The Cambridge handbook of African Linguistics 9781108283991, 9781108417983, 2018012835

Table of contents :
Front Matter
Contents
Contributors
Preface
General Introduction
Abbreviations and Acronyms
1 - African Linguistics Conceptions and Scope
2 - African Linguistics in Europe
3 - African Linguistics in North and Francophone Africa
4 - African Linguistics in North-Eastern and Anglophone Africa
5 - African Linguistics in Luso, Hispanophone, and Southern Africa
6 - African Linguistics in the Americas, Asia, and Australia
7 - Comparative African Linguistics
8 - A Typological and Areal Perspective of African Languages
9 - Niger-Congo Linguistic Features and Typology
10 - Afroasiatic Linguistic Features and Typologies
11 - Features and Typologies in 'Nilo-Saharan' Languages
12 - Features and Typologies in ‘Khoisan’ Languages
13 - Sociocultural and Political Settings of Language in Africa
14 - Into and Out of Africa
15 - Language Endangerment and Language Documentation in Africa
16 - Situated Language Use in Africa
17 - Special-Purpose Registers of Language in Africa
18 - African Languages in Urban Contexts
19 - African Languages and Mobile Communication
20 - Multilingualism and Education in Africa
21 - Intellectualization of African Languages
22 - African Languages and Human Language Technologies
References
Index

Citation preview

The Cambridge Handbook of African Linguistics This book presents an in-depth and comprehensive state-of-the-art account of the study of ‘African languages’ and ‘language in Africa’ since its beginnings as a ‘colonial science’ at the turn of the 20th century in Europe. Compiled by 55 internationally renowned scholars, this groundbreaking account looks at past and current research on ‘African languages’ and ‘language in Africa’ under the impact of paradigmatic changes from ‘colonial’ to ‘postcolonial’ perspectives. It addresses current trends in the study of the role and functions of language, African and other, in pre- and postcolonial African societies. Highlighting the central role that the ‘language factor’ plays in postcolonial transformation processes of sociocultural modernization and economic development, it also addresses more recent, particularly urban, patterns of communication and outlines applied dimensions of digitalization and human language technology. H. EkkEHard Wolff is Professor and Chair (emeritus) of African Linguistics at Leipzig University. He publishes widely on descriptive, typological, comparative, and applied linguistics and sociolinguistics of African languages, and on oral literatures. He has published over 25 books including Sprachkunst der Lamang (1980), Referenzgrammatik des Hausa (1993), The Lamang Language and Dictionary (2 vols., 2015), Language and Development in Africa: Perceptions, Ideologies and Challenges (2016), Multilingual Education for Africa: Concepts and Practices (2016), Multilingualism and Intercultural Communication: A South African Perspective (2017), and A History of African Linguistics (to appear 2019).

Downloaded from https://www.cambridge.org/core. University of Toronto, on 06 Jan 2020 at 10:39:36, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

cambridge handbooks in language and linguistics

Genuinely broad in scope, each handbook in this series provides a complete state-of-the-ield overview of a major sub-discipline within language study and research. Grouped into broad thematic areas, the chapters in each volume encompass the most important issues and topics within each subject, offering a coherent picture of the latest theories and indings. Together, the volumes will build into an integrated overview of the discipline in its entirety.

Published titles The Cambridge Handbook of Phonology, edited by Paul de Lacy The Cambridge Handbook of Linguistic Code-switching, edited by Barbara E. Bullock and Almeida Jacqueline Toribio The Cambridge Handbook of Child Language, Second Edition, edited by Edith L. Bavin and Letitia Naigles The Cambridge Handbook of Endangered Languages, edited by Peter K. Austin and Julia Sallabank The Cambridge Handbook of Sociolinguistics, edited by Rajend Mesthrie The Cambridge Handbook of Pragmatics, edited by Keith Allan and Kasia M. Jaszczolt The Cambridge Handbook of Language Policy, edited by Bernard Spolsky The Cambridge Handbook of Second Language Acquisition, edited by Julia Herschensohn and Martha Young-Scholten The Cambridge Handbook of Biolinguistics, edited by Cedric Boeckx and Kleanthes K. Grohmann The Cambridge Handbook of Generative Syntax, edited by Marcel den Dikken The Cambridge Handbook of Communication Disorders, edited by Louise Cummings The Cambridge Handbook of Stylistics, edited by Peter Stockwell and Sara Whiteley The Cambridge Handbook of Linguistic Anthropology, edited by N.J. Enield, Paul Kockelman and Jack Sidnell The Cambridge Handbook of English Corpus Linguistics, edited by Douglas Biber and Randi Reppen The Cambridge Handbook of Bilingual Processing, edited by John W. Schwieter The Cambridge Handbook of Learner Corpus Research, edited by Sylviane Granger, Gaëtanelle Gilquin, and Fanny Meunier The Cambridge Handbook of Linguistic Multicompetence, edited by Li Wei and Vivian Cook The Cambridge Handbook of English Historical Linguistics, edited by Merja Kytö and Päivi Pahta The Cambridge Handbook of Formal Semantics, edited by Maria Aloni and Paul Dekker The Cambridge Handbook of Morphology, edited by Andrew Hippisley and Greg Stump The Cambridge Handbook of Historical Syntax, edited by Adam Ledgeway and Ian Roberts The Cambridge Handbook of Linguistic Typology, edited by Alexandra Y. Aikhenvald and R. M. W. Dixon The Cambridge Handbook of Areal Linguistics, edited by Raymond Hickey The Cambridge Handbook of Cognitive Linguistics, edited by Barbara Dancygier The Cambridge Handbook of Japanese Linguistics, edited by Yoko Hasegawa The Cambridge Handbook of Spanish Linguistics, edited by Kimberly L. Geeslin

Downloaded from https://www.cambridge.org/core. University of Toronto, on 06 Jan 2020 at 10:39:36, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

The Cambridge Handbook of African Linguistics Edited by H. Ekkehard Wolff

Downloaded from https://www.cambridge.org/core. University of Toronto, on 06 Jan 2020 at 10:39:36, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

University Printing House, Cambridge CB2 8BS, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06-04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781108417983 DOI: 10.1017/9781108283991 © Cambridge University Press 2019 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2019 Printed in the United Kingdom by TJ International Ltd, Padstow Cornwall A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Wolff, Ekkehard, editor. Title: The Cambridge handbook of African linguistics / edited by H. Ekkehard Wolff. Description: Cambridge, United Kingdom ; New York, NY : Cambridge University Press, 2019. | Includes bibliographical references and indexes. Identiiers: LCCN 2018012835 | ISBN 9781108417983 (hardback : alk. paper) Subjects: LCSH: African languages. | Linguistics—Africa. Classiication: LCC PL8005 .C36 2018 | DDC 496—dc23 LC record available at https://lccn.loc.gov/2018012835 ISBN 978-1-108-41798-3 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Downloaded from https://www.cambridge.org/core. University of Toronto, on 06 Jan 2020 at 10:39:36, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

Contents

page vii viii xi xii xvi xix xxiv

List of Figures List of Tables List of Maps List of Contributors Preface General Introduction List of Abbreviations and Acronyms Introduction 1 African Linguistics: Conceptions and Scope

1

H. Ekkehard Wolff

Part I Short Regional Histories of African Linguistics 2 African Linguistics in Europe Roland Kießling, Nina Pawlak,

Alexander Zheltov, and Arvi Hurskainen

3 31

33

3 African Linguistics in North and in So-Called Francophone Africa Abderrahman el Aissati, Yamina El Kirat El Allame,

and Philip Ngessimo Mathe Mutaka

59

4 African Linguistics in North-Eastern and So-Called Anglophone Africa Marie-Claude Simeone-Senelle, Wolbert G. C. Smidt,

Ronny Meyer, Angelika Jakobi, Amani Lusekelo, Bruce Connell, and Akinbiyi Akinlabi

73

5 African Linguistics in So-Called Lusophone and Hispanophone Africa and in Southern Africa Anne-Maria Fehn, Sonja Bosch,

and Inge Kosch

98

6 African Linguistics in the Americas, Asia, and Australia

G. Tucker Childs, Margarida Petter, Shigeki Kaji, Sun Xiaomeng, Yang Chul-Joon, and John Hajek

115

Downloaded from https://www.cambridge.org/core. City, University of London, on 21 Jan 2020 at 22:42:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

vi

Contents

Part II Comparative and Descriptive African Linguistics 7 Comparative African Linguistics Gerrit J. Dimmendaal 8 A Typological and Areal Perspective of African Languages

137

139 166

Bernd Heine 9 Niger-Congo Linguistic Features and Typology

Larry M. Hyman, Nicholas Rolle, Hannah Sande, Emily Clem, Peter S. E. Jenks, Florian Lionnet, John Merrill, and Nicholas Baier 191 10 Afroasiatic Linguistic Features and Typologies Ronny Meyer and H. Ekkehard Wolff 246 11 Linguistic Features and Typologies in Languages Commonly Referred to as ‘Nilo-Saharan’ Gerrit J. Dimmendaal, Colleen

Ahland, Angelika Jakobi, and Constance Kutsch Lojenga

326

12 Linguistic Features and Typologies in Languages Commonly Referred to as ‘Khoisan’ Alena Witzlack-Makarevich and

382

Hirosi Nakagawa Part III African Languages in Cultures and Societies 13 Sociocultural and Political Settings of Language in Africa

417

419

Alamin Mazrui 14 Into and Out of Africa: The Sociolinguistics of European and Asian Languages in Africa and of the African Diaspora

442

Rajend Mesthrie 15 Language Endangerment and Language Documentation in Africa Friederike Lüpke 16 Situated Language Use in Africa Sigurd D’hondt 17 Special-Purpose Registers of Language in Africa

513

Alice Mitchell and Ayu’nwi N. Neba 18 African Languages in Urban Contexts

468 491

Andrea Hollington

and Nico Nassenstein

535

19 African Languages and Mobile Communication: Between Constraint and Creativity Ana Deumert, Ivan Panović,

Dorothy Agyepong, and David Barasa Part IV Applied Perspectives in African Linguistics 20 Multilingualism and Education in Africa Kathleen Heugh 21 Intellectualization of African Languages: Past, Present, and Future Russell H. Kaschula and Dion Nkomo 22 African Languages and Human Language Technologies

Justus C. Roux and H. Steve Ndinga-Koumba-Binza References Index

555 575

577 601 623 645 782

Downloaded from https://www.cambridge.org/core. City, University of London, on 21 Jan 2020 at 22:42:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

Figures

7.1 7.2 9.1 10.1 10.2 10.3 10.4 10.5 10.6 11.1 11.2 11.3 12.1 12.2 12.3 14.1 15.1 15.2 15.3 15.4 15.5

The subclassiication of Semitic page 149 NeighbourNet graph of Eastern Sudanic languages 163 Vowel sequences in Proto-Gbaya 208 Afroasiatic families 251 Proposed genetic classiication of Chadic 257 Genetic classiication of Cushitic 260 Genetic classiication of Omotic 261 Genetic classiication of Semitic 262 Genetic classiication of Ethiosemitic 263 Nilo-Saharan subgroups 328 Koman subgrouping 333 The subclassiication of Eastern Sudanic 357 Cluster constraints 392 Dorsal feature dispersion in the root template 396 Third person free pronouns and genders in Tsumkwe Juǀ’hoan 406 Idealized ive-vowel system of L2 English East Africa 450 Schematic representation of language prototypes for three speakers 479 Schematic representation of the prototype emerging from the shared core of speakers 479 Schematic representation of the convergences and divergences between languages 480 Schematic representation of those inhabitants of Agnack who are iconized and erased from representing the village as ‘Baïnounk’ 481 Grafiti in the grounds of an abandoned hotel in Kabrousse, Casamance 489

Downloaded from https://www.cambridge.org/core. City, University of London, on 21 Jan 2020 at 22:42:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

Tables

7.1 The classiication of Afroasiatic by J. H. Greenberg and subsequent views page 142 7.2 The classiication of Khoisan by J. H. Greenberg and subsequent views 143 7.3 The classiication of Niger-Congo by J. H. Greenberg and subsequent views 143 7.4 The classiication of Nilo-Saharan by J. H. Greenberg and subsequent views 143 7.5 Linguistic isolates 144 7.6 Lexical cognates in some Bantu languages 146 7.7 Cognate sets in Bantu 147 7.8 Correspondence sets between four Bantu languages 147 7.9 Pronominal sufixes in Semitic 149 9.1 Coda consonants in Grassields Bantu 193 9.2 Realization of C2 consonants in Ibibio 196 9.3 Properties of the prosodic stem in Kukuya 197 9.4 Some NC languages with prosodic stem-like asymmetries 198 9.5 Shekgalagari consonants 199 9.6 Glide hardening in Jukunoid 200 9.7 Williamson’s featural analysis of Godié vowels 204 9.8 Degema ATR contrasts 205 9.9 Cishingini co-occurrence restrictions 207 9.10 Likelihood of nasal vowel counterparts to be absent 210 9.11 Proto-Bantu noun class preixes 218 9.12 Joola Eegimaa and Temne noun class preixes 219 9.13 Noni pronouns 223 9.14 Independent incorporative pronouns in Medumba 224 9.15 Possible Niger-Congo verb extensions 225 9.16 Comparison of past tenses in two dialects of Totela 228 9.17 Default subject afixes in Eleme 229 9.18 Tense marking in Eik 241 9.19 Niger-Congo features compared to other African stocks 244 10.1 Consonant system of Kabyle 272 Downloaded from https://www.cambridge.org/core. City, University of London, on 21 Jan 2020 at 22:42:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

List of Tables

10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 10.10 10.11 10.12 10.13 10.14 10.15 10.16 10.17 10.18 10.19 10.20 10.21 10.22 10.23 10.24 10.25 11.1 11.2 11.3 11.4 11.5 11.6 11.7 11.8 11.9 11.10 11.11 11.12 11.13 11.14 11.15 11.16

Consonant system of Ahaggar Tuareg Consonant system of WCh Standard Hausa Consonant system of CCh Lamang Diachronic types of noun plurals in Chadic Sub-morphemic pronominal pluralizer in Chadic Derivative preixes in Berber PNG marking with verbs in Berber Imperative and adhortative marking with verbs in Berber Preverbal PNG marking with verbs in WCh Hausa (Standard) Postverbal person/number marking with verbs in CCh Lamang Imperative and adhortative marking with verbs in WCh Hausa (Standard) and CCh Lamang TAM marking in Berber Figuig: verb root *-t-f- ‘to enter’ TAM marking in WCh Hausa: verb stem (grade 3) shìgá ‘enter’, verbal noun: shìgá-a ‘entering’ TAM marking in CCh Lamang: simple verb kla ‘take’, simple verbal noun: kl-ò ‘taking’ The verbal complex in Berber The verbal complex: linear ordering of functional slots in CCh Lamang The verbal complex: linear ordering of functional slots in WCh Hausa Consonant inventories for Proto-Cushitic, Omotic, and Geez Geez person indexes Inlection of main-clause verbs in Wolaitta Common derivational morphemes Ethiosemitic personal pronouns Proto-Cushitic personal pronouns Omotic possessive/basic pronouns Songhay varieties The consonant system of proto-Songhay Phonemic consonants shared across Koman Regular sound correspondences between B’aga and Koman Reconstructed free pronouns in Gumuz plus Daats’íin Free pronouns in Kadallu Free pronouns in proto-Koman Deictic directionals in Koman and B’aga Incorporated body part terms in B’aga and Koman Relator nouns < Body part terms The vowels of Mangbetu The vowels of Mangbutu-Efe The vowels of Avokaya, Logo, and Omi Vowel harmony in Ngiti nouns The vowels of Lendu-Dja The vowels of Sara

ix

272 273 273 278 285 285 287 287 288 288 289 290 290 290 291 294 294 301 306 308 309 322 323 323 328 330 334 335 335 336 337 339 341 341 344 344 345 345 346 347

Downloaded from https://www.cambridge.org/core. City, University of London, on 21 Jan 2020 at 22:42:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

x

List of Tables

11.17 11.18 11.19 11.20 11.21 11.22 11.23 11.24 11.25 11.26 11.27 12.1 12.2 12.3 12.4 12.5 17.1 17.2 17.3 18.1 18.2 18.3 18.4 19.1 22.1

The consonants of Ngiti Glottalized consonants in Ngiti Independent pronouns in Ngiti Subject pronouns in Ma’di Coreference pronouns in Lendu and Ngiti Coreference pronouns and referential demonstratives in Ngiti The consonants of Alur The widespread preposition kI/ki Case markers in Northeastern Nilo-Saharan Cognate case markers in Fur and Kunama Pronominal enclitics or sufixes in Northeastern Nilo-Saharan The ive Khoisan lineages Cross-Khoisan consonant chart for the root-onset Cross-Khoisan vowel phonemes and distinctive features Dorsal features in V1 and V2 of the Gǀui words (1) [ɟìā] ‘owner’, (2) [ʔúú] ‘send’, and (3) [kh́̀] ‘person’ Cross-Khoisan variation of V1 in terms of distinctive features Lexical items of an Ethiopian spirit possession language Lexical items of the Bafut court variety Lexical replacement strategies in avoidance registers Urban language use in Harare, Zimbabwe Phonological manipulation in Randuk: use of metathesis by Shamasha and mechanics Semantic manipulations in Yarada K’wank’wa Zimbabwean play language and youth language practices Number of articles on Wikipedias in South African languages The big seven group of African languages

347 348 350 352 353 354 359 364 368 369 381 385 390 394 395 395 520 522 527 549 550 550 551 559 629

Downloaded from https://www.cambridge.org/core. City, University of London, on 21 Jan 2020 at 22:42:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

Maps

7.1 8.1 9.1 9.2 10.1 10.2 11.1 11.2 11.3 11.4. 14.1 18.1 18.2

Linguistic isolates page 145 Phonological zones in Africa 183 Distribution of nasality 209 Number of surface contrastive tones in West African languages 214 Chadic languages 255 Chadic languages (insets) 256 The distribution of Nilo-Saharan languages 329 The distribution of B’aga and Koman languages 333 The distribution of Central Sudanic and Ubangian languages 351 The position of the verb in Eastern Sudanic languages and Kuliak 363 European Colonization of Africa in 1914 443 The architect René Schoentjes’s suggested division of Léopoldville from 1932 539 Linguistic gravity of Kiswahili dialects in cities of eastern DR Congo 543

Downloaded from https://www.cambridge.org/core. City, University of London, on 21 Jan 2020 at 22:42:19, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

Contributors

dorothy agyepong Doctoral student, Linguistics Section, School of African and Gender Studies, University of Cape Town, Cape Town, South Africa Colleen ahland SIL International, Dallas, TX, USA; Lecturer, Department of Linguistics, California State University, Long Beach, Long Beach, CA, USA akinbiyi akinlabi Professor, Department of Linguistics, Rutgers University, New Brunswick, NJ, USA Nicholas Baier Postdoctoral Fellow, Department of Linguistics, McGill University, Montreal, QC, Canada david Barasa Postdoctoral Researcher, Linguistics Section, School of African and Gender Studies, University of Cape Town, Cape Town, South Africa Sonja Bosch Professor, Department of African Languages, University of South Africa, Pretoria, South Africa G. Tucker Childs Professor and Chair, Department of Applied Linguistics, Portland State University, Portland, OR, USA Emily Clem Doctoral student, Department of Linguistics, University of California, Berkeley, CA, USA Bruce Connell Associate Professor, Linguistics and Language Studies Program, Glendon College, York University, Toronto, ON, Canada ana deumert Professor, Linguistics Section, School of African and Gender Studies, University of Cape Town, Cape Town, South Africa Sigurd d’hondt Associate Professor, Research Collegium for Language in Changing Society (RECLAS), Department of Language and Communication

Downloaded from https://www.cambridge.org/core. Université du Québec à Montréal, on 08 Jan 2020 at 10:31:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

List of Contributors

xiii

Studies & Centre for Applied Language Studies, University of Jyväskylä, Finland Gerrit J. dimmendaal Professor, Institut für Afrikanistik und Ägyptologie, Universität zu Köln, Cologne, Germany abderrahman El aissati Assistant Professor, Department of Culture Studies, Tilburg University, Tilburg, The Netherlands Yamina El kirat El allame Professor and Vice-Dean for Research and Cooperation, Faculty of Letters and Human Sciences, Mohammed V University, Rabat, Morocco anne-Maria fehn Researcher, Department of Linguistic and Cultural Evolution, Max Planck Institute for the Science of Human History, Jena, Germany; Institut für Afrikanistik, Johann-Wolfgang-Goethe Universität, Frankfurt am Main, Germany; Human Evolutionary Genetics Group, CIBIO/ InBIO: Research Center in Biodiversity and Genetic Resources, Vairão, Portugal John Hajek Professor of Italian Studies and Director of the Research Unit for Multilingualism and Cross-cultural Communication (RUMACCC), School of Languages and Linguistics, University of Melbourne, Melbourne, Australia Bernd Heine Professor and Chair (emeritus), Institut für Afrikanistik und Ägyptologie, Universität zu Köln, Cologne, Germany kathleen Heugh Professor, Research Centre for Languages and Cultures, School of Communication, International Studies and Languages, University of South Australia, Magill, Australia andrea Hollington Researcher, Global South Studies Center, Universität zu Köln, Cologne, Germany arvi Hurskainen Professor (emeritus), Institute of World Cultures, Faculty of Humanities, Helsinki University, Helsinki, Finland larry M. Hyman Professor of Linguistics, Department of Linguistics, University of California, Berkeley, CA, USA angelika Jakobi Senior Researcher (retired), Institut für Afrikanistik und Ägyptologie, Universität zu Köln, Cologne, Germany Peter S. E. Jenks Associate Professor, Department of Linguistics, University of California, Berkeley, CA, USA Shigeki kaji Professor, Department of Sociology, Kyoto Sangyo University; Kyoto University (emeritus), Kyoto, Japan russell H. kaschula Professor and NRF Research Chair for Intellectualisation of African Languages, Multilingualism and Education, African

Downloaded from https://www.cambridge.org/core. Université du Québec à Montréal, on 08 Jan 2020 at 10:31:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

xiv

List of Contributors

Languages Studies Section, School of Languages and Literatures, Rhodes University, Grahamstown, South Africa roland kießling Professor, Afrikanistik, Asien-Afrika-Institut, Universität Hamburg, Hamburg, Germany Inge kosch Professor, Department of African Languages, University of South Africa, Pretoria, South Africa Constance kutsch lojenga Associate Professor (retired), Linguistics Centre, Leiden University, Leiden, The Netherlands; Senior Linguistics Consultant, SIL International florian lionnet Assistant Professor, Program in Linguistics, Princeton University, Princeton, NJ, USA friederike lüpke Professor of Language Documentation and Description, School of Languages, Cultures and Linguistics, SOAS, University of London, London, UK amani lusekelo Senior Lecturer, Department of Languages and Literature, University of Dar es Salaam, Dar es Salaam, Tanzania alamin Mazrui Professor, African, Middle Eastern, and South Asian Languages and Literatures, Rutgers University, New Brunswick, NJ, USA John Merrill PhD, Department of Linguistics, University of California, Berkeley, CA, USA rajend Mesthrie Professor and NRF Research Chair in Migration, Language and Social Change, Linguistics Section, School of African and Gender Studies, University of Cape Town, Cape Town, South Africa ronny Meyer Maître de conférences (Amharique), Institut National des Langues et Civilisations Orientales (INALCO), and Langage, Langues et Cultures d’Afrique (LLACAN), Paris, France; formerly Addis Ababa University, Addis Ababa, Ethiopia alice Mitchell Postdoctoral Research Associate, Department of Archaeology and Anthropology, University of Bristol, Bristol, UK Philip Ngessimo Mathe Mutaka Professor of Linguistics, University of Yaoundé I, Yaoundé, Cameroon Hirosi Nakagawa Professor, School of Language and Culture Studies, Tokyo University of Foreign Studies, Tokyo, Japan Nico Nassenstein Assistant Professor, Institut für Ethnologie und Afrikastudien, Johannes Gutenberg-Universität, Mainz, Germany H. Steve Ndinga-koumba-Binza Lecturer, Department of Language Education, University of the Western Cape, Bellville, Cape Town, South Africa Downloaded from https://www.cambridge.org/core. Université du Québec à Montréal, on 08 Jan 2020 at 10:31:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

List of Contributors

xv

ayu’nwi N. Neba Head of Division Teaching and Teaching Staff, Department of Linguistics, University of Buea, Buea, Cameroon dion Nkomo Associate Professor of African Languages Studies, African Languages Studies Section, School of Languages and Literatures, Rhodes University, Grahamstown, South Africa Ivan Panović Assistant Professor, Linguistics and Multilingual Studies, School of Humanities, Nanyang Technological University, Singapore Nina Pawlak Professor, Department of African Languages and Cultures, University of Warsaw, Warsaw, Poland Margarida Petter Professor, Department of Linguistics, Universidade de São Paulo, São Paulo, Brazil Nicholas rolle Postdoctoral Fellow, Program in Linguistics, Princeton University, Princeton, NJ Justus C. roux Professor (emeritus), Extraordinary Professor, Department of Afrikaans and Dutch, Stellenbosch University, South Africa Hannah Sande Assistant Professor, Department of Linguistics, Georgetown University, Washington, DC, USA Marie-Claude Simeone-Senelle Directrice de Recherche (emerita), Langage, Langues et Cultures d’Afrique (LLACAN), CNRS, INALCO, Université Sorbonne Paris Cité, Villejuif, Paris, France Wolbert G. C. Smidt Director of ‘Ethiomap’, Forschungszentrum Gotha, Universität Erfurt, Erfurt, Germany; PhD advisor and teacher in the PhD programme ‘History and Cultural Studies’, Mekelle University, Mekelle, Ethiopia Sun Xiaomeng Professor, School of Asian and African Studies, Beijing Foreign Studies University, Beijing, P. R. China alena Witzlack-Makarevich Assistant Professor, Linguistics Department, The Hebrew University of Jerusalem, Israel H. Ekkehard Wolff Professor and Chair (emeritus) of African Linguistics, Institut für Afrikastudien, Universität Leipzig, Leipzig, Germany Yang Chul-Joon Humanites Korea (HK) Research Professor, Institute of African Studies, Hankuk University of Foreign Studies, Yongin, South Korea alexander Zheltov Professor and Chair/Head of the Department, Department of African Studies/Department of African Ethnography, St. Petersburg State University/Museum of Ethnography and Anthropology, St. Petersburg, Russia

Downloaded from https://www.cambridge.org/core. Université du Québec à Montréal, on 08 Jan 2020 at 10:31:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

Preface

African linguistics has come of age as a separate academic discipline that, about 130 years after its inception in predominantly German-speaking academia, and 25 years after establishing the international conference series of the World Congress of African Linguistics, had not yet seen the publication of a comprehensive and solitary ‘handbook’ of its own. Therefore, in October 2015, Helen Barton of Cambridge University Press approached the editor with the proposal to design and compile such a volume for their prestigious Cambridge Handbooks in Language and Linguistics series. Beginning in November 2015, the editor inally assembled 55 experts who represent a remarkable mix of generations, professional experiences, genders, current afiliations, and origins. The task was to sketch out the history, the state of the art, and promising perspectives of African linguistics in the early 21st century and to relect on past and current research priorities and recent changes of paradigm. The authors completed their chapter manuscripts between November 2016 and November 2017. The African linguistics community had long deplored the scarcity of foundational literature for their ield. With only few exceptions, representatives of the irst and second generations of Africanists between 1885 and 1970 did not record their thoughts on the theoretical or methodological, not to speak of political and ideological, foundations of their concerns with languages in Africa, apart from individual cases subscribing to national colonial projects. There were hardly any publications before the 1980s, which would describe in greater detail the emergence of African linguistics as an autonomous academic discipline, or outline its speciic research methodology and priorities. A notable exception is volume 7 of Current Trends in Linguistics (Linguistics in Sub-Saharan Africa) of 1971. The Cambridge Handbook of African Linguistics ills a gap and adds – with a rather wide focus – to the dynamically growing literature on African languages and on language issues in Africa, which, however, tend to have rather narrowly focused perspectives.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.024

Preface

xvii

Researchers working in the ield increasingly realized that the peoples speaking African languages in situ remain in dire need of academic guidance, which would give science-based input to postcolonial linguistic, educational, and sociocultural planning. As editor, I wanted to ensure that a ‘handbook’ of African linguistics worth its title not only broadly covered descriptive, typological, and historical-comparative issues of individual languages or groups of languages but also relected the manifold ramiications of language into matters of cultural, social, political, and economic life in Africa and followed these ramiications across the divides of generations and genders, rural and urban communities, educated ‘elites’ and under-educated ‘masses’, and relected the differences between indigenous African and other languages. Such a broad approach not only relects my own background in Germanspeaking Afrikanistik, but also mirrors the transdisciplinary programme of the World Congress of African Linguistics (WOCAL) since it was irst convened in 1994. The structure of the present handbook does exactly this. The handbook includes abridged regional histories of African linguistics, depicting its emergence towards the end of the 19th century as a ‘colonial science’ in Europe and following its global spread in the 20th and 21st centuries. In order to provide a fuller picture, it is accompanied by A History of African Linguistics, written by the same authors and published by Cambridge University Press, which contains more detailed and more fully referenced historical accounts. The editor expresses his gratitude to Cambridge University Press, in particular to Helen Barton and all members of the team. We had just inalized production of my book Language and Development in Africa: Perceptions, Ideologies and Challenges (Cambridge University Press, 2016) when Helen approached me with the idea of compiling The Cambridge Handbook of African Linguistics. She accepted my proposal to have a sister volume, A History of African Linguistics, published in a parallel manner, in order not to overload the handbook with important hitherto unavailable information on the various regional histories of African linguistics, as was admonished by one anonymous reviewer of the original proposal with good reason. I am grateful for her conidence in me to shoulder the double project, and for the continuous and eficient support from her and by her team. I am deeply grateful to all contributors for their professional and focused cooperation, and their concern about observing necessary deadlines. Most authors and co-authors delivered within a time span of twelve to less than eighteen months from the time they accepted the invitation, some continuously improving their chapters in professional dialogue with the editor and colleagues, occasionally comparing notes with authors of other chapters in the project. I was happy to see that many authors followed the editor’s advice to team up with co-authors in order to secure a both topically and geographically wide representation. In fact, geographic author representation ranges from Canada and Finland in the North to Australia, Brazil, and South Africa in the South, from the West Coast of

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.024

xviii

Preface

the United States, via several European countries and Israel, to Russia, China, South Korea, and Japan in the East. It also involves good representation of the African sub-regions: from Rabat, Morocco, in North Africa, via Addis Ababa and Mekelle, Ethiopia, in North-Eastern Africa, Buea and Yaoundé, Cameroon, in West and Central Africa, Dar es Salaam, Tanzania, in East Africa, to Pretoria, Grahamstown, and Cape Town in South Africa. Fortunately, there were only very few withdrawals by prospective contributors, for suddenly emerging unfavourable personal circumstances; none of these withdrawals was critical because other excellent authors took over on rather short notice, yet delaying the completion of the whole book manuscript for several months. However, we lost two originally planned chapters, one on language policies and planning and one on the early history of standardization and literacy development of African languages; both issues are dealt with in other chapters, even though not with the originally planned depth and detail. Finally, I am very grateful to both Helsingin Yliopisto (in particular the Department of World Cultures) in Helsinki, Finland, and to Rhodes University (in particular the Faculty of Humanities and the School of Languages and Literatures) in Grahamstown, South Africa, which each hosted me for periods of ive to six months during preparation stages of this handbook. My Visiting Professorship in Helsinki (2016–2017) was co-sponsored by the DAAD Johann Gottfried Herder Program, and my stay in Grahamstown (2017–2018) was made possible by Rhodes University under a Hugh Le May Fellowship with support from the Andrew W. Mellon Foundation. All this support is gratefully acknowledged. Back at Rhodes, after an earlier six-month research visit in 2014–2015 under the DAAD Johann Gottfried Herder Program, I enjoyed the privilege of a congenial and stimulating academic environment, sharing with my local colleagues a focus on and passion for the intellectualization and re-empowerment of African languages.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:21, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.024

General Introduction

The Cambridge Handbook of African Linguistics presents retrospective as well as prospective views on the scientiic study of African languages on the one hand, and of language in Africa on the other. These are two fundamentally different but related perspectives, which both lie at the core of African linguistics as an autonomous academic discipline. A team of 55 expert authors stemming from or working in Africa, the Americas, Asia, Australia, and Europe came together in order to compile the present volume. Primary focus is on the more than 2,000 African languages, which together amount to almost one-third of all living languages on our planet. This focus covers various perspectives on how these languages can be analysed and described, grouped and classiied, and on how they have inluenced each other in geographical neighbourhood over time through language contact/multilingualism – all this on the basis of established and theory-guided scientiic methods and procedures. Additionally, the book also looks at other languages, namely those that were brought to Africa, as much as at language varieties that emerged within Africa, in historical times. Further, this handbook deals with aspects of language use in various cultures and societies in Africa. It addresses issues of language ideologies and attitudes, and describes how these weigh in on views of language(s) in Africa among professionals and in a wider public, both within and outside of Africa. This handbook updates its outstanding predecessor Linguistics in SubSaharan Africa (Vol. 7 of Current Trends in Linguistics, ed. Thomas A. Sebeok) of 1971, and complements pre-existing introductory volumes to the study of African languages and linguistics, like, for instance, Pierre Alexandre, Langues et langage en Afrique noire (1967; translation: An Introduction to Languages and Language in Africa, 1972); Edgar A. Gregersen, Language in Africa: An Introductory Survey (1977); Bernd Heine, Thilo C. Schadeberg, and Ekkehard Wolff, Die Sprachen Afrikas (1981); Jean Perrot, Gabriel Manessy, and Albert Valdman, Les langues dans le monde ancien et moderne (2 vols., 1981);

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.025

xx

General Introduction

Bernd Heine and Derek Nurse, African Languages: An Introduction (2000; translation: Les langues africaines, 2004); Ngessimo M. Mutaka and Pius Ngwa Tamanji, An Introduction to African Linguistics (2000); and G. Tucker Childs, An Introduction to African Languages (2003). As the present handbook neared completion, two other ambitious publications appeared in print: The Routledge Handbook of African Linguistics (2018, edited by Augustine Agwuele and Adams Bodomo), and The Languages and Linguistics of Africa (2018, edited by Tom Güldemann). Evolution has made Africa the cradle of humanity and the home of human language. Here it evolved and diversiied, and from here it spread across all inhabited continents. This alone makes African linguistics a ield of prime relevance in the overall study of the history of humankind. Furthermore, and since times immemorial, Africa has been and is a hub of multilingualism, which has increasingly complex ramiications into practically all aspects of social, cultural, political, and economic life. Current research into aspects of (applied) African (socio)linguistics, therefore, opens new and in-depth perspectives on the nature of multilingualism in general as much as on its variant manifestations in territorial, sociocultural, individual, and institutional perspectives. These are likely to challenge the prevailing ‘Northern’ and Eurocentric ideological bias in public as much as in academic discourse on Africa and, more generally, on the Global South. In this handbook, expert authors treat African linguistics as being much more than just a geographically focused sub-ield of modern linguistics. No doubt, African linguistics owes much of its theoretical and methodological foundations to developments in general and theoretical linguistics. Vice versa, however, it increasingly contributes challenging insights from the analysis of African language data to the testing and reinement of theoretical assumptions and methodological tools in general linguistics. More than that, African linguistics represents a unique ield of studies that rests on, but also overcomes the limitations of, narrow and so-called hard-core linguistic approaches, with a wide interface to neighbouring social and cultural sciences. On the one hand, African linguistics honours traditional approaches to language in terms of ideology-laden theoretical constructs like, for instance, ‘heritage language’, ‘mother tongue’ / ‘home language’, and ‘(ex-)colonial language’. Consequently, African linguistics works with two complementary approaches to the central object of study. Providing genuinely new empirical data, it delivers – at times monolectal – grammars for hitherto under- or totally undescribed languages, based on a thorough analysis of utterances from individual members of a speaker community. It also deals with both abstract and highly standardized reference systems, like in the case of established African ‘standard languages’. More recently, it explores dynamically changing actual language use in communities of practice, and does so under currently fashionable terms like ‘(trans-)

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.025

General Introduction

xxi

languaging’ and linguistic ‘superdiversity’, thereby calling into question our received notions about ‘named languages’ and ‘multilingualism’. On the other hand, African linguistics allows for various trans- and interdisciplinary perspectives, viewing languages as being essentially embedded in the cultures and societies of African peoples, with which they continuously entertain dynamic interaction. Since the times of the teaching and writing of one of its German founding fathers Diedrich Westermann (1875–1956), African linguistics (in German: Afrikanistik) entertains very close if not essential links with social and cultural anthropology (in German: Völkerkunde) and (oral) history, in addition to phonetics and both general descriptive and comparative linguistics. With this in mind some like to refer to Westermann’s legacy as having established some kind of ‘comprehensive African linguistics’ (in German: Gesamtafrikanistik) that reaches far beyond the scope of hard-core linguistics research on African languages. Third, and in terms of applied science, African linguistics links languages as resources to the aspirations of their speakers in their quest to master their daily routines and to meet the social, cultural, political, and economic challenges of sustainable, including mental, decolonization and of what is sweepingly called ‘development’. African linguistics, thereby and apart from individual language structures and genealogical classiication into language families and branches, addresses the lingering effects of colonialism in terms of continued linguistic and cultural imperialism and the onslaught of globalization. These have a strong impact on language choice and language use by individuals and sociocultural groups of practice in Africa, shaking the foundations of so-called traditional cultures and societies, which entail constant adaptations of both patterns of language use and properties of the languages themselves. The present handbook relects this broad approach to African linguistics, which may appear innovative to some readers but is quite familiar to others, depending on the reader’s academic socialization. Obviously, no handbook can ever be complete with regard to coverage of all potentially interesting and relevant subject matters, the present one being no exception. Readers and reviewers are encouraged to identify gaps and shortcomings and do their best to compile complementary publications in order to create a fuller picture. The contributors and the editor of The Cambridge Handbook of African Linguistics, however, take pride in having cleared the ground and laid foundations, so to speak, for other expert authors to build and expand on the description of African linguistics worldwide. In Part I, the handbook provides abridged surveys of where, since when, and how African linguistics became entrenched in academia on this planet, from the precolonial to the current postcolonial era. The historical overview includes not only the countries of the former Western European colonial powers, where African linguistics originally emerged as a ‘colonial

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.025

xxii

General Introduction

science’ but also their neighbours in Eastern, Central, and Northern Europe, in addition to destiny countries of the transatlantic slave trade in the Americas, the former African colonies and current independent states themselves, and, of more recent vintage, places in Asia and Australia. Part II constitutes the main body of the present handbook. It addresses the core issues of African linguistics, namely the descriptive analysis, comparison, history, and classiication of African languages. Relecting established mainstream lines of research, the handbook offers approaches to African languages within the framework of their partly proven and partly still hypothetical or contested genetic classiication, that is, as members of language phyla and families based on the still valuable yet critically reviewed referential classiication by Joseph H. Greenberg in his seminal work The Languages of Africa (1963). However, in their presentations, the authors share an innovative triple perspective on (a) cross-linguistic language typology and (b) the impact of language contact in terms of areal linguistic approaches, yet (c) not neglecting unilineal language history as reconstructed by classic comparative methods. Part III enlarges the scope of the present handbook to encompass bird’seye views on the overall linguistic situation in Africa, relections on language ideologies and attitudes, and discussions of patterns of language use and the interdependencies of languages with ever-changing surrounding cultures, social norms, and practices in Africa. This includes a focus on the more recent and dynamically increasing impact of urbanization, digitalization, and mobile communication. Part IV, rather selectively, addresses applied perspectives that link African linguistics to issues of societal transformation, cultural modernization, and economic development, through focusing on languages and multilingualism as resources for overcoming mass poverty, academic underperformance, and technological marginalization. Crucially, this involves language-in-education matters, language intellectualization and re-empowerment, and reaping beneits from human language technology. Like presumably most if not all contributors to this volume and many more Africanist colleagues across the planet, the editor believes in a comprehensive approach to status, acquisition, corpus, and opportunity planning for languages in Africa. He believes in providing science-based assistance for designing and implementing mother-tongue-based multilingual language policies for education across the continent, which would serve the ultimate beneit of the peoples who speak African languages as irst or second languages in their homes and workplaces. Unfortunately and to this day, African and non-African intellectuals outside professional linguistic circles, including the political elites, tend to neither listen to nor understand what we Africanists are trying to tell them. We do so in countless academic publications, via formal recommendations from within learned societies, and by resolutions emanating from high-class professional meetings, some of us willing to embrace language activism and

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.025

General Introduction

xxiii

risking degradation from the ranks of ‘pure’ academia. But clearly, for the sustainable empowerment of Africa’s ‘human resources’ to become able to take into their own hands postcolonial social transformation, cultural modernization, and economic development, only adequate education will do the job. ‘Adequate education’ obviously involves effective multilingual communication competencies in both endoglossic and exoglossic languages, because only this will enable African learners to compete successfully – both locally among each other, but also globally with members of their age cohort across the planet. Only mother-tongue-based multilingual quality education will overcome individual limitations of the ‘linguistic jail’ (Ouane 2003) of the mother tongue and of sociocultural ‘underdevelopment’, and will inally turn Africa into a knowledge-producing continent on equal scale with other continents, rather than remaining only knowledge-consuming at the mercy of the ‘North’. The African as much as the global arena is characterized by almost ubiquitous multilingualism in the higher domains of 21st-century verbal communication and knowledge production. Africans from all walks of life must be enabled to not only survive in the prevailing postcolonial and largely underperforming educational systems, but come out on top of qualitatively optimized systems, and thus attain equal opportunities with age-mates in other parts of the world. Targeting this far-reaching goal and supporting the linguistic dimension of endeavours along the way lie at the heart of matters that this handbook is about, apart from introducing and describing African linguistics as a thriving academic discipline across the globe. The Editor

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:35, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.025

Abbreviations and Acronyms

The abbreviations listed below are used across chapters in upper- or lowercase, often in small CAPITALS. Occasionally, the same abbreviation or symbol refers to different categories, or the same categories are represented by different abbreviations or symbols, in different chapters. The abbreviations and symbols are those originally used by the authors of the individual chapters. .I . II ↓ H *L Ø # 1, 2, 3; 1/2/3 1/2/3/4/5 A A AA ABS ACALAN ACC AD ADEA ADJ AFF AFLaT ALT-i ALUPEC ALUSTP ANN

set I person markers (A and S) in Gwama set II person markers (O and S) in Gwama downstepped high tone loating low tone zero marked (unmarked) boundary irst, second, third person agreement class in East and West !Xoon, Tsumkwe Juǀ’hoan agent argument of transitive clause aorist Afroasiatic absolutive African Academy of Languages accusative preverbal particle ad ‘non-realized’ (Berber) Association for the Development of Education in Africa adjective; adjective-deriving formative afirmative African Language Technology African Language Technology Initiative Alfabeto Uniicado para a Escrita do Cabo-Verdiano Alfabeto Uniicado para a Escrita das Línguas Nativas de S. Tomé e Príncipe status annexus

Downloaded from https://www.cambridge.org/core. University of New England, on 07 Jan 2020 at 16:31:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

List of Abbreviations and Acronyms

ANTICAUS APPL ASAFAS ASC, ASSOC ASP ASR AST ATR ATTR AUX BA BAKITA BEN BFSU BLARK BLR BRICS C C, CG CALL CAR CAUS CC CF CF, CFG CIBIO/InBIO CIDA CL1 CL2 CLARIN CLO CMS CNRS CNST CNT CODESRIA COLL COM COMP COP CPL CPT CRLD

xxv

anticausative applicative Graduate School of Asian and African Area Studies associative (marker) aspect automatic speech recognition African Speech Technology advanced tongue root attributor auxiliary baccalaureus artium, bachelor of arts National Swahili Council benefactive Beijing Foreign Studies University Basic Language Resource Kit Bantu Lexical Reconstructions Brazil, Russia, India, China, South Africa consonant common gender Colloquium on African Languages and Linguistics Central African Republic causative geminate consonant clause focus centrifugal Research Centre in Biodiversity and Genetic Resources Canadian International Development Agency class 1 gender class 2 gender European Research Infrastructure for Language Resources and Technology Cahiers de Littérature Orale Church Missionary Society Centre National de la Recherche Scientiique construct state continuous Council for the Development of Social Science Research in Africa collective comitative complementizer copula completive (aspect) centripetal Centre for Research on Linguistic Diversity

Downloaded from https://www.cambridge.org/core. University of New England, on 07 Jan 2020 at 16:31:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

xxvi

List of Abbreviations and Acronyms

CSIR CTexT CVB D, DECL D1 DAT DCL DD1 DD2 DEF DEL DEM DET DH DIA DIR DIST DO DOBES DRC DST DSTR DU EL ELDP EP ERG EXC F, Fm F1 F1, F2, etc. F2 FAC, FACT FLAS FOC FUG FUT FV GALI GEN GILLBT GP GR.1

Centre for Scientiic and Industrial Research Centre for Text Technology converb declarative proximal demonstrative dative declarative deictic directional ‘ventive’ deictic directional ‘itive’ deinite (marker) Documenting Endangered Languages demonstrative determiner digital humanities Dictionnaire Interactif pour Jeunes Apprenants directional distal demonstrative direct object Documentation of Endangered Languages Program Democratic Republic of the Congo distal distributive dual état libre Hans Rausing Endangered Languages Documentation Programme external possession ergative exclusive feminine irst formant future tenses (F1 hodiernal (near)future) second formant factitive, factative Foreign Language and Area Studies focus centrifugal future inlectional inal vowel Gabonese Languages through Internet genitive Ghana Institute of Linguistics, Literacy, and Bible Translation general possession grade 1 (Hausa)

Downloaded from https://www.cambridge.org/core. University of New England, on 07 Jan 2020 at 16:31:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

List of Abbreviations and Acronyms

H H HAB HLT HUFS i/ii IAI ICT IDPH IKS ILCAA IMM IMP INALCO INC INDE INDEF INE INEAS INF INFL INS, INST INT INTR IO IPF, IPFV IRR ISCED ISE ITER ITR JALL JOLAN JUNCT JUSS JWAL L LACITO LAN LBT LCRC LLACAN LLS LOC

xxvii

high tone human habitual human language technology Hankuk University of Foreign Studies tone class International African Institute information and communication technology ideophone Indigenous Knowledge System Research Institute for Asian and African Languages and Cultures immediacy marker imperative Institut National des Langues et Civilisations Orientales inclusive Instituto Nacional de Desenvolvimento da Educaçao indeinite (marker) Instituto Nacional de Estatistica Institute of Near Eastern and African Studies ininitive inlection instrument(al) intensiier intransitive indirect object imperfective irrealis Instituto Superior de Ciências de Educaçã Instituto Superior de Educaçao iterative intransitive Journal of African Languages and Linguistics Journal of the Linguistic Association of Nigeria juncture jussive Journal of West African Languages low tone Langues et Civilisations à Tradition Orale Linguistic Association of Nigeria Lutheran Bible Translators Language and Culture Research Centre Language, Langues et Cultures d’Afrique LexTorah Language series locative

Downloaded from https://www.cambridge.org/core. University of New England, on 07 Jan 2020 at 16:31:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

xxviii

List of Abbreviations and Acronyms

LOG LREC M M, Msc MA MaLEX MDG MED MED MID MIT MoI MPhil MT MT MuST N N NARR NC NEG NELIMO NEPAD NGO NH NINLAN NLP NLU NOM NOM NON-DEF NPST NRF NTR NTS NUFU

NZ O, OB, OBJ OBL OF OGL OLAC P

logophoric pronoun Language Resources and Evaluation midtone masculine magister artium, master of arts Malawi Lexicon Project Millennium Development Goals medial (distance) medial/middle voice middle Massachusetts Institute of Technology medium of instruction magister philosophiae, master of philosophy machine translation system mother tongue Multilingual Speech Technology noun nasal consonant narrative Niger-Congo negation Núcleode Estudos de Línguas Moçambicana New Partnership for Africa’s Development nongovernmental organization non-human National Institute for Nigerian Languages natural language processing National Lexicography Unit nominalizer (for verbs) nominative non-deinite non-past National Research Foundation neutral non-topical subject National Committee for Development-Related Research and Education (Nasjonalt Utvalg for Utviklingsrelatert Forskning og Utdanning) nominalizer object oblique out-of-focus marker Oxford Global Languages Open Language Archives Community patient argument in transitive clause

Downloaded from https://www.cambridge.org/core. University of New England, on 07 Jan 2020 at 16:31:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

List of Abbreviations and Acronyms

P, PERS P1, P2 etc. PanSALB PART PASS PAST PB PERF, PFV, PF PhD PL PL, PLA, PLUR PLAGR PNC PNG POSS PossN PossPro PRED PREP PRES, PRS PRO PROG PROH PROX PST PST1 PTCP Q R RCD RCLT REAL REC REC. PST RED REL REM. PST REV RFL RMA RTR S, SBJ, SU, SUBJ SADiLaR SAL SARChI

xxix

person past tenses (P1 immediate past, P3 distant past) Pan-South African Language Board discourse particle passive past tense Proto-Bantu perfective philosophiae doctor; doctor of philosophy plural pluractional plural agreement Proto-Niger-Congo person/number/gender possessive possessive noun possessive pronoun predicative particle preposition present tense pronoun progressive prohibitive proximal past (tense) recent past participle question (marker), interrogative root reductive-causative-downward Research Centre for Linguistic Typology realis reciprocal recent past reduplication relative (clause marker) remote past reversive relexive Language Resource Management Agency retraction of the tongue root subject South African Centre for Digital Language Resources Studies in African Linguistics South African Research Chair Initiative

Downloaded from https://www.cambridge.org/core. University of New England, on 07 Jan 2020 at 16:31:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

xxx

List of Abbreviations and Acronyms

S-Aux-OVX SCN SeLA SELAF SEQ SG SGL SIDA SIL SLTU SM SNG SOAS SOV SOVX SSC SSHRC STAT SUNY SUSO SVC SVO SVOX T TAM TGG TNS TOP TP TQ TR TTS TUFS TWB UBS UCB UCLA UIL UNICEF UPSID UQAM USC

subject-auxiliary-object-verb-any other constituent serial verb connective Scientiic e-Lexicography for Africa Société d’Études Linguistiques et Anthropologiques de France sequential-conjunctive singular singulative Swedish International Development Cooperation Agency Summer Institute of Linguistics Spoken Language Technologies for Under-resourced Languages subject marker singular verb stem School of Oriental and African Studies subject-object-verb subject-object-verb-any other constituent same subject converb Social Sciences and Humanities Research Council stative State University of New York Uniied Orthography of a Uniied Language called Shona serial verb construction subject-verb-object subject-verb-object-any other constituent tense node tense-aspect-mood (system) transformation-generative grammar tense topicalization (marker) tense phrase term question transitive text to speech system Tokyo University of Foreign Studies Translators without Borders United Bible Societies University of California, Berkeley University of California, Los Angeles UNESCO Institute of Lifelong Learning United Nations Children’s Fund UCLA Phonological Segment Inventory Database Université du Québec à Montréal University of Southern California

Downloaded from https://www.cambridge.org/core. University of New England, on 07 Jan 2020 at 16:31:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

List of Abbreviations and Acronyms

Ṽ V V VEN VSO VV WALS/SLAO WOCAL X

xxxi

nasal vowel verb, predicate vowel venitive verb-subject-object long vowel West African Linguistic Society/Société Linguistique de L’Afrique Occidentale World Congress of African Linguistics clausal complement

Downloaded from https://www.cambridge.org/core. University of New England, on 07 Jan 2020 at 16:31:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991

1 African Linguistics: Conceptions and Scope H. Ekkehard Wolff

1.1

What Is African Linguistics?

‘African linguistics’ is a potentially misleading term. It links up with the German label Afrikanistik (or Afrikalinguistik), under which African linguistics was irst established as an autonomous discipline at the turn of the 20th century in imperial Germany, and continues to be referred to also in countries that implicitly or explicitly follow the German model. In order to deine ‘African linguistics’, precision is required, irst, for members of academic communities like in, for instance, Africa and North America, in which African linguistics is not established as an independent scientiic discipline. Second, in European and particularly in German-speaking academia, there is considerable confusion whether Afrikanistik today best translates into English as ‘African linguistics’ or as ‘African studies’ (see below).1 Irrespective of certain irritations here and there, however, there exist clear conceptions of what African linguistics is in terms of a largely autonomous ield of scientiic research and academic teaching and learning as it is relected, for instance and most recently, in the scope of the World Congress of African Linguistics (WOCAL) series since 1994. The present introduction deines and delimitates African linguistics in this 1

The term Afrikanistik was originally coined in analogy to pre-existing Orientalistik, that is, the study of ‘Oriental’ languages, literatures, cultures, and history, which were accessible mainly through philological approaches to written documents. For African languages, however, written texts were almost ubiquitously absent. At the time, there were hardly any existent African literatures in the narrow sense to be studied. As for African cultures, their study came under the research focus of pre-existing Völkerkunde (ethnography). In addition, Africa had long since been considered a ‘continent without history’, not the least because of the absence of ‘historical documents’, which, if written in any African languages, would be accessible to philological and historical research. Quite naturally, therefore, the unique research focus remaining for Afrikanistik was the study of spoken African languages as embedded in cultures and societies that, in turn, relied predominantly on oral transmission for any content that was considered culturally and historically significant for European scholarship. Hence the synonymous usage of the labels Afrikanistik, Afrikalinguistik, and Afrikanische Sprachwissenschaft since the field’s beginnings in imperial Germany some 130 years ago.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

4

INTRODUCTION

sense and thereby justiies the compilation of The Cambridge Handbook of African Linguistics.

1.1.1

Ambiguity of the Term

Most readers are likely to agree on what ‘linguistics’ refers to, namely the scientiic study of human language per  se and of all human languages, whether living or extinct, in terms of formal features and conveyance of meanings, and also regarding usage in given contexts. It is less clear what ‘African’ refers to in the collocation of the terms ‘African’ and ‘linguistics’. ‘African’ can refer to at least three different qualities. A irst reading (a) would assume some kind of ownership, that is, linguistics as studied and conducted by Africans. A second reading (b) would refer to location of activities, that is, linguistics as studied and conducted in Africa. A third reading (c) could refer to the quality of the object of study in terms of historical origin and geographic distribution, that is, linguistics studying languages of African origin and/or languages most widely, if not exclusively, distributed in Africa. Reading (a) would have ethnic/nationalist if not racial implications; for instance, German linguistics could indeed mean linguistics as studied by Germans. As a matter of fact, national cultural history may lead to the development of particular ways of doing science that one might wish to refer to as national, in this case German, ‘school(s) of thought’.2 In this sense, African linguistics would link up with ‘black linguistics’ dealing with ‘black languages’ (cf. Makoni et  al. 2003). This, however, does not mean that African (or ‘black’) languages can be studied only by Africans (or ‘black’ people), like there is no scientiic reason to disallow non-Germans to study Germanistik, that is, German language, literature, and culture. In the 21st-century global culture of science, exclusive ownership claims of this kind would be considered unethical. ‘African linguistics’, therefore, is hardly ever construed in this vein. With reading (b), African linguistics would refer to territorial aspects in terms of location of the researchers, that is, linguistics as operating in African institutions of higher learning and research, whether by African or non-African researchers. In the same way, German linguistics could indeed mean the study of languages at academic institutions in Germany independent of the nationality or geographic origin of the researchers. This again would tie up with reading (a) in so far as it would attempt to account for the existence of idiosyncratic regional geographic cum cultural and historical variants of scientiic development in Africa, as opposed to the rest of the world. This reading is not too far off reality when one considers the speciic needs and conditions 2

In fact, Afrikanistik, that is, the particular German-origin approach to the study of African languages, represents such a ‘national’ school of thought in which, however, German-speaking academics from beyond the national borders of Germany, such as those from, for instance, Austria (Vienna), the Soviet Union (Leningrad), the United Kingdom (London), and South Africa (Pretoria), also were involved from the early times.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

5

of academic work on languages in Africa, as alluded to by Ngessimo Mathe Mutaka, for instance, in Chapter  3. However, both (a) and (b), though semantically plausible, do not happen to have any wider currency in the global scientiic community; this remains the privilege of reading (c). In its most established and widespread usage, ‘African linguistics’ is simply short for ‘linguistics of African languages’. This means that African linguistics refers to a well-deined object of research, of which there are again two readings. The more narrowly conceived reading of the speciic object of study would be ‘African languages’, individually or in scientiically relevant groupings, such as language families, language types, or members of areal convergence zones (sprachbund), and so on. Alternatively, and in a wider perspective, the reading of the speciic object of study would be ‘language and languages in Africa’, that is, the study of the role and functions of human language per se as much as of individual languages, African or non-African, in African cultures and societies. These readings hinge on what we deine as ‘African language’ on the one hand, and as ‘non-African language in Africa’ on the other (see below). Semantic irritation concerning the meaning of ‘African linguistics’ may further stem from comparing it with the labels ‘American linguistics’, ‘European linguistics’, and ‘Australian linguistics’ as in current usage among insiders and deined in popular encyclopaedias (such as Wikipedia). Most of the time, when we speak of African linguistics, we refer exclusively to the study of ‘the languages of Africa’ (which is another expression still to be deined). American linguistics, however, has two readings: it refers to the study of ‘the indigenous languages of the Americas’, but also to the history of linguistics particularly in the United States (cf. Tucker Childs in Chapter  6). Australian linguistics and European linguistics, on the other hand, appear to refer only to the study of linguistics on these continents and not to the linguistics of European or Australian Aboriginal languages.3 Sensitive critics may perceive persisting colonialist undertones in some of these usages and, for instance, sense some kind of belittling when representatives of the ‘North’ refer grosso modo to African affairs without apparently considering internal complexity, as they would do when speaking of Germanic, Romance, or Slavic linguistics rather than sweepingly referring to ‘European linguistics’. As a matter of fact, in current usage among experts, African linguistics follows the model of other complex linguistic ields like, for instance, Indo-European linguistics, which tends to be immediately subdivided into sub-ields like Balto-Slavic, Celtic, Germanic, Indo-Iranian linguistics, and so on. African linguistics, too, is commonly subdivided according to linguistic subgroupings on different levels of inclusion and partly specialized methodology, such as, for instance, Afroasiatic linguistics, Bantu linguistics, Chadic linguistics, and

3

The label ‘Asian linguistics’ does not appear to have any currency at all.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

6

INTRODUCTION

Khoisan linguistics, sometimes shortened to attributive expressions like Bantuistic, Chadistic, and Cushitistic, or allowing for specialized labels like Berberology and Egyptology.

1.1.2

Defining ‘African Language’

If ‘African linguistics’ refers to the study of African languages, how do we deine ‘African language’? This question already provides a broad entry into some of the major issues that African linguistics deals with, beyond – as some might presume – simply applying theoretical approaches and methodological tools from general linguistics to natural language data stemming from Africa. Deinitions would have to relect language-related problems deriving from the social, cultural, historical, political, and economic coexistence, if not rivalry, among ‘African’ and ‘non-African’ languages on the African continent, that is, the linguistic ecology of languages.4 The need for deinition opens up challenging questions like whether, for instance, Malagasy, Afrikaans, and Arabic are ‘African languages’, and whether one would want to include ‘Nigerian English’ and localized variants of French (like FPA: français populaire d’Abidjan or français populaire de l’Afrique) and Portuguese as spoken in Africa among ‘African languages’, that is, under the same label as we would subsume, for instance, Tamazight, Hausa, Kiswahili, and Khoekhoegowab. One may wonder what the scientiic or practical gain, or loss, would be of such inclusions or exclusions, in terms of relevant generalizations and signiicant increase of knowledge. In any case, deining ‘African language’ may evoke ideologies, stereotype cliché, and prejudice that inform language attitudes, which are held by political, economic, cultural, and educational stakeholders both within and outside Africa. Apart from ideology and having immediate repercussions on individual speakers, speaking an ‘African language’, in particular when accompanied by low competencies in a particular ‘non-African language’, may decide on speakers’ inclusion or exclusion from access to power and to national resources in postcolonial Africa. To the extent that language matters are of great ideological and political impact in the multilingual, multicultural, and multiethnic societies of postcolonial Africa, African linguistics functions as an applied science in the context of empowerment and disempowerment of languages and of the people who speak them. This entails, at times, fuzzy borders with language activism. What we deine as African language(s) has immediate repercussions on the triple deinition alluded to here of what African linguistics is about, namely (a) about ‘African languages’, (b) about ‘language in Africa’, and

4

Linguistic ecology, following its early conceptions as suggested by Voegelin et al. (1967) and in particular Haugen (1972), would appear to describe quite well the broad focus of African linguistics following the German Afrikanistik model, even though, to the best of my knowledge, this term has never been used in early German discourse.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

7

(c) about ‘languages in Africa’.5 In fact, there is no unanimous agreement on what to refer to as ‘African language’. Probably, most experts would claim that African languages, in a rather narrow sense of the term, are those that, to the best of our current knowledge, have originated – and are most widely, if not exclusively, spoken – in Africa. Under a slightly wider deinition, the languages of African slaves deported across the Atlantic and the Indian oceans, and whose languages, like in the Americas, have somehow survived the traumatic excision from the continental African context, fall under this deinition, like the pidgins and creole languages for which sources in African languages can be assumed. This deinition locates African linguistics indeed in the neighbourhood of ‘black linguistics’ by closely connecting languages with the ultimate origins of their speakers.6 This deinition would pose problems with Afrikaans, which most experts would classify as an Indo-European language (i.e., a variety of Nederlands/Dutch) even though its origins and its speakers are located almost exclusively in Southern Africa. It is often overlooked not only that Afrikaans is the mother tongue of a section of ‘whites’ in Southern Africa, but that the majority of it speakers were labelled ‘non-white’ (‘coloured’) in racist terminology under the apartheid regime in South Africa. Clearly, the notion of ‘black’ linguistics would face its limits. A similar problem arises with Malagasy, which experts classify as an Austronesian language, but which is widely if not exclusively spoken on Madagascar and adjacent islands that most people would consider to be part of Africa, and by people who today would be considered Africans. It would also pose a problem with Arabic, which was brought into Africa in historical times following the expansion of the Rashidun and Umayyad Caliphate in the 7th century CE and is now spoken in 15 or more African countries by people of very diverse origins and appearances. 5

Note that, for practical and systematic reasons, the notion of ‘African language’ shall be restricted to living or extinct spoken human languages, that is, to the exclusion of sign languages. This is based on the independence of sign languages from spoken languages. African sign languages are most profitably treated scientifically in the context of sign languages worldwide, even though they are also given room under the umbrella of the World Congress of African Linguistics.

6

This, however, raises the – in some quarters controversial – question whether reference should be only to subSaharan (‘black’) Africa or whether Northern and North-Eastern Africa should be included. The position taken in this handbook is that, from a linguistic point of view based on the distribution of language families across the African continent as a whole, there is no reason to separate sub-Saharan Africa from the rest of the continent. This position challenges the traditional practice to consider the languages of Northern and North-Eastern Africa, particularly (Ancient) Egyptian, Arabic, and the Semitic languages, and often Berber (Tamazight) languages as well, to be the domain of ‘Oriental studies’ rather than of African linguistics. Three historical facts have motivated traditional practice to include them in Oriental studies rather than African linguistics: First, as a result of geographic proximity to Europe, the study of (Ancient) Egyptian, Semitic, and Berber languages predates that of (sub-Saharan) African languages. Second, the formerly so-called Hamitic languages in Northern and North-Eastern Africa were considered to be ‘naturally’ within its domain because of their genetic affiliation with the Semitic languages (cf. Greenberg’s Afroasiatic), which lie at the core of ‘Oriental studies’. Third, the study of these languages rests largely on written documents, often from languages that are no longer spoken. Modern African linguistics has challenged the Orientalists’ monopoly, yet accepts overlapping research interests and the academic autonomy of both Egyptology and Semitic studies.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

8

INTRODUCTION

Probably the most widespread deinition of the term ‘African language’ among experts makes very convenient implicit reference to the seminal classiication of The Languages of Africa by Joseph H. Greenberg (1963), who had identiied four major language phyla in Africa: Niger-Congo (formerly also Niger-Kordofanian, Congo-Kordofanian), Afroasiatic, Nilo-Saharan, and Khoisan. The more recent research-based criticism of Greenberg’s classiication notwithstanding, one could consider, by way of convenience, all languages subsumable under any of the four Greenbergian language phyla to be ‘African languages’. This would immediately exclude Malagasy as an Austronesian language. Afrikaans either would be excluded as an IndoEuropean (Germanic) language or would be treated along with other creole languages that can be found on the African continent and adjacent islands, and which show non-unique genetic afiliation to both African and nonAfrican language families. Arabic, on the one hand, could be included as being an Afroasiatic language; it could also be excluded, on the other, for the reason that we know exactly when in historical times this language entered the African continent as corollary to the military and cultural expansion of Islam.7 Some linguistic experts would take issue with this somewhat narrow deinition of ‘African language’. They would claim that many languages of undisputed non-African origin, stemming from Europe or Asia, even if only of recent vintage in Africa, deserve to be included. They would argue that their inclusion would be justiied by their considerable political and cultural impact, despite the fact that there are hardly any relevant communities of native speakers on the continent for some of these languages (see Chapters 13 and 14).8 This would mean that not only English, French, Portuguese, and Spanish but also Dutch, German, Italian, and even a plethora of Indian languages like Marathi, Bhojpuri/Bihari, Awadhi, Rajasthani, Tamil, Gujarati, Hindustani, and Telugu should be considered ‘African languages’ by now. Clearly, there is little scientiic gain in such an over-inclusive

7

The question becomes more intricate in view of the fact that Arabic belongs to the Semitic language family within the Afroasiatic phylum, which as such may have originated ultimately in Africa. (This scenario assumes that speakers of pre-Semitic languages quite early migrated out of Africa, and much later re-migrated as speakers of Semitic languages into Ethiopia, forming the modern Ethiosemitic languages, or spread, as speakers of Arabic, across larger parts of Northern and Eastern Africa.) Since other authors, however, assume the origin (Urheimat) of Afroasiatic as a whole to have been outside Africa, the status of all Afroasiatic languages as African languages by origin could be considered unsettled. If, however, we strictly adhere to the Greenbergian system of reference, then all Afroasiatic languages should be considered ‘African languages’, including the Semitic languages like Arabic, Akkadian, Ugaritic, and others, which, however, would be counterintuitive. An intermediate position to take would be to consider as ‘African languages’ only those Afroasiatic languages that are/were spoken almost exclusively on the African continent. This would include the vast majority of them anyway, and also the so-called Ethiosemitic languages, but would exclude the Semitic languages that are/were exclusively spoken in the present countries of the Middle East including the Arabian Peninsula and adjacent islands. Still, Arabic would be part of both.

8

As a rule, European languages have only small, if any, native speaker communities in Africa. An exception to this rule would be Southern Africa with regard to English and Afrikaans mother tongue speakers and possibly larger groups of speakers of Indian languages. Mother tongue speakers of Arabic and Malagasy, however, have a prolonged presence in Africa and count in the millions.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

9

deinition of ‘African language’, so I have long since suggested working with the more useful distinction of ‘African languages’ in the narrow sense as deined above on the one hand, and with the complementary notion of ‘languages in Africa’ on the other. We can study the latter group of languages rewardingly in terms of diaspora linguistics, as we study African languages in the Caribbean and South and North American diaspora. The research focus would then be on the changes that non-African languages have undergone in Africa (such as speaking about Nigerian or East African Englishes, or earlier Cape Dutch changing into modern Afrikaans, etc.). Much of this discussion links up with the central question of whether at all, or in which way and to what extent, Africa as such represents a ‘linguistic area’ characterized by common, if not exclusive, linguistic features (see Chapter 8).

1.1.3

Giving ‘History’ to Africa

There is one important aspect of African linguistics that is often overlooked in academic and public discourse. At the time of the emergence of African linguistics in academic circles in Europe at the turn of the 20th century, Africa and her peoples were widely considered, in a characteristic Eurocentric perspective, to be without ‘history’. Such harsh judgement was based on the apparent lack of written records and documents from most parts of Africa, particularly in local or other languages, on which all historical science was supposed to be based. Part of the Eurocentric worldview at the time was that nothing in the humanities was really worth scientiic study unless it had (written) ‘history’. It was therefore nothing less than a revolutionary contribution to world historiography when Carl Meinhof (1857–1944), for instance, succeeded in applying ‘historical methods’ of language comparison and reconstruction to African languages. This ‘proved’ beyond doubt that African languages had history in the same way that Indo-European languages had history, that is, that ‘Proto-Bantu’ in Africa essentially compared to, for instance, ‘Proto-Germanic’ in Eurasia. By extension, the peoples who spoke such languages together with their cultures and societies also must have ‘history’. Scientiically, this put Africa on equal footing with the Old World – thanks to African linguistics research! Language history in the early days of African linguistics, like in historical Indo-European linguistics, which provided the model, was concerned almost exclusively with language ‘genealogy’. Proto- or ‘parent’ languages were seen to have branched into ‘daughter’ languages that later subbranched into ‘granddaughter’ languages and so on. The common model was the ‘family tree’, which provided apparently clear illustrations of language history in terms of families, sub-families, branches, sub-branches, and so forth. Alternative models, like the ‘wave model’, were less current but challenged the exclusivity if not validity of the family tree model

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

10

INTRODUCTION

from the beginning. The predominance of the genealogical approach is still witnessed today for Africa, namely in the impact of Greenberg’s classiication presented in The Languages of Africa as the pervasive system of reference. Historical African linguistics of the 21st century, however, has meanwhile turned its perspective towards ‘areal’ and ‘contact’ linguistics, that is, approaches more akin to the old wave model. Assumptions about language history in Africa that were long based, and almost exclusively so, on genealogical relationships thus come under closer scrutiny and criticism. This is now based on insights gained from discovering unexpected (and partly massive) reciprocal inluence of languages on each other that are not genealogically related but are spoken in the vicinity of each other, allowing for considerable degrees of individual or societal multilingualism over considerable periods of time. This makes historical African linguistics as burning and important a research issue now as some 130 years ago (see Chapters 7 and 8) when ‘history’ was discovered to pertain to African peoples and their languages too.

1.1.4

‘Critical’ African Linguistics

From its beginnings, African linguistics has carried a critical impetus geared towards pre-existing notions and preconceptions. It would take issue, more or less successfully, with Eurocentric approaches virulent at their times. Apart from the inherent deconstruction of fundamentally racist stereotype notions about (black) Africans representing ‘primitive’ peoples if not ‘savages’ or ‘beasts’, living in ‘pagan’ darkness, devoid of sophisticated ‘civilization’ and without ‘history’, speaking ‘tribal gibberish’, African linguistics gave not only history (see above) but also ‘proper languages’ to Africa. In the irst half of the 20th century, it boosted descriptive, irst of all pre- and early structuralist approaches to language, which increasingly became free of automatic adherence to grammatical models that were current in (Neogrammarian) Latin-based Indo-European historical linguistics, playing a role comparable to the study of American Indian languages in the Boas and Bloomield traditions in North America. This continued into more recent generativist and poststructuralist periods in the second half of the last century, when African linguistics, on both sides of the Atlantic Ocean, challenged mainstream ‘universalist’ linguistic approaches that were commonly ridiculed elsewhere as ‘let’s take any language, say English’. African linguistics, as much as it was proiting from advances in general and theoretical linguistics, always traded in its share towards reshaping and developing theories and methods to the beneit of general or theoretical linguistics (see further below). Far from rejecting traditional comparative linguistic methodology, current African linguistics is the driving engine of a revived interest in issues concerning geographic linguistics and language contact scenarios. These are likely to

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

11

shake some of our more traditional assumptions regarding the interface of genetic and typological linguistics that allowed us to speak of almost monolithic ‘languages’ that, from case to case, might have been ‘contaminated’ by interference from neighbouring languages, but otherwise remained ‘pure’ and were handed down, as ‘ancestral code’, to the following generations. More recent language typology approaches, not the least from African hotbeds of linguistic contact, stimulate revisions of established concepts, which, as empirical research in Africa suggests, do not match the sociolinguistic realities on the ground. In more recent sociolinguistic theoretical discourse regarding language ‘endangerment’, ‘attrition’, and ‘loss’, new data from Africa pertaining to language loyalty and language shift, language maintenance, and language revitalization gain challenging importance in view of still poorly researched manifestations of actual multilingual practice, negotiating linguistic and other identities. These new insights from Africa tend to shake apparently well-founded received notions of ‘(heritage) language’ (and ‘dialect’), ‘speech community’, ‘ethnolinguistic group’, and others that were originally developed in and for American and Australian scenarios. Most recently, African sociolinguistic research has become a testing ground for ‘new’ concepts labelled ‘trans-’ or ‘polylanguaging’, making references to linguistic ‘superdiversity’ and the existence of ‘supervernaculars’, and even suggesting doing away with received notions such as ‘named language’ (see Chapter 15).

1.2 1.2.1

African Linguistics as an Autonomous Academic Discipline Criteria

Accepting African linguistics as an autonomous academic discipline in the overall organization of sciences, rather than considering it simply a sub-ield of general linguistics or of African studies (see further below), requires explanation. It is justiied according to at least the following ive criteria: 1. Possession of a well-deined object of research that as such is not in the focus of any other established discipline of science 2. Relevance of the particular object of study, both qualitatively and quantitatively 3. Institutional history, that is, the establishment of speciically designated professorial positions and chairs, departments, or institutes (‘seminars’) at institutions of research and higher education 4. Acquisition of global recognition beyond the place and country where it was irst established 5. Possession of a speciic methodology

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

12

INTRODUCTION

In regard to (1), in the previous section we deined the object of African linguistics uniquely and in a threefold manner as ‘African languages’, ‘languages in Africa’, and ‘language in Africa’, albeit largely to the exclusion of sign languages. Focus on African languages as such, individually or as an ensemble of languages, does not form a core issue in any other science, not even in modern general or theoretical linguistics. In the colonial period, when African linguistics was established in Germany before modern linguistics was even invented, the languages of the world were typically and systematically studied as large ensembles according to either their genetic afiliation (e.g., Indo-European linguistics) or their geographic distribution (e.g., Oriental languages, African languages). Hence Afrikanistik was established as a younger sister discipline to Orientalistik, illing an obvious gap in the catalogue of specialized sciences. Relating to (2), Africa is the cradle of humanity and the place of origin of human language, which should give ‘residual’ African languages high research priority in the humanities. Today, Africa is the home of almost one-third of the world’s living languages, 2,139 of 7,097 according to the Ethnologue (Lewis et  al. 2016). The number of languages alone provides justiication for establishing a specialized academic discipline that would focus on the plethora of languages on the oldest continent inhabited by humans. Pertaining to (3), if we take the beginning of regular teaching of African languages at universities to be the starting point African linguistics has a record as a separate academic discipline of more than 130 years in Germany. Regular teaching of African languages began as early as 1885 in Berlin and 1895 in Leipzig. This coincided with the peak of European colonialism. African linguistics as an autonomous academic discipline started as a ‘colonial science’, at least in early 20th-century Germany, where the irst senior academic positions – ‘extraordinary’ or ‘ordinary professors’/ chairs – were created before World War I to be illed by leading igures in the ield. Hans Stumme (1864–1936) was appointed Professor of Hamitic and African Languages in Leipzig in 1900. Carl Meinhof (1857–1944) was named Professor of African Languages in Berlin in 1905, and beginning in 1909 held the worldwide irst chair dedicated to the study of African languages, in Hamburg. Diedrich Westermann (1875–1956) took over as successor to Meinhof in Berlin in 1909. After World War I, academic positions were also created in other European countries and other parts of the world, even though this in some places would not happen until after World War II or even after African independence from European colonialism around 1960 (see Chapter 6).9 Such senior positions formally created 9

African independence from European colonialism coincided with or entailed other ruptures relevant to African linguistics. In Western Europe, a paradigmatic shift occurred from colonial to non- if not anti-colonial ideology that was part and consequence of the worldwide anti-imperialist, anti-colonialist, and anti-establishment student revolt in 1968–1969. In the former socialist countries of Eastern Europe and the Soviet Union, traditional philological approaches based almost exclusively on written texts now met with the presence of students from Africa, who became available as native speaker informants of African languages. Parallel developments in the United States

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

13

the foundations for the establishment of a new academic discipline with independent ‘titular’ university departments and institutes (or ‘seminars’). Academically, this could be justiied only following pre-existing research traditions into African languages with a fair number of proliic scholars, sometimes in the service of Christian missions or colonial authorities, paving the ground with their publications and at least occasional teaching (see Chapter 2). Eventually African linguistics also found homes on African soil (see Chapters 3 to 5). Clearly, the primary motivations for allowing this new discipline to establish itself were political and economic, targeting the unquestioned ultimate goals of securing colonial rule and assisting exploitation in terms of material and human resources by former colonial powers. In the German-speaking context and implicitly based on Humboldtian ethnolinguistic language philosophy, knowledge of African languages was considered the key to a better ‘understanding’ of African peoples and their cultures, not the least to ensure smooth administration of the colonial territories. Further, Europe-based Christian churches and missionary societies were keen to convert the ‘pagans’ in the overseas territories and, in the Lutheran-Protestant tradition, do so through the vernaculars of the new proselytes. Comparable interests persist today, given the activities of Bible translators and NGOs like SIL International (formerly Summer Institute of Linguistics), which happen to be, in many parts of the world, the only more or less professional linguistic experts working ‘in the ield’. The missionary line of tradition concerning alphabetization, literacy, and post-literacy work under the umbrella of applied African linguistics can thus be considered unbroken. Exoglossic monolingual national language policies favouring the ex-colonial language, supported by membership in postcolonial agencies of linguistic and cultural imperialism, such as the Commonwealth of Nations, the Organisation Internationale de la Francophonie, and the Communidade dos Países de Língua Portuguesa, manifest persisting linguistic and cultural hegemonic dominance. In economic terms, centre-periphery relations with the former colonial master persist, for instance in the publishing business for textbooks and literatures in European languages to the beneit of non-African publishing houses. This happens at the expense of viable publishing in African languages locally in Africa and constitutes a second chain of unbroken ‘extractive’ tradition. A third unbroken chain of tradition is the continuous ebb and low of young academics and ield workers, again largely from non-African places of origin and/or academic afiliation, in search of linguistic and/or cultural data for their degree projects. To the extent that such activities have no return or beneit for the speakers of the African languages under research, such as providing or assisting with adequate were the emerging interest in ‘black studies’ in the context of the civil rights movement, political concern with postcolonial Africa under the bipolar geopolitics of the ‘Cold War’, and the dominant impact of structuralist and generativist schools of thought in North America on global linguistics, including African linguistics in Europe.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

14

INTRODUCTION

orthographies, dictionaries, grammatical descriptions, and relevant postliteracy materials, this may as well be considered part of the tradition of former colonial extraction of (immaterial) resources from Africa. In some quarters in Africa, this may and occasionally does create intellectual crises of ‘ownership’, when young African academics complain about expatriate researchers ‘stealing’ their languages. A relatively more recent line of motivation for studying the linguistics of African languages, often initiated from outside Africa (and to no minor extent motivated by vague feelings of guilt), stems from concerns about critically endangered biological species as much as likewise endangered languages and cultures, and their timely ‘documentation’ before extinction. Its roots lie in ‘Northern’ encyclopaedic scholarly traditions of collecting, describing, and taxonomically classifying, and possibly preserving, ‘exotic’, rare, or simply unknown species. This had been a concern already in early colonial if not precolonial periods, which can be associated with outstanding scholars and activists of great impact like Wilhelm (1767– 1835) and Alexander (1769–1859) von Humboldt, for instance. On the other hand, while exploiting traditional ‘ethnoknowledge’ about healing effects of plants has high commercial value for the modern pharmaceutical industry, knowledge of the local languages, which actually encode the local knowledge, is largely considered irrelevant, at least outside linguistic expert circles. In regard to (4), African linguistics as invented by German-speaking academics under the regime of imperial Germany’s colonial aspirations has long since left its erstwhile regional and political conines. It has been globalized under a new postcolonial paradigm and has established itself as a legitimate ield of study outside the European countries that look back on a colonialist past (see Chapter 2). African languages are studied not only in European countries without any colonial tradition in Africa, but also in the Americas, Asia, and Australia (see Chapter 6). Major African languages in the catchment areas of universities have become objects of study also in Africa, where African linguistics, usually as part of general linguistics or located alongside departments of modern languages, often mirrors the sustained impact of the former colonial masters (see Chapters 3 to 5). In some parts of the world, African linguistics has its own titular departments or named sections within larger institutes. In other parts, it may be hosted by institutions dedicated to linguistics in general or inds itself within conglomerates like foreign studies, world cultures, African and Asian studies, and so forth. Interestingly and characteristically, African languages are hardly ever grouped together with ‘modern languages’ in African institutions of higher education, as that particular label tends to be reserved for the (often ex-colonial) languages of mainly European provenance.10 10

An interesting case with regard to such language classification systematics is provided by South Africa’s Rhodes University, whose School of Languages encompasses sections for African languages (especially isiXhosa), Afrikaans/Nederlands, French, and German, but not English, which is blessed by an independent department.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

15

Since 1994, globalized African linguistics has been blessed with its own triennial World Congress of African Linguistics (WOCAL) series. Worldwide academic institutions in support of African linguistics host the conference, held alternatingly in Africa, Europe, America, and Asia, returning to Africa on every other occasion. In its irst 24 years of existence, participants of ten WOCAL events have met in the various regions of Africa (South: Swaziland 1994, West: Togo 2000, East: Ethiopia 2006, Central: Cameroon 2012, North: Morocco 2018), Europe (Germany: Leipzig 1997, Cologne 2009), North America (United States 2003), South America (‘Special WOCAL’ Brazil 2008), and Asia (Japan 2015). In this sense too, one may consider African linguistics to have come of age and to have established itself globally as an autonomous academic ield. Finally, and relating to (5), African linguistics characteristically possesses a speciic methodology, which stems from its inherently anthropocentric focus, its interdisciplinary approach to issues that are simultaneously also of interest to linguistics, literary science and philology, sociology, political science, economics, pedagogy, and history, and its inherent nature as an applied science (see further below for a more detailed account of methodology). Only on a shallow level, therefore, might African linguistics look like simply a regional sub-ield of general linguistics that applies tools of the latter to data from African languages.

1.2.2

Languages and Dialects

Apart from current critical approaches from within modern sociolinguistics to the received Eurocentric and ideology-laden somewhat ‘monolithic’ concept of ‘(named) language’ (see Chapter  15), there is considerable irritation in public as well as academic discourse on what constitutes a ‘language’ (often insinuating the notion of ‘proper language’) versus a ‘vernacular’, ‘dialect’, ‘patois’, ‘language variety’, ‘lect’, and so on, by terms that may carry, at times quite obvious, pejorative connotations. With regard to Africa, this irritation accounts for another uncertainty in underinformed quarters: Are we dealing, in the African context and based on ‘Western’ ideological positions and attitudes, with ‘proper languages’ or just with ‘tribal dialects’ (cf. Wolff 2016a)? This leads to much confusion about igures concerning the exact number of African languages, which hinges exactly on what to count as an African language and what would be subordinate language varieties (vulgo: dialects), that is, how to distinguish, classify, and label speech forms associated with higher and lower ranks on a prestige scale.11 The irritations stem from the fact that there are Here as elsewhere, particularly in the South African context, a negative classification of languages prevails: LOTE (languages other than English). 11

Minimal numbers of African languages are often given in the neighbourhood of 1,250, while maximal numbers quoted are in the range of 3,000. This uncertainty is not a problem of African linguistics in particular, but more generally relates to the definition of what to count as one ‘language’. Whereas common usage would refer to,

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

16

INTRODUCTION

complementary deinitions for what constitutes a ‘language’ as opposed to, for instance, a ‘dialect’ or, more vaguely, a language ‘variety’. In systematic terms, linguists distinguish linguistic, sociological, and political deinitions of language. Linguistic deinitions of language are based on the criterion of mutual intelligibility and are of two kinds, one based on historical linguistic and the other on sociolinguistic notions.12 In historical linguistic terms, the family tree model of quasi ‘natural’ diachronic language change over generations of speakers would foster increase of divergence among ‘dialects’ (in terms of pre-existing mutually intelligible varieties) of ‘parent languages’ to the point of being no longer mutually intelligible. In this case, one would speak of the emergence of ‘daughter languages’ – former dialects of a parent language would become daughter languages in their own right. Such inherent language change towards increased divergence is accelerated by historical events like migrations of speaker communities including contact with other languages, or facts like relative isolation from ‘sister’ dialects or languages. In sociolinguistic terms, any linguistic utterance carries markers of idiolectal, sociolectal, and/or dialectal variation that would reveal information about the speaker. This could refer to gender, age, education, status, and others in terms of the speaker’s individuality (hence we speak of idiolectal variation), features of group membership, and/or sociocultural speciication (hence we speak of sociolectal variation), or could refer to place of origin or residence in the sense of ‘local’ or ‘regional’ accent (hence we speak of dialectal or chthonolectal variation). In this sense, ‘dialect’ is a regional or local variant of a given ‘language’. Both linguistic deinitions of language versus dialect imply a hierarchical order in which ‘language’ is of a higher order than ‘dialect’. Problems arise in cases of a so-called dialect continuum in which neighbouring dialects are mutually intelligible, but with differences accumulating over distance, more distantly separated dialects are no longer mutually intelligible. Geographically distant dialects would then have to be considered different ‘languages’ without, however, being able to establish where a ‘language boundary’ would have been crossed in what actually is a chain of adjacent mutually intelligible dialects. for instance, German as one language, closer scrutiny reveals that this may be true just for ‘Standard German’, while the 16 or more so-called German dialects, by the criterion of mutual intelligibility (see below), should each be considered separate even though closely related languages (cf. Lewis et al. 2016); for such fairly common linguistic situations, the term ‘macro-language’ has been suggested. In any case, counting ‘languages’ remains controversial. Note that any such discussion of numbers rests on a highly controversial notion of language as a bounded distinct entity, a notion that stems from ‘Northern’ experience and ideologies. It belies recent sociolinguistic insights into daily ‘multilingual’ performance in Africa as seen from ‘within’, where speakers make use of several more or less elaborate ‘linguistic resources’ at their disposal, rather than accessing discrete complete languages (‘codes’), among which they happily ‘switch’ (hence the popular term ‘codeswitching’). This is currently discussed under terms like ‘trans -’ or ‘polylanguaging’. 12

‘Mutual intelligibility’ describes the observation that speakers of different varieties of languages are able to understand and be understood by speakers of other varieties; in such cases we speak of the same language having different idiolects, dialects, and sociolects (or simply ‘lects’).

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

17

The sociological deinition of language is based on the criterion of language status within a society or polity. Reference is usually to standard languages that have elevated (high) formal status as the ‘oficial’ or ‘national’ language of a country. Such language tends to be fully standardized and to have a long tradition of writing and literature. It stands opposed to non-standard languages, that is, ‘vernacular languages’ (often also called ‘dialects’ or ‘patois’) and ‘sub-standard languages’ that are then associated with low status for being neither standardized nor written, and therefore with no or hardly any tradition of literacy. Again, there is a hierarchical order implied, in which ‘languages’ (i.e., standard languages) rank higher than ‘dialects’ (i.e., non-standard languages, vernaculars), in that they are considered more prestigious and powerful. This type of hierarchy accounts for socioculturally relevant patterns of ‘diglossia’ and ‘polyglossia’ in Africa (see below). The political deinition of language is based on the criterion of international borders. Boundaries may run right through territories inhabited by speakers of the same language – so-called cross-border languages – possibly dividing dialects that remain mutually intelligible. Still, speakers on both sides of the border claim that they speak different ‘languages’, which they associate with the different national territories and polities. Hence the popular tongue-in-cheek deinition of language as a ‘dialect with an army and a navy’.

1.2.3

African Languages: The Greenbergian System of Reference

Despite its appearance more than half a century ago and the substantial modiications that have been suggested since then, the classiication proposed by Joseph H. Greenberg in The Languages of Africa (1963) still serves as a system of reference in terms of postulated  – partly proven, partly disproven, and partly still hypothetical or contested  – language families or phyla. One reason for its persistence is that the system is very simple, setting up only four major genetic units (major families or phyla) into which all known African languages should be able to be classiied as stemming from one of the four proto-languages: Niger-Congo, Afroasiatic, Nilo-Saharan, and Khoisan. Subsequent African linguistics research has questioned a number of Greenberg’s units, reassigning individual languages or language groups in terms of inclusion or exclusion from the four Greenbergian phyla, even allowing for a fair number of ‘isolates’, languages or small language groups that are not assumed to be part of any of Greenberg’s genetic units. The other reason for the persistence of the Greenbergian system of reference is that, despite the modiications that have been suggested, there is as yet no comprehensive re-classiication of African languages that would be generally accepted by the scientiic community to replace it, and if only for practical reasons of convenience.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

18

INTRODUCTION

Greenberg’s classiication was based on the assumption of genetic relationship among the languages within each phylum. Members of same families and phyla would be assumed to have ultimately sprung from the same proto-language (which eventually would and could be duly ‘reconstructed’, mainly by applying the ‘comparative method’ to language pairs). Greenberg himself did not make use of the ‘comparative method’, which was originally developed in early Indo-Germanic linguistics in order to provide ‘proof’ of genetic relationship. He used a different approach known as ‘mass comparison’ (more recently referred to as ‘multilateral comparison’). For methods and problems regarding language classiication in Africa, see Chapters 7 and 8. Language history and classiication have been in the focus of African linguistics since its beginning, leading to constant revisions of our assumptions about, if not ‘proof’ of, postulated language families and phyla, their sub-families or branches, sub-branches, and language groups. The Greenbergian classiication of 1963 still serves as the reference system, the fact notwithstanding that it has undergone several revisions that would even question the overall validity of one of his phyla, namely Khoisan. Today, Greenberg’s Khoisan is viewed as representing not a valid genetic unit, but rather at least three separate units and some likely language isolates (Chapter 12). Further, the validity of his genetic unit Nilo-Saharan is questioned as far as the inclusion of all of his suggested sub-units is concerned, now also allowing for possible isolates (Chapter 11). Much less doubtful would appear to be Greenberg’s Niger-Congo and Afroasiatic phyla as valid genetic units, yet with doubts voiced as to the inclusion of the Omotic language family (which Greenberg treated as part of Cushitic) within Afroasiatic (Chapter 10) and Mande and other smaller units within Niger-Congo (Chapter 9). A number of African languages are spoken in regions that, for many decades, had been inaccessible to researchers due to security constraints. Only in recent years have more and better data from and analytical insights into such languages prompted many of the revisions of the 50-year-old Greenbergian classiication. In addition, a renewed interest in ‘areal’ approaches to language comparison attempts to identify convergence processes in zones of massive language contact. This also has led to re-tabling questions concerning the validity of some of Greenberg’s genetic units. There are long lines of historical continuity that would relate insights and oversights of Diedrich Westermann (1911, 1927, 1935a) concerning, for instance, purportedly genetic linguistic features of the erstwhile ‘Sudanic’ languages, with the fairly recent hypothesis of a large linguistic contact zone now called ‘Macro-Sudan belt’ across much of the Sahel Zone south of the Sahara Desert (Güldemann 2008). This has more recently placed linguistic geography (cf. Heine & Nurse 2008) at the core of African linguistics, following leads by Greenberg (1959, 1983), Larochette (1959), Ferguson

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

19

(1970, 1976), Meeussen (1975), and Heine (1975, 1976) towards identifying linguistic areas (contact zones, sprachbund areas) in Africa. Information on Greenberg’s four language phyla and the major issues currently under debate concerning the genetic status of languages with regard to his postulated genetic units is summarized as follows. NIGER-CONGO, with uncontested member families of the so-called Atlantic, North-Volta-Congo (Kru, Gur), and South-Volta-Congo (Kwa, Benue-Congo) languages. Doubts are raised against the inclusion and position of Dogon, Ijoid, and the Mande, Kordofanian, and Ubangi languages. In terms of number of languages, Greenberg’s Niger-Congo is the largest phylum in Africa and possibly also worldwide, with about 1,500 individual languages. Its languages are spoken south of the Sahara from the Atlantic Ocean in the west across the continent to the Indian Ocean in the east and as far south as the Cape region in South Africa, thereby covering more than half of the African continent. Its name derives from two of Africa’s large river basins: the Niger basin in the west and the Congo basin in the south. The remarkable geographic spread is partly attributed to the so-called Bantu expansion, the large-scale migration of speakers of early Bantu languages from inland areas east of the Bight of Benin (in today’s Nigeria and Cameroon) into Central, Eastern, and Southern Africa over the last 2,000 to 3,000 years. AFROASIATIC, with its uncontested member families or languages Ancient Egyptian (extinct), Berber/Tamazight, Chadic, Cushitic, and Semitic; contested remains the inclusion of the Omotic languages (classiied as ‘West Cushitic’ by Greenberg). Doubts have also been raised as to whether Beja should be treated, following Greenberg, as part of Cushitic, or rather as a separate unit within Afroasiatic. In addition, the position and genetic status of Ongota is questioned; also, the genetic status and position of Meroitic (extinct) with regard to Afroasiatic remains unclear. The Afroasiatic phylum follows Niger-Congo in terms of number of languages, counting some 300 languages that are spoken over most parts of Northern and NorthEastern Africa, but also in adjacent regions of the Near and Middle East – hence the designation Afroasiatic to mean ‘spoken in both Africa and Asia’. Languages of this phylum pride themselves on some of the longest traditions of writing in human history, dating back some 5,000 years, whether hieroglyphic (Ancient Egyptian in Africa) or cuneiform (Semitic languages in Mesopotamia). Worth mentioning are also the Berber abjad script known as Tifinagh (with likely sources in Libyco-Berber and Punic scripts from about 3rd century BCE), which is still practised among the Tuareg and is also used – in modernized varieties – in Moroccan schools, and the Ethiopic abugida script (irst used for Geez from the irst centuries CE, now in use for all Ethiosemitic languages). There is still no generally accepted position among specialists as to where the origin (Urheimat) of the Afroasiatic languages should be assumed to have been, whether in Africa (Eastern Sahara, west of the Nile Valley) or in the Near East (Levant). Hence there

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

20

INTRODUCTION

is no unanimous agreement on assumable major scenarios of migration. Either pre-Semitic speaking populations moved out of Africa, or the predecessors of all Afroasiatic families originated outside and later migrated into Africa including part of Semitic (Ethiosemitic languages, later Arabic), but with a fair number of Semitic languages staying behind and spreading over most parts of the Arabian Peninsula and adjacent islands. NILO-SAHARAN with about 200 languages remains disputed in terms of its overall validity as a genetic unit and for including a number of likely ‘isolates’ and the Songhay languages, the latter having also been described as likely creoles in works by Robert Nicolaï and others. More or less accepted language families or groups within Nilo-Saharan are Berta, Central Sudanic (including Sara-Bagirmi, Bongo, Modo-Baka, Moru-Madi, Mangbutu-Efe, Kresh-Aja, Lendu), Eastern Sudanic (including Nubian, Nilotic), Fur, Maba, and Saharan languages. Less widely accepted potential members are Songhay as the most western potential outlier along the Niger River, and Koman, Gumuz, Kadu, as well as Kuliak and Kunama languages along the eastern periphery of the area covered by Greenberg’s Nilo-Saharan languages. Greenberg’s KHOISAN phylum has been factually dissolved as a valid genetic unit of still about 25 living languages mainly in Southern but also in East Africa. Currently, Hadza and Sandawe (Greenberg’s ‘East African Khoisan’) are each considered ‘isolates’, while Khoe, Tuu, and Kx’a (Greenberg’s ‘South African Khoisan’) are considered separate and unrelated families. Kwadi in Angola, whose inclusion among the Khoe languages remains disputed, is assumed to be extinct. In addition to Greenberg’s more or less established three major African language phyla Niger-Congo, Afroasiatic, and Nilo-Saharan, there may be more potentially separate language phyla or ‘isolates’ whose independent status as genetic units may still await inal proof. Among those suggested are ex-Niger-Congo groups like Kordofanian, Mande, Dogon, Ijoid, and Ubangi; ex-Cushitic/ex-Afroasiatic Omotic; ex-Nilo-Saharan groups like Songhay, Koman, Gumuz, Kadu, Kuliak, and Kunama; and the ex-Khoisan groups Khoe, Kx’a, Tuu, Sandawe, and Hadza. A number of languages still lack adequate descriptions in order to develop genetic hypotheses for them or verify their position as ‘isolates’: Meroitic (extinct) in Sudan, Ongota, Gomba, and Shabo in Ethiopia, Bayot in Senegal, Bangi-me in Mali, Mbre in Côte d’Ivoire, Mpre in Ghana, Jalaa in Nigeria, Laal in Chad, and Kwadi (extinct?) in Angola. Uncertain genetic status is further attributed to Aasax in Tanzania, Imeraguen in Mauritania, Kara (Fer?) in Central African Republic, Oblo in Cameroon, Defaka in Nigeria, and Dompo in Ghana. Summarizing the state of the art: Whereas Greenberg’s four language phyla may continue to provide a convenient system of reference, they do not always represent valid genetic units of African languages. Unless genetic relationship can be ‘proven’ by the application of unequivocally accepted methods of language comparison, we may have to envisage the existence of up to 20 or more such units (including ‘isolates’) that are,

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

21

to the best of current knowledge, not genetically related to each other. Whether, in the end, some or all of them, or possibly all human languages, stem from ultimately one source (monogenesis hypothesis; ‘Proto-Human’ would have been an African language, no doubt) lies beyond the reaches of any hitherto known and generally accepted method of proof.

1.2.4

Languages in Africa: Beyond the Greenbergian System of Reference

A number of languages that are spoken in Africa by both large and small numbers of speakers do not meet the deinition of being an ‘African language’ within the Greenbergian system of reference. A irst group concerns Semitic languages, most of all Arabic and the so-called Ethiosemitic languages in Ethiopia. Technically and historically speaking, these languages have their origin outside Africa and thus may not qualify to be called ‘African languages’. To the best of our present knowledge, the predecessors of the about 20 modern Ethiosemitic languages migrated into Africa from South Arabia about 2,800 years ago. Arabic came with the Islamic expansion in the 7th century CE and is now spoken in several rather divergent regional varieties in at least 15 African countries, in a number of them as the oficial language. A second case is that of Malagasy, an Austronesian language that was brought to Madagascar, the Comoros, and Mayotte across the Indian Ocean during the irst millennium CE. It is the oficial language of Madagascar. A third group is that of Indo-European languages in Africa. Apart from early periods, in which Greek, Latin, Old Persian, and even Vandalic played a role in Egypt and North Africa, they were brought to African coasts and later into the interior by European missionaries and colonialists from the early 16th century CE onwards. Some of these languages had acquired oficial language status in the colonial territories; this was maintained even after independence, like in the cases of Dutch (later Afrikaans), English, French, Portuguese, and Spanish, less so Italian and German. As oficial languages, however, they are still spoken only by formally educated minority sections of the populations, and this mainly in urban agglomerations, in public administration, and in the formal sector of economy. They have rather limited value as lingua francas for inter-ethnic communication for the masses of the populations, who generally prefer African lingua francas. Some of these languages underwent ‘Africanization’ and developed various types of, for instance, African Englishes. Dutch (or rather the variety called Cape Dutch) in South Africa underwent creolization in contact with indigenous Khoe and other languages in ethnically and linguistically mixed families to give birth to Afrikaans. Afrikaans is usually still considered an Indo-European language, like its ancestral Dutch, even if heavily inluenced by African adstrata. Others consider it a creole that emerged in Africa and could therefore justiiably be called an ‘African language’.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

22

INTRODUCTION

A fourth set of imported languages is historically linked to more recent immigrant groups from Europe and Asia, mostly during the colonial times. Under British colonial rule and particularly in Southern and Eastern Africa, settlers (e.g., from Germany) and forced contract workers (mainly from India, but also from other countries in South-Eastern Asia) were shipped to Africa to assist the colonial exploitation of resources. These immigrants brought along their languages, which may persist in Africa today among minority communities of practice (see Chapter 14). A ifth group of languages is the so-called pidgins and the creole languages found mainly along the coasts of the Atlantic and Indian oceans and their islands, and occasionally inland. Some of them emerged in contact with European languages used for trade along shipping routes, like Krio, (Wes Cos) Pidgin, the Creoles of Cape Verde and Guinea-Bissau along the Atlantic coast, and Sechellois and Mauritian Creole in the Indian Ocean. Some emerged in contact with Arabic, like Juba-Arabic (South Sudan) and Nubi (Uganda), some by contact between African languages, like Sango (CAR) and possibly Songhay (West Africa: Niger valley), and Fanakalo in South Africa. Some experts also rank Afrikaans among the creole languages of Africa. More recently emerging ‘hybrid’ language varieties that are mainly used in the urban agglomerations by young males, some of which originally associated with secretive jargons among street gangs in the context of juvenile delinquency, could also be grouped here, like Camfranglais, Camtho, Franlof, Indoubil, Nouchi, Sepitori, Sheng, Tsotsi-/ Flaaitaal (see Chapter 18).

1.2.5

Language in Africa: Bi- and Multilingualism, Di- and Polyglossia, the Challenges of Language Planning, and the Management of Multilingualism

Arising from the plurality and diversity of languages and language varieties in Africa is the need to study not only culturally and socially motivated patterns of language use (see Chapters  16 to 19) but also the ideologies behind language attitudes to one’s own and other languages. In Africa, languages may either peacefully co-exist or enter into fatal rivalry, depending on the various domains of human communication. Against this backdrop, African linguistics studies the empirical foundations underlying language planning, language policy design and strategies regarding policy implementation, and the overall relevance of language to issues of ‘development’ (Chapter 13). These issues are central to the research agenda of the emerging sub-ield of Applied African Sociolinguistics (Wolff 2012, 2016a). Multilingualism is another core topic in African linguistic research. The African continent is characterized by various instantiations of multilingualism and could justiiably be referred to as the mother continent of multilingualism, apart from being the cradle of humankind. The term ‘multilingualism’ as used here refers to the coexistence and/or use of more than one language either on a given territory, in social and cultural Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

23

institutions, by communities of practice, or by individuals.13 Since the most common type of multilingualism is assumed to be bilingualism – the use of two languages – some authors make a distinction between bilingualism and multilingualism, the latter thus referring to the use of three or more languages. In the African context, it is important to distinguish between bilingualism in general terms and ‘oficial bilingualism’. The latter refers ‘to a language policy that restricts national communication to two oficial languages, such as French and English in Cameroon, or English and Afrikaans in pre-democratic South Africa. Under such oficial language policy, a person considered to be bilingual could be construed to mean “speaking the two oficial languages”, irrespective of any other languages this person may be able to speak luently’ (Wolff 2016a:320). Further, we need to distinguish diglossia from polyglossia (or extended diglossia). In fact, the terms ‘bilingualism’ and ‘diglossia’ both translate as ‘use of two languages’. Diglossia, however, refers to a ‘[s]ituation in which two languages, or two different language varieties, occupy markedly different positions on a status and prestige scale: one is considered to be of high(er) status and prestige, the other of low(er) status and prestige; such status differences inform language attitudes’ (Wolff 2016a:322). Polyglossia is a ‘[t]erm expanded from the earlier notion of diglossia to describe high vs. low status and prestige hierarchies . . . between languages in a multilingual context. In African post-colonies, the ex-colonial languages are associated with high status and prestige, while the indigenous African mother tongue languages tend to be associated with rather low status and prestige’ (Wolff 2016a:341). With regard to multilingualism, we distinguish at least four relevant levels in Africa: territorial, institutional, individual, and sociocultural multilingualism. • Territorial multilingualism refers to ‘[t]he existence of more than one language on a given territory (local, regional, national, cross-border international) either in relative mutual isolation of several monolingual pockets (multi-monolingualism) or as partly overlapping distribution of different languages across the territory’ (Wolff 2016a:338). • Institutional multilingualism refers to ‘[t]he use of more than one language, depending on language policy, in public or private institutions, such as government administration and communication (oficial language), education (Medium of Instruction), media, law’ (Wolff 2016a:338), giving room to ‘special purpose languages’ for religious practices, higher education, and so forth. 13

Note that our understanding of ‘multilingualism’ hinges on our preconception of language. Either – and in traditional terms – we think of it as a more or less monolithic ‘heritage language’, which would be describable in terms of clearly defined phonological, grammatical, and semantic categories and representations, or we consider language an obsolete and ideology-laden concept, and speak of rather ‘fluid’ patterns of usage. This would imply the notion of manifold access to several resources within an individual linguistic repertoire, which in more recent sociolinguistic research is described as ‘(trans)languaging’ (Chapter 15) and/or representing ‘superdiversity’, particularly in urban contexts, not the least supported by the use of digital and mobile media of communication (Chapters 18 and 19).

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

24

INTRODUCTION

• Individual multilingualism refers to ‘[a]n individual’s use of more than one language, acquired informally, simultaneously or successively as irst languages or second languages or, often through formal education, as foreign languages’ (Wolff 2016a:337). • Sociocultural multilingualism would be a term applicable to socioculturally accepted patterns of multilingual communication (also discussed in terms of stable or instable multilingualism) among communities of practice, in which monolingualism is not (or no longer) practised. In its applied sociolinguistic dimension, African linguistics addresses issues of ‘comprehensive’ language planning (a notion introduced by the Nigerian sociolinguist Ayo.[rinde] Bamgbos.e) and ‘multilingualism management’ (a notion that still awaits theoretical and methodological elaboration). This would be based on the ‘language as resource’ paradigm (Ruíz 1984) and acknowledge the fact that, in Africa and in a comprehensive way, language planning is social planning. Language planning in a narrower perspective refers to the ‘[d]omain of applied linguistics and sociolinguistics pertaining to research into and the provision of solutions to sociocultural problems related to language issues within societies and states, such as design and implementation of language policies . . ., multilingualism management, standardisation and harmonisation of language(s). The classic division of language planning comprises status planning, corpus planning and acquisition planning’ (Wolff 2016a:335). In this context, multilingualism management would somewhat vaguely refer to the yet ‘[u]nder-researched domain of language planning regarding the implementation of multilingual language policies for public or private institutions to the beneit of the speaker communities’ (Wolff 2016a:338). Multilingualism tends to entail polyglossia in the sense of associating different languages with different degrees of prestige and power in the polity. This, in a parallel fashion, allows for either rather free or restricted access to political power and to economic resources for the speakers, if not ‘owners’ of the languages. In this sense, all language planning in multilingual societies is likely to interfere with the balance of power among communities of practice; hence language planning amounts to social planning.

1.3 African Linguistics and African Studies In Anglo-American usage, ‘African studies’ serves as a cover term to include, in principle, all scientiic approaches to cultures and societies in Africa, including languages: African studies is the study of Africa, especially the continent’s cultures and societies (as opposed to its geology, geography, zoology, etc.). The ield includes the study of Africa’s history (pre-colonial, colonial, post-colonial), demography (ethnic groups), culture, politics, economy, languages, and

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

25

religion (Islam, Christianity, traditional religions). A specialist in African studies is often referred to as an ‘Africanist’. (Wikipedia)

In German-speaking academic circles, where African linguistics was irst established some 130 years ago, there tends to be confusion between African linguistics and Anglo-American African studies, which – by some – are both considered equivalent to German Afrikanistik. This confusion is sometimes carried over into international debate, not the least due to careless usage of terms among members of the trade themselves. When the study of African languages developed into an autonomous academic discipline at German-speaking universities at the turn of the 20th century, it became known under three synonymous German terms: Afrikanistik (coined in analogy to pre-existing Orientalistik), Afrikalinguistik, and Afrikanische Sprachwissenschaft.14 The founding fathers Carl Meinhof and Diedrich Westermann used these terms interchangeably. Clearly, Afrikanistik was concerned with the study of languages in Africa, irst of all. African cultures, on the other hand, were the domain of pre-existing Völkerkunde (ethnography), allowing for a special ield of Afrikanische Völkerkunde (ethnography of Africa). Spoken language was at the fore in African language studies, since – with only a few known exceptions – literatures written in African languages were practically non-existent, in particular philosophical, religious, and historical documents that philologists would study, comparable to pre-existing ‘Oriental studies’. Hence there developed a strong interest in the phonetics of African languages.15 To the extent that knowledge of African languages would allow access to cultural and historical information encoded in them, including specimens of oral poetry, tales, proverbs, and so on, this also became the object of study for any scholar of Afrikanistik. Westermann in particular saw the study of African languages (Afrikalinguistik) intimately intertwined with the study of African cultures (as object of Völkerkunde) and precolonial history, so he is often seen as the founder and spiritus rector of a unique German tradition of a comprehensive interdisciplinary approach to the history and culture of African peoples through their languages (sometimes referred to in German as Gesamtafrikanistik). His unconditional support of early German colonialism in Africa made him include aspects of applied linguistics as much as ethno- and sociolinguistic approaches (for a conceptual history of German Afrikanistik, see Wolff 1981c, 2013a, 2014a). The term Afrikanist (a parallel formation to Orientalist) gained currency also in English (Africanist) and French (africaniste). Taking Westermann’s 14

Note that German Sprachwissenschaft is synonymous with Linguistik, both meaning ‘scientific study of language(s)’, so both terms, Afrikalinguistik and Afrikanische Sprachwissenschaft, translate the same as ‘scientific study of African languages’.

15

Note that in 1909 Carl Meinhof, when he took over the world’s first Chair of African Languages at the Colonial Institute in Hamburg, a forerunner of Hamburg University (founded 1919), he immediately initiated the establishment of a phonetics laboratory under the roof of his department; its first director (1910–1949) was Giulio Panconcelli- Calzia (1878–1966).

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

26

INTRODUCTION

idealistic notion of comprehensive Afrikanistik even further, it soon became used to refer to any kind of scholar whose research had something to do with Africa – what today many would refer to as African studies (see the quote from Wikipedia above). Clearly, this extended notion of ‘Africanist’ is not co-referential with the much earlier notion of German Afrikanist as a specialist in the ield of African languages, including the study of texts told and relevant documents written in African languages.16 It is this wider notion of German Afrikanistik that, translated into English usage as ‘African linguistics’, has informed the comprehensive concept behind the WOCAL and addresses descriptive, typological, comparative, areal, sociocultural, political, economic, historical, and literary aspects of ‘African languages’ as much as of ‘language(s) in Africa’ (Wolff 2012). It is also this wider notion of both original German Afrikanistik and WOCAL that informs the structure and content of this handbook.

1.4 African Linguistics as a Sub-field of General Linguistics We deine African linguistics by both its object of study and the choice of its primary theoretical and methodological tools. Its object of study, language and languages in Africa, makes it a regional science with a geographic focus on the African continent. By its primary theoretical and methodological tools, which stem from modern theoretical and applied linguistics, African linguistics is a sub-ield of general linguistics. For ‘regional sciences’, again, it is quite common to operate with trans-, multi-, and interdisciplinary approaches to their objects of study. Accordingly, we should best look at African linguistics as both a comprehensive trans- and interdisciplinary ield within general linguistics and a regional science focused on the African continent. As a sub-ield of linguistics, African linguistics is both part of and complementary to modern linguistics. On the one hand, it makes use of theories and applies methods that are being developed in linguistics under a universal perspective, pertaining to all human languages on our planet. On the other, it feeds back insights gained from applying these theories and methods to empirical data from African languages. This creates important cross-fertilization between African linguistics and the more abstract ields of general or theoretical linguistics. In particular, the study of typologically diverse African tone systems, of languages with ATR vowel harmony, syllable weight, logophoricity, and marked nominative morphosyntactic alignment, have had important impacts on linguistic theory in

16

With African linguistics available as a specialized science, African studies is often conceived as not including the scientific study of African languages, that is, beyond occasional learning and teaching of some of the major African lingua francas, for instance, at African studies centres. This has prompted the mockery among experts in African linguistics that ‘African studies is anything that you can read about Africa in English’.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

27

general. There appears not to exist any single linguistic feature that was widespread in Africa and only here, but not beyond. Almost all typological features found in African languages can also be found in languages outside Africa, despite observable areal clustering of features in various parts of the world (Haspelmath et  al. 2005). However, linguistic ‘Africanisms’, features that show remarkably high if not almost exclusive distribution in African languages across genetic boundaries between language phyla (socalled areal or sprachbund features), do exist. This applies to, for instance, phonetics and phonology (Clements  & Rialland 2008), frequency of verbal derivative sufixes and nominal modiiers that follow the noun (cf. Heine & Leyew 2008), focus strategies, SOVX word order, clause-inal negative marking (Creissels et al. 2008), and characteristic polysemies, comparative constructions, and diminutive expressions (Heine & Leyew 2008); see Chapter 8. African linguistics’ ultimate goals are not to gain insights into human language capacity as such (according to the Saussurean notion of langage), nor the validity of linguistic universals, which remains in the domain of overarching theoretical linguistics. Rather, the ield aims at gaining deeper insights into global linguistic diversity, albeit with a restricted focus on Africa, by describing and comparing individual languages (de Saussure’s notion of langue) and by developing or sharpening tools for the description and comparison of individual languages; this is achieved by resting the abstract analyses on the empirical data provided by parole in Saussurean terms. This approach is often referred to as ‘descriptive/descriptivist’ or ‘typological’. Most African languages still lack comprehensive descriptions by modern linguistic standards. Some languages are furthermore considered substantially ‘endangered’ by attrition and potential loss through language death, so that sound linguistic descriptions are a desideratum of high priority for documentation purposes, both archival and as reliable bases for scientiic comparisons. Occasionally, speaker communities request support from experts in African linguistics to ‘rescue’ or even ‘revitalize’ the languages of their ancestors. African linguistics, in its narrowest sense, studies individual African languages and groups of languages in terms of their phonetics, phonological and grammatical (i.e., morphological and syntactic) systems, and lexicon. Characteristically, researchers in African linguistics study their objects in the wider context of the languages’ assumed genetic afiliation according to the Greenbergian system of reference, that is, as languages belonging to Niger-Congo, Afroasiatic, Nilo-Saharan, or Khoisan. Depending on their individual academic background, researchers in African linguistics could also choose a more hard-core linguistic approach to certain sub-ields, like phonetics, phonology, morphology, syntax, semantic, pragmatics, and others, and position themselves within theoretical linguistics rather than African linguistics. Or, they could choose to approach African language data from a universal typological perspective, which would also position

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

28

INTRODUCTION

them close to general linguistics and not necessarily exclusively to African linguistics in the narrower sense.17 Accordingly, the ‘classic’ linguistic domains of phonetics and phonology, morphology and syntax, and lexicon, but also semantics and pragmatics, do not constitute separate chapters in this handbook. They are dealt with embedded in wider linguistic, if not extra-linguistic, contexts, such as genetic afiliation of languages and linguistic typology, sociopolitical, pragmatic, and cultural use of languages, and applied dimensions of linguistics. Quite often, however, 21st century researchers in African linguistics overcome such more or less artiicial divisions associated with (post)structuralist linguistics. They tend to address African language data at the interface of general theoretical and potentially universal notions about human language (langage), how language is used by speakers in given situations (parole), and how variant usages relect and constitute, or question, what we consider to be individual languages (langue). African linguistics, in its wider sense, therefore, studies individual African languages and groups of languages in terms of the various interlocking research domains of socioand ethnolinguistics, pragmatics, and more recently emerging ‘cultural linguistics’. Contributions to these ields by African linguistics are illustrated in Part III. Following the more recent research paradigm of ‘language as resource’ (Ruíz 1984), African (socio-)linguistics studies, among other things, the linguistic dimension of all discourse on how to achieve ‘development goals’ in order to overcome ‘underdevelopment’ in Africa to the beneit of all African peoples. A central issue here is that of the role of languages, African and non-African, in education from kindergarten to university (Chapter 20). In this way, African linguistics is distinctly anthropocentric and emancipatory in its ‘applied’ dimension, part of which pertains to the notion of ‘intellectualization’ of languages based on studying concept formation and the practical use of lexicography (Chapter 21). This applied dimension, not unlike in the beginnings of African linguistics in colonial times, links it to issues also discussed in terms of language activism. The shared concern is the empowerment of languages as much as of the speakers of disempowered languages, vis-à-vis the hegemonic linguistic and cultural domination of African languages by imported languages such as Arabic and the languages of the former colonial masters, but also of more recent hegemonic African lingua francas. In most higher domains of verbal communication, however, one of the major factors that perpetuates material and immaterial ‘underdevelopment’ in Africa is the hegemonic dominance of non-African languages at the expense of African languages (cf. Wolff 2016a).

17

According to this ambiguous situation, some or all of their contributions would be found mentioned in handbooks and learned journals devoted to fields of general or theoretical linguistics rather than in a volume like this, which focuses on African linguistics in the narrower sense.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

African Linguistics: Conceptions and Scope

29

In more recent times and under the impact of ever increasing urbanization and the digitalization of communication in Africa, speciic changes in language use in urban agglomerations and through digital media have come under a new research focus (Chapters 18 and 19). Under this impact, African linguistics must also take issue with received notions of ‘language’ and the ideologies behind. These may relect uncritical purist notions concerning a hypothetical ‘ancestral code’, or they may relate to outdated ‘standardized’ language varieties as encoded in standard grammars written by non-African missionaries during colonial times, sometimes a hundred or more years ago. The received monolithic notion of language must be scrutinized and challenged in view of various and ubiquitous multilingual practices (code-switching, -mixing, -meshing, translanguaging, etc.) under the regime of what presently is occasionally referred to as linguistic ‘superdiversity’ (Chapter 15). African linguistics, further, faces the wider challenges of information technology and digitalization of the humanities in general (Chapter 22), including its enormous job creation potential for emerging language industries in large-scale multilingual Africa. As a regional science, African linguistics is intrinsically pluri- or multidisciplinary, if not interdisciplinary, that is, involving, at times in eclectic and additive ways, theoretical notions and methods from neighbouring sciences outside linguistics, such as sociology, social and cultural anthropology, history, political science, economics, and others. It studies, in principle, all language-related questions of scientiic interest as far as these are related to Africa. This reaches from the aesthetics of oral poetry (cf., for instance, Wolff 1980) via secretive language use and ‘manipulations’ (Storch 2011) to the postcolonial class divide between an ‘elitist’ oligarchy and ‘underprivileged masses’ based on patterns of language choice and proiciency. These approaches are akin to literary science and philology on the one side, as well as to the ethnography of communication and the sociology of language on the other (Chapter 16). In plain words, African linguistics in this broader sense studies the pragmatics and sociocultural relevance of language as such in African societies and cultures; within general linguistics, such topics have been referred to as ‘hyphenated sub-ields’ like sociolinguistics, ethnolinguistics, cultural linguistics, and applied linguistics.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:48:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.001

2 African Linguistics in Europe Roland Kießling, Nina Pawlak, Alexander Zheltov, and Arvi Hurskainen

This chapter looks, in a summarizing way, at the emergence and development of African linguistics in Europe, both in countries that were involved in colonial activities in Africa and in those that were not. For a fuller account in separate chapters, including extensive bibliographical references, see A History of African Linguistics (Wolff 2019).

2.1

European Countries with a Colonial Past in Africa (Roland Kießling)

This section presents a condensed historical overview of decisive periods in the study of African languages and linguistics as conducted by the former European colonial powers (Belgium, England/the United Kingdom, France, Germany, Italy, the Netherlands, Portugal, Spain), as representing those countries in which Afrikanistik/African linguistics was ‘invented’ as an autonomous academic discipline.1 Setting out from a concise account of the leading igures and the national institutions involved, it traces the development of research priorities, their academic underpinnings, and their political motivations, from the early exploration years up to the current stage of modern linguistics  – with the ultimate goal of identifying regional ‘schools’ of African linguistics and their (dis)continuities.2

2.1.1

General Introduction

While recordings of word lists, idioms, and attempts at monographic descriptions of African languages based on the model of Latin grammar 1

Despite not being a colonial power, but given the historically and academically close links with developments in Germany and the common language, Austria is treated alongside Germany in this section.

2

This section roughly follows earlier synopses of the history of Africanistics, as contained in Möhlig (2000), Doneux (2003), Wolff (1981c, 2013a, 2014a). For more detail, see the extended version in Kießling (2019).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

34

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

alongside translations of Christian-religious texts, authored by missionaries and explorers, reach back to the 16th century, the discipline termed Afrikanistik, that is, African linguistics, was established as late as in the 19th century in those European countries with a colonial past, more precisely England, France, Germany, the Netherlands, and Belgium. Italy, but in particular Portugal and Spain, played lesser roles. Austria linked up to developments in Germany without being a colonial power itself. In intimate intertwining with world history, research targets and ideological orientations of African linguistics have changed drastically over the past 200 years. The present chapter is organized according to decisive periods as follows: After starting from the establishment of the discipline of ‘Africanistics’ in the precolonial era in Section 2.1.2, Section 2.1.3 traces research priorities, their academic underpinnings, and their political motivations in the colonial period. Section 2.1.4 unfolds the proliferation of Africanistic discourse in the postcolonial period, as enshrined in (national) histories of leading igures, the (national) institutions involved, regular journals, and academic conferences, up to the current stage informed by modern linguistics and characterized by convergence via advanced globalization in research agendas and densiication of interdisciplinary networking, as leshed out in Section 2.1.5.

2.1.2

Precolonial Era

The precolonial study of African languages was mainly carried by humanitarian motives, that is, the Christian mission, the abolishment of slavery, and the resocialization of liberated slaves. As a necessary precondition for a sustainable dissemination of the gospel, basic Christian literature had to be translated into African languages. Before this task could be achieved, the languages, most of them unwritten, had to be analysed thoroughly for their lexical and grammatical structures and to be ‘reduced to writing’. Inevitably, the intense study of African languages also revealed irst glimpses of the immense treasury of oral literatures enshrined in these languages. As a consequence, early pioneers in missionary service came up not only with remarkable monographic studies of individual African languages, but also with anthologies of oral literature. Leading igures at this early stage were Johann Ludwig Krapf (1810– 1881), founding father of Swahili studies, Heinrich Barth (1821–1865), pioneer in the languages of the Western Sudan, Sigismund Wilhelm Koelle (1823–1902), compiler of the Polyglotta Africana, Wilhelm Heinrich Immanuel Bleek (1827–1875), founder of modern Bantuistics, Johann Gottlieb Christaller (1827–1895), pioneer in West African tonology, and Simon Leo Reinisch (1832–1919), founder of Cushitic and Omotic studies. At this early phase, the study of African languages was regarded as part of ‘Oriental studies’. While today Africanistics has become established as a discipline in its own right, the traditional Orientalistic symbiosis

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

35

still lingers on in institutional labels such as the School of Oriental and African Studies (SOAS) in London and the Institut National des Langues et Civilisations Orientales (INALCO) in Paris.

2.1.3

Colonial Era

Following the ‘Scramble for Africa’ and the continent’s division among European colonial powers in the Berlin Congo Conference of 1884–1885, new agendas and applied perspectives began to emerge for Africanistics. In order to warrant a neat and effective administration of their newly acquired colonies, it became necessary to learn more about those African cultures that had ended up in their respective domains. Thus, the ensuing boom of the study of African languages was due to the practical need to teach the basics of African languages for wider communication to new target groups such as colonial service cadets, administrators, and military oficers, in addition to businesspeople and missionaries. In order to meet these requirements, new institutions were founded, such as the Seminar für Orientalische Sprachen (1888) at the Friedrich-WilhelmsUniversity (later Humboldt University) in Berlin and the Hamburgisches Kolonialinstitut (1908) (after 1919 University of Hamburg). The necessity for colonial powers to build an eficient administration in their respective colonies fuelled the interest in African languages to different degrees, though, depending on divergent ideologies in language policy, spinning out in a spectrum between the poles of assimilation versus indirect rule. The policy of assimilation, as pursued by Portugal, Spain, France, and Italy, versus that of indirect rule, which was characteristic of England, the Netherlands, and Germany, produced different effects with respect to academic interests in African languages. While the ‘Romance’ colonial practice was basically geared towards exporting language and culture to the colonies and assimilating the populations, the ‘Germanistic’ colonial practice rather aimed at indirect control of local hierarchies. Under the former approach, indigenous languages had no practical importance but rather posed an obstacle to the project of ‘civilization’. As a consequence, their exploration had no immediate point, apart from satisfying the esoteric pleasure of Orientalists. Therefore, the establishment of specialized chairs for African languages and linguistics was rather late in France, Italy, Portugal, and Spain, as compared with Germany and England , whose indirect rule principle made it necessary for many administrators to acquire full luency in local vehicular languages in order to avoid the risk of manipulation by translators and acquire respect from local chiefs. This was a powerful incentive to produce a considerable amount of early lexical compilations and practical grammars. Throughout the colonial period, the development of Africanistics was dominated by two (partly antagonistic) leading igures in Germany, Carl Meinhof (1857–1944) and Diedrich Westermann (1875–1956), who

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

36

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

succeeded in establishing it as an academic discipline in its own right, beyond the service role as a colonial language school. They deined three proper academic agendas: (a) description and analysis of African language structures under strictly linguistic perspectives, (b) language classiication and historical reconstruction, (c) understanding of the organization of thought, which underlies African poetry, art, religion, and law, as is manifest in language. An important achievement in the struggle for descriptive adequacy under (a) was the insight to develop analytical categories from the language itself, without imposing European preconceptions of grammar onto African language structures. Among many monographs conducted in this spirit, probably the most impressive ones are Westermann’s monumental studies on Ewe. With respect to classiication (b), it was Meinhof who laid the foundation of modern Bantuistics. In the transfer of Neogrammarian inductive methods of historical linguistics, he succeeded in reconstructing the sound system, grammatical structure, and vocabulary of Proto-Bantu, the hypothetic predecessor of modern Bantu languages. Westermann, working on the considerably more heterogeneous zone of the Sudan, recognized important relations of several ‘Sudanic’ languages to Bantu, pre-shaping Greenberg’s concept of the overarching phylum of Niger-Congo. Aim (c) indexes the holistic approach of a culturally integrated African linguistics, to which most European Africanists still subscribe today, and which continues in the spirit of Wilhelm von Humboldt’s philosophy of language. According to this view, language forms the central key to the understanding of all other aspects of human cultures. This conviction is most prominent in Westermann’s work, which includes anthropological, sociolinguistic, and ethnolinguistic aspects. As a consequence of the holistic Humboldtian approach to language, the scope of Africanistics was expanded to include issues such as language planning. In addition, new ields of practical application emerged from the claim to contribute to educational projects of colonial cultural policy such as ‘Eingeborenenlenkung’, that is, the guidance of Africans in the process of cultural transformation, which was under way in the colonies as a confrontation of African and European civilizations to an unprecedented extent. Ideologically, this claim was based on a mixture of humanitarian ideals including a neo-romantic valorization of African cultures on the one side and a focus on Eurocentristic beliefs in superiority on the other side. This merged in a paternalistic attitude, which saw the European masters, by virtue of their superior position, responsible for counterbalancing the exploitation of the colonies with efforts at preserving and developing African languages. In England, the beginnings of Africanistics go back to Alice Werner (1859–1935) at SOAS, founded in 1917. With the establishment of a separate department of African languages and cultures at SOAS in 1937 under Ida C. Ward (1880–1949) and Archibald Norman Tucker (1904–1980), the new discipline of Africanistics visibly emerged from the Orientalistic shadow.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

37

Independent from SOAS, the International African Institute (IAI) had already been founded in 1926 and had a considerable impact through later publications such as the Handbook of African Languages and the language map of Africa. Due to the priority of assimilation in the French colonial project, the institutionalization of the study of African languages in France began only in postcolonial times. Before that, research into African languages developed within the conines of Orientalistic circles, most prominently represented by Marcel Cohen (1884–1974), or remained the hobbyhorse of individuals such as Maurice Delafosse (1870–1926) and Henri Labouret (1878–1959).

2.1.4

Postcolonial Era

With the independence of African states from the beginning of the 1960s, most colonial powers retained economic links to their former colonies, under new agendas of cooperative partnership in development. While the macro-regional focus, inherited from colonial times, could be seen to linger on in France, Italy, Belgium, and the United Kingdom, Africanistic research in Germany followed a less constrained regional course of orientation from the beginning, due to colonial disruption after World War I. In general, the development of Africanistics in the postcolonial era is characterized by massive adoption of modern linguistic methods, the proliferation of interdisciplinary discourses, and intensiication of institutionalized international cooperation. Methodological reinement, empirical rigour, and analytical accuracy advanced the quality of descriptive studies of many African languages and produced major discoveries, which fed into the general debate on language typology, crucially expanding the image of the human language capacity in all domains. Empiricism and pragmatism together with the Humboldtian undercurrent allowed for a growing range of cross-disciplinary ramiications towards sociology, philology, cultural anthropology, and genetics  – targeted at the study of multilingualism, language ecology, language standardization, oral and written literatures, and the reconstruction of the genetic relationship of languages, historical contacts, and migration history. In terms of genetic classiication, the American Joseph H. Greenberg had elaborated the widely accepted model, which assigns all African languages to four overarching phyla, that is, Niger-Kordofanian (today NigerCongo), Afroasiatic, Nilo-Saharan, Khoisan, with detailed dendrograms of genetic sub-branching for each of them.3 Fuelled by the progress in broad-scale synchronic description, Greenberg’s classiication was put to the test by low-level reconstructions based on the historical-comparative method, including Bantu, Plateau, Kwa, Gur, Eastern Cushitic, Eastern 3

For a detailed account of the history of the classification of African languages up to the 1970s, see Köhler (1975).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

38

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

and Southern Nilotic. Eventually, these efforts led to a higher resolution and reinement in sub-branching as well as to reshufling of subgroupings or branches at lower levels within the phyla, while the study of language contact phenomena such as pidginization and creolization processes increasingly came to raise fundamental doubts about the adequacy of the unilinear genealogical model.

2.1.4.1 England In England, Africanist expertise was concentrated in London at SOAS, which advanced to become the most inluential centre of Africanistics in Europe after World War II. Following the recommendations of the Scarborough report, the African department of SOAS was considerably expanded by recruitment of staff from the ranks of colonial oficials as well as former missionaries, such as David W. Arnott (West African languages), Frederick W. Parsons (Hausa), Lyndon P. Harries (Swahili), Ewan C. Rowlands (West African languages), Gordon Innes (West African languages), and Edward Ullendorff (Ethiopian studies) (Brown 2016:128, 131–132). This expansion was part of SOAS’s ambition to strive for large-scale academic coverage of world cultures. According to the general organization at SOAS, Africanistics followed the ‘cultural studies’ model; that is, linguistics was balanced with anthropology, history, law, and the social sciences. The postwar expansion of SOAS was followed by a period of contraction after the 1980s, which hit the African department particularly hard in the form of a reduction of staff to nearly half its size of the 1950s (Brown 2016:213–214). An outstanding igure was Malcolm Guthrie (1903–1972). While his monumental oeuvre on Comparative Bantu clearly marks the end of the Meinhof era in Bantuistics, the impact of his methods and hypotheses remains limited, except for his concept of Common Bantu and his practical referential classiication of Bantu languages. His comparative series form the basis of the Bantu Lexical Reconstructions (BLR) website housed at Tervuren (Bastin et al. 2002). Another outstanding researcher was South Africa–born Archibald N. Tucker, who became an expert on the non-Arabic languages of the Sudan and played a major role, together with Margaret Bryan and Diedrich Westermann, with regard to the Handbook of African Languages series, which was most inluential for language classiication in the pre-Greenbergian era of African linguistics. Wilfred Howell Whiteley (1924–1972) established and consolidated an East African focus. As secretary of the East African Swahili Committee and editor of the journal Swahili, he was an active editor and translator of Swahili literature and critical observer of the process of Swahili standardization and its impact on the linguistic landscape in East Africa, thus pioneering in East African sociolinguistics. Roy Clive Abraham (1890–1963) was a linguistic all-rounder who contributed in remarkable ways, including the compilation of dictionaries,

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

39

to the study of Tiv, Idoma, Yoruba, and Hausa, but also to Amharic and Somali. A strong tradition of Hausa studies, initiated by George Percy Bargery, has been continued by F. W. Parsons, among whose important contributions was the analysis of Hausa verbs in terms of a ‘grade system’, Philip J. Jaggar (see especially his comprehensive grammar of 2001), and Graham Furniss (1995), an expert on Hausa literature, both oral and written. A tradition of Cushitic studies was founded by Bogumił Witalis ‘Goosh’ Andrzejewski (1922–1994) in his groundbreaking work on Somali and Oromo grammar (Samatar 1998) and Somali poetry; it was continued by David Appleyard (Agaw), Martin Orwin (Somali), and Richard Hayward (Eastern Cushitic, Omotic). SOAS has edited an impressive range of journals, including African Language Studies (1960–1980), African Languages and Cultures (1988–1997), and, the oldest one, the Bulletin of the School of Oriental and African Studies (since 1917), which continues in print today.

2.1.4.2 France While individual African languages of wider communication had been taught in France since colonial times for largely practical purposes, a broader institutionalization of African linguistics – beside a long-standing Orientalistic tradition of Berberology – started only in the 1960s, namely with the establishment of a series of specialized chairs at the École nationale de langues orientales vivantes (ENLOV), today’s Institut National des Langues et Civilisations Orientales (INALCO). This included senior academic positions for Bantu (Pierre Alexandre), Fulfulde (Pierre Francis Lacroix), Hausa (Claude Gouffé), and Mande (Maurice Houis, Gérard Dumestre) – beside two chairs for African linguistics (Paris: Serge Sauvageot; Aix-enProvence/Nice: Gabriel Manessy) (Thomas  & Behaghel 1980). Manessy (1923–1996) is outstanding in two respects: as founder of Gur studies and as initiator of the study of language contact and multilingualism in Africa (Nicolaï 2001) – which lives on in the sociolinguistic orientation at Nice (Robert Nicolaï, Carol de Féral). Claude Gouffé, irst professor of Hausa at INALCO (1960–1994), established a strong school of Chadicists including scholars such as Sergio Baldi, Henry Tourneux, Véronique de Colombel, Bernard Caron, and the late Jean-Pierre Caprile and Daniel Barreteau. With respect to research, Africanistics became institutionalized in largely independent research units (“laboratoires”) such as Langues et Civilisations à Tradition Orale (LACITO) and Langage, Langues et Cultures d’Afrique Noire (LLACAN), both afiliated with the National Research Council (CNRS). Principally devoted to the interdisciplinary study of languages and cultures of oral tradition worldwide, LACITO has, since its foundation in 1976, substantially contributed to the description, documentation, and analysis of under-documented languages of Africa. This

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

40

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

is particularly due to the prominent participation of leading Africanists such as Jacqueline Thomas and Luc Bouquiaux, who authored a compendium for documenting and analysing unwritten languages (Bouquiaux & Thomas 1976), which has widely been used as a guideline for linguistic ieldwork in Africa, long before guides to linguistic ieldwork came to lourish. LACITO’s focus on orality also fostered a tradition of the study of oral literatures and the verbal arts. Since 1994, major Africanist expertise has been concentrated at LLACAN, which manifests, by virtue of its thematic organization along the axes of grammatical analysis, typology, reconstruction, and literature and linguistic practices, a truly Humboldtian perspective with an impressive coverage of language families across Africa. Beside Paris and Nice, the research centre Laboratoire Dynamique Du Langage, established in 1994 in Lyon, has also developed a strong focus on African linguistics with Denis Creissels and Jean-Marie Hombert specializing on the exploration of language diversity from a typological and a historical perspective. Following the irst major outlet for French Africanist research, the Bulletin de SELAF (Société d’Études Linguistiques et Anthropologiques de France), several specialized Africanist journals have emerged since the 1970s, including Cahiers de Littérature Orale (CLO), the Mande journal Mandenkan, and the general Africanist journal Linguistique & Langues Africaines.

2.1.4.3 Germany In Germany, new departments (Cologne, Bayreuth, Frankfurt, Mainz) were installed in addition to the already existing ones in Berlin, Leipzig, and Hamburg, resulting from the postwar East-West divide and federal particularism. While there is considerable curricular overlap, for example, with regard to teaching big vehicular African languages, all departments have developed complementary research proiles, the larger ones with an internal differentiation according to macro-regions and language families and, eventually, with additional chairs with dedicated specializations, regional or disciplinary or both. As for the Western part of Germany, in Cologne Africanistics is supplemented by cultural anthropology and goes together with Egyptology in curricular symbiosis. Africanistics in Hamburg harbours a chair for Ethiopian studies (Gerhardt et  al. 2008). Bayreuth has installed a chair for literatures in African languages, which adds a unique note to the concert of African literature studies in European ex-colonial languages as established elsewhere in Germany. Among the leading igures, following Carl Meinhof and Diedrich Westermann, in (West) German Afrikanistik rank, mostly German- but also some Austria-born, August Klingenheben (1886–1967), Johannes Lukas (1901–1980), Ernst Dammann (1904–2003), Emmi Kähler-Meyer (†), Oswin Köhler (1911–1996), Carl Hoffmann (†), Anton Vorbichler (1921–1999), Inge Hofmann (†), Franz Rottland (1934–2014), Thomas Geider (1953–2010), Herrmann Jungraithmayr, Bernd Heine, Wilhelm Möhlig, Ludwig Gerhardt, Ekkehard Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

41

Wolff, Norbert Cyffer, Gudrun Miehe, Rainer Voßen, Raimund Kastenholz, Rudolf Leger, Dymitr Ibriszimow, Anne Storch, Roland Kießling, Gabriele Sommer, Tom Güldemann, Klaus Beyer, Henning Schreiber, Christfried Naumann, and many others. Intensiied contacts to and cooperation with African socialist countries such as Tanzania, Mozambique, and Ethiopia during the GDR period (1945– 1989) in the former Eastern part of Germany entailed a strong involvement in practical issues of an applied Africanistics at the departments in Berlin (van der Heyden 1999) and Leipzig (Brahm & Jones 2009:313; Geider 2009). In terms of institutional organization within the faculty, both departments stand out today against the majority of West German Africanistics in that they already integrate multidisciplinarity at the department level (Wolff 2014a:47–48). Leading Africanists associated with these former GDR institutions are Hans Stumme (1864–1936), Siegmund Brauner, Rainer Arnold, Karsten Legère, Irmtraud Herms, Hildegard Höftmann, Brigitte Reineke, Ines Fiedler, Rose Marie Beck, and others, including, at Leipzig University, also August Klingenheben (between 1930 and 1936) and Ekkehard Wolff (from 1994 to 2009). Afrika und Übersee, the oldest Africanistic journal worldwide, has been published since 1910 in Hamburg.

2.1.4.4 Austria Outside colonial contexts, Austrian Africanistics in Vienna emerged from Orientalist traditions in symbiosis with Egyptology (Sommerauer 2010). The foundation was laid by Leo Reinisch (1832–1919): holding a chair for Egyptology since 1868, he had developed interests in the Horn of Africa, providing an enormous number of descriptive studies of hitherto undocumented Cushitic, Omotic, and Nilo-Saharan languages (Kießling 2008a). Institutional recognition was achieved by Reinisch’s successor Hermann Junker (1877–1962), who established the name Institut für Ägyptologie und Afrikanistik in 1923 and appointed his direct successor Wilhelm Czermak (1889–1953) as professor for Egyptology and Africanistics in 1925, who focused on the interface of language, psychology, and cultural history. The succession of Czermak seems to have been marked by an explicit separation of Egyptology on the one side and Africanistics on the other side, since 1969 represented by Hans G. Mukarovsky (1922–1992). Starting out from Atlantic languages, Mukarovsky established a focus on West African languages with a special interest in comparative Niger-Congo and historical language contact between Mande and Chadic. Beginning in the 1970s, Africanistics was constantly expanded in Vienna, while the West African focus shifted to Saharan under Mukarovsky’s successor Norbert Cyffer.

2.1.4.5 Netherlands In the Netherlands, the Vakgroep Afrikaanse Taalkunde at the University of Leiden established itself in the 1960s with a core group of eminent Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

42

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

scholars such as Jan Voorhoeve, Achille Émile Meeussen, John Stewart, Paul Newman, Paul Polydor de Wolf, Tom Cook, Thilo Schadeberg, Maarten Mous, and Felix Ameka. Jan Voorhoeve (1923–1983) crucially shaped the evolving ield of tonology by his seminal contributions to the analysis and representation of tone in Cameroonian languages. He was founding member of the Benue-Congo Working Group and organized the international Grassields Bantu project ‘in which linguists from the Netherlands, France, the U.S. and Cameroon jointly tried to tackle the Bantu Border enigma’ (Elias 1983:107). The Benue-Congo focus had already added considerable weight by Paul Polydor de Wolf ’s (1936–2003) dissertation on the noun class systems in Benue Congo. In a wider Niger-Congo perspective, John Stewart’s (1926–2006) conceptual work on tonology and ATR vowel harmony, based on Kwa data, became extremely important for both synchronic and diachronic phonology (Mous 2007). The Niger-Congo focus was continued in various strands in Kordofanian and Bantu (Thilo C. Schadeberg), Kwa (Felix Ameka), Gur, and Adamawa (Stefan Elders). An Afroasiatic track of research has been growing to include Berber (Kossmann), Cushitic (Mous, Stroomer), and Omotic (Amha). Since 1970, Leiden has hosted the annual Colloquium on African Languages and Linguistics (CALL), the major venue of international Africanistics in Europe. It also edits the Journal of African Languages and Linguistics (JALL), founded in 1979 by Paul Newman along with Thilo C. Schadeberg.

2.1.4.6 Belgium While African languages had been taught at the universities of Ghent and Leuven since the 1920s, the main centre of Belgian Africanistics evolved at the Royal Museum of Central Africa (Musée Royal de l’Afrique Centrale) in Tervuren with Achille Émile Meeussen (1912–1978), leading Bantuist and founder of the Lolemi programme on comparative Bantu, which enabled him to provide condensed revisions of Meinhof’s Bantu reconstructions. Continued by Jean Doneux, André Coupez, Claire Gregoire, and Yvonne Bastin, this centre formed the starting point of the website of BLR, which presents a database with several thousand Proto-Bantu reconstructions (Bastin et al. 2002). Since 1962, Tervuren has edited the journal Africana Linguistica with a special focus on Bantu.

2.1.4.7 Italy Italian Africanistics at the universities of Naples (L’Orientale), Rome (Sapienzà), and Torino is regionally focused mainly on the languages of the Horn of Africa, as a result of Italian colonial ambitions in Eritrea, Ethiopia, and Somalia. Deeply rooted in the philological Orientalistic tradition, which is manifest in the focus of the main journals Rivista di Studi Orientali

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

43

(Rome) and Rassegna di Studi Etiopici (Rome, Naples), most Italian Africanist scholars, still today, are either Semiticists or Cushiticists (Lamberti 2003). The early extension of scholarly scope beyond Ethiosemitic, such as in the oeuvres of Carlo Conti Rossini, Enrico Cerulli, and Martino Mario Moreno, was probably spurred by divide et impera motives in the colonialist struggle with the Amharas, who had subjugated most of these groups. The Cushitic tradition continues with leading Italian Africanists such as Giorgio Banti, Mauro Tosco, Marcello Lamberti, Annarita Puglielli, and Graziano Savà. Focused research on other languages is linked to the names of Maddalena Toscano (Swahili, Zulu), Elena Bertoncini Zúbková (Swahili), and Sergio Baldi (Hausa) in Naples.

2.1.4.8 Portugal and Spain As European colonial powers with the longest colonial engagement in Africa, both Portugal and Spain are still at the beginning of establishing Africanistics as a discipline. Due to a rigorous politics of assimilation and elitarian education in Portuguese, Portugal was rather occupied with disseminating Portuguese in its colonies (Bunk 2000), so that engagement with local African languages started only very late, as compared to the other European colonial powers (see Chapter 5).

2.1.5

Current Trends

The current situation of Africanistics in most former colonial powers is characterized by a wide proliferation in thematic scope with expanding ramiications into various disciplines such as linguistic typology, historical linguistics, cognitive linguistics, anthropology, sociology, political science, discourse studies, philology, and media sciences – the common denominator being interest in how communication works in essentially multilingual Africa. Monographic language studies continues to form a priority. While considerable descriptive progress has indeed been made during the past 50 years, many African languages still remain uncharacterized or only poorly described. For historical reasons, these white spots on the map are unevenly distributed across Africa. While the languages in former British colonies are comparatively well investigated, large gaps of linguistic documentation remain in the former Portuguese colonies. Language endangerment and death, resulting from large-scale reorientation of linguistic minorities towards African languages of wider communication, upgrade the urgency of this descriptive task considerably. The debate of classifying African languages continues. However, the focus has shifted from deining larger phylogenetic units in the Greenbergian sense to providing methodologically accurate reconstructions at lower levels and to the exploration of historical relations via distribution of linguistic features within more restricted areal conines.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

44

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

On the synchronic level, comprehensive descriptive coverage in areas such as North-Eastern Bantu allowed for the adoption of rigid dialectological principles and inspired the development of a new method, dialectometry, for an accurate calculation of synchronic linguistic distance across dialect continua. Since independence, necessities of language planning, with respect to both corpus and status, established a thread of critical discourse on standardization of African languages, African language policies in general, and practical implications of linguistic diversity management, especially in the educational sector. Postcolonial sociolinguistic realities bring the issue of multilingualism vehemently to the fore, both under the perspective of its social functions as well as for the phenomena it generates. Inspired by Carol Myers-Scotton’s work, code-switching practices have come to be studied systematically, while their consolidation in emergent new codes in African megacities gains attention, for example, in urban vehicular languages in general and in juvenile codes in particular. Furthermore, the study of African multilingualism also includes its written manifestations – multilingual literacy. Studies on the type of multilingualism imported into European megacities by transcontinental migration increasingly come to converge in a shift of paradigm away from the rather static and compartmentalized Western view of languages as normatized systems towards a more dynamic and interactional view. Multilingualism focuses on language practices such as multilingual languaging and crossing and the deliberate choices speakers make when they draw on their repertoires in different types of multilingual settings. Africanistics claims to make crucial contributions to this ield as well, given the fact that communication in most African settings is characterized by a much higher degree of language diversity and density of interaction, as resulting from internal migration, and a much lesser degree of language standardization and normatization, which goes hand in hand with more relaxed attitudes regarding the luidity of codes, revealing various types of underlying ideologies of multilingualism that challenge Eurocentric framings of academic discourse on multilingualism. While the classiicatory debate about modifying and reining Greenberg’s dendrograms continues to be fuelled by low-level reconstructions on the basis of the rigid application of the historical-comparative method, the adequacy of the unilinear-genealogical model has seriously been called into question, resulting in the development of alternative representations of language history, such as via the stratiicational model as applied to Bantu. As the outlines of a model of historical sociolinguistics crystallize in Thomason and Kaufman’s seminal work on structural relexes of different types of language contact, reconstructions have been proposed for various contact areas in Africa, including Nubian, Northern Songhay, and Southern Cushitic, which qualify as sociolinguistically more adequate due

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

45

to their effort at counterbalancing divergent changes by due consideration of language convergence with support from ethno- and sociohistorical evidence. More recently still, social network models have been introduced to Africanistics to come to terms with mechanisms of linguistic innovation at the micro level. One domain of particular historical and typological relevance, which has proited substantially from Africanistic contribution in terms of data analysis and concept formation, is grammaticalization. In semantics, the elaboration of speciic lexical ields is studied in an interdisciplinary perspective, for example, ethnobotanics, smithing, pottery, and traditional crafts. Particularly in Bantu, the relatively high resolution of lexical coverage across a vast area, combined with an advanced stage of reconstruction, allows for ine-grained studies pertaining to entire lexical ields such as mammals or to individual concepts, including bone, oil palm, and canoe. The comparative perspective in enterprises such as these increasingly upgrades the need to develop models of semantic change by generalizing from the observation of recurrent patterns of polysemy, in an attempt to fertilize historical linguistics from an African perspective. While the exploration of ethnobotanic taxonomies follows semasiological lines of inquiry, another trend, rather onomasiologically oriented, emerges from the cognitive approach, investigating the expression of concepts within a semantic domain, for example, experience and (loco-) motion. Inspired by the paradigm of ethnopragmatics, the study of African languages expands to linguistic routines and the ways they are framed by speciic cultural patterns of communication. Extensions into the realm of text linguistics focus on either the structures of dialogic discourse or the architecture of narratives. This latter ield links up with the study of oral literature and verbal art. While collections have been compiled since the beginning of the 19th century, principled scientiic analysis started only in the 1970s. As literacy and written literature in African languages keep expanding, the study of creative writing in languages such as Swahili has evolved as a comparatively new ield. Still more recently, media studies has emerged. The spectrum here reaches from ilms to music, ‘speaking’ objects, written messages on clothes such as leso and kanga in Eastern Africa, manuscripts, and urban linguistic landscaping. Topics such as health care and development, which could effectively be tackled via discourse, media, and metaphor analysis as well as via ethnopragmatical methodologies, allow for fertile interdisciplinary crossover between Africanistics, African studies, ethnology, Anglistics, and Romance studies. Last, but not least, the indispensable relevance of the study of African languages for sustainable economic development in Africa is articulated ever more forcefully to gain the attention of stakeholders and decision makers in both the European donor countries and African recipient

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

46

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

countries, demanding the overdue linguistic turn in development discourse (Wolff 2016a) by pointing to the gross underdevelopment of communicative aspects in the underpinnings of failed technologically driven development projects. After the postwar and early postcolonial period of institutional expansion, the new millennium is characterized by an accelerated pace of compression and contraction for Africanistics. This seems to be caused by two converging trends in academia: the transnational institutionalization of teaching and research and a rigorous economization of learning, inspired by neoliberal austerity policies, which tend to measure the value of academic disciplines by parameters such as student demand, student output, and immediate impact factors. While European network programs (SOCRATES, ERASMUS) contribute to facilitate staff and student mobility, the transcontinental globalization of training and capacity building had already been established in Africanistics, as most centres have increasingly attracted African linguists working on their own languages, such as Ayọ Bamgboṣe (Nigeria), Florence Dolphyne (Ghana), and Leonidas Sibomana (Rwanda). In terms of research, funding formats have become increasingly clustered, on both national and European levels. Thus, a series of projects on the documentation of endangered African languages has been realized in the context of globalized research in internationalized programmes such as DoBeS (Documentation of Endangered Languages) and ELDP (Endangered Languages Documentation Programme). In agreement with the Humboldtian approach, Africanistics has increasingly contributed to cooperative schemes of interdisciplinary research such as the German Sonderforschungsbereiche (i.e., Special Research Programmes) and various graduate schools. Apart from symposia dedicated to individual language groups or areas, regular Africanist conferences have become installed at the national level. The oldest one is the annual Conference of African Languages and Linguistics (CALL) in Leiden. Another one is the Afrikanistentag in Germany/Austria. Unique in its dedication to a single African language is the annual Swahili Colloquium in Bayreuth.

2.1.6

Conclusion

The common denominator in the history of Africanistics, which sets some Western European countries apart from other world regions, is the impact of colonialism on the formation of academic institutions and  – to some extent – on concept formation in the discipline itself. The colonial project lent a strong incentive of practical application to Africanistics that was absent elsewhere. Having grown from the earlier missionary project, the colonial boom of interest in African languages was fuelled by the practical purpose of effectively gaining and exerting administrative (and military) control

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

47

over the colonies and economically exploiting them. Crucial differences in conceptual underpinnings of colonial policies entailed different effects on the development of Africanistics. While the ‘Romance’ colonial policy of assimilation kept the interest in African languages conined to small groups of individuals, the ‘Anglo-German’ indirect rule approach produced an earlier incentive for the acquisition of comprehensive knowledge in African vehicular languages across a larger group of colonial staff, crystallizing in institutions of imperial training such as SOAS in London (Brown 2016), the School of Oriental Languages in Berlin (Stoecker 2008), and the Colonial Institute in Hamburg (Meyer-Bahlburg & Wolff 1986; Paul 2008). Africanistics grew from the philological tradition of Oriental studies, but with the rise of structuralism its methodological pacemaker soon became linguistics. Due to the constant reinement of descriptive tools, Africanistics even advanced to become one of the conceptual pacemakers itself in various ields, for example, in tonology. In contrast to the US focus on theory, the Africanist orientation in the former colonial powers is rather towards empiricism, which has spurred the proliferation of interdisciplinary discourse. Borderlines between particular national schools of Africanistics become increasingly blurred and submerged by globalized research agendas and international networking. Practical applications of Africanistics have changed from serving administrative necessities of governing in the colonial ‘project’, being replaced by challenges of development discourse in the postcolonial period, that is, the need to inform and advise stakeholders and decision makers both in the former colonial powers and in Africa about basic concepts of communication and linguistic diversity management. While traditional threads of discourse such as basic research on individual languages, the documentation of endangered languages, and language classiication issues continue, the scope of Africanistics has widened considerably since the 1990s to include ‘research on language as resource for African individuals, social groups and even whole states’ (Wolff 2016a:21–22), resulting in a shift of paradigms that characterizes Africanistics as an emergent discipline dedicated to the study of communication in African languages at large.

2.2 2.2.1

European Countries Without a Colonial Past in Africa Eastern Europe (Nina Pawlak & Alexander Zheltov)

Eastern Europe and Russia have much in common in terms of the development of African studies focused on African languages, although the region, which has a geopolitical rather than geographic connotation, was divided into distinct sub-regions several times throughout history.4 As a result, 4

Note that for academic reasons Austria, which is a country without a colonial past in Africa, is treated together with Germany in the preceding section of this chapter. For more detail, see the extended version in Kießling et al. (2019).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

48

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Russia stands apart on the map of contemporary ties and relationships within the region, whereas Eastern Germany, despite being a part of the Eastern Bloc after World War II, is more deeply rooted in the development of African studies in Western Europe. In all the countries of the region, an interest in African languages (and Africa in general) was initially part of a more general interest in other people’s languages and cultures. Since the Middle Ages, groups from Eastern European countries and Russia led scientiic expeditions that were conducive to doing research on African languages. Notes accompanying geographic and ethnographic observations (rarely full descriptions) left by travellers and adventurers in the period from the 17th century to the beginning of the 20th have so far not been fully investigated. The basis for national scholarly activities in the ield of African studies that included studies of languages grew out of the attention paid to the Orient, later extended to the countries of Northern Africa. The history of African linguistics as grounded in the tradition of Oriental studies goes back to the mid-18th century in Russia, and the 19th century in Poland, Hungary, and the Czech Republic (former Czechoslovakia). Amharic and Geez  – Ethiopian languages  – were the irst languages to be studied in depth by Europeans. Initially, the teaching of these languages was a supplementary discipline within Christian theology and Bible research. A dynamic period in the development of scholarly activities in all countries under discussion began after World War II. Earlier, prior to the end of colonialism in Africa, African languages did not igure prominently in academic ields in any of these countries, although some activities were undertaken in terms of teaching and research. Especially after the Year of Africa 1960, various scientiic research centres focused on Africa were created in almost all of the so-called socialist countries of the region. Their research interests concentrated mostly on history, politics, and ethnography, less commonly on anthropology, but they also succeeded in arousing interest in the teaching of African languages. When initiated as a separate academic ield within the scholarly system (such as in the Soviet Union and Poland), African studies was located at the biggest state universities, which had a long tradition in the ields of the humanities and philological studies. In all the countries of the region, research activities have been divided into two main ‘streams’, namely so-called Academy science as opposed to university (higher school) science. The institutions of Academy science do mainly research work, whereas the teaching of languages is conducted at universities. Aside from Amharic, the languages more commonly taught and investigated include Hausa and Swahili. Some other languages are also offered, more regularly at Russian institutions, and occasionally also at others. Contact with African languages in the process of teaching and access to language data in the course of research work have for a long time been connected with scholarship systems for African students, who

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

49

were offered studies in these countries. Students from Africa (studying medicine or other subjects) were engaged as native speakers in teaching and research programmes. In modern times, cooperation with African academic institutions has provided opportunities for sending students to the region in which the language of their interest is used. Academic degrees (at the bachelor’s and master’s levels) are offered in Oriental (also termed African and Oriental) studies, while at the doctoral level one can achieve a PhD in linguistics. The speciicity of African studies in particular countries lies in their ties with various regions of Africa. The political involvement of the earlier socialist countries in Africa has had some impact on the development of studies of the main languages of Ethiopia, Nigeria, Tanzania, Mozambique, and South Africa. Missionaries have also made a contribution to the knowledge of the minority languages in other countries (Togo, Zambia, Zaire or the DRC, Madagascar), but the achievements in the ield of African linguistics in East European countries and Russia rest mostly on exceptional academics who were inluential researchers and had a signiicant impact on the development of studies on African languages both at the institutional level and in terms of research inspirations.

2.2.1.1 The Former Soviet Union and Russia Compared to African studies in the countries that had their colonies in Africa, African studies in Russia at the early stage of their development drew from the academic tradition of Oriental studies rather than from the practical needs and were concentrated on Egypt and Ethiopia. The history of African studies in Russia dates back to St. Petersburg, where the Faculty of Oriental Languages was established in 1854 (later transformed into the Oriental Faculty). The studies originated from the Semitic and Ancient Egyptian tradition and focused on Egyptian monuments (V. S. Golenishchev), also Coptic and Ethiopian manuscripts (O. E. Lemm, B. A. Turaev, P. K. Kokovtsev) and were also connected with Old Egyptian, Arabic, Geez, and Amharic languages (I. Y. Krachkovsky, N. V. Yushmanov, among others). Later Russian studies on African languages, both at the institutional level and in terms of research inspirations, have a basis in the activities of the leading Soviet Africanist, Dmitri Olderogge (1903–1987), the “founding father” of African studies in the Soviet Union. After his return to St. Petersburg (illo tempore Leningrad), after studying in Western Europe in late 1920s, he became the Head of the African Department in the Museum of Anthropology and Ethnology and together with a famous Semitist, N.  V. Yushmanov, and an Assyriologist, A.  P. Riftin, he put forward an initiative to reorganize the Department of Semitic Languages into the Department of Semitic and Hamitic Languages, which was a part of Leningrad (present-day St. Petersburg) State University. In the 1934– 1935 academic year two groups of students were admitted to this department and studied Hausa, Amharic, Swahili, and Zulu. The department

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

50

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

received its modern name Department of African Studies in 1950–1951. In January 1934, the Workshop on African Languages at the Research Association of National and Colonial Problems was organized in Moscow and became the irst Russian conference on African languages. Studies in languages initiated in St. Petersburg became a signiicant contribution to the development of African language study, not only in Moscow but also in other centres outside the Soviet Union (e.g., Warsaw). Quite a number of graduates of Leningrad/St. Petersburg University worked or are currently working in Moscow, Paris, Hamburg, and elsewhere. The Department of African Studies headed by D. A. Olderogge from 1946 to 1987 became a university unit that offered language courses (Swahili, Hausa, Amharic, Geez, Bamana, Maninka, Kinyarwanda), whereas the Institute of Anthropology and Ethnology (also headed by Olderogge) coordinated lexicographic works on the irst African language dictionaries to be published in the Soviet Union, such as the Swahili-Russian Dictionary (1961), Hausa-Russian Dictionary (1963), Luganda-Russian, Russian-Luganda Dictionary (1969) and grammatical sketches of Hausa, Swahili, Luganda, and Zulu. Common knowledge of Russian in the Eastern Bloc secured easy access to the dictionaries and teaching materials in Russian also in other countries. The ongoing tradition of teaching and studying African languages in Leningrad (now St. Petersburg) helped to establish related institutions in Moscow. In 1960 an independent unit  – the Chair of African Studies (headed by N.  V. Gromova) – came into existence as a branch of the Institute of Oriental Studies at Moscow University, renamed in 1972 as the Institute of Asian and African Studies. It offers courses in Swahili, Hausa, Amharic, Zulu, Fula, and Afrikaans. A graduate of Leningrad University, N.  V. Okhotina, who was a specialist in Bantu languages, established in 1965 the Department of African Languages in the Institute of Linguistics of the Academy of Sciences of the Soviet Union in Moscow. She headed the department until 1988, when she was succeeded by V. A. Vinogradov, a recognized specialist in the phonology and typology of African languages. Among other publications of the department, the inluential series Fundamentals of African Linguistics is worth mentioning. All these Leningrad/St. Petersburg and Moscow institutions have been working in close cooperation in publishing and organizing conferences. An example of such cooperation is a work on the historical comparative vocabulary of Afrasian by a group of Moscow specialists in Afrasian languages that have gained international recognition (A. G. Belova, A. S. Chetverukhin, A. Y. Militarev, V. Y. Porkhomovsky, O. V. Stolbova) headed by the famous Leningrad/St. Petersburg scholar I. M. Dyakonov. The period of very dynamic social changes in the 1990s strengthened the linguistic orientation of African studies and led to much more intensive and regular contact with many centres abroad. Many students and practically all the teachers visited Africa (earlier it was either impossible or rather problematic). The most important role in this respect was played by the Mande project organized and headed by V. F. Vydrine. This project Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

51

is still ongoing. Six issues of the St. Petersburg Journal of African Studies (1993– 1997) were published in English. This journal, edited by V. F. Vydrine, enabled research by Russian scholars to be read by foreign colleagues. Among some recent important projects, two are worth mentioning: corpora of Manden languages available online and a huge (1,140 pages) volume on the Mande language (both projects are headed by V. F. Vydrine). In 2012–2014 the purpose of a linguistic expedition to the Adamawa province in Nigeria, organized and headed by A.  Y. Zheltov, was to provide a description of certain Adamawa languages. Among other activities is the conference on African studies in memoriam of D. A. Olderogge (the so-called Olderogge readings), which takes place in St. Petersburg every two years and the conference on African languages in Moscow. Many graduates of Russian universities specializing in African linguistics work abroad; for instance, K. I. Pozdniakov and V. F. Vydrine are professors at INALCO, D. Idiatov and T. Nikitina are researchers at LLACAN, Paris, D. G. Bondarev works on Kanuri manuscripts, and D. A. Nosnitsin collaborated with the Encyclopedia Ethiopica project in Hamburg, among others.

2.2.1.2 Poland Polish interest in Africa can be traced back to the 16th century, and for a long time it was focused on ethnographic studies. Expeditions to Africa in the second half of the 19th century and the beginning of the 20th century (e.g., by Jan Czekanowski to Central Africa in 1907–1909) extended the focus to languages, but linguistic data were not abundant, with the most elaborate source being the grammar of Bakwiri by Szolc-Rogoziński that was the result of an expedition to Cameroon in 1882–1890. Some attempts at conducting research on African languages were undertaken in Kraków in the irst decades of the 20th century. They were initiated by Roman Stopa with his dissertation on clicks (1935) and courses in Swahili, Hausa, Ewe, and Khoisan at the Jagiellonian University. African studies as an academic ield oficially started in 1945 and was afiliated with Arabic studies, established in 1919. In the Institute of Oriental Studies (a chair until 1972), there were regular courses on Swahili and Cushitic languages. In 2000, the chair of Afroasiatic linguistics headed by Andrzej Zaborski (1942–2014) was established. It became a centre of comparative linguistics in the Semitic and Afroasiatic ields. The journal Folia Orientalia is a forum for researchers from around the world, and it has been published since 1959 by the Oriental Committee, Polish Academy of Sciences  – Kraków Branch. It is devoted to Oriental studies and publishes contributions to African linguistics, most frequently on Afroasiatic languages and linguistics. African studies in Warsaw is also a part of Oriental disciplines. This ield was initiated in 1950 by Stefan Strelcyn (1918–1981) and focused on studies of Semitic languages (Geez and Amharic). Courses in other African languages (Hausa and Swahili) were made possible when Nina Pilszczikova, Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

52

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

a former student of D.  A. Olderogge from St. Petersburg, and Rajmund Ohly from Kraków joined the staff. Since 1977, three languages, Amharic, Hausa, and Swahili, have been taught as the main languages in the programme offered by the Department of African Languages and Cultures. In addition, courses on literature (including Ajami writings) and those on African linguistics in a comparative perspective are offered. African linguistics is one possible specialization at the master’s and doctorate levels. Close cooperation with academic institutions involved in African studies at European and African universities (mainly Bayero University, Kano) has enabled progress in teaching and research. At the Warsaw centre, individual research on Bantu, Semitic, and Chadic linguistics is inclined towards topics relevant to theoretical aspects in the ields of phonology, syntax, ethnolinguistics, cognitive linguistics, and lexicography. This research is presented mostly in foreign languages (English, French), but scientiic works are also published in Polish, in the periodical Afryka, for example. The journal Studies of the Department of African Languages and Cultures (published since 1984) presents the results of scholarly works in linguistics and in other disciplines related to Africa.

2.2.1.3 Former Czechoslovakia (the Czech and Slovak Republics) The former Czechoslovakia, and now the Czech Republic, has a long and rich tradition of African studies, which includes the documentation of early Czech-African contacts made by travellers and missionaries in the 18th, 19th, and 20th centuries, but studies on languages are not so widely developed. Better known are the achievements of some prominent researchers dealing with African languages and linguistics, namely Petr Zima from Charles University in Prague, who has prepared descriptions of two African languages, Hausa (1973) and Songhay (1997, in cooperation with Robert Nicolaї), and made a signiicant contribution to linguistic studies on areal inluences and language contact in West Africa. Another scholar, Václav Blažek, is a historical linguist afiliated to Masaryk University (Brno), who in his research work includes data from Afroasiatic and NiloSaharan languages. At the institutional level, African studies is among the Oriental disciplines offered by the Institute of Near Eastern and African Studies (INEAS). The institute was founded in 1993 to continue the tradition of Oriental studies at the Charles University in Prague, dating back to the 19th century. At present, African studies is offered as research programmes only. Some initiatives (seminars, lectures on Africa and on African languages in particular) are undertaken by institutions representing the Czech Academy of Sciences or other academic units (e.g., Centre for the Studies of Contacts and Conlicts). The journal Linguistica Brunensia, published by the Faculty of Arts of the Masaryk University in Brno since 2009, focuses on general linguistics and invites contributions on African languages made from this perspective.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

53

2.2.1.4 Hungary The development of African studies in Hungary has been motivated by an interest in learning about the Arab world and studying Arabic sources, which include data related to early Hungarian history. African language data can be found in some missionary sources, such as those compiled by the Jesuit father István Czimmermann (1849–1893), who wrote linguistic studies on Mozambique, but African languages have not become the subject of scholarly interest on a large scale. Teaching of Oriental disciplines in Hungary started with Semitic languages when a relevant department was established at Budapest University in 1873. Also today there are separate units focused on Egyptology as well as Semitic and Arabic studies that deal with questions relating to the peoples of North and North-Eastern Africa. African studies is not distinguished as an academic specialization at any level, but focus on African languages is possible within linguistic studies. Research on Africa is also conducted in the Africa Research Centre, Faculty of Humanities at the University of Pécs and in the Hungarian Academy of Sciences. In the Institute of Linguistics, at the Hungarian Academy of Sciences, research on comparative linguistics within Afroasiatic languages with a focus on the relationship between Egyptian and other branches of the family is conducted by the renowned scholar Gábor Takács. The only Hungarian scholarly journal on Africa, the Afrika Tanulmányok (Africa Studies), has been published in Pécs since 2007.

2.2.2 The Nordic Countries (Arvi Hurskainen) 2.2.2.1 Early Initiatives The motivation for studying African languages in the Nordic countries emerged initially as part of missionary activities. There was a need to communicate using local languages. Missionaries had to learn the languages, and this was made possible by producing grammars and dictionaries. The work of missionaries also included the creation of orthographies and production of teaching materials for schools. Finally, their contribution extends to such achievements as the translation of the Bible or its parts into local languages. Very early in the ield’s history, Rasmus Rask from Denmark wrote an introduction to the Accra language on the Guinea Coast (modern Ghana). Finland made its contribution in northern Namibia with regard to local Bantu languages. The earliest major achievement was the translation of the Bible into Oshindonga by Martti Rautanen. The irst draft of the whole Bible was ready in 1920 and was published in 1954 in London. Later, Toivo Tirronen worked on teaching materials, such as a grammar and dictionary of Oshindonga. Tirronen also published the Ndonga-English dictionary. Based on the material left by Tirronen, Lahja Lehtonen, in cooperation with Eljas Suikkanen, edited the English-Ndonga dictionary.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

54

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

2.2.2.2 Current Academic Contributions The following paragraphs present country-by-country descriptions of current Nordic academic contributions to African language studies.

Sweden Research on African languages in Sweden is concentrated in Uppsala and Gothenburg. In Uppsala the emphasis has been on concrete language teaching (Swahili), while in Gothenburg the emphasis has been on research on African languages in general, without extensive African language teaching. In Uppsala, the research has mostly been at the hands of Abdulaziz Lodhi, an immigrant from Tanzania. He began work as a lecturer of Swahili in 1974 in the Faculty of Arts, Department of Linguistics and Philology (currently Faculty of Languages, Department of Semitic Languages). He became Associate Professor in 2002 and Full Professor in 2008, retiring in 2012. Swahili teaching has continued, led by a lecturer in general linguistics, Swahili teaching being part of the duties. Lodhi also studied various aspects of Swahili language and culture, including foreign inluences on the language. He pioneered Swahili-Swedish lexicography and contributed to the compilation of Oxford dictionaries of Swahili. The irst professor studying African languages in Gothenburg University (Faculty of Arts, Department of Oriental and African languages) was Tore Janson. Initially having studied Latin, he shifted to African languages, concentrating on comparative Bantu studies. After retiring in 2001, he afiliated with Stockholm University. Karsten Legère followed Janson in the professorship from 2001 until retirement in 2010. Having received his education in Leipzig during the socialist regime, he worked as a researcher at the University of Dar es Salaam and later in Windhoek at the University of Namibia. The major achievement of Legère is the Languages of Tanzania Project (LoT), carried out in cooperation with the University of Dar es Salaam. He has also worked on endangered languages, in part in collaboration with Bernd Heine and Christa König. The Akie, a southern Nilotic language in Northern Tanzania, and Vidunda, a Bantu language in Central Tanzania, have been special foci. Legère was a partner in SIDA-funded linguistic cooperation projects with Maputo (2001–2003) and Dar es Salaam (2001–2010). The following researchers are based in Gothenburg. Laura Downing followed Legère as a professor of African languages. She studied information structure in Bantu languages. She also headed a research project studying the prosodic system of Somali. Christina Thornell had worked within primary education in Central African Republic (CAR) in various pedagogical tasks. Beginning in 1998, she worked in a postdoc position at the University of Gothenburg. In her PhD dissertation she describes the Sango language and its lexicon from sociolinguistic and lexicosemantic perspectives. Thornell also conducted linguistic research on Mpiemo and

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

55

investigated the knowledge and use of wild plants. Later Thornell studied a language and dialect cluster, which she terms Ukhwejo. Helene Fatima Idris studied the status and use of African languages in Sudan, in environments where the Arabic language dominates, with a focus on the Nyala area (in Darfur) and Khartoum. She also participated in the study of language policies in selected African countries. Malin Petzell studied Kami, an endangered language in western Tanzania, as part of a larger project of endangered languages. In another project, Petzell studied tense, aspect, and mood (TAM) systems in selected East Ruvu Bantu languages (Kaguru, Kami, Kwere, Kutu, Luguru, and Zaramo). Eva-Marie Bloom Ström studied morphosyntactic variation in Xhosa dialects in the Eastern Cape area. The project focused on certain grammatical constructions, such as relative clause formation, the expression of focus, and the so-called temporal mood  – areas where variation is expected. Tove Rosendal’s study aimed at identifying the reasons behind code-switching, the alternation between Ngoni and Swahili, in the Ruvuma Region in south-western Tanzania. The work of Jouni Maho, compiling very extensive bibliographies and other collections of sources and making them available on the web, must be mentioned.

Norway In Norway, African language studies are concentrated mainly at the University of Oslo and the Technical University of Norway in Trondheim. In Oslo, pure linguistic studies as well as studies on sociolinguistics and language policy are represented. In Trondheim, courses on Swahili language and culture have been given since 1980. In addition, the Department of Linguistics has been involved in supporting linguistic research in Africa by providing scholarships for African students and in various research projects in Africa. Rolf Theil Endresen at Oslo University was appointed Associate Professor of African Languages (Hausa, Fulfulde) in 1978 and Professor of African Languages in 1994. Around 2000, the position was changed to Professor of General and African Linguistics in the Department of Linguistics and Philosophy (currently Department of Linguistics and Scandinavian Studies) of the Historical-Philosophical faculty. Endresen studied varieties of Fulfulde and the Bantoid Nizaa (Cameroon). The study on phonology included, for example, Nizaa, Kafa, and Koorete. He carried out grammatical studies on Fulfulde and Subiya. Endresen supervised nine PhD students, seven of them from Africa, before retiring in 2017. Kjell Magne Yri at Oslo University studied mostly Ethiopian languages, with special focus on Amharic and Sidaamu/Sidama Afoo. In Amharic, he compared the relative and genitive constructions and their assumed semantic relatedness. The research areas of Ingse Skattum are sociolinguistics, language policy, multilingualism, and orally transmitted literature.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

56

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

In Africa, she has studied mainly Bambara and its use in various roles. She has supervised several PhD and MPhil students in the above-mentioned ields. She, together with Birgit Brock-Utne and others, studied the introduction of national languages into primary schools in Mali, and its consequences on training. Marit Lobben is a scholar of the Chadic language Hausa and has worked within major linguistic theories such as cognitive grammar, grammaticalization theory, and the prototype theory of linguistic categorization. She wrote her PhD dissertation on the polysemous syntactic relationship of causative and benefactive constructions. At the Technical University of Norway in Trondheim, Professor Assibi Amidu has given courses in Swahili language, culture, and history since 1980. He has also published extensively on various aspects of Bantu languages, using Swahili as a test language. Lars Hellan at the Technical University of Norway has contributed to the development of language technology of African languages through various projects. Together with his colleagues Dorothee Beermann and Mary Esther Kropp Dakubu (from Ghana), he has extended the work to multilingual environments. An important initiative is the construction of TypeCraft, a user-driven database for the creation and retrieval of small corpora of interlinear glossed text. Norway has been a counterpart and a principal funder in such projects as Computational Lexicography, Typology, and Adult Literacy (LegonTrondheim Linguistics Project) in cooperation with the University of Ghana, sponsored by NUFU, for the periods 1996–2000, 2002–2007, and 2007–2009, and MaLEX (Malawi Lexicon Project), computational documentation of three Malawian languages, in cooperation with the University of Malawi, sponsored by NUFU, for the period 2007–2012. The main target here was dictionary compilation, although only one (for Ciyawo) has been published so far, and the project has produced two PhDs.

Denmark In Denmark, African linguistic research rests on the shoulders of individual researchers. Karl-G. Prasse (University of Copenhagen; born 1929 in Hamburg, Germany) stands out as Denmark’s eminent scholar of Tuareg (Berber), including grammatical descriptions and studies of their oral literature. He also compiled a Tuareg-French dictionary containing about 40,000 entries and covering the two main Tuareg dialects of Niger, namely those of the Azawagh and Ayr regions, supplemented by a number of notes from the dialects of the Kél-Geres and Mali. Torben Andersen from Aalborg University worked for a number of years as a lecturer in the Department of Communication and Psychology (Faculty of Humanities) studying African languages as part of his duties. In 2012, he became Associate Professor in Comparative Linguistics. He has studied Nilotic languages, carrying out ieldwork, for example, in Sudan.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

African Linguistics in Europe

57

The individual languages that he has studied include Dinka, Berta, Kurmuk, Jumjum, Mayak, Anywa, and Päri. On Dinka, he studied verbal directionality and argument alternation, interrogative sentences, and auxiliary verbs. On Mayak, he studied number inlection, vowel harmony, and vowel alternation. He also compared the phonology and morphophonology as well as morphosyntax of Anywa and Päri. William McGregor, at the Department of Linguistics, Cognitive Science and Semiotics at Aarhus University, has done research on the Khoisan language Shua.

Finland Full-time teaching and study of African languages at Helsinki University started in 1980, when Arvi Hurskainen was appointed as lecturer of Bantu languages. The subject was enlarged to ‘African studies’ in 1989, when a professorship was also established. Hurskainen was appointed to this position, where he continued until retirement in 2006. The lectureship of Bantu languages was continued by Raimo Harjula, and from 2004 onwards by Lotta Aunio (formerly Harjula). Hurskainen’s contribution to African linguistics lies largely in the ield of language technology. The principal African language in developing technology was Swahili. Among early outcomes of this research was the morphological analyser of Swahili. The work continued by developing the disambiguation component. Later on the emphasis was on developing rulebased machine translation systems between Swahili and English as well as on other applications facilitated by rule-based language technology. Major outcomes of these developments include the analysed corpus of Swahili (Helsinki Corpus of Swahili 1.0, Helsinki Corpus of Swahili 2.0),5 a Swahili spelling checker, a dictionary compiling system on the basis of a text corpus, an intelligent interactive language learning system, a vocabulary compiler, and a bilingual corpus tagger. Another application of language technology was the programme for converting disjoint writing into conjoint writing, as is the case with such languages as Kwanyama and Northern Sotho. Sponsored by the Ministry of Foreign Affairs, Hurskainen, together with the Institute of Kiswahili Research at University of Dar es Salaam, carried out the Swahili Language and Folklore research project between 1988 and 1992. The aim was to collect linguistic data on Swahili dialects in Tanzania as well as to collect folklore in these language varieties. Hurskainen also initiated the periodical Nordic Journal of African Studies in 1992 (www.njas .helsinki.i; since 2018: www.njas.i).

5

http://urn.fi/urn:nbn:fi:lb-2014032624 (main page of the corpus), http://urn.fi/urn:nbn:fi:lb-2016011301 (annotated version), http://urn.fi/urn:nbn:fi:lb-2016011302 (non-annotated version).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

58

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Lotta Aunio wrote her dissertation on the Ha language in western Tanzania, describing various aspects including tone. Later, she worked on a comparative study of Bantu nominal tone, comparing tone in Ikoma, Nata, and Isenye languages in western Tanzania. She also initiated the Mara Project, where she, together with other researchers, studied more extensively the languages of the Mara area in Tanzania, including a recent study of syllable weight and tone. Hurskainen was followed on the chair by the German Axel Fleisch, whose dissertation was on the morphosemantic analysis of the Lucazi language. Fleisch has also studied word order variation in two Berber dialects, Taqbaylit (Kabyle) and Tashelhit (Shilha). He found that although Berber is considered a VSO language, pragmatics trigger the emergence of relatively stable discourse-conigurationality, without giving rise to a VSO > SVO shift. In another study of Berber languages, Fleisch elucidated the problem of classifying these languages along the twofold typology as either verbframing or satellite-framing, concentrating mainly on Tashelhit. Fleisch has also contributed to the study of theories and methods of African conceptual history. In 2018, Fleisch left Helsinki to take up a professorship in Frankfurt, Germany. Doctoral and postdoctoral research in Helsinki covers a wide range of topics. Based on the morphological analyser of Swahili, Seleman Sewangi from the University of Dar es Salaam wrote his dissertation on computerassisted extraction of terms in speciic domains. Based on the same analysis system, Wanjiku Ng’ang’a from Nairobi University wrote a thesis on word sense disambiguation of Swahili using machine learning technologies. Gregory Kamwendo from the University of Malawi wrote his dissertation on language policy and health services in a Malawian referral hospital. Riikka Halme wrote her dissertation on the tone system of Kwanyama. The dissertation of Don Killian was on Uduk phonology and morphosyntax.

Acknowledgements Roland Kießling expresses sincere thanks to Guillaume Segerer, Maarten Mous, Friederike Lüpke, Lutz Marten, Alessandro Bausi, and the editor for valuable comments and assistance locating relevant sources. Nina Pawlak and Alexander Zheltov are grateful to Valentin Vydrin and Andrey Shluinsky for their help in getting access to data on the contemporary situation of African linguistics in Russia. Arvi Hurskainen expresses thanks to the following scholars for their help in locating information: Torben Andersen, Laura Downing, Axel Fleisch, Lars Hellan, Karsten Légere, Abdulaziz Lodhi, Malin Petzell, Ingse Skattum, Rolf Theil, Ekkehard Wolff, and Kjell Magne Yri.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:24:06, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.002

3 African Linguistics in North and in So- Called Francophone Africa Abderrahman El Aissati, Yamina El Kirat El Allame, and Philip Ngessimo Mathe Mutaka This chapter looks, in a summarizing way, at the emergence and development of African linguistics in so-called Francophone Africa, both in countries of sub-Saharan West and Central Africa and in North African countries. For a fuller account in separate chapters, including extensive bibliographical references, see A History of African Linguistics (Wolff 2019).

3.1 3.1.1

North Africa (Abderrahman El Aissati & Yamina El Kirat El Allame) Introduction

In this brief overview, the reader will ind information on the history and development of language studies in North Africa in general, with focus on studies on Amazigh, widely known as Berber, and varieties of spoken Arabic. In North Africa and the Middle East, Arabic linguistics has the lion’s share of attention, mostly due to the position of Arabic in the Islamic World. The part on Arabic linguistics is presented for the region as a whole, partly because issues around Arabic were seldom local in nature, as this language has continuously played a major role in a Pan-Arab ideology, covering the Middle East and North Africa. Readers interested in more linguistic details and extensive bibliographical references on the languages of the area are referred to El Aissati and El Kirat El Allame (2019). For ease of exposition, we have divided this overview into three parts for each country: the precolonial period, the European colonization period, and the independence period. Although these periods differ for the respective countries in the area, their signiicance in shaping different kinds of research interests remains very comparable. The Islamic period was a warranty for interest in Arabic, the colonization period brought about more interest in local varieties, with continuation of interest in Arabic Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

60

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

and Islam in general,1 while independence allowed local universities to have their own research agendas, in many cases starting with a hostile attitude towards languages other than Arabic and sparking debates on national languages and languages of education. The Arab nationalism has been using Arabic as a pillar in building new postcolonial states, usually, but not always, with Islam as a common denominator for countries where Arabic is – mostly – the only oficial language. Before dealing with work on linguistics in North Africa, we provide a short overview of the demographics of the countries in the area. Since it is not common to conduct language censuses in this region, we mostly present estimates from the literature. The only country in the region with an oficial population census including languages is Morocco. With a population close to 33 million, oficial statistics (2014) report a percentage of Amazigh speakers that amounts to 26.7, a decline of 1.5 percent in ten years. Close to 90 percent of Moroccans are reported to speak Moroccan Arabic (widely known as ‘Darija’), and 0.9 percent speak Hassaniyya.2 In Algeria, the number of speakers of Amazigh is estimated to be around 25 percent of a population of about 42 million, but no oficial statistics are available. Tunisia counts between 1 percent and 5 percent of Amazigh speakers out of a population of about 11 million, mostly concentrated in the island of Djerba and small enclaves in the southern part of Tunisia, and again without any oficial statistics. Libya’s oficial policy was never concerned with languages other than Arabic until the 2011 uprising. The number of Amazigh speakers is estimated at about 5 percent of the population, or about 135,000 (of the total Libyan population of about 6 million in 2017), although most are Amazigh-Arabic bilinguals (LaVerle 1989). In the western part of Egypt, a concentration of Amazigh speakers is found in the Siwa Oasis. Coptic, a previously widely spoken language in Egypt, has ceased from oral communication and is restricted to religious circles (the Coptic Church). The percentage of Egyptian Arabic speakers is close to 100 percent.

3.1.2

Arabic and Indigenous Languages in Precolonial North Africa

In the predominantly Muslim society of North Africa, religion has played a prominent role in studies of Arabic, witness the many institutions that were founded by Muslim rulers, such as the Al-Azhar mosque in Egypt, the Azzaytouna mosque in Tunisia (ca. 737 CE), and the Al-Qaraouine mosque in Morocco (ca. 859 CE), all of which were built in then non-Arabic-speaking areas. These have had Arabic grammar as a main

1

The Congrès International des Orientalistes, which held its inaugural session in 1873 in Paris, is a good illustration of colonial powers’ interest in the ‘Orient’. The proceedings of many of its sessions are available online (e.g., first session at http://gallica.bnf.fr/ark:/12148/bpt6k65804267/f1.image; accessed December 28, 2016).

2

See http://rgph2014.hcp.ma/attachment/611748/ (accessed March 2, 2017).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

African Linguistics in North and in So- Called Francophone Africa

61

component in their syllabi. Numerous works trace the history of the Arabic linguistic tradition from different perspectives and offer comprehensive accounts of the phonology, morphology, and syntax of the language, dating as far back as the eighth century (e.g., Owens 2006, 2013). Because of the availability of such work, we will allow more space for treatments of other languages in the region, especially indigenous colloquial varieties. A look at the existing references reveals that most of the published works focus on the grammar of the Arabic language, its vocabulary, and its glossaries. Most of the studies were meant to codify the Arabic language and teach ‘proper’ Arabic to new converts. Some of these references are still basic, such as Sibawih’s al-kitab (Baalbaki 2008).

3.1.2.1 Egypt In Muslim Egypt, one of the oldest treatments of a sociolinguistic issue deals with the shift from Coptic to Arabic, recorded in a text known as the ‘Apocalypse of Samuel of Qalamun’, which, though originally written in Coptic, has been preserved only in its Arabic translation (Cauwenbergh 1914). The text has no precise date of authorship, but clearly addresses the results of contact between Arabic and Coptic, which started taking place in the irst century of Islamic expansion. A Coptic-Greek bilingual lexicon was written in the Middle Ages (see Sidarus 1978 and Hamilton 2006 for a general overview of this period). Coptic was known to Western scholars already in the late 15th century (Emmel 2004, cited in Grossman et al. 2015:4; for a review of studies on the Coptic language, see Hamilton 2006:193–272). Coptic is a good illustration of the role of religion in fortifying the position of a language, if the language in question is associated with this religion. As Sidarus (1978:125–126) notes, ‘the majority of Coptic philologists will be found to be eminent biblical scholars’. Nubian was also the subject of study by a number of authors, such as Lepsius (1880; see Jakobi & Kümmerle 1993). Other languages of Egypt did not receive as much attention in this period.

3.1.2.2 Libya The languages spoken in Libya have received less attention in linguistic research than those spoken in other parts of North Africa (see Behnstedt & Woidich 2013:322, though their observation was made about lack of substantial dictionaries in the area, and Benkato & Pereira 2016). Before the introduction of Arabic in Libya in the seventh century CE, Berber was already in use in the area. An Indo-European descendant, Domari, is still spoken in Libya, as is the Nilo-Saharan Tedaga, which is reported to have a few hundred speakers in the southern part of the country, close to the borders with Chad (Scheele 2016). Benkato and Pereira (2016:149) report that they could not ind a single work on the Libyan varieties of Hausa and Tebu, nor on the Greek variety still spoken by the Gritliya community in eastern Libya.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

62

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

While Arabic has attained a prominent place in Egypt, partly through its omnipresence in educational institutions, in Libya it seems that it did not beneit from such institutional support. This is most likely due to the absence of any signiicant government headquarters in the area at the time of the Muslim caliphates. Until 1914, at the fall of the Ottoman Caliphate, there were no major Islamic schools, nor any known publishing houses or higher education centres in the country. However, a few publications can be found dating back to the 19th century, such as Rohlfs (1872) on numerals in Ghadames Berber and Newman’s (1882) book on Libyan vocabulary.

3.1.2.3 Tunisia The history of old Tunisia (the Ifriqiyya province for the Muslim conquerors) testiies to its importance in the Muslim world. One of the oldest learning centres established by Muslims in North Africa is the El-zituna (spelling variants: Ezzituna and Al-Zituna) mosque (founded in the beginning of the eighth century CE), where subjects such as theology, jurisprudence, and grammar were taught. Arabic grammar was taught right from the start and later on other languages as well. None of the references cited in Youssi (1989) and Boukous (1989) on Tunisian Arabic and Berber, respectively, predates the colonization period.

3.1.2.4 Algeria The precolonial history of Algeria is rather similar to that of Tunisia with regard to interest in language matters. The indigenous languages of the country, which fell under the Ottoman administration until its colonization by the French in 1830, were not treated in any reference we could ind. Studies on Arabic grammar that appeared in other parts of the Muslim empire were also relevant for Algeria. Algerian Arabic, however, just like Algerian Berber varieties, did not igure in any studies prior to 1830 (cf. Youssi 1989).

3.1.2.5 Morocco In Morocco, some interest in languages other than Arabic was recorded in the pre-colonization period. There are about 200 Berber (Tachelhiyt) manuscripts dating back to the end of the 16th century (Boogert 1998:10). Ibn Tunart authored an Arabic-Berber dictionary of the 12th century, on which a number of more recent Arabic-Berber dictionaries are based. The best known of these is the dictionary by Al-Hilali, signed by its author in 1655. The Berber variety in which these two dictionaries are written is Tachelhiyt (southern variety in Morocco). Other than this, there were a few works on Moroccan languages, such as Blanc (1906), which are translations of popular folk stories, and an introductory book to Western Arabic dialects by Dombay (1800), which contains a bilingual Arabic-Latin word list.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

African Linguistics in North and in So- Called Francophone Africa

3.1.3

63

The Colonial Period

The interest of the European colonizing powers in North Africa starting from the 19th century would result in numerous ethnographic accounts of indigenous populations of the area. This interest is characterized by the attention that local languages and populations would receive, and it initiated an accompanying interest in Standard Arabic by local intellectuals, mostly as a result of direct contact with Western languages and cultures. Such contact has induced major efforts to revive Arabic mainly through translation from other European languages, resulting in the widening of the functional spheres for Arabic (Suleiman 2008).

3.1.3.1 Egypt In the face of other languages, mainly Turkish, French, and English, Arabic was to be a vigorous symbol in Egyptian nationalism and would continue to function as such in Pan-Arab nationism. An interesting case noted by Mitchell (1991:133) regarding the defence of Arabic is worth mentioning here. Napoleon had prepared a proclamation to Egyptians in Arabic during his invasion of Egypt in 1798, written by French Orientalists, which was to be reproduced in full by a famous chronicler, Al-Jabarti, followed by pages of commentary pinpointing language errors, and as such showing the ignorance of French authorities and their deception techniques. Almost a century later, another Egyptian scholar, Ahmad Luti Al-Sayyid, expressed his opinions about the differences between colloquial and Standard Arabic in Egypt and pleaded for an ‘Egyptianization of Fush’a [the Standard]’ (Suleiman 2008:32). In the second decade of the 20th century, a period when Egyptian nationalism attained its apogee, serious debates were carried out about which language should be the symbol of Egyptian nationalism. Advocates of a return to Coptic had their arguments as did those who pleaded for the use of colloquial Egyptian Arabic, next to those who saw that Standard Arabic should be the sole language to fulil this need (Suleiman 2008:32–33). Colloquial Egyptian Arabic did receive some attention from Western scholars in the 19th century, of whom the oldest known is Spitta (1880). The earliest known study of Egyptian Arabic, where the colloquial is recognized as a separate language different from Standard Arabic, was by the Egyptian Yusuf al-Maghribi, where the author discussed the impurity of Egyptian Arabic, which despite its heavy loans from Coptic remains an Arabic variety (Zack 2009). Egyptian languages other than Coptic and Arabic have received much less attention. Miller (1996) provides an overview of works on Nubian, Beja, and Berber in Egypt. Nubian has had a bigger share in linguistics and literature, partly because of its importance in terms of number of speakers. Interested archaeologists and the massive displacement of Nubian populations have also played a major role in this interest.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

64

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

3.1.3.2 Libya The low interest in Libyan languages continued throughout the 20th century. During the Italian protectorate (1911–1943) and the British and French protectorates (up to 1951), there were no major educational institutions where Libyan languages (and literature) would be studied inside the country, as was the case in Morocco and Algeria. A few works were done on the Arabic of Libya, especially on the variety of Tripoli, and to a much lesser extent on the Arabic of Benghazi. Regarding Libyan Jewish Arabic varieties, there seems to be very little work done on them. Other varieties that have been scarcely studied are those of southern Libya (Fezzan region). A number of works cited in the annotated bibliography of Benkato and Perreira are language learning manuals and dictionaries that were very likely provided for the beneit of the Italian army. Berber languages of Libya were also treated in a few publications, but to a much lesser extent than Arabic varieties.

3.1.3.3 Tunisia (1881–1956) Tunisia fell under the protectorate of France in 1881. Before this period, there was already some academic interest in North Africa since Algeria had been a protectorate of France for almost half a century by that time. Interest in the languages of Tunisia can be seen in a number of publications in the Revue Tunisienne. This journal was published by the Institut de Carthage (Association Tunisienne des Lettres, Sciences et Arts) between 1894 and 1913 (20 issues). A complete index of the contents of the different issues has been published online by the French National Library (http:// gallica.bnf.fr/ark:/12148/bpt6k580392q/f1.image). The oldest reference to a work on the Arabic of Tunisia in the bibliography of Youssi (1989) is Stumme (1899), while the oldest reference we could ind to a study of Berber in Tunisia is Calassanti-Motylinski (1885). One of the earliest studies on Tunisian Berber varieties is Basset (1891).

3.1.3.4 Algeria (1830–1962) Studies on Algerian varieties of Amazigh in general date back to the irst third of the 19th century. Yet the most signiicant studies started in 1858, with General Hanoteau’s publication ‘Essai de Grammaire Kabyle’. All these studies follow the French school. Their quality depends on the identity and vocation of the author, whether professional grammarian, military oficer, proselyte missionary, or non-professional lover of exotic languages (Boukous 1989). The teaching of Amazigh started at the Ecole supérieure des lettres in the 1880s, and a certiicate of Amazigh language (Brevet de langue berbère) was created in 1885. Some Algerian publishers specialized in the publication of the works of Amazigh and Arabic, such as Jourdan, Carbonel, and la Typo-Litho. A database of Amazigh documentation was put in place by les Pères Blancs (Boukous 1989) at Fort National.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

African Linguistics in North and in So- Called Francophone Africa

65

At the beginning of the period of colonization, work on Algerian languages was almost exclusively conducted by military oficers and missionary priests. Specialized academics followed later on (Chaker 1982). A few indigenous authors joined in the research on local languages, such as Ben Sedira and Boulifa (see El Aissati  & El Kirat El Allame 2019 for detailed references). Among the contributions of colonial dialectology in Algeria is the elaboration of the contours of the Algerian linguistic map. According to Basset (1952:4), a third of the Algerian population speaks Amazigh and is subdivided into four main language areas.

3.1.3.5 Morocco (1912–1956) The colonial period in Morocco between 1912 and 1956 witnessed the continuation of interest in Arabic studies, but opened the doors to studies of Berber and Moroccan Arabic. The École supérieure de langue arabe et de dialectes berbères de Rabat edited the journal Les Archives Berbères between 1915 and 1920. For this purpose, a special committee was created, Comité d’études berbères de Rabat, with the goal of centralizing the works carried out in the different Berber regions, and obtaining practical results concerning the organization and the administration of Berber tribes (Lyautey 1915). Les Archives Berbères would appear in a joint publication with the Bulletin de l’École supérieure de langue arabe et de dialectes berbères de Rabat. Starting from 1921, the publication would bear the name of Héspéris. Archives berbères et Bulletin de l’Institut des Hautes Etudes Marocaines. In 1960, it would join with Tamuda under the name Héspéris-Tamuda. Tamuda was established in 1953 in Tetouan, a major city in the north that fell under the Spanish protectorate. In addition, a dedicated institute was founded for the study of Moroccan languages under the Spanish protectorate, namely the Instituto General Franco de Estudios y Investigación Hispano Árabe in Tetouan, with publications such as Cuadernos de Estudios Africanos, Revista de Tropas Coloniales (later to become Africa), Revista Rifeña, and Tamuda (Boukous 1989:122).

3.1.4

Linguistics in North Africa after Independence

After independence, language planning would play a major role in the national policies of these countries. Interestingly, unlike in other countries of the Maghreb, where the discussion was very heated about Arabicizing the countries through lessening the impact of European languages, in Egypt a discussion was going on regarding the suitability of Egyptian Arabic to be the symbol of the nation’s unity. Awad (1947, 1980) was a fervent defender of the idea that Egyptian Arabic is a language in its own right and can fulil all tasks that Standard Arabic can since it has its own phonology, morphology, syntax, and lexicon. Another scholar, Taha Hussein, proclaimed a different view that was common outside Egypt,

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

66

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

especially in other countries of the Maghreb. Hussein advocated the adoption of Fush’a not only as a language of religion but also as the ultimate symbol of the nation, and offered the Coptic Church his help to ‘improve’ the use of Arabic in its services (Suleiman 2008:37–38). Quite different from the other countries dealt with in the present chapter, Egypt did see a rise of indigenous linguistic activities immediately after its independence (cf. El Aissati & El Kirat El Allame 2019). However, lesser used languages in Egypt would remain under-represented. Libya joined the Arab League in 1953, before the French had completely withdrawn their military personnel from the country. The irst university founded was the Libyan University (al-jami’a al-libiya) in 1955. Compared to the other countries of North Africa, Libya had less scholarship on its languages in modern times, especially by Libyan researchers. In the irst few decades after independence, work on Libyan languages remained a task fulilled mostly by a few foreign researchers. Concerning Berber, independent Libya did not recognize its existence, let alone allow it to be a topic of research. The small number of Berber speakers in Libya can also be a reason why it has received so little attention from indigenous linguists themselves. We were unable to ind a single dissertation or publication by a Libyan linguist/ethnographer earlier than Yedder’s 1982 research. European-based researchers, especially from Italy, continued to write on Libyan languages and cultures. Tunisia bears some resemblance to Egypt concerning the interest of indigenous researchers in the languages of their country, and the position of colloquial Arabic in a multilingual society. A number of authors, such as Abdesslam (1956) and Khmiri (1958), took up the issue of the relationship between Standard and Tunisian Arabic. Others dealt with the relationship between Arabic and French in education and other sectors of society (e.g., Hamzaoui 1970). Language contact and borrowing was also on the agenda of indigenous authors (e.g., Ghazi 1958 and Baccouche 1966; for details, see El Aissati  & El Kirat El Allame 2019). Topics in linguistics proper, especially phonology and morphology, also had a share in the interest of indigenous linguists. Tunisian Berber has attracted very little interest in linguistic research, possibly relecting the small percentage of speakers in the country (Boukous 1989). The most recent reference to a work on Tunisian Berber we could trace is Gabsi’s dissertation (2013). After the independence of Algeria in 1962, research on Algerian colloquial languages was not a priority in the newly created universities. In 1962, the Berber Chair was cancelled, even before the creation of the University of Algiers. Some interest in Algerian Arabic could be discerned, as can be seen from publications around this time. The irst two decades after the independence of Morocco in 1956 were marked by low production in Berber studies inside Morocco and a fervent discourse on the importance of Arabic. Galand (1989:66) points out that Berber studies has gained international interest since 1956. Referring to

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

African Linguistics in North and in So- Called Francophone Africa

67

studies conducted during the protectorate, he rightly observes that their merit lay more in documenting the different varieties than in contributing to theoretical linguistics. Paradoxically, the irst generation of Berber linguists was more tempted by theory and seldom provided large corpora on which their work was based, a fact that Galand deplores as it risks fragmenting our knowledge of Berber in general (1989:70). Unlike Tunisia or Egypt, there were no considerable publications on Moroccan Arabic, neither was the multilingual nature of the sociolinguistic landscape a topic of research. Only in the 1970s did Moroccan researchers start tackling linguistic issues relating to local varieties of Berber and Arabic (see El Aissati & El Kirat El Allame 2019 for details).

3.2 3.2.1

Official French-Speaking West and Central Africa (Philip Ngessimo Mathe Mutaka) Introduction

The year 1960 symbolizes the period when most African oficially Frenchspeaking countries got their independence. In the period that preceded independence, colonizers came with missionaries, whose main objective was to ‘spread the word of God’. Thus they became interested in learning African languages in order to start translating the Bible into them. Some of the early dictionaries and grammar books were thus essentially written by missionaries, more precisely Christian missionaries. With the school system also came the creation of universities in French-speaking African countries. The early teachers essentially came from the countries of the colonizers, France for Senegal, Côte d’Ivoire, and Cameroon, Belgium for the Democratic Republic of the Congo. The students who were interested in linguistics were then given scholarships to study in Western countries, particularly Belgium, France, Germany, the Netherlands, and the United States. Because these students wrote their doctoral theses (doctorat de 3e cycle, doctorat d’état, PhD) at these Western universities, they automatically represented the various schools of linguistics as found in Western countries in African countries. Not only for Francophone Africa, one needs to emphasize the role of SIL International, a nonproit organization dedicated to promoting languagebased development within indigenous communities through linguistic research, translation, and practical services, primarily in order to foster and spread Evangelical Christian texts and scriptures. In this work, SIL International has been the natural collaborator of African linguists working on their own languages within their countries, through the creation of language committees, as in the case of NACALCO (National Cameroonian Language Committees) in Cameroon. Linking up with the truly African WALS/SLAO (West African Linguistic Society/Societé linguistique de l’Afrique Occidentale), it has also become instrumental by publishing JWAL

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

68

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

( Journal of West African Languages), a research tool of this oldest academic linguistic society in West Africa. In the sections to follow, primary focus will be on the works of African linguists publishing on African languages, whether afiliated with African authorities and universities or working at academic institutions outside Africa. Outstanding and inluential contributions by non-African linguists will also be mentioned as far as they pertain to so-called Francophone Africa.

3.2.2

African Linguistics in French-Speaking West and Central Africa

The history of African linguistics in oficially French-speaking West and Central Africa has partly been treated in Clements (1989) and Doneux (2003). They contain a Western perspective on the early history of African linguistics until the 1980s, including the history of African linguistics in so-called Francophone countries. This chapter complements the work by Clements and Doneux, mainly based on what I was able to ind on the internet on relevant authors and their works, in order to represent what has been achieved by contemporary African linguists. Another valuable published source is Bostoen and Maniacky (2005). It includes a number of doctoral dissertations, dictionaries, and grammars written by African and non-African researchers on African languages. For the categorization of the works, I borrowed from the domains of linguistics as synthesized in Clements (1989).

3.2.2.1 Historical Linguistics, Comparative Linguistics, Classification Among the African linguists to be listed under this heading, mention must be made, irst of all, of Capo’s works (1988, 1989) in which he explores a number of languages spoken in Benin. Kadima (1969) is a seminal contribution on the class system of Bantu languages. Bakume Nkongho Ojong (2012) deals with the reconstruction of the proto-language of the Nyang dialects of Cameroon. Other important works under this heading are mostly done by Africanists from Western universities either in the form of dissertations (e.g., Voeltz 1977; Pozdiniakov 1978; Stallcup 1978; Grégoire 1990; Janssens 1993; Vydrine 2001) or as part of their research on African languages (e.g., Welmers 1958; Caprile & Jungraithmayr 1978; Bouquiaux 1980; Hyman  & Voorhoeve 1980; Bastin 1983; Hombert 1988; Creissels 1991; Egner 1992; Botne 1994).

3.2.2.2 Monographic Descriptive Grammars Works in terms of descriptive grammars are often doctoral dissertations, including those by Djarangar Djita (1989), Lébikaza (1999), Bokula (1966), and Takassi (1996), or works whose authors have beneited from

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

African Linguistics in North and in So- Called Francophone Africa

69

a Humboldt grant (from Germany), as is the case of Tamanji (2009) and Akumbu and Fogwe (2012) for their work on their grammars on Bafut and Babanki, respectively, or a Research Institute for Asian and African Languages and Cultures (ILCAA) grant (from Japan), as in the case of Diouf (2001). Works by Africanists from Western universities include Benoist (1969), Coupez (1955), Dugast (1971), Meeussen (1971), Stappers (1973), and Vansina (1959).

3.2.2.3 Descriptive Linguistics Much work has been published on phonology and morphology (e.g., Coupez 1955; Vincke 1966; Guarisma 1969; Leroy 1977; Rialland 1979; Coulibali 1983; Achie Brouh 1996; Gbéto & Möhlig 1997; Essono 2000; Akumbu 2012). Grégoire (1976, 1990) is the only author who speciically mentions phonetics and tonetics in the titles of her dissertations. Some of Motingea Mangulu’s works were published in Tokyo, which implies that he has been receiving grants from Japan for his linguistic research (e.g., Mangulu 2005, 2008, 2010, 2012). This is an opportunity to point out that ILCAA in Japan has been sponsoring linguistic studies on African languages as well. Although works on syntax, semantics, and lexicon consist mostly of papers (see Bassène  & Creissels 2011; Sair  & Bassene 2017; Bassene  & Sair 2017), I would like to point out that there are books that do not readily pop up in Google searches of authors. This is certainly the case for Edmond Biloa’s syntax books (see Biloa 1995, 1998, 2004, 2005) and Bassong (2013).

3.2.2.4 Sociolinguistics, Dialectology, Language Planning This is one of the areas in which several African linguists have published, particularly in sociolinguistics and language planning. The topics explored include language contact (e.g., Bitjaa Kody 2004; Heine et al. 2005), transborder languages (e.g., Chumbow 1999), bilingualism (e.g., Essizewa 2011), ethnolinguistics of individual names (Mangulu 2014), tales and proverbs (Dugast 1971; Kamwangamalu et  al. 2013), and narratives of pygmies (Bahuchet 1989).

3.2.2.5 Applied Linguistics, Contrastive Linguistics, Orthography Orthography is a major domain that African linguists grapple with as they face the need to design writing systems for the numerous languages on the continent. Norbert Nikiéma is one of the rare US-trained linguists who speciically specialized in orthography as he wrote his doctoral dissertation on orthography (Nikiéma 1976). Tadadjeu and Sadembouo (1984) attempted to harmonize the orthography of the more than 200 languages spoken in Cameroon. Other works within the domain of applied linguistics include Diki-Kidiri (2008), Kamwangamalu (2016), Mba and Sadembouo (2012), Nikiéma and Kinda (1997), and Nikiéma (1979, 1980, 1982).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

70

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

3.2.2.6 Miscellaneous African scholars in particular to be mentioned here are many. There is Alidou’s work on gender and Islam (Alidou 2005; Alidou & Sikainga 2006); Adouakou (2009) and Adouakou and Ahoua (2009) on the learning of languages spoken in Côte d’Ivoire; Ahoua (2002) on tales in Côte d’Ivoire; Bokamba and Bokamba (2005) on the learning of Lingala; Musanji (2011) on oral texts and the future of African cultures; Gbéto (2000) on loanwords from European languages into languages spoken in Benin; Tchagbale and Lallemand (1982) on tales of Togo; Ndoma (1977, 1998) on literacy and language planning in the DRC; Mbaya (2005) on linguistic attitudes in Senegal; and Afeli (2003) on multilingualism in Togo. A number of works have been published by ILCAA in Japan, including Kagaya (1992) and Yukawa (1992) on vocabulary; Kagaya and Nobuko (2006) on a bibliography of African studies; Kawada (1997) on sounds; Traoré (2000) on tribalism in Africa; Kiango (2000) on Bantu lexicography; and Goyvaerts (2000) on conlict and ethnicity in central Africa.

3.2.2.7 Linguistic Atlas Several oficially French-speaking countries have promoted linguistic atlases that may have been initiated by Western linguists, but African linguists have been active collaborators in the compilation and elaboration of these atlases. This is the case of the linguistic atlases of Cameroon and of the DRC, in which the late professor Marcel Kadima Kamuleta was involved. Issa Takassi (1983) was responsible for publishing a sociolinguistic atlas for Togo.

3.2.3

Outlook

The irst African students to study African linguistics were sent to Western universities to be trained as linguists, which led to a perpetuation of Western ideology on linguistics. As some of them returned to their home countries, they found themselves as the agents of the different Western schools of thoughts that they were now enforcing in their own schools or universities. Those who worked in the administration soon realized that their local governments did not give a high priority to the kinds of studies they had done in Europe or the United States. They were faced with an agenda that they were not prepared for, that is, to undertake language planning studies, produce literacy materials for their local populations, identify the speciic languages in their territories, and eventually put them in linguistic atlases. Further, insuficient funding of linguistic studies by local African governments is a serious impediment to the production of studies that would relect a real perspective of African linguistics that is in consonance with the genuine needs of their countries. Most of the time, because studies are funded by foreign organizations, African linguists who do research under

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

African Linguistics in North and in So- Called Francophone Africa

71

such grants are somehow obliged to produce the type of work that these organizations require. The issue of the nature and dissemination of the material produced by African linguists for local needs can also explain why quite a few of them are not widely known beyond their countries. This may also partly explain why some of the brightest linguists decide to leave their countries and work abroad. This, however, does not have to be interpreted as something negative, as they may produce better work for the beneit of their country while working abroad. Another serious problem that African students in African linguistics face is employment. The areas of linguistics as taught in Western universities and reproduced in African universities cannot provide enough jobs. Recently, the rector of the University of Yaoundé 1 decided that the different departments must propose tracks that give students who graduate from those departments the possibility of being directly employed in the job market. Gabriel Mba (personal communication), from the department of African languages and linguistics at the University of Yaoundé 1, Cameroon, lists four different tracks that should be part of the training students receive in departments of African languages and linguistics: 1. The development of writing systems in local languages as well as lexical engineering, that is, new terminology for concepts not found in these local languages along with the creation and implementation of local language academies. 2. The development of didactic material and literacy material in a multilingual context and multicultural context. The aim of this track would essentially be to train students as writers or editors of literature (novels, tales, epics, etc.) in their native languages. This material would facilitate the transition of acquiring knowledge from local languages to national languages such as French and English. 3. Language pathology specialists (in French, linguistique clinique). The students in this training track would be able to identify language pathologies in both oral and written speech and give orientations to the patient. Such pathologies include dyslexia, dysorthography, dysgraphia, vocal troubles, articulation problems, luency troubles, and aphasias. Once the speciic trouble has been diagnosed, the student can then refer the patient to a medical practitioner or speech therapist. 4. Linguistics and socioeconomic development. The aim for this track would be to make sure that local development projects funded either by the government or by nongovernmental organizations are fully understood by the targeted communities. Explanations to the communities of such projects are best given in their local languages. That is why there is a need for students to be trained in translation and interpreting from oficial languages such as English and French into the various local languages.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

72

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Relevant in the future of African linguistics is community-based work as proposed for example in Rice (2011) and the volume edited by Shannon Bischoff and Carmen Jany (2018). Rice deines community-based research as follows: ‘It is research that is conducted by, for, and with the participation of community members  . . . community-based research aims not merely to advance understanding but also to ensure that knowledge contributes to making a concrete difference in the world’ (Rice 2011:189). This is a concern we share with work that is being promoted by Gabriela Pérez Baéz and her collaborators in Latin America. They have a project on data integration and tool development and on a database creation and maintenance that will contain archival and ethnographic collections for the purposes of knowledge recovery and implementation in language and knowledge revitalization efforts for their initiative of language and culture revitalization (personal communication).3 Such community-based work would enable African linguists to deeper explore topics related to African cultures that would it in the domains of sociolinguistics, anthropological linguistics, documentary linguistics, etc.

3

Pérez Baéz is Curator of Linguistics, Department of Anthropology, MRC 112, National Museum of Natural History, Smithsonian Institution.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:27:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.003

4 African Linguistics in NorthEastern and So-Called Anglophone Africa Marie-Claude Simeone-Senelle, Wolbert G. C. Smidt, Ronny Meyer, Angelika Jakobi, Amani Lusekelo, Bruce Connell, and Akinbiyi Akinlabi This chapter looks, in a summarizing way, at the emergence and development of African linguistics in so-called Anglophone Africa to the exclusion of Southern Africa (see Chapter 5), that is, countries in North-Eastern, Eastern, and West Africa. For a fuller account in separate chapters, including extensive bibliographical references, see A History of African Linguistics (Wolff 2019).

4.1 4.1.1

African Linguistics in North-Eastern Africa Djibouti (Marie-Claude Simeone-Senelle)

Two main ethnolinguistic groups are living in Djibouti, Somali and Afar, as well as an Arab minority on the northern coast and in Djibouti city. Each community speaks its language: Somali (afsoomaali) and Afar (qafaraf [ʕafaraf]) (both Cushitic), and a speciic Arabic variety. French was the only oficial language of the former colony. Presently French and Standard Arabic are the two oficial languages of Djibouti. Both are taught in public schools, but French is the only medium of instruction (MoI). Somali and Afar are oficially listed as national languages but not taught, while Djiboutian Arabic is considered a regional variety. Starting from the colonization until the 2000s, studies on Djiboutian languages were limited to Somali and Afar (Morin 2012: 49–55; Nilsson 2016). Spoken Arabic and French were hardly considered. The impact of linguistic contact and the emergence of new varieties are not yet the object of linguistic research. Until the early 21st century, Djibouti had no oficial institution for linguistic research, which was carried out by few local scholars, and by scientists belonging to European institutions.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

74

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Just before independence, research on the Afar language began through cultural associations working for the promotion of literacy such as the Union pour le développement culturel (UDC), founded in 1973. In 1976, UDC adopted the Latin alphabet developed by Dimis (Hamad b. ʕAbdallah b. Hamad) and Reedo (Gamaladdin Abdoulkadir, then director of the Afar Language Studies and Enrichment Center [ALSEC] in Samara, Ethiopia), which enabled writing in Afar and fostered collaboration between Djibouti and Ethiopia. Subsequently, glossaries, grammar sketches, and primers were prepared for educational purposes for terminology, pedagogy, and lexicography. In 1978, the Institut Supérieur d’Etudes et de Recherches Scientiiques et Techniques (ISERST) was founded. It included a Human Sciences section, where Didier Morin (1947–2017) launched linguistic research on Afar, Somali, Saho, and Beja between 1978 and 1982. In 2002, the Institut des langues de Djibouti (ILD) was created within the Centre d’Etudes et de Recherches de Djibouti (CERD), the successor of ISERST. ILD aims at contributing to a better knowledge of the national linguistic heritage by collecting and studying corpora of linguistic and traditional literature, and developing and publishing dictionaries. It organizes national and international meetings and collaborates with several institutions in France (CNRS, LLACAN, INALCO), in Ethiopia (ALSEC and the Afar Language Academy in Samara, and the University of Jigjiga), in Somalia (the Somali Language Academy in Mogadishu), and in Somaliland (Borama University). In 2004 an agreement of cooperation was signed between ILD and LLACAN. Currently, the Afar Section at ILD (Hassan Kamil) and LLACAN (Simeone-Senelle) are carrying out a collaborative project for a dialectal dictionary of halieutic vocabulary and studying the Arabic and French inluence on Afar spoken at the coast. The Institut des Sciences et des Nouvelles Technologies de Djibouti (ISNTD), established in 2001, is working on automatic natural language processing, that is, machine-aided translation and speech recognition, which so far concerns only the Somali language. The Centre de recherches de l’Université de Djibouti (CRUD) opened a linguistics section for the national languages, Somali, Afar, and spoken Arabic (vernacular and lingua franca) in addition to French and Standard Arabic. CRUD publishes the Revue Universitaire de Djibouti (RUD) and organizes interdisciplinary scientiic days each year. The new Institut de Recherche Indépendant de la Corne de l’Afrique (IRICA) was founded in Djibouti in June 2016 ‘to promote indigenous knowledge, provide community for researchers, and conduct applied research in the Horn of Africa’ (www.irica-dj.com). It is a multidisciplinary institute embracing natural sciences, earth and life sciences, and human sciences. It publishes the monthly online bilingual bulletin Recherche-Newsletter. To date linguistic research has been concerned exclusively with the teaching of mother tongues and the use of English by students in Djibouti.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

4.1.2

75

Eritrea (Wolbert G. C. Smidt)

In the 19th century, explorers and missionaries published the irst linguistic studies on Eritrean languages. Due to the accessibility of the Egyptian port Massawa, linguists also visited the region, the most important one being Reinisch in the 1870s. After the Italian Colonia Eritrea was created in 1890, research was mainly carried out by Italians, but also continued by Swedish missionaries. During World War II a new Eritrean press emerged, with important Eritrean personalities creating a modernized Tigrinya, followed by numerous linguistic studies. After independence in 1993, the Ministry of Education initiated a new period of research, as all languages of Eritrea acquired the status of administrative languages and were to be used for teaching. According to oficial accounts, nine languages are spoken in Eritrea, namely Tigrinya, Tigre, Kunama, Nara, Bilen, Saho, Afar, To-Bedawi (Beja), and Arabic. This chapter focuses on Tigrinya, also spoken in the regional state of Tigray in Ethiopia, and the other Eritrean languages (Afar is treated under Djibouti). From the 19th century onwards, Tigrinya started to become a written language. The Protestant missionary Isenberg commissioned the Adwa priest däbtära Matewos in Tigray to translate the Bible into Tigrinya, which later served as material for the studies of Praetorius (1871). The irst important Tigrinya text collection was the one by Kolmodin, documenting oral traditions of Ḥamasen. These texts belong to the inest examples of early Tigrinya literature (Kolmodin 1912, 1914). Likewise, the Catholic mission was active in the creation of literature and dictionaries (Francesco [da Bassano] 1918; Tigray: Coulbeaux & Schreiber 1915). The St. Chrischona Pilgrim Mission in Basel published Tigrinya texts and linguistic material prepared by Krapf. The French-trained Munzinger published the irst detailed word list of Tigre (1865), soon followed by von Beurmann (1868). Reinisch worked in Massawa, where he recorded numerous oral texts from Dankal in Afar; he also worked on Bilen, Saho, and Kunama. During the colonial period, dictionaries were produced (Perini 1893; Camperio 1894), as were publications in local languages (such as Rodén 1913). The Semitist Conti Rossini published documentations of texts, such as Tigrinya songs and poems, and linguistic studies. The most important publication of Tigre texts is Littmann’s volumes on his Princeton Expedition (1910–1915), on which basis Höfner inalized his Tigre dictionary (1962). After the colony went under British (later UN) administration in 1941, Eritrean writers emerged. The most remarkable one was Wäldä-Ab WäldäMaryam, a proliic Tigrinya writer who marked the emergence of a modernized Tigrinya based on the variant of Ḥamasen. He was also a member of the British Waʿəla qwanqwa Təgrəñña (Tigrinya Language Council), established in 1944 after Tigrinya became the administrative language. The grammarian abba Yaʿəqob Gäbrä Iyäsus, a Catholic priest (his grammar saw three editions, 1931–1948), became its president (Ghirmai Negash 1999). Tigrinya newspapers, such as Eritrean Weekly News, documented the discussions.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

76

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Several members retreated and developed their own style. This phase led to new Tigrinya grammars, dictionaries, and text collections, including the ‘Epic, Tales and Proverbs of the Elders’, which document traditional parlance (Yaʿəqob Gäbrä-Iyäsus 1948–1949). Especially Semitists, including Leslau, Ullendorff, Kane, Voigt, and Yaqob Beyyene, continued publishing Tigrinya texts and linguistic studies, but also studies on Tigre (Raz 1983). A quick rise in linguistic studies was observed after Eritrean independence, and after Tigray had become a regional state within Ethiopia. The richest Tigrinya dictionary is the one by Kane (2000). Eritrean linguists started to publish important and voluminous works on Tigrinya and increasingly on other languages (Musa Aron 2005; Tesfay Tewelde 2002). There are several recent academic dissertations on Tigrinya by Eritreans and Tigrayans. Linguists working for the Eritrean Ministry of Education documented dialects, identifying those suitable for education. Saho was studied within the Atlas of the Traditional Material Culture of the Saho project, especially by Vergari and Banti; Morin published studies on Afar in Eritrea; Saleh Mahmud worked on Tigre, Dawd Abushush on Nara and John Ashkaba on Kunama. Some dialects were identiied as separate languages (like Illiit and Sokodas within the Kunama language cluster; Dahlak within the Tigre cluster). Täkkəʾe Täsfay (1999) is the most important monolingual Tigrinya dictionary; he also produced a Tigrinya-English dictionary (2012). Increasing tourism and migration of Eritreans also led to the creation of Tigrinya dictionaries and language guides in other languages (e.g., in German: Zemicael Tecle 2012). Tigrinya literature and studies have also lourished in Tigray (Danəʾel Täk _ lu Rädda 2003–2004). A language conference in Mäqälä was organized by the Mah.bär bahli Təgray (Tigray Culture Association). In Tigray, the work of Kasa Gäbrä-Həywät and Amanuʾel Gankin (2007–2008) is the most remarkable dictionary, a further attempt to create a ‘high Tigrinya’, oriented at the central Tigrayan variant; they also produced a Tigrinya grammar (2003–2004). From the early 1990s, the Department of General Education at the Ministry of Education furthered linguistic research, as did the University of Asmara. In Tigray in Ethiopia, new universities established language departments, where local studies on dialectology were irst carried out. Studies of Eritrean languages appear in linguistic, philological, orientalist, and Ethiopianist journals, such as Aethiopica, Ethnorêma, Rassegna di Studi Etiopici, Africa, and the Journal of Eritrean Studies of the University of Asmara.

4.1.3

Ethiopia (Ronny Meyer)

One of the earliest linguistic activities by native Ethiopians was the creation of scripts, notably the Ethiopic script (or fidel) for writing Geez in about 300 CE, which was later modiied for other Ethiopian languages. About a

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

77

thousand years later, Muslim Ethiopians started to use Arabic Ajam scripts. Probably in the 17th century, the säwasǝw was developed as an indigenous grammatical tradition to preserve the knowledge of Geez, which ceased to be a spoken language in the second millennium CE but was retained as language of liturgy and as literary language. The teaching of Geez to new generations of church students resulted in a local grammatical tradition. The language of instruction in these church schools was and still is Amharic, the most promoted Ethiopian language (and pan-Ethiopian lingua franca) since the 20th century. The imperial government established an institution for developing the Amharic language in 1942, which eventually became the National Amharic Language Academy in 1972. It then changed name to the Academy of Ethiopian Languages under the Socialist Derg government (1974–1991), and inally became the Ethiopian Languages Research Center (ELRC) at Addis Ababa University, later renamed the Academy of Ethiopian Languages and Cultures. As an institution exclusively dedicated to research, it cooperates with local authorities towards the standardization of all Ethiopian languages. At the University College of Addis Ababa  – renamed as Haile Selassie I University in 1962 and as Addis Ababa University (AAU) in 1975 – Amharic and Geez courses were introduced in 1953–1954 through the efforts of Wolf Leslau (1906–2006) and then continued by Ethiopian scholars, like Abraham Demoz (1935–1994) and Hailu Fulass. In addition to the Department of Ethiopian Languages and Literature (DELL), a separate Department of Linguistics (DLING) was established at the Institute of Language Studies (ILS) in 1971 per the Ethiopian calendar (1978/1979 Gregorian calendar). The main concern in the DELL was teaching Amharic and its oral and written literature, while DLING did research on all Ethiopian languages, mainly from a sociolinguistic or a theoretical linguistic perspective. Between 2000 and 2012, the DELL split up into the departments of Amharic, Oromo, and Tigrinya. Several new programmes were established for Applied and General Linguistics (2004), Ethiopian Sign Language (2008), Documentary Linguistics, Experimental Phonetics and Philology (2007), and Computational Linguistics (with the School of Information Science). From the 1990s until approximately 2013, the numbers of both students and staff at DLING increased remarkably. Today, DLING has about 30 permanent staff members, almost all Ethiopians.1 Currently, DLING is involved in two large international collaborative research projects with the University of Oslo, sponsored by NORAD. The Linguistic Capacity Building project (2013–2019)2 focuses on the promotion of disadvantaged Ethiopian languages, while the Beyond Access project (2016–2020)3 is concerned with

1

See www.aau.edu.et/chls/academics/department-of-linguistics/ (accessed October 7, 2017).

2

See www.hf.uio.no/iln/om/organisasjon/tekstlab/prosjekter/Ethiopia/index.html (accessed October 7, 2017).

3

See www.aau.edu.et/chls/academics/department-of-linguistics/norad-beyond-access/ (accessed October 7, 2017).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

78

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

the improvement of mother tongue education in Ethiopia, South Sudan, and Sudan. Since the 1990s, 37 new universities have been built all over the country. A few of them started to offer linguistics (Mekelle, Wollo, Hawassa). Some regional universities established research centres for the languages spoken in their vicinity, like the Oromo Research Center (inaugurated in 2012 at the Adama Science and Technology University, since 2016 at Arsi University in Asella) and the Institute of Amharic Language Development (established in 2012 at Bahr Dar University). Other research institutions concerned with the study of Ethiopian languages are the Institute of Ethiopian Studies (IES), the Summer Institute of Linguistics (SIL), and the Centre Français des Études Éthiopiennes (CFEE). Research on Ethiopian languages is regularly disseminated through international and local journals, of which the following are published in Ethiopia: Journal of Ethiopian Studies, Ethiopian Journal of Languages and Literature, Zena Lissan: Journal of the Academy of Ethiopian Languages and Cultures, and ELRC Working Papers. The International Conference of Ethiopian Studies is a regular interdisciplinary forum for Ethiopian studies in general. The irst conference took place in Rome in 1959, followed by 18 other conferences, conducted in Europe, Israel, Japan, the United States, and Ethiopia. In the 1980s, an annual conference of the Institute of Language Studies (since 2012 under the auspices of the College of Humanities, Language Studies, Journalism and Communication) was established. Furthermore, Ethiopian linguistics is represented at regular meetings, like the International Conference of Cushitic and Omotic Languages, the Nilo-Saharan Linguistics Conference, the North Atlantic Conference on Afroasiatic Linguistics, and the World Congress of African Linguistics. Currently, the typological-descriptive approach predominates in Ethiopian linguistics. Many of the more than 100 in-house PhD dissertations at AAU have followed such a framework. Among Ethiopian linguists, Baye Yimam is the most inluential proponent of the formal approach. His generative grammar of (and in) Amharic is so famous that its concepts became the main component of Amharic grammar lessons at school. Applied linguistics, language planning, and standardization are central research issues. The introduction of mother tongue education in Ethiopian vernaculars presupposed a certain level of status and corpus planning, which typically was done by regional language experts, often supported by the Department of Linguistics or the Academy of Ethiopian Languages and Cultures. As a result, the amount of written material in vernacular languages increased substantially, which also includes their linguistic description by native speakers. During the last few years, several new research areas appeared, such as sign language, experimental linguistics, linguistic landscape, computer linguistics, and corpus linguistics. Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

4.1.4

79

Somalia (Ronny Meyer)

From 1880 to 1900, the Somali-speaking area was divided and colonized by Great Britain, Italy, and France; Ethiopia annexed the Ogaden. The former British and Italian colonies merged into independent Somalia in 1960; the French colony became independent Djibouti in 1977. Although Somali was spoken by more than 90 percent of the population, the choice of an oficial language became a political issue, as there was no consensus on a uniform Somali script. Consequently, the former colonial languages English and Italian as well as Arabic became oficial languages. Eventually, the military government under Siad Barre declared Somali the oficial language in 1973. The Somali language was standardized in a Roman-based script in 1972, although Cusmaan Yuusuf Keenadiid created an indigenous script (a phonetic syllabary), the Osmania, in the 1920s, which was the most frequently used script by native authors after World War II. Somali became the sole language in administration and pre-university education and was used in written mass media. At the university level, it was the MoI in the Department of Somali Language and Literature. The standardization of Somali was fostered through the then Academy of Culture and the National University of Somalia in Mogadishu. Linguistic research on Somali started with the work of Leo Reinisch (1832–1919), who  – through original ieldwork with native speakers  – published the irst comprehensive description of Somali in 1900–1904. Linguists who have substantially contributed to the study of Somali include Bogumil W. Andrzejewski (1922–1994), Giorgio Banti, Mara Frascarelli, Jacqueline Lecarme, Martino M. Moreno, John Ibrahim Saeed, and Mauro Tosco. Besides lexicography and the preparation of various bi- and monolingual dictionaries in Somali, central topics became the description and classiication of Somali dialects, and research in focus and discourse conigurationality, beside a variety of studies on morphology, syntax, and comparative studies. In sociolinguistics, studies on the standardization of Somali and on Somali scripts dominate. The Regional Somali Language Academy, founded in 2013 as a panSomali institution with its base in Djibouti, is concerned with the study of the Somali language and culture in Somalia, Djibouti, and Ethiopia. Somali linguistics constitutes part of the broader ield of Somali studies, with the Somali Studies International Association as its main body. The International Congress of Somali Studies was inaugurated in Mogadishu in 1980. Founded in 2014, the Hargeysa Cultural Center, with Jama Musse Jama as director, is an important institution for the study of the Somali language, which is supported by the Redsea Cultural Foundation.

4.1.5

Sudan and South Sudan (Angelika Jakobi)

The linguistic landscape of Sudan is characterized by at least three features: (a) the written traces of two indigenous ancient languages, Meroitic Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

80

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

and Old Nubian; (b) the unequal status of the Sudanese African languages on the one hand and of Arabic on the other; and (c) the linguistic diversity, as relected by some 130 languages representing three language phyla, Nilo-Saharan, Niger-Congo, and Afroasiatic. The most bewildering array of languages is found in the Nuba Mountains, in South Sudan and Darfur, and along the Ethiopian border. Unlike the 18th- and 19th-century travellers and explorers who usually collected more or less brief vocabularies from speakers they had coincidentally met, the Christian missionaries, who had entered Sudan in the mid19th century, aimed at preparing Bible translations and therefore tried to gain deeper insights into the grammatical structure of some languages. The irst grammars produced by missionaries were devoted to Dinka and Bari. In the second half of the 19th century linguistic research became more professional, either focusing on language description or comparison. The irst Sudanese languages that were compared and whose genetic relationship was recognized were Bari, Dinka, Nuer, and Shilluk (Müller 1876– 1888). The irst Nile Nubian grammar was published by Reinisch in 1879. Richard Lepsius (1880) noticed the relationship between the Nubian languages in the Nile valley and the Nubian languages of Kordofan. Another highlight of that period was Reinisch’s (1893–1894) four-volume grammar of Beja (‘Bed.auye’). In 1928 at Rejaf, the Anglo-British colonial administration organized a meeting of government representatives and missionaries who were engaged in educational work. Representatives of the Sudanese peoples were not invited. One of the outcomes of the Rejaf Conference was to adopt a restrictive language policy in southern Sudan. It differed sharply from the policy in northern Sudan where Arabic was promoted, while the use of the local African languages was strongly discouraged. Another objective of the conference was to adopt the International Phonetic Alphabet (IPA) for writing those southern Sudanese languages that were selected to be developed for elementary education. Due to the application of the IPA and Diedrich Westermann’s Shilluk grammar (1911) that served as a model, many missionaries were able to considerably improve their grammatical descriptions. On behalf of the International Africa Institute in London, Archibald N. Tucker and Margaret A. Bryan carried out a typological survey of The Non-Bantu Languages of North-Eastern Africa (1956). Other researchers contributed to this survey too. One of the most important scholars was Roland C. Stevenson due to his research on several Nuba Mountain languages (Stevenson 1956–1957). The major outcome of Tucker and Bryan’s work was two volumes published in 1956 and 1966. The latter – Linguistic Analyses – has become a standard reference work since it comprises grammar sketches of 33 language groups, some of which are still little known today.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

81

In the 1960s academics at Khartoum University developed an interest in the local spoken and ancient written languages. This interest was irst relected by the founding of the Sudan Research Unit, which was  later transformed into the Institute of African and Asian Studies (IAAS). This institution established several publication series; one of them, the Linguistic Monograph Series, made available several hitherto unpublished language studies. In the 1970s the IAAS initiated the Language Survey of the Sudan, a sociolinguistic project that was concerned with the use and social status of Arabic and the local African languages in selected communities. This project has continued to stimulate sociolinguistic inquiries, as shown in many Sudanese master’s and doctorate theses. Since the 1980s linguistic research in Sudan has received various impulses through series of publications, regular conferences, and research projects. The Occasional Papers in the Study of Sudanese Languages series, for instance, was established by the Summer Institute of Linguistics (SIL). Further research was stimulated through the triennial Nilo-Saharan Linguistics Colloquium and the publication series Nilo-Saharan – Linguistic Analyses and Documentation initiated by Franz Rottland and M. Lionel Bender. In 2011, another triennial conference series, the Nuba Mountain Languages Conference, was started by Thilo Schadeberg. The most recent series is Dotawo, an open-access journal of Nubian studies, edited by Vincent van Gerven Oei and Giovanni Rufini. It is also worth noting that Abeer Bashir, Suzan Alamin, Helene Fatima Idris, Maha Aldawi, Sawsan Abdelaziz, and Mona Hashim of the Department of Linguistics at Khartoum University engage in language documentation projects and support groups of speakers who want to develop orthographies for their languages. Research on both of the ancient written languages, Meroitic and Old Nubian, started with the groundbreaking works of the British Egyptologist Francis L. Grifith at the beginning of the 20th century. While Meroitic is currently studied by very few scholars, mainly Kirsty Rowan and Claude Rilly, Old Nubian has attracted a group of researchers including Vincent van Gerven Oei, Vincent Laisney, Adam Łajtar, Grzegorz Ochała, Giovanni Rufini, Helmut Satzinger, Alexandros Tsakos, Kerstin Weber, and Petra Weschenfelder.

4.2 4.2.1

African Linguistics in Eastern Africa (Amani Lusekelo) Introduction

The history of African linguistics in the Eastern African region surrounds the national languages, namely Bemba, Chewa (Nyanja), Ganda, Kinyarwanda, Shona, and Swahili. Historically, a few regionally dominant languages such as Haya, Luo, Ndebele, and Tumbuka are also important.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

82

PA RT I

4.2.2

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

The Contributions of Missionaries and Colonial Administrators (1840–1940)

The period of pre-scientiic studies in German and British East Africa begins with the spread of Swahili literacy, which is the result of the trade controlled by the coastal people from Zanzibar and Mombasa. In fact, the period between 1500 and 1650 marked the dominance of Arab traders, who spread both Swahili and Islamic civilization into the interior of Kenya, Tanzania, and Uganda. During the Portuguese invasion between 1650 and 1730, most of the old Swahili texts written in Arabic script were destroyed. The existing documents are presumably stored at the University of Dar es Salaam. The oldest Swahili texts in Arabic are letters stored at Goa (dated 1711–1728) and tendi ‘poems’ from Mombasa and Zanzibar (dated 1652–1728). After the 1750s, many famous poets emerged in Lamu, Mombasa, and Zanzibar, but particularly Mohammed Kijumwa from Lamu, Muyaka Haji al-Ghassaniy from Mombasa, and Fumo Liyongo from Zanzibar deserve mention. In German and British East Africa, the history of linguistics links up with the devotion of two missionary societies that participated in the writing of grammars, the compilation of dictionaries, and the translation of Christian texts. First, the Church Missionary Society, under the auspices of the University’s Mission to Central Africa, had the earliest inluence on the emergence of Eastern African linguistics through the Anglican and Lutheran churches of Mombasa and Zanzibar. Second, the Society of Missionaries of Africa (White Fathers) of the Catholic Church engaged later in the development of local languages in Eastern Africa. The scholarly linguistic works in Swahili began in the 1840s with the publications by the German priest Ludwig Krapf (CMS), who compiled a dictionary between 1844 and 1848 in Mombasa, Kenya, and a Swahili grammar book, published 1882 and 1850 respectively. From Zanzibar in Tanzania, the major contributions by the missionaries were vested in grammar books by the priest Edward Steere (CMS), and dictionaries and a grammar book by the priest Charles Sacleux of the White Fathers. The production in written form of oral works produced by Africans began with missionaries such as the priest William Ernest Taylor, who collected Swahili poetry such as Utenzi wa Fumo Liyongo; he published groundwork for Swahili in the 1890s. In the Buganda Kingdom, Ganda and Swahili were adopted as the languages of literacy and trade, mainly with the support of kabaka Mutasa (1856–1884). Priests Alexander Murdock Mackay (CMS) and George Lawrence Pilkington (CMS) translated biblical stories and religious hymns into Ganda. Indigene church clergymen contributed to the emergence of indigene scholars in German and British East Africa (1890s–1930s). In Uganda, Apollo Kaggwa (1864–1927) was converted to Christianity by CMS in 1886; he obtained literacy classes in Arabic, Ganda, and Swahili. Later, he devoted himself to the production of Christian materials in the Ganda language

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

83

for the kingdom of Buganda. His publications in the Ganda language contain important historical information. Another important person in the development of Ugandan languages was the Anglican clergyman Reverend Henry Wright Kitakule Duta (CMS). He translated sections of the Old and New Testaments into Ganda and documented the customs of the Baganda as expressed in the Ganda language. In Rwanda, Father Alexis Kagame (1912–1981) of the Roman Catholic Church promoted the Kinyarwanda language. As a graduate of Georgian University in Rome (1951–1955), Father Alexis Kagame returned home in 1958 and became a teacher at a Catholic seminary and later a professor at the University of Rwanda in 1963. Father Kagame collected numerous cultural and linguistic materials for Kinyarwanda and published important resources. In Tanzania, the Anglican Church groomed Mathias E. Mnyampala (1917–1969), who published important books about the Gogo, his ethnic group. Linguistically, Mnyampala’s publication is essential in the history of Swahili. In Northern Rhodesia (now Zambia), between 1885 and 1945 many mission stations were established in order to teach the gospel. To fulil this mission, early missionaries wrote grammars, compiled dictionaries, and translated religious texts into selected local languages. Consequently, Zambia became linguistically divided into four zones based on the predominant lingua francas: Tonga in the southern parts, Bemba in the north, Nyanja (Chewa) in the eastern parts, and Lozi in the west. During the colonial period, school materials were produced in these languages. At independence in 1964, two more languages (Kaonde and Luvale) became prominent. Consequently, for the linguistic situation in Zambia today, emphasis has shifted towards the promotion of Zambia’s seven national languages, Bemba, Nyanja, Tonga, Lozi, Lunda, Luvale, and Kaonde. In Southern Rhodesia (now Zimbabwe), the church organizations, later supported by the colonial government, developed selected languages, mainly Shona and Ndebele, for the period between the 1890s and the 1940s. In Mashonaland, the translations of the Bible and religious books into Shona were limited, based on the availability of missionaries to speak Shona. Each church translated the Bible and Christian texts into a different Shona dialect. For instance, in the southern areas, the Dutch Reformed Church and the Lutheran Church operated in the Karanga-speaking community, hence Reverend Andrew A. Louw of the Dutch Reformed Church translated the Bible into Karanga. Again, by 1898 the Methodist Church published Ivangeri ya Marako (the Gospel of Mark) in the Zezuru dialect. Between 1905 and 1908, E. H. Etheridge translated the New Testament into the Manyika dialect of Shona. In Nyasaland (now Malawi), the Scottish Missionary Society dominated the development of Malawian vernaculars. The mission opened up a printing press called Hetherwick Press in Blantyre in 1884 for the publication

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

84

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

of religious materials in the vernacular languages. Chewa/Nyanja received much attention because publications in it outnumbered those in other vernaculars. In northern parts of Malawi, the Livingstone Mission of the Church of Scotland contributed signiicantly to the growth of secular literature in Tumbuka, mainly before independence.

4.2.3

Language Institutions (1940–1970)

In German and British East Africa, the Inter-Territorial Language Committee for East African Dependencies (the Swahili Committee) existed between 1930 and 1970 with member territories being Kenya, Tanganyika, Uganda, and Zanzibar. It was established in order to standardize the language and encourage research in Swahili. It produced the dictionary of Standard Swahili by Johnson in 1935 and 1939. The Swahili Committee encouraged research on the grammar of the language. Many scholars were supported by the committee. The produced works include ones by Krumm (1940), Ashton (1944), and Whiteley (1969), among others. The oral wealth of Swahili people was compiled by Jan Knappert in important contributions in the 1970s. Indigene scholars, such as Shaaban Robert (1909–1962), were promoted by the Swahili Committee. Shaaban Robert became interested in Swahili poetry and novels. He composed a number of novels and poems that are used in the education system in Tanzania, for example, Kusadikika (1951), Kufikirika (1967), and Siku ya Watenzi Wote (1968). In Southern Rhodesia, there arose the need for a common Bible across the Shona dialects of Karanga, Korekore, Manyika, Ndau, and Nambya. The church organizations and colonial government needed a standard orthography for Shona language to be used in schools and church services across Mashonaland. Thus, the Standard Uniied Shona Orthography (SUSO) was conceived at the Southern African Missionary Conference for all church organizations working in Zimbabwe, which was held from 1903. The conference spearheaded the standardization and harmonization of the various Shona dialects. The development of SUSO was under the supervisions of Clement M. Doke (University of the Witwatersrand) and George Fortune (University of Zimbabwe). Rev. Andrew A. Louw of the Dutch Reformed Church and Father Barnes of the Catholic Church contributed much to the writing of the orthography. Beginning in 1981, both Ndebele and Shona became important languages used in colonial schools, particularly for the Zimbabwe Junior Certiicate Examination Level. In 1987, more minority languages were introduced at the lower level of education; Shona remained an important subject up to the tertiary level. Since it was successfully promoted, today Shona is a comparatively advanced language; it unites about 75 percent of Zimbabweans. In Malawi, the Chichewa Board was founded by the president Dr Hastings Kamuzu Banda (1898–1997). From 1958, Dr Banda became interested in

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

85

linguistics and the development of the dialect of Nyanja called Chewa into the national language. In 1968, Chewa was elevated to the status of national language. The Chichewa Board was established in 1972 and was declared a statutory board by 1977, with three responsibilities: (a) to develop and approve the new Chewa orthography rules, (b) to provide the national dictionary of the Chewa language, and (c) to develop the Chewa language. One of the merits of the board was the establishment of the standard orthography of the language. Also, the Chichewa Board checked the use of the language in newspapers and magazines as well as on the radio across the country (Malawi Broadcasting Cooperation) and in translations. The board managed to release the instructions for the standard orthography in 1991. A new horizon for all languages of Malawi emerged. The Chichewa Board became transformed into the Centre for Language Studies, a linguistically more inclusive academic and research institution. Also, the new language policy allowed promotion of all Malawian languages. Consequently, Malawi radio became multilingual with Chewa, English, Lomwe, Sena, Tonga, Tumbuka, and Yao.

4.2.4

Modern African Linguistics (1970–Present)

First, language in education was the primary theme of research. Since the 1970s, indigene scholars such as Abdulaziz (1972), Mbaabu (1973, 1978), and Kashoki (1978) have highlighted the importance of mother tongue education. The setbacks of English as MoI emerged in the 1980s and crystalized in the 1990s. The role of national languages (Bemba, Chewa, Ganda, Shona, and Swahili) and in relation to other native languages was also examined. Second, lexicostatistics was on the research agenda in the 1970s and 1980s. Its main target was the proper classiication of African languages through broader surveys of the lexicons. The motivation for the lexicostatistic surveys of Eastern African languages emanated from not only the shortfalls of the previous publications, but also the views of the community members. The non-Bantu languages in Eastern Africa were studied in the 1970s and 1980s by Lucia N. Omondi and Duncan O. Okombo, with a focus on Luo. The internal structure of Bantu languages formulated the third epoch. In the 1980s, the study of Bantu tones emerged in works by Massamba (1984), Batibo (1985), and Mtenje (1986). In the 1990s, Sam Mchombo from Malawi and Lioba Moshi from Tanzania then pioneered the study of the status of object nouns in Bantu languages. The application of the theory of predicate structure was examined further in relation to the functions of verb extensions and the resulting thematic roles. The next epoch emerged in the 1990s, focusing on the theme of the development and modernization of the major African languages, namely

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

86

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Bemba, Chewa, Swahili, and Shona. Scholars recorded the standardized vocabularies for the languages. Last, the promotion of endangered languages emerged in the 2000s. In Malawi, for instance, after the fall of Dr Kamuzu Banda, many minority languages were promoted. In Kenya, for instance, Rottland and Okombo (1992) discuss the situation of Suba society in relation to the inluence of neighbouring Bantu and lexiication by Luo in Kenya. In the Suba area, four language layers exist in this order, Suba, Swahili, Luo, and English, a situation that permits changing attitudes in favour of major languages such as Luo and Swahili. In Tanzania, scholars, for instance Josephat Rugemalira are engaged in language documentation through the writing of grammar books and dictionaries. Although the prominent igures and their research outputs have been highlighted already, three waves of indigenous scholars exist. In the 1960s, the famous indigene scholars included Mubanga E. Kashoki (University of Zambia), Mohamed H. M. Abdulaziz (University of Nairobi), and Francis X. Katamba (Lancaster University). These early scholars wrote about language in education (e.g., Abdulaziz 1972), lexicostatistical surveys (e.g., Kashoki 1978), the augment in Bantu (e.g., Hyman & Katamba 1993), and so on. The second wave of indigenous linguists emerged in the 1970s and 1980s. These included Lucia N. Omondi (University of Nairobi), Herman M. Batibo (University of Botswana), Sam A. Mchombo (University of California, Berkeley), David P. B. Massamba (University of Dar es Salaam), and Al Mtenje (University of Malawi). Numerous topics are dealt with by these researchers, such as the syntax of Luo, sound patterns in Swahili, lexical borrowing and historical linguistics, object marking in Bantu languages, the history of Swahili, tone shift due to reduplication, and others. The third wave of eminent scholars represents indigenous linguists who emerged in the 1990s. Perhaps the Zambian scholar Felix Banda (University of Western Cape) is representative. His research interests surround multilingualism and the MoI controversy, language use and identity in Africa, language and culture in urban Africa, and literacy in Africa.

4.3 African Linguistics in West Africa (Bruce Connell & Akinbiyi Akinlabi) 4.3.1

Introduction

The earliest known accounts of a West African language appear in the records of Portuguese explorers in 1523, in the form of vocabulary apparently from a precursor to present-day Akan (Kwa; Ghana). For the next two centuries little more was done outside the collection and publishing of brief word lists of coastal languages encountered by Europeans. The irst major developments in West African language study came beginning in the early years of the 19th century.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

87

Little has been written on the history of linguistics in West Africa. D. T. Cole’s ‘The History of African Linguistics to 1945’ (1971) is considered one of the main sources for the early history of linguistics in Africa, but his account is not particularly strong in its discussion of West Africa. Cole himself acknowledges knowing ‘Bantu linguistics, with no pretensions to knowledge of other areas, nor to being a historian’. P. E. H. Hair’s (1967) The Early Study of Nigerian Languages is an excellent study, though it is restricted in both its temporal and geographical scope, covering only up until about 1890 and with a focus on the work of the Church Missionary Society (CMS). Other work, inspired or initiated by other people and organizations, is omitted from his discussion. Since at the time much of the work carried out by the CMS was based in Freetown, Sierra Leone, information is also available in Hair’s account about the early period of language studies there. Information about the modern era can be found in works by Bamgbose (1994, 2007, 2011) and Emenanjo (2007), though not primarily intended as historical studies. Other sources, such as Yankah et al. (2012), the introduction to which discusses the growth of linguistics at the University of Ghana, and the occasional brief histories of linguistics departments found on departmental websites for some universities in the region, are also available. (For full references to these sources, see Connell and Akinlabi 2019.) This brief sketch of the history of linguistics in oficially Englishspeaking West Africa is an attempt to ill some of the gaps in our collective knowledge of how linguistics developed in the region. While space constraints permit discussion of the contributions of only very few individual scholars to the development of linguistics in the region, we note that from the beginning contributions have come from both African and nonAfrican scholars. Many of the most inluential African linguists received their training abroad, while other non-African linguists spent suficiently long periods of their careers in Africa as working linguists and training the early generation of modern African linguists that their contributions must be acknowledged. Elaboration of the discussion offered here can be found in Connell and Akinlabi (2019).

4.3.2

The Church Missionary Society, Samuel Ajayi Crowther, and James Schön

The British colony of Freetown was founded in 1792 with a settlement of over 1,100 freed slaves from the United States via Nova Scotia. The West Africa mission of the CMS was established in what is now Guinea (Conakry) in 1804 and in Freetown in 1816. Many of its missionaries to West Africa came from the Evangelical Lutheran Church in Württemberg, Germany, and were trained in seminaries in both Germany and Britain. These included James Schön, who was instrumental in establishing the mission in Freetown. Schön was also a linguist and later produced both

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

88

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

a grammar and dictionary of Hausa (1862, 1876); he also worked closely with Samuel Ajayi Crowther, the irst Yoruba linguist. Crowther (ca. 1809–1891) was born in Osogun, in what is now Oyo State in Nigeria. He was taken as a slave at the age of 12 in 1821; the ship on which he had been enslaved was liberated by the British, and he was taken to Freetown. Here Crowther met the CMS missionaries Kilham and Raban as well as Schön. In 1826 Crowther was sent for a year to London, where he studied at St. Mary’s Church School, studies that reportedly included Latin and Greek. On his return to Freetown he began working with Kilham and Raban. Hair (1967) reports that Crowther attempted to learn Temne and speculates that it was this experience, rather than helping Raban as his informant, that sparked his interest in studying his own language, Yoruba. In 1841 he was sent, along with Schön, on the Niger Expedition. Crowther’s selection for the mission was based in part on his being a Yoruba speaker, though since most of the country covered was not Yoruba speaking, it also involved his learning Hausa (Chadic; Nigeria); it was assumed that Hausa would be used in the interior. This proved not to be the case, and on the expedition Crowther was exposed to both Nupe (Benue-Congo; Nigeria) and Igbo (Benue-Congo; Nigeria), among other languages. As for the broader, evangelizing goals of the Niger mission, it is fair to say that while not all were realized, it was successful in studying the distribution of languages along the Niger River, up to and around the conluence with the Benue. In 1843 Crowther’s irst major work on Yoruba appeared, A Vocabulary of the Yoruba Language: Part I English and Yoruba; Part II Yoruba and English; to Which Are Prefixed Grammatical Elements of the Yoruba Language. A revised edition of the vocabulary appeared in 1852, with an expanded grammar published separately the same year. In 1864 he published A Grammar and Vocabulary of the Nupe Language, and in 1882 A Vocabulary of the Ibo Language. Crowther’s work also included many primers and numerous translations of parts of the Bible into these three languages. It is dificult to say to what extent Schön inluenced or inspired Crowther, or vice versa; Hair notes though that they remained friends and collaborators well into old age and further suggests Schön may have assisted Crowther in a number of ways in his work, ranging from linguistic points to proofreading. It may be noted that Crowther was not the only indigenous linguist or language scholar in Freetown at the time, though he was clearly the most important. Others, such as John Christopher Taylor, were engaged in writing primers and translating portions of the Bible. Taylor himself was Igbo, though born in Sierra Leone, and collaborated with Crowther on several occasions, including joining him on the Niger Expedition of 1857. For the most part these others were concerned not so much with the linguistic study of African languages as with developing a literature for their language. Crowther too was involved in these efforts, and he is frequently referred to as the ‘father of Yoruba literature’. He can also fairly

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

89

be honoured as the irst indigenous West African linguist. Crowther died in 1891 in Lagos. As for Schön, beyond collaborating with Crowther, he is also known for his work on Hausa; he produced both a grammar (1862) and a dictionary (1876) as well as other materials. Schön’s was not the only work carried out on languages situated to the north of the Niger and Benue rivers; work was carried out on Kanuri, especially by Sigismund Koelle, though Koelle was based in Freetown, where he completed his monumental opus, the Polyglotta Africana (1854). The work carried out by the CMS was not the only linguistic work done during the 19th century in this part of West Africa. East of the Niger, in the Cross River region, similar work was carried out by missionaries of the United Presbyterian Church of Scotland. The language that beneited most from these efforts was Eik (Cross River; Nigeria), which for a time rivalled, if not outshone, other West African languages with regard to development and description. Early important work on Eik includes, by Hope Masterson Waddell, A Vocabulary of the Efik or Old Calabar Language, with Prayers and Lessons (1852, revised in 1859), and, by Rev. Hugh Goldie, Principles of Efik Grammar (1857) and a Dictionary of the Efik Language (in Two Parts) (1862). Although there appears not to have been an indigenous linguist approaching the stature of Crowther, Waddell acknowledges King Eyo Honesty, Egbo Young, and Young Eyo Honesty as ‘the best native authorities’. At the same time, Johann Gottlieb Christaller of the Basel Evangelical Missionary Society was working in Ghana, with mission-educated David Asante and Theophilus Opoku, among other African scholars. Christaller published, in 1875, his A Grammar of the Asante and Fante Language Called Tshi (Chwee, Twi), followed in 1881 by his Dictionary of the Asante and Fante Language Called Twi. And earlier, English-Tshi-Akra Dictionary (1874) and A Collection of 3,600 Tshi Proverbs (1879) were also published. Both the grammar and the dictionary remain useful to this day. One of the most notable and striking aspects of his work from the perspective of linguistic scholarship was the implicit recognition of the tonal phenomenon ‘downstep’, though it was not until much later, with the work of Dennis Winston on Eik (1960) in particular, that downstep was more fully understood.

4.3.3

The Colonial Period

From the late 19th century up until the early 1960s, when the British colonies in West Africa gained their independence, linguistic research and the study of the languages of the region continued, though there is little evidence to suggest much in the way of contribution from African scholars. Here just brief mention is made of two notable Western scholars, Ida C. Ward and Diedrich Westermann. Ward’s The Phonetic and Tonal Structure of Efik (1933) began to rectify the weakness of the early work on

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

90

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Eik by Goldie and Waddell, who seemed unaware of tone. Ward’s analysis refers to a mid tone in Eik, along with high and low tones, though she recognized a relationship between high and mid tones (perhaps inluenced by Christaller’s work on Fante; see above). Ward also contributed numerous publications to journals such as the Bulletin of the School of Oriental and African Studies and, in 1952, published An Introduction to the Yoruba Language. The importance of Diedrich Westermann and his contributions to African linguistics cannot be overstated. His work spans several geographical and political regions of Africa and covers descriptive grammar (e.g., of Ewe in 1907) and phonetics (together with Ward he wrote Practical Phonetics for Students of African Languages, irst published in 1930), and he also established the foundations for the classiication of African languages (1911).

4.3.4

The Expansion of Modern Linguistics in the Postcolonial Period

Independence for Britain’s colonies in West Africa around 1960 ushered in a period we can conveniently refer to as ‘modern linguistics’ in West Africa, even though there was little immediate change in direction from the preceding few decades. There was however greatly increased incentive for linguistic research, partly due to the interests of new African governments and partly due to the contributions of the Ford Foundation, a private organization based in the United States that had the goal of advancing human welfare. The West African Languages Survey was established with a ive-year grant from the Ford Foundation. The purpose of the grant was twofold: the endowment of fellowships and grants for research on languages of the region and the organization of an annual West African Languages Congress. A fellowship programme sponsored research on speciic languages, some of which was published in the West African Language Monographs series. Several of these monographs took the form of descriptive grammars, based in some cases on modern linguistic theory, while other volumes included a phonetic survey by Ladefoged (1964), a dictionary of Grebo by Innes (1967), Crabb’s comparative study of the Ekoid languages (1969), and Hair’s treatise (1967), referred to above. The importance of the survey is seen not only in the linguistic research it supported and carried out during the ive years it was active, but in three other respects: its support for the training of African linguists, both as researchers and teachers; the emergence of departments of linguistics; and the founding of the West African Linguistic Society, which built on the activities previously carried out by the survey. When the survey was initiated there was no West African linguist with a doctorate in linguistics; by the end of its ive-year term, there were at least eight PhD linguists. When the survey was initiated there was no West African university with a linguistics department; by the time it evolved into WALS/SLAO in 1965,

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

91

departments had been established in Dakar (Senegal), Abidjan (Côte d’Ivoire), Accra (Ghana), Ibadan (Nigeria), and Yaoundé (Cameroon). The inaugural West African Languages Congress was held at the University of Ghana, Legon, in 1961. The congresses have continued under the auspices of the West African Linguistic Society / Société linguistique d’Afrique oriental (WALS/SLAO), since 1970 on a biennial basis. Host institutions alternate between anglophone and francophone countries. WALS/SLAO is itself by far the most important legacy of the West African Languages Survey. It was instituted at the end of the mandate of the survey, at the Fifth West African Languages Congress, held at the University of Ghana in 1965. Its main objectives were to foster linguistic research in the region and to establish an ongoing forum for the exchange of ideas among African language scholars in both anglophone and francophone areas of West Africa. It established working groups for areas of special interest, such as different language families of the region (e.g., Benue-Congo, Kwa, and Chadic), oral literature, and sociolinguistic problems. The impetus of the society resulted in the accelerated development of linguistics and linguistic studies, cooperation across the Francophone and Anglophone divide, and the establishment and publication of the Journal of West African Languages (JWAL), which irst appeared in 1964. The monograph series mentioned above, as well as the two-volume books Benue-Congo Word List and West African Language Data Sheets were also important early initiatives of the society.

4.3.5

Linguistics as an Academic Discipline

Modern linguistics began in both Ghana and Nigeria in a parallel, though by no means identical, fashion. In both countries, the ‘lagship’ university (the University of the Gold Coast, which became the University of Ghana, Legon, and the University of Ibadan in Nigeria) was afiliated with the University of London. Both had Phonetics Departments aimed at assisting students to achieve pronunciation of English at a level consistent with that found throughout educated communities elsewhere in the British Commonwealth, and at both universities this evolved into a department of linguistics: at Ibadan, the Department of Linguistics and Nigerian Languages (in 1962); and in Ghana, the Department of Linguistics and Ghanaian Languages (in 1964). The 1970s saw a period of growth and development in Ghana, both for linguistics programmes and for the discipline on the national stage. Programmes were expanded to include a wider range of course offerings as well as postgraduate degree and diploma programmes. The period also saw the irst Ghanaian become head of a department, in the person of Lawrence Boadi, and there was an increased focus on Ghanaian languages, especially Akan, Ewe, and Ga. Programmes in linguistics were also founded at three other Ghanaian universities: the Department of Modern Languages,

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

92

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Kwame Nkrumah University of Science and Technology, Kumasi (1973); the Department of Applied Linguistics, University of Education, Winneba (2003); and the Department of Ghanaian Languages and Linguistics at the University of Cape Coast. Outside university departments and their mandates, linguistics has had both an academic and non-academic presence in Ghana. The Linguistic Circle of Accra, which later evolved into the Linguistic Association of Ghana, was founded in 1967 by linguists from the department at Legon together with others from the Institute of African Studies at the University of Ghana. Language study also continued outside the academic world (though not totally separate from it) with the Ghana Institute of Linguistics, Literacy and Bible Translation (GILLBT) involved in language development for both education and evangelization. Similarly at Ibadan, after linguistics became a full department, both course offerings and degree options expanded. A postgraduate certiicate course in Phonetics and Linguistics was established and became a postgraduate diploma in 1965, renamed a Postgraduate Diploma in Linguistics in 1969. Ayo Bamgbose, after having earned his PhD in linguistics at Edinburgh, became professor in 1968 and head of department in 1969, thus earning the distinction of being the irst West African to become head of a linguistics department. The foci of linguistics at Ibadan, and subsequently elsewhere in Nigeria, were principally descriptive and historical, an appropriate approach in a country with over 500 languages, most of which were and are undocumented or undescribed. In keeping with this, the department at Ibadan (now Linguistics and African Languages) also published a mimeographed journal, Research Notes, which irst appeared in 1967. In the mid- to late 1970s a substantial expansion of the federal university system in Nigeria was undertaken, with many of the new universities having linguistics programmes. These were typically in departments of linguistics and Nigerian (or African) languages. Among the universities with linguistics programmes were the University of Nigeria (though the university itself predated the expansion), inaugurated in 1974, a fully-ledged department by 1981; the University of Calabar (1976); the University of Benin (1979); the University of Ilorin (1979); as well as Port Harcourt, Ife, Jos, Maiduguri, Amadu Bello University in Zaria, and Bayero University in Kano (BUK). Mention must be made of the Centre for the Study of Nigerian Languages (CSNL; now Centre for Research in Nigerian Languages and Folklore) at BUK, established in the early 1970s, supporting research on Hausa and Fulfulde, but also Kanuri and minority languages of northern Nigeria. The centre has published monographs, conference proceedings, a monolingual Hausa dictionary, and Harsunan Nijeria (Nigerian Languages), a journal that includes articles written in Hausa and in English. A third wave of expansion of linguistics in Nigeria came about in the early 1990s, following the redrawing of state boundaries in Nigeria

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

93

leading to the creation of new states. Among the new programmes was the Department of Linguistics and Nigerian Languages at the University of Uyo (1992), and later Adekunle Ajasin University in Akungba (1999). In total, over 30 Nigerian universities had linguistics programmes by the irst decade of the 21st century. The growth of academic linguistics programmes in Nigeria was accompanied by the birth of the Linguistic Association of Nigeria (LAN) in 1980. Its aims were to offer membership to and ensure permanent contact among all who are engaged in the scholarly study of the languages of West Africa/ Nigeria; to initiate and support scholarly linguistic research both for scientiic purposes and to assist with the linguistic aspects of development of West Africa/Nigerian educational and social development; to organize annual/biennial meetings, conferences, congresses; to establish and publish regularly the Journal of the Linguistic Association of Nigeria (JOLAN); to offer grants for research in these areas whenever practicable and desirable; and to cooperate with or establish association with international organizations. JOLAN was established in 1982 and continues to publish. Outside academic linguistics, the National Language Centre (established by the Nigerian Federal Ministry of Education) sponsored a series manuals entitled ‘Orthographies of Nigerian Languages’. Three manuals appeared in the series in the 1980s, covering languages that included Eik, Hausa, Igbo, Yoruba, Edo, Fulfulde, Kanuri, Tiv, Nembe, Ijo, Berom, Ibibio, Nupe, Idoma, Kalabari, and Igala. The National Language Centre has since been re-constituted as a university-level institution called the National Institute for Nigerian Languages (NINLAN). The functions of NINLAN include ‘research, teaching, documentation, and coordination of studies in Nigerian languages’ (www.ninlan-aba.edu.ng/). The school also offers undergraduate degrees in education, combining linguistics with Hausa, Igbo, Yoruba (or Kanuri).

4.3.6

Other Countries

While Ghana and Nigeria form the heart of linguistics in English-speaking West Africa, other countries also fall within the geopolitical realm of ‘oficially English-speaking West Africa’; these are Sierra Leone, to which we have already devoted passing attention though not in its modern, postindependence form, the Gambia, and the anglophone region of Cameroon, all of which are former British colonies. To these can be added Liberia, which, while English-speaking, was not a British colony but rather came into being through the resettlement of liberated slaves and freeborn African Americans from the United States. Despite its auspicious beginning with Freetown as the centre of linguistics and language study in precolonial West Africa, linguistics as an academic discipline in Sierra Leone was diminished in the latter part of the 20th century and the early years of the 21st. This can be attributed

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

94

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

to two main factors: Fourah Bay College, the main academic centre in Sierra Leone, did not receive the same attention following independence from either the former colonial rulers or the new government as did the University of Ibadan or the University of Ghana and did not fall within the immediate scope of the West Africa Language Survey in the same way as did the other universities. The second and perhaps more critical reason for the slow (or arrested) development of linguistics – and academics generally – in Sierra Leone was the civil war that devastated the country in the last decade of the 20th century. With the return of stability, linguistics has gradually begun to grow. The Gambia and Liberia too have seen little in the way of development of linguistics as an academic discipline. In the case of the Gambia, this may also be attributable to the lack of attention it received as an academic centre during the colonial era but also that it is the smallest country in Africa and essentially surrounded by francophone Senegal. While Cameroon is frequently considered to be part of Central Africa, its English-speaking region is a former British colony; it borders Nigeria, and many of its languages are found on both sides of the international frontier. Cameroonian linguists and linguistics programmes were also brought under the umbrella of WALS/SLAO from the outset, with the establishment of linguistics at Université Yaoundé I in the early 1960s. It is useful here to consider the English-speaking region of Cameroon together with the other parts of anglophone West Africa. Linguistics also exists at the University of Buea, the major academic centre in Cameroon’s anglophone region, established in 1993 explicitly to follow the English academic tradition. During the irst years of the 21st century, it established itself as an important centre for linguistic research, with a focus on languages of Western Cameroon and the Grassields region. In 2012 the Seventh World Congress of African Linguistics was hosted by the University of Buea, marking the irst occasion that the premier international meeting of African linguistics had been held in an English-speaking country of West Africa.

4.3.7

West African Linguistics in the 21st Century

Linguistics continued to grow and thrive in the early years of the 21st century in those parts of West Africa that are ‘English-speaking’. The collaborations witnessed between Africans and non-Africans in the decades following independence, both with Africans going abroad to study and work at major centres of linguistic research and returning to their home universities (Bamgbose, Boadi, Dolphyne, Emenanjo, and Essien, among others) and with non-Africans coming to study, teach, and do their research at West African universities and becoming de  facto Africans (Kropp Dakubu, Williamson), have led to the establishment of a solid academic linguistic tradition in the region. By the end of the 20th century, linguistics programmes were producing graduates capable in their own

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

95

right not only of training the next generation of linguists, but of contributing on the international stage. The days of linguistics departments or programmes being staffed mainly by expatriate linguists had passed. Instead, West African scholars based in American and European universities return regularly to West Africa to organize schools such as the two summer schools on language documentation held at the University of Winneba in Ghana (2008, 2010), and the biennial African Linguistics (summer) School held on a rotating basis at different universities and without regard for the anglophone/francophone divide. There are numerous African scholars who, in maintaining and developing programmes in their home universities, have also developed and maintained strong ties with universities and colleagues abroad.

4.3.8

Language Development

One of the distinguishing characteristics of linguistics in English-speaking West Africa since its beginnings in the 19th century has been a focus on language development. In Crowther’s time, this was motivated to a large degree by evangelization and the need to be able to write African languages in order to translate religious writings. In the modern era the concern with language development has continued, though in the hands of indigenous linguists the focus has been squarely on education and nation building. Ayo Bamgbose, having established himself in his early years with descriptive and theoretical work, dedicated the last half of his long career to issues pertaining to the role of language (and thus of the linguist) in society and the importance of supporting African languages. Inasmuch as most West African languages were unwritten, the main thrust of language development was orthography design and the production of bilingual dictionaries; in Ghana, for example, the Akan Orthography Committee was founded in 1952; in 1978 a common orthography began to be used in primary education; and in the 1980s a uniied orthography suitable for the three major dialects of Akan was agreed. Work on other Ghanaian languages, such as that of the Dagaare Language Committee, followed. In Nigeria, the three largest languages, Hausa, Yoruba, and Igbo, not surprisingly, have received the greatest amount of attention over the years. As already noted, development of Yoruba and Hausa began in the 19th century, and at the time these efforts inluenced Lepsius in his work developing a general alphabet for African languages. However, the substantial dialect variation found in Igbo made attempts to develop a common orthography for that language more dificult. The current oficial orthography has been in use since 1961, though debate has continued as modiications have been proposed periodically. Many other Nigerian languages have been the target of development efforts, far too many to mention here. It does bear mention though that

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

96

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

despite its auspicious beginning as the object of missionary linguists’ attention in the 19th century, and the early work of Ida Ward (1933), Eik has since received scant attention. On the other hand, two of its sister languages, Ibibio and Obolo, have proited by the attention of linguists. The same may be said of languages in the north of Nigeria, such as Kanuri. Discussion of language development in Nigeria would be incomplete without mention of the work of Kay Williamson. Although an expatriate linguist, Williamson spent her entire career in Nigeria, irst at the University of Ibadan as a founder member of the linguistics programme there and subsequently as a founder of the linguistics programme at the University of Port Harcourt. While at Port Harcourt she undertook what was known as the Rivers Readers Project, the development of orthographies and primers for many of the small languages of the then Rivers State. For this work, together with her role in training numerous Nigerian linguists, Williamson became known as the Mother of Nigerian Linguistics.

4.3.9

Language Endangerment and Language Documentation

In the last decade of the 20th century and the early years of the 21st, African linguists showed an increasing awareness of the threatened state of African languages. This concern led to the Summer Schools in Documentary Linguistics in West Africa, the irst in 2008 and another in 2010, both sponsored by the Hans Rausing Endangered Languages Documentation Programme at the School of Oriental and African Studies of the University of London, and both were held at the University of Education in Winneba, Ghana. They were organized by Felix Ameka and featured a cohort of international and African scholars as instructors; students were drawn from across West Africa, both anglophone and francophone countries, though all instruction was in English.

4.3.10 African Linguistics School A similar development, though oriented more towards theoretical linguistics, was the African Linguistics School founded by expatriate African scholars in collaboration with Western linguists and modelled on the Linguistics Institute of the Linguistic Society of America. The school was held four times up until the time of writing, each time in a different location. Its instructors were internationally recognized linguists from North America, Europe, and Africa. As with the School in Documentary Linguistics, students came from both anglophone and francophone countries, though in this case from all over Africa. Funding for the school came from a variety of sources: organizations such as the US National Science Foundation, the Carnegie Corporation of New York, and the Council for the Development of Social Science Research in Africa (CODESRIA); and different universities and linguistics associations. Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

African Linguistics in North-Eastern and So- Called Anglophone Africa

4.3.11

97

Conclusion

We conclude this brief sketch with an observation on what appear to be two general trends over the 200 or so years in which linguistics and language study has developed in this part of West Africa. The irst is an abiding concern with language development manifested irst, in the 19th century, through the desire for evangelization, and more recently in a recognition of the role language development has to play in social development and nation building. While in the modern era interest in theoretical linguistics has grown, it has not been at the expense of recognizing the role language plays in society. The second trend is cooperation between Africans of the region and non-Africans. The fruits of this collaboration have been descriptive works on local languages and theoretical contributions to general linguistics as well as practical outcomes. At the beginning of the 21st century this cooperation was still present, though with the balance shifted; only rarely now does one ind non-Africans in teaching positions in linguistics programmes in the region, which are staffed almost entirely by indigenes; it is more usual to ind West African scholars on the faculty of universities in North America, Europe, and other parts of the world.

Acknowledgements Marie-Claude Simeone-Senelle is grateful to Fatouma Mahamoud (Research Assistant at ILD), Hibo Moumin Assoweh (Senior Lecturer at UD, Director of CRUD), Amina Said Chireh (Teacher-Researcher at UD, Secretary of IRICA), and Mohamed Hassan Kamil (Researcher, Director of ILD). Amani Lusekelo gratefully acknowledges valuable information provided by David Massamba (University of Dar es Salaam), Felix Banda (University of Western Cape), Jane A. N. Udour (University of Nairobi), Nancy C. Kula (University of Essex), and Al Mtenje (University of Malawi). Bruce Connell and Akinbiyi Akinlabi acknowledge the contributions made by a number of scholars, colleagues who, through having been a part of the development and expansion of linguistics in West Africa, have lived the history they have discussed. Harrison Adeniyi, Nana Aba A. Amfo, Offeibea Awuku, Regina Caesar, G. Tucker Childs, Charlotte Fofo Lomotey, Ozo Mekuri Ndimele, Amo Ofori, and Ekkehard Wolff as editor all provided information not available to us through either published sources or our own recollections.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:31:33, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.004

5 African Linguistics in So- Called Lusophone and Hispanophone Africa and in Southern Africa Anne-Maria Fehn, Sonja Bosch, and Inge Kosch This chapter looks, in a summarizing way, at the emergence and develop­ ment of African linguistics in so­called Lusophone and Hispanophone Africa as well as in Southern Africa. For a fuller account in separate chap­ ters, including extensive bibliographical references, see A History of African Linguistics (Wolff 2019).

5.1 African Linguistics in So- Called Lusophone and Hispanophone Africa (Anne-Maria Fehn) 5.1.1

Background

This sub­chapter discusses the history and current status of African studies in the Portuguese ex­colonies Mozambique, Angola, São Tomé e Príncipe, Guinea­Bissau, and Cabo Verde, which have Portuguese as their oficial language, as well as in the former Spanish colony Equatorial Guinea, in which Spanish, French and Portuguese have oficial status. All presently lusophone countries were discovered by Portuguese seafarers during the 1400s and colonized in the subsequent centuries. Despite their growing economic importance, none of the colonies managed to attract a signif­ icant amount of settlers from the European motherland. Only after the Berlin Conference of 1884–1885 demanding ‘effective occupation’ from the participating nations can a change in policy, accompanied by limited investment in infrastructure and education, be observed (Keese 2007). With Salazar’s rise to power, the colonies became central to the self­ conception of the Estado Novo as a legitimate successor to the former glory of the Portuguese Empire. In 1951, the Portuguese colonies were rebranded ‘overseas provinces’, and their status as fully recognized parts

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

African Linguistics in Lusophone, Hispanophone and in Southern Africa

99

of Portugal was reinforced (Rocha 2017). During the 1960s, a new institu­ tion, the Junta de Investigações do Ultramar (now Instituto de Investigação Cientíica Tropical) was established in Lisbon to administrate and publish research undertaken in the Portuguese colonies. Notwithstanding the amount of scholarly works produced, research on non­European languages and cultures remained irmly in the hands of the colonizers. Africans had only limited access to the educational system, which, until independence, remained underdeveloped and focused on applied studies. Research under­ taken by researchers independent from the colonial framework was rare and remained restricted to missionary works and amateur studies receiv­ ing little international attention. The 1960s and 1970s were characterized by the outbreak of wars of inde­ pendence in Angola, Mozambique, and Guinea­Bissau. While Equatorial Guinea obtained independence from Spain in 1968, the Salazar regime in Portugal held on to its oversea territories until the peaceful revolution in 1974, which was accompanied by the dissolution of the colonial empire. Immediately after independence, violent civil wars continued to shake Angola and Mozambique, destabilizing both countries. Guinea­Bissau and Equatorial Guinea did not suffer a civil war but were subjected to autocratic politicians and frequent regime changes effectively preventing sustainable development, including a modern system of education and independent research institutions.

5.1.2

Mozambique

Apart from anthropological and linguistic research by scholars from non­ African countries, works from the ield of African studies produced in pre­ independence Mozambique are hard to come by. The irst grammars of local languages were drafted by missionaries: As early as the 17th century, Dominican priests wrote catechisms in Tete and Sena that by now have been lost (Moser 1986). Around the same time, an anonymous Jesuit wrote a description of Sena that was translated into German and republished by Schebesta (1919). Later works from the late 19th and early 20th century include grammars of Ronga (Farinha 1917) and Sena (Anderson 1897). A notable character in the research history of colonial Mozambique is the Portuguese scholar and colonial civil servant António Rita­Ferreira. Based on his travels around Mozambique, he extensively published about the country’s ethnographical landscape (e.g., Rita­Ferreira 1958, 1959, 1975). After independence, the country’s primary institution of higher educa­ tion became the Universidade de Lourenço Marques, which was founded during colonial times and was renamed Universidade Eduardo Mondlane in 1976. One of the irst academic departments devoted to the study of African cultures became the Centro de Estudos Africanos, with a strong focus on researching and promoting indigenous cultures for the beneit

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

100

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

of education and development. The linguistic diversity of Mozambique irst featured in the university’s curriculum when the linguist Eugeniusz Rzewuski, then a professor in Maputo (1977–1983), founded a department dedicated to the study of the country’s indigenous languages (Núcleo de Estudos de Línguas Moçambicana; NELIMO). Right after its establishment, the irst aim of the institute became the creation of a comprehensive bib­ liography of Mozambican Bantu, followed by the collection of expanded glossaries of scientiic and technical terms (Lopes 1998). After Portuguese had remained the sole language of education in post­ independence Mozambique, the Instituto Nacional de Desenvolvimento da Educaçao (INDE) collaborated with Stockholm University in establishing a mother tongue education project for the use of Bantu languages in adult education (1992–1997). Subsequently, NELIMO and INDE worked together in describing languages with the aim of creating standardized orthogra­ phies and teaching materials, such as manuals, primers, and supplemen­ tary readers (Lopes 1998). The description and standardization of Mozambican languages has also been supported by researchers from the Summer Institute of Linguistics (SIL) who have been collaborating with Mozambican Institutions like the Universidade Eduardo Mondlane since 1986. Their current research activi­ ties include work on Mwani, Sena, and Makonde, which is frequently cou­ pled with standardization and alphabetization endeavours (SIL 2017).

5.1.3

Angola

The irst descriptions of Angolan languages and peoples were produced by Catholic missionaries who came to the country during the 16th century. The earliest works in a Bantu language of the area were two catechisms in Kikongo, written by Gaspar da Conceição (1555) and Mateus Cardoso (1624); later, a grammar of Kimbundu was published as Arte de gramática da língua de Angola by the Brazilian Jesuit Pedro Dias in 1697 (Zwartjes 2011). During the 19th and early 20th centuries, academic research related to the ield of African studies was mainly carried out by colonial scholars sent from Lisbon. Works not incorporated into the administrative framework of Portuguese colonial studies are hard to come by. The German mission­ ary Karl ‘Carlos’ Estermann (1896–1976) lived and worked in southwestern Angola from 1924 to his death and devoted his time to the ethnographic documentation of the peoples of the area. Although his major work, the Etnografia do Sudoeste de Angola in three volumes (Estermann 1956, 1961a, 1961b), was published under the umbrella of the Junta de Investigações do Ultramar, it can be considered an independent work based in Angola rather than in Portugal. Another important character is José Redinha (1905–1983), who came to Angola as a civil servant in 1927. From 1936 to 1946, he travelled about

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

African Linguistics in Lusophone, Hispanophone and in Southern Africa

101

15,000 kilometres during various expeditions for territorial, ethnic, and cultural recognition, collecting more than 20,000 ethnographical arte­ facts. His best­known book, Distribuição étnica de Angola, was irst published in 1962 and has since been reprinted multiple times. In post­independence Angola, the main institution of higher education remained the University of Luanda, which in 1985 became Universidade Agostinho Neto. During the 1980s, various other institutions, particularly with a focus on teacher training, were founded throughout the country, with the Instituto Superior de Ciências da Educaçã (ISCED) in Luanda being of particular importance to the emergence of African linguistics in Angola. However, in the years following independence, the study of the languages of Angola was not undertaken at the university level, but remained in the hands of the Instituto Nacional de Línguas founded in 1976. The insti­ tute devoted its limited capacities to the applied study of the six national languages with the widest distribution (Kikongo, Chokwe, Umbundu, Kimbundu, Ganguela, Kwanyama), culminating in the creation of orthog­ raphies irst released in 1980. Only after the civil war, in 2004, was the irst department of African languages and linguistics established at an institu­ tion of higher education, ISCED, under the label Departamento de línguas e literaturas africanas. The relatively late interest in African languages in post­independence Angola is of course explained by the disastrous impact of the long­lasting civil war on Angolan society, but also by the limited importance of regional languages for governance and education, compared to other lusophone countries in Africa. Irrespective of their ethnic identities, more than 70 percent of Angolans speak and understand Portuguese, 39 percent even consider it their mother tongue (INE 2014). The study and development of indigenous languages, especially for educational purposes, was there­ fore not given priority until after the war. Recently, in the constitution of 2010, the ‘study, teaching and use of other Angolan languages’ has been promoted, along with the goal to ‘protect, value and dignify Angolan lan­ guages of African origin’ (UNICEF 2016). The Departamento de línguas e literaturas Africanas, along with other institutions offering training for teachers, is considered instrumental in implementing these policies, as it focuses on both the study and the teaching of the national languages. With the help of Pearson Education, a British publishing service, text­ books have been published in seven Angolan Bantu languages, and mother tongue education projects have been established throughout the country (UNICEF 2016). On the academic level, one of the main personalities promoting research on Angolan languages is Zavoni Ntondo, now a professor at ISCED and the Universidade Agostinho Neto, who has published on various Angolan Bantu languages (e.g., Ntondo 2006, 2015). ISCED­Luanda, as well as other academic institutions in Angola, also entertain a range of partnerships

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

102

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

with researchers and institutions from across the world. In the ield of African studies, recent collaborations include the Summer Institute of Linguistics (SIL) (e.g., Jordan 2016), as well as a TwinLab between ISCED­ Huíla in Lubango and CIBIO/InBIO: Research Centre in Biodiversity and Genetic Resources in Vairão, Portugal (e.g., Pinto et al. 2016).

5.1.4

São Tomé e Príncipe

The peoples and languages of the archipelago received only limited sci­ entiic attention during colonial times. While the existence of creolized varieties of Portuguese was mentioned from the 17th century onwards, written testimonies appeared only in the second half of the 19th century, especially in the works of São Tomean writers Fransisco Stockler and Almada Negreiros (Araújo & Hagemeijer 2013; Negreiros 1895). The irst linguistic descriptions of the Gulf of Guinea Creoles were provided by the creolists Schuchardt (1882, 1889) and Coelho (1880–1886). Santome was irst used as a literary language by São Tomean writer Francisco de Jesus Bonim, also known as Faxicu Bêbê Zaua, who from the 1920s onwards wrote texts for the journal A Liberdade and various pamphlets (Araújo & Hagemeijer 2013). The irst institution of higher education, the Instituto Superior Politécnico, was founded on the island of São Tomé in 1994 and reached the status of a university only in 2014. No information on the current availability of a department of African linguistics or related courses could be obtained. As the oficial language Portuguese is universally spoken in the archi­ pelago, few language planning activities have been undertaken with regards to the country’s minority languages. While a growing corpus on linguistic works on the Gulf of Guinea Creoles produced in Portugal, Brazil, and elsewhere has become available (e.g., Ferraz 1979; Hagemeijer 2011), only a small set of language materials to be used in mother tongue education or language revitalization was created. Based on experiences made in the Latin American context (Araújo & Agostinho 2010), linguists from Brazil and Europe have made propositions for the standardization of the country’s related creole languages, including a standard orthog­ raphy (Alfabeto Uniicado para a Escrita das Línguas Nativas de S. Tomé e Príncipe; ALUSTP), which was approved by the government in 2013. There are publications using it (e.g., Araújo & Hagemeijer 2013), and fur­ ther materials to be used in teaching the languages and training teachers are in preparation (see, e.g., Araújo  & Agostinho 2010; Agostinho et  al. 2016). At present, some attempts to revitalize Principense are under way, currently involving eight teachers who meet regularly to establish a pre­ liminary curriculum for teaching youngsters, albeit without considering the standard orthography and with no oficial teaching materials at hand (Agostinho et al. 2016).

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

African Linguistics in Lusophone, Hispanophone and in Southern Africa

5.1.5

103

Guinea-Bissau

During the inal decades of Portuguese colonialism, the ethnic and linguis­ tic heterogeneity of Guinea­Bissau attracted a limited amount of scholarly research, albeit closely monitored by the colonial authorities. In 1945, the governor of the colony, Sarmento Rodrigues, launched the Boletim Cultural da Guiné Portuguesa, a journal that reached 110 volumes from 1946 to 1973 and thereby became the major bulletin of cultural sciences in the Portuguese colonies. The journal was published by the Centro de Estudos da Guiné located in the capital Bissau and featured articles from various ields, ranging from agriculture to cultural anthropology and linguistics. Politically, no measures were undertaken to strengthen or develop the indigenous languages of Guinea­Bissau. Especially the Crioulo was sup­ pressed by the colonial authorities, who feared its potential as an African lingua franca uniting the multilingual ethnic landscape (Kohl 2011). After independence, Crioulo became the only language of Guinea­Bissau that has received attention from international researchers, combined with some minor attempts at language planning. The language is attested in the country since at least the 16th century and shares a common ancestor with the Crioulo dialects of Cabo Verde, where it most likely emerged before being spread to the mainland by immigrants (Jacobs 2010). It has been the subject of a number of publications dealing with aspects of its grammar (e.g., Couto 1994) and lexicon (Rougé 1988), with the Italian missionary Luigi Scantamburlo (e.g., 1981) acting as a major proponent of research on the language in Guinea­Bissau itself. Although Guinea­Bissau has a lan­ guage policy focusing on Portuguese as sole language of administration and instruction (Cá 2015), some attempts were made to introduce Crioulo, the country’s most widespread lingua franca, into the educational system. The Ministry of Education of Guinea­Bissau in cooperation with Dutch and Portuguese institutions launched an experimental project (1987– 1997) for the use of Crioulo as the medium of instruction for the irst two grades (Hovens 2002:253). Along similar lines, the aforementioned Luigi Scantamburlo started a bilingual teaching project in the Bijagós archipel­ ago (Cá 2015). An overview of ongoing projects using Crioulo as a language of instruction is provided by Martins et  al. (2016). Incanha Intumbo, a mother tongue speaker of Crioulo and student of Scantamburlo, is cur­ rently working at the University of Coimbra in Portugal and has published on his language (e.g., Intumbo 2006).

5.1.6

Cabo Verde

During the times of the Estado Novo, the use of the local language, the Portuguese­based creole Crioulo, in writing and education as well as in gov­ ernment buildings was strictly forbidden (Lobban 1995). Intriguingly, the stigmatization of the language during the colonization period did not pre­ vent its rise to a language of music, literature, and poetry. Crioulo­speaking

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

104

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

intellectuals like Jorge Barbosa (1902–1971), Eugénio Tavares (1867–1930), and Baltasar Lopes da Silva (1907–1989) created an impressive corpus of written literature in their mother tongue that deied the prohibition of the colonial authorities. Although Portuguese remained the country’s oficial language after inde­ pendence, there was a strong movement of people supporting the promo­ tion of Crioulo to the status of a co–oficial language. As a irst step towards this goal, linguists of Cabo Verdean origin attended a colloquium held by the Direcção Geral da Cultura in 1979 to work on the development of a uniied orthography (Veiga et al. 2000). This orthography was subsequently used by some writers, but did not become unanimously accepted, mostly due to dialectal differences in spoken Crioulo, which speakers found dif­ icult to unite in a single set of orthographic rules (Duarte 1998). Despite these persistent problems in standardizing the language, a group of lin­ guists headed by Manuel Veiga (b. 1948) keep ighting for the use of Crioulo in oficial contexts and as a language of instruction in schools and institu­ tions of higher education. In 1994, a new orthography was created based on the 1979 version and became known as ALUPEC (Alfabeto Uniicado para a Escrita do Cabo­Verdiano). ALUPEC was approved in 1998 and recognized by the government in 2005. Nevertheless, like its predecessor, it never man­ aged to satisfy all speakers and to this date is neither oficial nor mandatory. The comparatively big number of trained linguists with a Cabo Verdean background may seem surprising considering the country’s lack of a univer­ sity, let alone a department of linguistics, until the formation of the Instituto Superior de Educação (ISE) in 1995. However, it has to be kept in mind that the majority of Cabo Verdeans live in the diaspora and many scholars active in the debate actually received their training in Portugal, France, the United States, or Brazil. In consequence, there is a considerable corpus of linguistic descriptions, sociolinguistic treatments, and dictionaries on varieties of Crioulo from Cabo Verdean linguists, only a small subset of which can be cited in this chapter (e.g., Cardoso 1989; Delgado 2008; Mendes et al. 2002; Veiga 1996, 2000, 2002). Making use of these works, the Universidade de Cabo Verde, founded in 2006 as a union of smaller institutions of higher education such as ISE, offers a comprehensive course titled Línguas, Literaturas e Culturas – Estudos Cabo­Verdianos e Portugueses, which combines Portuguese and Cabo Verdean Linguistics with an overview of the rich literary tradition available in both languages.

5.1.7

Equatorial Guinea

Claretian missionaries began their work in Equatorial Guinea in 1883 and, from 1903 to 1947, published a bimonthly journal titled La Guinea Española that regularly dealt with indigenous cultures and languages. However, there was no actual interest in studying the latter for the purpose of mother tongue education. When by 1920 three Equatorial Guineans asked

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

African Linguistics in Lusophone, Hispanophone and in Southern Africa

105

for permission to teach in public schools, they were turned down as it was believed that they did not have the competence to teach in Spanish (Castillo­Rodríguez 2014). Other missionaries were less hostile towards the indigenous languages and created detailed descriptions and teaching materials (Bolekia Boleká 1988). The Baptist missionary John Clarke lived on Bioko (then Fernando Po) from 1840 to 1843 and wrote the irst descrip­ tion of Bube. His work was followed by the Methodist missionary William N. Barleycorn, who was himself of Equatorial Guinean descent and in 1875 created the irst Bube primer Bubi na English. The irst work in Castilian, the Primer paso a la lengua bubi, was authored by Joaquín Juanola in 1890 (Bolekia Boleká 1988), and during the 1920s, several linguistic works about the lexicon and grammar of the Bube language were published in Spanish (e.g., Aymemí 1942; Abad 1928). Other works that appeared during the colonial period include descriptions of Fang (e.g., Aranzadi 1962) and the Portuguese­based creole of Annóbon, Fa d’Ambô (e.g., Barrena 1957, 1965). Since independence, the teaching of languages at the country’s irst uni­ versity Universidad Nacional de Guinea Ecuatorial (established in 1995) has focused more on the acquisition of European languages of wider communi­ cation than on the study of the indigenous African languages (Johnson 2013). The Agencia Española de Cooperación Internacional is supporting the local cultural scene through the Centro Cultural de España en Malabo and the Centro Cultural de España en Bata, established in 1982 as Centro Cultural Hispano­Guineano. The organization further publishes the quarterly mag­ azine África 2000 and the monthly magazine El Patio, both of which feature works from local writers. Among the Equatorial Guinean intellectuals and scholars working abroad, Justo Bolekia Boleká (b. 1954) from the University of Salamanca is noteworthy for his poetry, essays, and linguistic work on his native language Bube (e.g., Bolekia Boleká 1988, 1991).

5.2 5.2.1

African Linguistics in Southern Africa (Sonja Bosch & Inge Kosch) Introduction

The aim of this sub­chapter is to give an overview of the history of African linguistics in Southern Africa (including Botswana, Lesotho, Namibia, South Africa, Swaziland) on Bantu and so­called Khoisan languages, from an inside­Africa perspective.1 The focus is on research activities 1

The term ‘Bantu’ was introduced in language studies as far back as 1857 by the German philologist Dr W. H. I. Bleek. Although this term became stigmatized in South Africa due to its use in contexts other than language, it is used in the international field of linguistics to refer to a particular family of languages that share common linguistic features. Khoisan today refers to five groups of non-Bantu click languages spoken in Southern and Eastern Africa. Traditionally, there is a basic distinction within South African so-called Khoisan, namely Khoe (formerly and derogatively ‘Hottentot’), and San (formerly and derogatively ‘Bushman’) according to Wolff (2016a:304).

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

106

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

and methodological approaches to the study of African languages, which were established and homegrown and are largely conducted in loco by African researchers with or without (further) outside stimuli at local universities and research institutions. We address the motivations for copying or maintaining European/Western ‘academic culture’ and higher education systems in order to ‘close the gap’ between the formerly so­called First World and what is now called the Global South, if not – in the long run – to establish Africa as a relevant region of global knowl­ edge production (which hitherto it is not, with the possible exception of South Africa). Such a perspective regards the impact from Europe and elsewhere as more or less imposed or solicited ‘academic fertilization’ for the growth of local and regional structures and institutions. Furthermore, stories of success and failure on African soil are included, suggesting a clear­cut dis­ tinction of the apartheid and post­apartheid periods. This sub­chapter is roughly divided into the following three historical periods  – these periods should be regarded not as ixed time spans but as approximate indications of the era that was dominated by a particular model: 1. The exploration years covering pre­scientiic (missionary, colonial) studies of African languages (ca. mid­1800s–1927). 2. Postcolonial linguistics as dealt with by Southern African institutions (ca. 1927–1975). 3. Modern linguistics relecting research priorities with regard to a vari­ ety of modern linguistic theories (ca. 1975–present).

5.2.2 Pre-scientific Studies (ca. mid-1800s–1927) 5.2.2.1 Background In Africa, Westerners in the person of missionaries were invariably responsible for the development of an orthography for the various African languages, the prerequisite for any further linguistic investiga­ tion. By the mid­1800s the spade work for the establishment of ortho­ graphies and spelling systems had roughly been done, empowering the missionaries to gain the ‘monopoly’, as it were, over the development of linguistic analyses and descriptions from a Western viewpoint. The mid­ 1800s is an arbitrary starting point to the pre­scientiic period, while its end can be associated with a more deinite milestone in 1927, to be dis­ cussed in the next section. There were several noteworthy publications on grammatical issues in African languages before the 1850s (e.g., Bennie 1826; Boyce 1834; Archbell 1837; Casalis 1841; Bryant 1849; to mention but a few), but since then a steady increase in the quality and quan­ tity of contributions can be observed. Colenso, for example, produced

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

African Linguistics in Lusophone, Hispanophone and in Southern Africa

107

a Zulu English Dictionary in 1861, which had a long­term effect on the development of the Zulu language. Apart from well­known languages, lesser known languages like the Khoisan languages also attracted the attention of scholars (Hahn 1857; Engelbrecht 1925; Vedder 1910, 1911; Doke 1925). Phonetic descriptions were and remained a topic of interest in the studies of the Khoisan languages due to the unique sound system of these languages.

5.2.2.2 Research Priorities The pre­scientiic or exploration period was characterized by the produc­ tion of word lists, dictionaries, and basic grammars by missionaries. Their primary objective was of a practical nature, namely to design lesson man­ uals for language learners, primers, elementary readers, and Bible trans­ lations, to facilitate the communication of the gospel. The translation of the Bible and grammatical work could not be divorced from each other. According to Welmers (1971:567–569), lesson manuals are academically interesting since ‘they can be classed as good descriptive grammars which happen to have a pedagogical organization’. The focus was on the acquisi­ tion of language and not on the scientiic description of individual linguis­ tic phenomena. Not being trained as language experts in African languages, the mis­ sionaries took recourse to patterns of grammars of European languages when drafting their manuals. It is therefore not surprising to ind some descriptions in the early grammars that are inaccurately imposed on the structure of the relevant African language. However, it was quite accept­ able for linguistics of contemporary and classical languages of the time to serve as paradigm for linguistics of African languages. Once the basics of a grammar had been laid down, missionaries with an interest in linguistics and other scholars invariably shifted their focus to the description of speciic linguistic phenomena. With regard to the study of African languages during this period, the following comment describes the developments: From the fourth decade of the 19th Century right on through into the 20th Century, in increasing numbers, the Bantu languages began to be studied in a more modern method with more attention to their intrinsic structure. It might be stated that the mediaeval Latin approach of the ‘Age of Brusciotto’ gave way to the imposition of modern European grammati­ cal methods, in which only to a certain extent was ‘Bantu grammar’ given any free play. Two decades of the 20th century had to pass before any real move was made for the 100 per cent treatment of Bantu languages accord­ ing to the genius of Bantu grammar. (Doke & Cole 1969:27).

Ziervogel (1956:1–2) adds in similar vein, One may take it as a rule that all these writers superimposed the grammar of their own mother tongue upon the particular Bantu language they were

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

108

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

dealing with. . . . Not only was it the accepted approach at the time, it is the only logical and practical one. The medium used in the exposition of Bantu languages was quite for­ eign to the structure of Bantu. It is a fact we have to consider that the language in which we clothe our thoughts has a deinite bearing on our approach to a subject.

5.2.2.3 Leading Figures Missionaries were often the leading igures in the exploration of linguistic phenomena. There were sporadic contributions by native speakers to the development of African languages during this period, but these pertained more to literature than to linguistic issues. Having had irsthand exposure to the structure of the language while compiling basic grammars, some of the missionaries developed into ardent linguists who advocated for the correct spelling and grammatical descriptions of the African languages. A case in point is Endemann (1876), who produced a grammar on the Sotho languages in which he advocated for a scientiic approach to the choice of orthographic symbols. His scientiically motivated symbols, however, were impractical and never became popular. His grammar and that of other emerging scholars, had vestiges of Western linguistic descriptions. Tertiary departments of African Languages in Southern Africa did not exist per se during the pre­scientiic period. A precursor to such depart­ ments was for instance the Department of Bantu Studies at the University of the Witwatersrand, which C. M. Doke joined in 1923 (Herbert 1993:4). In 1946 the Department of African Languages at the University of South Africa (Unisa) was established as an independent department after the evolvement of Unisa out of the University of the Cape of Good Hope, which had been functioning as an examining body since 1873 (Kosch  & Bosch 2014:51–52). Some noteworthy academic contributions on speciic linguistic aspects of African languages were produced during the pre­scientiic time, for example Bourquin (1922a, 1922b) on the preix of the locative and the adverb in Xhosa, and Eiselen (1924) on the change of consonants through a preceding i in the Bantu languages. As more and more languages were being reduced to writing and more information became available, interest in comparative study increased with scholars like Appleyard (1850), Bleek (1862 and 1869), Torrend (1891), and Meinhof (1899, 1906) rendering outstanding and leading contribu­ tions. Classiications of the Bantu languages were also attempted dur­ ing this period (e.g., Van Warmelo 1927). In Bleek’s Comparative Grammar (1869), one of the most important contributions was his classiication of the language types in South Africa according to ‘(a) a South African division of the sex­denoting languages (Nama, !Kora, Cape Hottentot and

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

African Linguistics in Lusophone, Hispanophone and in Southern Africa

109

Eastern Hottentot) and (b) a South African division of the Bantu family’ (Doke  & Cole 1969:64). Throughout this work, Bleek (1869) particularly compared and contrasted the ‘Hottentot’ sufix formation with the preix formation of the Bantu languages.

5.2.2.4 Scholars Not Born in Southern Africa Most of the scholars who produced material on the African languages were not born in Southern Africa but came as missionaries from Europe. They acquired the language of the people to whom they had been assigned by their missionary society and started producing written material with the help of local informants and converts. Many of these missionaries adopted a Southern African country as their new home.

5.2.3 Postcolonial Linguistics (ca. 1927–1975) 5.2.3.1 Background Contributors to postcolonial African linguistics in Southern Africa (ca. 1927 to 1975) were mainly African­born Africanists, with a broad train­ ing in linguistics. According to Ziervogel (1956:2), it was only since the early 1920s that scholars involved with the study of the indigenous South African languages were in fact South African­born, had lived in the coun­ try all their lives, and had acquired the Bantu languages at a young age. They generally focused on languages that they were well acquainted with from extended personal experience. Clements (1989:5) points out that this was in contrast with the European Africanists who spent time in Africa carrying out ield research. The European Africanists not only acquired the new languages they were exposed to but were also more holistic in their research involving history, society, and culture, true to the tradition of past pursuits of the missionaries, explorers, and colonial administrators. In the postcolonial period the uncertainties of the pioneer period dis­ appeared so that researchers in this era functioned in an academic envi­ ronment surrounded by political and social stability. Therefore linguistics became a focal point after the secondary role that linguistics of the African languages played during the pre­scientiic era (Griesel 1991:246). This period was marked by increased contributions from trained linguists, which ultimately completely replaced the participation of the missionar­ ies (Kosch 1993:29). Research during this period includes, beyond general grammar works, theses and scientiic articles in journals. Researchers were now mainly academics who were susceptible to and in contact with new linguistic approaches internationally, like, for instance, those of Meinhof (1899, 1906) and Guthrie (1948a). The language of publication was mainly English, occasionally Afrikaans, but seldom African languages. A few of the grammars published in the African languages included, for instance,

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

110

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Schwellnus’s (1931) Sepedi grammar (Tlhalosa-Polelo: Grammar ya Sesotho se se bolelwaxo dileteng tša Transvaal), Mojapelo’s (1960) Sepedi grammar (Popo-puo ya Sesotho), Nyembezi’s (1956) Zulu grammar (Uhlelo Lwesizulu), Pahl’s (1967) Xhosa grammar (Isixhosa), and Guma’s (1971) Sesotho grammar (An Outline Structure of Southern Sotho). Most grammar books that appeared in the vernacular during this period were the work of Europeans. In the case of the Khoisan languages, Dorothea F. Bleek published her Comparative Vocabularies of Bushman Languages in 1929, followed by a descrip­ tive grammatical sketch of the language of the /xam­ka­!e / (Bleek 1928– 1930). A re­classiication of the Khoisan languages was also attempted by Westphal (1962), whose interest in non­Bantu languages in Angola, Botswana, and Namibia resulted in research indings that the Khoi and the San languages belonged to totally different language families. Interestingly enough, Westphal’s so­called ‘splitter’ approach gained acceptance only in the late 1990s, mainly through the work of Tom Güldemann (W. H. G. Haacke, personal communication, March 7, 2017). Cole (1971:18) came to the conclusion that with the challenges experienced in the Khoisan lan­ guages at phonological level, ‘progress in morphological analysis becomes bogged down and retarded’. It will be shown in the discussion of the next era that scholars such as Snyman, Traill, and Haacke made a breakthrough in this regard. Nevertheless, Snyman’s main contributions to Khoisan studies were the publication of his master’s dissertation ‘An Introduction to the !Xũ (!Kung) Language’ (1970), in which he ‘cracked’ or deciphered the phonology of !Khung, and also his doctoral thesis ‘Zu|’hõasi Fonologie en Woordeboek’ (1975). Westphal was the supervisor of both these studies.

5.2.3.2 Research Priorities and Leading Figures In the early years of the 1920s the phonetic school became popular in linguistic studies. C. M. Doke, born in Bristol, England, in 1893 moved to South Africa with his parents in 1903 as a ten­year­old. He was an expo­ nent of this phonetic school in the ield of South African Bantu languages. His ‘Phonetics of the Zulu language’ (submitted for a DLitt degree in 1924) served as example for similar studies of various other Bantu languages throughout Africa. In 1927 Doke published his Textbook of Zulu Grammar, thereby revolutionizing Bantu linguistic description (Kosch 1993:28). Therefore the year 1927 is generally regarded as the watershed between the exploratory or pre­scientiic studies of African languages and the com­ mencement of a new era of grammatical description in the South African Bantu languages. Doke’s approach was basically functional as regards grammar, and purely phonetic as regards the deinition of a word. Doke (1945:79), who had the insight that a more suitable model of grammatical description for the Bantu languages was called for, described the major contribution of his own work as ‘a new grammatical classiication and

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

African Linguistics in Lusophone, Hispanophone and in Southern Africa

111

treatment designed to get away from the European and classical precon­ ceptions, and use a mould more naturally suited to the structure of the Bantu languages’. Doke’s model of grammatical description is regarded by Van Wyk (1993:21) as a paradigm shift in the African languages brought about by the discarding of inappropriate assumptions and terminology dating from earlier grammatical work in the classical tradition. Examples are superluous grammatical categories such as ‘case’, ‘preposition’, and ‘article’ (Kosch 1993:33). It is signiicant, however, that the descriptions of many Bantu grammars are to this day inluenced by Doke’s terminology, as well as his understanding of ‘moods’ and ‘tenses’, a relic of European language grammars (Kosch 1993:33). Regardless of such shortcomings, Doke’s approach, in which the word is rated highly as linguistic entity, deserves recognition ‘since it was the irst in which the identiication of words was based not on intuitions stemming from a European back­ ground, but on more scientiically motivated criteria’ (Kosch 1993:31). Overall, Doke’s theories strongly inluenced Bantu grammar for at least the next three decades. During the postcolonial period several attempts were made by linguists to improve on Doke’s scientiic investigations. Ziervogel came to the con­ clusion that the phonetic principle used by Doke to mark word bounda­ ries could not be applied across the board. In his publication A Grammar of Swazi (1952), Ziervogel therefore proposed a new approach to the iden­ tiication of the word in Bantu languages. Based on Meinhof’s view of the Bantu word, he adopted the morphological approach to the analysis of the Bantu word in terms of root plus preixes and sufixes. Later on, Ziervogel worked towards broadening his deinition by adding functional, prosodic as well as semantic principles. His work can also be regarded as a variation of Doke’s model. Ziervogel’s method of analysis did not have a particularly wide­ranging impact on Bantu grammatical works in general, although he applied it to various Southern African Bantu languages including Zulu, Swazi, Northern Sotho, and Northern Transvaal Ndebele. However, accord­ ing to Kosch (1993:39) the ‘School of Ziervogel’ can chronologically be regarded as the transition from the Dokeian period to the Van Wyk era. Cope (1957), in his work on the grammatical structure of Zulu, followed suit by deviating from Doke’s prosodic approach. Instead, he offered an alternative to the traditional ‘form and function’ analysis of Zulu by rig­ orously applying the principle of form. He chose to follow the syntacti­ cally based method propounded by the London linguist Malcolm Guthrie (1948a), who regarded the sentence as the most fundamental building block. Cope’s approach was also not far removed from Doke’s model and was moreover not widely acknowledged or actively applied beyond his own and his students’ publications (Griesel 1991:110). Van Wyk (1958) proposed a new theory with regard to word identiication in the Bantu languages particularly in Northern Sotho, as representative of

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

112

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

the disjunctively written Sotho languages, and in Zulu, as representative of the conjunctively written Nguni languages. Van Wyk’s aim was to address the main objection against Doke’s word identiication process, which was mainly of a phonological nature, and did not, for instance, make provision for anomalies such as monosyllabic words. He therefore propounded the analysis of words on different levels, with an emphasis on syntax. The grammatical description of Northern Sotho was more affected by Van Wyk’s theory than was the case of the analysis of the Nguni languages, which continued to be inluenced by the Dokeian structural approach inspired by Doke and Bloomieldian linguistics. Griesel (1991:89) regards van Wyk’s approach as a new model rather than a new paradigm, which did not completely substitute the Dokeian model in the research community, but rather formed an undercurrent. In addition, Kosch (1993:64) points out that for the most part ‘it can be said that the Bantu languages as a ield of scientiic investigation, beneitted from the theories put forward by Van Wyk. He had indicated the impor­ tant role of general linguistic science in providing a theoretical foundation which would ensure a more meaningful and systematic approach to Bantu linguistic study.’ Mother tongue speakers contributing to the scientiic study of the Bantu languages during the postcolonial period were few and far between. Scholars who attempted a scientiic treatment of Zulu and Southern Sotho in the medium of the languages were for instance Nkobi (1954) and Khaketla (1951). Referring to these works, Ziervogel (1956) points out that ‘they simply follow Doke without being in the least critical. Except for the medium, which is Zulu and Sotho respectively, they have brought us nothing new in two languages so well known.’

5.2.4 Modern Linguistics (ca. 1975–Present) 5.2.4.1 Background A gradual transition took place from about 1975 relecting research prior­ ities with regard to a variety of modern linguistic theories. The transition entailed a movement from structuralism to the modern period, with 1975 being adopted as an arbitrary boundary (Goslin 1983 in Kosch 1993:74). The irst of the more sophisticated contemporary theoretical models to offer a challenging new framework to Bantuists was the transformational­ generative model. In its wake followed a number of adaptations thereof as well as a variety of other approaches which soon superseded this frame­ work. In contrast to previous periods in the description of Bantu grammar, the modern period was not marked by the prolonged dominance of any one approach. In fact Bantu linguistic description as a whole opened up to a variety of international developments at approximately the same time, resulting in the co­occurrence of studies conceived within different frame­ works. (Kosch 1993:74)

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

African Linguistics in Lusophone, Hispanophone and in Southern Africa

113

5.2.4.2 Contributions by Various Institutions and Academics As mentioned above, African linguistic studies adopted divergent approaches in the years following 1975. In cases where linguistics was prac­ tised within a particular framework or philosophy, the impact was mostly localized and conined to certain universities. Individual researchers set the trend and mentored colleagues who in turn supervised postgraduate students who followed their approach with regard to their philosophy and theoretical inclinations. One could therefore speak of emerging regional ‘schools’ of African linguistics, which temporarily cast their research within the transformation­generative grammar (TGG) mould, or adapta­ tions thereof or other frameworks. At the University of Stellenbosch the theory of government and binding was applied to the study of isiXhosa (du Plessis, Visser). At the University of Pretoria the applicability of case grammar and relational grammar in Zulu and Northern Sotho was investi­ gated (Wilkes, Prinsloo). Research based on typological investigations was produced by scholars at Unisa (Hendrikse, Poulos, Bosch). A noteworthy contributor within the functional approach was Louwrens, who addressed issues like the distinction between old and known information, deinite­ ness and indeiniteness, and concepts like co­reference, focus, contrast, and discourse analysis. Discourse analysis narrowed the gap between pure linguistics and language as a social phenomenon. Traill (1977a:iii) remarked that ‘narrow linguistic research on the Khoisan languages is currently being undertaken by an extremely small number of individuals’. Two of the very few locally born scholars were Snyman (Unisa) and Traill (University of the Witwatersrand), who emerged as lead­ ing igures in the area of Khoisan research. Both had started publishing in the early 1970s. Snyman’s interest in the Khoisan group is relected in numerous publications in which he, as analytical ield linguist, dealt with language classiication based on lexical, phonological, and grammatical grounds (e.g., Snyman 1974, 1977). Traill was foremost a phonetician, spe­ cializing on the Khoisan language !Xóõ, renowned for having an extremely extended consonant inventory in comparison to any other existing lan­ guage. Traill’s publications (e.g., 1977b, 1985, 1986a) addressed the pho­ netics and phonology of !Xóõ and also contributed to comparative research on the sound systems in the Khoisan languages. Based in Namibia, ini­ tially at the Academy for Tertiary Education, and later at the institutional successor, the University of Namibia (UNAM), Haacke’s research interest ranges from grammatical descriptions of Nama (1977) to typological com­ parisons of northern dialects of Khoekhoegowab (2010). His main interest, however, remains tonology (Haacke 1999). The system he established for Khoekhoe has in essence been adopted by other Khoesanists, also for com­ parative work. Some scholars afiliated with the University of Botswana ventured into research on selected Khoisan languages, albeit with a more sociolinguistic focus (e.g., Batibo 2010a, 2015a; Chebanne 2010).

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

114

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Batibo’s contributions also cover aspects of phonology, tonology and morphology of languages in Botswana and Tanzania, including compari­ sons between linguistic features of Setswana and Kiswahili (1996, 2010a). Another scholar who has made contributions to the linguistic study of selected languages of Botswana and Čilubà is Lukusa, while Gabanamotse­ Mogara has added to existing knowledge on Naro, Ju|’hoansi and !Xóõ, languages that were the focus of her doctoral study (2011). Chebanne, also afiliated with the University of Botswana, has a special interest in Khoisan languages and is involved in the linguistic documenta­ tion of these languages of Southern Africa. To facilitate applied linguistics, computational linguistics and corpus linguistics came to the fore as essential technologies, leading to increased interest in interdisciplinary studies (see Chapter 22).

5.2.5

Conclusion

We presented three main phases in the development of the linguistic descriptions of African languages in Southern Africa. Academically based research output emerged during the missionary or pre­scientiic period from about the middle of the 1800s. Doke’s contribution in 1927 can be taken as a transition to the so­called postcolonial period when the dom­ inance of the Western framework was challenged with new insights regarding the unique structure of the African languages. As the previous phase, this period was also marked by seminal publications of prominent igures who had a long­term inluence on succeeding research. The advent of the modern period from ca. 1975 saw a deviation from the mainly struc­ tural investigation to more diversiied approaches such as TGG and sev­ eral related and other theories. This led to pockets of specialization within selected theories at different tertiary institutions. Whereas during the initial period of linguistic investigation missionar­ ies had a free hand in the study and teaching of the languages, in later periods the approach was shaped by changes in social and political condi­ tions. Over time African linguistics became a ield of specialization in its own right.

Acknowledgements Sonja Bosch and Inge Kosch are grateful to Wilfred Haacke for valua­ ble information on non­Bantu click languages research in personal communication.

Downloaded from https://www.cambridge.org/core. University of Texas Libraries, on 08 Dec 2019 at 16:49:28, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.005

6 African Linguistics in the Americas, Asia, and Australia G. Tucker Childs, Margarida Petter, Shigeki Kaji, Sun Xiaomeng, Yang Chul-Joon, and John Hajek

This chapter summarizes the history of the emergence and development of African linguistics in North America, Latin America, Asia, and Australia. For a fuller account in separate chapters, including extensive bibliographical references, see A History of African Linguistics (Wolff 2019).

6.1 6.1.1

A Brief History of the Study of African Languages in North America (G. Tucker Childs) Introduction

North American scholarship on African languages has never achieved the same intensity and breadth of research found in Europe, where the interest in African languages is long-standing and extensive.1 The reasons for these differences are historical and ideological, and perhaps institutional. The scholarship has, however, seen a long period after World War II and African independence, when the United States became the hub of (African) linguistics, following a brain drain of African PhD students from the United Kingdom to the States. In those days, Studies in African Linguistics (SAL) was considered the major journal in the ield, and the Annual Conference on African Linguistics (ACAL) was the African linguistics conference to attend. Leading universities were, for instance, UCLA, Indiana, Illinois, and Wisconsin. In recent years, African linguistics has seen a decline, including the closing down of many teaching programmes for African languages. Nonetheless, North American scholarship maintains a prominence in the ield, particularly on the formal-theoretical side. This work, however, has a focus different from purely descriptive work, 1

‘North American’ is used as a term referring to Canada and the United States; ‘America(n)’ refers only to the United States.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

116

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

often relying on secondhand accounts rather than original ieldwork, but it has nonetheless served as an important impetus for the study of African languages and linguistics.

6.1.2

Support for the Study of African Languages

In North America support for the study of African languages has come from a range of institutions, governmental and nongovernmental, steady and episodic. The ideologies motivating such study also cover a wide range. Motivations have ranged from spreading the word of God to saving dying languages. Nonetheless, the study of African languages in North America pales beside what has been done and what is being done in Europe. One thing that puts North America far behind Europe in terms of the breadth and longevity of European research on African languages is the lack of a colonial history. Canada also has nothing close to such a history, being a colony itself for many years and remaining in the Commonwealth. The closest thing the United States has to a colony is Liberia, a small country on the coast of West Africa, a relationship and history that require some explanation. One version of this story has it that Liberia was founded by freed American slaves, but in actual fact it was the American Colonization Society (established in 1816), a group of white Americans with mixed motives, who founded the state. One motive was the repatriation of free(d) African Americans, by whom many white Americans felt threatened. Another motive was to allow African Americans to return to their motherland, though few of them came from the area covered by the new state. The American Colonization Society, was one of several ‘Backto-Africa’ movements in US history motivated by sentiments dating back to the 18th century. The returnees or settlers, known as ‘AmericoLiberians’, held control over the indigenous peoples until a military coup in 1980. A civil war raged from 1989 to 1997 and again from 1999 to 2003, and Liberia was one of the three countries racked by the Ebola epidemic (2014–2016). Needless to say, little linguistic research was done during these periods of turmoil, but Liberia has been the site for much important research by Americans on African languages for it features the convergence of three important language families, Atlantic, Mande, and Kru, all typologically distinct groups, and has been the object of intense missionary work.

6.1.2.1 The Missionaries The role of Christian missionaries in the study of African languages is considerable (Welmers 1971) but relatively focused with regard to linguistics until recently. Previously limiting themselves to learning enough of the native languages to convert nonbelievers and translate the Bible, the

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

117

missionary effort of at least SIL International,2 the major North American organization, has now been realigned to deal with literacy and the documentation of endangered languages (www.sil.org/). SIL International is a faith-based non-proit organization committed to serving language communities worldwide as they build capacity for sustainable language development. SIL does this primarily through research, translation, training and materials development.  SIL works alongside ethnolinguistic communities and their partners as they discover how language development addresses the challenging areas of their daily lives – social, cultural, political, economic and spiritual.

They have also sought to monetize their products, for instance, limiting the number of free views of Ethnologue, their compendium of the world’s languages, as was not the case in the past. Founded in 1934 and led for many years by Kenneth Pike (see below for his contributions to African linguistics), SIL has trained many ine linguists of African languages, many of whom have contributed signiicantly to the ield. Nonetheless, there is some unease between missionaries and their non-missionary counterparts, American and African linguists. For example, a recent conference on African languages was relocated from the University of Texas at Arlington, a centre for SIL training, because of objections from the conference membership. Nonetheless, linguists have long depended on missionaries for support, especially in country but also in terms of ield technology (see Dobrin & Good 2009). An excellent summary and review of the complicated politics of missionary work can be found in Gilmour (2007) and some questions as to the future of missionary work are raised in Peterson and Allman (1999). I turn to some representative contributions by North American missionary-oriented linguists in the past and present. Kenneth Pike, a leader of SIL as mentioned above, was a pioneer in developing ield methodology. Although Pike’s work did not concentrate on African languages except for some work in Nigeria, his work on tone was crucial in identifying its phonemic importance (Pike 1948). Similarly, William Welmers had ties to much missionary work and was probably the most accomplished ield worker of his time, producing many irst descriptions of African languages beginning with his dissertation (Welmers 1946).3 His work in Africa began in Liberia, where he did work for the Lutheran Mission. He also trained many Africanists during his time at UCLA (two of whom joined the faculty). His magnum opus, African Language Structures,

2

The organization was formerly known as the Summer Institute of Linguistics; it originated as a summer training programme in Arkansas in 1934 (www.sil.org/about/history, accessed December 28, 2016).

3

His primary consultant was Mr Francis N.-K. (Kwame) Nkrumah, a leader of the independence movement in Ghana and its first head of state in 1957; his Penn dissertation was supervised by Zellig Harris, who also oversaw the studies of Noam Chomsky.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

118

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

where he recorded much of his personal experiences and discoveries in deciphering African languages, long served as an introduction to African languages (Welmers 1973). Although I have focused on SIL, there are other missionary groups that deserve mention with regard to work on African languages: the afiliated Wycliffe Bible Translators, the Lutheran Bible Translators (LBT), Pioneer Bible Translators, and United Bible Societies (UBS). William J. Samarin at the University of Toronto, whose work goes back to Samarin 1950,4 is a Pentecostal preacher and well known for his work on Gbeya and Sango (Samarin 1966, 1967a) and a ieldwork guide (Samarin 1967b). Current linguists associated with missionary work and producing important research on African languages are, for example, Lynell Marchese Zogbo (on Kru), Doris Payne (Nilotic), and John R. Watters (Bantu). More recent work based on African languages, speciically vowel harmony, comes from Miles Leitch and Rod Casali (e.g., Leitch 1997; Casali 2008), while important work on tone and labialvelars comes from Mike Cahill, the head of research for SIL (e.g., Cahill 2017). Relatively recent publications include Hopkins (1995) and Carlson (1994), which contain rich descriptions and irm grounding in linguistic theory.

6.1.2.2 Governmental Entities This section looks at the contributions North American governments have made to the study of African languages. I start with some American institutions. The Foreign Service has played a role in fostering research on African languages. Its language manuals sometimes represented the only research done on a language. Languages for which they developed manuals were such widely spoken tongues as Hausa, Swahili, Lingala, and Wolof. In 1958 the study of African languages, however, shifted into high gear. A number of ‘National Resource’ (Title VI) centres were established around the United States, as part of the reaction to the Soviets’ successful launching of Sputnik. An important aspect to all of these centres was the learning of foreign languages, especially for research abroad. The brief of the National Resource Centres was to provide the following for both graduate and undergraduate students: • instruction of ields and topics that provide full understanding of areas, regions, or countries • research and training in international studies • work in the language aspects of professional ields and research • instruction and research on issues critical to current world affairs5 A complete list of the Title VI African Resource Centres, located at some of the country’s most prestigious universities, is given in the 4

See https://tspace.library.utoronto.ca/handle/1807/67119 for a complete listing (accessed December 27, 2016).

5

See www2.ed.gov/programs/iegpsnrc/index.html for full details (accessed December 29, 2016).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

119

following list. Note that the centres were devoted to African studies, not African linguistics, and were sometimes coupled with African American studies. Title VI African National Resource Centres FY 2014–2017: Boston University: African Studies Center; Harvard University: Committee on African Studies; Howard University: Department of African Studies; Michigan State University: African Studies Center; University of Florida: Center for African Studies; University of Illinois, Champaign: Center for African Studies; University of Minnesota: African-American and African Studies; University of North Carolina at Chapel Hill: African Studies; University of Wisconsin–Madison: African Studies Program; Yale University: Council on African Studies. Because language was such a key component of African studies, the study of African languages was promoted either as an instrument of inquiry or as a linguistic object. Each centre awarded Foreign Language and Area Studies (FLAS) grants to students studying African languages.6 Related to this burgeoning interest in African languages was an interest in Africans themselves as students of African languages and as Africanist scholars in their own right. Africans were recruited to study at many North American universities such as those in America with centres mentioned above but also at Canadian universities such as UQAM, British Columbia, and Toronto. Several well-established scholars who decided to stay include, for example, Salikoko Mufwene, Eyamba Bokamba, Lioba Moshi, and John Mugane. Others returned to Africa where they inluenced the study of African languages in Africa, for example Gilbert Ansre, and younger scholars such as Philip Mutaka. Another important milestone in the study of African languages was the launching of Peace Corps. Established in 1961 at the beginning of the Kennedy administration, it posted its irst volunteers in two African countries (Ghana and Nigeria) by the end of the year. The number of countries served has grown to include nearly every country in Africa. Peace Corps in Africa has had a rocky history not just because of the instability and turmoil in various states, such as civil wars or coups d’état, but also due to political currents and various faux pas by volunteers. For example, in Nigeria a 1961 postcard from a volunteer occasioned calls for the expulsion of Peace Corps because of its unlattering description of living conditions in Nigeria. Peace Corps volunteers were trained in local languages using manuals developed by linguists and others, including themselves, former Peace Corps volunteers (e.g., Dwyer 1985). A number of them ended up doing research on African languages as professors in Europe, Africa, and North America. A list of a few of the early volunteers 6

In the interests of full disclosure I admit that I benefitted from four years and one summer of support from FLAS grants in Swahili and Kisi. I also benefitted from Title VI support at the Joint Center of Stanford–Berkeley for two years as the language coordinator overseeing the study of African languages at UCB.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

120

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

reads like a who’s who in African linguistics (I give only a handful of examples from the irst batch of Peace Corps volunteers in the sixties): Tom Cook, David Dwyer, John Hutchison, Will Leben, Paul Newman, and Russell Schuh.7 Thus the contribution of former Peace Corps volunteers was signiicant from its founding and continues to this day. The Canadian equivalent of the US Peace Corps is not nearly so ambitious. Known as CUSO (Canadian University Students Overseas), it was created in 1961. CUSO and the Canadian Volunteer Cooperation Program (CVCP, as part of the more general CIDA [Canadian International Development Agency]) place volunteers in only a limited number of countries.8

6.1.3

Leading Figures

As comparatist and typologist, linguist cum anthropologist Joseph H. Greenberg (1915–2001) stands out as one of the most inluential igures in North American African linguistics, to whom we owe a seminal re-classiication of The Languages of Africa (1963), which still serves as a convenient and widely accepted reference. William Welmers has been mentioned as central igure at UCLA, responsible for producing the next generation of African linguists consisting of Paul Newman, Russell Schuh, Larry Hyman, Talmy Givón, Charles Bird, and others. Charles Bird created African linguistics at Indiana, energizing the ield and also producing that early generation of African linguists, including John Hutchison, Eyamba Bokamba, and so forth. Similarly, Chuck Kisseberth at Illinois and Paul Newman at Indiana, after returning from positions in Nigeria and the Netherlands, recruited several students from Nigeria and Niger and supervised their PhD dissertations, including Mustapha Ahmad, now Vice-Chancellor in Kano and Professor Ousseina Alidou at Rutgers. Other relevant names that come to mind include Will Leben (Stanford), David Odden (Ohio State), and Poland-born Zygmunt Frajzyngier (Boulder).

6.1.4

Intellectual Currents from the Late 1960s and Onward

Black studies (African American studies) in the late 1960s led to the establishment of African American study centres or even departments, which often involved a component of Africanist study, including the learning of African languages. Both the Creolist Hypothesis (characterized below) and

7

In my department at Portland State University, we have four of nine tenure-track positions held by former Peace Corps volunteers, all of whom served in Africa (Guinea, Lesotho, Liberia [where I served 1970–1972], Tunisia).

8

In a 2005 review of the programme, there were interviews with volunteers only in Ghana and Burkina Faso, www. acdi-cida.gc.ca/inet/images.nsf/vLUImages/Performancereview6/$file/VCP%20English%20Final.pdf, albeit with the distinct programmes worldwide criticized in the report as having ‘no unified goal’ (accessed December 25, 2016).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

121

the Ebonics Controversy9 (African American Vernacular English – AAVE),10 that is, the view that AAVE was an African language (see Perry & Delpit 1998 for some discussion), promoted interest in African languages. Phoneticians and phonologists have always been interested in African languages because of their typologically unusual features (e.g., Ladefoged 1968; Lindau 1978; Traill 1973b), but other domains were slower to come around, such as studies of vowel harmony, syllable weight, serial verbs, and topic and focus (information structure). When they did, however, the scholarship was extensive. American interest in formal representations grew rapidly with the study of tone as a phonological phenomenon (e.g., Goldsmith 1972; Leben 1976),11 and later morphology using some of the same mechanisms (e.g., McCarthy 1981) and syntax (e.g., Bresnan and Mchombo 1987; a summary in Bresnan 1990).12 Because research in North America is more discipline- than areaoriented, particular topics rather than Afrikanistik in general have encouraged scholarship on Africa languages. Chief among these is tone. For example, Autosegmental Theory began with a paper (Williams [1971] 1976) later expanded into a thesis (Leben 1973) and was forwarded in another thesis (Goldsmith 1976). The study of tone continued under the leadership of such scholars as Larry Hyman (USC, UCB), Russell Schuh (UCLA), Doug Pulleyblank (UBC), Lee Bickmore (SUNY, Albany), and David Odden (Ohio State). Other areas of interest have been ideophones, noun class systems and verb extensions, argument structure in general, and general sociolinguistic topics such as language policy and multilingualism. Pidgin and creole studies grew interested in African and other substrate languages in reaction to universalist positions, especially as articulated in the work of Derek Bickerton (e.g., 1981, 1984). His proposal, known as the Language Bioprogram, posited semantic structures as something like innate Chomskyan syntactic structures, which manifested themselves in newly formed languages such as pidgins and creoles. These semantic structures appeared most notably as distinctions of tense, mood, and aspect. Much work was devoted to show the importance of the African substrate rather than innate structures. For example, in a detailed paper showing the importance of a Kru substrate in Liberian English, John Singler showed the persistence of a highly marked feature into the pidgin used by Kru speakers (Singler 1988). A second impetus for the study of African languages came from the Creolist Hypothesis (e.g., Rickford 1997, 1998), which advanced the thesis that the divergence found in AAVE was due to its creole past and its roots in Africa. A inal research initiative was a Canadian one, the Haitian Creole 9

The term ‘Ebonics’ was introduced in Smith (1975).

10

A sample characterization appears in Mufwene (2001).

11

A summary of the importance of African languages to phonological theory can be found in Goldsmith (1990).

12

An update appears in Henderson (2011).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

122

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Relexiication Project of Claire Lefebvre and John Lumsden at UQAM (e.g., Lefebvre 1998). The central claim is that Haitian is an African language, using French phonetic strings to be sure, but totally reanalysed on the basis of an African language. Their project led to a great deal of detailed syntactic work on West African languages (e.g., Lefebvre and Brousseau 2002). Other work on African languages has come from research initiatives inanced by SSHRC (Social Sciences and Humanities Research Council), such as the Relexiication Project, and various other ad hoc projects. For example, for many years Jean Lowenstamm and Jonathan Kaye at UQAM supported African graduate students and the study of African languages through such grants. Doug Pulleyblank is doing something of the same at the University of British Columbia. Before his retirement, Derek Nurse compiled a great deal of work on Bantu at Memorial University of Newfoundland, and there have been brief lorescences when an African language has been used for a ield methods course, including by Mende at the University of Toronto (Rice & Cowper 1984). This survey would not be complete without a repeated mention of Joseph H. Greenberg, who reconigured the classiication of African languages contra the prevailing orthodoxy of the time, especially by the Germans. Greenberg produced a set of widely accepted groupings that have withstood the test of time (Newman 1995), largely because he used purely linguistic criteria (see Dimmendaal 2011 for extended discussion). These contributions as well as others dealing with African languages (e.g., Greenberg 1977) signalled the importance of work on African languages in North America.

6.1.5 Studies in African Linguistics (1970) Studies in African Linguistics (SAL) was founded as an outlet for research on African languages, both descriptive and theoretical. The current characterization relects the sentiment of its founders. Studies in African Linguistics is a peer-reviewed, academic journal . . . [which] seeks to publish African language data and analysis that might not ind a place easily or suitably in more general journals. Contributions are not expected to adhere to any particular theoretical framework or linguistic theory . . . but should be data-oriented and of potential theoretical interest. Contributions may also take the form of short, descriptive grammatical sketches of endangered African languages. (http://sal.research.pdx.edu/, accessed December 26, 2016)13

SAL was created speciically in reaction to the preoccupation with formalism in the dominant generative paradigm and in an effort to raise the visibility of African languages. Early generative grammar was also preoccupied

13

This website also features electronic copies of all papers since SAL’s founding.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

123

with English (Harris 1995).14 After some time journals such as SAL and its European counterpart, Journal of African Languages and Linguistics, founded a few years later,15 provided much data and even theorizing for the burgeoning interest in linguistics. It was also the case that bastions of generative thinking such as MIT enrolled students of languages other than English, and the theory began to change. African languages came to the forefront of linguistic theorizing. Talmy Givón was the founder and irst editor of SAL; other early editors were Tom Hinnebusch, Larry Hyman, and Russell Schuh. All were associated with the Linguistics Department and the Center for African Linguistics, a Title VI centre for many years. Until 1992 the journal was housed and supported by UCLA. The journal provided a major and signiicant outlet for research on African languages.

6.1.6

Documentation of Endangered Languages (1990s and Beyond)

Americans and the American government have been concerned with the demise of many of the less widely spoken languages of the world (as articulated in, e.g., Hale et al. 1992). The interest in North America, however, has been primarily with Native American and First Nations languages, though support for the documentation of dying languages is also available for research on African languages. The forces at work on such marginalized languages have been well known for some time, including empire building, proselytizing (Childs 2007) and colonialization (Connell 2007), and lately threats of globalization (Bromber & Smieja 2004), but the extent to which African languages are threatened is underestimated (Childs, forthcoming). The study of endangered African languages has been supported by both private and governmental funding agencies available to researchers in North America, as listed below. • Documentation of Endangered Languages Program (DOBES), funded by the Volkswagen Foundation (http://dobes.mpi.nl/) • The Hans Rausing Endangered Languages Documentation Programme (ELDP) (www.eldp.net/) • Documenting Endangered Languages (DEL) supported by the National Science Foundation and the National Endowment for the Humanities (www.nsf.gov/funding/pgm_summ.jsp?pims_id=12816) • Social Sciences and Humanities Research Council (SSHRC) of Canada (www.sshrc-crsh.gc.ca/home-accueil-eng.aspx) 14

As has often been waggishly pronounced, Universal Grammar in its early days looked a lot like English.

15

As pointed out by the editor, an American linguist, namely, Paul Newman, a major contributor to African linguistics, was the initiator and first editor while at Leiden University, together with Jan Voorhoeve (a Dutchman) and Thilo Schadeberg (a German) (Newman 2010).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

124

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Other funding entities exist, but these are the ones that have funded major language documentation and sometimes revitalization projects in Africa.

6.1.7

Conclusion

Coming late to the game because of their non-colonial past, Canada and the United States have had to play catch-up. This fact, compounded by the distance of Africa from North America, has favoured more local research. Various government initiatives such as Title VI area centres and the establishment of Peace Corps in the United States have allowed for some connections with Africa to be created, and the funding for language documentation projects has helped further research on the continent. Whether these initiatives will continue into the future remains to be seen.

6.2 The Study of African Languages in Latin America (Margarida Petter) African linguistics in Latin America is still in its infancy. There are no centres, apart from the University of São Paulo (as of 1998), which offer courses in African linguistics that would study African languages as distributed and spoken in Africa. Research on African languages in Latin America focuses on manifestations of African languages, which were originally brought to the continent in the course of the transatlantic slave trade between the 15th and 19th centuries, in current or past usages in certain domains of cultural communication. A second focus is on the impact they had on the major languages spoken in the various Latin American countries, that is, most of all Portuguese and Spanish. The oficial languages of Latin American countries are originally nonnative varieties of Portuguese, Spanish, and French, resulting from intense linguistic contacts established in the context of colonization. Language contact involved the local indigenous languages, the languages of forced African diaspora, and, at a later stage, the languages of voluntary immigration from Europe and Asia. The huge population transplanted by human trade to Spanish America (1.5 million) and Brazil (5.5 million)16 resulted in (a) societies predominantly black, as some Caribbean countries, for instance, Haiti and the Dominican Republic; (b) countries with a signiicant proportion of African descent, such as Brazil, Cuba, and Colombia; and (c) nations with signiicant black minorities, like Venezuela, Peru, and Uruguay. The fact that these nations were formed in a very similar context, characterized by colonialism and slavery, has evoked an intense academic debate on the best name for this set of countries and societies, for which ‘Afro-Latin-America’ has been suggested. 16

www.slavevoyages.org/assessment/estimates (accessed March 16, 2016).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

125

The slave trade affected a vast area in sub-Saharan Africa, stretching from Upper Guinea along the West Coast to South-Central Africa. The origin of the deported people was at times mentioned in the naval documents, by indicating ports of departure or using generic names or, more rarely, names of ethnic groups. However, information on the origin was neither always correctly declared nor recognized by those who registered it. Mainly, the areas of origin of the human traficking were those regions in Africa where the languages belong to the Niger-Congo family. According to Bonvini (2008), the following languages were transferred to the Latin American region:17 a. From the West African region of high linguistic diversity came languages that belong to different groups, such as Atlantic, which includes languages like Fulani, Wolof, Manjaco, and Balanta. Other groups were Mande (Manding, Bambara, Maninka, Dyula); Kwa (Gbe: Ewe, Fon, Gen, Aja Jeje in Brazil); and Benue-Congo, namely the Defoid group, which includes Yoruba varieties that are known by the term Nagô-Ketu in Brazil and Anago/Lucumí in Cuba. b. In South-Central Africa, the languages of origin belong essentially to the Bantu subgroup, that is, along the Atlantic Coast of the current Republic of Congo, Democratic Republic of Congo, and Angola. In the case of Brazil, the catchment area was later extended to the Indian Ocean Coast (Mozambique). This region is characterized by a lesser number of languages, typologically fairly homogeneous, but spoken by most of the captives, such as Kikongo, Kimbundu, Cokwe, Umbundu, Luba, Lunda, Makua, Kwanyama, and Herero. The impact of African languages in Latin America, in addition to being the sources for considerable borrowing of African lexicon, provoked two prototypical linguistic situations: the emergence of Creole languages and the formation of restructured varieties of European languages. Quite surprisingly, African languages in Latin America have played only a small role in the creation of Creole languages: the IberianRomanic-based Papiamento (spoken in Aruba, Curaçao and Bonaire) of still unclariied sources (Spanish? Portuguese?), French-based Haitian Creole, and Spanish-based Palenquero in Colombia. Palenquero is the only Spanish-based Creole in South America in San Basilio de Palenque, a city located at the southwest of Cartagena, formed by runaway slaves of Bantu linguistic stock, that is, speakers of Kikongo. The Creole of Haiti has the biggest number of speakers among the Caribbean creoles. It is spoken by almost all the population of the country. It was formed by contact between French and West African languages (Fon, Ewe, Wolof). 17

The data about the African languages are from Bonvini (2008:30–31) for Brazil, but they can be extended, to a large extent, to many of the other Latin American countries.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

126

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

Cuba and Brazil, being the countries with a larger intake of enslaved black populations, did not develop a Creole language, despite hosting a great population of Afro-descendants. In these two countries, the type of language contact involved a greater impact of the European language of the masters on the language(s) of the enslaved. This corroborates the analysis of Mufwene (2008), who afirmed that it is not the mere fact of having a large population of African descent but socioeconomic factors that would favour the emergence of a creole. In Brazil, in the past the question was discussed whether the Portuguese language underwent creolization; this position was defended by some scholars (Guy 1981; Baxter 1992; among others), who today just afirm that in Brazil there was an irregular linguistic transmission. The Creolization hypothesis was contested by authors who do not deny the importance of the linguistic contacts, but argue that the notable changes in Brazilian Portuguese could already be foreseen to happen in the language when it was transplanted from Europe to America (Naro & Scherre 2007). In Cuba and Brazil, the social situation favoured the learning and use of Spanish or Portuguese language, both by the bozals, that is, Africans with little or no command of the language of the colonizers, and by the ladinos, Africans who arrived at the American colonies already speaking Spanish or Portuguese. They had learned these languages before shipment in the relatively long periods during which Africans of different ethnic groups would remain in the slavery warehouses and castles, while waiting for the loading of ships to be completed. Although today no African language has a community of speakers in Latin America, traces of these languages remain alive in two main spaces: Afro-Latin rituals and communities consisting predominantly of Afro-descendants.

6.2.1

Ritual Languages

In the Afro-Latin cults, African languages have a liturgical function and their use is restricted to the initiated ones and practitioners. Speaking of African ‘languages’ here refers to the use of a specialized lexicon related to the religious universe. In Brazil, the cults are generically referred to as candomblé. In each one of these cults, there is a type of organization and language, which distinguishes the different types of Candomblé, classiied in accordance with the ‘nations’, that is, according to the linguistic predominance represented: Nagô-Ketu, which is Yoruba based; Jeje-Mine, which is Ewe-Fon based; Angola, Congo-Angola, which is Bantu based (Kimbundu/Kikongo/Umbundu). Often, the cults are completely enounced in the African ‘language’, such as in the chants and in some ritual formulas. However, what is African is the lexicon, which is used in sentences constructed with the Portuguese language syntax (Petter 2012).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

127

In Cuba, the terms ‘Afro-Cuban religions’ and ‘of black-African matrix’ involve an immense variety of religious manifestations brought from the sub-Saharan Africa, syncretized with the Catholicism of the colonizers. This makes it a rather wide term that evokes hundreds of different manifestations. Diverse sub-Saharan languages mixed with the Spanish language and became an important part of the linguistic features of the four Afro-Cuban cults that persist until today in Cuba (Bernal 2009:60–61). These are Regla de Palo Monte, with cultural and linguistic models from the region of Congo, with predominance of Kikongo; Regla de Ocha and Ifá, known as Santería or Lucumi, with cultural and linguistic models of Yoruba, Ewe, and Fon, languages of the south of Nigeria and Benin; Regla Arará, with cultural and linguistic models of Ewe-Fon and Abakwa Society, a reconstitution in Cuba of the Egbo society among the Eik and Ibibio of Old Calabar, in Nigeria.

6.2.2

The Contemporary Communities with Afro-Latinate Speeches

Communities of people of African descent, in addition to the religious realm, form another realm of search for a free life. These comprise the black agricultural communities, or Quilombos, as they are currently called in Brazil. In Venezuela, they are called Cumbes. Palenques is the name by which they are called in Cuba and Colombia. The terms Cimarrón and Cimarronage, equivalent to Quilombola in Brazil, expand across all Latin America, meaning the resistance against oppression (Petter 2012). In these realms different African cultural codes were reconstructed and transformed and assumed the functions of secret languages; of communities of Creoles speakers (Palenquero); or communities in which, in the words of Lipski (2015:94) when referring to data from Afro-Hispanic speeches, persist ‘diluted features of what could have been authentic restructured dialects’. In Brazil, the so-called secret languages are sheltered in communities consisting of Afro descendants. Examples are Cupópia of the Cafundó, located in São Paulo (Vogt & Fry 1996), and, in Minas Gerais, language of the black of the coast (Língua do Negro da Costa), of Tabatinga (Queiroz 1998), and Calunga, the code used by some inhabitants of Patrocínio (Byrd 2006). In Panama, communities of black Congos, distributed throughout the Panamanian coast, conserve a ceremonial language with elements of the ‘bozal’ speech from the last centuries. According to Lipski (2015:101), from the point of view of its usage, the Congo speech resembles Cupópia and Calunga speech forms, based on the fact that it was also created with the intention of keeping secret the messages and to express revolt against the whites. Diverse social contexts reacted differently to the contact phenomenon, but, despite different circumstances, African languages and cultures, though transformed, remained a distinctive element of Latin American

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

128

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

identity. In the non-Creole regions, their linguistic features become most visible in the lexicon. It cannot be excluded that African languages also had an impact on the grammar of local varieties of Spanish and Portuguese (Petter 2011:92). Presently in Brazil, many studies are being undertaken on language contact in the domain of syntax (Negrão  & Viotti 2008, 2014; Avelar et al. 2009; Avelar 2015; Galves 2015). With regard to Afro-Hispanic vernacular varieties, works to be consulted are, among others, Lipski (2011, 2015) and Sessarego (2015).

6.3 African Linguistics in Asia and Australia 6.3.1

Introduction

Africa is distant from most Asian countries and Australia. Unlike Asian language studies, no association related to African language studies has yet been established in Asia; therefore, it is relatively dificult to gather information about scholarship on African linguistics as a whole in Asia. In this regard, Japan is an exceptional case in that studies of African languages are actively carried out there. In terms of research on African languages, Japan is followed by China, South Korea, and Australia. In other countries, like India, at least important African languages such as Swahili are taught.

6.3.2

African Linguistics in Japan (Shigeki Kaji)

Among the Asian countries, Japan is where African language studies are pursued the most actively. More than 30 researchers specialize in African languages (not counting PhD students) and hold positions in linguistics or African studies departments, which are scattered all over the country. One important characteristic of Japanese African linguistics is that these researchers all conduct ieldwork, and they are more data oriented than theory oriented. Their research began after World War II, with a few exceptions here and there.

6.3.2.1 The Research Institute for Asian and African Languages and Cultures (ILCAA) Among the many universities and research institutes in Japan, the Research Institute for Asian and African Languages and Cultures (ILCAA) is by far the most important.18 This national interuniversity research institute was founded in 1964 at the Tokyo University of Foreign Studies (TUFS) on the recommendation of the Science Council of Japan. Its objective was irst of all to acquire direct knowledge of Asian and African languages and cultures through ieldwork. Young researchers were recruited and sent 18

For ILCAA publications, see www.aa.tufs.ac.jp/documents/publ/ILCAApubl2016.pdf.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

129

to various countries. Its publications on African languages are extremely valuable for the scholarship and noteworthy for the number of volumes published and the variety of languages discussed. Yukio Ishigaki (1931–1983), a professor at ILCAA, was among the irst generation of African language researchers in Japan. He studied Amharic and other Semitic languages and Somali. Tsuneo Morino (1933–2003), a Swahili specialist, compiled six volumes of a Swahili dictionary in collaboration with Hisashi Nakajima, of the Osaka University of Foreign Studies. Akio Nakano (1937–2008) was an eminent ieldwork scholar whose work covered all the areas of northern Africa and who studied Hebrew, various Arabic dialects, Berber, Amharic, Tigre, Somali, and so on. Yasutoshi Yukawa (1941–2014) worked successively at ILCAA, the University of Tokyo, Kumamoto University, and Teikyo Heisei University. He did research on more than one hundred Bantu languages all over the Bantu area. His main research focus was on tonal matters. Shuji Matsushita studied Hausa and other Chadic languages. Ryohei Kagaya is a phonetician who investigated the Khoisan languages of Botswana and the Bantu languages of Eastern and Southern Africa, such as Kikuyu and Shona. Shigeki Kaji investigated the Bantu languages of DR Congo and Uganda, such as Tembo, Hunde, Tooro, and Nyoro, among others. Daisuke Shinagawa, who recently joined ILCAA, is a specialist in the Chaga languages of Tanzania. Yuko Abe, whose research is the Bende/Tongwe languages of Tanzania, moved from ILCAA to Tokyo Women’s Christian University 2017. Masaaki Kadoya, a research associate of ILCAA, studies Malila and Nyiha of Tanzania. Hiroshi Nakagawa, a specialist in Khoisan languages, works at the TUFS, and his wife, Hitomi Ono, also specializes in Khoisan languages at Reitaku University. Kiyoshi Shimizu graduated from the TUFS and studied Nigerian languages. He moved to Kumamoto University of Japan upon leaving Vienna University. Masayuki Nishie (1937–2015) worked at the TUFS before moving to Waseda University, the school he graduated from. He wrote many excellent ield essays based on his experience in Swahili, Masai, Kikuyu, and others. Haruko Sakaedani graduated from the TUFS and studies Egyptian Arabic from pedagogic and sociolinguistic points of view.

6.3.2.2 Other Universities and Research Organizations The linguistics department of the University of Tokyo is the oldest in Japan and has produced many specialists in African languages. However, it is not a centre of African language studies. Researchers have studied at other universities, especially ILCAA. Kyoto University, the second-largest national university after the University of Tokyo, has a long tradition of African studies, and its Department of Linguistics produced several specialists in African languages. Masaoki Miyamoto, whose area was Swahili and the African novel, occupied the chair in Swahili when the Department of Swahili and African

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

130

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

languages was irst created in 1986 in Japan at the Osaka University of Foreign Studies. Kazuhisa Eguchi (1942–2008), of the National Museum of Ethnology, devoted his energy to the Fulfulde language and folktales. Shigeki Kaji irst got a job at ILCAA, and then moved to the Graduate School of Asian and African Area Studies (ASAFAS) at Kyoto University. He now works at Kyoto Sangyo University. Osamu Hieda studied Nilotic languages such as Datooga, Acholi, and Kumam. He was a member of the Department of Swahili and African languages at the Osaka University of Foreign Studies. Later he joined ILCAA. Maya Abe, of the ASAFAS, studies Ma’a, a mixed language of Tanzania, from descriptive and sociolinguistic points of view. Shuichiro Nakao, a researcher in Juba Arabic, has moved from the ASAFAS to Osaka University. Toshiro Kamiya, a specialist in Zulu and related languages, works as a research administrator at Kyoto University. Yukitoshi Sunano, now at Kumamoto Prefectural University, engages in the study of multilingualism in Africa. The Department of Swahili and African Languages at Osaka University is the sole department where undergraduate students can specialize in Swahili. There are currently three scholars in this department, namely Nobuko Yoneda, Junko Komori, and Keiko Takemura. As instructors of Swahili, they do research on other languages: Yoneda on Bantu Matengo and Herero, Komori on Yoruba and Kerewe of Tanzania, and Takemura on the Swahili dialects of Zanzibar. Katsuhiko Shiota holds a part-time position at Osaka University. He is a specialist in Chadic languages who studies Hausa and Bura as well as Yoruba. Kumiko Miyazaki, a graduate of the university, studies Zanizibar Swahili as she lives there. The Osaka University of Foreign Studies merged with Osaka University in 2007. Tadahisa Goto (1911–2010), who was a professor of English, published the irst Swahili grammar written in Japanese. Tenri University is a religious-afiliated university and it missionizes in Africa from its base in Brazzaville. Toshio Jikihara compiled a series of African language textbooks and vocabularies, such as those on Lingala and Lari. Yoichi Wazaki (1920–1992), an anthropologist at Tenri University, produced a signiicant Swahili-Japanese dictionary based on his long-time ield experience in the Lake Eyasi area of Tanzania. Mitsuo Kowaki, of Kyoto Sangyo University, studies Ethiopian languages. Yoichi Tsuge, of Kanazawa University, investigates the Omotic Ari language of Ethiopia. Hideyuki Inui, of Yamaguchi University, conducts a typological study of Ethiopian languages on a ield survey basis. Jun Ikeda, of the University of Tsukuba, studies the Semitic languages of Ethiopia, such as Mesqan of Southern Ethiopia. Kyoko Koga, of Kochi University, studies Akan of Ghana. Motomichi Wasaka, of Meisei University, works on the Semitic and Cushitic languages of Ethiopia, such as Wolaitta. Hiroshi Yoshino, who now has a position at an academic recruitment company, studies the East Cushitic languages of Ethiopia, such as Ale. Kazuhiro Kawachi, a professor of English at the National Defense Academy, does ieldwork on Sidaama of Ethiopia and Kupsabinyi of Uganda from a theoretical Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

131

syntactic point of view. Sumiyo Nishiguchi of Tokyo University of Science studies Dhaasanac, a Cushitic language of Ethiopia and Kenya. Yukiko Morimoto, of Humboldt University of Berlin, conducts theory-based Bantu studies. Yasuko Nagano-Madsen, a professor of Japanese at the University of Gothenburg, Sweden, carries out phonetic studies on Asian and African languages including Mpiemo of the Central African Republic. Ritsuko Miyamoto is a professor at Akita University. After studying Fulfulde of Nigeria and Suba of Kenya, she now concentrates on the sign language of Kenya. Nobutaka Kamei, of Aichi Prefectural University, extensively studies sign languages in Africa. Malagasy languages/dialects are also studied in Japan by Ritsuko Kikuzawa, of the National Museum of Ethnology; Nobukatsu Minoura, of the TUFS; and Noa Nishimoto, of Kyoto University. They investigate those languages/dialects from a wide Austronesian perspective.

6.3.3 African Linguistics in China (Sun Xiaomeng) 6.3.3.1 Introduction Systematic African studies in China began only in modern times and were not institutionalized until the founding of the People’s Republic of China. At the initial stage, the motivation for these studies was pragmatic and politically motivated rather than academic. Following this logic, African linguistics in China was quite marginalized, and academic achievements in this ield are more limited than, for example, studies of Sino-African relations and contemporary African issues. Nevertheless, many achievements are worth mentioning and some of the aspects will be highlighted in this chapter.

6.3.3.2 African Languages Teaching in China Undoubtedly, African language teaching in China has been quite impressive. From a historical perspective, both Beijing Foreign Studies University (BFSU) and the Communication University of China have done pioneering work, which can be considered the precursor for African linguistics. Since 1960, Swahili, Hausa, and Zulu have been taught despite a lack of adequately trained lecturers, published language-teaching materials and dictionaries. The motivation also shared a parallel path, China dispatching several batches of diplomats to acquire African vernacular language proiciency, mainly in Swahili, Hausa, Amharic, Lingala, Somali, and Malagasy. With the turn of the millennium, a fast development of Sino-African relations was witnessed, African studies in China became consequently reinforced. More universities joined the initiative of language programmes. The most comprehensive work is located at BFSU, which has developed into a university of international reputation for language teaching and research, covering the most important languages of the African continent: Swahili, Hausa, Amharic, Malagasy, Zulu, Somali, and Yoruba. The substantial increase of African languages teaching makes a strong contribution to this ield. Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

132

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

6.3.3.3 Publications and Achievements During the initial stage of African linguistics in China, focus was placed on the translations of works by foreign scholars, introducing general information about African languages. Chinese scholars also started to publish some introductory papers, all of which laid a foundation for further research in Chinese academia. Considering the importance of African language teaching in China, the publication of teaching materials including textbooks, dictionaries, and grammar books is an integral part of the achievements in this ield, providing resources for language instruction to support students’ language competence. In more detailed research, some of the prominent typological features in African languages are analysed in comparison with Chinese language and dialects. Other prominent topics in current research include multilingualism, language policy, language and nation building, language and development, and language and Sino-African relations.

6.3.3.4 Challenges and New Trends Despite the achievements mentioned above, challenges remain. These include, irst, that China should reorient African linguistics in terms of disciplinary setting and capacity building of expertise. Second, due to language barriers, linguists in China frequently use secondary sources. Third, the range of thematic foci and disciplinary coverage should be expanded, ieldwork with vernacular language ability is much needed, and it is necessary to establish systematic and in-depth research agendas. The vulnerable nature of the subject needs long-term strategic planning for development, more government funding, and relevant policy assurance. It is also very inspiring to mention some new trends. On the national level, China has set the goal of offering all the oficial languages of countries that have established diplomatic ties with China, stimulating considerable capacity building in African languages. BFSU has taken this initiative forward, and Tigre, Shona, Tswana, Comorian, Kinyarwanda, Kirundi, Wolof, Sesotho, Berber, and Creole language programmes are on the university’s agenda and will be offered in the coming ive years. Meanwhile Chinese young scholars have shown their academic capacity inside and outside of China, and their research is built signiicantly on the command of vernacular African languages. Several other students are also engaged in African linguistics as part of transnational education, either as Chinese students abroad or as African students in China. China’s growing engagement in Africa will generate more demand for African linguistics. It is a very promising ield, and there are many opportunities for future development and collaboration.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

133

6.3.4 African Linguistics in South Korea (Yang Chul-Joon) 6.3.4.1 Introduction The institutionalization of African linguistics in Korea coincided with the establishment of the Department of Swahili in 1983 at Hankuk University of Foreign Studies (HUFS). The principal motivation for establishing the department was practical, in order to train Swahili experts. Unlike China and Japan where African languages are taught at national universities, HUFS is a private university where presently 45 languages of the world are taught. In the initial years of the department, emphasis was placed on improving the skills of reading, writing, listening, and speaking Swahili. Swahili was not a popularly known language among the general public in Korea. However, signiicant achievements were realized in Swahili studies, including the publications of essential teaching materials and academic articles and books. Swahili studies further acted as a harbinger for the addition of more African languages and a systematic study of African linguistics.

6.3.4.2 From the Heyday of Swahili Studies to African Linguistics The Department of Swahili was reorganized in 1988 as the Department of African Languages, adding two more African languages, namely Hausa and Zulu. It encompassed three divisions: East African, Southern African, and West African Studies. The concomitant departmental course syllabus readjustments in 2008 were a further advancement in studying African languages. The teaching of three African languages facilitated various other disciplines such as anthropology and literature to take roots since an erudite knowledge of these languages constitutes a fundamental prerequisite for delving into these disciplines. Scholars from anthropology and literature joined efforts in the teaching of African lingua francas and signiicantly contributed to the scientiic dissemination of these languages by participating in the publication of instructional materials and dictionaries. Gradually, research orientation shifted to broader perspectives in terms of comparative-historical linguistics, linguistic typology, language contact, and related topics.

6.3.4.3 Major Research Topics Owing to the limited number of scholars involved in the study of African languages, research topics covered are not as varied as at European and North American universities. Scholars have carried out theory-oriented as well as ield-based research on comparative-historical linguistics, language typology and universals, grammaticalization, language contact, and descriptive

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

134

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

linguistics. In addition, research on corpus-driven and ield-based variation analysis and the relationship between language and society has been conducted. With the introduction of Hausa, research on the Chadic languages of the Afroasiatic language family has been made possible.

6.3.4.4 Prospects and Challenges African linguistics in South Korea faces several challenges that deserve to be addressed for the sustainable promotion of the ield. First, the elicitation of primary data and collection of information by ield research is of signiicant importance for quality research in order to ensure originality and novelty and avoid tautological reproduction. Second, continued institutional support needs to be guaranteed to ensure that African linguistics contributes to the advancement of humanities as a whole. Third, scholars need to address thematic diversity in African linguistics and accommodate the latest research trends and current topics in African linguistics. Fourth, language is socioculturally situated as the result of social, cultural, and political processes. This calls for a ine-grained knowledge of African society and culture. Last, African languages referred to as Nilo-Saharan and Khoisan need to be included in the research agenda. A promising prospect is that the general public has become aware of the importance of studying African languages. It is expected that Amharic and Kinyarwanda will be included in a syllabus and taught at HUFS in the near future, in close collaboration with the Ministry of Education.

6.3.5

African Linguistics in Australia (John Hajek)

It is not surprising that African linguistics has not been a particularly visible discipline in Australia. Australian-based linguists have for the most part been focused on languages of Australia and the wider Asia-Paciic, as well as on English and other European languages. As a result, there has never been in Australia an academic programme or institute dedicated speciically to African languages. Instead, Australian activity in African linguistics must be drawn together through a series of different threads, some of which are noted here. In the irst instance, there is the activity of a small number of individuals who have worked on African languages in addition to their work on other languages for which they are typically much better known. In some cases, these are overseas linguists with Africanist experience who at some point took up academic positions in Australia, especially in the earlier days of linguistics.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

African Linguistics in the Americas, Asia, and Australia

135

These include (a) Patrick McConvell, Kevin Ford, and Paul Black, who completed PhDs on African languages overseas in the early 1970s before eventually working on Australian languages at the School of Australian Linguistics at Batchelor and elsewhere; (b) Edith Bavin, who irst published on Lango before arriving in Australia in 1981; and (c) Keith Allan, who spent extended periods at African universities before arriving at Monash University in 1978. Mary Laughren (PhD, Nice, 1973) appears to have been the irst Australian to have completed a doctoral dissertation on an African language, in this case Senufo, before returning to Australia. More recent contributions to different aspects of African linguistics from other non-African linguists permanently employed in Australia include Steven Bird, Bruce Donaldson, and John Hajek (University of Melbourne), Simon Musgrave (Monash University), and Aniko Hatoss (University of New South Wales). A few linguists of African origin previously or currently resident in Australia are well known for their work in African linguistics. Felix Ameka, originally from Ghana, completed his master’s (1986) and doctorate (1991) on Ewe at the Australian National University in Canberra, before taking up a permanent appointment in Europe. Paulin Djité was for many years (1989–2011) at the University of Sydney and then at the Western University of Sydney. During this time he published extensively on language policy issues in Africa and elsewhere before returning to Côte d’Ivoire. Likewise, Ibrahima Diallo, currently at the University of South Australia, and Finex Ndhlovu at the University of New England have been very active in the areas of African language policy and identity issues. Mengistu Amberber at the University of New South Wales has published extensively on different aspects of Amharic. Since the late 1990s, the most consistent source of activity on African linguistics in Australia at a more general level has been through the Research Centre for Linguistic Typology (RCLT), based irst in Canberra (1996–1999) and then at La Trobe University in Melbourne (2000–2011). Its activities have continued partly through the Centre for Research on Linguistic Diversity (CRLD) at La Trobe University, but more directly through the Language and Culture Research Centre (LCRC, irst established as the Language and Culture Research Group in 2009) at James Cook University. Well-known Africanists, including Felix Ameka, Mechthild Reh, Anne Storch, Gerrit J. Dimmendaal, Mauro Tosco, Paul Newman, and Zygmunt Frajzyngier, amongst others, have been regular visitors to the RCLT and LCRC, contributing also to annual international workshops and the edited volumes that have resulted from these. The three centres have also hosted a number of trained Africanists as postdoctoral fellows, such as Birgit Hellwig (Goemai, Katla) and Yvonne Treis (Baskeet), who continued to work primarily on African languages while in Australia, while others,

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

136

PA RT I

S H O RT R EG I O N A L H I STO R I E S O F A F R I C A N L I N G U I ST I C S

for example, Knut Olawsky (Dagbani) and Renée Lambert-Brétière (Fon), continued to publish on African languages while working on grammatical descriptions of non-African languages.

Acknowledgements G. Tucker Childs thanks his many Canadian colleagues, in particular Bruce Connell, Doug Pulleyblank, and Sharon Rose, for their assistance on this chapter, especially with the section on Canadian research projects.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:33:52, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.006

7 Comparative African Linguistics Gerrit J. Dimmendaal

7.1

Introduction

The historical-comparative study of African languages involves at least two dimensions: irst, the identiication of genetic relationships between languages; and second, the application of the comparative method to languages assumed to be genetically related, in order to establish historical changes between them. Section 7.2 addresses the successes and drawbacks of this continent-wide endeavour to date, while Section  7.3 discusses some of the results emerging from in-depth studies of well-established language families. Internal reconstruction and dialectology complement the historical-comparative method, as discussed in Sections  7.4 and 7.5, respectively. Distinguishing between borrowing and shared inheritance from a common ancestor is not always easy, as shown in Section  7.6, where different contact scenarios between genetically related as well as unrelated languages are discussed within an African context. While, traditionally, linguists have represented genetic relationships in terms of family trees, more recently network analyses have come into vogue. Whereas the former cladistic approaches relect tree-branching or divergence, the latter rhizotic models visualize convergence between genetically related languages, as shown in Section  7.7. This chapter concludes with some observations (Section 7.8) on possible future directions for the ield.

7.2

The Historical Comparison of African Languages

The gradual identiication of different languages outside Europe, mainly a result of European colonialism during the 18th and 19th centuries, led to the publication of collections of vocabularies for African languages. For example, Simon Pallas’s Linguarum totius Orbis Vocabularia Comparativa, a study published in two volumes in 1787 and 1789 and containing word

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

140

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

lists for Amharic, Fulani, Igbo, Khoekhoe, or Mandinka (and other African languages), was made possible with inancial support from Catherine the Great, empress of Russia, who initiated this comparative endeavour (Fodor 1975:7). During her reign, word lists were distributed via Russian ambassadors, and subsequently also reached Dutch, British, and Spanish colonies. Of course, European investigators were not the irst scholars in the ield. Scientiic knowledge had declined in the Middle Ages in Europe, until the 12th century, when scholars in Southern Europe recovered ancient Greek science through successive translations from Arabic and Hebrew, often with commentaries from scholars speaking these latter languages. These contributions to mathematics, medicine, or astronomy are well known, but scholars from the Middle East were also among the earliest investigators in comparative linguistics. The awareness of a relationship between languages, resumed today under the label ‘Semitic’ on the one hand and Berber on the other, probably dates back to as early as the 10th century, when a physician by the name of Judah ben Quraysh, who lived in Morocco, had apparently already identiied structural similarities between them (Becker 1984, quoted in Frajzyngier & Shay 2012a:4). The French scholar Postel, who was familiar with Arabic and Hebrew texts, published a study in 1538 on the origins of the Hebrew language, thereby de  facto identifying Semitic, although this label was not used as a common denominator for the family until the end of the 18th century. Today, we know that Semitic is part of a larger macro-family or phylum, called Afroasiatic, which includes not only Berber but also Chadic, Cushitic, Egyptian, and Omotic. As shown in the elegant survey by Hayward (2000), formal similarities, particularly involving case marking, conjugational features of the verb, personal pronouns, and plural markers point towards the common origin of these groups as Afroasiatic languages. Grammatical features tend to be more conservative (and less subject to borrowing) than lexical features, and so it should not come as a surprise that there are few obvious lexical cognates for Afroasiatic as a whole, given the time depth involved for this phylum; see also Chapter 10. Similarities between two or more languages are particularly signiicant when they can be observed in basic vocabulary, involving words for body parts or verbs such as ‘drink’, ‘eat’, or ‘walk’. Such lexemes are less likely to be borrowed, although borrowing is possible, as discussed in Section 7.6. By taking on a ‘wide-angle’ perspective, that is, by comparing a large number of languages at once rather than proceeding pairwise in a stepby-step fashion, it is usually easier to identify lexical (and grammatical) similarities, as the following example with the word for ‘goat’ in Swahili, a language of wider communication in East Africa, and Tima, a distantly related language spoken in the Nuba Mountains in Sudan, should help to illustrate. (1)

Swahili mbuzi

Tima cɪˊmɪ̀ɪ̀

‘goat’

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

141

Multilateral (or mass) comparison of vocabularies between a range of languages helps to identify similarities that are not obvious when comparing just two languages. Only when comparing other languages closely or more distantly related to Tima (and classiied as members of the Kordofanian branch of Niger-Kordofanian by Greenberg 1963), does it become obvious that the two words in Swahili and Tima are in fact related. The oldest generation of Tima speakers refers to ‘goat’ as cI´mɪ`ð. This form is remarkably similar to the word for ‘goat’ in languages like Tegali and Tumali, which are also spoken in the Nuba Mountains. (2)

e-mbʊt yimbʊte

‘goat’ (Tegali) ‘goat’ (Tumale)

The Swahili form is widespread across a range of languages now considered to form the Bantu family within Niger-Congo (compare the examples for ‘goat’ in (1)), and Meeussen (1969) reconstructs *-búdi ̧ for their common ancestor, Proto-Bantu. In his contribution on the grammatical reconstruction of Proto-Bantu, Meeussen (1967:102) reconstructs a (homorganic nasal) preix as well as a pre-preix or augment *ji- for the noun class to which this lexical root belonged. The emerging original Proto-Bantu form for ‘goat’, *ji-m-búdi ̧, is remarkably similar to the form in the Niger-Congo languages spoken in the Nuba Mountains in Sudan, illustrated in (2). When one observes hundreds of roots and grammatical morphemes with similar forms and meanings, especially involving basic vocabulary, a borrowing scenario becomes extremely unlikely. Nevertheless, disentangling borrowing from shared inheritance is not always easy, as further discussed in Section 7.6. Apart from genetic inheritance and borrowing, chance is another factor explaining the occasional lexical similarity between two languages. Sound symbolism (including onomatopoetic words) may also result in similarities, without necessarily relating to a common historical origin. As shown by Blasi et al. (2016), there are interesting and rather unexpected worldwide sound-meaning biases, for example the frequent occurrence of the vowel ‘e’ and the consonant ‘l’ in the word for ‘tongue’: the corresponding forms in Swahili, u-leme, and Tima, k̀ĺŋ̀̀, conirm this. It was on the basis of multilateral comparisons of basic vocabulary, together with grammatical patterns (for example verb paradigms), that Koelle (1854) published the most extensive historical-comparative study of African languages to that date, involving a comparison of more than 150 languages, based on interviews with liberated victims of the Atlantic slave trade in Sierra Leone and Liberia. Koelle (1854) identiied 11 language families, alongside several ‘unclassiied and isolated languages’. During subsequent decades and in the irst half of the 20th century, a range of scholars aimed at continent-wide classiications of African languages. Among them was Lepsius (1880), whose contribution is characteristic of the era in which it was written. Lepsius distinguished between Bantu, constituting a so-called Southern zone, as a genetic group using Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

142

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

preixes, and Hamitic and Semitic languages as languages of the so-called Northern Zone, with sufixes and gender distinctions. Based on the dominant hypothesis of those days that languages develop from morphologically complex into more reduced systems, Bantu languages were assumed to represent an original stage, whereas languages of the so-called Middle Zone were assumed to be a mixture of Bantu and Hamitic languages, so-called ‘Mixed-Negro languages’. This Middle Zone contained ‘proto-African’ languages and iniltrating Asian languages, but also Bushman and Hottentot. For reasons of space, it is not possible to discuss the contributions of other scholars during the 19th and 20th centuries here. The interested reader is referred to Childs (2003:19–54), Doneux and Rey (2003), Dimmendaal (2011:75–92, 307–331), and the companion volume to the present handbook (Wolff 2019) for detailed discussions. It is probably fair to say that, with few exceptions, much of the classiicatory work on African languages during the irst half of the 20th century was characterized by a mixture of genetic and typological classiications and marred by racist connotations. It was the late Joseph Greenberg (1915– 2001) who tried in a range of articles between 1949 and 1954, published as a monograph in Greenberg (1955), to break with this tradition, and to arrive at a continent-wide genetic classiication using multilateral (or mass) comparison without the misleading ‘noise’ of typology and racial or cultural considerations. While he established 16 independent families in his earlier studies, he hypothesized four phyla (stocks, or macro families) in his follow-up study, Greenberg (1963): Afroasiatic, Khoisan, Niger-Kordofanian, and Nilo-Saharan. Tables 7.1–4 summarize this classiication, as well the critiques and alternative hypotheses formulated by various scholars in response to Greenberg’s seminal contribution. In addition, the tables provide information on the application of the comparative method (discussed below) to subgroups within these hypothesized macro-families. As these tables show, there is disagreement among scholars about the proper classiication of a range of groups. Nevertheless, Greenberg’s (1963) Table 7.1 The classification of Afroasiatic by J. H. Greenberg and subsequent views Greenberg (1963)

Disputed affiliations

Afroasiatic

• The genetic affiliation of • All subbranches, i.e., Berber, Omotic (called Western Chadic, Cushitic, Omotic, Cushitic in Greenberg 1963) Semitic is disputed by • Historical development of some scholars Egyptian known through • The Eastern branch within historical documents; the Omotic is not part of latest stage of Egyptian, Omotic, according to Coptic, is still used as a some scholars liturgical language

Berber Chadic Cushitic Egyptian Semitic

Comparative method

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Table 7.2 The classification of Khoisan by J. H. Greenberg and subsequent views Comparative method

Greenberg (1963)

Disputed affiliations

Khoisan

• Northern, Central, and Southern • Central Khoisan Khoisan assumed to constitute independent families by specialists • Hadza is a linguistic isolate • Sandawe probably forms a genetic unit with Kwadi and Central Khoisan

Northern Khoisan Central Khoisan Southern Khoisan Sandawe Hadza

Table 7.3 The classification of Niger- Congo by J. H. Greenberg and subsequent views Greenberg (1963) NigerKordofanian (referred to as NigerCongo in more recent studies)

Disputed affiliations

Comparative method

West Atlantic • The genetic unity of Atlantic • Mande (Greenberg’s West Atlantic), Gur in particular the status of the Kwa Bijogo group, is disputed • Benue-Congo • Mande and Ubangian (called • Adamawa‘Eastern’ in Greenberg 1963) • Eastern may constitute independent • Kordofanian families (although the Mba group within Ubangian may be part of Niger-Congo) • • Dogon as an independent family rather than as a subgroup of Gur • Ijo (and Defaka) as an independent family rather than as a subgroup of Benue-Congo • Kadu (called ‘Tumtum’ in Greenberg 1963) as an independent family • Kordofanian probably consists of a number of distantly related Niger- Congo branches rather than forming a single branch

Northern branch and the Bak branch within Atlantic Gur Mande Kwa Bantoid and other subgroups of Benue-Congo Subgroups of Kordofanian

Table 7.4 The classification of Nilo-Saharan by J. H. Greenberg and subsequent views Greenberg (1963) Nilo-Saharan

Disputed affiliations

Songhai • Songhai (Songhay) as an Maban independent family Fur • Koman (=Coman), as an Saharan independent family Chari-Nile • Mimi of Nachtigal and Mime Koman of Gaudefroy-Demombynes as isolates rather than as members of the Maban group • Gumuz as an isolate rather than as a member of Koman • Kuliak (treated as a branch of the Eastern Sudanic subbranch of Chari-Nile in Greenberg 1963) as an independent family

Comparative method • • • •

Songhay Maban Central Sudanic Eastern Sudanic as a branch of Nilo-Saharan, with sub-branches: Daju, Nilotic, Nubian, Surmic, Taman • Koman and B’ata

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

144

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

classiication still forms the heuristic basis for conferences and workshops on African languages, as well as for the classiication of monographs and subject indexes in university libraries. Presumably, all human languages are related; one cannot prove that they are not. The point is to present evidence that they are, and scholars disagree on what constitutes ‘suficient’ evidence. Some historical linguists mention 12,000 years as an upper limit for the establishment of genetic relationships between languages, as much of the inherited lexicon and grammatical structure will have been replaced within this time span. A classiication that treats, for example, Fur as part of Nilo-Saharan (as Greenberg did in his 1963 publication), rather than as an independent family (as in Greenberg 1955) in and by itself is not very interesting unless this helps to explain grammatical phenomena historically. As argued in Chapter  11 on Nilo-Saharan, several grammatical features in Fur do indeed ind a natural historical explanation when comparing these with cognate forms in Kunama and other groups, all of which are hypothesized to belong to the same phylum in Greenberg (1963). Over the past decades, several linguistic isolates, that is, languages without any obvious relatives, have been identiied as well. These are listed in Table  7.5 (see also Map 7.1). Hadza, for example, was classiied by Greenberg (1963) as a Khoisan language, but the comparative evidence for this claim remains rather weak.

7.3

The Comparative Method

Apart from multilateral (or mass) comparisons, which were used in the past in order to come to grips with the classiication of the roughly 2,000 languages spoken on the continent, the comparative method has been applied successfully to well-deined language families, which may or may not belong to larger phyla or stocks such as those hypothesized by Greenberg (1963). This classical method was irst applied within Indo-European, but it Table 7.5 Linguistic isolates Bangi Me (Mali) Dompo (Ghana) Ega (Ivory Coast) Gola (Liberia, Sierra Leone) Gomba (Ethiopia) Gumuz (Ethiopia, Sudan) Hadza (Tanzania) Irimba (Gabon) Jalaa (Nigeria) Kujarge (Chad) Laal (Chad) Limba (Sierra Leone, Guinea) Luo (Nigeria)

Mawa (Nigeria) Meyobe (Benin, Togo) Mimi of Decorse (Chad) Mimi of Nachtigal (Chad) Mpra (Ghana) Ongota (Ethiopia) Oropom (Kenya, Uganda) Rer Bare (Ethiopia) Shabo (Ethiopia) Sinyar (Chad) Sua (Guinea-Bissau) Weyto (Ethiopia) Wutana (Nigeria)

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

145

Map 7.1 Linguistic isolates

has also been applied successfully to a range of African language families, as summarized in the tables above. Some basic principles of this method and some of its problems are illustrated with examples from (Narrow) Bantu below. Koelle (1854) had identiied a group called ‘K̄n’go N’ǵla languages’ as one of 11 genetic groupings, but the systematic comparison of what would come to be called Bantu received an important impetus through Bleek (1862, 1869), who may be seen as the founder of comparative Bantu studies, as well as Meinhof (1899, 1906).1 Whereas Bleek and Meinhof had to rely on a small set of languages, modern researchers can rely on hundreds of synchronic and diachronic studies on Bantu languages, as a result of which it has become the best studied African language family. This subgroup of Niger-Congo plays a central role in the next sections. 1

Schadeberg (2003:144) points out that ‘Bleek (1856) recognized not only the unity of the Bantu language family but also its relation to, what came to be known as, Niger- Congo and Kordofanian after Greenberg (1963); he names as speakers of related languages “the Otsi, or Ashantee” (Akan), “the Bullum, and the Timneh of Sierra Leone”, “the wide-spread Fulah” (Fulfulde), “the Accra” (Ga), “the Wolof”, “the Ukuafi” (Maasai, some of whom speak Bantu languages), and “the Tumale of Darfur” (speaking a Kordofanian language).’

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

146

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Table 7.6 Lexical cognates in some Bantu languages Cognate lexeme

Swahili

Nyamwezi

Kinyarwanda

Luba-Kasai

Proto-Bantu

1 2 3 4 5

m-bwa m-buzi m-bega m-boo m-bu

bwaa, -waa -bʊli, -ßʊli -ßega -ßolo -bʊʊ

-bwa — -bega -boro -bu

-bwa -buʒi -beya -bolo -bu

*-búa *-budi ̧ *-bega *-bodó *-bu

dog goat shoulder penis mosquito

7.3.1

Structural Changes

Because, for most words in languages, the link between sound and meaning is arbitrary, their actual resemblance between two or more languages in a range of basic vocabulary words requires a historical explanation of common origin. The pioneers of Indo-European studies in the 19th century, who came to be known as the Neogrammarians, discovered that such formal resemblances are systematic, and can be covered by regular ‘sound correspondences’ between genetically related languages. From a modern perspective, it would be more appropriate to talk about ‘phoneme correspondences’, but phonology as a linguistic concept was only established in 1881, when J. Baudouin de Courtenay introduced concepts like ‘phoneme’ and ‘phonetic alternation’ in scientiic publications in Polish. This conceptual innovation of phonology is illustrated below with examples from the Bantu branch within Niger-Congo, and speciically from Swahili. In spite of the fact that Swahili has borrowed extensively from Arabic and Persian (Farsi), its core vocabulary was not affected in most cases, but shows clear cognates with other Bantu languages, as the following lexical comparisons with Kinyarwanda (in Rwanda) and Luba-Kasai (in the Democratic Republic of the Congo) illustrate. The reconstructed ProtoBantu forms are derived from Meeussen (1969), who based his reconstructions on a systematic comparison of more than 400 Bantu languages. The Swahili examples (Table 7.6) involve nouns preceded by a nasal class preix. When adding further examples, it becomes clear that in intervocalic contexts, that is, where a different noun class preix or a verbal preix ending in a vowel preceded, the original bilabial stop underwent different changes. Synchronic alternations in Swahili, as with the lexical root for ‘two’, for example m-bu m-bili ‘two mosquitos’ and wa-tu wa-wili ‘two persons’, suggest that this conditioning was relevant historically. Based on these lexical cognates, one can establish slightly complex correspondence sets (Table 7.7). Because these correspondences are recurrent and regular, they must be the result of a systematic restructuring of phonological systems. Synchronically, w and v are distinct phonemes in Swahili, that is, sound units that cannot (or can no longer) be treated as variants (allophones) of b. They may occur in the same position within a word (e.g., as a result of borrowing from Arabic, as with the word for ‘seven’, saba, which forms

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

147

Table 7.7 Cognate sets in Bantu

Cognate lexeme

Swahili

Nyamwezi

Kinyarwanda

Proto-Bantu Luba-Kasai (Meeussen 1969)

6

shine

-wa-

-ßal-

—

-bal-

*-bád-

7 8 9 10 11 12 13 14

courtyard stone thigh break rain excrement mould become rotten see

-wanja -we -weo -vunj-ik-vua -vi -umb-oz-

— -we — -ßɪɪn-zik— — -ßʊʊmb-ßol-

-baandza -buye -biero -vun-vura -vyi -buumb-bor-

-banza -bwe -belo -vung-vul— -bumb-bol-

*-bánjá (-) *-bédo *-bu̧nj*-bu̧´da *-bı´i̧ ̧ *-búmb*-bod-

-on-

-ßon-

-bon-

-mon-

*-bón-

15

Table 7.8 Correspondence sets between four Bantu languages Relevant lexemes

Swahili

Nyamwezi

Kinyarwanda

Luba-Kasai

1, 2, 3, 4, 5 6, 7, 8, 9 10, 11, 12 13, 14, 15

-b -w -v -ø

-ß, -b -ß -ß -ß

-b -b -v-b

-b -b -v -b (-m if a nasal follows)

a minimal pair with sawa ‘like, equal’). However, in inherited Bantu roots they go back to one phoneme, the voiced bilabial stop *b. The exact conditioning for these changes, including the loss of the original *b before non–high back vowels (as in lexemes 13, 14, and 15), becomes clear only once correspondences for vowels have been worked out as well and once additional Bantu languages are brought in for comparison. The *b >v shift involved a split whereby fricativization occurred before high close vowels (e.g., *-bu̧nj- > -vung- ‘fold’). For Proto-Bantu, a seven-vowel system can be reconstructed. The initial stop is still found in Bantu languages that have retained the original seven-vowel system (*i,̧ *i, *u ̧, *u, *e, *o, *a), for example Bobangi, where the root for ‘break’ is -bun. Swahili has ive vowels, due to a historical merger of *i ̧ and *i, and *u̧ and *u. As a result, *v is no longer an allophone of *b, as both can now occur before high vowels, thereby establishing their status as independent phonemes. Fricativization of intervocalic vowels is a ‘natural’, that is, frequently occurring, sound change cross-linguistically; see Dimmendaal (2011:22– 58) for a discussion of common sound changes, also showing that it is not only speakers’ articulations that play a role in the historical reinterpretation of sound systems, but also the acoustic interpretations of hearers. As shown by the examples from Luba and Kinyarwanda in Table 7.8 above, these languages underwent the same fricativization process as Swahili (and they also have a ive-vowel system). When two or more related languages

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

148

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

undergo the same change, whether phonological, morphological, or lexical in nature, this may be due to independent developments; alternatively, the innovation was shared, that is, it occurred before these languages emerged as separate languages. The more innovations two or more languages share, the more likely it is that they were shared, that is, that they occurred before these languages emerged as separate languages, thus pointing towards a closer genetic relationship. No velar nasal *ŋ can be reconstructed for Proto-Bantu (Meeussen 1969). Comparative evidence suggests that this was a shared innovation of languages classiied as Narrow Bantu, setting the latter apart from closely related groups such as Ekoid (with which Narrow forms the Bantoid branch within Benue-Congo, the latter in turn forming a primary branch of Greenberg’s Niger-Congo, which in turn forms a coordinate branch with Niger-Kordofanian in Greenberg 1963). Greenberg (1963:33–34) gives examples of cognates between Eik, a language belonging to the Cross-River branch within Benue-Congo, and Proto-Bantu, in order to illustrate regular tonal correspondences. But several of these cognates also illustrate the loss of the velar nasal as an apparently shared innovation of (Narrow) Bantu. (3)

Proto-Bantu *-tui ̧ *-dui ̧ *-nai ̧

Eik (Cross-River) -tɔŋ -dɔŋ -naŋ

‘ear’ ‘knee’ ‘four’

When comparing such lexical roots with cognates in Ekoid Bantu, one inds that the velar nasal still occurs in these languages, for example in Ejagham -t̂ŋ ‘ear’, -rüˆŋ ‘knee’ (data from Crabb 1965). The loss of the velar nasal is therefore a shared innovation of all the languages belonging to the Narrow Bantu branch within Bantoid (Dimmendaal 1988:233). It is not always easy to distinguish shared innovations from areal spreading (i.e., borrowing) of innovations, in particular for Narrow Bantu (Schadeberg 2003:156–159). Bound grammatical (function) morphemes are less subject to borrowing, and consequently are sometimes more relevant for genetic classiications. The following example from Semitic illustrates this point. Like in Bantu, subgrouping on the basis of shared phonological innovations is notably dificult in Semitic, but shared morphological innovations provide a more clear-cut picture. Akkadian manifests the highest degree of formal heterogeneity, with an initial ‘k’ for the irst person subject marker on verbs, and ‘t’ for the second person. In Arabic, paradigm levelling occurred, whereby the irst person marker underwent a sporadic change *k >t by analogy with the second person, whereas in Ethiopian Semitic, the irst person formed the basis for this levelling process, as shown by Hetzron (1976), who gives additional examples of shared morphological innovations, all pointing towards the following family tree for Semitic (Figure 7.1). For a more extensive discussion of morphological reinterpretations, the reader is referred to Dimmendaal (2011:93–114, 290–297, 365–370). Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

149

Table 7.9 Pronominal suffixes in Semitic

1SG 2SG:M

Akkadian

Arabic

Ethiopian Semitic

- (a-)ku - (a-)ta

-tu -ta

-ku -ka

Proto-Semitic East Semitic (Akkadian)

West Semitic

South Semitic

Central Semitic Arabo-Canaanite

Ethiopian

Epigraphic South Arabian

Modern South Arabian

Arabic

Aramaic Canaanite

Figure 7.1 The subclassification of Semitic

While the historical-comparative study of genetically related languages in Africa has shown that regular sound changes can usually be established on the basis of cognate roots and afixes, it is also clear that irregularities occur for a number of reasons. One reason, analogical levelling, has already been illustrated for Semitic above. Another reason, involving so-called ‘sporadic changes’ in individual words with a high frequency of usage, is the common tendency to pronounce such words faster than words with a lower frequency, which in turn results in the loss of segments in the former. This is probably the main reason why, cross-linguistically, one tends to ind irregular singular/plural alternations in basic vocabulary items. Tabuization of speciic words is another reason why sound changes (and thereby correspondences) are not always regular. As shown by Herbert (1990b), Bantu languages belonging to the Nguni group have replaced inherited Bantu consonants with clicks (which in turn were borrowed from neighbouring Khoisan languages) as part of a strategy known as Hlonipa, aiming at the omission of words sounding like the names of key igures in the society. Hence, the Proto-Bantu root for ‘extinguish, be extinguished’, reconstructed as *-dím ̧ by Meeussen (1969), has a relex -cima in the Nguni language Zulu, whereby ‘c’ (representing a dental click) replaced the original *d in this particular root. Frequency of usage is an important and therefore central parameter determining the (in)stability of both phonological and morphosyntactic phenomena in two ways. First, it affects speech rates (thereby triggering sporadic sound changes); second, frequency of usage may affect the stability of sounds or grammatical structures. Thus, irregular present/ past alternations in Germanic languages (e.g., English ind/found, German inde/fand, Dutch vind/vond) are extremely stable (in spite of the fact that Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

150

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

the approximately 160 irregular verbs in German, for example, have to be memorised by speakers). Because of their high frequency, they are acquired by children at an early age, and are thus remarkably stable historically. Frequency of usage may also be reinforced by patterns of multilingualism. This is presumably the main reason why certain sounds are more common in some areas than in others. For example, labiovelars (or labial-velars) like kp, gb, and ŋm are ‘marked’ on a worldwide scale, but they are quite common in languages across West Africa and Central Africa, particularly in Niger-Congo and the Central Sudanic branch of Nilo-Saharan. Language contact and corresponding patterns of multilingualism have contributed to their spreading. Consequently, labiovelars also spread into neighbouring Chadic languages in northeastern Nigeria (Greenberg 1983:6), for example in trade languages like Bacama. Once languages (or their speakers) move into areas where such phonetic features are absent, such sounds may disappear. For example, in geographical outliers such as the Niger-Congo languages in the Nuba Mountains in Sudan, only Katla and the closely related language Julut have labial-velars. In Tima, also closely related, these have become labialised velar stops (*gb> *g (w)>k (w)). (4)

Katla gb-ɔ´l̂c gb-ɘlana gb-̀jàŋ g-̀c g-̂ g-ʊ´nʊ̂

Tima k-ʌ´lìh kɘ` -ĺ́nʊ` k-̀r̀ŋ k-̂h k-́̀ k-ɔ´nɔ`

‘(kind of) spear’ ‘elder’ ‘leopard’ ‘head’ ‘dog’ ‘ear’

Whether sound changes are blind to the overall phonological system of a language is a matter of dispute. Historical changes in African languages with so-called advanced tongue root (ATR) vowel harmony, for example, suggest that system preservation sometimes does play a role. Within the Eastern Nilotic branch of Nilotic, languages constituting the Bari subgroup have a classical system with ive [–ATR] vowels and ive [+ATR] vowels. In the so-called Non-Bari languages (forming the other primary branch of Eastern Nilotic), the tenth vowel *ə shifted to and merged with *a in roots, but in sufixes it shifted to and merged with *ɔ or *o, depending on the language, thereby maintaining a system of alternating sufixes (Dimmendaal 2011). While loss of the central [+ATR] vowel ɜ, ʌ, or ə (depending on the language) is common in a range of Niger-Congo and Nilo-Saharan languages (where ATR harmony is endemic), another type of restructuring occurred in Western Nilotic Jumjum. As shown by Andersen (2006), the correspondence sets between Jumjum and the closely related language Mayak at irst

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

151

sight suggest a ‘lip-lop’ or ‘swopping’ rule in one of the two languages, as the vowels in cognate roots have opposite ATR values: (5)

Jumjum wiil buuy pɪɪk lʊʊm

Mayak wɪɪl bʊʊr pii luum

‘tail’ ‘shoulder’ ‘water’ ‘grass’

Based on lexical cognates and regular correspondences with other Western Nilotic languages, Andersen (2006) postulates the following regular sound changes (in the order presented below) for Jumjum (historical changes in Mayak are not further discussed here): (6)

*ɪ > i and *ʊ > u (raising of high [–ATR] vowels) *e > ɪ and *o > ʊ (raising of mid [+ATR] vowels)

Apart from ATR harmony, there is much to be said about tone as another prosodic property shared by many languages south of the Sahara. The number of tonal registers ranges from two to ive, with language families in the geographically peripheral zones (e.g., Atlantic or Omotic) also showing transitions between tone and pitch accent or stress systems. Again, for reasons of space, these rather fascinating historical changes cannot be further discussed here; see Chapters  10 and 11 for more information. Syntagmatic changes because of sound units assimilating to neighbouring segments in adjacent morphemes may result in new allomorphs, and thereby in morphological complications. The Perfective (or Anterior/Past) marker in Bantu is an example of this. Bastin (1983) reconstructs this sufix ̧ but also points out that in many Bantu languages it fuses with the as *-ide, lexical root or stem (i.e., root plus derivational sufixes), thereby causing ‘imbrication’ (or ‘overlapping’). Hyman (1995a) gives a detailed account of the various factors inluencing its formal realization in Bemba. Some of the allomorphs can be predicted on the basis of the root vowel and the oral or nasal character of the root-inal vowel, but a further range of fusion processes occur in this language, whose conditioning is also morphological in nature (examples adapted from Hyman 1995): (7)

Root ćteblimtémśkatfýalceesh-

‘do’ ‘tell’ ‘cultivate’ ‘cut’ ‘seize’ ‘give birth’ ‘cause to dawn’

Perfective stem ćt-il-e eb-el-e lim-in-e tém-en-e śkeetfyéelcees-esh

Thus, with some verbs, a phonologically transparent form occurs next to an imbricated form, as with ‘lash’ and ‘cause to have a child’ below; with

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

152

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

others, for example ‘want, ind’, the phonologically transparent form is the norm: (8)

Root byaat- ‘lash’ fýash- ‘cause to have a child’ fwaay- ‘want, ind’

Perfective stem byeet- ~ byaat-ilfyéesh- ~ fýash-ishfwaa-il (*fweey-)

Analogical levelling, as in the example ‘want, ind’ above, is usually based on more productive and phonologically transparent forms. When speakers no longer recognize the separate status of afixes, these may become incorporated into stems, as in English child-r-en, with the incorporated plural -r still found in cognate stems in Germanic languages such as German (kind/kind-er). This latter process, also known as increment, is common in Niger-Congo noun class systems, and appears to have occurred independently in languages belonging to different subgroups.2 For example, the Proto-Bantu noun class preix *ku- is no longer used with nouns in Swahili, and only occurs as an ininitive preix (ku/kw-) with verbs. The preix was incorporated into the noun stem with nouns formally occurring in this noun class, as in kwapa, from ProtoBantu *ku-japa ‘armpit’ (Meeussen 1980), which is treated synchronically in Swahili as a noun belonging to class 5, as shown by its agreement marker li-, rather than ku-. Africa is often seen as a ‘hotbed’ of noun classes and verbal extensions. Interesting as these morphosyntactic phenomena are in and of themselves, it is also time to get away from clichés about the continent, in order to start investigating the historical development of a range of other fascinating morphosyntactic phenomena. These include such varied topics as the historical development of case systems (including split ergativity), converbs, numeral classiiers, verbal compounding, and the questions of how derivational afixes become inlectional morphemes and vice versa, and how such afixes develop out of compounds.

7.3.2

Semantic Change

While considerable progress has been made in our understanding of phonological and, to a lesser extent, morphosyntactic changes, semantic changes remain less well understood, because they are frequently associated with the cultural speciics of a speech community. The ProtoBantu root *-b́nj́ ‘courtyard’, in Table 7.7 above, for example, has been retained with the same meaning in a range of Bantu languages, but its derived meanings (historically) include ‘family branch’ in Bolia, and ‘village where the chief resides’, as in Mbala (both languages being spoken in the Democratic Republic of Congo). Speciic semantic extensions may also 2

This parallel development of features that were not present in the common ancestor is also known as drift, convergent evolution, or homoplasy (Dimmendaal 2011:365–370).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

153

be common to areas manifesting phonological and morphological convergence, as Hayward (1991) shows for Ethiopia, where Cushitic, Omotic, and Semitic, as distantly related members of the Afroasiatic phylum, manifest identical lexicalization patterns: (9)

ayət’ hantuuta ʔeč’ eré

1. mouse rat; 2. biceps muscle (Amharic; Semitic) 1. mouse rat; 2. biceps muscle (Oromo; Cushitic) 1. mouse, rat; 2. biceps muscle (Gamo; Omotic)

(10)

k’ ädda waraabe duuk’k’́des

1. draw water; 2. copy, imitate (Amharic; Semitic) 1. draw water; 2. copy, imitate (Oromo; Cushitic) 1. draw water; 2. copy, imitate (Gamo; Omotic)

These examples illustrate the two most common types of semantic changes: metaphorical and metonymic extensions, which may occur within a speciic semantic ield (intraield changes) or between semantic ields (interield changes). These extensions also play a role when lexemes develop a more grammatical function, or when one speciic grammatical function is extended into another. Research within these latter domains has become particularly popular following the inluential contribution of Heine et al. (1991), who show that there is a cross-linguistic tendency to move from concrete to more abstract meanings, for example from spatial to temporal meanings. A central claim of grammaticalization theories is that grammatical markers tend to be drawn from speciic lexical domains, and that their functional development is unidirectional historically. The important point to keep in mind from a historical-comparative point of view is that such claims on speciic grammaticalization paths need to be supported by way of the comparative method; they cannot simply be assumed, otherwise such claims become circular. The studies in Devos and van der Wal (2014) present elegant examples of ‘less beaten grammaticalization path[s]’, as the editors call them. While ‘come’ and ‘go’ are common source items for tense and aspect markers cross-linguistically, something which has been known for a few centuries, a wide range of other grammatical extensions are attested. Carlson (2014) discusses the verb sa ‘go’ in the Gur language Supyire, whose function has been extended into that of an intensifying discourse marker (Carlson 2014:257): (11)

k’a it PERF

s̀

a

INT

SCN

p̀̀ be.big

‘it is really big’ Other functional extensions of movement verbs include the use of the verb for ‘go’ as a focus marker in the Bantu language Shangaci, as a passive auxiliary in the Romance languages Italian and Romanian (where it may also be used as a copula), and as a necessive (modality) marker in Indo-Aryan languages (Devos & van der Wal 2014).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

154

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

By using speciic morphemes in new morphological or syntactic environments, additional meanings may develop. However, apart from polysemy resulting from their use in new constructions, paradigmatic oppositions (so-called relations in absentia) remain important for the development of grammatical systems, as the following example from the Omotic branch of Afroasiatic shows. Omotic languages usually make a formal distinction between a formally marked declarative mood and a zero-marked Interrogative mood (alongside other mood distinctions expressing illocutionary force) on the verb. Amha (2001:288) gives the following examples from Maale: (12)

ʔatś person-M:NOM

zigiń yesterday

mukk-é-ne come-PERF-AFF:DECL

‘the man came yesterday’ (13)

ʔatś person-M:NOM

mukk-́ya come-PERF:Q

‘did the man come?’

In Sheko, which belongs to another subgroup within Omotic, a number of mood markers are found. They are not cognate with the marker -ne in Maale (their selection being determined by tense and modality), but the paradigmatic contrast is maintained (data from Hellenthal 2010:402, who points out that ‘[t]he absence of a modal marker thus distinguishes between declaratives and interrogatives’). (14)

a.

ṇ=m̄̄k-̄ 1SG=tell-put[Q] ‘shall I tell?’

b.

ṇ=m̄̄k-̄-m 1SG=tell-put-IRR ‘I will tell’

The presence of the mood marker helps to focus on the correctness of a statement. As shown in the studies edited by Crass and Meyer (2007), there is a wide spread etymological link between demonstratives, copulas, and focus markers in Ethiopian languages, a phenomenon that is also well known cross-linguistically. The use of the same morpheme for semantically related functions, that occur in different syntactic contexts (for example after a noun or verb), is known as heterosemy. According to current grammaticalization theories, there is always a unidirectional change from copula to focus marker (see Heine & Kuteva 2002:95–96). However, as Sasse (2007:48) points out based on evidence from Eastern Cushitic, ‘[c]hanges or extensions of meanings within a semantic “landscape” are never unidirectional. The fact that “normally” focus markers come from copulae does not preclude the possibility that it may sometimes be the other way around.’

7.4

Internal Reconstruction

Languages with rich morphophonological alternations, as manifested in alternative realizations of the same morpheme (from a semantic point

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

155

of view) in different morphological environments, usually provide additional means of reconstructing older stages of languages. The Surmic (Nilo-Saharan) language Baale, in southwestern Ethiopia, presents such a case. Baale does not allow the phonetic realization of stops in word-inal position. However, once sufixes are added, otherwise unpredictable consonants show up, conditioning the exact shape of, for example, number sufixes for nouns. (15)

Singular mɛɛlɛ́ keý aǵ

Plural mɛɛlɛ-k-ḱ keyi-c-ć aga-t-t́

‘axe’ ‘udder’ ‘tongue’

The root-inal ‘loating’ stop (represented as in the examples above) also manifests itself when other bound morphemes (sufixes or enclitics) are added, such as a possessive linker -a- followed by a pronominal possessive (-naand´I ‘my’ in the example below):3 (16)

mɛɛlɛ´-k-a-naand´I

[mɛ` ɛ` lɛ´ɣ̀ǹ̀ndɪ´]

‘my axe’

The latent inal stop is still pronounced in closely related languages like Didinga: mɛɛlɛk ‘axe’. As shown by these examples from Baale, simpliications at the phonetic level (loss of inal consonants) result in morphological complications (an increase in the number of allomorphs). Phrased differently, syntagmatic simpliications (e.g., phonetic erosion) frequently lead to paradigmatic complications. Analogical levelling may result, eventually, in the reduction of the number of allomorphs for a particular inlectional or derivational morpheme. Semi-productive or unproductive morphological alternations in categories involving open sets (usually nouns or verbs) may also provide keys to earlier productive patterns of afixation.4 Alamin (2012) describes such a situation for the Niger-Congo language Tima. In Tima, the noun-class system characteristic of other Niger-Congo branches (including other languages in the Nuba Mountains belonging to this phylum) has been reduced essentially to a singular/plural alternation involving one productive singular preix, kɪ-/ki- alternating with one productive plural preix, ɪ-/i-. The alternation between ɪ and i in these preixes is conditioned by ATR harmony. However, a number of singular/plural alternations involving closed sets of nouns point towards a more elaborate former noun-class system. (17)

Singular t̪-`ɔndɔ` t̪-ɛ´lmɛ`

Plural y-ɔ` ndɔ` y-ɛ´lmɛ`

‘road, path’ ‘aardvark, ant-eater’

3

The liaison or linking of otherwise hidden consonants is also attested in languages such as French or Turkish.

4

Of course, it is not universally true that verbs constitute open classes in languages. In Saharan languages like Kanuri, for example, event structures are expressed by basic (underived) verbs in less than 10 percent of the predications, remaining predications being expressed by a coverb and the light verb ‘do, say’.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

156

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Additional details on internal reconstruction as a method can be found in Dimmendaal (2011:141–151).

7.5

Dialectology

Speech communities are rarely homogeneous, and it is from pools of dialectal (and idiolectal) variation that language change is drawn. Once speciic innovations accumulate in number, be they phonological, lexical, or grammatical in nature, different lects may emerge. The geographical boundaries for speciic variation between dialects are called isoglosses (corresponding to isobars, as lines on a map connecting places of equal atmospheric pressure in geography). The more isoglosses coincide, the stronger the linguistic boundary. If there are clear geographical centres from which innovations spread, these may be called focal areas, with transitional and relic areas being less affected by these innovations; for further details on the analysis of language-internal variation, including the use of dialectometrical studies, the interested reader is referred to Dimmendaal (2011:153–178). Once Indo-Europeanists had established sound correspondences and sound changes between major languages in Europe, usually involving a comparison between standardized national languages (German, English, etc.) or the languages known from ancient written texts (Latin, Greek, Sanskrit), they started looking for the implementation of innovations in dialects of these languages. Detailed dialectal analyses of Indo-European languages at the beginning of the 20th century led to a number of interesting discoveries, including, for example, that homonymic clash due to the phonological merger of words with different meanings may result in lexical replacement. Gilliéron and Roques (1912) describe historical sound shifts which occurred from Latin to the Romance language FrancoProvençal. These sound shifts also resulted in the phonological merger of certain words. (18)

*cattus> *gat *gallus> *gat

‘cat’ ‘cock’

Interestingly, in dialects where these words had merged (as a result of sound changes), the word for ‘cock’ was always replaced by another lexeme, such as vicaire or faisan. In other words, homonymic clash had occurred to an extent which presumably caused communicative problems, as both words referred to domestic animals whose status in a traditional southern French household rather differed. Consequently, a ‘therapeutic’ change occurred, that is, loss by collision and lexical replacement through semantic widening of a related term in the same semantic ield. Homonymic clash lies at the historical basis of click insertion in Bantu languages of Southern Africa. As shown by Herbert (1990b), among others, these clicks result from unadapted borrowing from neighbouring Khoisan

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

157

languages (as an areal grouping), but also from the systematic substitution of consonants in Bantu words by way of click sounds, in order to make such words suficiently distinct from taboo words (as mentioned in the discussion of Zulu in Section 7.3.1). Homonymic clash between grammatical (as against lexical) morphemes is less well known or studied. When different functions are involved for two morphemes with the same form, this should not interfere with comprehension. But what happens if these morphemes are in paradigmatic opposition to each other or if there is semantic afinity? Bostoen and Mundeke (2011) describe such a case for the Bantu language Mbuun, spoken in the Democratic Republic of the Congo. In Mbuun, the Proto-Bantu causative sufix *-ici merged with the applicative sufix *-Id due to a convergent phonological evolution of the two morphemes. Synchronically, these derived verb stems are expressed by means of consonant alternation on the verb (ka-ĺm ‘leave, come out’ / ka-ĺmme ‘to remove, take off/out’; kalaam ‘to follow / ka-ĺemme ‘to follow for’). This syncretism gave rise to a system whereby ‘causative verbs which are not applicative in origin may adopt an applied object’ (Bostoen & Mundeke 2011:216). (19)

taar father

o-a-lɛlle SM1-PRS.PROG-shrink (.APPL)

mo-an NP1-child

m-byeel NP9-leg

‘my father is checking his step (= is slowing down) for the child’ Another interesting phenomenon, involving dialect change due to areal contact, is illustrated in example (20) below. Swahili, as a major contact language, spread along coastal zones (hence its name, derived from the Arabic word for ‘littorals, coasts, seashores’, saw̄ḥ il, with the preix ki- expressing a way or manner of speaking). The Swahili examples presented in the discussions above are based upon Zanzibar Swahili, which formed the basis for Standard Swahili as the oficial language of Tanzania. Zanzibar Swahili is a Southern Swahili dialect, as becomes clear from the examples in (20) (derived from Nurse 1982). (20)

Northern Swahili nd̪e nd̪aa -t̪eka -t̪aka

Southern Swahili nje njaa -cheka -chaka

Standard Swahili nje njaa -cheka -taka

‘outside’ ‘hunger’ ‘laugh’ ‘want’

In the case of words like ‘outside’, ‘hunger’, and ‘laugh’, the same forms are found in Standard Swahili. But the Standard Swahili form for ‘want’ is -taka rather than -chaka, probably as a result of borrowing from a northern dialect (possibly the variety spoken in Mombasa). Another isogloss distinguishing Northern and Southern Swahili concerns the expression of the perfective. Whereas Northern Swahili dialects have retained the perfective sufix (see the discussion in Section  7.3.1), Southern Swahili dialects (including Standard Swahili) have replaced the archaic Bantu aspect marker with a perfective preix -me-.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

158

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

An interesting historical feature of the expansion of Swahili is the fact that during the past thousand years it spread as far south as the coastal zones of Mozambique, where it had a strong impact on the lexicon of southern Bantu languages such as Makhwa or Koti. During the 19th century, Swahili also spread into the African interior, as a result of which it is now spoken in different varieties in Tanzania, Kenya, and neighbouring countries such as Burundi, Rwanda, Uganda, and the Democratic Republic of Congo. The spreading of this lingua franca in the latter country is particularly interesting because the regiolects are strongly associated with major cities in the eastern and north-eastern zones of the Democratic Republic of Congo, Bunia, Goma, Kisangani, Kivu, and Lubumbashi. As argued by Nassenstein (forthcoming) in his survey of western Swahili dialects, a ‘gravity model’ is the most itting template for the spread of linguistic innovations in this area. Such a model was irst proposed by Trudgill (1974) in order to explain the diffusion of innovations in English dialects, namely from the largest urban centres to smaller towns and villages. As further shown in Nassenstein (forthcoming), contact with other languages in the area played a central role in the development of western Swahili dialects. The most dramatic restructuring as a result of language contact occurred in Bunia Swahili, where acrolectal, mesolectal, and basilectal varieties are used by speakers as luid repertoires (Nassenstein, forthcoming). Whereas the acrolectal variety is similar to Standard Swahili, the basilectal variety supericially looks like a pidginised variety, due to its reduced bound morphology. In actual fact, however, it constitutes a replication of structures found in Central Sudanic languages in the area such as Lendu and Ngiti, which are also characterized by a low degree of bound morphology. For example, pronominal subject clitics (as shortened forms of independent pronouns) and tone (which is also a prosodic feature of the basilectal variety of Bunia Swahili) have replaced the typical Bantu preixes (also for pronominal objects) and former stress patterns. (21)

7.6

Basilectal Bunia Swahili mi-pa-li-on-o 1SG:S-NEG-PAST-see-2SG:O ‘I did not see you’

Standard Swahili si-ku-ona 1SG:S:NEG-2SG:O-see ‘I did not see you’

Genetic Inheritance versus Borrowing

The second phenomenon of historical interest in comparative linguistics, alongside shared inheritance from a common ancestor, is borrowing, as already illustrated in the preceding section. The transfer of features from one language into another is the second reason why two or more languages may show formal similarities. It is not always easy to distinguish between shared inheritance and borrowing, especially when languages

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

159

are closely related; an irregular sound correspondence, as with the root for ‘want’ in Standard Swahili illustrated in the preceding section, is one such indication. Other rules of thumb for the identiication of borrowing between two languages, including the direction in which the borrowing occurred, involve the identiication of phonological and morphological features that are unique to one of the two languages involved, as shown by Greenberg (1960) in an illustrative contribution on borrowing in the Chadic (Afroasiatic) language Hausa from the Nilo-Saharan language Kanuri. The following two examples illustrate some of the diagnostic features involved: (22)

Hausa r~ ̀b̀̀t́́ k̀r̀̀t́́

Kanuri rəvo-tə´ kə` r̀-tə´

‘write, writing’ ‘read, reading’

Since Hausa does not have voiced labiodental fricatives, it replaced the latter sound with a bilabial stop. This phonological adaptation, as well as the fact that in the Saharan language Kanuri there is a nominalizing sufix -tə´, strongly suggests that it was the donor language.5 The inverse process, folk etymology (whereby morphologically simplex forms are decomposed into meaningful units), is found in the Swahili word vi-rusi, from English virus; the corresponding singular in Swahili is ki-rusi (the addition of a inal vowel being due to a preference for open syllables in this Bantu language). Borrowing occurs to differing degrees, and may go way beyond lexical transfer to include grammatical borrowing, as shown by Thomason and Kaufman (1988) in their seminal study on language contact. Grammatical borrowing without lexical borrowing is also possible, but this requires a language ideology whereby speakers want to keep their language free from foreign inluences. Whether typological distances between languages play a role when grammatical morphemes are borrowed is a matter of dispute. But the different structural effects of Bantu languages in Kenya on two neighbouring Nilotic languages, Kalenjin (which forms a dialect continuum) and Luo, for example, suggests that this factor does play a role. Both Kalenjin and Luo restructured their verb along a Bantu pattern by adding tense marking, something that could be accommodated easily since their verbal templates were already very similar. But Luo also developed a Bantu type of noun class preixation system, whereas Kalenjin did not; Luo has only two number sufixes, whereas Kalenjin has over 20 different sufixation patterns; see Dimmendaal (2011:188–196), who also discusses the sociohistorical background to this kind of structural borrowing. As shown by Thomason and Kaufman, grammatical borrowing (alongside lexical borrowing) and the transfer of structural features may come about (a) as a result of shift-induced interference from the primary language

5

The form kə´rà ‘reading’ in Kanuri derives from Arabic qaraʕa with the same meaning.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

160

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

(L1) when shifting towards another, more dominant language (L2) or (b) without language shift (because the primary language is emblematic for the community), but with structural interference from a dominant ‘interlanguage’. This latter process, which is now commonly referred to as metatypy, is attested in Greek as spoken in Turkey, but also in African languages. As shown by Yigezu (2005) for the Surmic (Nilo-Saharan) language Baale in Ethiopia, even basic vocabulary may be borrowed in such cases. While mass comparison has yielded interesting irst results on genetic groupings for Africa as a whole, the same method may also lead to misinterpretations, if one fails to identify extensive borrowing. According to Greenberg (1963), Northern Khoisan and Southern Khoisan are genetically related. But a study by Traill and Nakagawa (2000) reveals that a range of lexical similarities  – these languages have little bound morphology that could be compared – is probably due to borrowing. The Kuliak languages in northern Uganda present a further example of intensive borrowing from Nilotic, which may have misled Greenberg (1963) to classify this group of languages as a subgroup of the Eastern Sudanic branch of Nilo-Saharan; see Chapter 11 for further details.6 Language contact phenomena may be divided into four prototypes: (a) borrowing; (b) pidginization and creolization; (c) ‘language mixing’; and (d) language loss. These are not mutually exclusive, of course, as borrowing may occur with the second, third, and fourth phenotypes. Some authors emphasize the (presumed) unique nature of creoles (e.g., Bakker et al. 2011, and Ethnologue, including the last updated version by Lewis et  al. 2015). Like various other authors, the present author has argued against this so-called ‘exceptionalism’ (Dimmendaal 2011:213–236; see also Haspelmath 2013 for a succinct criticism). First, pidginization and creolization as restructuring processes come by degrees, depending on factors such as the number of primary language speakers and the intensity of their interactions in multilingual groups. Second, morphological simpliication does not occur only when ‘emergency languages’ develop between multilingual communities; the same processes are attested with speakers of morphologically rich languages encountering speakers of languages with limited bound morphology; see the discussion of Bunia Swahili in Section 7.5. ‘Language mixing’, as a label describing a third type of contact phenomenon, tends to be avoided by Africanists due to its association with scholarly contributions from the end of the 19th century and the beginning of the 20th century, and speciically those involving claims about the link between language, race, and culture. Moreover, as shown by Velupillai

6

Dixon (1997) claims that such lexical and grammatical diffusion processes typically occur during so-called equilibrium periods. These follow so-called punctuation, whereby languages develop into different varieties over a relatively short time span as a result of geographical expansion (due, for example, to technological innovations). The interested reader is referred to Dimmendaal (2011:352–354) for a discussion.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

161

(2015), different contact situations can result in similar ‘mixed’ languages, and different ‘mixed’ languages may arise from similar contact situations. The term consequently does not refer to a proper taxonomic class. What is more, virtually all languages are ‘mixed’, that is, inluenced by other languages. Dimmendaal (2011:237–252) refers to such intertwined languages as syncretic languages. An interesting African case of syncretism, involving the structural merger of two languages, is described by Souag (2015) for Tadaksahak, a variety of Northern Songhay. Tadaksahak, which is spoken by nomadic groups in Mali that are culturally close to Tuaregs (who speak a Berber language), shows a strong structural inluence from Berber, as manifested in extensive lexical and grammatical borrowings. Historically, the syncretic structure is the result of a shift in language solidarity from speciic varieties of Berber towards Northern Songhay. In Tadaksahak, for example, structural changes include the borrowing of a feminine plural ending -in, a centrifugal particle -an, a preposition inžin ‘like’, and the ‘exclusive use of Berber diathesis morphology, resulting in systematic suppletion for non-Berber verbs’ (Souag 2015:126–127), as illustrated in (23): (23)

Verb ḱs kaw

Causative š-ə´ɣdəš š-́kuš

Passive t-ə´ɣdəš t-́kuš

‘do’ ‘take out’

Interestingly, the presence of different languages in a speciic area (sometimes belonging to different, unrelated language families) does not necessarily lead to borrowing or convergence. The crucial point is whether bilingualism or multilingualism occurs. Apart from convergence (or ‘Sprachbund’) areas, there are a number of accretion (or residual) areas on the African continent, characterized by tremendous genetic diversity and typological disparity, such as southwestern Ethiopia and the Nuba Mountains in Sudan. In this latter area, more than 40 languages are spoken in an area roughly the size of Scotland, which belong to three different language families: Niger-Congo, Nilo-Saharan, and Kadu. Languages in this area present a microcosm of typological variation as found on the African continent as a whole. Languages in the Nuba Mountains vary in terms of constituent order, whether they prefer to mark syntactic relations on the verb (head marking) or rather on dependent categories (dependent marking), and whether they have extensive number-marking for nouns or (virtually) none at all. Whereas, in many parts of Africa, people switch between different languages on a day-to-day basis (a process frequently leading towards convergence between the languages involved), it is also clear that this does not apply to areas such as the Nuba Mountains. The use of Arabic as a contact language in this area these days, to an extent that it is gradually replacing local languages, is a fairly recent development. The highly restricted degree of bilingualism and localist language ideologies in the Nuba Mountains is the main reason for the absence of language union or convergence in this area (Dimmendaal 2015b:9–81). Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

162

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

One may also observe ‘fragmentation belts’, where language differences were maintained in the presence of extensive gene lows between people over an extended period of time, as shown by Veeramah et al. (2010) for the Cross River branch of Benue-Congo. While there are more than 60 Cross River languages (spoken mainly in south-eastern Nigeria), the different speech communities are fairly homogeneous in terms of Y-chromosomes and mtDNA markers. Genetic research may thus provide additional interesting insights into our understanding of language ecology. Currently, however, such research is interdicted in some countries (for example Ethiopia), as there is a fear that such indings may be instrumentalized politically.

7.7

Cladistic and Rhizotic Representations of Genetic Relationships

The classical family tree, as a metaphor, represents subgrouping based on the notion of shared innovations between related languages (as discussed in Section 7.3), and as such relects historical divergence. However, language change also involves convergence, between both unrelated and genetically related languages, as illustrated above. Scholars have been aware of such contact phenomena ever since the irst dialectological studies and contributions on creolized varieties of Indo-European languages appeared at the end of the 19th century. While family trees (i.e., cladistic models) continue to serve as a (metaphorical) representation of genetic subgroupings for languages, more recently, network analyses have become popular in order to represent genetic relationships. The integration of stochastic methods originally designed for biological applications has also made it possible to analyse large quantitative datasets for languages automatically. McMahon and McMahon (2005:122–138) give a detailed discussion of such methods, but their contribution also contains cautions and caveats (which are shared by the present author; see Dimmendaal 2011:276–280). In Chapter 11, a ‘classical’ family tree is presented for the Eastern Sudanic branch within Nilo-Saharan. Schnoebelen (2009) presents an alternative rhizotic model or NeighbourNet graph of Eastern Sudanic languages, which is reproduced in Figure 7.2. Here, shared similarities (which are not necessarily shared innovations in the classical, Neogrammarian sense) are visualized by way of a ‘clouding’ of genetic splits. Figure 7.2 shows, for example, that Eastern Nilotic Maasai is part of a ‘cloud’ containing Southern and Western Nilotic languages (two additional subgroupings that have been accepted by Niloticists for more than 50 years). As shown in the detailed historical-comparative studies by Voßen (1982) for Eastern Nilotic, there is a range of phonological, lexical and grammatical innovations clearly showing that Maasai

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

163

Teso_Nil Turkana_Nil

Maasai_Nil Nandi_Nil

Karimojong_Nil

Pokot_Nil Datooga_Nil Shilluk´ _Nil Anywa_Nil Seb´ei_Nil

Pari_Nil

Koegu_Sur

Dinka_Nil Nuer_Nil

Jur_Luwo_Nil Chai_Sur Mursi_Sur Tirmaga_Sur Me'en_Sur

Bari_Nil Acholi_Nil

Kuku_Nil Daju_Daj_Fur_Daj

Luo_Nil

Baale_Sur Majang_Sur

Lango_Nil

Shatt_Daj

Tennet_Sur

Didinga_Sur Murle_Sur Sha_Daj

Nyangi_Kul So_Kul

Ingessana_Eas Nobiin_Nub

Ik_Kul Temein_Tem

Nubian Hill_Nub

Midob_Nub Sungor_Tam

Nubian Kunuz_Nub Nubian Dongolese_Nub

Miisiini_Tam Nara_in_Ethiopia_Nar Tama_Tam

Nyimang_Nyi

Shabo_Sha

Figure 7.2 NeighbourNet graph of Eastern Sudanic languages

is an Eastern Nilotic language, closely related to the Lotuxo cluster, and forming a subgroup within Eastern Nilotic together with Teso-Turkana. But the network analysis in Figure  7.2 suggests something different. Network analyses are usually based on the counting of (presumed) cognates, although the actual calculations are far more complex than with traditional lexicostatistics.7 According to the present author, adherents of NeighbourNet graphs are mixing up the genesis of a language  – the 7

For most comparative linguists, lexicostatistics (or glottochronology) constitutes an outdated method for the genetic classification of languages; see Dimmendaal (2011:71–74) for a discussion.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

164

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

processes contributing to the present structure of a language  – with its genealogy – its inherited structures (an important terminological distinction going back to Manessy 1990). In this respect, such networks are to some extent ‘pre-Darwinian’ (like the botanical classiication by Linaeus), focusing on supericial typological similarities which actually evolved due to areal contact (borrowing).8

7.8

Future Directions

The traditional view on language change is that it is a more or less gradual process usually taking place through ‘the invisible hand’ (Keller 1994), that is, without speakers using the language being aware of these changes. But we now know that this characterization is incomplete. Abrupt creolization processes or the emergence of syncretic languages (as discussed above) show that dramatic changes may occur within a few generations. Moreover, the emergence of urban languages in a range of African metropoles show that ‘agency’ and deliberate language change are important too. These latter phenomena do not just constitute recent developments, but instead are continuations of ancient traditions, as becomes clear from the survey of language manipulation by Storch (2011). Current sociolinguistic investigations in an African context show that ‘trans-languaging’ (as discussed in some chapters in Part III of this volume) also results in new phonological and grammatical structures as a result of contact between African languages and ‘colonial’ languages. Nevertheless, the continued importance of grammars as abstracted and more or less bounded entities in the mind of speakers should be emphasized, as these are as much part of their competence as luid practices or repertoires are. If such grammatical systems were just constructs in the mind of scientist, it would be impossible to apply the comparative method, and thereby to establish genetic relationships and to reconstruct regular historical changes between them. After the numerous attempts at arriving at continent-wide classiications, and many empirically poorly founded claims about the unidirectionality of grammatical change, it is now time to go back to less ambitious – but ultimately intellectually more rewarding – applications of the comparative method to well-deined language families, in order to make progress in our understanding of language change. This would be in line with language typological research in the 21st century, which has also drifted away from large-scale comparisons and macro-areas, as such classiications do not do justice to the tremendous variation between languages. 8

Biologists would refer to these as ‘epigenetic features’. They involve modifications developed by plants and animals during their lives, rather than being part of those species’ inherited gene package.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

Comparative African Linguistics

165

Modern technology allows for new quantitative approaches towards the historical-comparative study of languages by connecting huge databases, as with RefLex, a database aiming at the exploitation of existing lexical documentation. As shown, for example, by Segerer (2009), such computerbased techniques allow for the search for cognates through extensive comparisons of databases. The ‘transition problem’, that is, understanding how a given variable (whether phonological, lexical, or grammatical) is selected as sociolinguistically signiicant and thus as generalized by groups of individuals, was one of the central questions in the research initiated by Labov in American cities. How and why linguistic innovations spread is a still largely unanswered question within the context of African linguistics, but initial fascinating results from social network analyses, such as Schreiber and Beyer (forthcoming), provide promising prospects. As a result of progress in our understanding of typological variation between languages, whether these are genetically related or not, we now know that parallel grammatical systems may come about without areal contact necessarily being involved. An example is the emergence of ergative case systems through the reinterpretation of a peripheral instrumental phrase as a core phrase marking agentivity (and volitionality) in languages spoken in Africa, Asia, or the Americas. Homoplasy (as it is called in evolutionary biology, where phenomena like the independent development of wings or poison as defence strategies in different species are instances of convergent evolution) is therefore another fascinating phenomenon of human language, where the comparative method as applied to different language families and language typology may be combined into a model of dynamic comparison, in order to explain recurring patterns of phonological, morphological, and semantic remodelling in languages in Africa and elsewhere.

Acknowledgements The author would like to thank Yvonne Treis and Ekkehard Wolff for useful comments and criticisms on an earlier draft, and Monika Feinen for the map and the improved representation of the NeighbourNet graph.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:35:51, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.007

8 A Typological and Areal Perspective of African Languages Bernd Heine 8.1 Introduction The aim of the present chapter is to provide a comparative overview of the roughly 2,000 African languages. This overview is restricted to three different perspectives. Section 8.2 deals with the typological landscape of the continent, while Section 8.3 looks into genetic and Section 8.4 into the areal relationship patterns among the languages of Africa. The concern of this handbook is with African languages, and this raises the question of what an African language is. Are Arabic or Afrikaans African languages since they are spoken in Africa by sizeable numbers of people? No conclusive answer will be provided in this chapter, but we will use the following convention: Languages belonging to families many of whose languages are spoken natively on the African continent are treated as ‘African languages’. This includes Arabic, which has been classiied as a Semitic language belonging to the ‘Afroasiatic’ family, and the majority of Afroasiatic languages are spoken on the African continent.1 But this convention rules out Afrikaans, which is commonly classiied as a Germanic, that is, as an Indo-European language, even if it came into being on the African continent. In addition to languages like Afrikaans we will also ignore other kinds of languages such as pidgin and creole languages, the various youth languages emerging in metropolitan areas of the continent such as Nairobi, Johannesburg, and Kinshasa, as well as the nearly 30 sign languages that have been identiied so far, since covering such languages would need a separate treatment.

1

As we will see in Section 8.3, the status of Afroasiatic as a family is not unproblematic; hence we are using inverted commas when referring to it as a genetic unit.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

8.2

167

The Typological Perspective

Overall, there seem to be few grammatical features that clearly distinguish Africa from the rest of the world; in this respect, the European continent, for example, is strikingly more ‘exotic’ than Africa, as the indings presented in the World Atlas of Language Structures (Haspelmath et  al. 2005) suggest. In an attempt to ind an areal identity of the African continent, one that sets Africa off from other regions of the world, Greenberg (1959) proposed a number of what he called ‘special’ features of African languages. The properties listed by him include in particular a number of lexical polysemies, such as the use of the same term for ‘meat’ and ‘wild animal’, of the same term for ‘eat’, ‘conquer’, ‘capture a piece in a game’, and ‘have sexual intercourse’, and the use of a noun for ‘child’ as a diminutive, or of ‘child of tree’ to denote ‘fruit of tree’. In the 1970s, students of African languages began to search for what – following Meeussen (1975) – tends to be called ‘Africanisms’ (Greenberg 1983:3; Heine & Leyew 2008), referring to properties that satisfy at least one the following set of criteria: (1)

a The property is common in Africa but clearly less common elsewhere. b It is found, at least to some extent, in all major geographical regions of Africa south of the Sahara. c It is found in two or more of the four African language ‘families’ proposed by Greenberg (1963).

Larochette (1959) presented a catalogue of linguistic properties characteristic of Congolese Bantu (Kikongo, Luba, Mongo), an Ubangian language (Zande), and a Central Sudanic language (Mangbetu), but a number of the properties proposed can also be found in other regions and genetic groupings of Africa. Building on this work, Meeussen (1975) proposed an impressive list of Africanisms, that is, phonological, morphological, syntactic, and lexical properties widely found in African languages across genetic boundaries. Another range of properties characterizing many African languages was presented by Welmers (1973). In 1983, Greenberg proposed distinguishing between areal properties that are exclusive to Africa though not found everywhere within it, and properties that are especially common in Africa although not conined to the continent (Greenberg 1983:3; cf. (1) above). As an example of the former he mentioned click consonants; as instances of the latter he discussed in some detail the following ‘characteristics’ (Greenberg 1983:4): (a) coarticulated labial-velar (or labiovelar) stops, (b) labial-dental (or labiodental) laps,2 (c) the use of a verb meaning ‘to surpass’ to express comparison,

2

Labial-dental flaps begin with the lower lip placed behind the upper teeth. The lower lip is then flipped outward, striking the upper teeth in passing. Its occurrence outside Africa is extremely rare, while in Africa it is found in more than 100 languages.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

168

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

and (d) a single term meaning both ‘meat’ and ‘animal’. He demonstrated that these four properties occur across genetic boundaries and, hence, are suggestive of pan-African traits, especially since they are rarely found outside Africa. Search for areal properties across Africa is associated to some extent with creole linguistics (see, e.g., Boretzky 1983). In an attempt to establish whether, or to what extent, the European-based pidgins and creoles on both sides of the Atlantic Ocean have been shaped by African languages, researchers of creoles pointed out a number of properties more widely found in Africa. Perhaps the most detailed study is that by Gilman (1986). He proposed a substantial catalogue of pan-African areal properties, arguing that a large number of African-like structures in Atlantic and other pidgins and creoles are best explained by substrate inluence. A number of the properties that are clearly more widespread in Africa than elsewhere are not considered here, for the following reasons. First, they appear to be genetically determined. The presence of gender or noun class systems is a case in point.3 Most instances of such systems to be found in Africa are presumably genetically inherited. This can be assumed to apply in particular to the noun class systems of the type found in many Niger-Congo languages of Western and Southern Africa.4 Perhaps surprisingly, we will also not consider presence or absence of click types as distinct phonemic units to be a relevant property, although it appears to be conined to Africa, and although it satisies most of the criteria proposed above. The reason for doing so is the following: The main goal of this section is to ind out whether African languages resemble one another more than they resemble other languages and what factors can be held responsible for such resemblances. To be sure, clicks occur in three of the four African ‘families’ distinguished by Greenberg (1963), not only in all Khoisan languages. Rather, they are also found in South African Bantu (Niger-Congo) languages such as Zulu, Xhosa, Gciriku (Dciriku), or Yeyi (see Maddieson 2003:31–37; Bostoen & Sands 2012 for more details), and in the Cushitic (Afroasiatic) language Dahalo. Still, their occurrence is geographically restricted to Southern Africa and three East African languages, that is, it is not representative of areal relationship in Africa as a whole. Furthermore, although Khoisan languages are amongst the phonologically most complex languages in the world, some of them distinguishing more than 110 distinct phonemes (see, e.g., Traill 1994), this fact is ignored here since it does not appear to be a characteristic of Africa as a linguistic area. Such complex sound inventories are restricted largely to the Kx’a, Taa, and !Ui languages, that is, to what Greenberg had classiied as Northern and Southern Khoisan. 3

Systems as they are found, e.g., in Bantu languages, are traditionally called ‘noun class systems’. We retain this term

4

As we will see in Section 8.2, Niger- Congo is possibly a language family, but there is so far no compelling evidence

here, although Dixon (2016:79) proposes to replace it generally by ‘gender systems’. that in fact it is.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

169

In the following we will discuss a catalogue of properties that have been proposed to be characteristic of Africa as a linguistic area (especially by Greenberg 1959, 1983; Larochette 1959; Meeussen 1975; Gilman 1986). Our selection is to some extent arbitrary in that we will ignore some properties that have been mentioned by other authors but where we are not entirely convinced that they are possible candidates for the status of ‘Africanisms’. And we are restricted to properties that are suggestive of areal relationship, that is, that are not conined to genetically deined language groupings. A general phonological property that has been pointed out by a number of researchers of African languages is the preponderance of open syllables and an avoidance of consonant clusters and diphthongs (Meeussen 1975:2; Gilman 1986:41), but both properties can also be observed in languages outside Africa. Furthermore, tone as a distinctive unit is characteristic of the majority of African languages, in many cases on both the lexical and grammatical levels. As the indings made in the world atlas show (Haspelmath et  al. 2005: chap. 13), a clear majority of the languages of the world do not use tone as a distinct feature of the phonological system. African languages are unique in that the vast majority of them are tone languages. There is only one other area in the world with a concentration of tone languages, but this area is largely restricted to South-Eastern Asia. Ignoring click consonants, there are a number of consonant types that are widespread in Africa but uncommon elsewhere (see Clements & Rialland 2008 for detailed treatment). Prime examples, among others, are coarticulated labial-velar stops such as kp and gb (Meeussen 1975:2; Greenberg 1983:4; Gilman 1986:41). There are also corresponding nasals and/or fricatives, but they do not show the wide distribution of stops, and their occurrence is largely predictable on the basis of stops (Greenberg 1983:4). The distribution of this property is clearly areally constrained: Labial-velar stops occur in a broad geographical belt from the Western Atlantic to the Nile-Congo divide (see Güldemann 2008; see Section 8.4.2.2 below), and they are also occasionally found outside this belt, e.g., in Katla and Giryama (see Welmers 1973:47–48). Still, they are found in three of Greenberg’s African ‘families’; only Khoisan languages have no labial-velar stops. Also, in Afroasiatic and Nilo-Saharan, their occurrence is restricted essentially to one branch each, namely Chadic and Central Sudanic, respectively (Greenberg 1983:7; Güldemann 2008). Perhaps even more characteristic are labial-dental (or labiodental) laps. They have been found in all African families except Khoisan, for example, in Chadic of Afroasiatic (Margi, Tera), Niger-Congo (Ngwe, Ngbaka, Ngbaka Mabo, Ndogo-Sere, some Shona dialects), and Nilo-Saharan (Kresh, Mangbetu) (Gregersen 1977:31; Greenberg 1983:4, 11; Clements & Rialland 2008; Güldemann 2008). Still, their occurrence is conined to a relatively small number of languages, and even there they show restrictions in their use as phonemic units; not infrequently, these sounds are

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

170

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

found only in special vocabularies such as ideophones. In their survey of 250 African and 345 non-African languages, Clements and Rialland (2008) found at least 70 African languages but not a single non-African language with such laps. A third type of consonants that is widespread in Africa can be seen in implosives, which – following Clements and Rialland (2008) – we deine as non-obstruent stops. To be sure, they exist in some non-African languages, such as the Indonesian language Auye (Mike Cahill, personal communication), but such languages are rare. Furthermore, word-initial prenasalized consonants, for the most part voiced stops, are widely found in Africa (Meeussen 1975:2; Gilman 1986:41), although they occur most of all in Niger-Congo languages. An outstanding property relating to the vowel system can be seen in the presence of cross-height vowel harmony based on distinctions of the tongue root position, commonly known as ATR (advanced tongue root) vowel harmony. It is widespread in Niger-Congo and Nilo-Saharan languages across the continent but appears to be rare outside Africa (see Clements & Rialland 2008 and Güldemann 2008 for discussion). Morphological properties that have been mentioned as areal characteristics of African languages include reduplication of nouns and adjectives, especially used to express distributive numerals (e.g., Swahili tano tano ‘ive each, in ives’; Gilman 1986:40). But distributive numerals expressed by reduplication exhibit the same kind of density of distribution in Southern Asia and are also found in other parts of the world, even if less commonly (Haspelmath et al. 2005: chap. 54). Within the verb phrase, many African languages are characterized by a wide range of derivational sufixes or clitics for functions such as relexive, reciprocal, causative, passive, stative, andative (itive), and venitive (ventive), and these elements can be combined in sequence (Meeussen 1975:2; Gilman 1986:43). However, these properties can also be observed in nonAfrican languages. A conspicuous feature of nominal morphology is the paucity of languages having case inlections. Those African languages that do distinguish grammatical case are mostly marked nominative, that is, in such languages it is the accusative rather than the nominative case that is unmarked; note that marked nominative languages are cross-linguistically exceptional. And a perhaps unique property of case systems in Africa is the presence of cases marked exclusively by tonal inlection, which so far has been found only in African marked nominative languages and apparently nowhere else in the world (König 2006, 2008). Another noteworthy areal feature concerns indefinite pronouns. Haspelmath distinguishes two major types of indefinite pronouns (e.g., ‘somebody’, ‘something’), namely interrogative-based indefinites (e.g., Russian kto ‘who’, kto-to ‘somebody’) and generic-noun-based

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

171

indefinites (e.g., Persian kæs-i [person-indefinite] ‘somebody’). The clear majority of the world’s languages prefer the former type, but in Africa this type is essentially absent: practically all African languages have generic-noun-based indefinites (Haspelmath et al. 2005: chap. 46). With regard to word classes, African languages have been said to be characterized by a paucity of adjectives and in a number of African languages adjectives are claimed to be absent altogether. Meanings typically expressed in non-African languages by adjectives are likely to appear as verbs of state in Africa (cf. Gilman 1986:40). On the other hand, there is a word class of ideophones that appears to be remarkably salient in many African languages (Meeussen 1975:3). While languages in other parts of the world have ideophones as well, African languages have been found to have them in distinctly larger numbers. Furthermore, ideophones expressing colour distinctions have so far been found only in Africa (Voeltz  & Kilian-Hatz 2001). With regard to argument marking, a morphosyntactic distinction between subject and object appears overall to be less problematic than in many other languages of the world, and in quite a number of African languages the two are morphologically unmarked (Creissels et al. 2008:87). In their arrangement of words, African languages of all four families exhibit a number of general characteristics, such as the following: While on a worldwide level languages with verb-inal syntax (SOV) appear to be the most numerous, in Africa there is a preponderance of languages with subject – verb – object (SVO) as their basic order: Roughly 71 percent of all African languages exhibit this order (Heine 1976:23; see also Gilman 1986:37). Furthermore, the placement of nominal modiiers after the head noun appears to be more widespread in Africa than in most other parts of the world. Thus, in Heine’s (1976:23) sample of 300 African languages, demonstrative attributes are placed after the noun in 85 percent, adjectives in 88 percent, and numerals in 91 percent of all languages. Another characteristic in the arrangement of meaningful elements relates to verbal structures: in most African languages, pronominal subject clitics or afixes precede the tense markers (93 percent), which again precede the verb (83 percent), while adverbs follow the verb (93 percent; Heine 1976:24). Accordingly, Dryer (2011) concludes in his survey of over 1,600 languages of worldwide distribution that languages in Africa exhibit a greater tendency to place modiiers after the noun than languages in other parts of the world. An arrangement of basic word order that occurs in a number of languages across the continent but which is fairly uncommon outside Africa concerns what nowadays tends to be referred to as ‘SOVX’ order. In languages having this order, the direct object precedes the verb but indirect objects and adjuncts follow the verb. SOVX languages are likely to have postpositions and to place the genitival modiier before its head while

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

172

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

other nominal modiiers follow the head noun. As we will see in Section 8.4.1.2, these languages all belong to type B of Heine (1976). With reference to information structure, front-focusing of nouns by means of some kind of cleft-construction has been mentioned, frequently used obligatorily in word questions, for example who went? is expressed by who is it who went? in many languages (Gregersen 1977:50–51; Gilman 1986:39). In addition to noun phrase focusing there is also front-focusing by means of verb-copying, where the verb appears irst in the focus position and is repeated in the main clause (Gilman 1986:39); the exact distribution of this phenomenon across Africa, however, is unknown. A striking characteristic of African languages can be seen in the widespread change of verb form, or the use of special auxiliaries, to express focus distinctions (Creissels et al. 2008:138). Logophoric marking constitutes another construction type that has been claimed to be speciically African. Logophoric pronouns indicate coreference of a nominal in the non-direct quote to the speaker encoded in the accompanying quotative construction, as opposed to its non-coreference indicated by an unmarked pronominal device (Güldemann 2008). Thus, whereas (2a) illustrates logophoric marking, (2b) is a plain, non-logophoric structure. (2)

Ewe (Kwa, Niger-Congo; data of the author) a é gblɔ bé yedzó. say that LOGleave 3.SG ‘Shei said that shei left’. b égblɔ bé édzó. say that 3.SGleave 3.SG‘Shei said that shej left’.

Logophoric structures are with very few exceptions concentrated in a large belt extending from the south-eastern corner of Ethiopia to the east up to the Niger River in the west and are found in three of the four language ‘families’ (Güldemann 2008; see Section 8.4.2.2). As already observed in Section 8.2, there are some common lexical polysemies, for example, involving activities that include verbs for ‘eat’, which are said to also denote ‘conquer’, ‘capture a piece in a game’, and ‘have sexual intercourse’ (Greenberg 1959); verbs for ‘die’, which tend to have many non-literal meanings in African languages such as ‘be in a painful condition’, ‘break down’ (cf. Meeussen 1975:4); verbs for ‘lie (down)’ also meaning ‘sleep’; or verbs for ‘hear’ (to a lesser extent also ‘see’) also denoting other kinds of perception, such as ‘smell’, ‘feel’, ‘taste’, ‘understand’ (Meeussen 1975:4–5). Meeussen (1975:4) furthermore notes that words for ‘good’ also tend to express ‘nice’, ‘beautiful’, and ‘ine’ in African languages. The status of some of these polysemies as cases of Africanisms, however, is far from clear. For example, meaning ranges expressed by verbs for ‘die’ in African languages may also be found in Australia or the Americas (Felix Ameka, personal communication), and much the same

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

173

applies to polysemies involving ‘hear’ (see, e.g., Evans & Wilkins 1998 for evidence on Australian languages). To conclude, while African languages are on the whole not dramatically different in their grammatical structures from other languages, there are nevertheless a number of features that set the continent off from the rest of the world as is also suggested by the quantitative evidence that is available on this issue (Heine & Leyew 2008).

8.3

The Genetic Perspective

The classiication of African languages in terms of their genetic interrelationship is complex.5 There is no generally or even widely agreed upon classiication. Some of the reasons for why this is so will be given in the present section; for a survey of many of the problems that are responsible for this situation, see Campbell and Poser (2008, sec. 6.8). In the present section we will irst deal with the classic work of Greenberg (1963) in Section 8.3.1 before turning to more recent work in Section 8.3.2, and in Section 8.3.3 we will draw some conclusions.

8.3.1

The Greenbergian Era

For roughly half a century, work on the reconstruction of African languages and their interrelationship has been based on the work of Joseph Greenberg (1955, 1963). What this work has established in particular are indings such as the following: a. The most easily accessible way of describing the historical relationship of these languages is by reconstructing their genetic relationship patterns. b. The multitude of African languages can be reduced to four genetically deined units, called families by Greenberg, also referred to as phyla. These units are Niger-Congo (Congo-Kordofanian or Niger-Kordofanian in Greenberg’s usage),6 Nilo-Saharan, Afroasiatic, and Khoisan. c. There are various methods available to the linguist for historical reconstruction. The task of the linguist is to choose that method that appears 5

In some recent traditions of African linguistics it has been proposed to replace the term ‘genetic (linguistic) relationship’ with ‘genealogical relationship’. There may be reasons for this proposal, such as distinguishing linguistic from biological relationship. Nevertheless, we will ignore this proposal for the following reasons. First, this is not the only term that is used in different academic disciplines, and as long as it is made clear whether one is concerned, e.g., with ‘morphology’ in the biological, the geological, or the linguistic tradition, there appear to have been few problems in the past. And second, ‘genetic relationship’ is a well-established term of comparative linguistics, be that in Africa or elsewhere, and introducing a new term for African linguistics might cause confusion when talking with colleagues working in other parts of the world.

6

Greenberg was initially restricted to ‘Niger- Congo’ (e.g., 1955:6), but later extended the family to also include the Kordofanian languages, referring to this combined grouping either as Niger-Kordofanian (e.g., 1963:149) or as Congo-Kordofanian (1963:162). We follow Williamson (1989) in referring to this grouping as Niger-Congo.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

174

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

to be best suited to solve a particular problem. Which method is most suitable in a given situation depends primarily on the time factor involved. Historical processes that happened within the last century require different methods of analysis from processes dating back to a thousand or thousands of years ago. d. Most of the work to establish genetic relationship in Africa, including that of Greenberg (1963), used the method of resemblances, which is based on the assumption that in order to establish that two or more languages are genetically related, or to determine the degree to which they are related, one simply needs to demonstrate that these languages share a suficient number of lexical and/or grammatical items that are similar in form and meaning. Greenberg was aware of the shortcomings of this method, which had been applied by generations of Africanists before him without much success, and reined it by introducing an additional component, namely that of ‘mass comparisons’, whereby comparisons are not restricted to individual languages but rather extend to groups of languages hypothesized to share genetic relationship. The main problem associated with this method is how the notions ‘suficient number’ and ‘similarity in form and meaning’ can be deined to the extent that they become part of an empirically viable tool of historical reconstruction independent of the intuitions of a single researcher. So far, there does not appear to be any satisfactory solution to this problem. e. On account of such problems, many students of African linguistics consider this method to be of doubtful value, and some would reject it altogether, arguing that reliable reconstructions of genetic relationship patterns can only be achieved by means of the comparative method. But so far it has not been possible to apply the comparative method appropriately to any of the four of Greenberg’s African language ‘families’. Greenberg’s (1963) classiication of African languages marked a milestone in the 150-plus-year-old history of language classiication in Africa. It differs from earlier classiications in the tradition of German (e.g., Meinhof 1910; Westermann 1911) and French linguists (e.g., Delafosse 1914) in that it is based on a methodology consistently applied across the roughly 2,000 African languages and, unlike that of some of his predecessors, this methodology is restricted to linguistic criteria. At the same time, this methodology has turned out to be largely inadequate for the purpose it was meant to serve, and we are inclined to side with the following appraisal by Campbell and Poser (2008:128) that Greenberg’s classiication is ‘badly in need of major reinvestigation and reworking’. These authors ind the Afroasiatic hypothesis to be the most convincing but are ‘reticent to accept traits that may have other explanations and thus are not fully persuasive of that relationship’. With regard to Niger-Congo, they conclude that many of the groups classiied as NigerCongo probably are genetically related, but the evidence presented for this Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

175

so far ‘is not especially heart-warming’, and they ind no compelling evidence for either Nilo-Saharan or Khoisan. Note, however, that none of Greenberg’s (1963) language ‘families’ has so far been ‘proven wrong’, and the classiication has remained valuable as a referential, rather than a historical classiication. Thus, the terms ‘NigerCongo’, ‘Afroasiatic’, ‘Nilo-Saharan’, and ‘Khoisan’ are still used today for convenient reference even by scholars who may reject the terms as standing for historically signiicant entities, and we adopt this convention here.7 Accordingly, the use of these terms in the following sections does not entail any claim on actual genetic relationship.

8.3.2

More Recent Work

Greenberg’s genetic classiication of African languages has been challenged over the last decades – with the effect that the four ‘families’ are no longer recognized as genetic units the way he had conceived them (see Dimmendaal 2008a). Language families numbering from 20 upwards are nowadays commonly distinguished, where some consist of genetic isolates, that is, having no more than one language or dialect group. Compared to Greenberg (1963), the new families proposed rest on a more critical analytic stance, on the one hand, and on a richer data base, on the other. But more often than not, they are not based on a methodology that is intrinsically different from that employed by Greenberg (1963), namely the method of resemblances and mass comparisons. Accordingly, their empirical signiicance is not entirely satisfactory. The following example may illustrate this observation. After Greenberg (1963:149–160) had suggested that there is a genetic link between the Kordofan languages spoken in North-Eastern Africa and the Niger-Congo languages of Western and Southern Africa, Dimmendaal (2008a:842) inds at least partial conirmation for this classiication in the Kordofanian language groups Heiban and Talodi as well as Rashad and Katla: Nevertheless, their common origin as Niger-Congo languages is beyond any reasonable doubt. Not only do they have relexes of widespread, cognate noun-class markers as found elsewhere in Niger-Congo, they also manifest an amazingly detailed formal identity of verbal derivational markers with, for example, Proto-Bantu.

Since no further evidence is provided, this hypothesis can reasonably be interpreted only as being based on the following observations. First, it appears to rest on typological reasoning: both these Kordofanian languages and at least a sizeable number of Niger-Congo languages have noun classes and verbal derivational markers. This observation, however, is not

7

Interestingly, for two of Greenberg’s most controversial families there exist regularly staged conferences continuing today, namely the Nilo-Saharan Linguistics Colloquium and the International Symposium on Khoisan Languages and Linguistics.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

176

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

extremely helpful, for the following reasons: On the one hand, there are many Niger-Congo languages, such as Kwa languages of West Africa, that have neither noun classes nor verbal derivational markers. On the other hand, there are African languages that have both but that are uncontroversially ‘unrelated’ to Kordofanian or Niger-Congo languages. A paradigm example is provided by the !Xun language of Southern Africa, which has both a noun class system and verbal derivational markers (Heine & König 2015). The language is classiied by Greenberg (1963) as forming the northern branch of his ‘Khoisan family’, now re-classiied as belonging to the Kx’a family (Heine & Honken 2010). Second, this hypothesis asserts that the languages compared have ‘cognate noun-class markers’. This raises the question of what the notion ‘cognate’ stands for. Assuming that the term is intended in the sense of referring to linguistic items that have a common etymological origin then it would seem that use of the term in the present example is in need of further evidence: To our knowledge, there is detailed information for neither the Niger-Congo languages nor the Kordofanian languages concerned on what a ‘cognate noun-class marker’ is. And third, the author claims that there is ‘an amazingly detailed formal identity of verbal derivational markers with, for example, Proto-Bantu’. We are not aware of any of the Kordofanian languages concerned to which this claim can be said to apply at least to a signiicant extent, so more detailed information is needed. To be sure, there are some striking similarities in form and meaning between these Kordofanian and Niger-Congo languages. But what such similarities mean with reference to hypotheses on genetic relationship, or to historical relationship in general, is hard to tell at the present state of research. Such observations show that claims to the effect that NigerCongo and parts of Greenberg’s Kordofanian are genetically related must be taken with care (see also Campbell & Poser 2008:131). And what was observed on Kordofanian applies at least as much to the rest of Africa. Dimmendaal (2008a) proposes the following African language families: In summarizing the current state of knowledge, the following can be stated: apart from Afroasiatic, Niger-Congo, and Nilo-Saharan (the latter two in a modiied or ‘reduced’ form), the following languages or families can be identiied – Northern Khoisan, Central Khoisan plus Sandawe and the extinct Kwadi language, Southern Khoisan, Mande, Songhai, Ubangian, Kadu, and the Coman language group plus Gumuz. (Dimmendaal 2008a:844)

In addition to these 11, he inds 8 isolates, that is, single languages not related to any other languages,8 the total of language families thus being 19. This stance is seemingly very different from that surfacing from Greenberg’s work; as a matter of fact, however, it is not, for the following 8

Dimmendaal (2008a, n. 3) adds that the figure of eight isolates might not be exhaustive.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

177

reasons. First, three of the four of Greenberg’s language ‘families’ have been retained, even if modiied or reduced, and the fourth ‘family’, namely Khoisan, is split into three families that correspond largely, though not entirely, to three of Greenberg’s sub-families of Khoisan. Second, Dimmendaal’s classiication looks surprisingly similar to that proposed by Greenberg (1955:101): The number of families is similar, namely 16 African families distinguished by Greenberg (1955) and 19 by Dimmendaal (2008a), and the families consist on the one hand of larger groups of languages, such as Niger-Congo and Afroasiatic, and on the other hand of isolates. And third, the methodology employed by the two authors is roughly the same: both authors rely on some variant of the method of resemblances as their main tool of comparative analysis, even if supplementary techniques may be recruited in addition. The main difference can be seen in the data base at the disposal of the authors concerned: in the course of the last decades, a wealth of descriptions of African languages has become available and a number of new languages have been discovered – with the effect that a modern student of comparative African linguistics is able to draw on a much larger corpus of information than that available to Greenberg. In an attempt to remedy this situation, it has been proposed to eliminate some of Greenberg’s sub-families to arrive at an empirically more convincing classiication. To this end, some authors cut off Kordofanian, Mande, and/or Ubangi (= Greenberg’s ‘Eastern’ branch) from Greenberg’s NigerKordofanian family, or Songhai, Coman, and so on, from Nilo-Saharan (Dimmendaal 2008a). This resulted in intuitively more plausible hypotheses but does not resolve inherent methodological shortcomings. Methodological problems inherent in classiicatory work such as the one just looked at also arise in work on languages that Greenberg (1963) had classiied as belonging to the Khoisan family. One example, relating to Dimmendaal’s ‘Central Khoisan plus Sandawe and the extinct Kwadi language’ family, may sufice to illustrate the problems involved. In a fresh and promising attempt to re-classify this unit, Güldemann and Elderkin (2010) propose a new genetic grouping consisting of Khoe, Kwadi, and Sandawe. The 20-plus Khoe languages, classiied by Greenberg (1963) as the Central South African Khoisan branch of Khoisan, were demonstrated by Voßen (1997) to be in fact genetically interrelated, whereas the Sandawe language of central Tanzania was classiied by Greenberg as a separate branch of ‘Khoisan’, and the by now extinct Kwadi language of southwestern Angola was still unknown at the time Greenberg worked on his classiication. Güldemann and Vossen (2000:102) had concluded that Khoe, Kwadi, and Sandawe must be left unclassiied; Güldemann and Elderkin (2010), in contrast, postulate a Proto-Khoe-Kwadi language group and argue further that this unit shares genetic relationship with Sandawe. This is not a new hypothesis; that there are ‘relations’ between Sandawe and the Khoe

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

178

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

language Khwe of Namibia had been suggested in particular by Köhler (e.g., 1975:133–134). As Güldemann and Elderkin (2010) show, however, the three share resemblances in form and meaning that are not restricted to the lexicon but also include grammatical forms such as personal pronouns. But in addition, the authors argue that there are even some regular sound correspondences between the three units, thereby implying that it is possible to also apply the comparative method to the Khoe-KwadiSandawe group. They are able to present a number of noteworthy correspondences, such as the ones in (3). (3)

Correspondences between glottalized lateral clicks in Sandawe and Proto-Khoe (Güldemann & Elderkin 2010:41) Sandawe Proto-Khoe a ǁʼa˜:ḱ ‘to ight’ *ǁʼa˜´a˜´ ‘to ight’ b ǁʼôˆ ‘to sleep’ *ǁʼóm̀ ‘to sleep’ c ǁʼ́né ‘to be ripe’ *ǁʼáń ‘to ripen’

The presence of a glottalized lateral click (/ǁʼ/) in the languages concerned is in fact indicative of a regular correspondence between Sandawe and the reconstructed Proto-Khoe. One may wonder, however, whether the presence of such a correspondence is sufficient to establish a genetic relationship. In order to rule out factors like chance similarity and borrowing and to establish a plausible hypothesis on genetic relationship one might wish to know whether this regularity extends beyond the initial consonant unit – for example, whether the Sandawe vowel i in the first syllable of (3c) corresponds regularly to the vowel a in Proto-Khoe, or whether the correspondence between final m in Proto-Khoe and lack of the nasal in Sandawe of the kind found in (3b) is also regular. On the basis of the data provided this does not seem to be the case. Accordingly, correspondences such as the ones in (3) constitute valuable pieces of possible evidence but one may hesitate to accept them as being suficient to formulate a compelling hypothesis on a genetic relationship. Such a hypothesis would need to be supported by the presence of regular correspondences that are not restricted to a limited set of phonological units but rather extend to the phonological system at large, thereby making it possible to reconstruct lexical and grammatical units in a principled way. Our reason for insisting on the application of the comparative method is the following. Approaches akin to the one used by Güldemann and Elderkin (2010) have been applied time and again since the 19th century to make claims on a genetic relationship among African languages; one example may show this. In order to establish that the Mande languages of Western Africa (classiied by Greenberg as belonging to Niger-Congo) and the Chadic languages of North-Central Africa (belonging to Afroasiatic) are interrelated, Mukarovsky (1987) proposed a catalogue of over two hundred morphemes, including personal pronouns, showing form-meaning

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

179

resemblances, and he also found what he interprets as regular sound correspondences between the two language groups (Mukarovsky 1987:48–73) – yet, for good reasons this hypothesis is nowadays generally rejected by Africanists. In a similar fashion, Westermann’s (1927) and Mukarovsky’s (1976–1977) reconstruction of Niger-Congo languages and Ehret’s work on Afroasiatic (1995), Nilo-Saharan (2001), and Khoisan (2003) are all seminal pieces of work aiming at applying the comparative method to Greenberg’s four language families, but once again they were not based on a rigid application of this method and, hence, were not successful to the extent the authors had hoped for. To conclude, hypotheses on a Khoe-Kwadi-Sandawe family, or a common Mande-Chadic stock, are intuitively appealing on the basis of the data presented, but one may wonder whether the evidence adduced is really suficient to propose a compelling hypothesis on a genetic relationship. Considering the long history of comparative linguistics in Africa based on similar methodologies, it would seem to be preferable to insist on analyses  resting on a fairly rigid application of the comparative method in order to arrive at more reliable language classiications. Until this has been achieved, we prefer to treat units such as Khoe, Kwadi, or Sandawe, or Mande and Chadic in Western and North-Central Africa, as genetically unrelated.

8.3.3

The State of the Art

As the discussion in the preceding section may have shown, genetic language classiication in Africa has so far not made dramatic headway other than pointing out that a number of the hypotheses presented by Greenberg (1963), both on the family and the sub-family level, are not supported by appropriate evidence, and proposing a number of new stocks. Accordingly, whether there should be 19 (Dimmendaal 2008a:844) or as many as 30 (Nichols 1997:376), or even more families will remain a matter of dispute in the years to come. In the light of this situation it would seem that instead of the top-down approaches that have dominated much of the work in the past, comparative linguistics in Africa might be better off with a bottom-up approach to language classiication. Such an approach, which ignores large-scale comparisons but rather focuses on dialect clusters and closely related language groupings, is actually not unheard of in African linguistics. It was essentially the approach adopted in the Handbook of African Languages series between 1948 and 1966, where the basic units of classiication were the language and the language group rather than families of languages (see Westermann and Bryan 1952:7–8 for the principles of classiication employed). Accordingly, the largest taxonomic entities distinguished in Western Africa are Gur, Kru, Kwa, and so forth (Westermann  & Bryan 1952), rather than large-scale units such as Western Sudanic (Westermann

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

180

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

1927) or Niger-Congo, and rather than a macro-class like Nilo-Saharan, the largest entities distinguished in North-Eastern Africa are groups such as Maba, Tama, Fur, Kunama, Nubian, and so on (Tucker & Bryan 1956). Not all of the language groups proposed in the Handbook series are nowadays recognized as valid genetic units, but many of them are. In addition to small groups of closely related languages, there is also one group consisting of an exceedingly large number of languages that was treated in the Handbook series as one unit, namely that of the 300-plus Bantu languages of the southern half of Africa (Guthrie 1948b; Bryan 1959), for an obvious reason: these languages had already been established earlier as being genetically interrelated through a thorough application of the comparative method (Meinhof 1899; see also Guthrie 1967–1971 and Meeussen 1980). The authors of the Handbook series were not much concerned with language diachrony, but there is another line of research relying on a bottom-up approach that has demonstrated that by applying the comparative method to smaller groups of languages it is possible to arrive at empirically plausible hypotheses on genetic relationship patterns in Africa. This line of analysis has established, for example, that neither a macro-unit like Niger-Congo nor its branches may be viable entities to search for a linguistic relationship, while smaller groups allowing for sound historical reconstruction are. In this tradition, Dalby (1965) demonstrated that Greenberg’s Atlantic branch of Niger-Congo is not suitable for historical reconstruction, while its subgroups, such as the Mel languages of Sierra Leone and neighbouring countries, are. Other examples of similar smaller-group reconstructions are not hard to come by; they include the Buem languages of Ghana and Togo (Heine 1969), the Southern Nilotic languages (Rottland 1982), and the Eastern Nilotic languages (Voßen 1982) of eastern Africa, the Khoe languages of Namibia and Botswana (Voßen 1997), or the Kx’a languages of southwestern Africa (Heine  & Honken 2010). Such analyses involve as a rule smaller language groups, but there is also at least one larger-scale study by Stewart (2002), relating the Bantu languages of Southern Africa to the Potou-Akan group of West Africa. To conclude, macro-units as proposed by Greenberg and others may be intuitively appealing; but presumably due to the enormous time depth and internal diversity involved, sound reconstruction work has so far not been entirely successful. The situation is different in the case of smaller language groups, and it is hoped that future comparative work in Africa will concentrate on such groups before embarking on largescale comparisons.

8.4 The Areal Perspective Arguably the most important question that has vexed language reconstruction in Africa is whether a given linguistic resemblance between

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

181

languages or language groups is due to common inheritance, that is, a genetic relationship, or to contact-induced factors, that is, an areal relationship. The history of more than 150 years of comparative linguistics in Africa has shown that without knowing where genetic boundaries are, it remains notoriously controversial to establish where areal boundaries are. After having dealt with the former issue in Section 8.3, we now turn to the areal relationship. The present section provides an overview of some major achievements that have been made in the areal classiication of African languages. On the one hand, it will look at Africa at large, based on classiications that make claims on African languages as a whole (Section 8.4.1). On the other hand, it will be concerned with smaller groups of areally related languages and the major regions of the continent. Then in Section 8.4.3 our interest is with grammaticalization areas, that is, with geographically deined groups of languages that are hypothesized to have undergone the same grammatical development as a result of language contact. Two main types of areal classiications can be distinguished in African linguistics. On the one hand, they are based on a single property or domain of language structure that is argued to be diagnostic of an areal relationship; we will refer to them as single-feature classiications. Paradigm examples of such classiications will be presented in Sections 8.4.1 and 8.4.3. On the other hand there are feature-bundle classiications, involving a combination of several properties shared by a deined set of languages. Typically, the declared or implicit goal underlying feature-bundle classiications is to search for sprachbund-like linguistic areas (or convergence areas); see Section 8.4.2 for examples.

8.4.1

Macro-areas

In his seminal article on ‘Africa as a Linguistic Area’, Greenberg (1959; see also 1983) provides a sketch of an areal macro-classiication of the African continent. The description provided there is not highly speciic but provides a general overview of the areal dimension that language structures in Africa exhibit. We have alluded to this work in Section 8.2; in the present section we are concerned with the internal situation in Africa. We will be restricted to a couple of examples illustrating the kind of work that has been done on the continent-wide internal classiication of linguistic features (see also Heine & Fehn 2017; Dimmendaal 2017; Good 2017). The features examined concern phonology in Section 8.4.1.1 and word order in Section 8.4.1.2.

8.4.1.1 Phonology Can Africa be divided into signiicant, geographically deined phonological zones? This question has occupied researchers in some form or other for more than a century. A recent example is provided by Clements and

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

182

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Rialland (2008): the authors propose to divide the continent on the basis of phonological distinctions into the following zones (see also Map 8.1): (4)

A geographical classiication of African languages based on phonological features (Clements & Rialland 2008) North Africa: Its phonological properties coincide largely with those of Arabic and the Berber languages. East Africa: It encompasses Ethiopia, Eritrea, Djibouti, and Somalia; nearly all its languages are usually classed Afroasiatic. Sudanic Belt: It includes the vast savanna that extends across sub-Saharan Africa bounded by the Sahel on the north, the Atlantic Ocean on the west and southwest, Lake Albert on the southeast, and the EthiopianEritrean highlands on the east. Centre: It is almost exclusively Bantu-speaking and is characterized by the linguistic features typical of Bantu languages. South Africa: It includes semi-desert, savanna, and temperate coastal regions. Its phonological characteristics derive from those of the Khoisan and Bantu languages spoken within it. This zone contains some of the richest consonant and vowel inventories of the world’s languages (see Section 8.2). Rift Valley: It includes much of the eastern branch of the Great Rift Valley in northern Tanzania and southwestern Kenya. In this region, languages of all four of Greenberg’s ‘families’ are found.

These zones are primarily geographic and only secondarily linguistic in nature. No claim is made on genetic or areal implications of the classiication. The zones must be viewed as a irst approximation to areal phonology, where each zone stands for a prototype, or ideal type. This means, for example, that a number of the languages of the South Africa zone have more phonological features in common with languages of the Centre, and many features of the Rift Valley zone can also be observed in other zones.

8.4.1.2 Word Order Based on a survey of the order of meaningful elements in roughly 300 African languages, Heine (1976) proposed to classify the languages exhaustively into four types. The main features of these types are pointed out in (5); for more details, see Heine (1976:39–45). The typology is similar to but must not be equated with the classic threefold classiication proposed by Greenberg (1966). (5)

Main features of basic word order types in African languages (S = subject, V = verb, O = object; Heine 1976:39–45) Type A: SVO, predominantly head-modiier order Type B: Possessor-possessee order, SVO or SOV order, but adverbial phrases follow the verb Type C: VSO order Type D: Modiier-head is the predominant order, and adverbial phrases and most other participants precede the verb.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

183

Map 8.1 Phonological zones in Africa (Clements and Rialland 2008; adopt from Heine and Nurse 2008, p.37, Map 3.1)

Type A corresponds largely, though not entirely to Greenberg’s (1966) SVO type. As was observed in Section 8.2, this constitutes the default type in Africa, accounting for more than two-thirds of all languages of the sample. The remaining types exhibit a clearly areal pattern, being suggestive of some linguistically deined areas cutting across genetic boundaries. One such area is made up of type B languages, consisting of a large, contiguous part of West Africa, where Mande, Gur (Voltaic), and western Kwa languages are spoken. In addition to these languages, which are traditionally classiied as Niger-Congo, this area also includes Songhai, a Nilo-Saharan language cluster. What characterizes this area most of all is the presence of possessor–possessee word order syntax, which is not restricted to the noun phrase but also affects the structure of the clause (see Heine & Claudi [1998] 2001). A second large area consists of type D languages. It is located in NorthEastern Africa, extending roughly from Lake Chad to the west to the Horn of Africa to the east. The languages showing this type belong, on the one hand, to the Semitic, Cushitic, and Omotic group of the Ethiopian highland

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

184

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

area. On the other hand, they include a number of languages classiied as Nilo-Saharan, such as the Saharan and Nubian languages, Maba, Fur, Kunama, Nara (Barea), and others. Another, much smaller area consists of type C languages, called the Rift Valley Convergence Area, deined by the presence of verb-initial (VSO) syntax. The languages of this East African area belong to Greenberg’s (1963) Nilo-Saharan (Surma, Kuliak, Eastern Nilotic, and Southern Nilotic) and Khoisan (Hadza) ‘families’.9 Note that none of the four types is entirely restricted to the areas sketched. For example, type B languages are also found in the Kx’a, Taa, and !Ui families, all of which Greenberg (1963) had classiied as South African Southern Khoisan, or in Central Sudanic languages such as Moru, Madi, Ngiti, or Lugbara. Type C languages also include the Berber languages of northern Africa, a few Chadic languages, and the Kordofan language Krongo,10 and pockets of type D languages are located in Western Africa (e.g., Ijo) and Southern Africa (the Khoe family). Both the phonological classiication in Section 8.4.1.1 and the wordorder classiication of the present section are suggestive of signiicant macro-areal clusterings of linguistic features, but what the historical signiicance of these classiications may be is still largely unclear and constitutes a task of future research.

8.4.2

Linguistic Areas

There are by now a number of areal groupings in the world that have been deined as linguistic areas (or Sprachbunds), even if most of them are not entirely uncontroversial, especially since the isoglosses that make up these areas usually do not all coincide. Linguistic areas are generally hypothesized to be the product of language contact, yet hardly any one of them has contributed signiicantly to a better understanding of the history of the area concerned. Furthermore, their contribution to synchronic linguistics has also been moderate. Nevertheless, there are some areal groupings in Africa that have been proposed to be Sprachbund-like units, and in the present section we will be dealing with the most prominent ones.

8.4.2.1 The Ethiopian Language Area The most frequently mentioned and most widely recognized linguistic area is North-Eastern Africa, frequently called the Ethiopian Language 9

As pointed out in Section 8.3, it is a matter of much dispute whether the Khoisan languages of Southern and Eastern Africa do in fact form a language family, as claimed by Greenberg (1963). Hence we are using the term ‘Khoisan’, in the same way as ‘Niger- Congo’, ‘Nilo-Saharan’, and ‘Afroasiatic’, simply for referential convenience. An alternative, typologically based term for ‘Khoisan’ is ‘Non-Bantu click languages of Southern Africa’.

10

Classical Arabic also was a type C language, but to our knowledge, all modern Arabic varieties spoken on the African continent belong to type A.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

185

Area (or Ethio-Eritrean Sprachbund). Within roughly the last two millennia, the highlands of Ethiopia appear to have favoured cultural and linguistic exchange on a massive scale, with the effect that the languages of this region now share a number of linguistic properties (Ferguson 1976; Crass & Meyer 2008; see also Güldemann 2005a). The languages included in this area are mostly genetically interrelated, belonging to the Cushitic, Omotic, and Semitic branches of Afroasiatic, but some Nilo-Saharan languages are also included. First proposed by Ferguson (1970, 1976), the notion of an Ethiopian Language Area has not gone unchallenged (Tosco 2000b), but Crass and Meyer (2008) ind further support for it. The area is not restricted to the nation-state of Ethiopia but also includes languages spoken in Eritrea, Djibouti, Somalia, Kenya, and Sudan. Ferguson (1976) proposed 8 phonological and 18 grammatical features to deine the Ethiopian area, but Crass and Meyer (2008) add another 12 criteria in support of the grouping, most of which are morphosyntactic in nature. Half of these criteria concern processes of grammaticalization shared by the languages of the area and, hence, are suggestive of grammaticalization areas (see Section 8.4.3).

8.4.2.2 The Macro-Sudan Belt Over a century ago, Westermann (1911) proposed an essentially genetically deined stock, namely the Sudanic languages (Sudansprachen).11 This hypothesis was rejected by subsequent generations of Africanists, but, as Güldemann (2008) argues convincingly, many of the Sudansprachen of Westermann can more plausibly be re-classiied as an areal grouping, referred to by him as the Macro-Sudan Belt, which is similar to, but should not be confused with, the Sudanic Belt of Clements and Rialland (2008; see Section 8.4.1.1). Deining properties of this grouping are the presence of (a) logophoric markers, (b) labial-velar consonants, (c) vowel harmony of the ATR (advanced tongue root position) type, (d) a word order pattern SOVX (see Section 8.2), (e) another word order pattern V-O-NEG, where the negation marker (NEG) is placed clause-inally, and (f) labial-dental lap consonants. Not all the features are found in all languages of the belt, but there is a massive clustering of the features that appears to be distinctive.

8.4.2.3 The Kalahari Basin In the dry area of Central-Southern Africa, there is said to be another kind of linguistic area, portrayed as an instance of a refuge area where people have been living over centuries and probably millennia without much interference from outside. It is the homeland of traditional huntergatherer populations speaking Khoisan languages. But Güldemann (1997) 11

With the hedge ‘essentially’, we remind the reader that at that time no clear distinction between genetic and other kinds of linguistic relationship was made in pre- Greenbergian language classification in Africa.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

186

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

argues that the Kalahari Basin convergence area is not conined to languages conventionally classiied as belonging to Khoisan but also includes a Bantu language, namely Tswana (see also Güldemann & Fehn 2017).

8.4.2.4 The Tanzanian Rift Valley Area This linguistic area of central Tanzania was proposed by Kießling et  al. (2008). It includes languages of all four of Greenberg’s language ‘families’. These are (a) the West-Rift languages of Southern Cushitic (Iraqw, Gorwaa, Alagwa, and Burunge) of Afroasiatic, (b) the Datooga dialects of Southern Nilotic, belonging to Nilo-Saharan, (c) some Bantu languages (Nyaturu, Rangi, Mbugwe, and maybe Nyilamba, Isanzu, and Kimbu) of Niger-Congo, and (d) Sandawe and Hadza, two ‘click languages’ that Greenberg (1963) had classiied as ‘Khoisan’. The authors adduce a range of altogether 15 features to substantiate their Sprachbund hypothesis. Five of the features are phonological, they include presence versus absence of a lateral fricative /ɬ/, of ejective obstruents, or of a 7-vowel system. The largest number of features are morphological, being either structural (presence vs. absence of a preverbal clitic complex, of categories such as verbal plurality, applicatives, venitives, or the number of past and future tense distinctions) or formal (presence of subjunctive -ee or irrealis laa). Further criteria are syntactic, concerning the linear arrangement of constituents, or conceptual, involving polysemy and semantic transfer strategies (see Kießling et al. 2008, Table 6.5). Four of the features have the same distribution as some other feature, while the remaining 15 feature isoglosses do not match those of other features. A number of additional groupings, suggestive of contact-induced relationships, have been proposed (see, for example, the Southwest Ethiopia area of Sasse 1986), but have so far not received wider recognition as linguistic areas.

8.4.3

Grammaticalization Areas

Drawing attention to the problems that are notoriously associated with linguistic areas, Heine and Kuteva (2005) propose to study grammaticalization areas as an alternative, which they deine as groups of geographically contiguous languages that have undergone the same grammaticalization process as a result of language contact (Heine & Kuteva 2005:182).12 Grammaticalization areas, thus, are single-feature classiications (Section 8.4.1) that are concerned with what, following Weinreich ([1953] 1964), is referred to as replication, that is, the contact-induced diffusion of structures and/or meanings rather than that of phonetic/phonological or morphological material. 12

Grammaticalization is defined as the development from lexical to grammatical forms, and from grammatical to even more grammatical forms (Heine and Kuteva 2002:2).

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

187

As argued by Heine and Kuteva (2005:182–187) and Heine (2011), most linguistic areas that have been identiied so far can be analysed, at least to some extent, as consisting of grammaticalization areas (Heine 1994; Kuteva 1998, 2000; Stolz & Stolz 2001:1549). Contact-induced grammaticalization processes of this kind must have happened in Africa on a massive scale. While we lack appropriate historical records for reconstruction, grammaticalization theory allows reconstructing at least an outline of some of these processes (Heine 1994, 2011; Leyew & Heine 2003; Heine 2011). Two examples may sufice to illustrate the signiicance of such areas in Africa, where one concerns comparative constructions and the other relexive marking in Africa.

8.4.3.1 The Comparative of Inequality Perhaps the most widely discussed example is that of comparative constructions, more precisely of comparatives of inequality of the form ‘X is bigger than Yʼ. The corresponding conceptual schema in the majority of African languages, called the Action Schema, taking either of the forms ‘X is big defeats/passes Y’ or ‘X defeats/passes Y in size’, uses a verb meaning either ‘defeat’, ‘surpass’, or ‘pass’ to express comparison (Meeussen 1975:4; Greenberg 1983:4; Gilman 1986:39), as seen in the following example. (6)

!Xun (W2 dialect, Kx’a family; Heine & König:236)13 !xō má nǁāˈà !ˈālā gùmì. TOP be.big pass cow elephant ‘An elephant is bigger than a cow.’

The grammaticalization from either of these verbs to a grammatical marker denoting the standard of comparison (cf. English than) must have taken place in some form or other in many African languages across genetic boundaries. To be sure, this grammaticalization has also been observed in some other parts of the world. Especially in Mainland Southeast Asian languages (Sinitic languages, Thai, Lao, Vietnamese, and Hmong-Khmer languages) verbs for ‘to cross’ or ‘passʼ have given rise to standard markers of comparison (Ansaldo 2004:490–493).14 But beyond this Sprachbund, and outside Africa it is extremely rare, while at least two-thirds of the African languages that have been documented in some detail appear to have undergone this process in some form or other (Heine 1997:126–129, Table 6–5; Leyew & Heine 2003; Heine & Leyew 2008). 13

The symbols ! and ⃦ stand for an alveolar and a lateral click type, respectively.

14

For example, in the following example from Lao of Southeast Asia, the standard of comparison is expressed by the action verb kua1 ‘passʼ and, like in African languages, the standard (caw4 ‘you’) forms the sentence object: Lao (Tai-Kadai; Ansaldo 2010:933) khòoj5

suung3

kua1

caw4.

I

tall

pass

you

‘I am taller than you.’ Note that Ansaldo (2010) refers to the Action Schema as the ‘surpass type’.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

188

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

The Action Schema is clearly predominant in Africa and rare elsewhere, but this does not seem to apply to Africa as a whole. A different situation is found in the Ethiopian area discussed earlier (Section 8.4.2.1), where another grammaticalization process has been identiied (Leyew & Heine 2003). Rather than the Action Schema, the main conceptual template used for this process is the Source Schema (‘X is big from Yʼ), where the standard of comparison (‘thanʼ) is provided by a grammatical word, particle or afix expressing an ablative function (‘(away) fromʼ), as illustrated in example (7). (7)

Amharic (Semitic; Leyew & Heine 2003:50) Miste käïhïte tat’ rallä̌̌. my. wife from. my sister she. is. short ‘My wife is shorter than my sister.’

What most of the languages of the Ethiopian area in fact have in common is that the Source Schema provides the only or one of the main options for encoding comparatives. Note that this grammaticalization area includes languages of different stocks, such as Semitic, Cushitic, Omotic, as well as languages classiied as Nilo-Saharan. Accordingly, there is reason to hypothesize that this process was, at least to some extent, contact-induced (see Leyew & Heine 2003 for more details).

8.4.3.2 Reflexive Markers Providing a sketch of an areal classiication of the African continent, Greenberg’s (1959) examples of Africa-speciic phenomena include that of relexive marking (see Section 8.2). He observes that ‘he himself’ translates in African languages as ‘he with his head’ (1959:23). This example relates to expressions for what tends to be referred to as emphatic relexives, now more appropriately described as intensiiers (König & Siemund 2000), but one can generalize by saying that, in fact, a number of African languages have grammaticalized relexive pronouns (including intensiiers) by using expressions that are etymologically derived from terms for the body-part ‘head’, as in the following example. (8)

Hausa (Chadic; Heine 2011:55) Sun kashè kân-sù. they kill head-their ‘They have killed themselves.’ (i.e., ‘they have committed suicide’)

However, Greenberg’s example is in one respect not entirely satisfactory: in the majority of cases it is not a noun for ‘head’ that is employed in African languages. A survey of African languages carried out by Heine (2011; see also Heine 2000) suggests that by far the most common nominal meaning to encode relexive concepts in African languages is to use the noun ‘body’ as the complement noun in a propositional schema of the kind ‘X sees/hits/kills X’s bodyʼ to grammaticalize a relexive category, as

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

A Typological and Areal Perspective of African Languages

189

illustrated in the Yoruba example of (9). More than half of the 46 African sample languages for which a nominal source could be established use this schema. This can in fact be said to constitute an areal property of the African continent as a whole; it seems that in no other region of the world can such a process be observed at a comparable scale (Heine 2011:50, Tables 2, 3). (9)

Yoruba (Kwa; Awolaye 1986:4; Heine 2000:14) Nwosu ŕ ara rɛ. Nwosu saw body his ‘Nwosu saw himself.’

In sum, ‘body’-relexives can be said to constitute the default case in African languages that have grammaticalized relexive categories by means of nominal expressions.15 Compared to this grammaticalization process, the number of African languages using a ‘head’-relexive, that is, grammaticalizing a schema of the form ‘X sees/hits/kills X’s head’, is fairly small, amounting to less than one fourth compared to the languages using a ‘body-relexiveʼ (Heine 2011:50).16 Nevertheless, from an areal perspective within Africa, the former are of equal interest, as the following example may show (for more examples, see Heine 2011). As is argued there, there appears to be a grammaticalization area, referred to as the Western Sahel belt, where there is a clustering of languages having a ‘head’-relexive. The languages of this belt are spoken in the arid sub-Saharan region of West Africa, roughly between Senegal and Cameroon, belonging to three of the language ‘families’ proposed by Greenberg (1963), namely Niger-Congo, Afroasiatic, and Nilo-Saharan. These languages include Fulani (Fulbe), Diola, Chadic languages such as Hausa, Margi, Mina, Pero, Kwami, and Lele, Berber (Tamazight), and Songhai (Koyraboro, Koyra Chiini). Note that the languages spoken further south along the West African coast, such as the Kwa language Yoruba (see example (9)), do not belong to this belt, showing overwhelmingly ‘body’-relexives. There are two interesting non-linguistic correlates to the distribution of the Western Sahel belt. First, unlike the languages further south, the languages of the belt are spoken predominantly by Islamic societies and, second, the area where these languages are spoken roughly coincides with the Fulani expansion that took place in the course of roughly the last 800 years (see Heine 2011:56), but more research is needed on the signiicance of such correlations.

15

Note, however, that in quite a number of African languages, reflexive markers are etymologically opaque, that is, no grammaticalization pathways could be reconstructed. Such languages are ignored in this chapter.

16

There is a range of other body parts in addition, such as ‘soul’, ‘hand’, etc. (see Heine 2011:50–51), but their occurrence is highly restricted.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

190

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

8.5 Conclusions The concern of this chapter was with nearly one-third of the languages of the world. As we saw in Section 8.2, these languages do not differ dramatically in their grammatical form from languages spoken elsewhere if one ignores a few peculiar features such as the presence of click consonants in southern and eastern Africa. That it is nevertheless possible to distinguish African languages on a quantitative basis is argued in Heine and Leyew (2008). We saw in Section 8.3 that genetic language classiication in Africa has so far not made dramatic headway since Greenberg’s seminal classiication of more than half a century ago. What was achieved mainly is that a number of the hypotheses proposed by Greenberg (1963), both on the family and on the sub-family level, are not supported by appropriate evidence. Accordingly, whether there are 15 or 20 language families or more than 50 is hard to tell on the basis of the present state of research. Nevertheless, given that new technologies and quantitative approaches are likely to shape the future of comparative linguistics in Africa in novel ways, it is hoped that Campbell and Poser’s (2008:144) somewhat pessimistic conclusion is premature when they observe that ‘Greenberg’s African phyla involve such depth and internal diversity that they remain unproven and probably can never be demonstrated.’ Most studies of language use converge on observing that multilingualism and contact-induced language change are of paramount importance in Africa. Areal linguistics provides one window for approaching the results of language contact, and in Section 8.4 we presented a small collection of the indings that have been made so far. In doing so, we ignored a problem that has vexed much of the comparative work in Africa, namely whether a given feature shared by different languages or language groups is due to genetic or areal relationship. Some headway has been made towards answering this question. Still, this remains one of main problems to be tackled in future work on linguistic relationship among the languages of Africa.

Acknowledgements The present chapter has beneitted greatly from comments by Tom Güldemann and Edward Elderkin. I would also like to thank Haiping Long for his cooperation within the project on Mandarin Thetical and Reconstruction of Discourse Grammar (National Social Sciences Fund; 15BYY107), Guangdong University of Foreign Studies, and the University of Cape Town and Matthias Brenzinger for the academic hospitality received while working on this chapter.

Downloaded from https://www.cambridge.org/core. University of Liverpool Library, on 26 Aug 2019 at 08:51:10, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.008

9 Niger-Congo Linguistic Features and Typology Larry M. Hyman, Nicholas Rolle, Hannah Sande, Emily Clem, Peter S. E. Jenks, Florian Lionnet, John Merrill, and Nicholas Baier

9.1

Introduction

The purpose of the present chapter is to survey and discuss some of the prominent linguistic properties of Niger-Congo (NC) languages. This task is a dificult one for at least ive reasons. The irst is the sheer enormity of the group: consisting of approximately 1,500 languages (Simons & Fennig 2017), the NC linguistic stock is the largest widely recognized genetic group. Second, there is considerable variation in the linguistic features among NC languages, which spread from Senegal in the west to Kenya in the east, and down to South Africa. The third reason is that most of the common features that can be identiied are not limited to NC, but are also attested in other African stocks. Fourth, there is disagreement about which linguistic units should be considered NC, such as Mande, Dogon, and Ijoid. This is potentially quite important to the extent that these groups diverge typologically from the rest of NC in interesting ways. Particularly in the case of Mande, there is the question of whether its S-Aux-OVX word order has not spread to neighbouring NC subgroups in what we shall refer to as the Mandesphere. Finally, so much more is known about certain subgroups and the more major languages that one has to be careful not to generalize, for example, not to be Bantu-centric. The task of this chapter is therefore not only to present some of the recurrent typological properties of an agreed upon NC core, but also determine which of these can be traced back to an ancestral Proto-Niger-Congo (PNC) or are due to more recent areal contact. The strategy that will be followed here is to focus on phenomena that are widely attested, noting at the same time any obvious distributional limitations, whether by subgroup or geography. After introducing the basic syllable and word structure in Sections 9.2 and 9.3, we turn to the prosodic stem in 9.4 and segmental consonant and vowel inventories in 9.5. This is followed by vowel and nasal harmonies (9.6), tone (9.7), noun

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

192

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

classes (9.8), noun modiiers and pronouns (9.9), verb extensions (9.10), verb inlection (9.11), word order (9.12), serial verbs (9.13), and information structure (9.14). Because of space limitations, much of the discussion will be brief and only suggestive of the tremendous richness encompassed by NC languages. In the inal section (9.15) we summarize the most interesting distinguishing NC features covered in the preceding sections and indicate the extent to which each feature is found in the other African stocks beyond NC.

9.2 Syllable Structure Most scholars familiar with African languages would probably irst think of the NC languages as disproportionately favouring strict open syllables. This is found in most Mande, Kru, Kwa, Western Benue-Congo, and Bantu languages. The tendency towards simple syllable structure is not only against codas, but also against consonant clusters, whether tauto- or heterosyllabic. In many Bantu languages the only potential clusters consist of a homorganic nasal+consonant, which is usually interpreted as producing a complex syllable onset; for example, Luganda is syllabiied as [lù.gá:.ndà]. A complicating factor is that many NC languages have syllabic nasal syllables, which may contrast tone, such as Igbo m  ṕ ‘smallness’, n´nà ‘father’ (Emenanjo 1987:1–2), Haya m`bwà ‘it’s a dog’, n  tè ‘it’s a cow’ (Byarushengo 1977:6). NC open-syllable languages otherwise prohibit word-internal consonant clusters as well as word-inal consonants. There are, however, many exceptions. Many of these fall into two areas. The irst consists of languages in westerly subbranches of NC such as Atlantic, Gur, Kwa, and to some extent Mande, where the coda may be restricted to the velar nasal, for example, Maninka (Niokolo) kúŋ ‘head’, kàŋ ‘neck’ (Creissels 2013:13), Guinean Kpelle laˇŋ´ ‘to jump’, mànáŋ´ ‘cassava’ (Konoshenko 2014:239). The second area includes a number of linguistic groups in Nigeria including Cross-River, Central Nigerian (Kainji, Plateau, Jukunoid), and Bantoid, going into Grassields and certain Northwest Bantu languages in Cameroon. In many of these cases it is possible to show that vowel deletion has more recently produced closed syllables by comparing the resulting forms to Proto-Bantu (PB) reconstructions (Bastin et al. 2002), for example, Babanki [Bantoid; Cameroon] ə`bèm ‘belly’ (cf. PB *-bùmù), ə`kẃn ‘bean’ (cf. PB *-kúndè) (Akumbu 2008:21, 24); Nzadi [Bantu; DRC] ntsɔˆ n ‘shame’ (cf. PB *-cónì), mbɔˇŋ ‘brain, marrow’ (cf. PB *-bòngó) (Cohen 2011:256, 260). Such languages can also have internal (more rarely, inal) consonant clusters, Babanki báŋlí ‘be ripe’, də`mkə` ‘grumble’ (Akumbu 2008:16, 20). While there are other subgroups and sporadic cases of closed syllables throughout the NC stock, NC languages rarely allow the full set of onset consonants to appear as codas. Frequently missing are a voicing

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

193

contrast, fricatives, liquids, and palatals. With a few exceptions, languages that have implosives or labiovelar /kp gb ŋm/ typically disallow these as codas. In many languages codas are restricted to a subset of stops and/or nasals. Beyond this it is hard to generalize. Table 9.1 provides a sample of Grassields Bantu results from an unpublished survey of coda consonants in 114 Bantoid and Northwest Bantu languages. In the irst line of the table, P, T, K refer to a labial, coronal, and velar stop, respectively, whose phonetic voicing may vary according to context. While all of the languages in Table 9.1 contrast voiced vs. voiceless stops in onset position (with the exception of the labial within most Bamileke), there is no voicing contrast in coda position. Authors disagree on whether these should be transcribed [p, t, k] or [b, d, g]. Table 9.1 clearly shows that only a subset of consonants can occur as coda, as few as one consonant (the glottal stop in Mfumte) and as many as nine in Bum. As seen, most have oral stops and nasals, although either can be missing. Although the velar nasal is also a common coda consonant elsewhere in NC, the glottal stop is a particular feature of this area. Also observable is that oral versus nasal codas can have separate histories concerning which places of articulation are represented. Thus, no language has only labial /P/ as an oral stop coda, and no language has only coronal /n/ as a nasal coda. Finally, some of the Ring languages prefer oral continuant fricative and liquid codas at the expense of stops. Similar onset-coda asymmetries are found throughout West African NC languages, sometimes generalizing to stem shapes (see 9.4).

Table 9.1 Coda consonants in Grassfields Bantu Language

Subgroup

P

T

K ʔ

m n

Medumba Fe’ fe’ Yemba Bamena Ghomala Bamunkumbit Shingu Ba’angu Akum Limbum Mfumte Moghamo Mundani Ambele Weh Bum Oku Kom Babanki Lamnso Babungo Bamessing

Eastern Eastern Eastern Eastern Eastern Eastern Eastern Eastern Eastern Eastern Momo Momo Momo Ring Ring Ring Ring Ring Ring Ring Ring

p p p p p p p

t t t t

k k k k k

m n m n m m m m n

t t

ʔ ʔ ʔ ʔ ʔ

k ʔ m n

b p p p p

t t t t t

ʔ ʔ k ʔ k ʔ k ʔ k ʔ ʔ ʔ

m m m m m m m m m

n n n n n n n n

ɲ ŋ f

s

ŋ

l

r

h

ŋ

h

ŋ

h

ŋ ŋ ŋ ŋ ŋ ŋ ŋ ŋ ŋ ŋ ŋ ŋ ŋ

h

r

f f f f v

s s s s

h l l r h

w y Totals 7 7 6 6 5 4 5 3 4 1 7 5 5 w y 8 9 7 7 6 y 6 2 w y 4

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

194

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

9.3 Word Structure Turning to the morphology, NC languages vary from a mono- and bisyllabic, largely isolating type in Kru, Kwa, Ubangi, and part of Adamawa, to highly agglutinative structures in North Atlantic, Bantu, Kordofanian, and certain languages within Gur, Upper Kwa, and Northern Nigeria. While sufixation is perhaps more overtly widespread, especially in derivational morphology, NC languages may also exploit preixation, especially to mark inlectional categories on the noun and verb. In many cases afixation is tonal. Where preixation and sufixation are both present, the following Bantu verb structure based on Meeussen (1967) appears to apply to other branches of NC as well: word

(1) pre-stem

prefixes

stem

root

derivational suffixes

inflectional suffix

As can be seen, sufixes more closely ‘cohere’ to the root, forming a stem, while preixes are less cohering and attach at the word level. Like the outermost sufix of the stem, preixes are inlectional, as in the following Chichewa example (Hyman  & Mchombo 1992:358), where [ marks the beginning of the verb stem: (2)

Mchómbó a-na- [ máng-ír-ídw-á nkhûni Mchombo he-PAST-tie-APPL-PASS-FV irewood ‘Mchombo was tied irewood’ (APPL = applicative; PASS = passive; FV = inlectional inal vowel)

In this example, a- is the class 1 3sg subject preix and -na- is the recent past preix. The irst two sufixes are derivational ‘verb extensions’ : applicative -irsignals the benefactive NP ‘Mchombo’, while -idw- marks the passive. In most Bantu languages the verb must end with an inlectional FV, in this case -a. Nouns (and where present, adjectives) differ from verbs both in having less internal structure, but also in whether they are marked by preixes and/or sufixes. In the case of non-derived nouns whose stem is monomorphemic, there is no distinction between root and stem. If present, a NC language may mark noun classes by preixes, sufixes, or both: (3)

a. Moro [Kordofanian; Sudan] (Gibbard et al. 2009:107, 112) a-jén ‘mountain’ pl. e-jén ‘mosquito’ pl. ɲ-aŋwat̪a l-aŋwat̪a ð-áp:á ‘friend’ pl. r-áp:á e-tám ‘neck’ pl. ́-tám

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

195

b. Kulango [Gur; Ivory Coast, Ghana] (Elders 2007a:299–306) xɛɛn-ɔ` ‘man’ pl. xɛɛm-bɔ` nimee-yò ‘bird’ pl. nimee-hʊ̰̀ pʊʊ-rɔ` ‘ish’ pl. pʊʊ-nʊ` tɛ-gɛ` ‘hole pl. tɛ-hɪ ̰̀ c.

Noni [Bantoid; Cameroon] (Hyman 1981:9, 11, 12) lów-̄ ‘bean’ pl. ɛ̄̄-lów t̄m-é ‘axe’ pl. ɛ̄-t̄m f̄-t̄n̄ ‘squirrel’ pl. m̄-n-t̄né-m bv̄-dvú̄ ‘place’ pl. mɔ̄-n-dvúú-m

Derived nominalizations, on the other hand, may have internal structure. The stems of the following Luganda [Bantu; Uganda] deverbal nouns thus consist of a verb root plus any of the specialized ive derivational inal vowels /i e u o a/ (Ashton et al. 1954:373–380): (4)

-i

: -lim-kol-e : -sib-laaw-u : -tamiir-kyúuk-o : -sek-lóot-a : -sik-geend-záal-

‘cultivate’ ‘work’ ‘tie’ ‘castrate’ ‘be drunk’ ‘change’ ‘laugh’ ‘dream’ ‘inherit’ ‘go’

→ → → → → → → → → →

ò-mú-lím-í ò-mú-kóz-í ò-mù-síb-è ò-mù-lááw-è ò-mù-támììv-ú ò-mù-kyúùf-ú è-n-sék-ó è-kì-róòt-ó ò-mú-sík-á à-má-géénd-á

‘bear a child’→ ò-lù-záàl-á

‘cultivator’ ‘worker’ ‘prisoner’ ‘eunuch’ ‘drunkard’ ‘a convert’ ‘laughter’ ‘a dream’ ‘heir’ ‘outward journey’ ‘child-birth’

(agent) (patient) (experiencer) (result) (various)

Among other derivational processes, the combination of applicative -irplus inal -o derives nouns indicating the location where the verb action or state takes place: -fuumb- ‘cook’ → è-f-fúúmb-ír-ó ‘kitchen’, -lwaal- ‘be sick’ →è-d-dwáál-ó ‘hospital’. The examples in (4) also show that many of the meanings consistently line up with speciic noun classes, for example, human class 1 singular ò-mu-. In Kom [Bantoid; Cameroon], deverbal adjectives produce an agglutinated sequence of both preixes and sufixes (Hyman 2005:314): (5)

a. -béé-

‘be big’

→

̄-f̄-ɲwí̄n AUG-19-bird

b. -béé-lé- ‘be big’ [plural]

→

f̄-béé-né-fé 19-big-ADJ-19

‘big bird’

̄-t̄-bí̄ t̄-béé-lé-né-té ‘big kolanuts’ AUG-13-kolanut 13-big-PL-ADJ-13

In (5) the sufix -ń- derives adjectives from verbs. In (5a), the adjective ‘big’ carries both a preix f̄- and a sufix -f́, agreeing with the class 19 noun ̄-f̄-ɲ wí ̄n ‘bird’. In (5b) ‘big’ agrees with the class 13 noun ̄-t̄bí̄ ‘kolanuts’, which, being plural, carries over the pluractional sufix -ĺ- from the verb -b́́-ĺ- ‘be(come) big’; -ń- is an adjective-deriving

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

196

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

formative. The result is ive agglutinated morphemes with noun class preixation and sufixation. While there are even more elaborate morphological structures in certain NC languages, others have very little morphology, ultimately few, if any, preixes or sufixes marking noun classes or anything else, for example, in Gbe, Defoid, Edoid, and part of Nupoid (McWhorter 2016). This is part of the diversity problem alluded to in 9.1. While the initial V- or N- of nouns in languages such as Igbo, Yoruba, Nupe, and others are clearly relics of older noun class systems, they are almost largely frozen and unpredictable, such as, Yoruba ̄̌́ ‘work’, ̄jò ‘snake’, ɔ` rɛˇ ‘friend’ (Pulleyblank 1986:116).

9.4 The Prosodic Stem An important number of NC languages spoken in Western and Central Africa are characterized by stem-initial prominence, deined mostly (but not exclusively) on the basis of segmental distributional asymmetries, whereby the stem-initial C or CV sequence allows more segmental contrasts than any other position within the stem. A good example comes from Ibibio [CrossRiver; Nigeria] (Urua 1999; Akinlabi & Urua 2003; Harris [1994] 2004), where a prosodic stem can have any of the shapes CV, CVC, CVVC, CVCV, CVCCV, or CVVCV. (VV indicates a long vowel and CC a geminate consonant.) Within the maximally bisyllabic and trimoraic structure, the irst consonant (C1) position can be any of the 13 consonants /b t d k kp m n ɲ ŋ f s y w/, while the second consonant (C2) slot is limited to six possibilities, realized as in Table 9.2. Representative CVC, CVCV and CVCCV examples are presented in (6). CVC

(6)

‘drag’

CVCV tòβó ‘make an order’ ‘stop’ t`ɪɾé

sItté

‘vomit’ ‘do, perform’ ‘carry with hand’ ‘go’

fèɣé yòmó tʌ` nɔ´ sàŋá

sʌ` kkɔ´ dámmá ɲànná wɔ´ŋŋɔ´

/P/

:

dIp

‘hide’

/T/

:

dʌ̀t

/K/ /m/ /n/ /ŋ/

: : : :

kɔ` k nám bén sàŋ

‘run’ ‘talk noisily’ ‘discipline’ ‘walk’

CVCCV dáppá

‘dream’ ‘remove stopper’ ‘faint’ ‘be mad’ ‘stretch’ ‘turn’

Table 9.2 Realization of C2 consonants in Ibibio /P/ /T/ /K/ /m/ /n/ /ŋ/

: : : : : :

CV(V)C p t k m n ŋ

CV(V)CV β ɾ ɣ m n ŋ

CVCCV pp tt kk mm nn ŋŋ

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

197

In the CVC column we see that the inventory of six coda consonants are identical to those found in several of the Grassields Bantu languages in 9.1. What is signiicant is that C2 consonants exhibit the same six contrasts internal to the stem. In the CVCV column oral /P, T, K/ spirantize to [β, ɾ, ɣ], while they geminate as voiceless stops in the CVCCV column. One can also notice in the above that there are exactly six contrasting vowels in the irst syllable: [ɨ, e, ʌ, o, ɔ, a]. Whether occurring in an open or closed syllable, the two central vowels [ɨ, ʌ] are realizations of /i, u/ stem-internally (cf. dí ‘come’, kpù ‘be in vain’). With this in mind we can now point out that despite the four realizations of the second vowel in the word, these all derive from a single underlying vowel, presumably /a/, which is realized [e] after /i/ [ɨ] and /e/, [o] after /o/, [ɔ] after /u/ [ʌ] and /ɔ/, and [a] after /a/. For all of the above reasons, Akinlabi and Urua (2003) conclude that the Ibibio stem constitutes a single, maximally binary foot, whose irst syllable is strong and second syllable is weak. Crucially, a vowel preix falls outside the prosodic stem and thus has no effect on the C1 consonant. The same kind of observations have been made for languages both to the East and West of the Cross-River subgroup, to which Ibibio belongs. A widespread property of Northwest Bantu languages, such effects were irst thoroughly described by Paulian (1975) in Kukuya [Bantu; Congo] (cf. Hyman 1987, 2008), whose prosodic stem has the properties in Table 9.3. In Kukuya, stems may have from one to three syllables and be mono-, bi-, or trimoraic, taking one of the ive indicated tonal melodies. While a large consonant inventory is attested stem-initially, only six are attested in C2 or C3 positions, among which are the underspeciied consonants /P/, /T/, and /K/, which are realized as [b ~ β], [r], and [k~g~ɣ] respectively. Furthermore, out of the 36 (i.e., 6 × 6) possible C2-C3 combinations, only six are attested. These combinations may not include consonants produced at the same place of articulation or disagreeing in nasality, and must be either coronal C2 + non-coronal C3 or velar C2 + labial C3. Finally, since preixes fall outside the prosodic stem, they too have a very limited consonant inventory, viz. /P, K, l, m/. A perhaps even more dramatic case comes from closely related Tiene, which has the same ive prosodic stem shapes: CV, CVV, CVCV, CVVCV, and CVCVCV (Ellington 1977; Hyman 2010). While the ‘strong’ C1 position can be occupied by any of the 13 consonants /p t k b d f s v l y m n ɲ/, the other ‘weak’ C2 and C3 positions are again restricted. Table 9.3 Properties of the prosodic stem in Kukuya Five syllable shapes Five tonal melodies Six attested C2 or C3 Six C2-C3 combinations Prefix consonants

: : : : :

CV, CV.V, CV.CV, CVV.CV, CV.CV.CV L, H, LH, HL, LHL P, T, K, l, m, n C-n-m, C-T-K, C-l-K, C-l-P, C-K-P, C-T-P P, K, l, m

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

198

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Table 9.4 Some NC languages with prosodic stem-like asymmetries Family

Languages (references)

NW Bantu

Kukuya (Paulian 1975; Hyman 1987), Tiene (Ellington 1977; Hyman 2010), Basaa (Hyman 2008), Eton (van de Velde 2008) Ibibio (Harris [1994] 2004; Akinlabi & Urua 2003), Gokana (Hyman 2011a) Izere (Blench 2001; Hyman 2010), Birom (Blench 2005; Hyman 2010) Lua (Boyeldieu 1985), Kim (Lafarge 1978), Day (Nougayrol 1979), Mundang (Elders 2000), Mambay (Anonby 2010), Mumuye (Shimizu 1983), Dii/Duru (Bohnhoff 2010) Gbaya Kara Bodoe (Moñino & Roulon 1972) Konni (Cahill 2007), Koromfe (Rennison 1997) Guro (Vydrin 2010a), Mano (Khachaturyan 2015)

Cross River Plateau Adamawa

Gbaya Gur Mande

In CVCVCV stems, C2 must be coronal, C3 must be non-coronal (either labial or velar), and C2 and C3 must agree in nasality. Where a coronal sufix such as causative -es- would violate these constraints, it must be inixed. Thus compare dím-à ‘become extinguished’ versus dísèb-ɛ´ ‘extinguish (tr.)’, where /m/ has also been denasalized to agree with oral /s/. Since Proto-Bantu allowed a free distribution of consonants within the stem (Teil-Dautrey 2002, 2008; Hyman 2008), restrictions such as in Tiene have to be innovations. Such distributional asymmetries are especially common in BenueCongo, but are also attested in most if not all of the branches of NC, including Adamawa, Plateau, Gur, and as far west as Mande. While a more systematic typological and comparative study is still needed to map out the exact nature and distribution of prosodic stem effects, Table  9.4 presents a partial survey of such prominence asymmetries within NC. Since such asymmetries have also been noted in unrelated languages in close proximity with NC, for example, in Chadic languages such as Goemai (Hellwig 2011), Tumak (Caprile 1977), and Ndam (Bross 1988) and the isolate Laal (Lionnet 2014), areal factors must be taken into account as well. However, the similar properties noted in the former ‘Khoisan’ groups Khoe, Tuu, and Kx’a (e.g., Beach 1938; Traill 1985; Miller-Ockhuizen 2001; Nakagawa 2010) suggest that limiting the distribution of Cs, Vs, and tonal patterns may more generally be a derived property in languages that have undergone considerable simpliication and restructuring.

9.5 Segmental Phonology In this section we briely consider a few of the most noteworthy properties of consonant and vowel inventories in NC languages, starting with consonants.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

199

Table 9.5 Shekgalagari consonants Bilabial Dental Alveolar Stops, p ph b affricates Fricatives Nasals m Trill/tap Glides Lateral Clicks ʘ

9.5.1

th t d

ts tsh tsw tshw sz

n

Post-alveolar Palatal tʃ, tʃh dʒ tʃw, tʃhw ʃ ʒ ʃ w ʒw

r

c chɟ cw, chw

k kh g q qh

ɲ

ŋ

y ǀ

Velar Uvular Glottal

χ

h

w

l

Consonant Systems

NC languages vary considerably in the complexity of their phonetic and phonological consonant systems. We have already cited two languages, Ibibio and Tiene, that contrast a small inventory of 13 underlying consonants. Contrast this with the consonant system of Shekgalagari [Bantu] Botswana in Table 9.5, based on Lukusa and Monaka (2008:12). Among the 54 NC languages in UPSID (Maddieson & Precoda 1990), the size of their consonant systems ranges from 13 in Eik (closely related to Ibibio) and Klao [Kru; Liberia] to 43 in Oh̃h̃ Igbo [Igboid; Nigeria], which, like Shekgalagari, contrasts aspirated and unaspirated stops as well as labialized consonants. All but seven fall into the 18–32 consonant range. The exact number often depends on the interpretation, for example, whether to analyse Cw as a labialized consonant /Cw/ or a sequence of a consonant +/w/, and similarly for whether NC is a prenasalized consonant /NC/ or a nasal+consonant cluster. In addition, results will be different depending on the level of analysis. For example, based on Welmers (1962), UPSID reports a contrast between /p t k kw kp f s/ and /b d g gw gb v z/ in Liberian Kpelle [Mande; Liberia] despite the fact that voicing is almost completely predictable, mostly being derived after a nasal, for example, the H(igh) tone nasal preix ‘my’ in the following examples (where ̰ indicates nasalization): (7)

stem pólù

‘my’

túɛ´ kɔ´ɔ´ kpíŋ` fíí sṵ́á̰

́dúɛ´ ŋ´gɔ´ɔ´ ŋ´gbíŋ` m ´víí ́zṵ́á̰

‘his/her’

m ´ bólù b`bólù d`dúɛ´ g`gɔ´ɔ´ g`gbíŋ` v` víí z`zṵ́á̰

‘back’ ‘front’ ‘foot, leg’ ‘self’ ‘hard breathing’ ‘nose’

A complication arises in the ‘his/her’ forms, which begin with a L(ow) tone fully voiced obstruent and which Dwyer (1974) interprets as geminate. Alternations involving sonorant-initial stems show that the ‘his/her’

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

200

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

morpheme can be analysed morphophonemically as a L tone nasal: l̄̄ ‘mother’, n´̄̄ ‘my mother’, n`̄̄ ‘his/her mother’, ý́ ‘hand/arm’, ɲ´́́ ‘my hand/arm’, ɲ`́́ ‘his/her hand/arm’. The ‘his/her’ forms can thus be set up with underlying voiceless initials, for example, /m  -pólù/, /n  -túɛ´/, and so on, thereby reducing the size of the underlying consonant system. Kpelle illustrates several widespread properties of NC languages: The irst is the frequent presence of labial-velar /kp, gb/ (and less commonly /ŋ m/) in the western branches (see Cahill 1999; Clements  & Rialland 2008:42–44, Güldemann 2008:156–158). The second, characterizing also Bantu languages in the east, is the tendency to avoid voiceless consonants after a nasal, including voiceless prenasalized stops (typically missing from Chadic languages as well). In addition, alternations such as l̄̄ ‘mother’ versus ́̄̄ ‘my mother’ show a common relation between oral and nasal sonorants. It is not uncommon for voiced oral consonants to become fully nasal in the environment of a nasalized vowel. Thus compare the intensive derivation of the following forms in Kana [Cross-River; Nigeria]: bà ‘eat’ → bà-gàrà versus gɔ ~¯ ‘hide’ → gɔ̰̀ -ŋàná (Ikoro 1996:153). There may also be changes of nasal to oral, as in Grebo [Kru; Liberia], where /n/ denasalizes to [l] and transfers its nasality to the following vowel in fast speech: pone → ple˜ ‘rat’, kene → kle˜ ‘wickerwork basket’ (Innes 1966:15–16). Initial ‘consonant mutations’ are especially prevalent in Northern Atlantic languages, as seen in the following singularplural pairs from Fula (Arnott 1970): (8)

singular gor-ko mbaal-u haak-o

plural wor-ɓe ‘man’ baal-i ‘sheep’ kaak-e ‘leaf’

Among other common consonant processes is the ‘hardening’ of glides to fricatives or stops, particularly post-consonantally. This is well known in the Kirundi/Kinyarwanda and Shona subgroups of Bantu, but also in Northern Nigeria. Thus, Shimizu (1980:100) reports the realizations given in Table 9.6 in a number of Jukunoid languages and dialects. Such developments are also found in the Nupoid languages, which produce correspondences such as Gwari Kuta ̄pyá, Ganagana ̄pʃ ̄, and Nupe ̄tsw̄ ‘moon’ (Hyman  & Magaji 1971:7). In the last column in Table  9.6

Table 9.6 Glide hardening in Jukunoid py by fy vy mby

[pʃ] [bʒ] [fʃ] [vʒ] [mbʒ]

bw fw vw mw

[bg] [fk] [vg] [mŋ]

tw

[tk]

kw

[kp]

dw sw

[dg] [sk]

gw

[gb]

nw

[nŋ]

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

201

it can also be seen that labial-velars can result from a velar followed by [w], as is happening synchronically in Noni [Bantoid; Cameroon] kw̄n ~ kp̄n ‘irewood’, gwɔ´ɔ´n ~ gbɔ´ɔ´n ‘bamboo’, where the labialization is from an earlier *u- preix (cf. the plurals k̄n, gɔ´ɔ´n, which lack a preix). On the other hand a labial consonant + [w] can also result in a labial-velar, as in Aghem [Bantoid; Cameroon]: kí-bɛ´ ‘fufu’, pl. ó-gbɛ´; kI´-báʔ ‘rope’, pl. ó-gbáʔ (Hyman 2011b:15). An analogous property of liquid hardening is responsible for special realizations of /l/ before glides or high vowels. There are a number of languages where /l/ is realized as [d] before /i/, for example, Ciyao [Bantu; Mozambique] mil-a ‘swallow’ versus mid-isy-a ‘swallow a lot’; kul-a ‘grow big’ versus kud-isy-a ‘grow very big’ (Ngunga 2000:56). In Gbari [Nupoid; Nigeria], /l/ is realized as [ɗ] before /i/, for example, ɗ í ‘eat’. In Gwari Kuta /l/ has further developed into a palatal stop with offglide before both /i/ and /e/, such as gyí ‘eat’, gyè ‘sharpen’ (Hyman & Magaji 1971:14, 15). It is perhaps worth noting in this context that implosives are very common in NC, a commonality they share with many Chadic and Nilo-Saharan languages. The implosives /ɓ, ɗ/ are the most common, as in Dan [Mande; Ivory Coast] ɓ̄ ‘be’, ɗ ̄ ‘rain (n.)’ (Vydrin  & Mognan 2008:107, 128), and many languages have only these two, for example, Degema [Edoid, Nigeria] and Kalabari [Ijoid, Nigeria]. A few languages have a full set of contrasts, including Ega [unclassiied NC, Ivory Coast] which has the voiced plosives /b d J g gb/ and the implosives /ɓ ɗ ʄ ɠ ɠɓ/, for example, /ùgbò/ ‘cooking pot’ and /u`ɠɓò/ ‘mouse (sp.)’ (Connell et  al. 2002). Several NC languages have typologically rare voiceless implosives as well. Most notably, Sereer makes use of a phonemic voicing distinction for implosives at three places of articulation: /ɓ ɗ ʄ/ versus /ƥ ƭ ƈ/, exploited by the language’s consonant mutation system as well as for underlying lexical contrasts. Implosives are largely found in NC languages in the Sudanic Belt (which excludes most Bantu languages). Clements and Rialland (2008:58), however, demarcate an area within this macro-area where implosives are systematically absent, which includes languages within Dogon, Senoufo, Gur languages (Mòoré, Kabiyé), Kwa languages (Akan, Guang, Gbe), Yoruba, most Edoid languages, and westernly located Ijoid languages (e.g., Izon varieties). They note that ‘there is comparative evidence that earlier implosives shifted to non-implosive sounds, for example, *ɓ > b/v, *ɗ > d/ɖ/l in Central Gur (Manessy 1979)’. Some NC languages in West Africa make a distinction between fortis and lenis consonant series, on top of a voiced/voiceless distinction, where lenis consonants are indicated by or with . Stewart (1973) notes Ebrié [Kwa; Ivory Coast] fortis consonants as /p b/ and lenis consonants as /’p ’b/. Botma & Smith (2006) summarize the phonetic research on this distinction, noting that the ‘lenis stops of [Ebrié] are produced with less energy than their fortis counterparts, they have a shorter duration, they involve spontaneous voicing and they are realized as

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

202

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

non-exploded, often with an implosive quality’, and note comparable observations for lenis stops in Edoid (Elugbe 1980). Welmers (1973: 46–47) indicates that in the Edoid language Etsako (Ekpei dialect), fortis /kp gb/ occur with a suction sound, whereas lenis /kph gbh/ do not. We ind it unlikely that fortis/lenis distinctions have uniform phonetic realizations across NC. Fortis/lenis distinctions have played a role in a number of synchronic and diachronic proposals in early NC. Stewart (1973) reconstructs fortis stops (e.g., *t *d) versus lenis stops (e.g., *’t *’d) for ‘Proto Volta-Bantu’ which would encompass modern day Kwa and Bantu families, retained in Ebrié above. Leynseele and Stewart (1980) explicitly reconstruct fortis and lenis stops for NW Bantu, a distinction used to account for so-called ‘double relexes’ (Hedinger 1987:105–108; but see Bachmann 1989, Blanchon 1991, and Janssens 1993 for a different interpretation of the NW Bantu facts). Proto-Edoid is reconstructed with a full set of fortis and lenis consonants, including the labial series *p, *ph, *b, *bh, *ɓ, *m, *mh, *f, *v, largely maintained in the Ibilo dialect of Okpamheri (Elugbe 1989:297). This reconstructed set illustrates that fortis/lenis cannot simply be reduced to voicing, implosiveness, sonority, or frication. Fortis/Lenis oppositions are also proposed for Delta Cross (Connell 1994:8), for example, Abua (Gardner 1980). Leggbo, a related language in Upper Cross, also has both a fortis and lenis series of consonants, for example, d̄ ‘beat, pound’, dd̄ ‘whisper’. Fortis consonants are transcribed as double graphemes, and are articulated with greater force, have greater duration, have a shorter following vowel, and have differences in intra-oral air pressure proiles (Udoh 2004). This highlights the fact that the fortis series is consistent with an interpretation as gemination on a case-by-case basis. Relatedly, Clements and Rialland (2008:57) classify implosives as ‘non-obstruent’ stops, and explicitly relate them to the use of the imprecise term ‘lenis’ in the Africanist literature (cf. Clements & Osu 2002). Future work should seek to determine potential correspondences, such as whether fortis/lenis series correspond in different proto-languages, lenis consonants correspond to implosives, or fortis consonants correspond to consonant clusters supporting incipient gemination. There are a number of notable phonological contrasts not typically found in NC. NC languages do not frequently make use of breathy and creaky phonation types, although breathy consonants are found in the Oh̃h̃ variety of Igbo (Green  & Igwe 1963) and the Ibilo dialect of Okpamheri (Elugbe 1989:297). Sylak-Glassman (2014:55) notes that post-velar nonglottal sounds (i.e., uvular, pharyngeal, epiglottal) are largely absent in NC languages, although they are found in peripheral Bantu languages (e.g., Shekgalagari above, Nyaturu /ʁ/, and Datooga /q/, which has the realizations [q G χ ʁ]), in certain Senufo languages (Mamara), Wolof and Sereer within Atlantic, and in the Mande language Susu. Further, clicks in NC are also only found in the Southernmost Bantu languages, such as in SothoTswana and Nguni groups. This is uncontroversially due to sustained contact with ‘Khoisan’ languages (Gunnick et al. 2015). Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

9.5.2

203

Vowel Systems

While some NC languages have the unremarkable ive-vowel system /i, e, u, o, a/, for example many Bantu languages, it is much more common to ind vowel systems involving a greater number of vowel quality distinctions and, particularly in certain western areas, nasalized vowels. Some NC languages also contrast vowel length. Among the 54 NC languages in UPSID, the size of their vowel systems ranges from 5 in one language (Zulu [Bantu; South Africa]) to one language (Dan [Mande; Ivory Coast]) with 19 – in fact, 21 in Eastern Dan (Vydrin & Mognan 2008). While Zulu is alone, four other languages expand to six vowels by adding either /ə/ (Dagbani, Tarok) or /ɔ/ (Bisa, Mambila). Much more common is the oral vowel system in (9a), where the languages on the right are listed roughly from west to east: (9)

a. i

u

e ɛ

o ɔ

(Senadi, Ga, Ewe, Lelemi, Yoruba, Eik, Birom, Noni, Gbeya, Sango)

a b. i ɪ ɛ

u ʊ ɔ

(Kpelle, Teke, Jomang)

a

In the languages in (9a), there is tense/lax or advanced tongue root (ATR) contrast only among the mid vowels, for example, Mbembe [Bantoid; Cameroon] bí ‘ask’, b́ ‘call’, bɛ´ ‘dog’, b̄ ‘thing’, b̄ ‘they (human)’, bɔ` ‘beg’, bá ‘legs’ (Richter 2014:17). In contrast, the vowel system in (9b), which has an ATR contrast only in the high vowels, is much rarer, especially outside Bantu. Thus, in Kinande [Bantu; DRC]: -lím- ‘extinguish’, -lIm- ‘cultivate’, -lem- ‘fail’, -lúm- ‘be very lively’, -lʊ´m- ‘bite’, -tóm- ‘put aside’, -lam- ‘recover from a disease’ (Mutaka & Kavutirwaki 2011). Although both systems have been considered, a consensus seems to be forming that (9b) should be reconstructed at the Proto-Bantu stage (Schadeberg 1995:73; Stewart 2000/2001:46). Among the larger ATR vowel inventories are those in (10a) which have nine vowels, lacking an ATR contrast only for /a/, and those in (10b) which have ten vowels dividing into two sets: (10)

a. i ɪ e ɛ

u ʊ o ɔ

(Joola, Aizi, Tampulma, Isoko, Kalabari)

a b. i ɪ e ɛ

ɜ a

u ʊ o ɔ

(Ogbia, Amo, Kokumono, Degema)

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

204

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Table 9.7 Williamson’s featural analysis of Godié vowels [expanded] [front] [high] [mid] [low]

i̘ e̘

i̘ ə̘

[constricted]

[round]

[front]

u̘ o̘

ɪ̙ ɛ̙

a

[round] ʉ̙ ʌ̙

ʊ̙ ɔ̙

Other systems also exist lacking one or another of the above vowels, for example, Igbo /i, ɪ, e, u, ʊ, o, ɔ, a/, which lacks /ɛ/ and Lokaa /i, e, ɛ, u, o, ɔ, a, ɜ/, which lacks both /ɪ/ and /ʊ/, but has /ɜ/ (Williamson 2004:129). Several Kru languages have developed central vowels with ATR distinctions, for example, Godie (Marchese 1983a), as represented by Williamson (2004:132) using Lindau’s (1978) pharynx features Expanded and Constricted (see Table 9.7). While /a/ occurs with both sets of vowels, as in many ATR systems (see 9.6.1), it functions with the [−ATR] or contracted vowels in certain phonological processes. However, the cooccurrence of interior vowels with ATR in Godié and other languages in the Kru/Mande contact area is quite unusual. As ATR typically correlates with a distinction along vocalic height (F1) more so than backness (F2), languages which show ATR contrasts and harmony tend not to have interior vowels, and vice versa. Rolle et al. (2017) show that this inverse relationship is statistically signiicant within a sample of African languages. Many NC languages in Central Africa have interior vowel phonemes, that is, /ɨ ʉ ɯ ɘ ə. . ./, for example, within Grassields/ North Bantoid, Ubangi, Adamawa, Platoid, and Kainji. The same Rolle et al. study shows that the development of interior vowels is an areal phenomenon that cuts across the different phyla in the area. In some of these languages, expansion of vowels along the front/back dimension correlates with fewer distinctions along height, for example, the Jukunoid language Etkywan (Icen) in Eastern Nigeria, which has the vowels /i e ɨ a u o/, as well as their nasal counterparts (Shimizu 1980:72). (Nasalized vowels will be treated in 9.6.2.)

9.6 Vowel and Nasal Harmonies African languages are known for different types of vowel and nasal harmonies, which will be treated in the following two subsections.

9.6.1

Vowel Harmonies

NC languages are well known for having several types of vowel harmony. The most famous type is ATR harmony, as mentioned above, which has an Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

205

Table 9.8 Degema ATR contrasts [+ATR]

high mid low

[−ATR]

front

central back

front

central back

i̘/ i e̘ / e

u̘ / u o̘ / o

i̙ / ı e̙ / ɛ

u̙ / ʊ o̙ / ɔ

a̘ / ɜ

a̙ / a

extremely rich phonological and phonetic literature (catalogued in Casali 2008). In this type, vowels form two mutually exclusive groups: in the ATR set vowels are articulated with an advancement of the tongue root widening the pharyngeal cavity; in the other, vowels are articulated with retraction of the tongue root (RTR). Degema is an example with a full set of contrasts (Kari 2004:370). Assuming ATR as a binary feature, [+ATR] is indicated with [  ̘] and [−ATR] with [ ̙]. This distinction is often times transcribed as tense vowels versus lax vowels. Both transcriptions are shown in Table 9.8. In Degema, these vowel sets cannot co-occur, whether within roots or across morphemes. As a result, the nominalizer o- has two allomorphs, [+ATR] /o-/ and [−ATR] /ɔ-/ (Kari 2008:xxxi): (11) [+ATR] [−ATR]

verb ɗeginɜ tɛvtɛ´v

‘be old’ ‘be short’

nominal o-ɗégínɜ ɔ-tɛvtɛ´v

‘one that is old’ ‘one that is short’

[+ATR] vowels canonically have lower irst formant frequency (F1) than their [−ATR] counterparts. As F1 is also the primary cue for tongue height, it has led to the convention of transcribing ATR with different vowel heights, for example, [e] versus [ɛ]. Casali (2008:510) further notes that in languages where ATR is additionally cued through voice quality, ‘[+ATR] vowels have been characterized . . . as having a “breathy”, “deep”, “mufled”, or “hollow” quality, while [−ATR] vowels have been described by terms such as “bright”, “choked”, “brassy”, or “creaky”’. ATR harmony is found throughout NC in virtually every major family as summarized in Casali (2003:312), including Atlantic, Mande, Gur, Kru, Kwa, Ijoid, Benue-Congo (Bantu, Defoid, Edoid, Igboid, Nupoid), Ubangi, and Kordofanian. Notably, it is largely absent in the centrally located families Gbe, Grassields/Bantoid, Kainji, Platoid/Jukunoid, and Adamawa. ATR harmony is a robust areal feature of the Sudanic Belt, common also in Nilo-Saharan, although less so in Afroasiatic. This ATR macro-area can be further divided into a Western and an Eastern ATR zone, with the largest concentration of NC languages with ATR in the Western zone. In some languages between these zones, there has been a loss of ATR harmony, correlating with extensive reduction of morphology and vowel mergers. Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

206

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

One example is Edo where even mid vowels can co-occur, for example, the name /ɛ` dó/ ‘Edo people, Benin City’. Casali (2003, 2008) diagnoses speciic sub-types of ATR, for example, whether the language has two heights in high vowels (/i/ vs. /I/), whether the plus or minus value is dominant in phonological processes, whether roots are ATR dominant, among other dimensions. Casali is liberal in what he considers an ATR system, including languages such as Yoruba with /i e ɛ a ɔ o u/ where /e o/ do not co-occur with /ɛ ɔ/, but high vowels /i u/ can occur with either set. In such 7-vowel systems, there is no clear cross-height harmony present as in 9-vowel ATR systems. Rolle et  al. (2017) classify this 7-vowel type as ‘trace ATR’ systems to distinguish them from other ATR systems. Such trace systems are found in many Mande languages, Western Kru family, Dogon family, Gbaya family, and Idomoid. A number of other vowel harmonies occur throughout NC as well. Height harmony, which is not reducible to ATR harmony, is quite common. Hyman (1999) presents an extensive survey of Bantu vowel harmony, illustrating symmetrical and asymmetrical height harmony in 7V languages. Under symmetrical harmony, the Proto-Bantu high vowels*I, *ʊ lower to mid [ɛ, ɔ] when adjacent to mid vowels. Under asymmetrical harmony, Proto-Bantu *ɪ is lowered after both mid vowels, whereas *ʊ lowers only when following *ɔ. Asymmetric height harmony of Nyamwezi’s vowel system /i I ɛ a ɔ ʊ u/ is shown below with the applicative sufix /-ɪl-/ and the reversive transitive or ‘separative’ sufix /-ʊl-/ (Maganga & Schadeberg 1992:155): (12)

Root V i I

u ʊ ɛ ɔ a

Appl /-Il-/ -shik-Il-a -βIt-´I l-a -zug-´I l-a -gʊl-Il-a -lɛk-ɛl-a -βɔn-ɛ´l-a -ap-´I l-a

‘arrive+appl’ ‘pass+appl’ ‘cook+appl’ ‘buy+appl’ ‘let, leave+appl’ ‘see+appl’ ‘collect honey+appl’

Rev. Trans. /-ʊl-/ -βis-ʊ´l-a ‘ind out’ -pIInd-ʊl-a ‘overturn’ -gub-ʊ´l-a ‘take off lid’ -shʊʊn-ʊl-a ‘show teeth’ -zɛɛng-ʊl-a ‘build’ -hɔng-ɔ´l-a ‘break off’ -gaβ-ʊl-a ‘divide’

Another type of height harmony affects the low vowel /a/. In a number of NW Bantu languages such as Babole [Congo], /a/ becomes [ɛ] in the context of /ɛ/ and [ɔ] in the context of /ɔ/ (Leitch 1996:123): (13) (a)

(b) (c)

Root vowel /i/ /e/ /u/ /o/ /a/ /ɛ/ /ɔ/

Imperative sìl-á kèl-á tsùm-á kòh-á sál-á hɛ´k-ɛ´ kɔ´s-ɔ´

Passive /-am-a/ -sìl-ám-á -kèl-ám-á -tsùm-ám-á -kòh-ám-á -sál-ám-á -hɛ´k-ɛ´m-ɛ´ -kɔ´s-ɔ´m-ɔ´

Applicative /-el-a/ -sìl-él-á ‘sharpen’ -kèl-él-á ‘make’ -tsùm-él-á ‘dip’ -kòh-él-á ‘take’ -sál-él-á ‘do’ -hɛ´k-ɛ´l-ɛ´ ‘cut’ -kɔ´s-ɛ´l-ɛ´ ‘gather’

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

207

In (13a) the input sufix vowels are realized without change. On the other hand, the /a/ of the inal vowel and the passive sufix become [ɛ] after the root vowel /ɛ/ in (13b) and [ɔ] after the root vowel /ɔ/ in (13c). The applicative forms in the last column show that Babole also harmonizes /e/ to [ɛ] after both /ɛ/ and /ɔ/, which then triggers the /a/ → [ɛ] change on the inal vowel. Cases of /i/ → [u] rounding harmony are found in other languages, for example, Punu [Bantu; Gabon] (Fontenay 1980; Hyman 2008:329–330): (14)

a. -bíng-asan-a → -bíng-əsən-ə -bund-igil-a → -bund-igil-ə

‘roll (sth.)’ ‘slander’

(general, ‘default’ inal /-a/)

b. -bíng-asan-i -bund-igil-i

‘roll (sth.)’ ‘slander’

(present, subjunctive /-i/)

c.

→ -bíng-isin-i → -bund-igil-i

-bíng-asan-u → -bíng-usun-u ‘be rolled’ (passive /-u/) -bund-igil-u → -bund-ugul-u ‘be slandered’

In Punu, sufixal vowels are limited to /i u a/. An /a/ or sequence of sufixal /a/s is realized [ə] word-inally, as in (14a), but as [i] before inal /-i/, as in (14b). (14c) shows that both /i/ and /a/ assimilate to [u] before a inal /-u/. Rounding harmony is widespread also outside of Bantu and found, among others, in Tommo So [Dogon; Mali] (McPherson  & Hayes 2016), Alladian [Kwa; Ivory Coast] (Duponchel  & Mel 1983), Kalabari [Ijoid; Nigeria] (Akinlabi 1997), Samba-Leko [Adamawa; Nigeria] (Fabre 2002), Iceve-Maci [Bantoid; Cameroon] (Cox 2013), Zande [Ubangi; Central African Republic] (Tucker 1959; R. Boyd 1995), and Southwest Gbaya [Ubangi; Central African Republic] (Moñino 1995:86–92, 99–104). Many NC languages do not show vowel harmony per se, but have cooccurrence restrictions which limit possible combinations of vowels. Cishingini [Kainji; Nigeria] has neutral vowels /i u/ and non-neutral vowels /ɘ o e a/ (Stark 2000). Non-neutral vowels may co-occur with neutral vowels, but may only co-occur with an identical non-neutral vowel (Stark 2000:57). This is summarized in Table 9.9. Further, although Gbaya languages do not have clear active harmony, they also have robust vowel co-occurrence restrictions, many of which can be reconstructable and are diachronically stable. Moñino (1995:378) Table 9.9 Cishingini co-occurrence restrictions V1

i u

ɘ

e o a

V2 i √ √ √ √ √ √

V1 u √ √ √ √ √ √

ɘ √ √ √ * * *

V2 e √ √ * √ * *

V1 o √ √ * * √ *

a √ √ * * * √

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

208

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

i

u

e

o

ε a Figure 9.1 Vowel sequences in Proto-Gbaya

reconstructs the Proto-Gbaya vowel system, and provides the following restrictions in Figure 9.1. Solid lines indicate vowel sequences that cannot occur in either order, while dashed lines indicate vowel sequences that cannot only occur in the order of the arrow. Thus, there is no word that can be reconstructed with the shape CiCu, but there are words of the shape CuCi. Finally, a number of NC languages show a propensity towards having identical vowels within roots or stems. For example, while Berom [Platoid; Nigeria] has a seven vowel system, Bouquiaux (1970:98–99) writes that approximately 80 percent of all disyllabic stems have identical vowels, and similarly in C’Lela [Kainji; Nigeria], Dettweiler (2015:28) notes that approximately 61 percent (158/260) of disyllabic stems have identical vowels. Similar facts are seen in Ubangi languages, for example, BandaNdele where 47 percent of CVCV words have identical vowels (Sampson 1985:141). This may be due to an areal phenomenon in Central Africa, as similar identical vowel propensities are found in the vicinity in non-NC languages, such as in the Sara-Bongo-Bagirmi family as well as in some Central Chadic languages.

9.6.2

Nasal Harmonies

Contrastive nasal vowels, nasal harmony, and prenasalized consonants play a large role in NC languages. Rolle (2013, 2015) presents a survey of 473 West African languages (of which NC are the majority) and inds that 53 percent have contrastive nasal vowels (252/473), 40 percent lack contrastive nasal vowels (191/473), and 6 percent are ambiguous (30/473). We can compare this to cross-linguistic worldwide averages for contrastive nasal vowels, which range between 19 percent and 26 percent depending on sampling. The geographical distribution of these patterns is given in Map 9.1, from Rolle (2015). As can be seen from Map 9.1, contrastive nasal vowels are found in a continuous belt from Guinea to the Central African Republic. They are common in Mande, Kwa, Dogon, Adamawa, Gbaya, and Ijoid; mixed in Kru, Gur/Senoufo, Benue-Congo, and Ubangi; and largely absent in Atlantic to the west and Bantu and Kordofanian to the east.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Ambiguous No nasal vowels / Has nasal vowels / Map 9.1 Distribution of nasality – Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

210

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Contrastive nasal vowels are used as evidence for classifying the Sudanic Belt as a macro-area (Clements  & Rialland 2008; Güldemann 2008). Within this area, however, oral vowel zones exist in a number of areas dominated by NC languages: among North Atlantic languages in Senegal, in Eastern Kru languages in Ivory Coast, among Gur and Kwa languages in northern Ghana, and among Benue-Congo languages in the Nigeria/Cameroon borderlands. This last area extends into the Bantu spread zone, which generally lacks nasal vowels. Incipient contrastive nasalization is found, however, in a small number of Bantu languages (Maddieson 2003:23), for example, Umbundu (Schadeberg 1982), Teke (Paulian 1975; Hombert 1987), and Bitam Fang (Medjo 1997:63–64). Cross-linguistically, it is extremely common for the set of nasal vowels to be a subset of the oral vowels. This is true in most NC languages as well, for example, Esan [Edoid, Nigeria] has oral vowels /i e ɛ a ɔ o u/ but nasal vowels /̃ ɛ˜ ̃ ɔ˜ ̃/. As seen in the scale in Table 9.10, based on Rolle (2015) the most common missing vowels are /e˜ ̃/, supporting previous observations in Africanist literature, and aligning with wider typological and phonetic research. For languages that have a full set of nasal vowel counterparts, these are disproportionately found at the transition zones between oral vowel and nasal vowel zones, suggesting recent transphonologization of *VN ˜/. Still in some other languages, contrastive nasal or *NV sequences as /V vowels are marginal and found in only a few lexical items, for example, in Sango Véhiculaire (Moñino 1988), Gbaya-Mbodomo (L. Boyd 1997), and Biali (Sambieni 2005:25). Contrastive nasal vowels have been reconstructed for Proto-Mande (Vydrin 2012), in some stage following Proto-Kwa (Kropp Dakubu 2012:32), and Proto-Gbaya (Moñino 1995). Stewart (2000, 2002) speculates that nasalized vowels occurred as far back as his ‘Proto-PotouAkanic-Bantu’. However, the loss or gain of contrastive nasal vowels can be quite sudden. Nasal vowels interact with nasal consonants in complex ways in NC. In a certain number of Mande, Kru, Kwa, Gur, and Benue-Congo languages, nasal vowels are contrastive but nasal consonants are not (Clements  & Rialland 2008:47), a typologically unusual pattern. In these languages, nasal consonants are conditioned variants of oral consonants in the presence of a nasal vowel. An example is from Akan,

Table 9.10 Likelihood of nasal vowel counterparts to be absent most likely to be absent mid-close e˜ õ

>

+ATR low > near-high >

mid-open >

high and Least likely to low be absent

ə˜

ɛ˜ ɔ˜

˜ı u˜ a˜

ɪ˜ ʊ˜

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

211

where /b d/ become [m n] adjacent to nasal vowels (Schachter & Fromkin 1968:72): (15)

Underlying /b̃/ /d̃/ /j̃/ /w̃d̃/ /h̃/

Surface [m̃] [ñ] [ȷ˜̃] [w ˜ ̃ñ] ˜ ̃] [h

‘give’ ‘and’ ‘receive’ ‘scrape’ ‘fear’

This can be interpreted as a type of nasal harmony, which is extremely common in NC. In such harmony systems, nasality spreads from a sponsor onto other segments, subject to their ability to bear nasality and the presence of transparent and opaque segments. Nasal consonant harmony can take place even in the absence of contrastive nasal vowels. For example, in Yaka the perfective sufix deriving from Proto-Bantu *-ile has a number of allomorphs, including height harmony and the variants -ini/ -ene when preceded by a nasal consonant earlier in the verb stem (Hyman 1995b:6–9): (16)

a. kúd-idi kás-idi

‘chase (s. o.)’ ‘attach’

ték-ele sób-ele

b. tsúm-ini kún-ini

‘sew’ ‘plant’

kém-ene ‘groan’ kóny-ene ‘roll up’

c. mák-ini ‘climb’ míítuk-ini ‘sulk’

mék-ene nók-ene

‘sell’ ‘change’

‘try’ ‘rain’

The examples in (16c) show that the triggering nasal need not be in the immediately preceding syllable. Finally, prenasalized stops occur in most Bantu languages, and are reconstructable at the Proto-Bantu stage (Meeussen 1967), for example, *búmb- ‘mould’, *gènd- ‘go’, *táŋg- ‘read’. Prenasalized stops are not particularly common in the rest of NC, although North Atlantic languages often have them, and they operate within the consonant mutation systems, for example, marking subject number agreement on verbs in Fula (Merrill 2013:6): (17)

singular war rew yolb ʔand

plural mbar ndew njolb ngand

‘kill’ ‘follow’ ‘be loose’ ‘know’

As seen from such alternations, prenasalization can affect the realization of the consonant, for example, w ~ mb, r ~ nd. Besides having a stopping effect, prenasalization can also result in voicing, for example, Ciyao lù-tèwù ‘beard’, pl. n-dèwù (Ngunga 2000b:66). In fact, many languages only allow prenasalized voiced stops.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

212

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

9.7 Tone It is well known that NC languages, and African languages in general, tend to be tonal in Welmers’s (1959:2, 1973:80) sense that pitch is a contrastive exponent of at least some morphemes. The prevalence of tone in diverse languages within the NC family supports a view where tone is reconstructed to PNC. Almost all NC languages are tonal, including controversial inclusions (Mande, Dogon, Ijoid). Non-tonal NC languages are geographically peripheral and are assumed to have lost their tone either through natural tone simpliication processes (cf. Childs 1995b) and/or inluence from neighbouring non-tonal languages (cf. Hombert 1984:154–155). This includes not only Swahili in the East, but also Northern Atlantic (Fula, Seereer, Wolof, etc.), Koromfé [Gur] (Rennison 1997:16), and (outside NC) Koyra Chiini [Songhay] (Heath 1999b:48), due to a possible effect of contact with Berber or Arabic, either directly or through Fula (Childs 1995b:200). There is one striking geographic exception: certain inland Bantu languages in Mozambique, Malawi, and Tanzania (e.g., Sena, Tumbuka, Nyakyusa) have lost the inherited Proto-Bantu tonal contrasts. As is the case with several of the other major properties considered in this chapter, tone is not limited to NC within Africa, rather it is a sub-Saharan African areal feature (see 9.15). Beyond possessing the shared property of being tonal, NC languages show considerable variation in the speciics of the tone systems they exhibit. First, NC languages contrast anywhere from two to ive tone heights, as in Dan [Mande] (Vydrin 2008:10): (18)

Extra-high: High: Mid: Low: Extra-low:

ka˝a˝ káá k̄̄ kàà ka `` ``a

‘scabies’ ‘you pl.’ (present negative pronoun) ‘you pl.’ (prospective aspect pronoun) ‘scratch’ (in conjoint construction) ‘reed’

Second, some allow their tones to combine to form tonal contours, while others do not. Third, some have a tonal contrast on every tone-bearing unit, while in others tonal contrasts are limited to certain positions within the word or phrase. Fourth, NC languages differ in the extent to which they exploit tone for the purpose of distinguishing lexical versus grammatical morphemes. It is quite rare for a tonal NC language to have only grammatical tone, as in Chimwiini (Kisseberth  & Abasheikh 2011:1992), or to not contrast tone on nouns. However, a number of NC languages do not have a tonal contrast on verb roots, for example, Kisi [Atlantic] (Childs 1995a:171–194), Konni [Gur] (Cahill 2000:), Kulango [Gur] (Elders 2008:75), Cicipu [Kainji] (McGill 2009:133), Zande [Ubangi] (R. Boyd 1995), and all of Edoid (Elugbe 1989:299), not to mention many Narrow Bantu languages such as Kimatuumbi (Odden 1998:190), where tones are assigned by the inlectional morphology (tense-aspect-mood-negation). Most NC languages

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

213

exploit grammatical tone at least for this latter purpose, for example, marking tense/aspect (as in two of the above Dan examples). Finally, NC languages differ in the extent to which phrasal tones are affected by syntactic conigurations: In some languages there may be special rules that apply within certain prosodic domains, for example, the phonological phrase, or in speciic contexts, for example, the associative (possessive) construction, as in Igbo (Welmers 1963:441–442): (19)

a. ìtè ‘pot’, ànyí̙ ‘our’ → ìté ↓ányí̙ ‘our pot’ b. ìtè ‘pot’, àtó̙ ‘three’ → ìtè àtó̙ c.

ìtè ‘pot’, àtó̙ ‘three’ → ìté átó̙ ↓

from: /ìtè + ´ + ànyí/

‘three pots’ ‘third pot’

from: /ìtè + ´ + àtó̙/

In (19a) the L-L noun ‘pot’ becomes L-H, and the L-H pronoun becomes ↓ H-H, a sequence of two high tones realized on a downstep. As shown to the right, these changes are caused by a H ‘associative’ tonal morpheme which marks the relation, hence ‘pot of us’ (see 9.9). The H tonal morpheme is lacking in the noun + numeral construction in (19b), and therefore no tonal change takes place. When the associative morpheme occurs between the noun and numeral, as in (19c), the meaning is ‘third pot’, where ‘pot’ is ‘associated’ with the number ‘three’. The Igbo examples in (19a, c) also highlight another tonal property: languages may have downstepped tones or not. While the most common contrast is restricted to H-H versus H-↓H sequences, languages have been reported that contrast H versus ↓H after L, as well as downstepped L and M tones, particularly, but not exclusively, in Grassields Bantu (Hyman 1979). The above variations in tonal properties are not randomly distributed within NC, but may depend on either genetic or areal factors. Drawing from a database of 662 tonal languages compiled by the irst author, 243 of which are spoken in Africa, we discuss in the following subsections how each of the above tonal features is distributed within Africa.1

9.7.1

The Number of Surface Tone Heights

NC languages vary in the number of non-contour tones in their surface tone inventory. For the purpose of comparison, surface tones were encoded in order to avoid differences of analysis which could affect the number of claimed underlying tones. This surface-based count includes downstepped tones as constituting distinct surface levels. As seen in Map 9.2, which also includes West African non-NC languages, several micro-areas within Africa are deined by the number of level tones distinguished in the surface tone inventories (cf. Clements & Rialland 2008:73). 1

The tonal systems were entered with a numerical system we designed to encode equivalencies across distinct traditions of tonal orthographic representations. This coded database allowed us to explore the distribution of numerous tonal sub-features, including the number and identity of level tones, downsteps, and contour tones, as well as the direction of contour tones.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

M

T

Jukunoid

J

K

M

N

Delta Cross

Edoid

Idomoid

Gur

D

E

F

G

Y

Yoruboid (Defoid)

Kwa

Bantoid, including Bantu

Central Sudanic

Ijoid

Abadan

Accra

G

Map 9.2 Number of surface contrastive tones in West African languages

Nupoid

Mande

W

S

O

Igboid

I

Chadic

C

Kru

G

G

Burkina Faso

H

Benin

N

C

Abuja

Kano

C

N i g e r

5 tone heights

4 tone heights

3 tone heights

2 tone heights

I O E

E

to

E DD I

C

A T

C C G

C C

0

uin ea

Yaounde

Gabon

Libreville

lG

T

300

T

S

Kinshasa

600 miles

1000 km

T

T

T

Democratic Republic of the Congo

S

Central African Republic

A S

S

Bangui

S

500

750

C h a d

400

N’Djamena

200

500

CC C C C C C C A C C C

100

250

J T T T T T T T T T T Cameroon T T T T

C

Douala

ri a

F DD D

N i g e r i a Lagos

Y

Niamey

M Togo Cöte Ghana M d’Ivoire W K M Yamoussoukro W G K W W W K W Kumesi W Lomé W W K

Bamako

Songhay

Liberia

M

M

H

Monrovia

K

Sierra Leone M

Guinea

M

M

Adamawa

Conakry

GuineaBissau

M

Senegal

A

The Gambia

Dakar

M a l i

0

o

M a u r i tan ia

C

g o

n

Eq ua

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

215

Despite the fact that such a high proportion of African languages have tone, there are subgroups of languages that pattern together based on the number of distinct surface tones. As seen in Map 9.2, language families for each of the 243 languages surveyed are represented with letters. As also indicated in the legend, each letter is shade-coded for the number of surface level tones. Although truncated on the map, two-toned languages are the most common south of the Macro-Sudan Belt (Güldemann 2008), where most Bantu languages are spoken. Taking a closer look at the Macro-Sudan Belt itself, we see additional areas of similarity. For example, the languages spoken near Lake Chad tend to have 2–3 tone levels, while those in the Kru-Mandesphere and in the Nigeria-Cameroon borderland area tend to have 3–4, occasionally more.

9.7.2

Additional Tonal Inventory Features

Using the same database, we considered other tonal features as well, including the identity of tones in three-tone-height languages, whether each language has a mid or downstepped tone, and the number and identity of contour tones in each language. The (non-)intersection of these additional features provides extra evidence for certain areal patterns across Africa. For example, languages like those in the KruMandesphere, which tend to have high numbers of level and contour tones also tend to lack downstep. Also, perhaps expectedly, languages with a higher number of level tones potentially have more contour tones. In addition to making generalizations about tonal features and whether they pattern together, we can also propose micro-areas of contact and spread based on the distribution of tonal features. These include (a) the area south of Lake Chad, which shows three level tones, usually low, mid, and high, but tends to lack contours and downstep; (b) the Nigeria/Cameroon border, where there tend to be more surface level tones and contours than in closely related Bantu languages in Southern Africa, and where there tends to be at least one downstepped tone; (c) the Ghana/Togo/Burkina Faso cluster where there are often three surface level tones, including downstep, but where there is a lack of contour tones; (d) the southern part of the Mandesphere, where there are many surface level tones and contours, including complex contour tones. The Mandesphere is made up primarily of Mande and Kru languages, though Atlantic, Gur, Kwa, Dogon, and non-NC Songhay are also interwoven into this area. The working hypothesis of Vydrin (2002) is that ProtoMande had two level tones. Proto-Kru, however, is reconstructed with four level tones (Marchese 1979; Marchese Zogbo 2012). Those Mande languages geographically situated nearest to Kru tend to have larger tonal inventories

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

216

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

than those further north. These distributional indings are consistent with Vydrin’s (2009) indings, where he uses the number of tonal heights, in addition to a number of other linguistic features, to show that there is a micro-area of linguistic features where Mande languages come into contact with Kru. Thus, in Ivory Coast, the Mande languages Toura and Dan, which border Kru languages, have 4 and 5 tone heights, respectively. Moving slightly farther north away from Kru languages, however, the Mande language Mahou has only 3 surface tone heights.

9.7.3

The Function of Tone

As mentioned above, tone can have either a lexical or grammatical function. Like many tonal languages cross-linguistically, tone typically differentiates lexical items in NC, as well as in African languages in general. In Guébie, a four-height Kru language of Ivory Coast (Sande 2017), tones are marked with numbers 1–4 where 4 is high. The following examples show tone distinguishing lexical morphemes: (20)

a. si2 si3 si31 si23 si32

‘trees’ ‘laugh’ ‘snails’ ‘sweep’ ‘dry’

b. nə2 nə41 nə31 c. gbala3.4 gbala4.4

‘mouth’ ‘hole’ ‘curse’ ‘climb’ ‘sew’

However, in the same language, tone can play a grammatical role. Still in Guébie, (21) shows that tonal marking on the verb differentiates perfective from imperfective aspect: (21)

a. e4 li3 Ja31 1.SG eat.PFV coconuts ‘I ate coconuts’ b. e4 li2 Ja31 1.SG eat.IPFV coconuts ‘I eat coconuts’

Interestingly, in the neighbouring Mande language Gban, a similar grammatical distinction is realized on the tense auxiliaries which index the subject’s person and number (Zheltov 2005:24), which can be nicely captured with tone features: Present

(22)

SG

1st pers. 2nd pers. 3rd pers.

I2 ᷉

PL

u2 2 aa2 ɛɛ 1 ɛ ɔ1 [-raised]

Past SG I4 ᷉

PL

u4 [+upper] ɛɛ aa4 ɛ3 ɔ3 [-upper] [+raised] 4

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

217

In other languages tone has been found to mark person, tense, negation, transitivity on the verb, and singular-plural distinctions on the noun, for example, Noni [Bantoid; Cameroon] (Hyman 1981:10): (23)

bìe bì̄

‘ish’ ‘goat’

pl. pl.

bíe b̄e

(L vs. H) (LM vs. M)

In addition to the above morphological function, tone may interact with the syntax or information structure (see 9.14). The presence of tone does not preclude important interfaces with intonation (see Downing & Rialland 2016).

9.8

Noun Classes

One of the most notable features of NC languages as a whole is their systems of noun classiication. Some form of noun class system is found in the majority of NC languages, and noun classes have traditionally been cited as one of the deining features of the phylum, for example, ‘this feature was and is the best non-lexical diagnostic for genealogical classiication in the Niger-Congo domain since Westermann 1935’ (Güldemann 2011:131). In this section we will irst examine what are often thought of as ‘typical’ NC noun class systems, before briely surveying the rather impressive diversity of systems encountered within the phylum as a whole. Discussion of NC noun class systems inevitably begins with Narrow Bantu languages, being the best studied and most commonly cited. Proto-Bantu (cf. Meeussen 1967) made use of around 20 noun class prefixes obligatorily marked on the noun itself, a system kept generally intact in most of the descendant languages. Elements agreeing with the noun (adjectives, verbs, etc.) also take a class-appropriate prefix, the form of which is for some classes somewhat different from that seen on the noun itself. An illustrative example of class agreement from Herero [Namibia] (Möhlig & Kavari 2008:209) is seen in (24). (24)

oka-tí oka-t̪ít̪í k-á-kóyóka 12-stick 12-small 12.SUBJ-IMPERF-be.broken ‘the small stick is broken’

Most classes are grouped into a default singular/plural pair, as indicated in Table 9.11. However, individual nouns can make use of non-default singular/plural pairings, and some classes serve double duty as a both a singular/mass and plural class. For example, class 6 *ma- is both the plural of class 5 and the class for liquids, and class 14 *bʊ´- is used for abstract

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

218

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Table 9.11 Proto-Bantu noun class prefixes sg. classes

pl. classes

locative classes

1 3

*mʊ` *mʊ` -

2 4

*bà*m`ɪ-

16 17

*pà*kʊ `-

5

*ì-

6

*mà-

18

*mʊ` -

15 7 9 11 12 19

*kʊ` *kı``*N *dʊ` *kà*pì-

8 10 13 14

*bì`*N *tʊ` *bʊ` -

concepts (e.g., *bʊ -ntʊ ‘humanity’) in addition to being a diminutive plural class. Schadeberg (2011) cites this ‘conlation’ of class and number as a deining characteristic of NC languages more broadly. For several classes, one can identify a prototypical semantic ield that the class marks: classes 1/2 *mʊ -/bà- for people, 3/4 mʊ -/m`I for plants, 9/10 *N -/N - contains many animals, and so on. However, the semantic membership of most classes is not particularly strict, and often the class of a noun will differ even among very closely-related languages. A common feature of Bantu languages and other NC languages with similar class systems is the possibility of a single root appearing in multiple classes with different but related meaning. This is often a productive way of forming diminutives, augmentatives, and evaluatives of various sorts, for example, Lusoga ò-mú-tì ‘tree’ (cl. 3) → à-ká-tì ‘small tree’ (cl. 12), ò-gú-tì ‘big tree’ (cl. 20), è-cí-tì ‘short, fat tree’ (cl. 7), ò-lú-tì ‘tall, thin tree’, ́-í-tì ‘abnormal or defective tree (e.g., diseased, uglylooking)’ (cl. 5) (L. Hyman, personal notes). Outside of Bantu, we ind many examples of typologically similar noun class systems, not only in the Benue-Congo languages (those most closely and demonstrably related to Bantu), but also within much more distant families. The noun class systems of Joola Eegimaa and Temne (Atlantic) in Table 9.12 function in basically the same way as in Bantu, even though the class markers themselves are in most cases not obviously cognate with the Bantu markers. This typological proile of noun classiication – namely, a rather large inventory of classes marked by obligatory preixes on the noun and agreeing elements – is often taken as proto-typical for NC, and is assumed by many scholars (e.g., Schadeberg 2011) to have existed in much the same form in a putative proto-language. However, when taken as a whole, there is a rather impressive typological range of noun class systems within NC, many of which do not it this typological proile, especially outside of Benue-Congo.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

219

Table 9.12 Joola Eegimaa and Temne noun class prefixes Joola Eegimaa (Sagna 2010)

Temne (Schlenker 1864)

sg. classes

sg. classes

aefujubu-/bagañu-/ñalocative classes tdn- (temporal)

pl. classes

pl. classes

bug-

i-/a-

maŋ-

esugumuu-

taa-/ŋa-

maɛ-/yɛ-/yatrǝnapǝa-/ŋa-

collective classes bafamuma-

da-/rakǝo-/uɔ-/ulocative classes ro-/dora-/rǝ-

The position of the marker itself is subject to widespread geographical variation. While preixes are overall more common, noun class sufixes are found in many families including Gur, Kru, Bantoid and a number of languages within Atlantic such as Fula (Arnott 1970) and Kisi (Childs 1985). Languages that make use of both sufixes and preixes are common in Atlantic, including Mbulungish, Baga Mboteni, and the Bullom languages (Sapir 1971). In some languages like Wolof and the Cangin languages within Atlantic, noun classes are marked only on agreeing elements. In these languages, marking on the noun is either entirely absent, or only identiiable as an inert historical vestige, ‘fossilized’ on certain nouns. The number of classes varies widely across languages. Some have rather more than Bantu, like the Bainunk languages (Cobbinah 2013) and Kobiana (Voisin 2015) within Atlantic, each with over 30. But most have less, sometimes signiicantly so: nine in Kisi (Childs 1985), ive in Palor (D’Alton 1983). The conlation of number and noun class is not a feature of all NC languages. Often, as in many Atlantic languages, number is a grammatical category distinct from noun class. Furthermore, some languages do not possess a particularly large number of plural classes – Wolof has only two plural versus eight singular classes, and Fula has three for its ifteen singular classes (ignoring diminutives and augmentatives). The degree to which elements agree with the noun also differs greatly from language to language. Some languages like the Northern Atlantic group (including Wolof and Fula) show no verbal agreement with noun class. Agreement within the noun phrase is more common, but can be lacking entirely as in Dagbani (Gur) (Olawsky 2004), which distinguishes nine distinct sufixes only on nouns (and adjectives, which

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

220

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

are nouns). A number of languages show no noun class system at all, most notably in Mande and Ijoid, subgroups whose NC status has been questioned (Dimmendaal 2011:320, 323). Much of this typological diversity could be explained as the result of the erosion of an originally Bantu-like system, and this certainly did occur in a number of families (see Good 2012). However, especially at the highest levels of the phylum, it is far from certain that such a system was original, given the typological diversity of the modern class systems. One oft-proposed idea (e.g., Kießling 2013) is that some or all of these systems originated as a system of noun classiier or numeral classiier words, becoming grammaticalized in different ways in different families. This could explain not only some of the typological variation, notably preixes versus sufixes, but also the apparent noncognacy of many of the markers among more distantly related families. Nonetheless, some particular class markers have an extremely wide distribution, such as the personal plural *ɓa and liquid *ma, both of which are found across Atlantic as well as Bantu, and in many families in between. Needless to say, this heavily suggests the presence of some system of noun classiication at the earliest stages of NC. The relevant questions concerning PNC noun classes are the following: How many? How semantically coherent were they? How were they realized (preixes, sufixes)? How do they correspond in the daughter languages and groups? With respect to this last question, there continues to be interest in explaining how and why the Narrow Bantu nasal N (V)- preix classes correspond to non-nasal (C)V- in most non-Bantu NC languages. Even within Narrow Bantu, the agreement markers closely resemble the noun preix shapes (and tones) found on nouns outside of Narrow Bantu, for example, class 3 noun preix *mʊ  - versus agreement preix gʊ -, class 4 noun preix m`I - versus agreement preix g I´-, and so on. For overviews and different views, see Hyman (1981, 2017), Miehe (1991), and references cited therein. Surveys of noun class systems in different NC subgroups include de Wolf (1971) for Benue-Congo, Creissels and Pozdniakov (2015) for Atlantic, and Miehe and Winkelmann (2007, 2012) for Gur, many of whose noun classes are identiied – even by number – with those found in Bantu.

9.9 The Noun Phrase As outlined in Heine (1976), African languages vary considerably in the structure and order of elements within the noun phrase. In all NC languages a noun can be modiied or expanded by a possessive pronoun or full noun phrase, a demonstrative, and/or numeral. The word order can vary considerably. In some head-inal languages such as Kalabari [Ijoid; Nigeria] (Harry &

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

221

Hyman 2014), all modiiers precede the noun except for the plural marker, the deinite marker, certain higher numerals, and the quantiier ‘all’: (25)

a. PossN + N PossPro + N Determiner + N Numeral + N

tʊ` ɓɔ` námà `I ná↓má mɪ´ nàmá sɔ´ná nàmà

‘the child’s animal/meat’ ‘my animal/meat’ ‘this animal/meat’ ‘ive animals’

b. N + Plural N + deinite N + quantiier

námá ámɛ´ɛ` námá mɛ´ námá mámgbà

‘animals’ ‘the animal/meat’ ‘all the animals/meat’

At the other end of the spectrum – and much more common in NC – are languages where most or all modiiers follow the head noun. Such is the case in many Bantu languages such as Luganda (Ashton et al. 1954): (26)

a. N + PossN N + PossPro N + Demonstrative N + Numeral N + Adjective N + quantiier

ènnyámá yá ómwáàná ènnyàmá yààngé ènnyàmá ènó ènnyàmà ttáànó ènnyámá émbísì ènnyàmá yònnâ

‘the child’s meat’ ‘my meat’ ‘this meat’ ‘ive meats’ ‘raw meat’ ‘all the meat’

b. ‘every’ + N

bùlì n`nyámá

‘every meat’

While Bantu languages sometimes allow preposed modiiers, especially demonstratives (van de Velde 2005), possessive pronouns and ‘every’, for example, bùlì in Luganda, it is sometimes hard to tell if these are appositional (‘this one meat’, ‘mine meat’, etc.) or true modiiers (Rijkhoff 2002:272–276). The situation is much clearer in the obligatorily mixed noun phrase word order which obtains in a wide range of West African NC languages, particularly in Cameroonian Bantu (Mous 2005b) and the Mandesphere. In these languages possessives and at least some determiners precede the head noun, while other modiiers (e.g., numerals, adjectives) follow. This is illustrated below from Tommo So [Dogon; Mali] (McPherson 2014:58–59): (27)

a. gàmmà gɛ´m gàmmà nɔ´ gámmá tààndú

‘black cat’ ‘this cat’ ‘three cats’

b. gámmá m ´mɔ

‘my cat’

c.

‘my uncle’

mí bábé

d. íí=gɛ gàmmà íí=gɛ bàbè

‘the child’s cat’ ‘the child’s uncle’

The expected noun-modiier order is observed in (27a) involving an adjective, a demonstrative, and a numeral. The other examples are complicated

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

222

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

by the fact that Tommo So makes an alienable/inalienable distinction found especially in the Mandesphere. With an alienable noun, the possessive pronoun follows, as in (27b), while the possessive pronoun precedes an inalienable noun such as the kinship term in (27c). Possessive nominals precede both alienable and inalienable nouns, as in (27d). As (27b, c) also show, it is common for possessive pronouns to have different forms when modifying alienable versus inalienable nouns. Thus compare the following possessed ‘free’ versus ‘relational’ nouns from Liberian Kpelle [Mande; Liberia], slightly modiied from Welmers (1969:74): pɛ´rɛ´ ‘house’ (free)

(28) ‘my’ ‘your sg.’ ‘his/her’ ‘our’ ‘your pl.’ ‘their’

ŋá pɛ´rɛ´-í í pɛ´rɛ´-í ŋà pɛ´rɛ´-í kú pɛ´rɛ´-í ká pɛ´rɛ´-í `dí pɛ´rɛ´-í

pólù ‘back’ (relational) m ´-bólù í pólù `bólù kú pólù ká pólù `dí pólù

Not only are the pronouns different in the irst and third person singular forms, but only Welmers’s ‘free’ nouns can take the -í speciic sufix: ŋá pɛ´rɛ´-í ‘my house’, ŋá pɛ´rɛ´ ‘a house of mine’, m´-bólù (*-í) ‘my back’. The voicing of the /p/ of /pólù/ ‘back’ is due to a preceding nasal preix which drops out in the L tone third singular form. The nasality is however kept when the initial consonant is a sonorant, for example, l̄̄ ‘mother’, n̄̄ ‘his/her mother’. The Kpelle forms have a parallel in ka versus Ø marking in Bambara, a language of the Manding Group of Western Mande (examples from Welmers 1963:436, as modiied by Valentin Vydrin, personal communication): (29)

‘Free’ nouns ́ ká bón cɛ` ká bón à ká f ìni mùso ká f ìni

‘my house’ ‘the man’s house’ ‘his cloth’ ‘the woman’s cloth’

‘Relational’ nouns ́ bólo ‘my hand’ cɛ` bólo ‘the man’s hand’ à dén ‘his child’ mùso dén ‘the woman’s child’

While only a minority of NC languages make a free/relational or alienable/ inalienable distinction, many if not most show at least traces of so-called associative markers when connecting nominals. Welmers exempliies the pervasive Bantu -a with Swahili examples ki-pande ch-a nyama ‘piece of meat’, nyumba y-a udongo ‘house of stone’, ma-ji y-a chumwi ‘salt water’ (water+assoc.+salt), m-tu w-a Utete ‘a person from Utete’, which show that associative -a fulils typical genitive functions. It should be noted that certain NC languages have introduced nouns meaning ‘thing’ or ‘unit’ into associative constructions, sometimes to emphasize the possessor, for example, Igbo ìtè m´ ~ ìtè ŋ` kè m´ ‘my pot’ (cf. ŋ` kè m´ ‘mine’) (Emenanjo 1987:93). Only a small number of NC languages have been reported with numeral classiiers, the most extensively documented being Kana [CrossRiver; Nigeria] (Ikoro 1994).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

223

Niger-Congo Linguistic Features and Typology

While possessives, demonstratives, and numerals are universal, not every NC language has article-like determiners or an easily identiiable class of adjectives. Luganda stands at one end of the spectrum in both regards. First, there is the augment known also as the initial vowel or prepreix in Bantu studies, where it has various functions (de Blois 1970; Hyman & Katamba 1993): è-kí-tábó ‘a/the book’ versus kì-tábó ‘it’s a book’ (see also 9.14). While some NC languages have only a restrictive set of adjectives, for example, exactly eight in Igbo, falling into four groups of opposites (Welmers  & Welmers 1969:321): ó̙ má ‘good’, ó̙ jó̙ ↓ó̙ ‘bad’, ójí↓í ‘black, dark’, ó̙cá ‘white, light’, úkwú ‘large’, ́tà ‘small’, ó̙ hú̙↓rú̙ ‘new’, ócyè ‘old’, Luganda has a more extensive set of underived adjectives, including the eight in Igbo (Ashton et  al. 1954:161). In addition, Luganda productively derives adjectives from verbs: Many transitive verbs produce a past participle-like adjective by adding the inal vowel -e, while intransitives form a present participle-like adjective by adding -u (with labiodentalizing of the preceding obstruent): -fuk- ‘pour’ → -fúk-è ‘spilt, poured’; -guy‘deceive’ → -guy-ê ‘deceived, beguiled’; -búùb- ‘mould’→ -búùmb-è ‘moulded’ versus -jeem ‘rebel, revolt’ → -j́́m-ù ‘rebellious’; -tamiir- ‘get drunk’ → -támììv-ù; ‘drunken’; -kólòòngok- ‘be slender, thin’ → -kólòòngòf-ù ‘long and slender’ (Snoxall 1967). We have already mentioned possessive pronouns in the context of free versus relational nouns. While most NC languages have distinct forms for the six person-number combinations, a few have inclusive and exclusive irst person plural forms. Others may merge two of the plural pronouns in at least certain functions. As mapped out by Güldemann (2003a; 2008:154), there is a wide area in his Macro Sudan Belt where logophoric pronouns are found. Noni [Bantoid; Cameroon] not only has distinct singular and plural logophoric pronouns used in reported speech (‘the childi said that s/hei fell’), but also a contrast between singular non-logophoric co-referential versus non-co-referential pronouns (Hyman 1981:15, 20), see Table 9.13. Finally, a number of NC languages, especially Bantoid and Northwest Bantu, but also Southern and South-Western Mande (Vydrin 2010b), have what Cysouw (2009:182) terms ‘incorporative pronouns’. The most elaborate such system may be Medumba [Bantoid; Cameroon] (Déchaine  & Keupdijio 2016). In this language the oblique or independent pronouns are Table 9.13 Noni pronouns

subj/obj cl. 1 poss. cl. 2 poss. cl. 7 poss.

1sg

2sg

3sgj

3sgi

LogSg

1pl

2pl

3pl

LogPl

me¯ wɛ` m bɛ¯m ` kɛ¯m `

wɔ` wɔ` bo¯w ` ko¯w `

wvù wè bêw kêw

— — bêŋ kêŋ

we¯n we¯n bɔ̄-we¯n-ɛ´ ke¯-we¯n-ɛ´

bèsèn wèsèn bɔ` sɛ´sɛ` n kèsɛ´ sɛ` n

bèn wènè bɔ` nɛˆ n kènɛˆ n

bɔ´ (wù)bɔ` bɔ̄bɔ´ɔ´ lɛ` ke¯bɔ´ɔ´ lɛ`

bɔ` we¯n bɔ` we¯n bɔ̄ -bɔ` we¯n-ɛ´ ke¯-bɔ̀ we¯n-ɛ´

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

224

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Table 9.14 Independent incorporative pronouns in Medumba 2sg

2pl

3sg

1sg + mə` bə´ g -ù 1pl + (*bə´ g) bə´ g -ù 2sg + 2pl + 3sg + 3pl +

mə` - bə´ g -à-bín bə´ g- bə´ g -à-bín

mə´ bə´ g ùbínyí(*bú-)

3pl

bə´g

bə´ g bîn bîn bû bû

-yí -yí -yí -yí -yí -yí

mə` bə´ gùbínyíbú-

bə´ g bə´ g bín bín búbú-

-à-bú -à-bú -à-bú -à-bú -à-bú -à-bú

mə` ‘1sg’, ù ‘2sg’, yí ‘3sg’, bə´g ‘1pl’, bín ‘2pl’, and bú ‘3pl’. In the independent incorporative pronouns in Table 9.14, the above plural pronouns form the base that determines the overall (plural) person to which additional singular or plural component pronouns can be compounded, for example, me-we-you sg., you sg.-you pl.-them. The apparent uniqueness of Medumba is the tripartite structure of such pronouns. In neighbouring languages the irst column would not be present. A form such as bə´g-yí (we-him/her) is necessarily dual, i.e., ‘him/her and me’, while a form such as bə´g-à-bú (we-them) would be ambiguous between ‘them and us’, ‘them and me’, and ‘him/her and us’. In Medumba there are more possibilities to disambiguate the person and number of the parts, but obligatorily with the plural pronoun in the middle.

9.10 Verb Extensions Two properties of NC that are often considered together are noun classes and the derivational sufixes known as verb extensions (causatives, applicatives, etc.). This can be seen, for example, in Williamson and Blench’s (2000) tabular summary of each sub-branch. One difference, however, is that noun classes are almost strictly limited to NC, while extensions occur in all four of Greenberg’s (1963) African language phyla (Dimmendaal 2000a:187–188). Since such extensions are typically short, consisting of one or two segments, and often resemble each other even across unrelated language families, it is sometimes dificult to establish cognacy. Still, Table  9.15 presents an early attempt by Voeltz (1977) to propose reconstructions at the PNC stage. Even though Voeltz’s proposals are tentative, it is clear from Table 9.15 that the extensions shared by Proto-Bantu and distant Moore have to be reconstructed at a level that includes BenueCongo and Gur (see also Elders 2007b). Further efforts to reconstruct at the PNC level have been hampered by the fact that several of the subgroups do not have verb extensions, while others have extensions that may or may not be cognate to those found in Proto-Bantu. For example, the causative extension has [l],[ɗ] or [n] or [nd] in several North Atlantic languages,

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

225

Table 9.15 Possible Niger- Congo verb extensions

applicative causative contactive passive reciprocal reversive (tr.) reversive (intr.) stative/neuter stative/positional

Proto-Niger-Congo (Voeltz 1977)

Proto-Bantu (Meeussen 1967)

*-de *-ci, *-ti *-ta *-o *-na *-to *-ko *-ke *-ma

*-ɪd- [-ɪl-] *-ic-i- [-is-] *-at*-ʊ*-an*-ʊd- [-ʊl-] *-ʊk*-ɪk*-am-

Moore (Gur) (Canu 1976) -d ‘locatif’ -s ‘causatif’ -b ‘être dans un état’

-g ‘inversif’ -m ‘positionnel’

e.g., Sereer -and, Wolof -al (Becher 2000:31), while the more Bantu-looking Fula/Sereer -it- and Noon/Ndut -is- instead have reversive and intensive functions. While the Bantu applicative *-Id- has multiple functions marking benefactives, recipients, locatives, circumstantials (‘for some reason’) and sometimes instruments, the situation can be quite different in other sub-branches. Thus, Atlantic languages typically differentiate between benefactive/recipient and circumstantial extensions, for example, Ndut -iɗ- versus -aʔ- (Becher 2000:31), while Kordofanian languages distinguish benefactive/recipient and locative/applicative extensions, for example, Koalib – (V)ccÉ versus (V)t̪À, the latter having spatial, but especially malefactive readings (Quint 2010:297). Atlantic and Kordofanian languages are particularly rich in extensions, having also ventive (towards the speaker) and itive (away from the speaker) sufixes, often lacking in other subbranches. What this suggests is that the proto language might have had a quite different system from Proto-Bantu, perhaps with more distinctions rather than fewer. The extensions in Table 9.15 all have in common that they are valencerelated. Some license an additional argument or other constituent, for example, locative, circumstance, while others demote or otherwise remove an argument, such as the passive. The ‘stative’ or ‘neuter’ extension is better identiied as a ‘decausative’ (Creissels et  al. 2008:314n). NC languages that have a valence-increasing applicative extension that licenses a benefactive or recipient argument will typically allow unmarked double objects, as in the following Luganda examples (L. Hyman, personal notes): (30)

a. à-fúúmb-á 3SG-cook-INFL

ómúpúùngá rice

‘he is cooking rice’

ábáànà òmùpúùngá b. à-fúúmb-ír-á 3SG-cook-APPL-INFL children rice ‘he is cooking rice for the children’

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

226

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

The causative may also increase the valence: (31)

a. à-ly-á 3sg-eat-INFL

òmùpúùngá rice

‘he is eating rice’

b. à-lí-ìs-á ábáànà òmùpúùngá 3SG-eat-CAUS-INFL children    rice ‘he is feeding the children rice’

Such languages typically have only a few ditransitive verbs that can take two unmarked objects without an applicative sufix. This seems always to include the verb ‘give’. NC languages vary considerably in the status that they accord to verb extensions. At the one extreme, many NC languages allow for multiple ‘stacking’ of verb extensions: (32)

a. Sereer [Atlantic; Senegal] (J. Merrill, personal notes) a    up    -t    -ik   -t            -ir  -oox    -k   -a     apeel 3p.SM bury-REV-GOAL-INST.APPL-REC-RFL      -FUT-INFL shovels ‘they’ll go unbury each other with shovels’ b. Cicipu [Kainji; Nigeria)] (McGill 2009:209) zzá nnà ù- tób -ìl -ìs -ìs -u-wò -wò -nò = mu sháyì person REL 3SG-cool-PL-CAUS-CAUS-V -ANTICAUS-APPL-PERF =1SG tea ‘the person who has caused tea to become cooled down in a forceful and iterative fashion for me’ c.

Moro [Kordofanian; Sudan] (Rose 2013:49) owːa g  -ubəð-i   -tʃ -ən -ə´   -ŋó woman SM.CL -run -CAUS -APPL -PASS -PERF -3SG.OM ‘the woman was made to run away from him’

d. Kinande (Bantu) [Bantoid; DRC] (Philip Mutaka, pers. comm to Nurse  & Philippson 2003b:9) tu-né-mu-ndi-syá-tá-sya-ya -ba -king -ul -ir -an -is -i -á =ky-ô -them-close -REV-APPL-REC-CAUS-CAUS-INFL it we-TENSE/ASPECT COMPLEX ‘we will make it possible one more time for them to open it for each other’

The linear order of such stacked extensions has received considerable recent attention, particularly in Bantu (Hyman 2003b; Good 2005, 2007; McPherson & Paster 2009, among others). The results point to a conlict between ixed (‘templatic’) orders versus order determined by the relative scope of the different extensions  – or even phonological considerations (Hyman 2010). At the other extreme, a language may have only one or two extensions, which may not be productive. Often original valence-related verb extensions evolve into aspectual sufixes (Gerhardt 1988:5), for example,

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

227

marking various types of pluractionality, as in Babanki [Grassields Bantu; Cameroon] (Akumbu 2008; Harriet Jisa, personal notes): (33)

tsɔ´ʔɔ´

‘jump’ →

tsɔ´ʔ-mə´ tsɔ´ʔ-kə´ tsɔ´ʔ-lə´ tsɔ´ʔ-tə´

‘jump one after the other’ ‘jump time and again’ ‘jump across things’ ‘jump gently’ (= attenuative)

In NC languages which have only lexicalized extensions or none at all, the prior functions must be taken over by other means. In the absence of an applicative extension to mark a benefactive or recipient referent, NC languages either allow unmarked double objects, adpositions, or serial verbs, sometimes alternatively or combined in the same language (XP = the theme; YP = the beneiciary or recipient): (34)

a. Verb YP XP: Nzadi [Bantu; DRC] (Crane et al. 2011:155) bì ó súm mwàán òŋkàáŋ ‘we bought the child a book’ 1pl PST buy child book b. Verb XP to/for YP: Limbum [Grassields Bantu; Cameroon] (Fransen 1995:259) wìr bí fàʔ nì Tàr̄ ‘we will serve [work for] the Lord’ we FUT0 work for lord c.

Verb YP with XP: Koshin [Bantoid; Cameroon] (Ousmanou 2014:309) mə̄ fà wə bə´ ndì ‘I give you some water’ I give you with water

d. Verb XP give (to/for) YP: Mundani [Grassields Bantu; Cameroon] (Elizabeth Magba, personal communication) ‘she cooked food for him’ tà lè la̹a̹ èghidzi ŋa abua tò she PST3 cook food give to him

9.11

Verb Inflection

As in other areas of the grammar, there is considerable variation in verb inlection. The NC word structure presented in 9.3 suggests that the verbs should have both inlectional preixes and sufixes, the latter following any verb extensions that may be present. This structure is exuberantly realized in the Kinande example in (32d) above. While Kinande represents the extreme at one end of a continuum of inlectional marking, most other sub-branches are not this complete. On the one hand it is quite common for subject-verb agreement, tense, aspect, and perhaps negation to be either preixed to the verb or to combine to form a ‘tensed pronoun’ or subject-tense-aspect-mood-polarity ‘S/TAM/P morph’ (Anderson 2012). On the

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

228

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Table 9.16 Comparison of past tenses in two dialects of Totela

a.

Prehodiernal past

b.

Hodiernal past

c. d.

Hodiernal Posthodiernal

Perfective Imperfective Perfective Imperfective Future Future

Namibian Totela

Zambian Totela

na-SM-a-R-a ka-SM-R-a SM-a-R-a SM-la-R-i mo-SM-R-e ka-SM-R-e

SM-a-ka-R-a ka-SM-R-a SM-a-R-a SM-na-R-a SM-la-R-a na-SM-la-R-a

other hand, the most analytic NC languages have only limited such inlection, ultimately none at all. These may relect the same categories analytically, that is, as separate words. It is safe to say that most NC languages are rich in ‘auxiliaries’, whether monomorphemic or fused (Anderson 2011). One of the striking properties of tense-aspect-mood (TAM) marking is how much it can vary, even among closely related languages. This can be seen, irst, in a comparison of the agglutinative structures of two mutually intelligible forms of Totela [Bantu] (Crane 2011:110) in Table 9.16. As seen, not only can the TAM markers be different, but they also can differ in whether they precede or follow the subject marker (SM). The inlectional inal vowel sufix that follows the root (R) may also differ. Similar differences are noted between the following closely related Kegboid [Ogoni] Cross-River languages, which use different habitual auxiliaries (cf. Bond 2006:191–193): (35)

nɔm a. ò ʔórò ba-i 2 HAB eat-2PL meat ‘you pl. used to eat meat’

(Gokana)

b. ò-bere kɛ-a-i m ` bó slaughter-HAB-2PL goat 2-ANT ‘you pl. used to slaughter goats’

(Eleme)

The separation of person and number marking is also noteworthy: Without the -i sufix on the main verb, the subject would be second person singular. Table 9.17 summarizes the default subject afixes in Eleme (Bond 2010:4). Concerning the different TAMs that are marked, NC languages, especially Bantoid, are known for distinguishing multiple degrees of past tense, as in Totela (Table 9.16), but also of future tense, as in Yemba (Bamileke-Dschang) [Bantoid; Cameroon], which has ive degrees of future (Hyman 1980b:228): (36)

F1 : F2 : F3 : = F4 : F5 :

à↓á táŋ àà ↓pìŋ↓ŋ´ táŋ àà ↓lù↓ú táŋ ` ʔ↓é táŋ àà ↓̌u ↓ à á láʔé ↓táŋ à↓á fú↓táŋ

‘he is about to bargain’ ‘he will bargain (later today)’ ‘he will bargain (tomorrow)’ ‘he will bargain (after tomorrow, some days from now)’ ‘he will bargain (a long time, e.g., a year or more from now)’

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

229

Table 9.17 Default subject affixes in Eleme

1st person 2nd person 3rd person

singular

plural

m ` -, n`-, ŋ` -, ŋ` m `ò-, ɔ` è-, ɛ` -

rɛ¯-/nɛ¯ò-, ɔ` - . . . -i è-, ɛ` - . . . -ri

In addition to the tonal coniguration on the subject pronoun, the F2-F5 tenses are marked by an auxiliary, which except for F5 can be related to a Yemba main verb: pìŋ ‘return’, lù ‘get up’, ̌ùʔ ‘come’, láʔ ‘spend the night’. Thus, F2 derives from ‘return & bargain’, while the two variants of F3 derive from ‘come & bargain’ and ‘spend the night & bargain’. These markers can in fact be combined either as main verbs or indicators of ‘relative’ time: (37)

↓ àà lùù pìŋ↓ŋ´ táŋ F2 bargain ‘he will bargain later tomorrow’ 3SG.F1 F3 F3 return ‘he will return tomorrow and bargain’ get-up F2 ‘he is about to get up and later today bargain’ get-up return ‘he is about to get up and return and bargain’

The relative interpretations of F2 and F3 are thus ‘the same day’, and ‘the next day’. However, as pointed out for Chibemba, in many cases ‘the attitude of the speaker is the deciding factor, not the mechanical division of days’ (Sharman 1956:31). While languages like Totela and Bamileke-Dschang are very rich in tenses, other NC languages have been claimed not to have tense at all. One controversial case is Igbo, which has been claimed by several scholars only to have aspect (see Obiamalu 2015:49–55 for a recent perspective on the different views). There certainly are cases where tense is underdifferentiated, at least in certain aspects. Thus Ikoro (1996:171–172) interprets the Kana habitual marker we᷇̄ as ‘express[ing] customary or generic action which used to take place in the past, or still takes place at present, but not at the moment of speaking’. In other words, it expresses what is usually referred to as present AND past habitual, hence Ikoro’s translations: (38)

` -we᷇̄ dʒìge a. m 1SG-HAB snatch

‘I normally snatch (it)’

` -we᷇̄ fà b. m 1SG-HAB weed

‘I used to weed (it)’

A number of languages have a verb form identiied as ‘factitive’ (also ‘factative’) which can be interpreted differently according to the nature of the verb: With an active verb the time reference is usually past, while with a stative verb it is (or can be) present. The tradition in Igbo has been to view

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

230

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

the -rV sufix as marking such a factitive, as seen in the following examples (Emenanjo 1987:169): (39)

a. chí↓kéè Chike

sì̙-rì̙ tell-FACT

b. àdáà Ada

chò̙-rò̙ é↓gó ‘Ada wants money’ want-FACT money

àsí̙ lie

‘Chike told lies’

The controversy comes from whether one should analyse forms such as the above as having the same -rV sufix. While the Kana sentences in (38) show that certain aspects do not distinguish tense, the different time references of the Igbo -rV sufix(es) has indicated to most scholars that tense is subordinated to the Aktionsart of the lexical verb. While the above discussion has been centred around the inlectional features of person, number, tense, and aspect, NC languages also very often express negation within the inlectional morphology. Nowhere is this property more evident than in Bantu. As exempliied by the following examples from Chichewa, the Bantu preixal sequence lends itself to a ‘slot-iller’ interpretation (FV = inal inlectional vowel): (40)

a. main (root) clause: NEGATIVE-SUBJECT-TENSE-ASPECT-OBJECTsitidzá-ngo- mú- mény-á ‘we will not just hit him’ NEG-1PL.SUBJ-FUT-just-3SG.OBJ-hit-FV b. subordinate clause: SUBJECT-NEGATIVE-TENSE-ASPECT-OBJECTti- sa- dzá-ngo- mú- mény-á 1PL.SUBJ-NEG-FUT-just-3SG.OBJ-hit-FV

While the negative preix precedes the subject preix in main clauses, it follows it in relative, adverbial and other subordinate clauses, and occurs with the same form in the negative ininitive (ku-ḿny-a ‘to hit’, ku-saḿny-a ‘to not hit’) (see Kamba Muzenga 1981 and Maho 2007 for further elaboration of the template). While there is considerable variation across Bantu languages, some of which adopt other strategies (Devos & van der Auwera 2013; Gibson & Guérois 2016), Güldemann (1999) argues that the order in (40b) predates that in (40a). Further evidence of the thorough integration of negation into the inlectional morphology is that the same tense or aspect can be marked quite differently in the afirmative versus negative, as seen in the following near future (F1) forms from Lusoga [Uganda] (L. Hyman, personal notes): (41)

a. tùnà-à-gézémúl-á 1PL.SUBJ-F1-F1-perspire-FV b. tí-

tú- ú-gézèmùl-é

‘we will perspire’

‘we will not perspire’

NEG-1PL.SUBJ-F1-perspire-FV

It should irst be pointed out that the -a- and -u- morphs are realizations of the underlying F1 empty mora assimilated to the preceding vowel. There

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

231

are three more important differences to note: (a) the F1 preix -nà- is present only in the afirmative; (b) the FV is -a in the afirmative but -e in the negative; (c) the tones that are assigned to the verb are different: in the afirmative the stem is all H tone (underlying toneless, receiving its tones from a inal H% boundary tone); in the negative a H is placed on the irst mora which is followed by Ls (the inal H is due to a H% boundary tone). With respect to this last point, it is extremely common throughout NC for the afirmative and negative forms of the same TAM to have different tones. Despite the widespread integration of negation into the inlectional system, many NC languages instead (or also) realize negation through syntactically independent markers. In Ubangi and Adamawa languages the negative marker is generally clause-inal (R. Boyd 1989:208), as in Mundang (Adamawa). (42)

a. mè 1SG

kó-k̄ see-DET

b. mè 1SG

kó-k̄ see-DET

c.

ɓè

‘I have seen him’

(Elders 2000:441)

PERF

ȳ

‘I have not seen him’

NEG

mò ràk ȳ n̄ 2SG eat NEG Q

‘Have you not eaten?’ (Elders 2000:491)

The last example shows that the question marker n̄ follows the negative marker ȳ.

9.12

Sentence Structure

Although they do not utilize case marking on nouns, NC languages make a clear distinction between subject and object through word order and often pronoun forms and subject-verb agreement. Converging with a general tendency throughout the continent (Heine 1976; Creissels et al. 2008), the basic sentence word order is SVO in NC. The most notable deviations from this pattern are the head-inal patterns found in two areas: (a) in Mande and neighbouring languages in the Mandesphere; (b) in Ijo and certain Cross-River, Bantoid, and a few NW Bantu languages, for example, Tunen (Mous 2003). Particularly striking is the S-Aux-OVX order that occurs in an area that Güldemann (2008:151) calls the Macro-Sudan Belt (cf. Creissels 2005). This pattern is signiicant for its mixing of head initial (VO) and head inal (OV) properties. While the surface order S-Aux-OV(X) is sporadically widespread within this zone, its deeper syntactic properties are heterogeneous, suggesting distinct subtypes of the pattern (Sande et  al., forthcoming). In one type represented in the more westerly Mande area, S-Aux-OV(X) is a general organizing principle of the syntax, while in many others it limited to speciic syntactic environments which may differ from language to language.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

232

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

In Kru languages, if an auxiliary is present, it immediately follows the subject, and the primary object precedes the verb. In the Kru language Guébie (Ivory Coast), the word order is S-Aux-OV in most environments, including most main clauses and subordinate clauses. As seen in (43) and (44), the verb follows the object within a verb phrase (Sande 2017). (43)

e4 ji3 Ja31 li3 1SG FUT coconut eat ‘I will eat a coconut’

(44)

gba1 e4 ka3 tɛlɛ3.3 kɔklalɛ3.2.2 e4 wa2 1SG want.IPFV that 1SG IRR snake touch ‘I want to touch the snake’

Guébie also shows numerous other head-inal properties, such as postpositions, Genitive-Noun word order, and manner adverbs, which precede verbs within verb phrases. When there is no overt segmental auxiliary, however, SVO order emerges, as in (45). This occurs in present imperfective and basic perfective contexts. (45)

e4 li3 ɟa31 1SG eat coconut ‘I ate a coconut’

According to the clause structure proposed by Sande (2017:94–102) and Sande et al. (forthcoming), the VP is head-inal, but the tense phrase (TP) is head-initial, realized by the future auxiliary in (46a). When no auxiliary is present in T, the verb moves to that position, resulting in the SVO order in (46b) (cf. Koopman 1984): (46)

TP

TP

e4 I

e4 I

T ji3 will

VP DP

T′

DP

T′

DP

V+T

VP

li3 eat.PFV

V

DP

V

li3

(a)

a31 a coconut

a31

eat

a coconut (b)

Guébie is representative of Kru languages in general in that it alternates between S-Aux-OV and SVO word order. However, Kru languages vary in whether anything can follow a clause-inal verb. For example, in Guébie only heavy complement clauses follow the verb, while in certain Western Kru languages like Grebo (Innes 1966) complement clauses as well as some postpositional phrases follow the verb when an auxiliary is present. Either

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

233

way, S-Aux-OV seems to be the default word order across Kru, where SVO is a derived structure. Mande languages are like Kru languages in that S-Aux-OV word orders are pervasive. This basic word order is illustrated below for Marka Daing (Burkina Faso), a close relative of Bambara in the Manding group, for the past and future tense auxiliaries (P. Jenks, personal notes): (47)

a. wúrú-↓ú ↓ní ʃwó-↓ó ɲìmì

dog-DEF PST meat-DEF eat ‘The dog ate the meat’ b. wúrú-↓ú ↓ná ʃwó-↓ó ɲìmì

dog-DEF FUT meat-DEF eat ‘The dog will eat the meat’ As in Kru, S-Aux-OV is a general organizing principle of Mande languages, which can be modelled by adopting the same syntactic analysis in (46). Also like Kru languages, Mande languages are primarily postpositional and have genitive noun order in noun phrases. However, there are two important syntactic differences between Kru and Mande. First, when there is no overt segmental auxiliary, such as the present habitual, SOV order is found in Mande: (48)

wúrú-↓ú ʃwó-↓ó ɲìmì dog-DEF meat-DEF eat ‘The dog eats the meat’

This fact illustrates the general observation that Mande languages typically have less V-to-T verb movement than their Kru counterparts, although it is attested. Because of this, SVO is not an alternate word order in Mande languages. The second observation is that other than nominal primary objects, phrasal modiiers and verbal arguments in Mande follow the verb rather than preceding it. This means that secondary objects (which are postpositional phrases), embedded clauses besides relative clauses, and adverbial modiiers all occur after the verb. Thus, Mande has in a sense a canonical S-Aux-OVX pattern, with particular emphasis on X. It is an interesting question how to model these word order facts, whether these modiiers and arguments are all rightwards adjuncts of VP or a higher projection, as suggested by Nikitina (2009, 2011), or if there are some right-branching structures in the VP domain. While Guebie and Daing are good representatives of the most common syntactic patterns in their respective families, both families are more diverse, and contain languages which are somewhere in between the two patterns described above. For example, some languages might allow some but not all complements to precede the V, or allow certain nominalized clauses to precede the V in addition to noun phrases themselves.  These

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

234

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

patterns are not limited to Kru and Mande languages. For example, Creissels (2005) notes that nearby Senufo and Songhay languages share the general S-Aux-OVX pattern characteristic of Mande. Outside of Mande and Kru, the distribution of S-Aux-OV(X) is typically restricted to speciic syntactic environments. For example, in Gwari [Nupoid; Nigeria], which contrasts SVO and S-Aux-OV word order, the latter is found only with the completive aspect auxiliaries lá/kú, deriving from the singular/plural forms of the main verb ‘take’: (49)

wo kú à-shnamá 3SG take:PL PL-yam ‘s/he takes yams’

(SVO)

(50)

wo kú à-shnamá si buy 3SG CPL:PL PL-yams ‘s/he has bought yams’

(S-Aux-OV)

(Hyman & Magaji 1971:56)

Unlike the general S-Aux-OV order in Mande and Kru, other auxiliaries like the today past continuous auxiliary ɓ́̄ require S-Aux-VO word order in Gwari: (51)

wo ɓé̄ si à-shnamá 3SG T.PST buy PL-yam ‘s/he was buying yams’

(S-Aux-VO) (H&M 1970:54)

This is a crucial difference, because the S-Aux-OV order is not a general characteristic of all clauses with auxiliaries. In fact, the two types of auxiliaries can co-occur: (52)

wò ɓei kú à-shnamá si (S-Aux-Aux-OV) 3SG BY.PST CPL:PL PL-yams buy ‘s/he has bought yams’ [before yesterday]

(H&M 1970:57)

This indicates that OV order in Gwari is tied to the presence of kú in (50) and (52) and not just a general requirement that OV word order follows auxiliaries. Sande et  al. (forthcoming) argues that the Gwari VP is head initial, and that OV is derived. We ind a similar situation in Fongbe [Kwa; Benin], with a limited number of environments showing OV order. The dominant word order is SVO and there is a class of auxiliaries that requires S-Aux-VO order (Lefebvre & Brousseau 2002): (53)

Kɔ` kú xò àsíbá Koku hit Asiba ‘Koku hit Asiba’

(54)

Lili nɔ` ɗù gba`ɖé (S-Aux-VO) Lili HAB eat corn ‘Lili (habitually) eats corn’

(SVO)

(p.247)

(p.94)

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

235

However, certain aspectual verbs are followed by OV word order: (55)

àsíbá ɖò [[ ví ɔ kpɔ ́n ]] wɛ` ] Asiba be.at child DEF look.at.NZ POST ‘Asiba is looking at the child’

(56)

ùn jɛ` [[ nú ɖù ] jí ] fall thing eat.NZ on 1SG ‘I began to eat’

(p.215)

(p.215)

Following Lefebvre and Brousseau (2002), Sande et al. argue that the OV constituent is a nominalized VP (cf. Khachaturyan & Nikitina 2016 for Mande). The OV word order is expected on this analysis because noun complements precede their selecting head. Apparent ‘Aux-OV’ word order is therefore epiphenomenal. It is not a deep property of the grammar of Fongbe. Marginal S-Aux-OV(X) is found elsewhere in NC as well. Childs (2005) argues that several Atlantic languages show an S-Aux-OV(X) pattern. However, it appears that these patterns are similar to the Gwari and Fongbe type in that they are limited to certain types of constructions (cf. Creissels 2005). For example, in Noon, Childs shows that S-Aux-OV word order is only found when the object is a pronoun. The same is true of Kana [Cross-River]: (57)

maa zìb légbò 1SG.PROG hit Legbo ‘I am hitting Legbo’

(58)

maa yé zìb 1SG.PROG 3SG hit ‘I am hitting him’

(SVO)

(Ikoro 1996:226)

(SOV)

This suggests that, as in Gwari, OV is derived by object shift. It is well known that in many languages, object shift applies differentially to only a subclass of nominals, with pronouns being the most likely to shift crosslinguistically. In addition to the ive case studies discussed above, Sande et al. (forthcoming) surveyed 55 languages spoken throughout the Macro-Sudan belt, based on the diversity sample used by Clements and Rialland (2008). The results of the survey show that there is a strong cluster of head-inal properties and S-Aux-OV(X) order in the Mande area – the Mandesphere – encompassing Ivory Coast, Liberia, Sierra Leone, Guinea, Mali, and Burkina Faso. Outside of this cluster, some NC languages such as Gwari and Kana show S-Aux-OV(X) order in limited syntactic environments. While S-Aux-OV(X) seems to be prevalent in a number of NC subgroups throughout West Africa, it does not span the entire Macro-Sudan Belt, nor would we want to reconstruct it in PNC (cf. Hyman 2011b). Although S-Aux-OV was likely present in Proto-Mande and some neighbouring families (Creissels 2005), there appear to be three diachronic sources of S-Aux-OV (cf. Claudi 1993, 1994): (a) from a SVOV serial verb construction, as in Gwari (cf. Heine &

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

236

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Reh 1984); (b) from a transitive SV-OV, where the OV is a nominalized VP, as in Fongbe; (c) from S-Aux-VO with fronting of the O for defocusing, as in Aghem (cf. Güldemann 2007): (59)

fíl á mɔ` zɪ´ kɪ´bɛ´ án ↓sóm friend 3PL Y.PST eat fufu in farm ‘the friends ate fufu in the farm’ (S-Aux-VOX) (yesterday past) ‘the friends ate fufu in the farm’ (Watters 1979:147)

(60)

á mɔ` bɛ´ ↓kɪ´ zɪ´ án ↓sóm fíl friend 3PL Y.PST fufu DET eat in farm ‘the friends ate fufu in the FARM’ (S-Aux-OVX) (not somewhere else) ‘the friends ate fufu in the farm’ (Watters 1979:148)

As others have noted (e.g., Dimmendaal 2008b:306), some S-Aux-OV languages, including Aghem, allow more than the direct object to precede the verb, which is likely more prevalent where preposing is associated with defocusing. Also of note is that preposing may also be associated with negation, perhaps exclusively so, as in Leggbo [Cross-River; Nigeria] (Good 2003; Hyman 2003a).

9.13 Serial Verbs Another widespread property of NC languages are serial verb constructions (SVCs). Largely restricted to the middle part of area, that is, Ghana to Cameroon, SVCs can be defined as a series of verbs within a single clause which do not have a marker of morphosyntactic dependency, show argument sharing, and whose verbs can function as independent verbs outside of an SVC (Ameka 2001). An example is provided from Isu [Bantoid; Cameroon] (Kießling 2011:44) involving four verbs in series. (61)

↓ ú tú ɲi tswí dzɔ´ŋi tsə´ sɔ`ʔɔ` 3SG. P3 escape enter descend return IMM also V2 V3 V4 V1 ↓ á↓nə´ tsəˇm k-ə´ to 7.deep.pool 7-D1 ‘She also dived back into the water of the deep’

á↓nə´ to

↓

mwí 6a.water

SVCs are common throughout the Kwa and non-Bantu Benue-Congo families, and also found in Gur, for example, Dàgáárè (Hiraiwa  & Bodomo 2008) and Ijoid, for example, Izon (Williamson 1965; Carstens 2002). Contrasting with these are the highly synthetic languages without SVCs found in Atlantic, Kordofanian, Bantu, and certain Benue-Congo branches (e.g., Kainji) (cf. (32) above). Foley and Olson (1985:50–51) offer a list of shared phonological and grammatical properties among languages with certain types of SVCs (e.g., valence-increasing), particularly in West

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

237

Africa and Southeast Asia, which include (a) phonemic tone, (b) having many monosyllabic words, (c) isolating morphology, and (d) verb medial word order (SVO). They view this cluster as interrelated, where ‘phonological attrition causes syncope of segments or syllables’, which leads to ‘the gradual loss of the bound morphemes . . . [and] a new device for valence adjustment must be found’ (Foley  & Olson 1985:51). In many languages, SVCs ill this functional gap, provided that SVCs already exist in some capacity in the language, for example, as simple motional/directional SVCs. There thus is at least a diachronic relation between phonological development, the loss of head-marking morphology, and the extended development of serial verbs. Even so, NC systems which display increased analyticity outside of this Kwa/Benue-Congo zone do not necessarily comply with other aspects of this typology. For example, SVC systems are not found in the Bantu language Nzadi which has lost much of Proto-Bantu morphology (Crane et  al. 2011:4), nor in the more analytic Kru and Mande families (Marchese 1989:135; Dwyer 1989:57). To some extent, the areal distribution of SVCs within NC will depend on the criteria used to deine them (cf. Hopper 2008 for discussion of the ‘take & verb’ construction as an incipient SVC in English). Serial verb constructions are used to express a wide range of meanings. In Degema [Edoid; Nigeria], SVCs are used to express exhaustion/ completion of a situation, directionals, benefactives, verbal comparison, comitatives, instrumentals, accompanimentals, refusal, simultaneousness, abilitatives, consequentials, and event coordination (Kari 2003). Examples are below of benefactive expressed with ke ‘give’ in second verb position (V2) and verbal comparison expressed with tul ‘reach’ in V2 (Kari 2004): (62)

(i) imó nú. Ohoso o=méné ké=n what that Ohoso 3SG=do give=FAC ‘What was it that Ohoso did for him?’ (ii) O=sóm=n ú̩s̄ o=túl beauty 3SG=reach 3SG=be.good=FAC ‘He is as handsome as you.’ (Kari 2004:157)

ó.yi? him

(p.19)

wó̩=̄n you=FAC

(p.157)

A common type of SVC involves the verb ‘take’ or ‘carry’, typically in V1 position. These may have metaphorical uses in SVCs, for example, in the Esan case in (63a) below, where it combines with èghò·nghò·n ‘happiness’ to mean ‘with happiness’ or ‘happily’. Further, in (63b) it is used to introduce an object. This is remarkable as the Esan verb wo·n ‘to drink’ is a transitive verb already and may readily occur with the object directly. (63)

(a) ̀lékà riè èghó.nghó.n gbè Eleka take happiness\IPFV dance ‘Eleka is dancing with happiness’

ìkhié.n dance\IPFV

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

238

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

riè àyó.n (b) ̀lékà ghâ Eleka PreV take alcohol\IPFV ‘Eleka is getting drunk’

wó.n drink

Such data as in (63b) illustrate that SVCs in NC, especially with ‘take’, may originate as offering something different from the constructions with which they compete and may ultimately replace (cf. Lord 1993; Hopper 2008:271–274). For example, consider the function of ‘take’ as a ‘custody transfer’ verb in Mungbam [Bantoid; Cameroon]. Lovegren (2013:222) describes the use of the coverb mə¯ ‘take’ with the verb m̄ ‘drink’ as indicating that the drinking cup is not in the addressee’s hand, and that ‘the only situation where a simple imperative mu᷆ ‘drink!’ is felicitous is a case where the addressee is already holding a drinking cup’. (64)

jə´ à mu᷆ ‘take and drink!’ a. mə¯ take.IRR ascend.IRR 2s.TOP drink.IRR [cup is on the loor and has to be ‘ascended’ by addressee] b. mə¯ mu ‘take and drink!’ take.IRR drink.IRR [cup is within reach and at the level of the addressee’s hands, in front of him] c.

mu᷆ ‘drink!’ drink.IRR [addressee is already holding a drinking cup]

These above usages illustrate the development of SVCs to express functional relations, such as valency-changing, which in other languages may be expressed through verbal morphology and adpositions (recall the NC verb extensions in Table 9.15 above). Frequently, SVCs take on tense/aspect meanings as well, for example, grammaticalization of the verb ‘inish’ in V2 position to indicate completion or exhaustion. Grammaticalization paths are often language-speciic, for example, in Isu the verb fyí ‘exit’ extends to mean ‘do intensively, excessively’ in an SVC, and the verb diáŋí ‘pass through’ extends to ‘immediately, without delay’ (Kießling 2011:144, 167). Further, the existence of two verbs within a clause presents a number of logical possibilities for where the locus of inf lectional marking should be: on V1, V2, or both. Numerous languages within NC show complex patterns, the most interesting of which involve cases where inf lection is split across the verbs, or found repeating fully or partially on both. Examples are seen in (65a) from Ogbronuagum [Cross-River; Nigeria] where the aspectual marker mə- ~ ma- is found on both verbs in the SVC (Kari 2000:55) and (65b) from Baoule [Kwa; Ivory Coast],

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

239

where the perfective suffix -lì is on both verbs (Creissels  & Kouadio 2010:176): (65)

a. eédi tɛ´ɓɛ´n´I a mʊ´ lórú ꜜɛ´ɛ´má ɪ´nam aamɪ´ mə ´ -ɗeeri 2SG that in town animals 1SG ASP-know that tell má-tɛ a nɛ´ɛ´ma ASP-surpass you something ‘That will show you that I’m the wisest in the animal kingdom’ b. kuàkú fà-lì mín klè-lì show-PFV Kouakou take-PFV 1SG ‘Kouakou showed me to to Akissi’

Ákísí Akissi

Such inf lectional concord is found even in cases where the two verbs do not refer to the same subject, for example, in the example from Akan [Kwa; Ghana] in (66) where the verb gu ‘f low’ is marked with f irst person singular agreement, even though its semantic argument is the object aburow ‘corn’ (Schachter 1974:258, tones not marked): (66)

me-de aburow mi-gu msum 1SG-take corn 1SG-low water.in ‘I pour corn into water’

Finally, we understand verb compounding to be a special type of serialization in which the two verbs are obligatorily adjacent, often forming a single ‘word’. Verb compounding is not common in NC, but it is represented in a number of languages. A famous case is in Igbo, such as in (67) below where the verbs t ụ́ ‘throw’ and fù ‘be lost’ form a single compound marked by a single tense inf lection (Lord 1975:25): (67)

↓ ó. tụ́-fù-rù 3SG throw-be.lost-TNS ‘He threw paper away’

ákwụ́ kwó. paper

Verb compounds are also found in Isu (Kießling 2011), and to a limited extent in Degema with the verb ki· ye ‘give’ (Kari 2003, 2004).

9.14

Information Structure

As throughout the continent, NC languages are very rich in marking information structure, particularly focus distinctions. These can be realized through syntactic structures (word order variations, clefts, etc.), prosody (particularly tone), or most signiicantly, morphology. In many NC languages focus has a major effect on the grammar. This can be seen in

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

240

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

the following sentences from Aghem [Grassields Bantu; Cameroon] from Watters (1979:146–147): (68)

a. fíl á mɔ` zí kí-bɛ´ á↓zɔ´ɔ´ friends SM P2 eat 7-fufu yest.

‘the friends ate fufu yesterday’

b. fíl á mɔ` zí á↓zɔ´ɔ´ bɛ´ ↓kɔ´ ‘the friends ate fufu YESTERDAY’ friends SM P2 eat yest. fufu DET c. fíl á mɔ` á↓zɔ´ɔ´ zí friends SM P2 yest. eat

bɛ´ ↓kɔ´ ‘the friends ate FUFU yesterday’ fufu DET

d. à mɔ` zí á-fín bɛ´ ↓kɔ´ á↓zɔ´ɔ´ ‘the FRIENDS ate fufu yesterday’ ES P2 eat 2-friends fufu DET yest. e. fíl á mɔ` zí nô bɛ´ ↓kɔ´ á↓zɔ´ɔ´ ‘the friends ATE fufu yesterday’ friends SM P2 eat FM fufu DET yest. f.

fíl á máà zí bɛ´ ↓kɔ´ á↓zɔ´ɔ´ ‘the friends DID eat fufu yesterday’ friends SM P2.FOC eat fufu DET yest.

The even focus with neutral word order is shown in (68a), where the object kí-bɛ´ ‘fufu’ consists of a class 7 preix kí- plus the stem bɛ´ ‘fufu’ and appears in the immediate after verb (IAV) position. (68b) shows that the IAV position can be exploited to focus constituents other than the direct object, in this case the temporal adverb á↓zɔ´ɔ´ ‘yesterday’. Note also that the bare object must occur with a determiner, in this case semantically empty class 7 kɔ´ (before which the noun undergoes preix deletion). In (68c) á↓zɔ´ɔ´ is defocused into the immediate before verb (IBV) position thereby stranding ‘fufu’ in the contrastive IAV position. (68d) shows that when the subject is focused, it too is realized in IAV position, leaving the expletive subject marker à in its normal, preverbal place. In (68e) the lexical meaning of the verb is contrasted by placing the focus marker (FM) /nô/ after it. Finally, in (68f) truth-value focus is achieved by the substitution of the focused tense marker máà which also must be used if the verb is inal in an afirmative main clause (Watters 1979:151): (69)

a. bvú dogs b. *bvú dogs c.

bvú dogs

ñíŋ run

‘the dogs ran/did run’

ti mɔ` P2

ñíŋ run

‘the dogs ran’

ti

ñíŋ run

ti

↓

SM

máà P2.FOC

↓

SM ↓

SM

mɔ P2

nô ‘the dogs RAN’ FM

Although unmarked word order appears to hold in (68f), the form of ‘fufu’ is still bɛ´ ↓kɔ´, indicating that ‘fufu’ is not in the focused IAV position at a deeper level of analysis.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

241

Table 9.18 Tense marking in Efik [−focus]

[+focus]

Past Present

-kVØ

Future

dì

-ma -mVkɛ´-

[−progressive] [+progressive]

-yɛ´-`

The interaction between tense marking and focus is quite widespread within NC. Table  9.18 based on Welmers (1968) and Cook (1976) shows the different markers of the three tenses in Eik [Cross-River; Nigeria]. As implied by the term, the focused forms of a tense indicate that an inlectional feature or truth value is being focused, while such focus is lacking in the non-focused forms. Another property, however, concerns the tightness of the link to the constituent that follows the verb, for example, the object. With non-focused tense marking, the verb typically forms a tight bond with what follows, which may be required, as in the case of Aghem mɔ` , while with a focused tense marker there often is a prosodic break after the verb. The ability to appear at a constituent end correlates with the ability of a [+focus] marked verb to appear inally within an afirmative main clause, as with Aghem máà. For this reason the comparable distinction in Kirundi was termed ‘conjoint’ versus ‘disjoint’ by Meeussen (1959), terms which have been found useful in describing analogous distinctions in numerous other Bantu languages (see Van der Wal 2017 and other chapters in the same volume). As in Eik, the markers can be quite different segmentally and tonally. In Simakonde, the conjoint [−focus] past tenses are marked with an -ile sufix, while the corresponding disjoint tenses are marked with an -ndi- preix and an -a sufix (Manus 2017:242). Such contrasting [±focus] variants are typically restricted to afirmative main clause present and past tenses. Relative and adverbial clauses typically take the [−focus] tense forms (as these are backgrounded clauses), as do negative verb forms. However, both negatives and imperatives can otherwise act as if they are [+focus], for example, in Aghem and Kru (Marchese 1983b). In (69a) we saw that the [+focus] sequence máà ñíŋ can appear inally, while the [−focus] sequence mɔ` ñíŋ cannot. While the [±focus] distinction is neutralized in the corresponding negative, the latter can appear inally, suggesting that it is inherently [+focus], therefore not needing any constituent to ill the IAV position: bvu´ ↓tí káá ñíŋ ‘the dogs didn’t run’. Consider also the contrast between the following afirmative and negative today past tense verb forms in Haya [Bantu; Tanzania] (Hyman & Watters 1984:260): (70)

a. afirmative

bá-á-kôm-à

‘they tied’

b.

bà-à-kòm-à kátò

‘they tied Kato’

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

242

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

c. d.

negative

tì-bá-á-kòm-à

‘they didn’t tie’

tì-bá-á-kòm-à kátò

‘they didn’t tie Kato’

Both (70a) and (70c) show that the afirmative and negative verbs both have H tones when occurring inally in their clause. When an object follows in (70b), the afirmative verb loses its H tones by virtue of being out of focus, while the negative verb is not affected in (70d), again revealing its inherent focus. For more on what he termed ‘predication focus’, see Güldemann (2003b). Besides morphological marking and tonal effects, focus constructions vary considerably across NC languages. Besides exploiting the IAV position, even closely related languages can differ in placing the focused constituent at the left versus right edge of the utterance: (71)

Nupe sisi    Musa si eyì   o (cf. Kandybowicz 2008:105) buy.RED Musa buy  corn   FOC ‘Musa BOUGHT corn’ (he didn’t steal it) Gwari Musa si yaywì sisi nu᷆ (cf. Hyman & Magaji 1971:122) Musa buy corn  buy.RED FOC ‘Musa BUYS corn’ (he doesn’t steal it)

The above sentences from Nupe and Gwari [Nupoid; Nigeria] also exemplify a common repetition strategy for focusing on the lexical meaning of the verb, here in reduplicated form. There has been considerable recent interest in focus in NC languages too numerous to mention here, but see for example Fiedler and Schwarz (2010) and several articles on Ghana-Togo-Mountain languages in the Journal of West African Languages 36 (2009). There is of course much more going on than focus in the packaging of information in NC languages. This includes attitudinal or evidential marking, as in the case of the Luganda complementizer ntì, which introduces reported speech versus mbù, which ‘introduc[es] hearsay information’ (Snoxall 1967:203). A particularly interesting set of clause-inal ‘marqueurs de proposition’ occurs in a number of Adamawa and Ubangi languages (Bohnhoff & Boyd 2003:3). These markers fuse mood, aspect and polarity, taking different shapes, depending on whether the utterance is perfective/ imperfective, indicative/ ‘injunctive’, and afirmative/negative. The following examples illustrate two injunctive markers occurring at the end of an afirmative versus negative imperative in Dii [Adamawa; Cameroon] (Bohnhoff & Boyd 2003): (72)

` hò̝ lig míí a. à̝m 2SG.INJ look.at house.DEF my ‘look at my house!’

nu

(p.9)

INJ.AFF

b. à̝m ` làà kaa-lí sá 2SG.INJ go village.DEF-LOC INJ.NEG ‘don’t go to the village’

(p.8)

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

243

Numerous different allomorphs are used not only according to the features with which they fuse, but also depending on the word class of the constituent that immediately precedes the marker, for example, verb, noun, possessive pronoun. The last issue to be discussed is how questions are expressed in NC. WH elements generally follow the same patterns as in focused constituent structure. Thus, many languages front the WH word or place it in IAV position, often as an enclitic in Bantu. Clements and Rialland (2008:74ff.) establish a number of patterns by which yes-no questions are expressed. Many West African languages express yes-no questions through what Rialland (2007, 2009) calls a inal ‘lax question prosody’. While many NC languages have question particles or end questions in a H pitch, a inal L tone is quite widespread in West Africa, which may in turn be accompanied by lengthening and/or breathiness (Clements & Rialland 2008:79). This is illustrated in Adele [Kwa; Togo, Ghana] (Cahill 2015b:31), where both length and pitch lowering mark yes-no questions: (73)

a. ɛ` ɛ´ ɛ` ɛ´

kòrò ɛ` kpɛ` ɛ´ ‘s/he ate dog’ kòrò ɛ` kpɛ` ɛ` ɛ` ‘did s/he eat dog?’

b. è è

dé dé

fótù fótùù `

‘s/he is holding camera, photo’ ‘is s/he holding camera, photo?’

In (73a) the long [ɛ` ɛ´] rising tone acquires a L pitch with even longer duration, while the inal L of fótù ‘camera, photo’ becomes even lower [`ù], also with lengthening. Finally, although falling outside the heading of information structure, NC languages are rich in packaging information in terms of expressive particles and interjections and often have phonologically and semantically defined sub-systems of ideophones (Samarin 1965, 1971). This is illustrated from Yoruba [Defoid; Nigeria] in (74) (Courtenay 1976:19–20): (74)

a. wɔ´kɔ̄wɔ` kɔ̄ pɛ´tɛ¯pɛ` tɛ¯ ̌úk̄̌ùk̄

‘zigzag’ ‘muddy, soggy’ ‘disorderly’

(same vowel throughout)

b. pát̄pàt̄ jás̰̄ jàs̰̄

‘covered in warts’ ‘rough’

(different irst and second vowel)

As Courtenay points out, the above identiiable pattern involves the reduplication of a bisyllabic structure onto which a H-M-L-M tone pattern is imposed. She further indicates that such ideophones ‘always have [the] connotation of “irregularity”’ (19). As with other properties, this is one that characterizes more of Africa than NC (Creissels et al. 2008:126–127). For recent perspectives on the description and typology of ideophones, see Dingemanse (2011, 2012, forthcoming) and the papers in Voeltz and Kilian-Hatz (2001).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

244

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

9.15 Summary Although the preceding sections have attempted to discuss some of the most prominent characteristics of NC languages, we began by pointing out how dificult a task this presents. NC is vast and varied, and much of what is found in it is not restricted to NC, but rather spills over into (often adjacent) parts of Africa. We have reiterated the oft-made observation that NC languages are best distinguished from other African languages by their noun class systems. Although many of the other properties we have covered have a greater African distribution, we have not addressed the speciic lexical items, lexical semantics, and grammaticalization processes that have been variously identiied as ‘Africanisms’ (Greenberg 1959; Meeussen 1975; Heine & Leyew 2008). In Table 9.19 we recapitulate some of the NC properties discussed above and broadly indicate the extent to which they are found in Greenberg’s (1963) other three stocks. In the last column we also include Proto-Bantu for reference. The numbers we present are arrived at from our personal research and our understanding of the literature as well as consultations with other Africanists. A more ine-tuned summary would assign integers to the various subgroups of each stock. However, if the above is accurate, the properties that have 3 in NC, but only add up to 4 combining Nilo-Saharan, Afroasiatic, and Khoisan are (a) open syllables only, (b) labiovelars, (c) noun classes, and possibly (d) predicate focus marking. This distribution of features suggests that these properties are better diagnostics of NC languages than the other features considered throughout this paper. While many linguistic features, including

Table 9.19 Niger-Congo features compared to other African stocks Property Open syllables only Labiovelars ATR harmony Tone Stem-initial prominence Noun classes Incorporative Pronouns Verb extensions Pluractional marking S-Aux-OV Serial verb constructions Predication focus marking Case

Niger-Congo Nilo-Saharan Afroasiatic Khoisan 3 3 3 3 3 3 2 3 3 3 2 3 1

2 2 3 3 2 1 ? 2 3 2 1 1 3

1 1 2 2 2 1 ? 3 3 1? 1 2 3

1 1 1 3 3 2 1/3 2 3? 2? 2/3 ? 1

Proto-Bantu 3 1 2 3 1 3 1 3 2 2 1 2 1

3 = widespread; 2 = found in certain geographic zones or subgroups (possibly areally induced); 1 = rare or unattested (possibly a small number of languages through contact). Where two values are given for ‘Khoisan’, the irst refers to Khoe, the second to Tuu/Kx’a. (Our thanks to Denis Creissels and Gerrit Dimmendaal for their expert input, which we think has resulted in a more accurate, although rough, assessment.)

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

Niger-Congo Linguistic Features and Typology

245

those discussed here, are found more in NC languages and throughout Africa than typologically expected, we conclude that noun classes are still the primary distinguishing feature of NC languages.

Acknowledgements This chapter grew out of a group research project titled ‘Areal Features and Linguistic Reconstruction in Africa’ funded by a Mellon Project Grant from the University of California, Berkeley and a University of California Humanities Research Institute grant, 2015–2016, with Nicholas Rolle as the Graduate Student Researcher. The project culminated in a workshop with the same title held in conjunction with the 47th Annual Conference on African Linguistics, Berkeley, March 23–26, 2016, further supported by NSF Travel Grant 1546957 (Peter Jenks, PI; Larry Hyman, co-PI). Besides the co-authors, who contributed sections to this report, we would like to thank the other members of the team, Pius Akumbu, Geoffrey Bacon, Matthew Faytak, Maria Khachaturyan, and Spencer Lamoureux, for their contributions to the project and the workshop. Finally, we especially thank Maria Khachaturyan and Valentin Vydrin, who provided important input concerning the discussions of Mande, and Denis Creissels and Gerrit Dimmendaal, who either conirmed or suggested changes to the summaries in Table 9.19.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 01 Aug 2019 at 05:16:25, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.009

10 Afroasiatic Linguistic Features and Typologies Ronny Meyer and H. Ekkehard Wolff

10.1 Introduction The chapter presents general characteristics of Afroasiatic languages (Sections 10.1.1–3), summarizes selected results of comparative linguistic studies (Section 10.2), and compiles state-of-the-art generalizations concerning the typology of contemporary languages in two large areal groupings: North-Central Africa and Northeast Africa (Section 10.3). Sections 10.1 and 10.2 have been jointly prepared by both authors. Regarding Section 10.3, primary responsibility for dealing with Berber and Chadic languages under 10.3.1 lies with H. E. Wolff, while R. Meyer takes primary responsibility for dealing with Ethiosemitic, Cushitic, and Omotic languages under 10.3.2.

10.1.1

General

Afroasiatic (occasionally also Afro-Asiatic), formerly called ‘HamitoSemitic’, represents the language family of the longest diachronic perspective from the earliest attested written monuments of humankind, but also includes numerous unwritten languages, some of which only became known quite recently. The situation of Afroasiatic linguistics is exceptional in the African context for several reasons: (1) The languages of this phylum are indigenous to an area covering two neighbouring continents, Africa (mainly non-Semitic languages) and Asia (only Semitic languages). Relecting their past and present geographic distribution, two regionalized approaches to their study have developed: Oriental philology (including Semitic and Egyptian linguistics) and African linguistics. (2) Within African linguistics, Afroasiatic languages are often considered ‘less African’ than other African languages, the latter being most

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

247

commonly associated with ‘black’ or ‘sub-Saharan’ Africa. This relects uncertainties regarding the ultimate origin of the Afroasiatic languages, whether in Africa or in the Near East, which allow for competing scenarios of early migrations out of or into Africa.1 (3) Afroasiatic languages account for the longest tradition of linguistic scholarship, not the least based on ancient documents in the classical Semitic languages, ever since Jewish and Islamic scholars recognized the structural and lexical similarities between Hebrew, Aramaic, and Arabic in the Middle Ages. This view was later extended to encompass evidence from African languages like Geez and Amharic (Ethiosemitic), as irst established by Hiob Ludolf (Hammerschmidt 1965; Hayward 2000a) in the 17th century. (4) These classical Semitic languages are also the languages in which the earliest and fundamental documents concerning the three big monotheistic Oriental religions were written: Judaism, Christianity, and Islam. This fact had, and still has, a salient impact on the intensity and number of studies relating to texts in these languages. To this must be added the rich tradition of Egyptian scholarship, which began following the Napoleonic military expedition to Egypt (1798–1801) and the decipherment of the hieroglyphic inscriptions by Jean-François Champollion (1790–1832) in the 1820s.2 (5) Afroasiatic linguistics consequently found itself caught between more or less traditional Oriental philology approaches on the one hand, and modern linguistic approaches on the other. Semitic scholarship remained very inluential, leading to a ‘Semiticist bias’, which blurred the view on African Afroasiatic languages for a very long time – until both African and general linguistics entered the picture and submitted Semitic languages, and to a lesser extent also (Ancient) Egyptian, to the same analytical procedures as any other (living) language, in Africa or elsewhere. The heavily biased and, therefore, partly detrimental impact of former ‘Hamito-Semitic’ scholarship on Afroasiatic linguistics, particularly with regard to the vast majority of the languages of this family that are spoken in Africa, still tends to impede work on these languages, as was already pointed out more than 40 years ago by Newman (1984). It still forces researchers to take issue with received and often over-generalized concepts that are virulent in the literature 1

This debate suffers – accidentally and unhappily – from reminding us of the notorious but linguistically wrong

1 01 n-fN F-

‘Hamitic’ theory (see below), which lingers on in some contemporary academic cycles and public opinion also in Africa. 2

Interestingly, the sensational and highly valued early ‘civilization’ of Ancient Egypt, based on script, was located in Africa, and, therefore, (Ancient) Egyptian could rightfully be called an African language. But this fact was disregarded as somehow accidental in early European scholarship. Rather conveniently, Egyptology was academically administrated under the umbrella of pre-existing Oriental studies. Later and in many quarters, emerging autonomous African linguistics largely followed this pattern and considered (Ancient) Egyptian, often also extending this view to the Berber languages of North Africa, to lie outside its scope, leaving the study of these languages to Orientalists rather than Africanists.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

248

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

and need to be addressed, either explicitly or implicitly, in modern analyses and descriptions. All of these facts and observations motivate the uniquely extensive treatment of Afroasiatic languages and linguistics in much of the available literature on African languages to be reviewed, in both Orientalist and Africanist traditions, in general as much as for the present chapter. Here, we are dealing with Afroasiatic languages in Africa. We do so in two areal groupings: North-Central Africa and Northeast Africa. Since Ancient Egyptian is extinct as a spoken language, and most of the classical Semitic languages were and are spoken outside Africa (with the exception of Arabic), and both Egyptian and the classical Semitic languages have developed a philological tradition of their own outside African linguistics proper, they are not dealt with here in any detail. We do, however, consider the Berber languages of North Africa and the Ethiopian Semitic languages to be legitimate domains of African linguistics research. Within the Greenbergian (1963) classiication system (Chapter  1), Afroasiatic is the only phylum straddling other continents besides Africa, namely adjacent parts of Asia (hence the name Afroasiatic) and of Europe. With some 370 languages (Lewis et al. 2016), Afroasiatic is the second largest phylum in Africa. Its vast geographic spread, from Morocco to Iraq, and from Malta to central parts of both Nigeria and Tanzania, is paralleled by a high degree of typological diversity. This relects, partly at least, a long history of language contact with different neighbouring language groups, whether within Afroasiatic (such as Berber/Chadic and Cushitic/ Ethiosemitic, possibly Ancient Egyptian/Berber in very early periods) or with other genetic units like Niger-Congo languages in West Africa and Nilo-Saharan languages in Northeast Africa. Accordingly, Afroasiatic languages form part of the debate on areal convergence zones in Africa, such as the Macro-Sudan belt (Güldemann 2008), the Chad-Ethiopia macro-area (Heine 1975:41), the Ethiopian linguistic area (Crass  & Meyer 2008), and the Tanzanian Rift Valley area (Kießling et al. 2008); see Chapter 8. A further hotbed of language contact includes Berber languages (Tuareg) and their impact on the so-called Northern Songhay languages (Chapter 11). Afroasiatic languages share, to varying extent, lexical items inherited from the proto-language. Such common vocabulary will not be discussed; for this the reader is directed to Cohen (1947), Diakonoff et al. (1993; 1993– 1997), Ehret (1995), Kogan (2015), and Orel and Stolbova (1995).

10.1.2

Language Vitality

Afroasiatic exhibits both classical (often extinct) and modern (living) languages. The classical languages, speciically Egyptian and part of Semitic, no longer function as native languages (some are still used for religious purposes, for example, Classical Arabic, Geez, or Coptic), but are known

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

249

only from written records codiied centuries ago. Some other, unwritten languages simply vanished. Few of them are documented partially, for example, Ethiosemitic Mesmes (M. Ahland 2010) and Gafat (Leslau 1956), or Ongota (Birale) (Savà 2003; Tosco 2000a), or their former existence can be deduced from other sources, as for the Omotic Naga (Azeb 2012:424) or Southern Cushitic Aasáx and Kw’adza (Nurse 1988a; Petrollino & Mous 2010). Many living languages are currently threatened by extinction, for example, Shabo (Bender 1994b; Lemi 2017), whose classiication is impeded by insuficient data. Beside Arabic (Semitic), which is spoken by about 300 million people in various countries, only a few other Afroasiatic languages have more than 10 million native speakers, like Hausa (Chadic, 34 million or more), Amharic (Semitic, 25 million), Oromo (Cushitic, 25 million), Somali (Cushitic, 15 million) (Lewis et  al. 2016). Several languages have between 1 and 9 million native speakers, for example, Sidaama, Afar, Hadiyya (Cushitic), Tigrinya (Semitic), Wolaitta, Gamo (Omotic), but most languages are spoken by fewer than 1 million people, often fewer than 100,000 or even 50,000 people. Usually, speakers of smaller Afroasiatic languages are bilingual. In addition to their native language they also speak the regional lingua franca as second language, like Arabic in the Berberophone Maghrib, Amharic in Ethiopia, or Hausa in Nigeria. Bilingual speakers may mix two languages into a new language, as often argued for Ma’a/Mbugu, which combines a Cushitic lexicon with Bantu grammar (Mous 2004a; cf. also Greenberg 2005 for arguments against mixed languages). Other mixed languages probably exist in the contact zone between Oromo and Somali in Ethiopia, for example, Girirra (Mekonnen 2016: chap. 1), and Rer Bare or Adona (Bender 1975; Braukämper 2002:15, 137). Except simpliied Hamar (Lydall 1976:397) and the debatable pidgin origin of Amharic (Bender 1983b contra Girma 2009), pidgins and creoles typically involve Arabic: (Ki-)Nubi in Kenya and Uganda, Shuwa Arabic in Nigeria, Juba Arabic in South Sudan, or Turku in Chad (cf. Versteegh 1984:117–119; Manfredi 2017).

10.1.3

Writing and Script

Afroasiatic differs from other African language families through an early use of writing. The irst written records in Egyptian date back to 3200 BCE, and in Akkadian (Semitic) to the middle of the 3rd millennium BCE. Egyptian is codiied in hieroglyphs, that is, logograms conveying semantic information, and grammatical and phonetic values (Schenkel 1994; Wilson 2004; Rogers 2005:§6). The monumental hieroglyph scripts were modiied to cursive forms (i.e., the Hieratic and Demotic scripts), and eventually replaced by a modiied Greek alphabet, the Coptic script, around 300 CE (Grossman  & Richter 2015:81–85). Akkadian was written in a cuneiform script, originally used for Sumerian. It is a syllabic script, but also contains

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

250

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

logographs and graphemes in grammatical function (Krebernik & Nissen 1994; Rogers 2005:§5; Finkel 2010). Records of other classical Semitic languages (Aramaic, Hebrew, Arabic, Sabaic, etc.) go back to the irst millennium BCE. These languages were written in a consonantal (or abjad) script, in which certain graphemes optionally mark long vowels (Günther & Ludwig 1994; Daniels 1997). The Ethiopic script for Geez, which evolved between 100 and 300 CE, was initially an abjad that then was modiied to an alpha-syllabary (or abugida) in 350 (cf. Meyer 2016a). Amharic has also been written in this script since the 14th century, other Ethiopian languages since the 19th century. Tuareg (Berber) also uses an alphabetical script, based on an antique scriptural tradition – Libyco-Berber – which itself had been inspired by the Phoenician alphabet (Pichler 2007; Casajus 2015).3 Most Chadic, Cushitic, and Omotic languages, and several modern Semitic languages acquired writing either relatively recently, or not at all. Beside the Ethiopic script, Afroasiatic languages of Ethiopia and Eritrea have been reduced to writing in modiied Roman or Arabic scripts, the latter known as ajäm (Wedekind 1994; Pasch 2008; Mumin 2009; Watten-Rothlin 2009) and as ajami in West Africa. Occasionally local scripts were invented, but remained limited in use (cf. Tosco 2015 for Somali; Hayward & Mohammed 1981 for Oromo; Ahmed 2005 for Harari; Awagana 2011 for Hausa).

10.2 Comparative Afroasiatic and Genetic Classification 10.2.1

Overview

Hebrew and Arabic have been studied at European universities since the 14th century CE (Robins 1967:96). Geez and Amharic are known in Europe since the 16th century; their irst grammars and dictionaries were published by Hiob Ludolf in the second half of the 17th century (Hammerschmidt 1965:257–259). Most Afroasiatic languages were not discovered until the late 18th century (Sasse 1981a:132), irst by missionaries and travellers. Egyptian was deciphered in the 1820s (by Jean-François Champollion), and the Akkadian clay plates later in the 19th century. The linguistic afinity between Hebrew, Aramaic, Arabic, and Geez had long been recognized, but it was only in 1781 that A. L. von Schlözer grouped them as Semitic, referring to biblical Shem (Sasse 1981a:131). It soon became apparent that they were related to certain non-Semitic languages spoken in Africa for which F. Müller and K. R. Lepsius introduced the term Hamitic in 1876/1880, which is derived from the name of Shem’s younger brother Ham (Genesis 5:32). Subsequently, Hamito-Semitic became the cover term for Semitic and all African (even non-Afroasiatic) languages with a nominal gender system (Sasse 1981a:131–132). 3

For an overview about scripts and their function in Africa, see the contributions in Kootz and Pasch (2009).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

251

In the mid-19th century, the so-called Hamitic Hypothesis was developed (Sanders 1969), which divided the African population in two races. The ‘Hamitic race’ was connected to Caucasians, who assumingly migrated from the Middle East to Africa in ancient times; they were considered culturally superior to the native ‘Negroid race’ in sub-Saharan Africa. Consequently, linguistic features (particularly nominal gender) and nonlinguistic anthropological and racial attributes (e.g., mode of subsistence, skin-colour, type of hair) merged in the genetic classiication of North and East African languages, as propagated in Carl Meinhof’s Die Sprache der Hamiten (1912) which remained inluential until the appearance of Joseph Greenberg’s studies in the 1940s and thereafter. Meinhof’s notion of ‘Hamitic languages’ had already been questioned by A. Klingenheben and D. Westermann in the 1920s and 1930s and by Cohen (1947); but it was Greenberg (1950a) who inally established that Meinhof’s ‘Hamitic’ does not constitute a valid language family, but is rather a non-linguistic (racial) concept. Consequently, he proposed the term Afroasiatic – which had already been used by M. Delafosse in 1912 but remained unpopular – for former Hamito-Semitic (Greenberg 1950a:57) in order to avoid ‘Hamitic’ entirely. Later, terms such as Afrasian (in Russian linguistics), Lisramic, and Erythrean/Erythraic were coined as alternative designations, but none gained general acceptance. Greenberg’s (1950a:55) Afroasiatic family encompassed ive coordinate families, namely (i–v) in Figure 10.1, to which later Omotic (vi) was added whose status is still controversial. Fleming (e.g., 1969) established the Omotic family in the 1960s, whereas in earlier (and concurrently opposing) classiications it constitutes West Cushitic (for details, see Lamberti 1991:554–558; Petrollino 2016). Recently, Theil (2012:376; 2006) argues for an independent Omotic family outside Afroasiatic. Within Afroasiatic, separate branches were proposed for (North Cushitic) Beja (Hetzron 1980:96–98), (unclassiied) Ongota (Fleming 2006:135), and (East Sudanic) Meroitic (Rowan 2006; Rilly 2010). Even for the six branches in Figure  10.1, there is no consensus about their subgrouping (Peust 2012:225–227). Moreover, according to the Nostratic hypothesis, Afroasiatic may belong to an even larger super-family together with Indo-European and other language families (Dolgopolsky 1998; Bomhard 2008; Campbell & Poser 2008:§9.4).

Afroasiatic

i. Semitic

ii. Berber

iii. †Egyptian

iv. Cushitic

v. Chadic

vi. Omotic

Figure 10.1 Afroasiatic families

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

252

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

There is no consensus about the time (estimates range between 8,000 and 18,000 BCE) and the place at which Proto-Afroasiatic evolved (McCall 1998; Blench 2006:§6). One of the two main hypotheses suggests the Urheimat between the south-eastern Sahara and the Red Sea, implying a subsequent dispersal and movement out of Africa (Diakonoff 1965; Bender 2003). The other hypothesis, forwarded by A. Militarev (following A. Dolgopolsky), assumes an origin in the Levant, from where speakers of all Afroasiatic language families (except Semitic) moved into Africa (Diakonoff 1988:23– 25; Militarev 1996, 2002). As for Ethiosemitic, it is traditionally assumed that its speakers are descendants of Old South Arabian groups who crossed the Bab-el-Mandeb strait in the second half of the irst millennium BCE, and settled in the Ethiopian Highlands where they intermingled with Cushitic (and Omotic) speakers (Hetzron 1972:122–125; Marrassini 2011). The evidence and time frame for such a scenario has been questioned by some authors (e.g., Drewes 1958; Hudson 1977; Weninger 2011a:1115), but supported by others (e.g., Kitchen et al. 2009). There is a wealth of linguistic publications on Afroasiatic (cf. Hodge 1971; Sasse 1981a; Hetzron 1987; Appleyard 1999; Takács 1999; Frajzyngier  & Shay 2012a). The foundational work on reconstructed Afroasiatic grammar is Diakonoff (1988); a recent attempt towards a typological approach is Frajzyngier and Shay (2012c). Grammatical overviews on Afroasiatic families are found in Heine et al. (1981), Hayward (2000a), Frajzyngier and Shay (2012c), and Edzard (2012b), and descriptions of individual languages in Kaye (1997, 2007). Semitic languages are outlined in Hetzron (1997) and Weninger (2011b), while Woodard (2008) and Streck (2008) particularly deal with ancient Afroasiatic languages. Uhlig et al. (2003–2014) contains grammar sketches of most Afroasiatic languages in Ethiopia.

10.2.2 Main Sub-families/Branches of Afroasiatic 10.2.2.1 Ancient Egyptian Egyptian represents different diachronic periods of a single extinct language, spoken in the northern Nile valley for more than 4 millennia until ca. 1400 CE (Grossman & Richter 2015). It ceased to be spoken as a contemporary African language, but is retained for religious purposes in the Coptic Church. Through long periods of specialized philological research, Egyptology has established itself as a separate academic discipline. For further and recent information, see Grossman and Richter (2015), Haspelmath (2015), Loprieno and Müller (2012), Reintges (2018).

10.2.2.2 The Berber Languages The Berber languages, with the exception of Ancient Egyptian, constitute the most homogeneous family within Afroasiatic, to the extent that they are often referred to as a single language (especially in French scholarship)

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

253

with a plethora of extremely disparate ‘dialects’. They are sometimes subsumed under the term Amazigh languages or Tamazight ( *a-u (in diachronic analysis?) or sufix -uu and change of tone melody (in synchronic analysis). In contrast, in Berber Tashelhiyt ta-mazir-t becomes pl. ti-mizar ‘country’ by internal vowel pattern change only: -a-i > -i-a. Internal vowel pattern changes create so-called ‘broken’ plurals; internal and external devices of pluralization may combine, for example, Semitic Geez nigus ‘king’, pl. nägäs-t ‘kings’. External marking devices may relect, as frequently in Chadic, defunct former determinatives such as *-n, *-t, *-k, *-ɗ, *-i (Schuh 1983), which tend to fuse and petrify with both singulars and plurals (see Wolff 2004b).

10.2.3.3

Verbal Grammar

Major typological and partly controversial features still under research are conjugational patterns, like preix versus sufix conjugation, the nature of aspect/tense systems, and the interplay between thematic derivational and inlectional verb formations, which are best discussed at the level of individual families within Afroasiatic.

10.2.3.3.1 Verbal Plurality (Pluractional) The most widely spread marker is so-called ‘internal a’, sometimes accompanied by consonant replication. These formatives and their original pluractional function, in some divisions of Afroasiatic, have been transferred into the aspectual and derivational sub-systems. Therefore, on typological grounds, they may be compared to Semitic verb stems of the type qātala (internal a) and qattala (gemination/replication), even though synchronically these forms represent very different categories if, for instance, compared to pluractionals like in Chadic Lamang kla ‘take’, kala (internal a) ‘take many’, kalala (replication) ‘take many here and there’ (Wolff 2015a).

10.2.3.3.2 Systems of Tense, Aspect, and Mood (TAM) There is little agreement on the fundamental categories underlying the proto-language’s TAM system and on whether its basic structure was triadic or binary. Dichotomous sub-systems would include indicative/realis versus subjunctive/irrealis, state versus action, transitive versus intransitive. Aspectual apparently dichotomous notions such as punctual versus durative, and perfective versus imperfective, however, could relect a trichotomic system in which a basic unmarked category (aspect/tense-less) stands opposed to two marked fundamental categories: perfective and imperfective. An open question remains whether any and which of these formations were actually nominal rather than verbal, or even something in between, such as participial or gerundial. Likewise, there is debate over the original marking devices for the expressions of aspect, tense, and mood. The proto-language may have marked these categories through both

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

268

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

internal inlection in the root-and-pattern system, as well as preixal and/ or sufixal person marking. Where internal marking by vowel pattern or inix (cf. internal a) and/or consonant replication is involved, the ultimate origin of these forms could be in the pluractional rather than the aspectual sub-system. Relexes of the ancestral preix conjugation are maintained in Semitic and Berber, to a lesser degree in Cushitic. Chadic appear to have lost these to a large extent, Omotic and Egyptian have lost the preix conjugation completely. Despite considerable divergence in the systems of modern languages, scholars of Afroasiatic tend to agree that the verbal systems of all Afroasiatic languages can ultimately be traced back to the common proto-language; they immediately point out, however, that Egyptian underwent a unique development. Nominalized verb formations such as verbal nouns, participles, and predicative adjectives may also harken back to the proto-language for expressing state rather than action. This predicate form is commonly referred to as the ‘stative conjugation’ and uses sufixes for indicating the person, number, and gender of the subject. Some conjugational paradigms of today’s languages (and also of extinct ones such as Egyptian and Semitic Akkadian) show sufix conjugation, including practically all paradigms in Egyptian, the Akkadian stative, the West Semitic perfective, and the qualitative in Kabyle (Berber), which, however, may be independent innovations in the various families.

10.2.3.3.3 (Suppletive) Imperatives of Motion Verbs A common Afroasiatic feature is the irregularity of imperative forms of the verbs ‘come’, ‘go’, ‘take’, and ‘give’ by using speciic suppletive forms (cf. Newman 1980; Cohen 1984). For instance, in Proto-Chadic the normal verb stem and the imperative form of ‘come’ are reconstructable as *(-)sə and *ya respectively; in Berber Kabyle the corresponding stems are as and eyya; in Egyptian they are nn and mn; Semitic Amharic has mätt’a (3sm.pfv) and na (2sm.imp).

10.2.3.3.4 Verbal Derivation Verbs in Afroasiatic languages can be modiied to indicate semantic nuances and qualities of action through so-called ‘stirpes’ or ‘themes’ by either root modiication (inixes, such as ‘internal a’ to mark pluractional) or derivative afixes. Common are indication of repeated action (iterative), action caused to happen (causative), action with a demoted agent (passive), action affecting the subject (relexive), or action mutually affecting subject and object (reciprocal). Derivational markers may combine. Partial or complete reduplication of the verb base indicates repeated action and should be considered to represent pluractional formations – in this shape possibly borrowed from other African languages rather than being an inherited verbal derivative.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

10.2.3.4

269

Personal Pronouns

Systematically and etymologically, personal pronouns are considered to be very stable across most of the sub-families, the least so in Omotic. Verb conjugation displays very similar afixes to indicate person, gender, and number of subject. For example, the 3sf preix *ta- or *ti- is found in Berber (t-dawa ‘she healed’), in Chadic (Hausa tá-sàyáa ‘when she bought’), in Cushitic (Beja ti-dbíl ‘she collected’), and in Semitic (Akkadian ta-prus ‘she divided’). The corresponding sufix is found in conjugations in Berber (-t), Egyptian (-tj), Cushitic (-t), and Semitic (*-at). Many languages maintain the masculine/feminine gender distinctions in both 2nd and 3rd person, singular and plural. In Cushitic and Omotic, however, it is commonly limited to the 3rd person singular. Separate but formally similar sufixed pronouns are used to indicate object and possessor.

10.2.3.5

Word Order

Classical Semitic, Egyptian, and Berber use VSO order, but almost all Cushitic, Omotic, and Ethiosemitic languages use SOV order. Chadic languages usually have SVO order, while Central Chadic languages somewhat exceptionally have VSO order. If, however, as Igor Diakonoff suggested (e.g., 1965), Proto-Afroasiatic had ergative alignment, in which subject and object as traditionally construed are not valid concepts, then such simple formulas would not be suitable to relect the syntax of the proto-language. The issue remains open for further research.

10.3

Typologies of Contemporary Afroasiatic Languages in Africa

In this section, we will present, in more detail, typological generalizations concerning contemporary Afroasiatic languages in two sub-regions, that is, Northern Africa and northern parts of West and Central Africa versus the Horn of Africa region in Northeast Africa. The sub-regional approach is justiied by the fact that the languages of the Northeastern sub-region have long been considered to form part of a sprachbund (cf. 10.3.2.1). The question whether Chadic, Berber, and possibly Ancient Egyptian could also be seen as part of some linguistic convergence area has never been seriously considered, even though authors like I. M. Diakonoff, C. Ehret, C. Hodge, A. Militarev, P. Newman have suggested some closer ties between varying pairs among these three families. We will exclude Arabic, the hegemonic language over most parts of Northern and some parts of North-Eastern Africa, being a Semitic language of much more recent vintage in the area with, however, considerable impact on at least the lexicon of many Berber and, directly or indirectly, on some Chadic languages as well.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

270

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

10.3.1

North-Central Africa

In this sub-section, we look at Afroasiatic languages in the Maghrib and the western and central Sahara, that is, the Berber (or Amazigh) languages, as well as those along the southern fringe of the Sahara Desert in the wider Lake Chad Basin – the Chadic languages. Whether Libyco-Chadic or ChadoBerber form a closer unit, either genealogically within Afroasiatic or by a long history of contact, will not be addressed. Our treatment of Berber relies heavily on Kossmann (2012) including practically all examples; for shorter overviews, see also Hayward (2000a) and Wolff (1981b, 2008b). For Chadic, we acknowledge the recent treatments by Frajzyngier and Shay (2012b:236–341) and Schuh (2017).7 For shorter overviews, see also Hayward (2000a) and Wolff (1981b, 2008c).

10.3.1.1 Phonological Features in Berber and Chadic 10.3.1.1.1 Suprasegmental Phonology Berber languages, like Semitic, are not tone languages; some languages use contrastive stress (Tuareg;8 Eastern Berber). Chadic languages, like Cushitic and Omotic, are tonal. They distinguish two or three pitch levels for contrastive tone both in lexical and grammatical functions. Assuming that Proto-Afroasiatic was non-tonal and likely had a pitch-accent system, synchronic tone systems, as in Chadic, may historically derive from pitch-accent systems. In some Chadic languages, voiced obstruents in syllable onset position have a pitch lowering role on the syllable (‘tonal depressors’), while voiceless consonants may have the opposite effect, that is, pitch raising, which may account for raised low and lowered high emerging as a third intermediate pitch level. On this basis, Wolff (1983c, 1987a, and elsewhere) proposed a theory of tonogenesis for Chadic.9 In some languages, toneless morphemes occur, which receive their pitch realization either by polar tone to, or by tone copy from, a neighbouring syllable (e.g., WCh Hausa). Occasionally, even (voiced) fricatives may carry pitch/tone and constitute a phonetic syllable, for example, CCh Lamang z`ɗál ‘man, husband’. Not all lexical roots must be tonal; in some languages

7

Differences between the treatment of Chadic presented here and in both Frajzyngier and Shay (2012b) and Schuh (2017) testify to (a) persisting uncertainties regarding diachronic developments and (b) choice of descriptive categories used by the authors. For reasons of space, examples stem mainly from two languages that one author is particularly familiar with, i.e., WCh Hausa (cf. Wolff 1993; Newman 2000) and CCh Lamang (Wolff 2015a). Readers are advised to cross-check with Frajzyngier and Shay (2012b) and Schuh (2017), whose examples rest mainly on languages that, in turn, they are particularly familiar with, and to consult individual monographic descriptions by other authors.

8

In the absence of a generally accepted common language name for the various varieties spoken by Tuareg, the ethnonym is used; note that twa:reg is a plural form referring to people rather than to language.

9

At least one language has an intermediate split system: CCh Lamang is tonal regarding grammatical morphemes and verbal grammar, but is largely non-tonal with traces of pitch accent in its nominal grammar where it also shows traces of tonal depressors (Wolff 2015a); its closest sister language Hdi, on the contrary, is tonal throughout (Frajzyngier & Shay 2002). Interestingly, both languages are also linked by ‘tone reversal’ (Wolff 2013d).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

271

they receive their pitch realization by default or by automatic assignment of grammatical tone.

10.3.1.1.2 Segmental Phonology Syllables Generalizations concerning syllable structures depend on a clear distinction between consonants and vowels, which is debatable on the level of both diachronic and abstract synchronic phonological analysis for both Berber and Chadic languages (cf. also the notion of ‘weak radicals’ in Semitic linguistics). Given the almost consistently doubtful status of short central vowels adds to the systemic dificulties of phonological analysis in Berber, but also in Chadic where consonant clusters tend to be dissolved by inserting epenthetic (usually short and centralized) vowels. In both Berber and Chadic, vowel-initial syllables are acceptable. WCh Hausa is rather exceptional in requiring a consonantal onset and thus inserts an initial glottal stop or /h/ (Newman 1976).

Consonants Consonant systems in Berber languages are diverse; they all distinguish short/long and voiced/voiceless consonants. Consonant length is also discussed in terms of fortis/lenis or tenseness (cf. Galand 2002), or biphonemic gemination (e.g., Saïb 1977). Consonant lengthening and shortening both occur in morphophonological processes, allowing for alpha-switch rules regarding the features voice and spirantization. Typical accompanying features of Berber consonants are pharyngealization, giving rise to so-called ‘emphatic’ consonants that have a lowering effect on adjacent vowels. Other additional features are spirantization (i.e., changing short stops into fricatives, often accompanied by fronting) and labialization (of velars and uvulars); these processes would appear to be areally conined (Northern Berber). Further and due to long-term language contact, loan phonemes from Arabic enlarge the consonantal inventory of individual Berber languages. For illustration, see charts of consonant systems for Kabyle and Ahaggar Tuareg in Tables 10.1 and 10.2 (slightly modiied from Kossmann 2012:23–24; rare phonemes are given in parentheses). Chadic languages are also not uniform with regard to their consonantal inventories and differ remarkably from Berber inventories. They share, however, triadic sets of obstruents (voiceless-voiced-glottalized) which are reconstructed for three of the four points of articulation for PC (labials, alveolars, palatals, but not for velars). In addition, modern languages tend to have a fourth set of prenasalized stops, which, however, are not reconstructed for PC (Newman 1977a). Lateral continuants (l, ɮ) are attested widely in Central Chadic and only rarely in West Chadic, they are absent in East Chadic. The consonants p/f and often b/v do not regularly contrast. Rare occurrences of labial-velars (CCh Bacama, Afade) as much as of +ATR vowel harmony (WCh Tangale, ECh East Dangla), are due to areal contact

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

272

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Table 10.1 Consonant system of Kabyle

Labial Vl. plosive Vl. fricative Vd. plosive Vd. fricative Phar./labial. vl. plosive Phar./labial. vl. fricat. Phar./labial. vd. plosive Phar. /labial. vl. fricat. Nasal Glide Rhotic Phar. rhotic Liquid Phar. liquid

Interdental Dental

f, ff θ (b), bb β ð bbw ð̣ m, mm w

(t) s, ss (d), dd z, zz (t. ), t. t. s., s.s. (d.) z., z.z. n, nn

Prepalatal Palatal Velar (cˇcˇ) š, šš (gˇgˇ) ž, žž

(k) ç ʝ

(š.), (š.š.) çw (ž. ), (ž. ž.) ʝw

Uvular

Pharyngeallaryngeal

q x, xx

ħ, ħħ

gg

ɣ, (ɣɣ) ʕ, ʕʕ (kw), kkw qw, qqw xw, xxw (gw), ggw ɣw, (ɣɣw)

y

h, hh

r, rr r., r.r. l, ll (l.l.)

Table 10.2 Consonant system of Ahaggar Tuareg Labial Vl. plosive Vl. fricative Vd. plosive Vd. fricative Phar. vl. plosive Phar. vd. plosive Phar. vl. fricative Nasal Glide Rhotic Liquid

f, ff b, bb

m, mm w, (ww)

Dental t, tt s, ss d, dd (z), zz t. t. d. z., z.z. n, nn

Palatal

Velar

Uvular

k, kk

qq (x), (xx)

(š), (šš) gy, ggy ž, žž

( ɲ) y, (yy)

Laryngeal

ɣ

(ŋŋ)

h, (hh)

r, rr l, ll

(cf. Macro-Sudan belt; Güldemann 2008). For illustration, see charts of consonants systems in Tables  10.3 and 10.4 for WCh Hausa (Standard Hausa: orthographic, acc. to Newman 2000:392) and CCh Lamang (semiorthographic, acc. to Wolff 2015a:1:61, modiied).

Prosodies As regards long components (‘prosodies’) reaching beyond the syllable or morpheme, Berber recognizes pharyngalization in language-speciic ways, while in Chadic, particularly in the central branch, palatalization (‘Y-prosody’ fronting vowels and/or palatalizing consonants; cf. R. M. Newman (1977), for example) and labialization (‘W-prosody’ rounding vowels and/or labializing consonants) may function as prosodies. The assumption and presence of such prosodies has immediate repercussions on the type of underlying synchronic structures and diachronic reconstructions. In terms of

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

273

Table 10.3 Consonant system of WCh Standard Hausa

Obstruents

Voiceless Voiced Glottal Voiceless Voiced Glottal

Sonorants

Labial

Coronal

Palatal

Velar

Labialvelar Palatalvelar

(f, fy) b ɓ f, fy

t d ɗ s z ts n

c j ‘y sh (j)

k g ƙ

kw gw ƙw

m l

ky gy ƙy

Laryngeal

‘ h

r r ̃a y

w

a Apical tap or roll (trill), as opposed to retrolex lap r; not represented as such in Standard Hausa orthography.

Table 10.4 Consonant system of CCh Lamang Post-alveolar

voiceless voiced prenasalized glottalized Affricates voiceless voiced prenasalized Fricatives voiceless voiced Lateral fricatives voiceless voiced Sonorants nasals Liquids: trilled lateral Approximants Obstruents Stops

Pre-alveolar Alveolar

Unrounded

Rounded

p b mb ɓ

k g

ŋg

kw gw ŋgw ~ ŋw

h gh

hw ghw

ŋ

[ŋw]

t d nd ɗ [~ ’] ts [~ c] dz [~ j] ndz ~ nz s [~ sh] z sl zl n r l

f v

m

w

y

’

phonetic realization, not only genuine vowels, but also underlying sonorants and approximants (as well as epenthetic vowels) may occur in syllable nucleus position, also meaning that they can carry tone. Internal and proto-language reconstructions for Chadic, in particular, are complicated by such observations, cf. (C = ‘radical’ consonant/sonant/vocoid): (1) WCh Hausa

synchronic

underlying/diachronic

to be compared to

[d óo k ìi] CV CV

*d á w k ìi C1V C2C3V

pl

d á w -áa- k íi C1VC2- V- C3V

gloss

‘horse’

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

274

PA RT I I

(2)

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

synchronic underlying/diachronic to be compared to gloss pluractional: w-a-m a ‘to marry’ CCh Lamang [uma] *w m a C1-V-C2V VCV C 1 C2 V PC *a (w)ku ‘goat’ [ógò]/agu/ *+w/ a g w / (Newman 1977a:27) VCV VC1C2

Vowels Berber vowel systems are not uniform, ranging between 3 vowels i, a, u (e.g., in Tashelhiyt) and 7-vowel systems (with additional ə, ɐ, e, o; e.g., in Ghadames and Tuareg), the central vowels ə, ɐ tending to be shorter in duration. Northern Berber has three ‘plain’ vowels (i, a, u) plus one central vowel ə, which as in Chadic often, but not always, has doubtful phonemic status. When the status of schwa is non-phonemic, and again as in Chadic, central ə occurs with consonantal syllabicity (along a sonority scale) and by epenthetic vowel insertion (cf. Kossmann 2012:28–33). Chadic languages tend to restrict the distribution of both consonantal and vocalic segments within the phonological word, that is, whether they are allowed to occur in initial, medial, or inal position. WCh Hausa, for instance, disallows any vowel in initial position but requires a consonantal onset; it disallows short e and o in medial position, but allows all 5 short and 5 long vowels to occur in inal position. Generally, ‘[i]n many languages . . . vowels are not fully contrastive, the distinction between i and u, ə and i, and/or ə and u being neutralized in speciic phonological environments’ (Newman 1977a:12). For most Chadic languages, traditional vowel systems with 5 or more short (plus sometimes also long) monophthongs plus diphthongs are accepted as phonemic, like for WCh Hausa /a, e, i, o, u, aa, ee, ii, oo, uu, ai, au/. Languages mainly of Central Chadic (sub-branch A) can be analysed as having maximally two phonemic vowels /a, ə/, or just one /a/, or none at all, depending on the level of abstractness of analysis (Wolff 2017e). In many Chadic languages, the mid vowels e and o are clearly of secondary status, ‘being restricted to loanwords, being recently derived from diphthongs, or being conditioned variants of other vowels’ (Newman 1977a:12). Since distinctive vowel length occurs in some but not all languages, individual Chadic languages have been described with up to 14 and 17 vowels, like ECh East Dangla (Ebobissé 1979) and Tumak (Caprile 1975), others with maximally one vowel phoneme /a/, as CCh Moloko (Bow 1999) and deep-level analyses of CCh Wandala (Wolff  & Naumann 2004) and Lamang (Wolff 2015a). Nasalized vowels are extremely rare (ECh Tumak). Some languages allow symmetrical systems of short and long vowels, others have asymmetrical systems (i.e., not all short vowels may also occur long) or have no long vowels at all; in some languages, however, only the low vowel /a/ appears both short and long (e.g., WCh Miya). Abstract and/or diachronic language-speciic analysis may reveal that long vowels represent underlying sequences of segments, either identical short vowels or diphthongs (with subsequent monophthongization), like in the above example of WCh Hausa dóokìi < *dáwkìi (pl dáwáakíi) ‘horse’.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

275

In synchronic descriptions, the vowel ə (schwa) is treated as a phonemic vowel for some languages, and as fully predictable epenthetic vowel for other languages. In diachronic perspective, the phonological status of schwa remains obscure. Uncertainty about the inventory of vowels also interferes with descriptions of phonological processes. Both umlaut, that is, vowel changes triggered by (distant) assimilation to vowels in neighbouring syllables, and ablaut, namely, grammatically conditioned vowel changes, are widely discussed, yet it is not always clear whether one is dealing with one or the other, particularly with regard to so-called ‘internal a’, which may represent either ablaut (origin: *-a- pluractional verb inix) or umlaut (origin: verbal and nominal sufixes). Diachronic features of the Proto-Afroasiatic root-and-pattern morphology may survive where paradigmatic changes of vowel sequences can be observed, resembling so-called ‘broken plural’ formations like (3)

(4)

WCh Hausa

Berber

gúlbíi (*gulb-) kúrmìi (*kurm-) dúutsèe (*duwts-) sg sg

-a-i > pl -i-a -a-i > pl -u-a

gúlàabée kúràamúu pl dúwàatsúu

pl

pl

’stream’ ‘copse, jungle’ ‘stone’

ta-mazir-t, pl ti-mizar ’country’ a-gadir, pl i-gudar ‘wall’

One may also be dealing with reciprocal umlaut that harmonizes all vowels within the same word (and/or under the impact of prosodies, for example, CCh Lamang ógò < w/agu/ ‘goat’, éwé < j/awi/ ‘mouth’). Under-analysis of so-called ‘internal a’ as verbal inix (marking pluractional) on the one hand, which is often treated as ablaut, and umlaut assimilation of root vowels to /a/ by a vowel of a sufix (marking plural) of nouns on the other (cf. Newman 1990), explain further analytical irritation with regard to comparative Chadic and Proto-Chadic reconstructions.

10.3.1.2 Nouns and Nominal Morphology 10.3.1.2.1 Simple and Derived Nouns Berber more than Chadic languages conform to the inherited root-andpattern system, that is, they display abstract consonantal roots, which carry the basic lexical meaning, and sets of vocalic patterns, which carry grammatical meaning. In Berber, a trilateral root √-k-r-s- (the hyphen representing slots potentially illed by vowels), which connotes the general idea ‘tie/tying’, can be made into a verb, as in tə-kras ‘she ties’, as well as into a feminine noun t-akərris-t ‘knot’. This has parallels in WCh Hausa: a biliteral root √ t-r-, which connotes the general idea ‘collect/gathering’, derives verb forms like táar-àa ‘collect sth.’, táar-óo ‘collect and bring’, tàar-ú ‘assemble’, but also nouns like táarìi ‘heap’, tàaróo ‘assembly’, táarúu ‘large ishing net’. On the level of roots, nouns are not distinguishable from verbs.

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

276

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Nouns may be simple or derived; compounding is rare. Derived nouns often have agentive, instrumental, and locative meanings. Agentive derivations in both Berber and Chadic use the still productive Proto-Afroasiatic preix *m(V)-, for example. (5)

Berber Tashelhiyt a-mzdaɣ < *m-zdaɣ ‘inhabitant’ < √ z-d-ɣ ‘live in’

(6)

WCh Hausa

‘parent, father’ < √ hayf- ‘give birth’

má-hàif-íi

The same preix is also used for instrumental and locative derivations in Chadic, but not in Berber, which uses the preix a- instead: má-háƙ-íi ‘tool for digging holes’ < √ haƙ- ‘dig a hole’ má-háƙ-áa ‘mine’

(7)

WCh Hausa

(8)

Berber Tashelhiyt a-smsd < *s-msd ‘whetstone’ < √ m-s-d ‘sharpen, whet’ a-sks < *s-ks ‘place for pasturing’ < √ k-s ‘graze’

Nouns may be preceded by a limited and partly lexicalized set of derivative markers to encode plurality, and expressions for ‘son(s) of’, ‘daughter(s) of’ to indicate group membership; in Berber Zenaga as in West Chadic Hausa, the latter (also) marks diminutives. (9)

Berber Figuig

(10)

Berber Zenaga

id- w-alli pl-ann:m-brain at-t-mira sons-ann:m-beard

‘several brains’ (cf. alli ‘brain’) ‘a group of bearded persons’

‘small ram’ (cf. əgrər ‘ram’)

aɣ-əgrar-t son-ram-sg.f

Compare similar constructions in Chadic:10 (11)

(12)

WCh Hausa

sú Múusáa assoc.pl Moses ɗá-n Íngìlà son-gen England ɗá-n ƙàrámí-n yáaróo son-gen small-gen boy

CCh Lamang lá-ghìlì lá-dzə` vó

‘Moses and the others’ ‘Englishman’ ‘a very small boy’

‘thieves’ tōrē

‘to lift up’

(33)

Masa Zime-Mesme

*bà *tí

>*b=bà > pà >*t=tí-a > *tté > dé

‘to love’ ‘to call’

(34)

WCh Kanakuru

*mut > múrì

>*mut=t > mútè

‘to die’

Pluractionals may also be formed by partial or complete reduplication of root material, both in terms of rightward and leftward reduplication, with or without ‘internal a’, for example, (35)

CCh Lamang: complete reduplication of verb root √ kla ‘to take’ >k=a=la > k=a=la+k=a=la ‘to take (many times/things)’ take:pla.pla take:pla

(36)

WCh Hausa: partial reduplication of verb root (left- and rightward) √ haif- /hayf-/ ‘to give birth’ >*hay=hayf> hàhhàifáa ‘to give birth (many times)’ give birth:pla

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

284

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

See also *hay=a=fgive birth:pla

> *hay=a=f=yafgive birth:pla.pla

> hàyàyyáfàa

For some verbs, pluractionals are represented by suppletive stems. Diachronically non-transparent pluractional formations (involving added or morphophonemically modiied root consonants, but also borrowed suppletive stems), such as found in some West Chadic languages, are likely due to language contact with neighbouring Benue-Congo languages (Wolff & Gerhardt 1977). As for Berber: Tuareg uses (complete) verb stem reduplication to derive pluractional verbs stems. As pluractionals are nowhere else to be found in Berber, we are most likely dealing with borrowing from either Chadic (Hausa) or Songhay, which both have reduplicative pluractionals.

10.3.1.3.2.2 External derivation: forming plural agreement verb stems Pluractionals must not be confounded with so-called plural agreement verb stems (Newman 1990), which are externally derived by sufix to overtly mark number agreement between subject plurality and verb stem (a device that allows number distinctions to be neutralized with subject marking pronouns, as, for instance, in CCh Gisiga); examples are from Newman (1990:108, 113). (37)

a. WCh Karekare

b. CCh Gisiga

nà ngàtaa kò 1sg fall compl mù ngat-ʔan kò 1pl fall-plagr compl kə 2pers kə 2pers

kaɗ kill kəɗ-am kill-plagr

‘I fell’ ‘we fell’

‘you (sg) kill’ ‘you (pl) kill’

Newman (1990) reconstructs a verb sufix *-(a)n for plural agreement marking in Proto-Chadic. It remains an open question whether there is some diachronic relationship to the sufixal nasal found with 2nd and 3rd person plural subject marking elsewhere in Afroasiatic (Wolff 2011e:33), which, if established, would indicate that plural agreement marking on verbs was a property of Proto-Afroasiatic (cf. Newman 1990:118f.). (38)

Cushitic Rendille

ti-gis 2pers-kill ti-gas-en 2pers-kill(:pla?)-2pers ( -kə ni

sú=n

c. g.

(Ø, -*ɗ i)

-ha=ŋ(i)

Table 10.7 Derivative prefixes in Berber ssmm-/nntt-/ttwa-/etc.

‘factitive’ derivation with transitivizing and causative readings ‘middle’ derivation with reciprocal readings ‘passive’ derivation (passive and medio-passive)

pluralizer *N in pronoun shapes *CVN(V) (cf. Frajzyngier & Shay 2012b:266; see Table 10.6). There are no plural agreement stems in Berber. One may wonder, however, whether plagr marking in Chadic has historically anything to do with the sufixed elements of ambiixal plural person marking elsewhere in Afroasiatic.

10.3.1.3.3 Extended Verb Stems Berber and Chadic languages share the potential for forming extended (‘derivational’) verb stems, but differ considerably in their morphology. Verbal derivation in Berber is limited and exclusively by preixing (see Table 10.7). Derivational preixes may combine, rarely more than two at a time; only Tuareg appears to allow almost unrestricted combinations (Kossmann 2012:38). (39)

Berber Iwellemeden

nə-mə-təwə-sə-ssə-gru med-med-pass-caus-caus-understand:a ‘make understand each other’

Chadic languages, in particular those of the Central Chadic branch, tend to be rich in inventories of (exclusively sufixal, at times clause-inal particles) derivative morphemes that modify the verb, both grammatical in terms of argument structure, and other semantic nuances, including locative-directional speciications. Extension afixes, etymologically originating from nouns, prepositions, and possibly deictic or anaphoric pronouns, relect transparent and opaque grammaticalization paths. Where extension systems have been studied in greater detail in Central Chadic,

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

286

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

researchers noted an intimate relationship between the use of extension afixes and categories akin to aspect, nowadays often described in terms of specificity/referentiality. This puts extended verbs semantically into the neighbourhood of perfective (macro-)aspect, and – by analogy – simple verbs into the neighbourhood of imperfective (macro-)aspect. This was irst noticed for CCh Margi by Hoffmann (1963) and was conirmed, in more detail and complexity, by other researchers for other Central Chadic languages (e.g., Smith 1969; Wolff 1983b/2015a). In WCh Hausa descriptive grammars, the common Chadic verb extension system hides behind the label ‘grade system’ introduced by the London school of Hausa studies (see Parsons 1960). The Hausa ‘verb grades’ (more recently re-numbered 0 through 7, Newman 2000) relect an amalgamation of phonological classes of verbs, stem preparation for anaphora and object pronoun incorporation, and derivative extensions (Wolff 1993). The extensions relect some of the more widely spread derivative categories of Chadic verbs: applicative, benefactive (dative), causative, completive, efferential (itive), separative-partitive, source/agent/ subject-affectedness, ventive, and others (see ‘verbal complex’ below).

10.3.1.3.4 Inflection I: Use of Clitic/Bound Pronominal Subject and Object, and of Independent Pronouns In both Berber and Chadic languages, the independent pronouns are mostly used in non-verbal sentences, but also in topic position with verbal predicates. Pronominal clitics tend to be distinguished both formally and according to their functions as subject, direct and indirect object with verbs, and possessives with nouns. Berber and Chadic differ in terms of using gender distinctions with pronouns. Whereas Chadic generally has no gender distinctions in the plural (and many Central Chadic languages not even in the singular), in Berber it may apply to all persons except 1sg. In Berber, all verbs must mark person, number, and gender (PNG) of subject by preix, sufix, or circumix according the particular PNG combination. Verbal subject inlection always distinguishes gender in the 3sg/pl, but does not do so in 2sg, but maintains gender distinction in 2pl. Sets of subject PNG markers differ according to functional conjugational paradigms, such as stative, future, imperative, and so on and across the language family. Typically, preixal, sufixal, and circumixal marking has the following distribution according to PNG, while the so-called stative, where it is present, has sufixes throughout (cf. Kossmann 2012:45, based on Ghadames; see Table 10.8). Imperatives and adhortatives show variant PNG marking, involving dual and/or inclusive-exclusive distinction for 1pl in, for instance, Tashelhiyt and Ghadames (Kossmann 2012:45ff.), reminiscent of one of the ‘areal’ features that are diagnostic for the so-called Macro-Sudan belt (Güldemann 2008) south of the Sahara Desert (see Table 10.9).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

Afroasiatic Linguistic Features and Typologies

287

Table 10.8 PNG marking with verbs in Berber

1sg 2sg.cg 3sg.m 3sg.f 1pl 2pl.m 2pl.f 3pl.m 3pl.f

general

‘stative’

___-*a˘ɣ t-___-*ə d y-___ t-___ n-___ t-___-a˘m t-___-ma˘t ___-a˘n ___-na˘t

___-*a˘ɣ ___-*ə d ___ Ø ___-a˘t

___-it

Table 10.9 Imperative and adhortative marking with verbs in Berber

Imperative

Adhortative

2 sg.cg 2 pl.m f 1 cg.dual/excl 1 pl.incl.m f

Ghadames

Tashelhiyt

___Ø ___-a˘t ___-ma˘t n-___ n-___-a˘t n-___-ma˘t

___-aɣ ___-ataɣ ___-amtaɣ

Chadic languages relect different word order types, the majority being of SVO type. Some Central Chadic languages are VSO. This links up with the observation that most Chadic languages use preverbal subject marking by pronoun, while others use post-verbal pronominal subject marking, which again allows for medial position between reduplicated verb roots, like in CCh Wandala. The nature of the pronouns, whether clitic or bound, varies between languages. In some languages, clitic or bound pronouns are replaced by (at least diachronically) independent pronouns, completely or in certain conjugational paradigms only. WCh Hausa, for instance, uses diachronic independent pronouns (Newman & Schuh 1974), which synchronically are reanalysed as portmanteau morphemes encoding not only person, gender, and number, but also perfect I (absolute/in-focus perfect) when preceding the verb. Other TAM paradigms use transparent preverbal ‘person-aspect/tense markers’ in Standard Hausa in, for instance, the perfect ii (relative/out-offocus perfect; historicus): {High tone and Ø/-kà}; in the subjunctive/jussive/ optative and ‘neutral’ (Newman 2000:593) marking is {Ø} (see Table 10.10). Pronoun shapes of 1st and 2nd person in Chadic, both singular and plural, are considered rather conservative with regard to retaining similarities with those in other families of Afroasiatic. Third person pronouns tend to etymologically relect diachronic determiners, which Schuh (1983) had reconstructed for Proto-Chadic as *n, *t, *k, *ɗ, *i (see WCh Hausa 3sg.m and 3sg.f in Table 10.10).

Downloaded from https://www.cambridge.org/core. Nottingham Trent University, on 06 Sep 2019 at 17:40:54, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108283991.010

288

PA RT I I

C O M PA R AT I V E A N D D E S C R I P T I V E A F R I C A N L I N G U I S T I C S

Table 10.10 Preverbal PNG marking with verbs in WCh Hausa (Standard) perfect I singular náa káa cg kín yáa cg táa

1 2m f 3m f

perfect II plural mún kún

singular ná-Ø ká-Ø kí-kà yá-Ø tá-Ø

sún

subjunctive plural mú-kà kú-kà sú-ká

singular n` kà kì yà tà

plural mù kù sù

Table 10.11 Postverbal person/number marking with verbs in CCh Lamang [-inclusive] 1 2 3

[-plural] -i, -yo