin cihui de juli jisuan fangfa tansuo [How to Distinguish Languages and Dialects: A Quantitative Approach]’, Yuyan Zhanlue Yanjiu [Chinese Journal of Language Policy and Planning] 3 (2): 50–8. Raphelengius, Franciscus (1983). ‘349. 84 05 18 R. Franciscus Raphelengius sr. (Antwerpen) aan Lipsius (Leiden)’, in M. A. Nauwelaerts and S. Sué (eds), Iusti Lipsi Epistolae 2. Brussel: Paleis der Academiën, 121–3. Rask, Rasmus (2013 [1818]). Investigation of the Origin of the Old Norse or Icelandic Language. New edition of the 1993 English translation by Niels Ege, with an introduction by Frans Gregersen. Amsterdam Classics in Linguistics, 1800–1925 18. Amsterdam & Philadelphia, PA: John Benjamins. Rastell, John (1566). A treatise intitled, beware of M. Iewel. Antuerpiæ [Antwerp]: ex officina Ioannis Fouleri. Ravis, Christian (1646). Orthographiæ et analogiæ (vulgò etymologiæ) Ebraicæ delineatio juxta vocis partes abstractas I. consonas. II. vocales. III. Accentus . . . Amstelodami [Amsterdam]: apud Joannem Janssonium.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

324



Ravis, Christian (1650). A generall grammer for the ready attaining of the Ebrew, Samaritan, Calde, Syriac, Arabic, and the Ethiopic languages . . . London: printed by W. Wilson for Tho. Slater and Tho. Humington. Ravis, Christian [praeses et auctor] (1656). Disputatio de Dudaim Rubenis . . . Ubsaliæ [Uppsala]: excudebat Johannes Pauli. Ray, John (1678). A collection of English proverbs . . . The second edition enlarged by the addition of many hundred English . . . Cambridge: printed by John Hayes . . . for W. Morden. Regis, Riccardo (2012). ‘Note su koinè’, Romanische Forschungen 124 (1): 3–16. Rhellicanus, Johannes (1543). In C. Iulii Cæsaris dictatoris uiri disertissimi, et Auli Hirtij, seu Oppij, Commentaria de bello Gallico, ciuili Pompeiano, Alexandrino, Africano, & Hispaniensi, annotationes . . . Basileæ [Basel]: per Hieronymum Curionem. Ribier, Guillaume (1666). ‘Remarques sur ladite ville de Soleurre et sur la diversité des langues d’Allemagne’, in Guillaume Ribier (ed.), Lettres et mémoires d’estat des roys, princes, ambassadeurs et autres ministres, sous les regnes de François premier, Henry II. et François II. . . . A Paris: chez François Clouzier . . . et la vefve Aubouyn, 1: 25–8. Rice, Eugene F. and Anthony Grafton (1994). The Foundations of Early Modern Europe, 1460–1559. Second edition. New York: Norton. Richey, Michael (1743). Idioticon Hamburgense sive Glossarium vocum Saxonicarum quae populari nostra dialecto Hamburgi maxime frequentantur . . . Hamburgi [Hamburg]: apud Conr. Koenigium. Richey, Michael (1755). Idioticon Hamburgense oder Wörter-Buch, zur Erklärung der eigenen, in und üm Hamburg gebräuchlichen, Nieder-Sächsischen Mund-Art. Jetzo vielfältig vermehret . . . Hamburg: verlegt von Conrad König. Rider, John (1589). Bibliotheca scholastica: A double dictionarie . . . Oxford: printed by Joseph Barnes. Rijcke, Theodorus (1684). ‘Dissertatio de primis Italiæ colonis et Æneæ adventu’, in Lucas Holstenius, Notæ et castigationes posthumæ in Stephani Byzantii Ἐθνικά, Quæ vulgo Περὶ πόλεων inscribuntur . . . , ed. Theodorus Rijcke. Lugd[uni] Batavorum [Leiden]: Apud Jacobum Hackium, 399–467. Rizzo, Silvia (1990). ‘Petrarca, il latino e il volgare’, Quaderni petrarcheschi 7: 7–40. Robins, R. H. (1967). A Short History of Linguistics. Longman Linguistics Library 6. London: Longman. Rocher, Rosane (2001). ‘The Knowledge of Sanskrit in Europe until 1800’, in Auroux et al. (2001: 1156–63). Roest, Bert (2015). Franciscan Learning, Preaching and Mission c. 1220–1650: Cum scientia sit donum Dei, armatura ad defendendam sanctam Fidem catholicam . . . Leiden & Boston, MA: Brill. Rollin, Charles (1726). De la maniere d’enseigner et etudier les belles lettres, par raport à l’esprit & au cœur. Volume 1. A Paris: chez Jacques Estienne. Ronsard, Pierre de (1565). Abbregé de l’art poëtique françois. A Paris: chez Gabriel Buon. Rosier-Catach, Irène (1997). ‘Roger Bacon and Grammar’, in Jeremiah Hackett (ed.), Roger Bacon and the Sciences: Commemorative Essays. Studien und Texte zur Geistesgeschichte des Mittelalters 57. Leiden, New York & Cologne: Brill, 67–102. Rotolo, Vincenzo (1973–4). ‘L’opinione di F. Filelfo sul greco volgare’, Rivista di studi bizantini e neoellenici 20–1 (n. s. 10–11): 85–107. Rousseau, Jean-Jacques (1781 [1755]). ‘Essai sur l’origine des langues, où il est parlé de la mélodie & de l’imitation musicale’, in Œuvres posthumes de Jean-Jaques Rousseau . . . Volume 3. Genève [Geneva]: [s. n.], 3: 211–327.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



325

Rubin, Donald L. (1992). ‘Nonlanguage Factors Affecting Undergraduates’ Judgments of Nonnative English-Speaking Teaching Assistants’, Research in Higher Education 33 (4): 511–31. Ruddiman, Walter, ed. (1783). ‘To query XV.–“What are the causes of a difference in dialect among people who speak the same language?”’, The Edinburgh weekly magazine 57 (September): 334. Ruhig, Philipp (1745). Betrachtung der littauischen Sprache in ihrem Ursprunge, Wesen und Eigenschaften . . . Königsberg [modern Kaliningrad]: druckts und verlegts Johann Heinrich Hartung. Sabellicus, Marcus Antonius (1490). Suetonius cum commento. Venetiis [Venice]: per Baptistam de Tortis. Sacré, Dirk (1994). ‘Andreas Hoius’ sterfjaar (met een noot over zijn familie)’, Biekorf: Westvlaams archief voor geschiedenis, archeologie, taal- en volkskunde 94: 387–92. Sansone, Mario (1999). Da Bembo a Galiani: Il dibattito sulla lingua in Italia, ed. Francesco Tateo. Biblioteca di Filologia Romanza 41. Bari: Adriatica. Sapir, Edward (1921). Language: An Introduction to the Study of Speech. New York: Harcourt, Brace and Company. Sapir, Edward (1931). ‘Dialect’, in Encyclopaedia of the Social Sciences. New York: Macmillan, 5: 123–6. Saporta, Sol (1965). ‘Ordered Rules, Dialect Differences, and Historical Processes’, Language 41 (2): 218–24. Saumaise, Claude de (1643a). De Hellenistica commentarius, controversiam de lingua Hellenistica decidens, & plenissimè pertractans originem et dialectos Græcæ linguæ. Lugd[uni] Batavor[um] [Leiden]: ex Officinâ Elseviriorum. Saumaise, Claude de (1643b). Funus linguæ Hellenisticæ, siue Confutatio exercitationis de Hellenistis et lingua Hellenistica. Lugduni Batavorum [Leiden]: ex Officina Ioannis Maire. Saussure, Ferdinand de (1995 [1916]). Cours de linguistique générale, ed. Charles Bally and Albert Sechehaye. Édition critique préparée par Tullio de Mauro. Postface de Louis-Jean Calvet. Paris: Payot. Scaliger, Joseph Justus (1594). Epistola de vetustate et splendore gentis Scaligeræ, et Iul[ii] Cæs[aris] Scaligeri vita. Lugduni Batavorum [Leiden]: ex officina Plantiniana. Scaliger, Joseph Justus (1610a). ‘Diatriba de Europæorum linguis’, in Isaac Casaubon (ed.), Opuscula varia antehac non edita. Parisiis [Paris]: apud Hadrianum Beys, 119–22. Scaliger, Joseph Justus (1610b). ‘Diatriba de hodiernis Francorum linguis’, in Isaac Casaubon (ed.), Opuscula varia antehac non edita. Parisiis [Paris]: apud Hadrianum Beys, 123–6. Schäfer, Gottfried Heinrich (1811). Gregorii Corinthii et aliorum grammaticorum libri de dialectis linguae Graecae . . . Lipsiae [Leipzig]: sumtibus Io. Aug. Gottl. Weigel. Scheuerer, Franz Xaver (1995). Zum philologischen Werk J. A. Schmellers und seiner wissenschaftlichen Rezeption: Eine Studie zur Wissenschaftsgeschichte der Germanistik. Studia Linguistica Germanica 37. Berlin & New York: Walter de Gruyter. Schilter, Johann (1728). Thesaurus antiquitatum Teutonicarum, ecclesiasticarum, civilium, literariarum. Præfationem præmisit Joannes Frickius. Volume 3. Ulmæ [Ulm]: sumptibus Danielis Bartholomæi & filii. Schleicher, August (1861). Compendium der vergleichenden grammatik der indogermanischen sprachen. Two volumes. Weimar: Hermann Böhlau. Schleicher, August (1863). Die Darwinsche Theorie und die Sprachwissenschaft. Weimar: Hermann Böhlau.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

326



Schlieben-Lange, Brigitte (1992). ‘Reichtum, Energie, Klarheit und Harmonie: Die Bewertung der Sprachen in Begriffen der Rhetorik’, in Susanne R. Anschütz (ed.), Texte, Sätze, Wörter und Moneme: Festschrift für Klaus Heger zum 65. Geburtstag. Heidelberg: Heidelberger Orientverlag, 571–86. Schmeller, Johann Andreas (1821). Die Mundarten Bayerns grammatisch dargestellt. Munich: Karl Thienemann. Schmidt, Erasmus (1604). Tractatus de dialectis Græcorum principalibus, quæ sunt in parte λέξεως . . . VVittebergæ [Wittenberg]: imprimebat Laurentius Seuberlich. Impensis Samuel Selfisch. Schmidt, Erasmus (1615). ‘Discursus de pronunciatione Græca, contra νεόϕυτον’, in Erasmus Schmidt (ed.), Cyrilli vel ut alii volunt Johannis Philoponi opusculum utilißimum de differentiis vocum Græcarum quoad tonum, spiritum, genus, &c. Witebergæ [Wittenberg]: typis Richterianis, sumptibus B. Corvini, 213–55. Schmidt, Johannes (1872). Die verwandtschaftsverhältnisse der indogermanischen sprachen. Weimar: Hermann Böhlau. Schmitt, Charles B. (1983). Aristotle and the Renaissance. Cambridge, MA: Harvard University Press. Schnell, Rüdiger (2008). ‘Vom Nicht- und Missverstehen im Mittelalter: Zu Hindernissen sprachlicher und nichtsprachlicher Interaktion im Mittelalter’, in Moos (2008: 567–90). Schoppe, Caspar (1636). Consultationes de scholarum & studiorum ratione deque prudentiæ & eloquentiæ parandæ modis. Patavii [Padua]: apud Paulum Frambottum. Schottel, Justus Georg (1663). Ausführliche Arbeit von der teutschen Haubt Sprache . . . Braunschweig: gedrukt und verlegt durch Christoff Friederich Zilligern. Schrambke, Renate (2010). ‘Language and Space: Traditional Dialect Geography’, in Auer and Schmidt (2010: .87–107). Schrieckius, Adrianus (1615). Monitorum secundorum libri V . . . Ypris Flandrorum [Ypres]: ex Officina typographica Francisci Belletti. Schuchardt, Hugo (1900). Über die Klassifikation der romanischen Mundarten: ProbeVorlesung gehalten zu Leipzig am 30. April 1870. Graz: [s. n.]. Schuchardt, Hugo (1922). Zur Kenntnis des Baskischen von Sara (Labourd). Abhandlungen der Preussischen Akademie der Wissenschaften. Jahrgang 1922: Philosophischhistorische Klasse 1. Berlin: Verlag der Akademie der Wissenschaften. Schuchardt, Hugo (1928). Hugo Schuchardt-Brevier: Ein Vademecum der allgemeinen Sprachwissenschaft, ed. Leo Spitzer. Zweite erweiterte Auflage. Mit einem Porträt Hugo Schuchardts. Halle (Saale): Max Niemeyer. Schultens, Albert (1732). Oratio altera de linguæ Arabicæ antiquissima origine, intima ac sororia cum lingua Hebræa cognatione, nullisque seculis præflorata puritate . . . Lugduni Batavorum [Leiden]: apud Samuelem Luchtmans. Schultens, Albert (1737). Institutiones ad fundamenta linguæ Hebrææ . . . Lugduni Batavorum [Leiden]: apud Johannem Luzac. Schultens, Albert (1738a [1729]). ‘Oratio prima de linguæ Arabicæ antiquissima origine, intima ac sororia cum lingua Hebræa affinitate, nullisque seculis præflorata puritate . . . ’, in Schultens (1738b: separate pagination). Schultens, Albert (1738b). Originum Hebræarum tomus secundus . . . Lugduni Batavorum [Leiden]: apud Samuelem Luchtmans. Schultens, Albert (1738c). Vetus et regia via Hebraïzandi, asserta contra novam et metaphysicam hodiernam. Lugduni Batavorum [Leiden]: apud Joannem Luzac.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



327

Schultens, Albert (1739). Excursus tertius ad editionem primam et secundam dissertationis de lingua primæva, ejusque additamentum apologeticum. Lugduni Batavorum [Leiden]: apud Johannem Luzac. Schultens, Albert (1748). ‘Præfatio’, in Albert Schultens (ed.), Proverbia Salomonis . . . Lugduni Batavorum [Leiden]: apud Johannem Luzac, –. Schwartz, Christian Gottlieb [praeses] and Abraham Helm [respondens] (1702). Disputatio de causis dialectorum, speciatim Graecarum . . . Vitembergae [Wittenberg]: prelo Gerdesiano. Sechehaye, Albert (1908). Programme et méthodes de la linguistique théorique: Psychologie du langage. Paris, Leipzig & Geneva: Honoré Champion, Otto Harrassowitz & A. Eggimann. Séguy, Jean (1973). ‘La dialectométrie dans l’Atlas linguistique de la Gascogne’, Revue de linguistique romane 37: 1–24. Sergeev, Michail (2019). ‘Der Mithridates (1555) zwischen Sprachmuseum und neulateinischem Onomastikon: Einige Überlegungen zur Konzeption und zum Genre des Gessnerschen Handbuchs’, in Leu and Opitz (2019: 517–32). Sèrio, Luigi (1780). Lo vernacchio: Resposta a lo dialetto napoletano. Napole [Naples]: [s. n.]. Shapiro, Michael C. and Harold F. Schiffman (1981). Language and Society in South Asia. Columbia, MO: South Asia Books. Shelford, Robert (1635). Five pious and learned discourses . . . Cambridge: printed by the printers to the Universitie of Cambridge. Siebenborn, Elmar (1976). Die Lehre von der Sprachrichtigkeit und ihren Kriterien: Studien zur antiken normativen Grammatik. Amsterdam: B. R. Grüner. Silverstein, Michael (1971). ‘Whitney on Language’, in Michael Silverstein (ed.), Whitney on Language. Cambridge, MA & London: MIT Press, x–xxiii. Simon, Fabien Dimitri (2011). Sortir de Babel: Une République des Langues en quête d’une «langue universelle» à la Renaissance et à l’Âge classique? Rennes: unpublished PhD dissertation (Université Rennes 2). Skinner, Quentin (1969). ‘Meaning and Understanding in the History of Ideas’, History and Theory 8 (1): 3–53. Springer, Katharina B. and Ragnar K. Kinzelbach (2008). Das Vogelbuch von Conrad Gessner (1516–1565): Ein Archiv für avifaunistische Daten. Berlin & Heidelberg: Springer. Stachowski, Marek (2013). ‘Miechowita’s Knowledge of East European Languages (Mainly Hungarian, Lithuanian and Tatar) Based on His Tractatus de duabus Sarmatiis (1517)’, Studia Linguistica Universitatis Iagellonicae Cracoviensis 130: 309–16. Stammerjohann, Harro, ed. (2009). Lexicon Grammaticorum: A Bio-Bibliographical Companion to the History of Linguistics. Second edition, revised and enlarged. Tübingen: Niemeyer. Stanihurst, Richard (1577). ‘A treatise contayning a playne and perfect description of Irelande, with an introduction . . . ’, in Raphael Holinshed, The historie of Irelande from the first inhabitation thereof, unto the yeare 1509 . . . At London: imprinted for John Hunne, 2: 1–28. Stanton, Domna C. and Rebecca M. Wilkin (2010). ‘Volume Editors’ Introduction’, in Gabrielle Suchon, A Woman Who Defends All the Persons of Her Sex: Selected Philosophical and Moral Writings. The Other Voice in Early Modern Europe. Chicago, IL & London: The University of Chicago Press, 1–51.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

328



Stapleton, Thomas (1566). A returne of untruthes upon M. Jewelles replie. In Antwerpe [Antwerp]: printed . . . by John Latius. Stevens, Benjamin (2006–7). ‘Aeolism: Latin as a Dialect of Greek’, The Classical Journal 102 (2): 115–44. Stevin, Simon (1586). ‘Uytspraeck vande weerdicheyt der Duytschen tael’, in Simon Stevin, De beghinselen der weeghconst. Tot Leyden [Leiden]: inde druckerye van Christoffel Plantijn, by Françoys van Raphelingen, b.1–d.4. Stiernhielm, Georg (1669). Babel destructa, seu Runa Suethica . . . Stockholmiæ [Stockholm]: [s. n.]. Stiernhielm, Georg (1671). ‘De linguarum origine præfatio’, in Georg Stiernhielm (ed.), D. N. Jesu Christi SS. Evangelia ab Ulfila Gothorum in Moesia episcopo circa annum à nato Christo CCCLX. ex Græco Gothicè translata . . . Stockholmiæ [Stockholm]: typis Nicolai Wankif, a.3–f.4. Stiernhielm, Georg (1937). ‘Till David Blondel (Hösten 1651.)’, in Per Wieselgren (ed.), Samlade skrifter av Georg Stiernhielm. Svenska författare utgivna av svenska vitterhetssamfundet 8. Stockholm: Albert Bonniers, 3.1: 136–41. Stockhammer, Robert (2014). Grammatik: Wissen und Macht in der Geschichte einer sprachlichen Institution. Berlin: Suhrkamp. Strodtmann, Johann Christoph (1756). Idioticon Osnabrugense. Leipzig und Altona [Leipzig & Altona, Hamburg]: in der Kortenschen Buchhandlung. Swadesh, Morris (1950). ‘Salish Internal Relationships’, International Journal of American Linguistics 16 (4): 157–67. Swadesh, Morris (1952). ‘Lexico-Statistic Dating of Prehistoric Ethnic Contacts: With Special Reference to North American Indians and Eskimos’, Proceedings of the American Philosophical Society 96 (4): 452–63. Swadesh, Morris (1955). ‘Towards Greater Accuracy in Lexicostatistic Dating’, International Journal of American Linguistics 21 (2): 121–37. Swartling, Birger (1909). Georg Stiernhielm: Hans lif och verksamhet. Uppsala universitets årsskrift: Filosofi, sprakvetenskap och historiska vetenskaper 4. Uppsala: Akademiska bokhandeln. Swiggers, Pierre (1981). ‘The Word Šibbōleṯ in Jud. .6.’, Journal of Semitic Studies 26 (2): 205–7. Swiggers, Pierre (1984). ‘Adrianus Schrieckius: De la langue des Scythes à l’Europe linguistique’, Histoire Épistémologie Langage 6 (2): 17–35. Swiggers, Pierre (1997). Histoire de la pensée linguistique. Paris: Presses Universitaires de France. Swiggers, Pierre (2000). ‘La canonisation d’un franc-tireur: Le cas de Hugo Schuchardt’, in Wolfgang Dahmen et al. (eds), Kanonbildung in der Romanistik und in den Nachbardisziplinen: Romanistisches Kolloquium XIV. Tübinger Beiträge zur Linguistik 449. Tübingen: Gunter Narr, 269–304. Swiggers, Pierre (2009). ‘Le français et l’italien en lice: L’examen comparatif de leurs qualités chez Henri Estienne’, Synergies Italie 5: 69–76. Swiggers, Pierre (2017). ‘Indo-European Linguistics in the 19th and 20th Centuries: Beginnings, Establishment, Remodeling, Refinement, and Extension(s)’, in Jared Klein et al. (eds), Handbook of Comparative and Historical Indo-European Linguistics. Handbooks of Linguistics and Communication Science 41.1. Berlin & Boston, MA: De Gruyter Mouton, 171–210. Tavoni, Mirko (1982). ‘The 15th-Century Controversy on the Language Spoken by the Ancient Romans: An Inquiry into Italian Humanist Concepts of “Latin”, “Grammar”, and “Vernacular”’, Historiographia Linguistica 9 (3): 237–64.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



329

Tavoni, Mirko (1984). Latino, grammatica, volgare: Storia di una questione umanistica. Padova: Antenore. Tavoni, Mirko (1986). ‘On the Renaissance Idea that Latin Derives from Greek’, Annali della Scuola normale superiore di Pisa: Classe di lettere e filosofia 16 (1): 205–38. Tavoni, Mirko (1987). ‘Contributo all’interpretazione di De vulgari eloquentia I 1–9’, Rivista di letteratura italiana 5 (3): 385–453. Tavoni, Mirko (1990). ‘“Ydioma Tripharium” (Dante, De Vulgari Eloquentia, I 8–9)’, in Hans-Josef Niederehe and E. F. K. Koerner (eds), History and Historiography of Linguistics: Papers from the Fourth International Conference on the History of the Language Sciences (ICHoLS IV), Trier, 24 – 28 August 1987. Studies in the History of the Language Sciences 51. Amsterdam & Philadelphia, PA: John Benjamins, 1: 233–47. Tavoni, Mirko (1998). ‘1 Renaissance Linguistics: 1.2 Western Europe’, in Giulio Lepschy (ed.), History of Linguistics. Longman Linguistics Library. London & New York: Longman, 3: 2–108. Taylor, Daniel J. (1996). De lingua Latina X: A New Critical Text and English Translation with Prolegomena and Commentary. Studies in the History of the Language Sciences 85. Amsterdam & Philadelphia, PA: John Benjamins. Thomas, Jacques (1953). ‘Dialecte et patois: Esquisse d’une étude sémantique’, Romanica Gandensia 1: 93–117. Thryllitsch, Georg Friedrich [auctor et respondens] (1709). Suspiciones quasdam historicotechnicas de dialectis Græcis ex consideratione originum migrationumq[ue] Graecarum nationum collectas . . . eruditorum examini modeste subiicio . . . Thryllitius . . . Vitembergae [Wittenberg]: litteris Schroedterianis. Tillinger, Gábor (2013). ‘Langues, dialectes et patois—Problèmes de terminologie dialectologique: Réflexions sur la situation géolinguistique en France et la terminologie française’, Argumentum 9: 1–18. Tory, Geoffroy (1529). Champ fleury . . . A Paris: par Maistre Geofroy Tory de Bourges . . . et par Giles Gourmont. Trabant, Jürgen (2006). Europäisches Sprachdenken: Von Platon bis Wittgenstein. Munich: C. H. Beck. Trautmann, Thomas R. (1998). ‘The Lives of Sir William Jones’, in Alexander Murray (ed.), Sir William Jones, 1746–1794: A Commemoration. Oxford: Oxford University Press, 91–121. Tribulato, Olga (2014). ‘Dialectology (diálektos), Ancient Theories of ’, in Georgios K. Giannakis (ed.), Encyclopedia of Ancient Greek Language and Linguistics. Leiden & Boston, MA: Brill, 1: 457–61. Trissino, Gian Giorgio (1529). Dialωgω . . . intitulatω il castellanω, nel quale si tratta de la lingua italiana. [Vincenza]: [Tolomeo Ianiculo]. Trovato, Paolo (1984). ‘“Dialetto” e sinonimi (“idioma”, “proprietà”, “lingua”) nella terminologia linguistica quattro- e cinquecentesca (con un’appendice sulla tradizione a stampa dei trattatelli dialettologici bizantini)’, Rivista di letteratura italiana 2: 205–36. Trudeau, Danielle (1992). Les inventeurs du bon usage (1529–1647). Arguments. Paris: Editions de Minuit. Trudgill, Peter (1983). On Dialect: Social and Geographical Perspectives. Oxford: Basil Blackwell. Trudgill, Peter (2003). ‘Modern Greek Dialects: A Preliminary Classification’, Journal of Greek Linguistics 4: 45–64. Ursin, Georg Heinrich (1691). Grammatica Græca . . . Norimbergæ [Nuremberg]: sumptibus Wolfgangi Mauritii Endteri.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

330



Vandewalle, Carolus Borromaeus (1929). ‘Roger Bacon dans l’histoire de la philologie’, La France franciscaine 11 and 12: 315–410 and 45–90, 161–210. Van Hal, Toon (2010a). ‘Moedertalen en taalmoeders’: Het vroegmoderne taalvergelijkende onderzoek in de Lage Landen. Verhandelingen van de Koninklijke Vlaamse Academie van België voor Wetenschappen en Kunsten. Nieuwe Reeks 20. Brussels: Koninklijke Vlaamse Academie voor Wetenschappen en Kunsten. Van Hal, Toon (2010b). ‘“Quam enim periculosa sit ea via . . . ”: Josephus Justus Scaliger’s Views on Linguistic Kinship’, Beiträge zur Geschichte der Sprachwissenschaft 20 (1): 111–40. Van Hal, Toon (2013/14). ‘From Alauda to Zythus: The Emergence and Uses of OldGaulish Word Lists in Early Modern Publications’, Keltische Forschungen 6: 219–77. Van Hal, Toon (2015). ‘Friedrich Gedike on Why and How to Compare the World’s Languages: A Stepping Stone between Gottfried Wilhelm Leibniz and Wilhelm von Humboldt?’, Beiträge zur Geschichte der Sprachwissenschaft 25 (1): 53–76. Van Hal, Toon (2018). ‘The Alleged Persian-Germanic Connection: A Remarkable Chapter in the Study of Persian from the Sixteenth through the Nineteenth Centuries’, in Alireza Korangy and Corey Miller (eds), Trends in Iranian and Persian Linguistics. Berlin & Boston, MA: De Gruyter Mouton. Van Hal, Toon (forthcoming). ‘Ludolf ’s Language Laws: Pitfalls in Documenting and Comparing the World’s Languages’. Van Hal, Toon and Raf Van Rooy (2017). ‘ “Differing Only in Dialect”, or How Collocations Can Co-Shape Concepts’, Language & Communication 56: 95–109. Van Rooy, Raf (2013). ‘ “Πόθεν οὖν ἡ τοσαύτη διαϕωνία;”: Greek Patristic Authors Discussing Linguistic Origin, Diversity, Change and Kinship’, Beiträge zur Geschichte der Sprachwissenschaft 23 (1): 21–54. Van Rooy, Raf (2016a). ‘Struggling to Order Diversity: The Variegated Classifications of Greek Dialects before the Rise of Modern Linguistics’, Studies in Greek Linguistics 36: 465–73. Van Rooy, Raf (2016b). ‘The Ancient Greek and Byzantine Notion of Dialektos: A Criterial Approach’, in Carlos Assunção et al. (eds), History of Linguistics 2014: Selected Papers from the 13th International Conference on the History of the Language Sciences (ICHoLS XIII), Vila Real, Portugal, 25–29 August 2014. Studies in the History of the Language Sciences 126. Amsterdam & Philadelphia, PA: John Benjamins, 55–67. Van Rooy, Raf (2016c). ‘The Diversity of Ancient Greek through the Eyes of a Forgotten Grammarian: Petrus Antesignanus (ca. 1524/1525–1561) on the Notion of «Dialect»’, Histoire Épistémologie Langage 38 (1): 123–40. Van Rooy, Raf (2016d). ‘ “What Is a ‘Dialect’?”: Some New Perspectives on the History of the Term διάλεκτος and Its Interpretations in Ancient Greece and Byzantium’, Glotta 92: 244–79. Van Rooy, Raf (2018a). ‘Latin as a Variable Language: Livy’s Patauinitas through Early Modern Eyes’, in Astrid Steiner-Weber and Franz Römer (eds), Acta Conventus NeoLatini Vindobonensis: Proceedings of the Sixteenth International Congress of Neo-Latin Studies (Vienna 2015). Leiden & Boston, MA: Brill, 767–77. Van Rooy, Raf (2018b). ‘Regional Language Variation in European Thought before 1500: A Historical Sketch Reflecting Current Research’, Beiträge zur Geschichte der Sprachwissenschaft 28 (2): 167–216. Van Rooy, Raf (2018c). ‘«Plutarque dialectologue»: La pseudo-autorité de Plutarque dans le discours sur les dialectes grecs à la Renaissance (ca. 1400–1670)’, Revue belge de philologie et d’histoire 96: 1115–34.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi



331

Van Rooy, Raf (2019). ‘Διάλεκτος, Dialectus, Dialect: A Word’s Curious Journey from Ancient Greek to (Neo-)Latin and Beyond’, Latomus: Revue d’études latines 78: 733–70. Van Rooy, Raf (2020a). Greece’s Labyrinth of Language: A Study in the Early Modern Discovery of Dialect Diversity. History and Philosophy of the Language Sciences 2. Berlin: Language Science Press. Van Rooy, Raf, ed. (2020b). Essays in the History of Dialect Studies: From Ancient Greece to Modern Dialectology. Münster: Nodus. Van Rooy, Raf (2020c). ‘Tradition, Synthesis, and Innovation: An Early EighteenthCentury Dissertation on Dialects Presented in Wittenberg’, in Marti Hanspeter and Robert Seidel (eds), Early Modern Disputations and Dissertations in an Interdisciplinary and European Context. Leiden & Boston, MA: Brill, 536–54. Van Rooy, Raf (forthcoming). ‘A Latin Defence of Early Modern Greek Culture: Alexander Helladius’s Status praesens (1714) and Its Linguistic Arguments’, in Ioannis Deligiannis et al. (eds), Post-Byzantine Latinitas: Latin in Post-Byzantine Scholarship (15th–19th cent.). Turnhout: Brepols. Van Rooy, Raf and John Considine (2016). ‘Between Homonymy and Polysemy: The Origins and Career of the English Form Dialect in the Sixteenth Century’, Anglia: Journal of English Philology 134 (4): 639–67. Varchi, Benedetto (1570). L’Hercolano: Dialogo . . . nel qual si ragiona generalmente delle lingue, & in particolare della Toscana, e della Fiorentina. In Vinetia [Venice]: appresso Filippo Giunti, e Fratelli. Várvaro, Alberto (2008). ‘La tua loquela ti fa manifesto: Langue et identité dans la littérature médiévale’, in Moos (2008: 223–36). Vassallo, Stefano (2005). Himera città greca: Guida alla storia e ai monumenti. Beni culturali Palermo: Collana di studi e ricerche 8. Palermo: Regione Siciliana. Vergara, Francisco de (1537). De Græcæ linguæ grammatica libri quinque. Compluti [Alcalá de Henares]: apud Michaelem de Eguia. Versteegh, Kees (1986). ‘Latinitas, Hellenismos, ’Arabiyya’, Historiographia Linguistica 13 (2–3): 425–48. Villani, Francesco Paolo (2003). ‘Scythae: Un problema linguistico, etnografico e culturale dell’età moderna’, Linguistica 1: 443–91. Vives, Juan Luis (1531). De disciplinis libri XX. Antuerpiæ [Antwerp]: excudebat . . . Michael Hillenius. Vives, Juan Luis (1533). De ratione dicendi libri tres. De consultatione. Louanij [Leuven]: ex officina Rutgeri Rescij . . . Sumptibus eiusdem ac Bartholomęi Grauij. Voegelin, Carl F. and Zellig S. Harris (1951). ‘Methods for Determining Intelligibility among Dialects of Natural Languages’, Proceedings of the American Philosophical Society 95 (3): 322–9. Vogt, Joann (1974). The Linguistic Work of Friedrich Karl Fulda. The Hague & Paris: Mouton. Vrolijk, Arnoud, Jan Schmidt, and Karin Scheper (2012). Turcksche boucken: . . . The Oriental Collection of Levinus Warner, Dutch Diplomat in Seventeenth-Century Istanbul. Eindhoven: Lecturis. Vuidius, Robertus (1569). ‘De dialectis Græcis libellus’, in Pantaléon Bartelon, De ratione quantitatis syllabariæ liber . . . Parisiis [Paris]: apud Hieronymum de Marnef, & Gulielmum Cauellat, 136–48. Wachter, Johann Georg (1727). Glossarium Germanicum continens origines et antiquitates linguæ Germanicæ hodiernæ: Specimen ex ampliore farragine decerptum. Lipsiae [Leipzig]: sumptibus Iacobi Schusteri.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 27/8/2020, SPi

332



Wackernagel, Jacob (1876). De pathologiæ veterum initiis: Dissertatio inauguralis . . . Basileæ [Basel]: typis expressit Officina Schweighauseriana. Waldberg, Max von (1889). ‘Richey, Michael’, in Allgemeine Deutsche Biographie. Munich & Leipzig: Duncker & Humblot, 28: 436–9. Walper, Otto (1589). De dialectis Græcæ linguæ præcipuis: Attica; Ionica; Dorica; Æolica; & coronidis vice nonnulla de proprietate poetica . . . Francofurti ad Moenum [Frankfurt am Main]: ex officina typographica Ioannis Spiessii. Walton, Brian (1657). ‘In Biblia polyglotta, prolegomena’, in Brian Walton (ed.), Biblia sacra polyglotta . . . Londini [London]: imprimebat Thomas Roycroft, 1–102. Weinreich, Max (1945). ‘ ‫’דער יִיווָא און די ּפרָאבלעמען ֿפון אונדזער ציַיט‬, ‫י ִיווָא בלעטער‬a25 (1): 3–18. Weinreich, Uriel (1954). ‘Is a Structural Dialectology Possible?’, Word 10 (2–3): 388–400. Wellisch, Hans H. (1984). Conrad Gessner: A Bio-Bibliography. Zug: IDC. Wensinck, A. J. (1921). ‘Schultens (Albert)’, in P. C. Molhuysen et al. (eds), Nieuw Nederlandsch Biografisch Woordenboek. Leiden: A. W. Sijthoff, 5: 707–11. Whitney, William Dwight (1867). ‘Languages and Dialects’, The North American Review 104 (214): 30–64. Wichmann, Søren (2016). ‘How to Distinguish Languages and Dialects’, English draft version of Ran and Wichmann (2018). Wilkins, John (1668). Essay towards a real character, and a philosophical language. London: printed for Sa: Gellibrand, and for John Martyn. Wilkinson, Robert J. (2016). ‘Constructing Syriac in Latin—Establishing the Identity of Syriac in the West over a Century and a Half (c.1550–c.1700): An Account of Grammatical and Extra-Linguistic Determinants’, BABELAO 5: 169–283. Williams, John (1685). A discourse concerning the celebration of divine service in an unknown tongue. London: printed for Richard Chiswell. Wolf, Hieronymus (1578). ‘De orthographia Germanica, ac potius Sueuica nostrate’, in Johannes Rivius, Institutionum grammaticarum . . . libri octo . . . Augustae Vindelicorum [Augsburg]: Michael Manger excudebat, 594–615. Wolfram, Walt (1998). ‘Language Ideology and Dialect: Understanding the Oakland Ebonics Controversy’, Journal of English Linguistics 26 (2): 108–21. Wood, Robert (1775). An essay on the original genius and writings of Homer, with a comparative view of the ancient and present state of the Troade. London: printed by H. Hughs; for T. Payne . . . and P. Elmsly. Wurm, S. A. and D. C. Laycock (1961). ‘The Question of Language and Dialect in New Guinea’, Oceania 32 (2): 128–43. Wyss, Caspar (1650). Dialectologia sacra . . .Tiguri [Zürich]: typis Johannis Jacobi Bodmeri. Zimmermann, Klaus and Birte Kellermeier-Rehbein, eds (2015). Colonialism and Missionary Linguistics. Berlin, Munich & Boston, MA: de Gruyter. Zimont, Elizaveta (2020). ‘Variation diatopique dans la lexicographie français$néerlandais aux XVIe et XVIIe siècles: Réception et usage des marques diatopiques’, in Van Rooy (2020b: 48–67). Zwartjes, Otto (2012). ‘The Historiography of Missionary Linguistics: Present State and Further Research Opportunities’, Historiographia Linguistica 39 (2–3): 185–242. Zwinger, Jakob (1605). ‘Græcarum dialectorum hypotyposis . . . ’, in Johannes Scapula, Lexicon Græco-Latinum nouum . . . Basileæ [Basel]: per Sebastianum Henricpetri, .1–a.6.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

Index The index records (1) language families, languages, and their varieties, (2) historical figures and scholars cited in the main text (i.e. not in the footnotes), and in selective fashion (3) terms (in italics), concepts, subjects, and other entities. I have classified language families, languages, and their varieties in rather broad categories here in order to avoid having to make countless language/dialect decisions myself and introducing too many sublevels in the index (two maximum). Only those subdivisions have been introduced which are relevant to the argumentation of the book, which is why I have, for instance, refrained from dividing Slavic into East and West. abolition of dialects 131–2, 288 Abstandsprache (‘language by distance’) 271–2 Académie française 106 accent 161, 163, 178–9, 193, 195, 197, 279–80 accent 268 vs. dialect 268 accidental (superficial) variation 5–7, 9, 11, 21, 33, 59, 74, 76, 80, 104, 109–24, 148, 157–8, 161, 163–6, 174–6, 179, 195, 197, 203, 205, 226–7, 253, 265–6, 282, 298 Adams, James N. 276–7 ad fontes 28 Adelung, Johann Christoph 8, 190–2, 196 Aeolism 34, 178 Agard, Frederick B. 10, 259–60, 262 Ahačič, Kozma 67 Ahrens, Heinrich Ludolf 222 Alans 144 Albertus, Laurentius 101–2, 118 d’Alembert, Jean le Rond 204 Alexander the Great 20 Alinei, Mario 2, 47–8, 298 Althamer, Andreas 74 alphabet (letters) 9, 142–3, 174, 178, 186, 188, 195, 197–200, 203 American peoples, Scandinavian descent of 145–6 Amerindian languages 10, 256–7 classification of 281 Amerot, Adrien 83–4 ancestral language 164, 220, 251 ancient Greece 2, 4–5, 296, 299 Ancient Greek 4–5, 7, 17, 22, 28–30, 32–8, 43, 49, 51, 53–5, 59–60, 62, 65–6, 68–71, 74–5, 77, 79, 81–4, 88–90, 95, 100–1, 107, 111, 114–16, 120, 125, 127–8, 130, 132, 143, 145, 150, 153–5, 165, 171, 175, 177, 179, 183, 196, 207, 210–11, 213, 219, 222–3, 246

Aeolic 15, 21, 23, 34, 58–60, 64, 66, 69, 77, 82, 89, 102, 115–16, 141, 143, 175–8, 190, 222–3 Aetolian 16 in Asia Minor 114 Attic 15, 17, 21–3, 59–61, 64, 66, 69–70, 77–8, 82, 89, 102, 115–16, 123, 128, 141, 143, 175–6, 178, 190, 199, 231 and High German 206 as the norm of Ancient Greek 222 Biblical 101 New Testament 101 dialects 4–6, 15–16, 18–21, 23–4, 30–2, 35–6, 38, 51–2, 59–60, 63, 65–7, 70–1, 73–4, 77–8, 81–4, 88–90, 98–9, 102, 112, 115–16, 120, 126, 141–2, 148, 152–3, 155–6, 165, 174, 176–7, 179, 189–90, 193–5, 199, 205, 207–8, 211–12, 214, 216, 221–4, 296 in Byzantine scholarship 197 classification of 22 as corrupt 132, 155 dramatic usage of 16 extinction of 20 handbooks on 82–4, 102, 107, 125, 179, 194–5, 203, 206, 208, 211 literary status of 15, 18, 20–1, 23, 130–1, 157, 199, 205, 212, 296 modern study of 222 and provinces 98–9 rediscovery of 81–4 Doric 15–16, 18–19, 21–3, 59–61, 64, 66, 69, 77–8, 82, 89, 102, 115–16, 141, 143, 175–6, 190 broadness of 16–17 Lacedaemonian 16, 66 Rhodian 64 and Upper German 206

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

334



Ancient Greek (cont.) Hellēnismós as ideal 22 Hellenistic 103, 107 history of 115 Homeric 213–14 Ionic 15–17, 19, 22–3, 59–61, 64, 66, 69–70, 77–8, 82, 89, 102, 115–16, 141, 143, 175–6, 179, 190 as the ancestral Greek language 214 Chalcidian 16 and Low German 206 Koine 6, 20–2, 33, 59–61, 64, 68–71, 89, 102, 126–9, 141–3, 150, 154, 206 analogical 148 as normative variety 22, 126 vs. the other dialects 126, 148 poetic dialect 107 pronunciation 198–9 Byzantine 198 Erasmian 199 Renaissance interest in 51–2, 91 as a Scythian dialect 165 on Sicily 16 ancient Greek literature 15, 52, 55, 296 ancient Greek scholarship 199, 223, 296–7, 299 ancient Greek tribes 19–20, 60, 76, 101–2, 116 Hellenistic 103, 107 Anderson, Benedict 85 Antesignanus, Petrus 70–1, 91, 95–7, 101–2, 154–5 anthropology 248, 256–8, 294 Antikvitetskollegium (College of Antiquity) 159 antiquarianism 183, 199, 221 Apollonius Dyscolus 21 appropriation 5, 90–1 approved dialects 70–1 Aquinas, Thomas 4, 38–40, 43 Aquitanians 110, 145 archaism in dialects 7, 228, 294 Aristophanes 16, 20, 54 Aristotle 6, 18, 54, 57, 109–10, 119, 153 Aristotelian ontology (substance vs. accidents) 6, 34–5, 43, 109, 123, 148, 282 Aristotelianism and early modern universities 109 Armenian 171 Armitage, David 3 article (part of speech) 34, 74 articular pronoun 34 Ascoli, Graziadio Isaia 242, 253 Asdal, Kristin 3 Ashley, Kathleen M. 90 Asia Minor 16, 212–13 Auer, Peter 294

augment 223 Aurelius, Marcus 54 Ausbausprache (‘language by elaboration’) 271–2 Babel, Tower of 40–2, 111, 194, 196, 285, 289 confusion of tongues 194, 199 Bachmann, Andreas 157 Bacon, Roger 4–5, 28–44, 47–8, 61, 77, 81, 90, 95, 109, 296 Baggioni, Daniel 297 Bally, Charles 245 Baltic languages 59, 74, 186 Curonian 74 Latvian 60 Lithuanian 59–60, 74, 144 Livonian 74 Old Prussian 59, 74, 144 Samogitian 60 Yotvingian 60 Bantu languages 253 barbarismós 20 Barbiers, Sjef 265 basic vocabulary 8, 186–7, 192, 257, 260, 271 Basque 116, 145, 234–5, 238, 252 dialects of 238 and Iberian 235 Sara 238 Bayly, Anselm 98 ‘to be’ 187 Beauzée, Nicolas 212 Bede (the Venerable) 38–9 Belgians 110, 145 Bellay, Joachim du 50, 84 Bembo, Pietro 68 Benzelius, Erik, the Younger 210 Besnier, Pierre 132, 139–40 Bible 28, 30, 39–40, 86, 88, 91, 122, 200, 207, 220, 239 New Testament 101, 103, 153, 171, 208, 212 Acts 38, 153, 212 Gothic translation of 162 Mark 171 Matthew 39 Old Testament 19, 172 Genesis 40 Judges 39–40, 165 Proverbs 176 Psalms 38–9 polyglot 122 translation into German 124, 207 into Swedish 207 Vulgate 30, 38–9

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

 Bibliander, Theodore 58, 77–9, 98, 140, 194 bibliography 53 bidialectal speakers 250–1 bilingual speakers 250–1 biology 139, 226, 229 Blair, Ann M. 49, 87 Blondel, David 164 Bloomfield, Leonard 248 Blount, Thomas 112, 156 Boas, Franz 248 Boberg, Charles 279 Boccaccio, Giovanni 51, 202 Bodin, Jean 75 Boeckh, August 221 Bolzanio, Urbano 82 Bopp, Franz 219–21, 229 Borghini, Vincenzo 68 borrowing 285 and conceptual change 80 botany 52–3, 139 Bourgain, Pascale 37 Bovelles, Charles de 96 Bracciolini, Poggio 26 Bréal, Michel 241 Bregius, Johannes 208, 210 Brevini, Franco 201 bridging context 67 Bringuier, Octavien 241 Britons 111 Bruni, Leonardo 26 de Bry, Johann Theodor 138 Buchanan, George 111 bucolic poetry 15 Bureau de la statistique 224 Burgundians 144 Burke, Peter 49, 134, 158 Byzantine émigrés 81–2 Byzantine scholarship 33, 199, 223, 296–7, 299 Caesar, Julius 144–5 Calepino, Ambrogio 72 Calmeta 68 Calvet, Louis-Jean 294–5 Camden, William 102 Camerarius, Joachim 78 Campbell, Lyle 264 Canini, Angelo 73–4 case 142, 161 Catholicism 5, 86, 212 Caucasian languages 235, 254 Caxton, William 51 Celtic languages 20–1, 24–5, 51, 75, 144, 186, 190, 234 Breton 24

335

Galatian 24–5 Gaulish 24, 51, 146 in Gaul 20–1, 59, 110, 113–14, 144–5 Irish 73 dialects of 73 Trierisch 24–5 Welsh 25, 145, 284 Celts 110, 145 centralized government 69, 96, 99, 106, 128, 133, 157 Ceporin, Jakob 74 Challenge controversy 212 Chambers, J. K. 267–8 Chinese languages, see Sinitic languages Chomsky, Noam 11, 245, 263, 265, 268–70 Choniates, Michael 20 Christina, Queen of Sweden 8, 160, 166 chronology 183 Chrysoloras, Manuel 81–2 Chytraeus, Nathan 129 Cicero 172 citation practices 300 city-states 99 civilization 236 Clackson, James 22 classical languages 189 Clement  (Pope) 31 Clement of Alexandria 19, 55–7, 59, 64–5, 96–7, 108 Clenardus, Nicolaus 70, 95 Codro 66–7 cognition 265 Collegium theologicum 172 Colli, Vincenzo, see Calmeta collocation 5 Colombat, Bernard 57–8 Columbus, Christopher 49 Colvin, Stephen 16 common language 5, 50, 59–61, 67–72, 80, 96, 122, 125–30, 132–3, 150, 164, 204–5 as a compilation from different dialects 129 ethnic neutrality of 105 geographic centrality of 60–1 intelligibility of 128 communicative reach 6 composition (lexical) 285 conceptual pair, see language/dialect distinction conjugation 142, 161 Considine, John 85, 144, 162 consonant 190, 201, 285 cognate 190 Constantin, Emile 246 Constantinople 82 Fall of 52

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

336



Cooper, Thomas 72–3 Coquebert de Montbret, Charles-Étienne 224–5 Coquebert de Montbret, Eugène 224–5 core landscape (Kernlandschaft) 242 Coseriu, Eugenio 275, 281–2 Cottière, Matthieu 153–4 Council of Europe 290–2 countability of language 5, 88–90, 243, 296, 298 assumption of 284 specifically of dialects 89, 296, 298 creole 234 Crusius, Johannes 155 curiosity cabinets 87–8 customs 179 Cysouw, Michael 277, 282 Dacus, Johannes 34 Dahan, Gilbert 42 Dante Alighieri 25–7, 42, 99, 202 Darwin, Charles 226 De Laet, Johannes 145–6 declension 161 deep structure 259, 263–5 degenerare in dialectum 165 degenerate offshoot 8, 176–8, 182, 211–12 degeneration of language 167 degradation of dialects 289 Delbrück, Berthold 229 derivation 190 Descartes, René 204 dia- 275 diacritic marks 193 dialect 4, 11, 19–20, 58, 73, 75, 90, 98, 112, 153–4, 188, 220, 227, 249–52, 254–5, 266–8, 270, 275–85, 297, 299–300 abandonment of the term 251, 275, 299 country 133 compositions with 275–7, 282 daughter 136, 139, 247 definitions of 11, 251 derivations of 11, 275–8, 282 as a difficult word 285 in lay usage 283 naturalization as an English word 73 as a non-technical term 249–51 poorly defined 251 as a term for Greek only 154 dialect area 267 dialect area 276 dialect attitudes 17, 61 dialect boundaries 10, 239, 254, 300 arbitrariness of 10, 300 problematic nature of 25, 245, 250, 267, 276–7, 300

dialect chain 25 dialect chain 276 dialect classification 237 impossibility of 237 dialect cluster 260, 262, 283, 301 dialect continuum 231, 250, 257 dialect continuum 276 dialect convergence vs. divergence 252–3, 280 dialect convergence/divergence 276 dialect and corruption 98 dialect and etymology 157 dialect as a fictitious concept 239, 246 dialect geography 9, 228, 230, 243, 298, 300 dialect as an identity mark 70 dialect knowledge, utility of 207–8 dialect levelling 276 dialect loss 276 dialect and migration 138 dialect mixture 16, 51, 241 dialect as a popular concept 238 dialect preservation 210 dialect as a relative concept 223, 235, 238, 246, 279 dialect renaissance 281 dialect shift 276 dialect and social class 130 dialect, sociological view of 248, 274 dialect status, acceptance of 271 dialect term/word 276–7 dialect and writing 207 dialectal 275 dialecte 129, 131, 151–2, 212, 246–8 naturalization as a French word 151 dialectologia 9, 208, 210 dialectological tradition 9, 208, 211, 216 dialectological turn 10 Dialectologie 210 dialectology (institutionalized) 210–11, 225, 228, 230, 245, 263, 266–7, 273, 279, 281, 293–4, 298 continental European 279 dialectology 21, 208, 275 externalist form of 249 Gallo-Romance 241, 243, 247, 251 generative 245, 264 German 242 Italian 242 Montpellier school of 241 Parisian school of 242 structuralist 245, 247–55, 265, 298 Swiss 249 traditional 263, 267, 281 dialectometry 276

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

 dialects, existence of lack of 251 ‘only dialects exist’ 247 dialectus 4–5, 7–8, 24, 35, 53, 55–81, 84, 90, 95–8, 101–3, 107–8, 120–3, 130–1, 137–42, 153, 156, 167, 180–1, 194, 204 communis 60, 89, 150 egkhṓrios 116 ethnikḗ 116 generalis 115 naturalization as a Latin word 53, 62–4, 149 polysemy of 57–8, 80, 103, 121 primaria 116 princeps 115 principalis 116 propinquior 116 remotior 116 secundaria 116 specialis 115 topikḗ 116 urbica 116 vulgaris 60 dialégomai 18 Dialekt (Dialect) 216, 220, 222, 225–6, 235–6, 286 Hauptdialekt 116, 236 Nebendialekt 116 Unterdialekt 236 diálektos 2, 4–5, 17–21, 23–4, 27, 43, 55–7, 63, 65, 72, 76–7, 80–1, 90–1, 96, 98, 101–3, 108, 147, 149–50, 154–5, 208, 297 ethnic conception of 17, 19, 27, 297 Latinization of 90, 149 literary conception of 20, 27 and lower class speech 20 appropriation of 149–50 parátonos 20 as ‘particularity of tongue’ 18 polysemy of 153–4 spatial conception of 17, 19–20, 27, 297 dialetto 202, 280 dialettuzzolo 116 diaphasic 275 diastratic 275 diasystem 10, 247–51, 254, 259, 272 diatopic 275 Dictionnaire de Trévoux 197 dictionary 185 dialect 199 Diderot, Denis 204 ‘to differ only in dialect’ 5, 74–6, 144–6, 164, 166, 220, 226, 290, 292 expansions of the phrase 109–13 digamma 34

337

diglossia 278 Diogenes Laertius 17 Diogenes of Babylon 17–20, 98 diversitie 73 Driessen, Anton 173, 175 Drusius, Johannes 103 dynamism of language 239 École des Hautes Études 247 edit distance, see Levenshtein distance (normalized) Egyptian 20, 164, 220, 253 Coptic 253 Eichhorn, Johann Gottfried 190 Eisenstein, Elizabeth L. 52 Eliot, Thomas 72–3 Eloy, Jean-Michel 278, 282 encyclopaedia 204 Encyclopédie 204, 212 ending 142, 161 English 2–4, 25, 33, 36, 38, 51, 69, 72–3, 75, 106, 133, 156, 175, 184, 214, 216, 227, 265, 285, 292–3 American 266, 285–6, 289 African American Vernacular 267, 290 on Martha’s Vineyard 266 in New York 266 Pacific Northwest 266 Pittsburgh 266 British 285–6 Eastern 89 London 133 Northern 89 Southern 89 Western 89 common language 89, 106, 127–8 of the court and intelligentsia 128 dialects of 89, 127–8, 156, 206, 290 diphthong change in 266 lexicon 271 Middle English 31 Devonish 72 East Midland 31 Kentish 72 Northern 32, 72 South Eastern 51 South Western 31 Southern 72 West Midland 31 Poetic 89 received pronunciation 280 as a Romance language 177 standard language 280, 290 variability of 51 Welsh 284

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

338



Enlightenment 8, 192, 201, 203–4, 206, 216, 286, 297 Ephraimites 39–40, 165, 239 Erasmus, Desiderius 65–6 Eskhult, Josef 173 Estienne, Henri 98, 130, 195–6 ethnography 183–4, 191–2 of Asia 220 Ethnologue 10, 258, 261, 277 éthnos 19, 101 etymology 177, 189, 296 European Charter for Regional or Minority Languages 278, 290–2 European Union 292 exegesis 4 explorations, age of 49–50 extinct languages 125 feudalism 289 Filelfo, Francesco 82 Flavio, Biondo 26 Florence 81 Franciscans 30 Franks 144 French National Convention 288 French Revolution 225, 288–9 Fulda, Friedrich Carl 9, 214–16, 224, 230 Furetière, Antoine 129, 131, 151–2, 155 Fust, Johann 91 Gabelentz, Georg von der 229–30, 286–8, 299 Galiani, Ferdinando 9, 201–3 Gallo-Romance 33, 35–6, 39–40, 96, 278 dialects 241 Franco-Provençal (Arpitan) 242, 278 Lyonnais 89 Savoyard 196 French (langues d’oïl) 4, 26, 33, 35–8, 69–71, 79, 84, 88–9, 96, 114–15, 117–18, 120, 128–30, 139, 151–2, 154, 165, 175, 177, 195–6, 214, 225, 241, 252–3, 269, 288–9 Burgundian 32, 35, 37 common language 70 and Corsican 252, 271 dialects of 70–1, 98, 117–18, 130, 161, 195, 211, 269 of the Île-de-France 51 lingua Gallicana 33 Norman 35, 98, 161 of Orléans 115 Parisian 37, 40, 70, 89, 128, 130, 161 patois 212 Picard 35–7, 40, 89, 98, 161, 278 at the royal court 70, 89

social variation in 130 standard language 88, 252 standardization of 96 Walloon 35–7, 117 Occitan (langues d’oc) 26, 241 Gascon 117 Limousin 89 Provençal 36, 89, 117, 161 Toulousian 161 Gatterer, Johann Christoph 8, 183–93, 196, 200, 203–4, 286, 297 Gauchat, Louis 242 Gaza, Theodore 82 Gedike, Friedrich 206–8 Geeraerts, Dirk 276 Geijer, Erik Gustaf 160–1 gender studies 276 generic vs. specific dialects 205 genius 180, 207 gens 101 geography 37, 183, 238–9, 275 ethno- 147 geologia 208 Gerald of Wales 25 Gerard of Huy 30 gergo 102 German 32–3, 47, 60–1, 68–9, 73–4, 79, 91, 97, 104, 115, 118, 122, 129, 150, 153, 165, 183–4, 190–1, 200, 205, 207, 210, 214–16, 220–1, 227, 233–4, 237, 267, 272 Austrian 186 Bavarian 61, 68–9, 186, 286 dialects 221 and Bible translation 207 Büchersprache 185 common language 105, 124 dialects of 78, 89, 101–2, 115, 124, 155, 161, 198, 206, 212, 214–15 as conservative 185 in parallel to Greek 206 and social class 155 Franconian 115, 153, 186 Hamburg 210 High 102, 118, 123, 132–3, 150, 206–7, 210, 287 and Attic Greek 206 at the imperial court 105 Low 102, 118, 129, 132, 153, 207 and Ionic Greek 206 of Meissen 115, 153 Osnabrück 210 Palatine 286 Plattdeutsch 286–7 of the Saxon chancellery 69, 124

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

 Schriftdeutsch 286 Silesian 153 in Speyer 105 standard language 69, 261 Swabian 68–9, 120, 153, 186, 286 Swiss 47, 61, 73, 157, 261 Thuringian 115, 153 Ubian 68–9 Upper, and Doric Greek 206 Germanic languages 25, 42, 49, 60, 75, 100, 113, 116, 137, 144–5, 149, 157, 186, 190, 215–16, 225 Danish 214, 221, 261, 267, 292 as a dialect chain 25 Dutch 69, 100, 111, 119, 121, 148, 157, 201, 212, 216, 280 Brabantian 61, 120 common written 148 dialects of 148 Flemish 120 Guelderish 157 interlanguage (tussentaal) 280 kinship with Persian 201 as the primeval language 119, 121 English, see English German, see German Gothic 162, 165 as the Germanic mother language 214 grammar 215 lexical roots 215 Norwegian 267, 292 Old Saxon 165 protolanguage 200 in Scandinavia 267 Swedish 121, 160, 162, 165, 207, 227, 261, 267 and Bible translation 207 Dalecarlian 164 dialects of 161, 183, 210 norm 207 orthography 207 as a Scythian dialect 162 Västergötland dialect of 121, 207 Gessner, Conrad 4–5, 47–62, 64–5, 72–5, 77–9, 88, 91, 96–8, 113, 137, 140, 144, 296 Giard, Luce 146 Giese, Albert 9, 221–4, 230 Gileadites 39–40, 239 Gill, Alexander 89–90, 133 Gilliéron, Jules 10, 241, 251, 275 Gipper, Sonja 67 Glarean, Heinrich 59 glôssa/glôtta 19 glossonym 277, 282 Glottolog 10, 258, 277–8

339

glottopolitics 271, 293 Goebl, Hans 241–2, 276 Goldhagen, Hermann 98–9 Good, Jeff 277, 282 good language 98 Gooskens, Charlotte 258–9 Goropius Becanus, Johannes 6, 119–22, 257 Göschel, Joachim 281 Goths 144 Göttingen Historical Institute 183 Gousset, Jacques 173, 175 Grafton, Anthony 91 grammar 11, 34, 100, 147, 151, 259–60, 263, 267–8, 271, 279, 301 Byzantine 30, 126 comparative 185 Greek 34, 76, 82 of individual dialects 264 Latin 34 manuals 267 modistic (speculative) 25, 34 panlectal (polylectal) 267 universal (generative) 263, 265 as a universal art 34–5 Greek, see Ancient Greek, vernacular Greek (early modern), Modern Greek history of 152 Greek Orthodox Church 123 Greek studies 69, 71, 91–2, 126, 207, 211 Greene, Robert Lane 292–3, 299 Grégoire, Henri (Abbé) 288–9 Gregorius Pardus, see Gregory of Corinth Gregory of Corinth 19, 72, 82 Grenzsaum 272 Gretser, Jakob 194 Grimm, Jacob 220 Groddeck, Benjamin 181, 206 Grotius, Hugo 145 Gubbio tablets 51 Guillén de Brocar, Arnao 101 Guldi, Jo 3 gunpowder 28 Gutenberg, Johann 91 Habsburg 100 Halle, Morris 264 Haloinus, Georgius 99–100, 113–14 Ham 79 Hamitic languages 253 Handbook of dialectology 278–80 Harris, Zellig 10, 256–7, 262–3 Haspelmath, Martin 277 Haugen, Einar 281, 284 Heinsius, Daniel 103

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

340



Helladius, Alexander 122–3 Helm, Abraham 125, 132–3, 204–6, 211 Hemsterhuis, Tiberius 180 Henry, Victor 240–1 Heraclitus 239 heraldry 183 Herodotus 15–17, 19, 78 Hervet, Gentien 55–7 Hesychius 20 Heylyn, Peter 156 Hickey, Raymond 279 Himera (Sicily) 16 Hinskens, Frans 279 Hippocrates 54, 70 Hirschi, Caspar 85 historiography 8, 15, 147, 151, 159, 183–92, 204, 214, 220, 294 and statistics 189 universal 183 history, sense of 146 Hockett, Charles F. 254–6, 258, 266 Hof, Sven 121, 207–8 Hoius, Andreas 97, 154–5 Homer 15, 21, 54, 153 language of 213–14 Horace 23 van der Horst, Joop 88 Hottinger, Johann Heinrich 7, 141–3, 213, 249 Hovdhaugen, Even 43 Howell, James 114, 116–17, 138–9 Howes, Thomas 214 Howlet, Richard 73 humanism 2, 5–6, 26, 87–8, 100, 296–7, 299 Hyde, Thomas 198 Iberian 235 and Basque 235 Ibero-Romance 97 Catalan 261 as a Spanish dialect 252 Spanish 32–3, 51, 69, 79, 115, 120, 128, 139, 175, 261 Castilian 51, 128 of Toledo 115 ideal speaker 245, 265, 280 idiolect 226, 248, 255–6, 275, 277 idiom 5, 35–6, 42–3, 76–8, 179, 194, 197 idiom 227 idioma 4, 31–8, 42–3, 76–7, 90, 131 idíōma 77 idiome 212 idiota 38 Ihre, Johan 183

imperialism, ideology of 294 Indo-European languages 219–21, 225, 228, 230–1, 237, 245, 252, 298 as dialects 248, 250 inductive method 28, 31 info-lust 87–8 information explosion 4–5, 84, 87–8, 91, 296 innateness of language 263 intellectual history 2–3 Internet 11, 293 isogloss 241, 249–51 Italic languages 223 Latin, see Latin Umbrian 51, 226 Italo-Romance 26, 99, 202, 280 Bolognese 99 variation in the city of Bologna 99 Corsican 251–2, 271 Italian 26, 33, 68–9, 79, 81, 91, 115, 120, 128, 139, 165, 175, 251–2, 269, 271 common 202 as corrupted Latin 174 dialects of 81, 161, 202, 212 Florentine 115 standard language 280 Tuscan 128, 161, 202, 251, 280 Lombard 161 Neapolitan 161, 201–2 Roman 238 Sicilian 280 Venetian 161, 238, 280 Jablonski, Paul Ernst 212–13 Jäger, Andreas 166 Japheth 79 jargon 102, 254, 278 jargon 132, 212, 275 Jaucourt, Louis de 97 Jerome 24–5, 30, 39, 49, 98 Jesus Christ 49 Jiménez de Rada, Rodrigo 41–2 John the Grammarian (also known as John Philoponus) 19–22, 33, 77, 89, 193–5, 199, 203 Jones, William 219–21 Joseph, John E. 85 Kals, Jan Willem 116, 180–2 ten Kate, Lambert 116, 148–9, 151 Keller, Raffael 53–4 kharak-root 16–17, 19 kharaktḗr 16 kharássō 17

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

 Kirchmaier, Georg Caspar 155, 166 Kirchmaier, Georg Wilhelm 206 Kloss, Heinz 271–3, 281–3 koine 71 koinḗ 60, 84, 89 as a language/dialect criterion 252 Kretzschmar, William A., Jr 273–4, 276, 281–2, 300–1 Kristiansen, Gitte 276 Kristiansen, Tore 279 Kümin, Beat 99 Kunstsprache 15 Kurath, Hans 242, 267 Labov, William 266–7, 279, 281, 290, 293 language 11, 19, 73, 127–8, 130, 220, 227, 252, 254–5, 266–70, 277–8, 282–4, 297, 300 common 127 in lay usage 283 mother 136, 138, 150, 152, 196, 247 as a non-technical term 268–9 poorly defined 251 radical 112, 156 language/dialect distinction absence of 17 alternatives for 270 ambiguity of 215, 229, 250, 279, 284, 287, 294 ambiguous attitudes towards 248, 282–3 arbitrariness of 9–10, 182, 189, 226, 228, 231, 248, 298, 300 Aristotelian conception of 6–9, 33–4, 109–24, 147–9, 156–7, 163–4, 166–7, 174–6, 182, 195, 197, 205, 215–16, 227, 230, 253, 264, 282 attributed to the Greeks 120 auxiliary status of 162, 182, 230, 296 and communicative reach 122–3 consonants vs. vowels 201, 203 convenience of 287 and cut-off point 253, 260–1, 271, 283, 300 debate about (or lack thereof ) 249, 280–1, 298–9 de-Hellenization of 7, 133–4, 151–7, 167, 176, 190, 194, 203, 211, 297, 299 diagnostic criteria for 6, 107, 123, 147, 165, 167 as a discursive strategy 7, 143–6 dismissal of 252, 268–70, 273–4, 278–80, 282–3, 295 emergence of 85, 90–1 uncoordinated nature of 90 ethnic conception of 6, 33, 59, 101–8, 147–9, 151, 156, 205, 224

341

as a flexible matrix 150–1, 228, 242, 254–5 future of 299–302 historical status of 12 implicitness of 248 interpretations of 92, 147–51, 156, 167, 202, 211–12, 214, 224, 228, 242, 248, 280, 282, 289, 297, 299 language-historical (genealogical) conception of 7–9, 58, 61, 78–9, 136–46, 148–9, 151–2, 164–5, 167, 176–7, 182, 189, 205, 213–14, 216, 219–22, 225–7, 229–30, 247–9, 251, 298 mother/offshoot distinction 213 lay conceptions of 11, 186, 252, 254, 256, 259, 266, 272, 276, 283–95 and linguistic features 163, 247, 256–7, 259, 268, 271, 276 non-scientific nature of 268 normative conception of (analogy vs. anomaly) 6–8, 125–7, 148, 176, 247, 251, 279, 286, 294 as an observational artifact 273 political activations of 11, 287–93 politicization of 11, 105–6, 158, 288–9 problematic nature of 291 and preconceived ideas 147 quantification of 258, 260–2 relativity of 247–8, 252 as a remnant of earlier thought 273, 282 as the responsibility of linguists 252, 301–2 scepticism about 212–16, 224–5, 228, 230, 240–2, 247, 249, 282, 284, 297–8, 300 scientific redefinition of 251–4, 260, 282–3, 298, 300 as a self-evident given 1, 12, 87–8, 92, 113, 123, 149, 166, 203, 220, 229, 242, 298–9 and social class 60, 134 (socio)political conception of 7, 11, 157–8, 270, 274, 282–3, 289–90, 293–5, 298, 300 rejection of 287, 292, 294 spatial conception of 5–9, 33, 59, 95–101, 147–9, 151, 156, 165, 176, 228, 230, 266–7, 276, 279, 282, 298 subjective conception of (status) 6, 125–35, 148–9, 151–2, 157, 165, 195, 202, 225, 249, 279, 288–90, 293, 298, 300 subsumption of dialect under language 18, 47–8, 57, 59, 62–4, 67, 69, 72–3, 76, 80, 92, 96, 101–3, 107–9, 120, 128, 147, 154–5, 205, 223, 297 supplementing of 11, 211–12, 248, 252, 254–5, 270–3, 280, 282 and syntax 202–3 tenability of 211

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

342



language/dialect distinction (cont.) as a useless abstraction 10, 216, 240, 270, 274, 293, 300 value of 215, 229, 246, 258, 270 language/dialect status determination of 10, 195, 273, 277–8, 287 forced 291 by laypeople (subjective perception) 256–7 debates about 290–3, 298 preconceived ideas about 298 language atlas 239, 298 language attitudes 43 research 293 language catalogue 4, 10, 47, 52–3, 75, 88, 113, 258, 277–8, 296, 298 language change 7, 111, 195, 197, 232, 235, 237, 252 causes of 164, 205–6, 285, 301 colonization 17, 201 isolation 286 language-external 221 migration 201 space 164, 201 time 164, 201, 205, 215 constant nature of 245 as a cyclical process 7, 161, 164, 200–1, 227, 249, 252 and ethno-geographic dispersion 20 gradual nature of 146, 161, 176, 181, 200–1, 205, 222, 226–7, 301 tree model (Stammbaumtheorie) 10, 231–3, 236 wave model (Wellentheorie) 10, 231–3, 236, 247 language classification 4–5, 7, 139, 221, 233, 281–2, 284, 296, 298 language cluster 260, 262, 283, 301 language community 287 language comparison and historiography 183–92 language contact 236–7, 239, 249, 253, 301 language and diplomacy 131 language and ethnic history 191–2 language family 229, 231, 241, 277 language history 7–8, 79, 122, 136–46, 149, 162, 167, 215–16 language as living organism 226, 242 language mixture 234, 237, 239 language origin 162, 166, 194–5 language and politics 269 language questions 127 questione della lingua 51, 68 language as a social fact 266, 271, 275 language and state 105–6, 212

language system 10–11, 245, 247, 253–4, 265 as closed 265, 273–4, 276, 279, 301 as complex and self-organizing 273, 282, 300–1 homogeneity of 10, 250, 280 horizontal vs. vertical varieties 253–4 idealized 269, 281 mixture of 269 subsystem 253–4 regional 253–4 language variation 111, 116, 119, 139, 149, 232, 235, 241, 247, 249, 252, 255, 263, 265–6, 270, 274–5, 278, 280–1, 285, 289–90, 296, 298, 300–1 awareness of 16–17, 27, 64 and boundaries 88, 231, 236–9, 254, 281, 300 arbitrariness of 238–9, 300 fuzziness of 246, 272, 300 categorization of 88, 289 need for 270 in one city 99, 266–7 classes of 22, 178–80, 182, 193–203, 206, 268, 297 as a continuum 88, 114, 239, 254 curiosity about 51, 79, 183 different levels of 115, 122–3, 205 discovery of 49 evaluation of 61 and external factors 11, 212, 247–8, 252–3, 258–9, 270–2, 275–6, 279, 287, 289, 298 culture 247–9, 254, 258–9, 267, 272 ethnolinguistic identity 258, 267 exclusion of 253 geography 248–9, 267, 276 history 267 perception 276 politics 247, 252, 254, 259, 267 school use 272 social class 248–9, 252, 259, 267, 276 style 276 time 249 gradational conception of 6, 99, 114–17, 123, 148, 187–8, 205, 229, 235–6, 246–7, 252 in the Holy Roman Empire 122 lack of interest in 24–5 in the New World 145–6 New Spain 145 origin of 200 universality of 5–6, 62, 99–101, 161, 227 as unrestrained 197 langue 10, 245–7 languoid 11, 277, 282 Larramendi, Manuel de 116 Lascaris, Constantine 82

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

 lāšṓn 103 Latin 2, 4–5, 15, 20, 22–6, 28, 30, 32–6, 38, 47, 49, 51, 53–4, 62–4, 66, 68, 72–3, 79–81, 84–5, 100–2, 111, 113–14, 129, 139, 145, 171–2, 174–8, 190, 198, 208, 210, 213, 219, 222–3, 233, 246, 276–7 as a daughter language of Greek 213 as a dialect of Greek 177–8, 223 dialects of 175 as a lingua franca 63, 80, 84, 145 literature 296 as a living language 33 Neo- 5, 63, 74–5, 85, 149 orthographies of 32–3 Paduan 100 periodization of 66 Praenestine 23 pronunciations of 32–3 regional variation in 23–4, 38, 206, 276–7 registers in 23, 78 social variation in 24, 100 as a uniform language 62, 66, 100 and vernacular 26 Vulgar 233, 237 vowel system 233, 235–6 lect 11, 275–6 communa- 275 docu- 11, 277, 282 ethno- 275–6 multi- 275 gender- 275–6 idio- 275 natio- 276 as a neutral term 275 regio- 275 socio- 275 techno- 275 lectometry 276, 281 Leibniz, Gottfried Wilhelm 185 Leskien, August 222, 232 letter mutations (permutatio litterarum) 21–2, 58–9, 92, 142–3, 161, 163, 165–6, 178–9, 187, 189, 191, 193–4, 196–7, 199–200, 205 and etymology 197 Leu, Urs B. 53–4 Leusden, Johannes 150 Levenshtein distance (normalized) 261 lexeme 276 lexicon 7, 9–10, 27, 111, 142, 161, 163, 165–6, 172, 179, 189, 193–200, 202–3, 221, 253–4, 257, 259–61, 268, 279, 286, 301 as language/dialect criterion 271

343

lexicostatistics 8, 10, 183–92, 257, 283, 286, 297, 300 cut-off point 10, 189, 283 léxis 17–18 lingua 4–5, 7, 31–9, 42–3, 57–8, 62–81, 90, 95, 102–3, 120–1, 131, 167, 181, 194 communis 5, 60, 67–8, 80, 84 as language of learned men 60 communissima 69 diversa 202 ipsissima 150 polysemy of 62 primigenia 113 principalis 61 tota 111 linguagium 60 linguistic corruption 177, 212 linguistic distance (Abstand) 271, 273, 282, 301 bimodal distribution of 10, 260–2, 283 as a criterion 200, 258, 271 quantification of 8, 10, 189, 257, 259–61, 281, 283 linguistic horizon 43, 49–52, 183, 296 linguistic kinship 191, 195, 220 degree of 187–90, 301 as a continuum 189, 193 as a language/dialect criterion 251–3 quantification of 256–7, 276, 301 linguistics 2, 11, 240, 244, 272, 276–8, 300–1 as an autonomous field of study 2, 9, 225, 229–30, 266, 293, 298–9 basic assumptions of 242 cognitive 276 computational 282 contact 234, 249 creole 234 descriptive 259 dialectological turn in 263 general 266 generative 11, 244–5, 259, 263–8, 280–2, 298 Minimalist programme 265 Optimality Theory 265 Principles and Parameters theory 265 geographical 246 historical-comparative 9, 139, 219–23, 225, 228, 230, 245, 248, 252, 298 history of xiii, 80, 225, 278, 299–300 vs. lay ideas 277, 280, 293–4 linguistics 211 mainstream 243, 245, 274, 280 missionary 49–50 modern 186–7, 219, 268, 299 and other disciplines 294–5

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

344



linguistics (cont.) pre-comparative 143–4, 146, 177, 186, 188 social turn in 253, 281 socio- 11, 186, 244, 259, 263, 265–70, 272–3, 276, 279, 281 externalist approach of 268, 281 structuralist 10, 244–60, 263, 266, 280–1, 298 in the USA 255, 275 textbooks 248, 281 typology 277 variationist 293 Lipsius, Justus 201 literary dialects 250 literary language 254 littera 198 Livy 100 loan translation 60 local dialect 248 locutio 39 lógos 208 Longobards 144 longue durée 3 loquela 38–9 Lord’s Prayer 157 Lucian 66 Ludolf, Hiob 157–8 Ludovici, Michael 157 Luther, Martin 69, 71, 86, 96–7, 118, 124 Lycaonian 212–13 as an Ancient Greek dialect 212 lyric poetry 15 Maaler, Josua 47 Macaulay, Ronald K. S. 266–7 Machan, Tim William 85 macro- vs. micro-parametric variation 265 Malay language 261 Indonesian 261 Malaysian 261 Malherbe, François de 127 Manutius, Aldus 99 de la Mare, William 30 Martial 62 Martinet, André 10, 249, 251–3, 282–3, 301 Martini, Johann Georg 117, 206 Mass, language of 122, 127, 212 mass education 11, 288–9, 293, 301 mass media 288–9, 293 Matthias de Miechow, see Miechowita, Maciej Max Planck Institute for Evolutionary Anthropology 277 Maxwell, Alexander 268, 270 May, Stephen 289

Mayer, Bartholomaeus 199–200 Mazzarella-Farao, Francesco 116 Meillet, Antoine 248, 252–3 Melanchthon, Philipp 68–9, 74, 80, 91 Merton, Robert K. 109–10, 155–6 Merula, Paulus 110–11, 113, 136–8, 140, 144–5 metalanguage 16, 18, 24, 31, 35, 80, 131, 136, 139, 147, 149–50, 167, 173, 238, 262, 278, 283, 289, 299–300 metaphor 178 family 139, 141, 215–16, 226, 298 Meyer, Georg 142 Meyer, Paul 241–2, 245–6, 274 Miechowita, Maciej 60 Milton, John 89 minority languages 271, 278, 290–2 Mischsprache 223 Mithridates , king of Pontus 52, 54, 113–14 Modern Greek 263, 280 dialects of 280 Cappadocian 280 Cretan 280 Cypriot 280 phonology 263 regional accents 280 standard 280 Mokyr, Joel 204 monogenesis of language 78–9, 187–8, 200 Morhof, Daniel Georg 132 morphology 27, 142–3, 163, 190–1, 195, 197–8, 200, 202, 221, 278, 301 inflection 199–200 as language/dialect criterion 198 morphosyntax (language/dialect criterion) 10, 253–4 Morpurgo Davies, Anna 228, 231–3, 240, 242 Morrhy, Gérard 83 Mosellanus, Petrus 72 Moser, Ingunn 3 Müller, Andreas 157 Müller, Max 227 multidialectal outlook 51, 84, 296 multi-linguis 113 Mundart 220, 222, 225–6, 235, 238, 286 Hauptmundart 116 Untermundart 235–6 musicality of language 239 mutual intelligibility 6–7, 9–10, 60, 117–23, 158, 164, 166–7, 176, 201, 215, 256–62, 271–2, 282, 286, 301 asymmetric nature of 258, 292 as a contextual factor 261

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

 as a language/dialect criterion 10, 60, 118, 147–8, 161, 229–30, 246, 248–9, 255–62, 267, 271, 287, 292, 294 arbitrariness of 229, 235–6, 246, 256, 287 in laypeople’s ideas 256, 287, 299 objectivity of 292, 299 rejection of 236, 257, 261, 271 utility of 246, 259 vagueness of 215 and good will 258–9 gradational nature of 235, 246, 258–9, 292 immediate 6, 119–22, 124 lack of 6, 16, 41, 75, 96–7, 118, 164, 269, 286–7, 301 in relation to social class 118 measuring of 271 objective definition of 229, 287 and previous knowledge 258 as strictly binary 122, 124, 164 as a surface phenomenon 259 testing 10, 256–9, 282 threshold 259 and uneducated speakers 287 mutual laughter 36, 43 mutual recognizability 272 Mylius, Abraham 7, 111–12, 121, 140 Myricaeus, Johann Caspar 121 mythology 185 Nahuatl (Mexican) 145 Napoleon 225 nation 6, 33, 43, 101–8, 241, 285 and dialect 101–5 natio 33, 43, 101 nation 212 nation-state 85, 267, 276, 288–9, 292–4, 301 national community 84–5 national language 85, 252, 289 nationalism 85, 287–9, 292, 296 native dialect 70 natural history 227 Renaissance invention of 296 natural language 247 development 228, 273 natural philosophy 31 nature 239 neogrammarians (Junggrammatiker) 9, 228–9, 234–5, 241 New Guinea, languages of 257 Newton, Brian 263–4 Nibbe, Johann Barthold 210 Nicholas of Lyra 40 normative linguistic thinking 22, 128 NORMs 228

345

number of languages (total) 55–6, 64–5, 88, 229, 261, 272, 284 infinity of 221 numeral 187 Oakland Ebonics controversy 11, 290, 293 obliteration by incorporation 109–10, 155–6 Ogilvie, Brian W. 47, 53, 296 optics 28 Oriental studies 173 orthography 202, 207, 267 Otfrid 185–6 Ottomans 171 pánta rheî 239 paradiastrophḗ 153 paragōgḗ 16 Paris, Gaston 241–2, 245, 274 parole 10, 245, 247 pars grammaticalis 198 parts of speech 74 Pascale, Carlo 131 Pasor, Georg 101, 208 Patavinitas 100 pathology 21, 59, 197 patois annihilation of 288–9 as corrupt 98, 128, 131, 212 patois 97–8, 132, 224–5, 227, 241, 251–2 negative connotations of 252 as pure 132 as remnants of feudal society 289 as supplement to the conceptual pair 252, 282 patriotism 69 and dialect description 210 Paul, Hermann 229, 253 Pelasgian (as the ancestral language of Greek) 214 Pennisi, Antonino 202 Perotti, Niccolò 62–3, 100 Persian (Indo-Iranian language) 49, 227 kinship with Dutch 201 Persons, Robert 196 Peter (apostle) 39 Peters, Manfred 57–8 Petrarch 51, 202 Peucer, Caspar 144 peuple 212 Pfeiffer, August 117, 206 Phillips, Edward 106 philology 2, 8–9, 11, 20–2, 27, 147, 151, 155, 159, 183, 194–5, 199, 204, 206–7, 211, 215, 219, 221, 233–4, 237, 245, 284, 289, 301 phōnē ̀ eggrámmatos 18, 20

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

346



phonetics 27, 202, 238–9, 253, 257, 268, 279 phonology 10, 259–60, 268, 271, 279, 301 and structuralism 249 phraseology 142–3, 161, 163, 174, 178–9, 195, 202 Phrygian 213 as an Ancient Greek dialect 213 physiology 265 Piccolomini, Enea Silvio (Pope Pius ) 25 Pillot, Jean 130 Pindar 22 Plato 16, 57, 153 Plautus 23 Plesch, Véronique 90 Pliny the Elder 75 Plutarch, pseudo- 21 poetical license 285–6 Polák, Václav 10, 251, 253–4, 257, 264 Polier de Bottens, Antoine-Noé de 181 Pollux, Julius 17 Polybius 75 polyglot 51, 113–14, 123, 143 Poole, Matthew 100 Pop, Sever 251, 253 popular knowledge 2 popularization of knowledge 285, 290, 293 popolo 102 populus 101 Porphyry 54 postcolonial studies 276, 294–5 Pott, August Friedrich 222, 229, 240 Prague linguistic circle 253 prehistory 8, 185 prescriptivism 290, 294 primary dialect 277 vs. secondary 116 primeval language 177, 181, 200 principle 196 print-capitalism 85 printing press 5, 51–2 commercialization of 5, 51 invention of 90–1, 296 and standardization 85–6 Priscian 34 Prodigal Son, Parable of 225 pronoun 187 demonstrative 34 Greek 82 pronunciation 7, 9, 33, 111, 157, 161, 163, 166, 174, 178–9, 187, 194–200, 202–3, 216, 221, 241, 268, 285 common 161 correct vs. incorrect 241 vulgar 157 propago 61, 137, 152, 196, 213

prosody 179 Protestantism 5, 85–6, 92, 207, 212 Proto-Indo-European 163, 221, 225, 231 as unitary 233 vowel system 245 protolanguage 173, 221, 280 reconstructed 225 prototypicality 276 proverbs 196 province (provincia) 98–9, 165, 167, 212, 215, 241, 247 provincial dialect 252 purity of dialects 294 Puttenham, George 130 quadrivium 28 Quechua 264 questione della lingua, see language questions questionnaire 245, 293 Quintilian 23–4, 59, 63, 72, 78, 90, 100 race 241 Raphelengius, Franciscus 201 Rask, Rasmus 219–21 Rastell, John 75–6 rationalism 8, 192, 203–4, 208, 211, 216, 286, 297 Ravis, Christian 8, 112–13, 141, 166, 195–6, 199 Ray, John 133 Reconquista 51 Reduction 187–8 Reformation 85–6 regional dialects 280 register 275 Reland, Adriaan 171 related dialects 6, 188–9, 202–3, 205, 222, 224, 235, 248, 253, 256, 264, 271, 283, 286, 300 related languages 187–90, 222, 224 Renaissance 2–4, 6, 27, 51–2, 62–4, 71, 80, 84–5, 87, 91, 106, 126, 146, 156, 167, 193–4, 230, 282–3, 289, 296, 298, 301 collector’s mania 87 conception of space 99 Italy 21, 24, 26 multilingual perspective of 52 scholarship 297, 301 Renan, Ernest 227 Republic of Letters 51, 63, 90–1 Rhellicanus, Johannes 58, 75, 79 rhetoric 15 Rhodes 15, 64 rhythm of language 239 Ribier, Guillaume 122 Rice, Eugene F. 91 Richey, Michael 210

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

 richness of dialects 130, 208, 210 Rider, John 73 Rijcke, Theodorus 213 Rocher, Rosane 220 Rollin, Charles 211–12 Roman Empire 15 patois of 212 Romance languages 26, 36, 38, 68, 113, 177, 205, 225, 234, 269 as dialects 234, 237, 239 Gallo-, see Gallo-Romance Ibero-, see Ibero-Romance Italo-, see Italo-Romance Rhaeto- 139 Romansh 234 Romanian 139 Sardinian 139 Romanticism 286 and the use of dialects 286 Ronat, Mitsou 268 Ronsard, Pierre de 71 roofing 64, 128, 201, 271, 279 low variety 201 roofless dialects 271 roots (of words) 9, 111–12, 142–3, 161, 163, 165–6, 187, 190–1, 195–6, 199, 203 Rosier-Catach, Irène 42 Ruhig, Philipp 144 sacred languages 30 salience 276 Salish languages 256–7 Sanskrit 219–20, 222–3, 226, 246 Hindi 227, 261 Urdu 261 Sapir, Edward 248–9, 257 Saporta, Sol 264 Sappho 15 Sardism (Sardismós) 23–4 Saumaise, Claude de 8, 103–5, 107, 115–16, 166 Saussure, Ferdinand de 10, 239, 244–8, 250, 273 Scaliger, Joseph Justus 7, 118, 136–41, 144, 146, 152–3, 196 Schäfer, Gottfried Heinrich 21 Schleicher, August 9–10, 225–7, 229–31, 233, 235, 237, 242 Schmeller, Johann Andreas 221 Schmidt, Erasmus 95, 153, 198, 206 Schmidt, Johannes 10, 231–4, 237–8 Schnell, Rüdiger 27 Schöffer, Peter 91 Schoppe, Caspar 105, 115–16 Schottel, Justus Georg 97, 105, 126, 132–3, 150, 152

347

Schuchardt, Hugo 10, 231–43, 249, 253, 274, 298 Schultens, Albert 8, 171–82, 189–90, 193, 196, 200, 204, 206, 208, 211–12, 297 Schultens, Hendrik Albert 173 Schultens, Jan Jacob 173 Schwartz, Christian Gottlieb 125, 132–3, 204–6, 211 science communication 301–2 Scots 111 Scythian hypothesis (Scythian as protolanguage) 162–3, 165, 219 Sechehaye, Albert 245 Séguy, Jean 276 semantics 5, 187, 197, 241, 301 of words (lexico-) 163, 174, 178–9, 188, 190, 195–6, 215, 253 as dialect-level variation 253 Semitic languages (also termed Oriental in the past) 36, 38, 49, 73, 116, 141–4, 162, 165, 172–7, 181, 183, 185, 189, 195–6, 199–200, 206–7, 211, 216 Arabic 28, 30, 49, 79, 112, 121, 141–2, 164, 171–3, 175–7, 179, 181, 185 antiquity of 198 Cairo 261 dialects 190 Moroccan 261 Aramaic (Chaldean) 28, 36–7, 111–12, 115, 141, 164, 175–6, 185, 199 Biblical 30 Galilean 39–40, 117 Jerusalemite 117 Judean 40 Canaanite 164 Ethiopian 112, 164, 185 Hebrew 19–20, 28–30, 34, 36–7, 39–40, 49, 53, 73, 79, 95, 98, 103, 111–12, 117, 121, 125, 141–3, 162, 164–5, 171–6, 179–81, 185, 198–9, 205, 239 alphabet 198 as a dialect 162 dialects of 165 etymology 173 lexicon 198 as the primeval language 79, 98, 119, 138, 141–3, 162, 172–3, 181 variation in 39, 98 Punic 73 Samaritan 112 Syriac 49, 73, 79, 112, 115, 121, 141, 175–6, 179, 185, 199 Sèrio, Luigi 201–2 sermo 38–9, 57, 120 Shem 79

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

348



shibboleth 39–40, 165, 238–9 Sinitic languages Cantonese 292 Chinese 264, 269 Dongshan 261 Fuzhou 261 Mandarin (Standard) 292 politically defined 269 Shanghainese 292 Skinner, Quentin 3 slang 102, 254 Slavic languages 37, 75, 190, 254, 272 Bosnian 261 Bulgarian 254 Croatian 261 Czech (Bohemian) 37 as one language 75, 102 Polish 37 Russian 37, 227 Rusyn (Ruthenian) 291 as a Ukrainian dialect 291 Slavonic 37 Sorbian 287 social media 293 sociology 259, 271–4, 282, 294 of science 109 sound changes 234 sound laws 223–4 without exception (ausnahmslose Lautgesetze) 228, 235, 241 speach 130 speech organs 285 Speroni, Sperone 50 spiritus 161 Spitzer, Leo 235 Sprachbund (union des langues) 253 Sprache 189, 216, 225–6, 235 Sprachlandschaft 249 Sprachmischung 234 Sprachphilosophie 188, 192 Sprog 221 Sprogart 221 Stammbaumtheorie 10, 231–3, 236 Stammwort 187 standard colloquial 248 lack of 33 language 7, 11, 71, 132–4, 149–50, 154, 250–2, 265, 278–80, 286, 288 concept of 22, 127–9 and dialect 11, 132–3, 214, 247, 279, 294 ideologies 11, 288–90, 294 imposition of 288 and mutual intelligibility 292

and nation-building 288, 293 and power 288, 290, 293 and progression 290, 293 regionalized 277 and social class 149, 285–6 literary 248–9 provincial 248 standard 122, 127, 214 sub- 248 standardization 5–6, 25, 43, 50–1, 61, 69, 84–6, 88, 91–2, 106, 115, 117, 128–9, 202, 267, 296 acceptance 50, 128, 157 codification 50, 85, 128–9, 133, 157, 267, 271 competition among languages 88 elaboration 50–1, 85, 128–9, 133, 271, 278, 282 and language planning 273 as language regulation 127 norm 6, 50–1, 68, 125, 133, 289 processes of 130, 149, 157 selection 50–1, 85, 115, 117, 128, 130, 134, 247, 296 competition among dialects 61, 85 Stanihurst, Richard 73 Stapleton, Thomas 126 statistics 214, 256, 260 stereotyping 85 Stevin, Simon 100 Stiernhielm, Georg 7–8, 109, 112, 116–18, 144, 151, 159–67, 176, 182, 195–6, 208, 297 Stobaeus, Johannes 54 Strabo 17, 20–1 Strodtmann, Johann Christoph 210 style 5, 76, 78, 174, 197, 202, 253–4 vs. dialect 205 subdialect 114, 116, 226, 236, 241 subdialect(us) 116–17, 153, 252, 275 substantial variation 5–7, 9–10, 59, 76, 80, 104, 109–24, 148, 161, 163–6, 174–6, 195, 198, 203, 227, 253, 265, 271 Suda 19 Suetonius 15–16, 64 surface structure 259, 263–5 Swadesh, Morris 10, 257, 259–60, 271, 281–3 syllable 161, 163 derivation 190 syntax 27, 77, 142–3, 161, 174, 178–9, 194, 196–7, 205, 265 as a language/dialect criterion 202 Tacitus 24–5, 75 Tavoni, Mirko 2 terminology 238, 278–9, 282, 289, 293, 295, 300 success of 299

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

 theology 28, 214 Thomas, Jacques 98 Thryllitsch, Georg Friedrich 206 Thucydides 16 Tiberius, Roman emperor 15–16 timelessness of language structure 253 tópos 19, 96, 101 Tory, Geoffroy 88–90 Tourtoulon, Charles de 241 transformation (generative concept) 263–4 translation theory 36, 42 travel 51 diary 277 Treuge, Michael Gottlieb 206 trilingual colleges 49 Collège Royal (Paris) 49 Collegium Trilingue (Leuven) 49 Trissino, Gian Giorgio 99 trópos 16 Trudgill, Peter 267–8 Tryphon 21 túpos 19 Twitter 293 typophilia 241–2 typophobia 241–2 Ulfilas 162 universal constraints 265 ranking of 265 universal language 113 universities, medieval 28–9 unrelated languages 187–90 Uralic languages 144 Urceo, Antonio, see Codro usage 100 bon usage 127 Valckenaer, Lodewijk Caspar 180 Valente, Luisa 42 Valla, Lorenzo 26 Van Hal, Toon 144 Van Keymeulen, Jacques 274, 279 Varchi, Benedetto 68, 99, 102 variability of human culture 42 of language 17, 215–16 variety 250, 268, 275 flexibility of 268 more neutral than dialect 250, 268 Varro, Marcus Terentius 59 Vergara, Francisco de 97 vernacular 252 vernacular dialects 51, 73, 88–9, 128, 148, 183, 207, 210–11, 216, 228, 296

349

book-length treatments of 211 in parallel with Greek dialects 88–9, 155, 206 vernacular Greek (early modern) 118, 122 dialects of 118 ecclesiastical dialect 123 variation in 152 vernacular languages 5, 25–6, 31, 33, 43, 50–1, 61, 66, 68–71, 75, 84–6, 88, 91–2, 96, 106, 117, 127, 129, 139, 149, 157, 165, 183, 196, 250 emancipation from Latin 85 and nations 106 vices of dialects 130 vir trilinguis 30, 49 Vives, Juan Luis 66, 69, 91, 128 Voegelin, Carl F. 10, 256–7, 262 vowel 190, 201 contraction 197 Vuidius, Robertus 194 vulgare illustre 26 Wachter, Johann Georg 9, 200–1, 203 Walper, Otto 125–6, 194 WALS Online 278 Walton, Brian 122 wanted notices 285 Warner, Levinus 171–3 Weidmann, Sandra 53–4 Weinreich, Max 1, 249, 302 Weinreich witticism 1, 158, 249–50, 268–70, 282, 288, 292, 302 attributed to Noam Chomsky 270 misuses of 268, 270 Weinreich, Uriel 10, 247–51, 254, 256, 263–6, 283, 298 Wenker, Georg 228, 286, 298 Whitney, William Dwight 225–8, 230 Wichmann, Søren 260–1, 282 Wilkins, John 132 Williams, John 122, 127–8 Williram 185–6 Wolf, Hieronymus 89 Wolfram, Walt 290 Wood, Robert 128, 213–14 Wyss, Caspar 132, 208–9 Yiddish 1, 249 YIVO (Yiddish Scientific Institute) 1 Zapotec 50 Zeno 15–16 zoology 52–3 Zygomalas, Theodosius 118

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 25/8/2020, SPi