Aspects of (Post)Colonial Linguistics: Current Perspectives and New Approaches 9783110436907, 9783110442229

Citation preview

Aspects of (Post)Colonial Linguistics

Koloniale und Postkoloniale Linguistik Colonial and Postcolonial Linguistics Edited by Stefan Engelberg, Peter Mühlhäusler, Doris Stolberg, Thomas Stolz and Ingo H. Warnke

Volume 9

Aspects of (Post)Colonial Linguistics Current Perspectives and New Approaches Edited by Daniel Schmidt-Brücken, Susanne Schuster and Marina Wienberg

ISBN 978-3-11-044222-9 e-ISBN [PDF] 978-3-11-043690-7 e-ISBN [EPUB] 978-3-11-043402-6 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2016 Walter de Gruyter GmbH, Berlin/Boston Printing and binding: CPI books GmbH, Leck ♾ Printing on acid free paper Printed in Germany www.degruyter.com

Contents Daniel Schmidt-Brücken, Susanne Schuster and Marina Wienberg Preface | vii Liesbeth Zack  Arabic language guides written for the British Army during the British occupation of Egypt, 1882–1922 | 1 Murad Suleymanov  Baku Russian: Colonial heritage in a post-colonial era | 27 Eeva Sippola  Postcolonial language ideologies: Writing in Chabacano | 53 Martina Drescher  Hybridized discourse markers in Cameroonian French? The example of déjà | 79 Andrei A. Avram  Kolonial-Deutsch and Français Tirailleur | 101 Janneke Diepeveen and Matthias Hüning  The status of Dutch in post-colonial Suriname | 131 Susanne Mohr  From Accra to Nairobi – The use of pluralized mass nouns in East and West African postcolonial Englishes | 157 Astrid Alexander-Bakkerus  A colonial grammar or arte | 189 Kazuko Matsumoto  The role of domain and face-to-face contact in borrowing | 201 Norbert Schaffeld  Representing bicultural knowledge systems and epistemic decolonization | 229 Marina Wienberg  Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 247

vi | Contents

Index of Authors | 265 Index of Languages | 269 Index of Subjects | 271

Preface The new field of Colonial and Postcolonial Linguistics is continuing to gain traction in the international scientific community – this volume is a testament to that fact. It is comprised partly of papers given at the First International Bremen Conference on Language and Literature in Colonial and Postcolonial Contexts 2013, and partly of original contributions. The internationality of the volume’s contributors and the themes discussed both speak to the global relevance of a discipline that deals with the descriptive and critical investigation of linguistic phenomena related to colonial circumstances and postcolonial effects of language structures and language use. In the meantime, the Second Bremen Conference was held in 2014, the preparation for its proceedings as much in preparation as the third conference in 2016. These conferences as well as the edited volumes resulting from them are activities and results of the project “Koloniallinguistik – Language in Colonial Contexts” of the University of Bremen, funded by the Excellence Initiative of the German federal and state governments. In this volume’s peer-reviewed studies, the diversity of Colonial and Postcolonial Linguistics is presented in a showcase of its current perspectives and new approaches. The first contribution by Liesbeth Zack deals with “Arabic language guides written for the British Army during the British occupation of Egypt, 1882–1922”. The author examines five Arabic language guides for British soldiers used in a context of military hierarchy and linguistic diversity, the focus being on vocabulary, grammar, and usefulness of the books in question. Murad Suleymanov’s study on “Baku Russian: Colonial heritage in a post-colonial era” takes up on the perspective of the linguistic situation in the post-Soviet of Azerbaijan. Once a lingua franca, Russian has lost influence due to its speakers’ emigration. As a now-minority language in Baku, the Baku variety of Russian is increasingly subject to language change on lexical and syntactic levels. This leads to the development of a new distinct dialect, the author argues. In her paper on “Postcolonial language ideologies: Writing in Chabacano”, Eeva Sippola focuses on literacy of creole speakers in the postcolonial context of the Philippines. Under the premise of a sociolinguistic methodology, the author studies orthographic practices in the Chabacano language and the creative reappropriation of the former colonizers’ languages to express statements of creole language ideologies. Martina Drescher’s article tackles the question of “Hybridized discourse markers in Cameroonian French? The example of déjà”. In the postcolonial context of Cameroon, speakers of the local French variety use the adverb déjà as a discourse-structuring unit showing semantic ambiguity that is not easily resolved, as the author shows in analyses of media and face-to-

viii | Preface

face data. The contribution of Andrei A. Avram takes a look at the planned colonial languages “Kolonial-Deutsch and Français Tirailleur”. He can show that there are structural similarities in both languages with respect to vocabulary, morphology, and syntax. As constructs of colonialist language planning, the author shows that both languages reflect racist stereotypes. In their paper on “The status of Dutch in post-colonial Suriname”, Janneke Diepeveen and Matthias Hüning focus on the diglossic relation between the colonialist language Dutch versus the English-based creole Sranantongo in the South-American country and former Dutch colony Suriname. The uses of both languages – Dutch as the language of public life and Sranantongo as a widespread lingua franca – are analyzed against the background of a post-colonial national identity. In her article “From Accra to Nairobi – The use of pluralized mass nouns in East and West African postcolonial Englishes”, Susanne Mohr scrutinizes in a multilevel analysis the uses of mass nouns in their differing contexts and with respect to areoversals in those African varieties of English. In her paper on “A colonial grammar or arte”, Astrid Alexander-Bakkerus takes an in-depth look at two 18th century manuscripts in Spanish, Quechua, and Jebero, respectively. The author argues that both codices – that contain a vocabulary and a grammar of the colonized languages – must be seen as a whole to form a colonial grammar. A study of language contact in colonial and postcolonial contexts, Kazuko Matsumoto’s paper examines “The role of domain and face-to-face contact in borrowing”. The author analyzes phenomena of borrowing in Palauan with respect to aspects of hierarchy, linguistic assimilation, and language contact of Palauan with Spanish, German, and Japanese. In his article “Representing bicultural knowledge systems and epistemic decolonization”, Norbert Schaffeld examines a song by a mixed Aboriginal and non-Aboriginal Australian band. The song is analyzed as intending to foster co-existence in a decolonized Australia, using mythological concepts and metaphors. Marina Wienberg’s article “Moanus or Titan? On Ludwig Cohn’s amateur linguistics” deals with linguistically interesting documents which are stored in archives or private collections. Numerous language descriptions, word lists and other linguistically furlednotes have remained unpublished to this day. This paper presents some linguist notes written by Ludwig Cohn, a typical amateur linguist of the colonial period, and his language-related ideas. We would like to thank this volume’s peer-reviewers as well as all of the participants of the First International Bremen Conference on Language and Literature in Colonial and Postcolonial Contexts who contributed to the ongoing consolidation of Colonial and Postcolonial Linguistics by taking part in

Preface | ix

stimulating discussions. We further thank Dr. Cornelia Stroh for the editorial assistance and great professional care with which she attended to this volume. Daniel Schmidt-Brücken, Susanne Schuster & Marina Wienberg Bremen, July 2015

Liesbeth Zack

Arabic language guides written for the British Army during the British occupation of Egypt, 1882–1922 Abstract: This paper deals with Arabic language guides created for the British army during the British occupation of Egypt (1882–1922), when officers were sent to Egypt and the Sudan to head the Anglo-Egyptian army. They needed to communicate with Egyptian soldiers in Egyptian Arabic, as well as correspond with officials in Standard Arabic. This situation was complicated by the fact that Turkish was also used in the Egyptian army. This paper explores five language guides which were written for the British army, comparing them on points of vocabulary, Arabic grammar, and their general usefulness for the British officers. Keywords: Egypt, British occupation, Arabic, language guides, army

1 Introduction1 During the British occupation of Egypt (1882–1922), British officers were sent to Egypt and the Sudan to head the Anglo-Egyptian army. There they were confronted with a complex language situation: the Arabic dialects were used in communication with the soldiers, Standard Arabic was used in written correspondence with Egyptian officials, and some Turkish military terminology was used. These officers were not given any official training in Arabic before setting off for the East, although they did have to pass an Arabic exam. Soon after the British took over Egypt’s administration, language guides aimed at these military personnel started to appear. This paper looks at five of these guides, describing the Arabic language as presented by them, and comparing the military terminology provided.

|| 1 I would like to thank Manfred Woidich, Caroline Roset, and Margreet Dorleijn for their comments on an earlier version of this paper. The research for this paper was done as part of a research project entitled “The making of a capital dialect: Language change in 19th century Cairo”, which was funded by a VENI grant from The Netherlands Organisation for Scientific Research (NWO). || Liesbeth Zack: University of Amsterdam, Faculty of Humanities, Arabic Language and Culture, Spuistraat 134, 1012 VB Amsterdam, Netherlands. E-mail: [email protected]

2 | Liesbeth Zack

2 Egypt and its language situation in the 19th century The language situation in the Arab world is one of diglossia, where there are two varieties of the same language: a high and a low variety. In the case of Arabic, there is an official language, Classical Arabic or, for the modern period, Modern Standard Arabic, which is based on the language of the Quran and is used for writing and formal speech, but which is nobody’s native language. The dialects are used in daily language situations such as informal conversation. When the term Egyptian Arabic is used in this paper, I intend the Arabic of Cairo, the capital of Egypt, unless otherwise specified. The language situation in Ottoman Egypt was complicated by the fact that the language of administration was Ottoman Turkish. This was no longer the case in the period under discussion in this article, because in 1858 Arabic was introduced as the official language of administration: a slow process which was not completed until the end of the 19th century (Toledano 2003: 158). However, the Turkish language still played a role in the army during the British occupation, as will be expanded upon below. The second half of the nineteenth century saw a great interest from foreign scholars in Egyptian Arabic, which was reflected in the publication of grammars, conversation guides, collections of stories and proverbs, etc. One of the reasons for this interest was the renewal of Egypt’s relations with Europe under Egypt’s ruler Khedive Ismail. One of the earliest grammars of Egyptian Arabic from this period is the Grammatik des arabischen Vulgärdialektes von Aegypten (1880) by the German scholar Wilhelm Spitta, who was director of Egypt’s National Library. This was a solid work which set a high standard for other works to come. However, there were also French, Italian, and of course English language guides and grammars.2 With the British occupation of Egypt, a whole new genre came into being: language guides for the army.

|| 2 A small sample: Finch-Hatton (1873), Haggenmacher (1892), Nallino (1900), Thimm (1898), Vollers (1890), Vollers & Burkitt (1895), Willmore (1901).

Arabic language guides written for the British Army | 3

3 The British in Egypt In the second half of the 19th century, Egypt’s rulers initiated numerous large projects in order to modernize and westernize Egypt’s infrastructure and architecture. A few examples are railroads, the Suez Canal, opened in 1869, and the building of new neighborhoods in Western style. This eventually caused severe financial problems for Egypt, when they had to default on loans from England and France. At the same time, discontent was growing in the Egyptian army. The higher ranks were reserved for the Turco-Circassian elite, while Arabic speaking Egyptians could only reach the rank of colonel. Displeasure with this system, as well as discontent about the growing European influence on Egypt, led to the revolt of the Egyptian army under the leadership of Colonel Ahmad Urabi in 1881 (Al-Sayyid Marsot 2007: 86). The British, responding to a call for assistance by the Khedive, decided to undertake measures in order to safeguard their financial assets. After bombarding Alexandria, defeating the Egyptian army, they assumed control of Egypt’s administration. The British occupation of Egypt ended officially with Egypt’s independence in 1922. However, British military presence lasted until 1956, when president Abd al-Nasser decided to nationalize the Suez Canal, which had been the last part of Egypt still in British hands. After the takeover of Egypt’s administration in 1882, British administrators, armed forces, and specialists such as medical doctors and engineers, moved to Egypt. These British governmental employees needed to be able to communicate with Egyptians, creating a need for Arabic language guides. These were written with different purposes and with different user groups in mind, such as the army, the court system, the railways, or simply residents and travelers. Because of the wide scope of these books, this paper will focus on only one type, namely the Arabic language guides written for British officers.

4 The language situation in the army Sir Evelyn Wood, appointed sirdār (Commander-in-Chief) of the Egyptian army in 1882, was given the task of reorganizing the army. High ranks were given only to British officers, while the lower officers and soldiers were recruited from among the Egyptian population (Baring 1908 II: 473–477). White (1899: 293) described the language situation in the Egyptian army as follows:

4 | Liesbeth Zack

British officers join the Egyptian army with the rank of bimbashi (major). As the medium of intercourse with their Native subordinates is Arabic – the words of command being given in Turkish – the first duty of an Anglo-Egyptian officer is to acquire a knowledge of that difficult language. Official correspondence between English officers is, of course, carried on in English; but, between them and the Native officers or non-commissioned officers, Arabic is employed. Orders are given out both in English and Arabic.

This shows the complexity of the language situation: conversation with Egyptians was in (Egyptian) Arabic; official correspondence with Egyptian officers was in (Modern Standard) Arabic, while the commands were given in Turkish.3 There was no official system in place for teaching officers Arabic, but they did have to pass examinations, as this account by Hulme-Beaman (1898: 72) reveals: At the present day there are hundreds of officers and others in Her Majesty’s Service having a very fluent acquaintance with the [Arabic] language, but at that period I was one of about half a dozen official Britons with a real “possession,” as the French call it, of both the written and spoken tongue. One of the rules of the new Egyptian Army was that the English officers should pass an Arabic examination, and I therefore offered to teach and prepare any of them for the test.

As this quotation shows, some officers had already learned Arabic elsewhere. Hulme-Beaman himself had learned Arabic in Constantinople from a Persian teacher, and during his subsequent stays in Beirut and Damascus. Horatio Kitchener, the later sirdār (1892–1899) of the Anglo-Egyptian army, had learned Arabic while on survey in Palestine in the 1870s (Jerrold [1916]: 21). Wingate, Kitchener’s successor as sirdār, had learned Arabic while stationed in India and Aden in 1881–1883, and passed the Higher Standard Arabic Examination in 1883 (Daly 1997: 9). However, when he arrived in Egypt, Wingate knew neither colloquial Arabic, nor the Turkish commands used in the army (Daly 1997: 17). All this led to confusing situations, as described by Scudamore (1925: 69): […] with the exception of Kitchener, who was already proficient, and Wingate – a linguist ingrain – none of [the officers] knew any Arabic. This indeed produced a curious anomaly in the earlier ceremonial parades of the new troops. There was a native Egyptian Brigadier, Schudi Pasha, who had been educated in Berlin, and it chanced that Lord Grenfell – then Major and Brigadier – knew some German. Thus Sir Evelyn had to give his orders, as Commander of the Forces, in the one language common to his two Brigadiers – i.e. Ger-

|| 3 Besant (1934: 162) mentions: “The words of command were Turkish. The only modification to this was the substitution of the word ‘march’ (which was pronounced ‘marsh’) for ‘yuru,’ the latter not being conducive to smartness of movement as a word of executive command”. Note that yuru should be yürü.

Arabic language guides written for the British Army | 5

man; Schudi afterwards giving his words in Arabic, Grenfell in English, and the four English commanders in Turkish […].

For soldiers of the British Army of Occupation,4 the situation was even more complex, as Lamothe (2011: 139) points out: Whereas British officers in the Egyptian Army studied colloquial Arabic, could speak French, and knew the Turkish words of command, soldiers in the British Infantry Division of the Anglo-Egyptian army – the Warwicks, Lincolns, Camerons, Grenadiers, and others – had no fluency in Arabic whatsoever.

A logical result of this situation was an established need for language guides to teach the British military Arabic, both written and spoken, and Turkish military terminology.

5 Language guides for the army This paper will discuss five language guides intended for use by the military. In order of date of publication:  Tien, Anton. 1882. Egyptian, Syrian, and North-African hand-book: A simple phrase-book in English and Arabic for the use of the armed forces and civilians. London: W. H. Allen & Co.  Green, Arthur Octavius. 1883–1885 (1909). Practical Arabic grammar for the use of English officers in Egypt. 2 vols. Vol 1 1883, vol. 2 4th ed. 1909 (1st ed. 1885). Cairo: Boulack Printing Office, Oxford: Clarendon Press.  Mosconas, Demetrius. 1884. English & Arabic dictionary accompanied by dialogues & useful notes for the use of the British Army of Occupation. Cairo: s.n.  Watson, C. M. [1885]. English-Arabic vocabulary and dialogues for the use of the Army and Navy. London: Her Majesty’s Stationery Office.  Scudamore, Frank. 1915. Arabic for our armies: Words and phrases with their equivalent in Colloquial Arabic (phonetic pronunciation) in daily requirement by H.M. Forces serving in the Near East. London: Forster Groom. There are other language guides in which military terminology is mentioned, for instance Thimm’s Self-Taught series, but the abovementioned five books were selected because they were written especially for the army.

|| 4 This numbered around 5,000 soldiers (Mak 2011: 18).

6 | Liesbeth Zack

5.1 Tien, Anton. 1882. Egyptian, Syrian, and North-African hand-book: A simple phrase-book in English and Arabic for the use of the armed forces and civilians The Reverend Anton Tien (1835–1920) was born into an ancient Lebanese Maronite family in Beirut. He converted to Anglicism and went to England to study theology at St. Augustine’s College in Canterbury. He served in the Crimean War, and after his ordination joined the Anglican mission in Constantinople.5 In 1879 he published a curious book, entitled The Levant interpreter: A polyglot dialogue book for English travellers in the Levant. Keeping in mind that the Levant encompasses modern Syria, Lebanon, Palestine, and Jordan, all of which are Arabic-speaking countries, it is surprising – to say the least – that this book contains conversation in Turkish, Greek, and Italian, but not in Arabic. However, because this book does not deal with either Egypt or the army, we will not further consider it here. In 1885, Tien published a Manual of Colloquial Arabic which, despite its title and the preface which promises to focus on the dialects of Syria and Egypt, predominantly discusses Standard Arabic grammar. Tien’s Egyptian, Syrian, and North-African hand-book is, as indicated in the title, a small book, but it is more than a phrase book, as it also contains an English-Arabic vocabulary. As the title indicates, it was intended for use in Syria, Egypt, and North-Africa. Anybody familiar with colloquial Arabic knows that these three dialect families differ greatly. Tien’s only comment on this subject is (Tien 1882, after title page): The great difference between the Syrian and Egyptian Arabic is that the latter hardens the soft consonants, e.g.: – Syria kawi jib shajarah

becomes " "

Egypt gawi gib sagarah

strong bring tree

He does not comment on any other differences except for this phonological remark. The pronunciation he gives here for Egypt is ambiguous: gawi is used in the countryside, as in Cairo this would be pronounced ʾawi.6 However, the g in

|| 5 See Boggis (1907: 195–196, 228) and Journal of the Royal Asiatic Society (1921: 487). 6 The transcriptions from the language guides are quoted as in the originals, but a standardized transcription is provided where necessary. This transcription follows the system of the Encyclopedia of Arabic Language and Linguistics (Versteegh et al. 2006 I: viii). A macron indi-

Arabic language guides written for the British Army | 7

gib and sagarah indicates the pronunciation of Cairo, as this sound would be pronounced j, ǵ, ž or even d in rural areas (Woidich 1996: 334). The book contains no further explanation of the Arabic pronunciation, except that “the pronunciation of the Arabic words in Roman letters is similar to that of Italian” (Tien 1882, after title page). Stress is not indicated. The transcription is not very accurate: for instance, Tien uses the letter p in several words, though p does not exist in Arabic. One example is “wait patiently – ospor” (Tien 1882: 59) which should be uṣbur. The voiceless pharyngeal fricative ḥ is transcribed as h, making no distinction between it and the voiceless glottal fricative h, while the voiced pharyngeal fricative ʿ is rendered with ’, e.g. “a soldier – ’askari” (Tien 1882: 53), which is also used for the glottal stop, thus creating non-existing homonyms, e.g. “ask – sa’al” (Tien 1882: 93) versus “to cough – sa’al” (Tien 1882: 33), the latter being saʿal in correct transcription. The Arabic is a mishmash of Syrian and Standard Arabic. For instance, the pronouns (Tien 1882: 1) are given in Standard Arabic, while the names of the months (Tien 1882: 4) are the Aramaic names of the Babylonian calendar, which are used only in the Levant but certainly not in Egypt or North Africa, where English- or French-derived names of the months are used. (1)

Tien (1882: 69) “kho [sic] qadah xud qadaḥ take.IMP.SG.M glass ‘take a glass of milk’

halîb” ḥalīb milk

“kho” is a typo and should read “khod”. “qadah” and “halîb” are both Levantine or Standard Arabic, the Egyptian equivalents being kubbāya and laban. (2)

Tien (1882: 80) “a’tiridoon an tashraboo a-ti-rid-ūn an ta-šrab-ū Q-2-want-IND.PL.M CNJ 2-drink-SBJV.PL.M ‘do you wish to have tea?’

ashshai” aš-šāy DEF-tea

The second example is mostly Standard Arabic, with the question particle a, and the distinction between the indicative in the first verb and the subjunctive

|| cates a long vowel; a dot under a consonant indicates that the consonant is emphatic, except for ḥ, which is the voiceless pharyngeal fricative ħ. ʾ indicates the glottal stop ʔ; ʿ indicates the voiced pharyngeal fricative ʕ. ġ is the voiced velar fricative ɣ. ḏ and ṯ are interdental fricatives (ð and θ). š is the voiceless fricative ʃ and j is the voiced palato-alveolar affricate ʤ.

8 | Liesbeth Zack

ruled by the conjunction an in the second verb, a distinction which no longer exists in the dialects. However, the ti- prefix in ti-rid-ūn is dialectal, as this would be tu- in Standard Arabic. The book is divided into three sections. The first section contains word lists, ordered by topic. These lists provide useful words on subjects such as the division of time, the senses, parts of the body, and accidents and diseases, as well as a section on “Military Profession” (Tien 1882: 52–55). The second section contains dialogues, while the third is an English-Arabic vocabulary. The dialogues, six in total, seem rather haphazard, and show a great pre-occupation with eating and sleeping, as these were given the following titles: “Going to bed and rising”, “breakfast”, “dinner”, “eating and drinking”, and “at an inn” (the first dialogue is not titled). The first dialogue contains some phrases for military use (Tien 1882: 62–64), as does the last one (Tien 1882: 87–88). Some sample sentences from the military sections: “I see foot-soldiers advancing – ara il-’askar yotakaddam”, “to what regiment do you belong? – min ayn alay int?” (Tien 1882: 63), “I am shot – taqawast” (Tien 1882: 64), “we assure you that no harm will happen to you if you surrender – nahaqiq lak in laisa ’alaik khatar iza sallamt” (Tien 1882: 87), all displaying the same kind of simplified Standard Arabic.

5.2 Green, Arthur Octavius. 1883–1885 (1909). Practical Arabic grammar for the use of English officers in Egypt Arthur Octavius Green (1847–1924) was sent to Egypt with the military expedition as an engineer in the summer of 1882. In 1884, he joined in an expedition to the Sudan, where he was severely wounded during the battle of El Teb. Subsequently he was promoted to the rank of Major (Porter 1889 II: 65–72), and at the end of his service he had reached the rank of Colonel. He died in 1924 in Camberley, England.7 He is also the author of A Practical Hindustānī Grammar, and of A Collection of Modern Arabic Stories, Ballads, Poems and Proverbs, for the Use of English Officers in Egypt.8 According to its preface, Green’s Practical Arabic Grammar was first distributed as loose sheets to “the Officers and Men of the Army of Occupation, to the English Officers serving in the Egyptian Army, the Gendarmerie, and the Police” – 150 copies in total, before being published as a two-volume book in 1883– || 7 See Western Daily Press (1924). 8 As the latter is merely a new edition of Spitta-Bey’s collected stories (1880 and 1883), it will not be discussed here.

Arabic language guides written for the British Army | 9

1885.9 The work seems to have been rather successful, as it went through four editions (the last reprint was published in 1915). The book was not based on original material, but rather was copied from several existing works such as those by Forbes, Shidyāq, Wahrmund, Wright, and Nakhlah (Green 1883: Preface). The works of Forbes, Shidyāq, and Wright deal with Classical Arabic, while Wahrmund’s contains Classical Arabic as well as the Syrian and Egyptian dialects, and Nakhlah’s work is a conversation manual of Egyptian Arabic. Most of these works were themselves based on other, older works. The first part of Green’s book deals with grammar and introduces with the Arabic alphabet. The grammar is explained using Arabic script and transcription, while the exercises are only in Arabic script. The transcription is accurate, and indicates the ḥ and the emphatic consonants with dots underneath. The exercises consist of Arabic-English and English-Arabic translations. The keys to these exercises can be found in volume 2, which also contains “Selections from Arabic Authors and Newspapers”, “English and Arabic Letters and Manuscripts”, EnglishArabic and Arabic-English vocabularies, and a table comparing “Classical and Modern Arabic Forms and Expressions”. The author often shows Egyptian and Standard Arabic side by side, e.g. “hâza el-bait lak, el-bait da b’tâ‘ak this house belongs to you” (Green 1883: 36), which does not indicate that the first part is Standard Arabic and the second is dialect, or “‫ نحن‬naḥnoo, nahn vulgar. ‫ احنا‬eḥna we” (Green: 1883: 20), which does point out the difference between the two varieties. Green’s sample sentences are sometimes a mixture of both varieties, as the following examples illustrate. Colloquial features are underlined, and Standard Arabic ones are in boldface. The parts not underlined or bold faced are neutral: (3)

(4)

Green (1883: 22) “hâzee el-bint hāzī el-bint DEM.SG.F DEF-girl ‘this girl is pretty’

kwyeeseh” kuwayyis-eh pretty-SG.F

Green (1883: 86) “heeyeh gâliseh hiyyeh gālis-eh she sit.PTCP-SG.F ‘she sits near me’

bikoorbee” bi-qurb-ī at-vicinity-POSS.1SG

|| 9 The 1909 edition of vol. 2 will be used here, as the 1885 edition was not available to the author.

10 | Liesbeth Zack

In example 3, the demonstrative is Standard Arabic (Egyptian Arabic would be postposed di, as Green explains on p. 23), but the z instead of interdental ḏ is the Egyptian Arabic pronunciation. The vocabulary item kuwayyiseh is dialect as well. The pronunciation of the feminine ending -a as -eh (called “final imāla”) in kuwayyis-eh and gālis-eh is typical for colloquial Arabic. An interesting example is gāliseh in example 4: as a vocabulary item it is Standard Arabic, as is its form: CāCiCa(eh) would have become CaCCa(eh) in Egyptian Arabic. However, the pronunciation of g instead of j is Egyptian Arabic, as is the pronunciation of the feminine ending -eh. Although the title indicates that the book is meant for English officers in Egypt, the contents are for the most part intended for general use and do not contain specific army-related information. An exception is the section “English and Arabic Letters and Manuscripts” in part 2, for this contains letters dealing with army matters. The vocabulary in part 2 also contains army-related terminology. The letters are handwritten in different hands in Arabic and a transcription and translation are provided. The letters are written in Standard Arabic; however, the vocalization used in the transcriptions is mostly Egyptian Arabic, and the declensions are not transcribed.10 This is possible because the Arabic script is defective, meaning that short vowels are not written. Therefore, the text can be written in Standard Arabic but read using Egyptian Arabic pronunciation. (5)

(Green 1909: 92–93)

‫نتشرّف بان نخبر حضرتكم أنه قد صار تسفير الفين عسكري انكليزي من ھنا‬ “natasharraf b’inn nukhbir ḥaḍritkum innoh qadd ṣār tasfīr alfēn ʿaskarī inklīzī min hena” ‘It is an honor for us to inform you that 2000 English soldiers have travelled from here’. A transcription following Standard Arabic pronunciation would read: natašarrafu biʾan nuxbira ḥaḍratakum ʾannahu qad ṣāra tasfīru alfayni ʿaskariyyin inkilīziyyin min hunā. All considered, Green’s work was useful for people who wanted to learn Arabic, although sometimes it was confusing due to his mixture of Standard Arabic and dialect, probably caused by the variety of sources used for the work.

|| 10 Standard Arabic has declensions, while Egyptian Arabic has not.

Arabic language guides written for the British Army | 11

5.3 Mosconas, Demetrius. 1884. English & Arabic dictionary accompanied by dialogues & useful notes for the use of the British Army of Occupation Demetrius Mosconas (1839–1895) was a Greek Orientalist who came to Egypt as a young man. He was a student of the famous German Egyptologist Heinrich Brugsch.11 Mosconas worked for some time as an interpreter for Thomas Cook.12 In 1880, he is mentioned as consul of the United States at Suakim (a port in the North Eastern Sudan),13 while in 1883 he was living in Kassala in the Sudan, sinking wells for the Egyptian government (Williams 1884: 148, 151). On the title page of his English & Arabic Dictionary Mosconas calls himself “Interpreter to the Commissariat and Transport Staff”. In 1893, he was responsible for the reproduction of the Temple of Luxor at the World’s Columbian Exposition in Chicago.14 The book is dedicated to the British army: To those brave officers, non-commissioned officers & men of the British Army who took part in the Egyptian campaign of 1882 and the Soudan expedition of 1884 this book is dedicated as a slight mark of esteem for and appreciation of their services to Egypt in general and Cairo in particular in the hope that those who hesitated not to shed their blood in the cause of law and order will now take up and study the language of the people with whom they sojourn.

In the preface, the author briefly explains the language situation in Egypt and states the purpose of the work to be “the instruction of the British Army of occupation in order that the Soldiers composing it, may make themselves properly understood amongst its people.” Mosconas’ book, like Tien’s, pays no attention to typical Arabic consonants such as the ḥ and ʿ, or the emphatic consonants ṭ, ḍ, ṣ, and ẓ. When learners of Arabic would have used the pronunciation presented in these books, their Arabic would have been hard to understand for Arabic speakers. The lack of indications of stress must have made it even harder to get the pronunciation right. The first half of the book consists of three parts: parts one and three contain general information, part one on grammatical rules, and part three on “familiar phrases and conversation”. Part two is dedicated to military terminology and is

|| 11 Brugsch was director of the Cairo School of Egyptology from 1870 to 1879 (see Encyclopaedia Britannica). 12 See Carstens (2014: 486). 13 See Bonola (1880: 86). 14 See Delamaire (2003: 133) and Campbell’s Illustrated Weekly (1893: 128).

12 | Liesbeth Zack

entitled “Reconnaissance questions”. This is divided into eight sections, entitled “The road”, “a mountain”, “Ford Ferry”, “Bridge”, “Marsh”, “Fortress”, “The enemy”, and “Foragin [sic]”. The second half of the book is an English– Arabic dictionary. The grammar in part one gives a short explanation of Egyptian Arabic grammar, focusing mostly on verbs. The Arabic used in it is correct, as is that of the third part with the familiar phrases. However, the second part, containing the “Reconnaissance questions”, contains many grammatical errors and incongruities. Although at first this discrepancy in quality seems baffling, it is easily explained, because the grammar and familiar phrases are copied almost verbatim from Yacoub Nakhlah’s New Manual of English and Arabic Conversation (1874).15 As Nakhlah was a Coptic Egyptian and therefore a native speaker of Arabic, it is not surprising that the parts copied from his manual contain correct Arabic. However, Mosconas’ own Arabic was apparently far from perfect, shown by the examples from the “Reconnaissance questions”, which he could not have copied from Nakhlah and therefore would have had to have produced himself:16 (6)

(Mosconas 1884: xxvii): as-sikket di yirôohh ala as-sikkit di yi-rūḥ ʿala DEF-road DEM.SG.F 3SG.M-go to ‘where does this road lead to?’

fain fēn where

There are two problems here: first, the ending -it in the word as-sikkit which can only be used in the construct state; it should be as-sikka here. The second problem is the lack of agreement between the subject and the verb: the verb has the masculine prefix yi- instead of the feminine prefix ti-. Lack of agreement is a common problem in this book; another example is: (7)

(Mosconas 1884: xxx): el-kantârah tayib el-qanṭara ṭayyib DEF-bridge good.SG.M ‘Is the bridge good?’

|| 15 In the Preface, Mosconas declares that he has extracted “the most perfect method of the grammatical rules similar to those published by Mr Yacoub Nakhleh”. “Similar” seems somewhat of an understatement here. 16 The Dictionary is also for a large part copied from Nakhlah.

Arabic language guides written for the British Army | 13

The word qanṭara is feminine, as can be seen from its ending -a, and therefore the adjective should also be feminine: ṭayyiba. The article is also used incorrectly: (8)

(Mosconas 1884: xxxi): ez-zabit-hoom min-huwa eẓ-ẓabiṭ-hum mīn huwwa DEF-commander-POSS.3PL who he ‘what is the name of their commander?’

The article cannot be used in combination with the possessive suffix and should have been omitted. The verbs are especially interesting. In the imperfect, some of the verbs have the extra -u suffix in the first person plural, which is characteristic for the Western Arabic dialects of North Africa, while others do not have this suffix. The dialect of Cairo, like most of the other Egyptian dialects, belongs to the Eastern type. Tab. 1: Affixes of the 1st person imperfect in Eastern and Western Arabic

I write we write

Eastern Arabic

Western Arabic

a-ktib ni-ktib

ni-ktib ni-ktib-u17

The prefix ni- is ambiguous, as can be seen in the table, and can mean ‘I’ or ‘we’ depending on the dialect. In Mosconas’ work, the two varieties are mixed for the first person plural form, sometimes displaying the Eastern type, sometimes the Western type, even within one sentence: (9)

(10)

(Mosconas 1884: xxviii) “negdar nem-shoo ni-gdar ni-mš-u 1PL-can-ø 1PL-walk-PL ‘Can we go on foot?’

ala ʿala on

riglaina” riglē-na feet-POSS.1PL

(Mosconas 1884: xxix) “nikdarôosh ne’addy ala riglaina” ni-gdar-ū-š ni-ʿaddy ʿala riglē-na 1PL-can-PL-Q 1PL-cross-ø on feet-POSS.1PL ‘Can it be crossed on foot?’ [literally: ‘can we cross it on foot?’]

|| 17 The vowel of the n-prefix can vary from dialect to dialect.

14 | Liesbeth Zack

Note also the incongruity in the spelling of “negdar” and “nikdar” (nigdar). The 1SG form is always ni-, never a- in Mosconas (1884): (11)

(Mosconas 1884: xxvii): “nâraf nitkellim, ni-ʿraf ni-tkallim 1SG-can 1SG-speak ‘I speak very little’.

shoowâya” šuwayya a bit

It is unclear on which dialect Mosconas based the Reconnaissance part of the book. He lived in at least three different Arabic speaking regions: Alexandria (Delamaire 2003: 133); Cairo, where he studied; and Eastern Sudan. In the latter two, the aktib-niktib paradigm is used,18 while Alexandria used to have niktibniktibu (nowadays almost disappeared in favor of the forms used in Cairo) (Behnstedt 1980: 37–38). It is possible that this difference of dialectal input might have confused Mosconas, leading to the mixed information given in the book. Yet other characteristics point in the direction of Egyptian Arabic, e.g. “eh”, “ey” ē ‘what’ (Mosconas 1884: xxviii), where Sudanese has šinu, šīn,19 and the negation using ma-…š in “manish-fà-hem” ma-nīš fāhim ‘I do not understand’ (Mosconas 1884: xxvii). As the whole Reconnaissance part, the only original part of the work, is just six pages out of a book of more than a hundred, and is of dubious quality, it can be concluded that Mosconas’ work was not much of an addition to the existing collection of Egyptian Arabic text books.

5.4 Watson, C. M. 1885. English-Arabic vocabulary and dialogues for the use of the Army and Navy Sir Charles Moore Watson (1844–1916) joined the Royal Engineers in 1866 and was sent to Egypt in 1882. He assisted Sir Evelyn Wood in creating the new Egyptian army and was acting sirdār for some time. In 1886 he acted as Governor General of the Red Sea Littoral in Suakim (King 1916: 387–388). Watson had already joined General Gordon for a survey in the Sudan in 1874, and started learning Arabic there. Back in England, he continued to work on his Arabic and took lessons with Rizq Allāh Ḥassūn (Lane-Poole 1919: 71). Ḥassūn (1825–1880) was a native of Aleppo and lived some time in Russia and || 18 For Sudan, see Bergman (2002: 23), and for Egypt, see Woidich (1996: 338). 19 See Amery (1905: 396) and Bergman (2002: 377).

Arabic language guides written for the British Army | 15

England (Al-Ziriklī 2002 III: 19). Watson made good use of the Arabic he had learned and began instructing his fellow passengers while on board the ship en route to Egypt: I gave a lecture on the grammar and pronunciation this morning of over an hour to a large party [….]. The idea of poor me coming out in this line is rather amusing, but, as you know, “among the blind the one-eyed man is a prophet.20

After arriving in Egypt, his linguistic skills remained in demand, as he remarks in a letter to his wife: “I have become a regular dragoman!”21 It was therefore a logical decision to publish a work on the Arabic language. The English-Arabic Vocabulary and Dialogues starts with the Arabic alphabet and pronunciation of the consonants and vowels, followed by a short grammar (8 pages) dealing with the article, noun, adjective, pronoun, and verb. Short remarks are made about the difference between Standard Arabic and dialect. The pronouns (personal and possessive) mentioned are Standard Arabic (without dual and feminine plural, which exist in the standard language but not in Egyptian Arabic). The verbs are also in a simplified Standard Arabic (e.g. nasartu ‘I helped’, with the -tu suffix of Standard Arabic, as opposed to the -t suffix in Egyptian Arabic), not including the plural feminine forms and the dual forms. A note on page 17 mentions that “the Egyptian dialect has been adhered to in this vocabulary”. The second part of the work is the Vocabulary, which has word lists on general topics such as the numbers, the days of the weeks, etc., but also armyrelated vocabulary such as “Divisions of Troops” and “Arms and their Accessories,– Fortification Terms, &c.”. The dialect represented is not that of Cairo, as can be demonstrated in the two words for ‘very’: “kawi” and “jiddan” (p. 25). In “kawi”, the *q is written with k, perhaps representing the g of the countryside; it is pronounced ʾawi in Cairo. In “jiddan”, the *j is written with j (giddan in Cairo). Also “how? – kayf?” (p. 25) is rural, as Cairo has izzāy. The months (p. 30) are the Coptic ones, which also points to the rural background of Watson’s Arabic knowledge: the Coptic months were (and still are) only used in matters related to farming (e.g. when to harvest), whereas Western style months (yanāyir, fibrāyir, etc.) were used in daily life in the towns. The third part, Dialogues, starts with “Useful Sentences”, followed by “Reconnaissance Questions”, and other dialogues useful for the army, such as “Foraging”, and “On Patrol”, as well as some vocabulary relating to purchases,

|| 20 Letter from Watson to his wife, dated 8 August 1882, quoted in Lane-Poole (1919: 106–107). 21 Letter from Watson to his wife, dated 22 September 1882, quoted in Lane-Poole (1919: 131).

16 | Liesbeth Zack

etc. Watson’s “Reconnaissance questions” have been taken from Mosconas’ “Reconnaissance questions”, with some changes and additions.22 Both men lived in Suakim for some time, though we do not know exactly where Mosconas was in 1883, when Watson was residing in Suakim as governor. However, it is possible that they knew each other; or perhaps Watson had picked up Mosconas’ book while in Suakim. In any case, Mosconas’ influence on Watson becomes clear when comparing the two books: Tab. 2: Comparing part of the “Reconnaissance questions” of Mosconas and Watson Mosconas p. xxviii

Watson p. 60

what is the name of this mountain? is it steep? wooded?

esmoo-ey-el-gâbal di

is it high; low?

huwa âly, waty

huwa mâyil, fih ashgâr

is there a defile? what is it called? can a cart pass?

fih khanik esmoo-eh el-’arabieh y’akdar yeroôh ’alaih whither does it lead? et-tareêk yewoddi fayn is there any fort here? fih kala’a henâk can we go round yemkinna ni tajànnab (avoid) it? minnoo

what is the name of this mountain? is it steep? " wooded? " high? " low? is there a defile? what is it called? can a cart pass?

al jabal di ismuh ey?

huwa māyil? fih ashjār? huwa ’ali? " watī? fih chānik? ismuh ey? ’arabiah yakdir yaruh ’alaih? whither does it lead? at tarik di yawadi ila fain? is there any fort there? fih hisn henak? can we go round mumkin an natajannab (avoid) it? minhu?

The most obvious difference between the two texts in the table is the way the Arabic is transcribed. However, linguistically, the only substantial difference is the word for ‘fort’: “kala’a” qalʿa in Mosconas, “hisn” ḥiṣn in Watson. In the first sentence, Watson inverted the word order (both word orders are correct), while in the last sentence, Watson uses the Standard Arabic conjunction “an”. Like Mosconas, Watson mixes up Eastern and Western Arabic patterns of the verb for the first person: (12)

(Watson 1885: 56): “ana ārid” ana a-rīd I 1SG-want ‘I want’.

|| 22 E.g. Watson has added a section entitled “Examination of prisoners”.

Arabic language guides written for the British Army | 17

(13)

(Watson 1885: 57): “mā nakdirsh” mā na-gdir-š NEG 1SG-can-NEG ‘I cannot’.

As we see in these two examples, the 1st person singular is either indicated with a- or na-. The interesting thing, however, is that in the chapter “Arabic Grammar” Watson gives the paradigm ansur “I help” and nansur “we help” (p. 15) and thus, as Mosconas, does not follow his own grammar rules in the “Dialogues” section of his book. Watson uses na- to indicate the 1st person plural. However, the 1st person plural form na-…-u is not found in Watson’s work (see also Nakao 2013: 5–6): (9)

(14)

again (Mosconas 1884: xxviii): “negdar nem-shoo ala ni-gdar ni-mš-u ʿala 1PL-can-ø 1PL-walk-PL on ‘Can we go on foot?’

riglaina” riglē-na feet-POSS.1PL

(Watson 1885: 60): “nakdur namshi ’ala na-gdur na-mši ʿala 1PL-can-ø 1PL-walk-ø on ‘Can we go on foot?’

riglaina” riglē-na feet-POSS.1PL

Again as in Mosconas, there are issues with subject-verb agreement: (15)

(Watson 1885: 59): “’arabiah yakdir an ʿarabiyya ya-gdir an cart 3SG.M-can CNJ ‘Is it difficult for a cart?’

tamshi ’ala ta-mši ʿala 3SG.F-go on

at aṭ DEF

tarīk di?” ṭarīq di? road this?

The word ʿarabiyya is feminine, but the first verb is masculine, while the second is feminine. Note also the very free translation (literally it says ‘is it possible for a cart to go on this road?’). The conjunction an is only used in Standard Arabic, not in the dialects. The source text of Mosconas contains this same sentence twice, once correct, once incorrect (Mosconas 1884: xxviii): el-a’arabiyeh tegdar tim-shi (correct, both verbs have the feminine prefix t-) and el-‘arabieh y’akdar yeroôh ’alaih (incorrect, both verbs with masculine prefix y-). Perhaps Watson was confused by this and confused Mosconas’ two sentences.

18 | Liesbeth Zack

It can be concluded that for the Grammar part, Watson probably relied on the Standard Arabic he learned in England, while the rest of the book contains rural Egyptian Arabic. It is unfortunate that Watson did not cast a critical eye on the Vocabulary, which contains many errors. This can probably be explained by him having learned Standard Arabic from a teacher, while he had probably picked up the dialect in the streets of Egypt, without any formal instruction.

5.5 Scudamore, Frank. 1915. Arabic for our armies: Words and phrases with their equivalent in colloquial Arabic (phonetic pronunciation) in daily requirement by H.M. forces serving in the Near East Frank Scudamore (1859–1939) was the son of Frank Ives Scudamore (1823– 1884), the Post Office administrator responsible for nationalizing the British telegraph industry. In 1875, when Scudamore Jr. was 16 years old, his father was appointed by the Ottoman government to reorganize the Turkish post office (Perry 2004). The family spent some time in Constantinople. Scudamore reported his first war, the Russo-Turkish war, in 1877 (Roth 1997: 277), and in 1882 went to Egypt as a war correspondent (Scudamore 1925). Scudamore’s Arabic for our Armies is a booklet of 24 pages containing word lists on the subjects of “food and drinks, general wants, travel by river or sea, travel by land, Red Cross, camp terms and phrases, sentences and words, and numerals”.23 The Foreword (p. 3) starts with the interesting statement that “There are several forms of the Arabic language, and of them ‘Egyptian Arabic’ has been termed ‘bastard’”. Scudamore does not explain why Egyptian Arabic deserves this dubious honor. It is possible that “bastard” refers to the lack of respect felt for the dialects as opposed to Standard Arabic, but this is the case for all Arabic dialects, not exclusively Egyptian Arabic. Scudamore (1915a: 3) continues to say that The matter of pronunciation is extremely difficult. There are many complicated vowel 24 sounds, which cannot be rendered by any possible combination of English letters, and such endeavour to explain them as could be given in a tiny manual would confuse without helping the reader. All that is attempted here is to convey the sounds of Arabic words as nearly as possible […].

|| 23 Scudamore wrote similar booklets about Turkish, French and German, see the References. 24 He obviously means consonant, because the vowels in Arabic are rather straightforward, whereas the consonants are a whole different matter.

Arabic language guides written for the British Army | 19

This overview shows that Scudamore did not follow any logical system in reproducing Arabic sounds, let alone convey them as nearly as possible:  ʿ : this is rendered in a number of ways, none of them very accurate, e.g.:  e: “veal – e-idjl” [ʿijl] (p. 7),  dg: “make broth – udg-mel-marakă” [iʿmil maraqa] (p. 7) ,  zoŏ (?): “grapes – zoŏ-nab” [ʿinab] (p. 7),  y: “how far – kām bay-id” [kām baʿīd, grammatically also incorrect] (p. 15),  h: “forty – arr-bah-een” [arbiʿīn] (p. 24),  ø: “I want – eyes” [ʿāyiz] (p. 16).25  ḥ:  gh: “apple – tō-fagh” [tuffāḥ] (p. 8),  ø: “yesterday – embarra” [imbāriḥ] (p. 22),  k: “open the door – if tak26 el bab” [iftaḥ el bāb] (p. 20).  x:  h: “store house – māh zān” [maxzan] (p. 18),  kh: “chicken – farkh” [farx] (p. 7). It is unfortunate that he sometimes uses kh but leaves a space, as if kh is supposed to be pronounced as two separate letters: “a pillow – muk hād dā” [muxadda] (p. 9), “tobacco – dook hān” [duxān] (p. 10),  k: “thread, twine – kait” [xayṭ] (p. 9).27  j:  j: “kid – jad-ee” [jady] (p. 7),  dj: “auxiliaries – nădj-dā” [najda] (p. 18).  q, pronounced in Cairo as a glottal stop and in most of the rest of Egypt as g:28  c: “a lamp – can-deel” [qandīl] (p. 8),  k: “a water carrier – săk-kăh” [saqqa] (p. 9),  kh: “a spoon – mih-lākhā” [miʿlaqa] (p. 9),  g: “writing paper – wārrăg” [waraq] (p. 9).  p: does not exist in Arabic and is pronounced as b in loan words, but is used by Scudamore: “orange – portoo-gal” [burtugāl] (p. 8). || 25 Scudamore has a habit of using the spelling of familiar English words to convey the sounds of the Arabic, for instance “alive – high” [ḥayy] (Scudamore 1915a: 17), “dinner – rudder” [ġada] (Scudamore 1915a: 20). 26 Notice the same word with a completely different transcription on p. 21: “open! – efftăh!”. 27 The same word, but translated as “string”, is given on p. 10 with the spelling “khā-ĭt”. 28 The following examples seem to indicate that Scudamore was not describing the dialect of Cairo. However, this is contradicted by the word “ā-we” in “cold (very) – bard (ā-we)” [bard ʾawi] (p. 19), which seems to indicate the glottal stop for the *q, like in Cairo.

20 | Liesbeth Zack

The indication of long and short vowels seems to be random. The incorrect use of spaces makes clear that Scudamore did not know the word boundaries. Often he uses spaces within words, e.g. “battle – mō har eb bey” [muḥāraba] (p. 18), but also spaces between words are given in the wrong places, e.g. “give me (something) to drink – gib lĕ shish rŭb” [gib li ši (a)šrab] (p. 19) where the š of the root ŠRB ‘to drink’ is attached to the previous word. Some words are transcribed in such far-fetched ways that it is almost impossible to recognize their Arabic counterparts: “roast – mack-swee” (p. 7) (mašwi?), “cheese – jā-boom” (p. 7) (jubn?), “artichokes – hark-ssoof” [xaršūf] (p. 8), “flour – dāgh-n” [ṭaḥīn?] (p. 8), “arm – sāthād” [sāʿid?] (p. 16), “bind! – hāss-ūb!” [ḥazzim?] (p. 16). Typos are also common, e.g. “a glass – koo-bāgā” [kubbāya] (p. 9), “fog – shab-ooza” [šabbūra] (p. 12), “the doctor – tā-beel” [ṭabīb] (p. 15). Incorrect translations and grammar are rife as well, e.g. “a horse – high wann (or) huss-an” [ḥayawān or ḥuṣān] (p. 13), ḥayawān meaning ‘animal’, not ‘horse’; “being without anything – sā kit” [sākit] (p. 22), sākit means ‘silent’, not ‘poor’; “to dig – očh-for; dig (command) – očh!” (p. 19): the imperative ‘dig!’ is uḥfur, no infinitive exists in Arabic. The -fur in uḥfur can in no way be deleted from the rest of the word, as Scudamore suggests here. It is clear that the many problems of this booklet, such as faulty translations, incorrect grammar, and botched transcriptions, cannot have done anything to improve communication between the British and the Egyptians, and must even have complicated it further.

6 Comparison of some military terminology This paragraph compares a sample of 18 terms from the five works. Most of these are military ranks, while others are words commonly used in the army. The term is followed by the page number. Tab. 3: Comparison of some military terminology English term Tien artillery barracks battalion brigade

Green vol. 2

tobjieh, ṭobgīyah 118 madf’ajieh 53 kishlah 53 qishlah, qishlāq 119 tabour 52 ōrṭah 119 liwa 52 liwā 122

Mosconas

Watson

Scudamore

tupgi, tupgieh 17 keeshlàk 22

tupjī 43

top-jee-ya 17

kishlah 44

kāz-lă 18

taboòr 23 –

ordā tabur 43 liwa 43

boo-look 17 lee-wā 17

Arabic language guides written for the British Army | 21

English term Tien captain29 colonel garrison30 general

infantry lieutenant major31 regiment rifle sergeant sergeantmajor soldier squadron staff

Green vol. 2

Mosconas

yoozbashi 52

qabṭān, kaptàn 41 yūzbāshī 123 mir-alaiy 52 mīr ālāi 125 miralay 48 ’oordi 53 ḥāmiyah 143 asàkeer-elmahàfezah 86 sari ’askar 52 mushīr, fereék 87 sarʿaskar 143

biyadeh’askar 53 milazim 52 binbashi 52 allaiy 52 shishkhanah ’askar 53 shawish 52 bash-shawish 52

Watson

Scudamore

yuzbashi 42

youz-bashi 18

miralai 42 urdī (camp) 46

meer allie 18 sĭch na, rā-bee-tā 18 emeer 18

biyādah 148 biyàdah 107

pasha 42, pasha ferik (general, lieutenant), pasha liwa (general, major) 43 biyada 44

mulāzim 153 binbāsha 155 orṭah; alāi 174 –

mulāzim 43 bimbashi 43 alai 44 sheshchanah 45

shāwīsh 178 bāsh shāwīsh 178 ’askari 53 ʿaskarī 181 boolook 52 ōrṭat sawārī 182 arkani harb 52 arkān ḥarb 182

moolazim 120 bembâshe 127 ’â’lay 177 sheshkhanah 180 – shawushbashi 191 asscari 203 bûlook 207 arkanel-harb207

pee-yādā 17 – bimbashi 18 allie 17 –

– – chawush bashi 43 – ’askari 42 āsker (army) 17 buluk 44 ta-boor 18 arkhan al harb 44 –

The various works agree reasonably well on a number of terms, such as ‘barracks’, where all mention a form of the (Ottoman) ‫ قشال‬qišla or ‫ قشالغ‬qišlāġ (see Redhouse 1880: 49a), or some ranks such as ‘lieutenant’ mulāzim (Arabic), ‘captain’ yuzbāši (Ottoman, from yüz ‘100’ and baš ‘chief’), and ‘colonel’ miralay (Ottoman, from mir, abbreviation of amīr ‘commander’, and alay ‘regiment’). The difference is mostly in the transcription used or in the word order (e.g. ‘sergeant-major’: baš-šawīš versus šawuš-bāša, from Ottoman baš ‘chief’ and çawuš ‘sergeant’). In ‘artillery’, Tien is the only one who mentions the Arabic “madf’ajieh” madfaʿiyya besides the Ottoman ṭubgiyya (topçu). More interesting are the words on which the various works differ considerably. These terms, for ‘battalion’, ‘garrison’, ‘general’, and ‘squadron’, will be briefly discussed here:

|| 29 Translations of the type qabṭān (Green, Mosconas) refer to the naval term, not the army rank. 30 Either ‘military post’ or the troops stationed there. 31 Major (bimbāši) was the lowest rank with which an English officer entered the Egyptian army. In the Sudan, where many native soldiers did not speak Arabic as their mother tongue, the British used a kind of pidginized Arabic called ‘Bimbashi Arabic’, which according to Nakao (2013) formed the basis for Arabic creole languages such as Juba Arabic.

22 | Liesbeth Zack

 ‘battalion’ and ‘squadron’: The words used by the five works for ‘battalion’ and ‘squadron’ all go back to three terms: bulūk, ṭabūr, and orṭa. These were originally terms used in the Ottoman Empire by the Janissary corps. When Sultan Selim III established the new army in 1794 as part of the Nizam-i Cedid (‘New System’), an orṭa consisted of 1602 men, divided in two ṭabūrs of 800 men each, which were again divided into smaller companies (bölük in Ottoman) of 100 men (Shaw 1965/1966: 176). The word sawārī mentioned by Green means ‘cavalry’.  ‘garrison’: In the word ‘garrison’, the difference in the translations can be partially explained by the meaning of the word in English. A garrison can either be a military post or the troops stationed there. Mosconas took the meaning of ‘troops’: his translation “asàkeer-el-mahàfezah” ʿasākir almuḥāfaẓa correlates to ‘soldiers of protection’. Green uses another Arabic word for protection, “ḥāmiyah”. Tien and Watson use the Turkish word for ‘camp’, ‫ اردو‬ordu (Redhouse 1880: 60). The first term given by Scudamore, “sĭch na”, is probably an incorrect representation of the word ‫ شحنة‬šiḥna ‘troop, garrison’ (see e.g. Steingass 1884: 531). The second term given by Scudamore is the Arabic ‫ رابطة‬rābiṭa ‘band; bond; confederation’.  ‘general’: For the term ‘general’, Tien and Green use the translation sar ʿaskar which is a mix of the Persian sar ‘head’ and the Arabic ʿaskar ‘soldiers’. The sar ʿaskar was the commander in-chief and minister of war of the Ottoman Empire (see Lewis). However, during the period under discussion, the function of commander-in-chief was called sirdār in Egypt,32 not sar ʿaskar. Green also uses the Arabic mušīr, which was used in the Ottoman army with the meaning of “Field Marshal”, the highest rank in the army (see Deny). Mosconas uses the Arabic farīq, which means ‘lieutenant general’.33 Scudamore uses amīr ‘commander’, which was not an official rank in the army. Watson gives pasha, which is incorrect, as pasha was an honorific title accorded to the commander-in-chief and generals, but not a rank.34 It is clear that most of the differences in terminology are caused by a choice the authors made from the available Ottoman and Arabic terms. Some confusion about the interpretation of the English terms played a role as well.

|| 32 See Lamothe (2011: xv), Amery (1905: 416), and Baring (1908 II: 283). 33 See Lamothe (2011: xv) and Wingate (1891: 209). Amery (1905: 146) gives ‘General of Division’, which coincides with Mosconas’ ‘general’. 34 See Lamothe (2011: xv) and Wingate (1891: 209–213). The latter gives an excellent overview of the composition of the new Egyptian army and an explanation of the origin of the ranks.

Arabic language guides written for the British Army | 23

7 Conclusion The complicated linguistic situation in Egypt during the British occupation is reflected in the language teaching materials that were available for British officers. These officers had to deal with an amalgam of Standard Arabic, Egyptian Arabic, and Turkish, with no formal system of teaching these languages in place at the time. The quality of the five books discussed in this paper varies greatly. This has to do with the different backgrounds of the authors, the variety of Arabic they had learned, their teachers and informants, and their (lack of) proficiency in the language. This makes them unreliable sources of information for dialectologists who are interested in the Arabic dialects of that time, although it does make them interesting from a historical and sociolinguistic point of view. The purpose of the works also differs: on one end of the spectrum we find a complete course of Arabic grammar, including Arabic script, intended to give a thorough knowledge of written and spoken Arabic (Green); on the other end, we have a simple word list containing basic vocabulary, intended for providing the bare minimum of Arabic needed for dealing with everyday situations (Scudamore). The military terminology presented in the five works is only partly standardized, as has been shown with a sample of terms. This has to do with several factors: the choice of either the Ottoman Turkish or the Arabic terms; the confusion about the exact meaning of English ranks and terms; and the reorganization of the Egyptian army which caused some terms to disappear.

Abbreviations CNJ DEF DEM F IMP IND M

conjunction definite demonstrative feminine imperative indicative masculine

NEG PL POSS PTCP Q SBJV SG

negation plural possessive participle question particle subjunctive singular

24 | Liesbeth Zack

References Al-Sayyid Marsot, Afaf Lutfi. 2007. A history of Egypt: From the Arab conquest to the present. Cambridge: Cambridge University Press. Al-Ziriklī, Khayr al-Dīn. 2002. al-Aʿlām: qāmūs tarājim li-ašhar al-rijāl wa-l-nisāʾ min al-ʿArab wa-l-mustaʿribīn wa-l-mustašriqīn. 15th ed. Beirut: Dār al-ʿilm li-l-malāyīn. Amery, H. F. S. 1905. English-Arabic vocabulary for the use of officials in the Anglo-Egyptian Sudan compiled in the Intelligence Department of the Egyptian Army. Cairo: s.n. Baring, Evelyn (Earl of Cromer). 1908. Modern Egypt. 2 vols. London: Macmillan. Behnstedt, Peter. 1980. Zum ursprünglichen Dialekt von Alexandria. Zeitschrift der Deutschen Morgenländischen Gesellschaft 130. 35–50. Bergman, Elizabeth M. 2002. Spoken Sudanese Arabic: Grammar, dialogues, and glossary. Springfield: Dunwoody Press. Besant, W. H. 1934. The early days of the Egyptian Army, 1883–1892. Journal of the Royal African Society 33,131. 160–168. Boggis, Robert James Edmund. 1907. A history of St. Augustine’s College, Canterbury. Canterbury: Cross & Jackman. Bonola, F. 1880. Compte-rendu des séances de la Société. Bulletin de la Société de géographie d’Égypte 9–10. 81–86. Campbell’s Illustrated Weekly. 1893. Professor Demetrius Mosconas. March, vol. III,1. 128. Carstens, Patrick Richard. 2014. The Encyclopædia of Egypt during the Reign of the Mehemet Ali Dynasty 1798–1952. Victoria: FriesenPress. Daly, Martin W. 1997. The Sirdar: Sir Reginald Wingate and the British Empire in the Middle East. Philadelphia: American Philosophical Society. Delamaire, Marie-Stéphanie. 2003. Searching for Egypt: Egypt in 19th century American world exhibitions. In Jean-Marcel Humbert & Clifford Price (eds.), Imhotep today: Egyptianizing architecture, 123–134. London: UCL Press. Deny, J. Mus̲h̲īr. Encyclopaedia of Islam, Second Edition. http://referenceworks.brillonline.com/entries/encyclopaedia-of-islam-2/mus-h-i-rSIM_5591 (checked 24/3/2014) Encyclopaedia Britannica. Heinrich Karl Brugsch. http://www.britannica.com/EBchecked/topic/82059/Heinrich-Karl-Brugsch (checked 24/3/2014). Finch-Hatton, Murray. 1873. A practical hand-book of Arabic for the Nile. London: W. H. Dalton. Forbes, Duncan. 1863. A grammar of the Arabic language, intended more especially for the use of young men preparing for the East India Civil Service: and also for the use of selfinstructing students in general. London: W. H. Allen. Green, Arthur Octavius. 1883–1909. Practical Arabic grammar for the use of English officers in Egypt. 2 vols. Vol 1 1883, vol. 2 4th ed. 1909 (1st ed. 1885). Cairo: Boulack Printing Office, Oxford: Clarendon Press. Green, Arthur Octavius. 1886. A collection of modern Arabic stories, ballads, poems and proverbs, for the use of English officers in Egypt. Cairo: s.n. Green, Arthur Octavius. 1895. A practical Hindustānī grammar. 2 vols. Oxford: The Clarendon Press. Haggenmacher, Karl. 1892. Grammatik des ägyptisch-arabischen Vulgärdialektes. Cairo: Boehme & Anderer. Hulme-Beaman, Ardern G. 1898. Twenty years in the Near East. London: Methuen. Jerrold, Walter. 1916. Earl Kitchener of Khartoum: The story of his life. London: J. Johnson.

Arabic language guides written for the British Army | 25

Journal of the Royal Asiatic Society. 1921. Notes of the Quarter. Vol. 53, no. 3, July 1921. 481–499. King, L. W. 1916. Obituary: Sir Charles Watson. The Geographical Journal 47(5). 387–389. Lamothe, Ronald M. 2011. Slaves of fortune: Sudanese soldiers and the River War, 1896–1898. Woodbridge: James Currey. Lane-Poole, Stanley. 1919. Watson Pasha: A record of the life-work of Sir Charles Moore Watson, K.C.M.G., C.B., M.A., Colonel in the Royal Engineers. London: John Murray. Lewis, B. Bāb-i Serʿaskeri. Encyclopaedia of Islam, Second Edition. http://referenceworks.brillonline.com/entries/encyclopaedia-of-islam-2/ba-b-i-seraskeri-SIM_0972 (checked 24/3/2014). Mak, Lanver. 2011. The British in Egypt: Community, crime and crises 1882–1922. London: IB Tauris. Mosconas, Demetrius. 1884. English & Arabic dictionary accompanied by dialogues & useful notes for the use of the British Army of Occupation. Cairo: s.n. Nakao, Shuichiro. 2013. Pidgins on the Nile: Tracing back the history of European influence on Arabic-based pidgins. Paper presented at the 10th AIDA Conference, Doha, Qatar, November 2013. Nakhlah, Yacoub. 1874. New manual of English and Arabic conversation = al-tuḥfa al-murḍiya fī taʿallum al-luġa al-ingilīziyya. Cairo: The Khèdive’s Press. Nallino, Carlo Alfonso. 1900. L’Arabo parlato in Egitto: grammatica, dialoghi, e raccolta di circa 6000 vocaboli. Milano: Ulrico Hoepli. Perry, C. R. 2004. Scudamore, Frank Ives (1823–1884). Oxford dictionary of National Biography. Oxford University Press, http://www.oxforddnb.com/view/article/24974 (checked 24/3/2014). Porter, Whitworth. 1889. History of the Corps of Royal Engineers. 2 vols. London: Longmans, Green. Redhouse, James W. 1880. Redhouse’s Turkish dictionary, in two parts, English and Turkish, and Turkish and English. London: Bernard Quaritch. Roth, Mitchel P. 1997. Historical dictionary of war journalism. Westport: Greenwood Press. Scudamore, Frank. 1915a. Arabic for our armies: Words and phrases with their equivalent in Colloquial Arabic (phonetic pronunciation) in daily requirement by H.M. Forces serving in the Near East. London: Forster Groom. Scudamore, Frank. 1915b. Turkish for Tommy and Tar: Words and phrases with their equivalent in Colloquial Turkish (phonetic pronunciation) in daily requirement by H.M. Forces serving in the Near East. London: Groom. Scudamore, Frank. 1915c. “Parley voo”!! Practical French phrases and how to pronounce them. For daily use by British soldiers. London: Forster Groom. Scudamore, Frank. 1925. A sheaf of memories. London: Unwin. Scudamore, Frank & Frederic Natusch Maude. 1916. Sprechen Sie Deutsch and “Parley Voo”!! Practical French and German phrases, and how to pronounce them. London: Forster Groom. Shaw, Stanford J. 1965/1966. The Nizam-I Cedid Army under Sultan Selim III 1789–1807. Oriens 18/19. 168–184. Shidyāq, Aḥmad Fāris. 1856. A practical grammar of the Arabic language: With interlineal reading lessons, dialogues and vocabulary. London: B. Quaritch. Spitta-Bey, Wilhelm. 1880. Grammatik des arabischen Vulgärdialektes von Aegypten. Leipzig: J. C. Hinrichs’sche Buchhandlung. Spitta-Bey, Guillaume. 1883. Contes arabes modernes: recueillis et traduits. Leiden: E. J. Brill, Paris: Maisonneuve. Steingass, Francis J. 1884. The student’s Arabic-English dictionary. London: Lockwood. Thimm, C. A. 1898. Egyptian self-taught (Arabic). London: Marlborough.

26 | Liesbeth Zack

Tien, Anton. 1879. The Levant interpreter: A polyglot dialogue book for English travellers in the Levant. London: Williams and Norgate. Tien, Anton. 1882. Egyptian, Syrian, and North-African hand-book: A simple phrase-book in English and Arabic for the use of the armed forces and civilians. London: W. H. Allen & Co. Tien, Anton. 1885. Manual of Colloquial Arabic comprising practical rules for learning the language, vocabulary, dialogue, letters and idioms, &c. in English and Arabic. London: W. H. Allen & Co. Toledano, Ehud R. 2003. State and society in mid-nineteenth-century Egypt. Cambridge: Cambridge University Press. Versteegh, Kees, Mushira Eid, Alaa Elgibali, Manfred Woidich & Andrzej Zaborski. 2006. Introduction. In Kees Versteegh, Mushira Eid, Alaa Elgibali, Manfred Woidich & Andrzej Zaborski (eds.), Encyclopedia of Arabic language and linguistics. Vol. I: A–Ed, v–x. Leiden: Brill. Vollers, Karl 1890. Lehrbuch der aegypto-arabischen Umgangssprache, mit Uebungen und einem Glossar. Cairo: s.n. Vollers, Karl F. & F. C. Burkitt. 1895. The Modern Egyptian dialect of Arabic: A grammar with exercises, reading lessons and glossaries, from the German of Dr K. Vollers with numerous additions by the author. Translated by F. C. Burkitt. Cambridge: University Press. Wahrmund, Adolf. 1861. Praktisches Handbuch der neu-arabischen Sprache. Giessen: Ricker. Watson, Charles Moore. 1885. English-Arabic vocabulary and dialogues for the use of the Army and Navy. London: Her Majesty’s Stationery Office. Western Daily Press. 1924. Two war veterans. 27 August 1924, p. 6. White, Arthur Silva. 1899. The expansion of Egypt under Anglo-Egyptian Condominium. London: Methuen. Williams, Josiah. 1884. Life in the Soudan: Adventures amongst the tribes, and travels in Egypt in 1881 and 1882. London: Remington. Willmore, John Selden. 1901. The spoken Arabic of Egypt. London: David Nutt. Wingate, Francis Reginald 1891. Mahdiism and the Egyptian Sudan, being an account of the rise and progress of Mahdiism, and of subsequent events in the Sudan to the present time. London: Macmillan. Woidich, Manfred. 1996. Rural dialects of Egyptian Arabic: An overview. Égypte/Monde arabe 27–28. 325–354. Wright, William & Karl Caspari. 1859–1862. A grammar of the Arabic language, translated from the German of Caspari, and edited, with numerous additions and corrections. 2 vols. London: Williams and Norgate.

Murad Suleymanov

Baku Russian: Colonial heritage in a post-colonial era Abstract: Almost a quarter of a century after the break-up of the Soviet Union, the linguistic legacy of Russian remains an inseparable part of identity for large groups of ethnically non-Russian population now living in the newlyindependent states whose political ties with Moscow have since loosened considerably. In some of them, such as Belarus and Kazakhstan, Russian has been able to hold on to its status as an official language. In others, such as Azerbaijan, it completely ceased to fulfill administrative functions, but continues to enjoy relative popularity in everyday life. A majority-Russian-speaking city well into the 1980s, the capital city of Baku is probably the only locale in Azerbaijan where the survival of Russian is ensured at least in the lifetime of today’s young generation. Owing to the city’s multicultural environment dating back to the 1880s oil boom which stimulated rapid growth and active immigration of non-Azeris in Baku, Russian soon surpassed its ethnic borders and came to be adopted as a first language by thousands of Azeris, Armenians and Jews growing up in the city in the post-World War II era. On the other hand, its special role as the first language of groups that did not identify as Russian and its insular position led to it acquiring a recognizable drawl and unique elements of vocabulary, clearly under the influence of Azeri and to a lesser extent Armenian and non-standard varieties of Russian. The break-up of the Soviet Union affected both the status and the spread of Russian in Azerbaijan. Hundreds of thousands of its speakers chose to emigrate for political and economic reasons, leaving behind a city where Russianspeakers are now a minority; but one willing to preserve its language and pass it onto the generation which has not in fact grown up or even lived under the Soviet rule. Baku Russian, however, found itself more isolated and therefore more and more susceptible to lexical and syntactic influence from Azeri, leading to the creation of a very distinct dialect. Keywords: bilingualism, language contact, Russian, creole, Azerbaijan

|| Murad Suleymanov: Collège de France, 52, rue du Cardinal Lemoine, 75005 Paris, France. E-mail: [email protected]

28 | Murad Suleymanov

1 Introduction Baku, the capital of Azerbaijan, is the largest city of the country with a population of over two million people, owing its growth to the country’s historically significant oil industry. It is unknown when exactly Baku was founded, but according to Sara Aşurbəyli, already in the first century AD a port town existed in the place where Baku is located today. At that time the Roman army briefly maintained a military camp nearby (Aşurbəyli 1992: 9–15). According to one tradition, it became the site of Bartholomew the Apostle’s martyrdom (Alexius II 2000: 495). A Zoroastrian temple built around a natural flame feeding off the region’s rich gas reserves has attracted pilgrims here from ancient times. A significant Zoroastrian community was still in existence here in the tenth century, two centuries after the region’s Islamization. From 1191 to 1538, the fortress of Baku was the capital of the mediaeval state of Shirvan, until conquered by the Safavid Empire, a state that encompassed modern-day Azerbaijan, Armenia, Iran, Iraq and Afghanistan. In the eighteenth century, Baku served as the capital of the semi-independent Khanate of Baku, and in the early nineteenth century it was conquered by the Russian Empire. Baku remained a relatively remote corner of the Caucasus and did not grow much out of the original fortress until 1859 when it became a provincial capital. The economic boom that followed the industrialization of oil extraction made Baku a destination for tens of thousands of immigrants from elsewhere in Russia and abroad. The arrival of immigrants drastically changed the cultural scene of the growing city. The multi-ethnic community that developed outside the walls of the mediaeval fortress would less than a century later transform into a distinct secular colonial supra-ethnic urban micro-society with its own behavioral and linguistic characteristics that remain understudied to this day. The author of this article directly witnessed the by-gone dominance of this micro-society in Baku and is himself a speaker of the language variety presented in this article, known as Baku Russian. These factors are a major source of the author’s interest in this particular subject. They prompted him to study this variety systematically later in his life, as well as to provide first-hand insight into the historical and sociological factors of the emergence and existence of the Russian-speaking community of Baku. This article is a modest attempt of a general description of this community and its language. The sociolinguistic part includes testimonials from individuals who grew up in Baku in the Soviet era and with whom the author held personal conversations. The empirical data used in examples is quoted from a corpus collected by the author during a three-and-a-half year stay in Baku from

Baku Russian: Colonial heritage in a post-colonial era | 29

2009 to 2013. The corpus consists in part of recordings of spontaneous speech and elicitations from volunteer first-language speakers of Baku Russian and in part of samples from the media available on the Internet, such as video clips with speech samples and written discourse in online forums. To avoid examples of first-language-influenced speech for those whose fluent command of Russian cannot be verified (e.g. online forum users), only examples that confirm the author’s personal field observations have been cited in this article.

2 Sociolinguistics of Russian in Baku Historically, the Abşeron Peninsula, including the fortress of Baku, did not display much linguistic diversity. Prior to the massive industrialization of Baku in the second half of the nineteenth century, the only widely spoken languages of the area were Azeri and notably Tat, an Iranian language closely related to Persian and gradually supplanted by Azeri due to language shift. According to Azeri historian Abbasqulu Bakıxanov, apart from six villages in the eastern part of the peninsula, the entire population of the Baku District1 in 1840 spoke Tat (Bakıxanov 1991: 24). In his 1842 description of a journey along the coastal regions of the Caspian Sea, Russian orientalist Ilya Berezin, Bakıxanov’s contemporary, mentions the existence of 38 villages in Abşeron (Bérézine 2006: 368). This gives an idea of the wide spread of Tat in the region still present in the midnineteenth century. However, when Boris Miller, a Soviet specialist on Iranian languages, published his findings on the state of the Tat language in 1926, the knowledge of Tat remained stable only in seven of the original villages; in two more, the language still lingered only in several families (Miller 1929: 5). Nowadays Abşeron Tat is spoken only in two villages: Balaxanı and Suraxanı. In the second half of the nineteenth century, the effects of language contact in the urban and rural parts of Abşeron went separate ways. Rural areas remained largely unaffected by massive waves of immigration from elsewhere in the Russian Empire which brought drastic changes upon the ethnic and linguistic make-up of the urban area of Baku. The dominant role of Azeri shrank as the city grew and developed. It yielded increasingly to Russian in many positions – not only as the language of the governing authorities, but also as a mode of interethnic communication. By 1897, ethnic Russians already constituted 33 %

|| 1 The Baku District (Bakinskij uezd) corresponded to the Abşeron Peninsula and the modernday Xızı District of Azerbaijan.

30 | Murad Suleymanov

of the city’s population. Within the next six years, this number grew by exactly a third and between 1903 and 1913 by a quarter, leading to their surpassing Azeris in numbers by the advent of World War I (Altstadt 1992: 32). Azeris would not become a majority in Baku again until the 1970s, when the city’s population exceeded 1 million. The table in the appendix shows changes in the ethnic make-up of Baku between 1897 and 1979 (Altstadt 1992: 32; Population of the Transcaucasian Krai, 1886; Soviet census results from 1926, 1939, 1959, 1970, and 1979). Interethnic communication, in turn, created conditions for widespread albeit unidirectional bilingualism, disfavoring the knowledge of Azeri. The latter’s lack of administrative functions spared non-Azeri civil servants from the necessity to possess any knowledge of it. In addition, job opportunities in the rapidly growing city also attracted people of backgrounds other than Russian in significant numbers. Consequently, non-Muslim immigrants of this period arriving mainly from or through European Russia (European Jews, Germans, Georgians, Poles, Greeks, etc.) tended to integrate and later assimilate into the more dominant and prestigious Russian-speaking culture and rarely knew Azeri. The secular intelligentsia that emerged in the Azeri society at the time was mainly educated in the so-called Russian-Muslim schools where Russian was the main language of instruction. During this period, when Azeri nationalism was still dormant, linguistic preferences of many graduates of such schools reflected the common attitudes of the city’s environment. Disdain of Azeri as the ‘rough vernacular’ by a portion of the Russian-educated ethnic Azeri urbanite middle class was gibbeted in the satirical press and in the early-twentieth-century works of Cəlil Məmmədquluzadə (Məmmədquluzadə 1906). Məmmədquluzadə is known to have once characterized the linguistic attitudes of the contemporary educated Azeri society thus: “My kinsmen that speak Russian with their living and Arabic with their dead…”2 The 1917 Russian Revolution, Azerbaijan’s brief independence from 1918 to 1920 and the post-World War I turmoil did not change much in the ethnolinguistic situation of Baku. Russian reaffirmed its dominating role after the installation of the Communist government. Despite the exodus or repatriation of many Russian-speakers of the previous waves (mainly civil servants, industrialists and entrepreneurs), the gap was quickly filled by scientists, technicians, teachers, military staff and other representatives of the Soviet ‘middle class’ sent on assignment by public institutions from Russian-speaking parts of the

|| 2 In Islamic societies, prayers for the repose of the dead are recited in Arabic. The original Azeri saying is “Dirisiylə rusca, ölüsüylə ərəbcə danışan xalqım mənim...”

Baku Russian: Colonial heritage in a post-colonial era | 31

USSR to Azerbaijan to provide the young republic with the specialists it needed. Similar to the original Russian-speaking newcomers, these migrants did not regard it as necessary to learn the language of the native population even in the second generation, as services in Russian were never unavailable or refused. Soviet ethnic policy did indeed favor and promote literature and education in the main languages of the Soviet republics, but those measures were primarily intended to meet the cultural needs of the native population. In 1979, less than 10 % of approximately half a million Russians in all of Azerbaijan possessed knowledge of Azeri (Altstadt 1992: 187). From the administrative point of view, Azeri enjoyed a semi-official status in Baku; like elsewhere in Azerbaijan. Signs, stamps, documents and application forms were usually bilingual, at least in the post-World War II period. On 21 August 1956, the Chair of the Presidium of Azerbaijan Mirzə İbrahimov signed an amendment to the constitution of the republic which granted Azeri the status of an official language (Bakinskij Rabočij 1956). In the 1978 versions of the constitutions of Georgia, Armenia and Azerbaijan, their principal languages (Georgian, Armenian and Azeri respectively) were mentioned as the only official languages of these republics. Azeri was the least Russified Turkic language of the USSR, and its vitality seemed obvious thanks to rich and diverse literature and the expansion of the Azeri-language domain in universities and the Academy of Sciences (Bennigsen & Wimbush 1985: 138). Nevertheless in Baku, despite secondary and postsecondary education being available in both Azeri and Russian, the quality of the Russian option was superior in its logistical base, methodology and teaching staff which often held diplomas of prestigious institutes. The relatively young tradition of secular education in Azeri could hardly compete with its well-established and continuously improving Russian counterpart. This reality tended to influence the decisions of those Azeri-speaking parents who chose to register their children in Russian schools. Given the fact that the phenomenon of assimilating into the Russian-speaking environment remained the case for traditionally non-Muslim residents of Baku, it is needless to say that attending Russian schools was viewed as a self-evident option for minorities such as Armenians, Jews and Ukrainians. Thus the generation of people of various backgrounds born during or immediately after the Second World War featured the first representatives of the massive linguistic community of the late- and post-Soviet-period known in Baku as the Russkojazyčnye (Russian for ‘Russian-speakers’), i.e. those who used exclusively or primarily Russian in their everyday lives. Curiously, the sociolinguistic dominance of Russian continued to grow even after the last massive wave of immigration of Russian-speakers from outside of Azerbaijan in the late 1940s. Furthermore, first signs of the sluggish out-

32 | Murad Suleymanov

migration of Russian-speakers to other republics around the same time, as well as the arrival of Azeris from ‘Azeri-speaking’ parts of the republic (whose level of spoken Russian was relatively low compared to the Azeris in Baku) in the early 1970s could not compromise the dominant role of Russian, as Russification of the younger generation of ‘middle-class’3 Baku Azeris who had attended Russian schools was well underway. This generation of urbanite Azeris remained more or less bilingual, still using Azeri at home with their parents, but increasingly opting for Russian within their circle of friends, including those who did know Azeri. In 1970, 29 % of urban Azeris (including residents of cities other than Baku) claimed to speak Russian fluently (Baskakov 1977: 164). According to Turovsky, by 1989 this figure constituted a little less than 60 % for Baku Azeris (Turovsky 1998: 117). Furthermore, the 1989 census counted 57,500 ethnic Azeris living in Azerbaijan who already considered Russian their first language (Grenoble 2003: 135). This figure may not represent the real picture, because in the case of Azeris who reported Azeri as their first language the census did not take into account whether they actually used this language in their households, nor did it offer to evaluate their degree of competence in Azeri, which may have been lower than that in Russian. Similar processes were observed within other major ethnic communities of Baku, sometimes even in earlier decades. Most Armenians living in Baku were descendants of immigrants from the Nagorno-Karabakh region of Azerbaijan where there was an administrative Armenian autonomy during the Soviet era and the viability of the Armenian language was very stable. The second and third generations of these immigrants tended to have a weak knowledge of Armenian, and their integration into the Russian-speaking culture went as far as them often opting to name their children traditionally Russian names, uncommon among Armenians in Armenia, and even adopting Russian surname endings (Shahnazarian 2014). In 1970, 40 % of the urban Armenians of Azerbaijan reported Russian as a second language as opposed to 10 % who spoke Azeri as a second language (Altstadt 1992: 189), though for many of those who reported Russian as a second language, it was probably used in everyday life far more often than Armenian. As of 1970, 58 % of ethnic Armenian schoolchildren in Azerbaijan (urban and rural) were receiving education in Russian (Baskakov 1977: 164). Leaping ahead, it should be noted that these trends continued even after the exodus of Armenians from Baku at the advent of the Nagorno-Karabakh War. According to sociologist

|| 3 The relation between language choice and class in a supposedly classless Soviet society will be addressed later in this article.

Baku Russian: Colonial heritage in a post-colonial era | 33

Nona Shahnazarian, in rural Armenian-populated towns in southern Russia, where many Baku Armenians had settled after the war, teachers of Armenian noted a complete lack of interest on the part of this group towards learning Armenian (Shahnazarian 2008: 102). Ashkenazi Jews who settled in Baku following the pre-Soviet wave of immigration had already been largely Russified, but a significant minority retained the knowledge of Yiddish and even competence in Hebrew in the early twentieth century. According to the Jewish Library in Baku’s 1910 report of books borrowed from 1907 to 1909, 7,134 books were in Russian, 600 in Yiddish and 389 in Hebrew (Kelner 2003: 40). In the pre-World War II period, when schools with instruction in Yiddish could still be found in the Soviet Union, this option was available to Baku Jews as well. As of 1927, only 11 % of Ashkenazi Jewish children attended school in Yiddish, and none of them above fourth grade (Konstantinov 2007: 118). In 1939, roughly 20 % of Ashkenazi Jews in Azerbaijan reported Yiddish as a first language, the rest mainly reporting Russian (Konstantinov 2007: 30). Heavy Russification was characteristic of most Jews living in the urban areas of the Soviet Union. It is noteworthy that the Ashkenazi Jewish community of Azerbaijan stood out socially against others elsewhere in the USSR: in 1939, 94 % of the Ashkenazi Jews in the republic lived in urban areas (mainly Baku; Konstantinov 2007: 24), in addition to having the second highest rate of university graduates (after the Ashkenazi Jews of Russia-proper), a trend that continued at least for the next 20 years (Konstantinov 2007: 90). Waves of Iranian Azeri immigration were also motivated by different factors: primarily economic ones for low-skilled seasonal day-laborers who arrived before the 1917 Revolution (some of them would later move back to Iran) and political ones for those who arrived after the Sovietization and were generally well-educated sympathizers of communism. Due to cultural and linguistic affinity, Iranian Azeris did not juxtapose themselves to the indigenous Azeris, like other immigrants did, and at the same time, were socially very well integrated. They remained, however, a rather urbanized community, often populating immediate ‘Russian-speaking’ suburbs of Baku, which contributed to a good knowledge of Russian on their part already in the second generation. It is noteworthy that the linguistic shift towards Russian among Azeris (or Armenians) did not at all signify a shift towards a Russian ethnic identity. The legal status of ethnicity reflected in every official identification card issued by the Soviet government, the state’s constant emphasis of ‘internationalism’ and its reinforcement through various artistic initiatives and establishments promoting folklore, the unofficially privileged status of the ‘titular’ ethnicities (i.e. those forming majority in each of the 15 republics) and the lack of any strategic steps on the part of the Soviet government to rid them of their identity (despite

34 | Murad Suleymanov

encouraging mass knowledge of Russian) had all led to a situation where people were generally quite conscious of their ethnic background and often strongly identified with it. This was especially true for Baku, where the four major ethnicities (Russians, Azeris, Armenians, and Jews) historically hailed from very different cultural backgrounds, rendering the language shift insignificant in determining their ethnic affiliation. The latter soon became the recipient of heavy complement from (but hardly losing grounds to) the new secular superethnic identity known as the Bakintsy (Russian for ‘inhabitant of Baku’), which essentially was the first step, albeit on a very limited scale, towards the formation of the ‘Soviet nation’, an idea that had gained popularity after Nikita Khrushchev acknowledged it at the 22nd Congress of the Communist Party of the Soviet Union in 1961. Contemporary Marxist–Leninist philosophers defined it as follows (“Soviet people” 1983: 169): The example of the Soviet people serves to demonstrate patterns and ways of gradual elimination of class and ethnic differences. The path to a common class-free and ethnicityfree communist mankind does not lie through mechanical rejection of class and traditional forms of social life, but through a full use of their potential, as well as full disclosure and high-level synthesis of the best and most progressive features associated with each class and each ethnic group. This process does not presuppose social leveling of public life or ridding it of its ethnic features.

The Bakintsy of the late Soviet period, thus, came to define themselves, besides geographical affiliation (born and raised in Baku), in terms urban manners (including aesthetic tastes and choices), cosmopolitanism (openness to outside culture, indifference to ethnic factors when making friends, socializing, dating, starting a family, etc. and/or being of multiethnic origin), high level of education, and linguistic factor (having a fluent command of Russian), therefore embracing the Russkojazyčnye but not at all being limited to them. The Bakintsy constituted, as anthropologist Tsypylma Darieva concludes, a social caste with its own behavioral code, etiquette and language (Darieva 2011: 154). Due to its development, the Russian-speaking population of Baku continued to grow culturally distant to the non-Russian-speaking population elsewhere in the republic. At the same time, the spoken Russian variety of Baku was likewise gaining distance to standard Russian: decades of contact with Azeri had left its mark on it in the form of a very recognizable ‘Azeri tune’ and unique vocabulary. One factor that contributed heavily to the consolidation of historically diverse ethnic communities in Baku under the umbrella of a single identity was the official Soviet policy of atheism. Like elsewhere in the Russian Empire, in pre-Soviet Baku religious differences between Shia Muslim Azeris, Eastern Orthodox Russians, Oriental Orthodox Armenians, and Jews were quite marked. In

Baku Russian: Colonial heritage in a post-colonial era | 35

the Soviet era, obligatory secondary education for all characterized by massive atheist propaganda led people in Baku to largely abandon their religious traditions or attach very little importance to them outside of certain rites (such as funeral proceedings, non-mixed cemeteries, circumcision, occasional baptism). Only seventeen mosques, three Armenian Orthodox churches, two Russian Orthodox churches, and one synagogue functioned in all of postwar Soviet Azerbaijan. The sociolinguistic processes of this period and the following decades, thus, demonstrate that the phenomenon of ‘speaking Russian’ was very well distinguished from the phenomenon of ‘being Russian’. It would be more accurate to assert that the urban culture of post-World War II Baku was not Russian, but rather Russian-speaking in essence. The position of Russian was further reinforced by mixed marriages. Cases of primarily Azeri men marrying Russian, Armenian and Jewish women became quite common in Baku in the second half of the twentieth century (but still relatively low compared to other large cities of the Soviet Union). In 1960, 7.6 % of all marriages recorded in two central districts of Baku were interethnic; by 1985, this figure fell to 7.0 % (Ždanko 1990: 427). In 1970, up to 60 % of all mixedmarriage couples in Azerbaijan included an Azeri spouse, and in 2/5 of those cases, the second spouse was Russian (Bromley 1977: 464–465). In mixed Azeri– Russian, Azeri-Armenian and Azeri-Jewish families living in Baku in this period, the primary language spoken at home was almost always Russian. The next generation of Baku Azeris showed even more preference for Russian, not only because many of its representatives completed their studies in Russian, but also due to them often having strong Russian-speaking or asymmetrically bilingual (with Russian as the dominant language) parents. In addition, recent immigrants from the Azeri-speaking provinces, which were still not as numerous as the local Russkojazyčnye, usually mastered Russian quickly. Ideally a classless society, the Soviet Union displayed quite obvious class division, especially in large cities, including Baku. In linguistic terms this was marked by Russification: the higher the class, the higher the asymmetry of bilingualism, favoring Russian. Children of high-ranking non-Russian members of the Communist Party apparatus often had very superficial knowledge of their heritage language. A 2012 interview carried out by the author suggests that even gender may have affected the phenomenon of bilingualism. A female informant, 60, recalled from her secondary school experience that boys who attended schools with Russian language instruction usually had a better command of Azeri than their female schoolmates. The informer explained this by the fact that due to the lower status of Azeri in comparison with Russian, it was regarded as the lan-

36 | Murad Suleymanov

guage of the ‘street culture’ that boys aspired to identify with, whereas girls would not find it as appealing. Soviet Baku fell under the definition of a colonial city in many terms. From the economic point of view, the exploitation of the city’s oil resources by the Soviet government without considering the ecological impact led to alarming levels of pollution in Baku and the nearby industrial city of Sumqayıt by the early 1990s. From the cultural point of view, in light of the existence of Azerilanguage publications, there was indeed an attempt to keep this language alive, but attempts to promote it were superficial at best. In Russian schools of Baku, weekly Azeri lessons were obligatory, but they lacked solidity and were reduced to reading and translating unsophisticated texts with bilingual vocabulary given in advance and completing simple exercises. Speaking Azeri outside of the Azeri class was not welcome. Once demoted, Azeri was underappreciated even by those whose parents still formed part of the Azeri speech community (similar to graduates of the Russian-Muslim schools of the early twentieth-century, as mentioned above). A middle-class female Azeri informant born in Baku, a fluent Azeri-Russian bilingual, who had begun speaking to her toddler only in Azeri in the mid-1980s (instead of Russian, which circumstances would have enabled him to pick up in the future anyway, according to her logic), reported bewilderment of her Russian-speaking ethnic Azeri friends at her choice. In 1989, according to Lenore Grenoble, 62 % of the population of Baku reported speaking Russian as a second language. This figure does not include those who reported it as a first language, of which ethnic Russians constituted a little less than 20 % at the time (Grenoble 2003: 135). In addition, this figure may have been much higher given that a portion of the Armenian population had left Baku by the time of the census due to disturbances caused by the Nagorno-Karabakh conflict. It should be noted, however, that Azeri was far from being considered an endangered language in Baku. Significant groups of the population living in the old neighborhoods did not speak much Russian and were still clinging to their historical language, forming Azeri-speaking ‘enclaves’ in the city center. Such were notably the Old City (İçərişəhər), the Outer City (Bayırşəhər) and Sovetskaya. In terms of the preservation of Azeri, the situation in the villages of the Greater Baku area was even less worrisome. Given the fact that pre-Soviet Russian-speaking immigration did not affect Abşeron outside of half a dozen industrial suburban communities, the traditional villages were not affected by Russification. Even in the Soviet era, the influence of the Russian-speaking population in those villages was extremely weak. The Perestroika and the subsequent fall of the Soviet Union initiated a remarkable change in the linguistic attitudes and habits of the population of Ba-

Baku Russian: Colonial heritage in a post-colonial era | 37

ku. The economic crisis, the violent attempt by Soviet troops to suppress the democratic movement in Baku in January 1990, the interior instability in Azerbaijan, the loss of the official status of Russian, uncertainty and personal security concerns, and the hostility of the newly-arrived Azeris expelled from Armenia and Armenian-controlled parts of Azerbaijan at the advent of the NagornoKarabakh conflict, later topped with deterioration in Russian-Azerbaijani relations, caused many Russkojazyčnye, including Russian-speaking Azeris, to leave the country, while Azeris from outside of Baku settled massively in the capital city either as refugees or IDPs, or in search of jobs after rural economies had fallen into stagnation. The Bakintsy culture suffered moral decline as its representatives dispersed throughout the world in search of new opportunities. The linguistic policy in post-Soviet Azerbaijan was mainly preoccupied with the revitalization of Azeri. At the beginning of the twenty-first century, even though Russian was no longer used in administration, it has still remained the household language for a significant number (but not the majority) of inhabitants of Baku, mostly belonging to the middle class. Today those belonging to the generation born to the Bakintsy families as the Soviet Union neared its demise are for the most part fluent bilinguals, but the tradition of introducing Russian to the children (even those born more than a decade after the fall of communism) from a very early age is still widespread. Darieva accurately characterizes the perception of the Bakintsy identity in the twenty-first century as a “nostalgic concept of the vanishing past”, often idealized as “united by a peaceful conviviality based on formal knowledge of distinctions between ethnic groups on the one hand and on romanticized interethnic intimacy on the other hand” (Darieva 2011: 135). In contrast, the Russkojazyčnye identity exists to this day, though its cultural connotation is not nearly as universal and elaborate as that of the Bakintsy and carries only distant traces of it, being used mainly to distinguish those who received secondary education in Russian. Still today the Russian language retains popularity among the Azerbaijani intelligentsia and the ‘elite’ (Karavaev 2008), which presupposes that it enjoys certain prestige and that a good command of Russian is still socially marked.

3 Russian variety of Baku – general overview Permanent Russian settlements first appeared in what is now Azerbaijan in 1833, but field work on their spoken varieties had not been undertaken until the second half of the twentieth century and still remains limited. Mikirtuni (1952) makes the first attempt at describing the phonology of the local Russian varie-

38 | Murad Suleymanov

ties by analyzing dialectal vocabulary based on data collected in 12 Russianspeaking villages of Azerbaijan. Katz (1963) offers a lexical insight into the variety of one single Russian village, Slavyanka, in western Azerbaijan. Aslanov (1967) concentrates on the phonetic adaptation of Azeri loanwords in rural Russian. More recently, the dialectal dictionary by Quliyeva et al. (2006) was published, consisting of approximately 6,000 lexical units collected by authors in 22 Russian villages across Azerbaijan, accompanied with examples from a spoken corpus. All of these works, however, dealt exclusively with the dialects of rural Russian communities of Azerbaijan (nowadays almost extinct due to emigration), whose origins, background, everyday culture and history of interaction with non-Russians were profoundly different from those of the urbanite Russians of Baku. While it is true that exposure to Azeri has been key in both cases, social factors exclude any chance of parallel linguistic processes. In contrast, the phenomenon of the Russian variety of Baku has not been addressed in academic literature. However, due to its phonetic, lexical and syntactic features, as well as enduring sociolinguistic role in Baku even a quarter of a century into Azerbaijan’s independence, Baku Russian certainly merits a thorough description. It is indeed a typical case of a minority language, as opposed to a pidgin, in that its ‘historical’ speakers (i.e. ethnic Russians of Baku) display the same speech patterns as Azeris who speak it fluently, but as a second language. What makes Baku Russian different from other forms of spoken Russian in the South Caucasus, however, is that the continuous dwindling of the ethnic Russian community does not necessarily presuppose the dwindling of the use of the Russian language. Despite heavy influences from the official local language, even today Russian continues to serve as the mother tongue and means of everyday communication for many non-Russian inhabitants of Baku, including those who lived most of their lives in independent Azerbaijan.

4 Phonetics Baku Russian is recognizable for its unique intonation pattern, which generally follows the intonation pattern of standard Azeri known for its ‘melodiousness’ (Budaqova et al. 1977: 50). Vowels in Baku Russian are reduced, though not to the same degree as in standard Russian. Speakers have a tendency to preserve the vowels /æ/ (), /œ/ (), /y/ () in Azeri borrowings (e.g. mählä – ‘neighborhood’, görmämiš – ‘scrounger’, ‘grab-all’), including in personal names which cannot be reanalyzed as first-

Baku Russian: Colonial heritage in a post-colonial era | 39

declension nouns (e.g. Mänzär, Könül, Sädagät, Vügar). On the other hand, names ending in a stressed /æ/ in Azeri (spelled as ) are usually adapted to fit the first declension in Russian and therefore undergo complete phonetic change: in every position switches to a ‘more Russian’ (sometimes ) and the stress notably shifts to the penultimate syllable: Lalə > Lala, Təranə > Tarana, Nazilə > Nazilja, Mədinə > Medina. Russian speakers from Russia tend to substitute the sound /æ/ in borrowed words of English origin with the sound /ɛ/ ( or ), e.g. flash (drive) > flēška or fleška, back-up > bēkap. Given that the sound /æ/ exists in the phonemic inventory of Azeri, Baku Russian speakers may preserve it in recent English borrowings, producing fläška (which sometimes passes to fljaška due to the proximity of before front vowels and the Russian palatalized ) and bäkap. The general feature concerning the consonants is the aspiration of voiceless stops

, and in syllable-initial and syllable-final positions, not typical for standard Russian, but typical for Azeri. Examples from Baku Russian include avt(ʰ)oš – ‘stunt driver’, k(ʰ)alat’ – ‘to kick up (a ball)’, p(ʰ)ap(ʰ)in dom – ‘excessive permissiveness’ (literally: ‘(one’s) dad’s house’). Another Azeri-induced feature common across all speakers of Baku Russian is the realization of the voiced affricate /dʒ/ (), which in standard Russian is pronounced as /dʑ/. Baku Russian speakers do not apply the usual standard Russian velarization to the second part and realize it like the corresponding Azeri voiced affricate (e.g. ǰana – ‘sweetie’ (term of endearment used for women), ǰaz – ‘jazz’, ǰigalit’ – ‘to cheat in a game’). In addition, some speakers, especially younger ones, would preserve the glottal fricative in Azeri borrowings, thus differentiating it from . For other speakers, shifts to , like in standard Russian (e.g. hajasyzka vs. xajasyzka – ‘female brawler’). Exceptions are lexemes that passed into local Russian very early, featuring archaic Russian transliteration of as (e.g. həmşəri > gamšara – ‘immigrant from Iran’, attested in 1886)4 and some proper names: Hüseyn > Gusejn, Həmidə > Gamida, Hövsan > Govsany, etc. Likewise, some speakers tend to preserve the voiced velar fricative in Azeri borrowings; others pronounce it as (e.g. gabyrganut’sja vs. gabyrğanut’sja – ‘to impose oneself’).

|| 4 Attested in Yagodynsky (1886: 164).

40 | Murad Suleymanov

5 Vocabulary The lexical inventory of Baku Azeri does not differ much from that of standard Russian, but differences can be pinpointed mainly on the level of slang. Origins of words that have no widespread cognates in Russian, spoken in Russia, are different. The most common source of non-typical Russian vocabulary is, as one may expect, Azeri. The use of Azeri words in Baku Russian does not necessarily presuppose code-switching, as some examples demonstrate a very advanced level of their integration into the Russian inflexional paradigm. In the examples (1) and (2), the borrowed nouns avara (‘slacker’) and šüvän (‘hue’, ‘outcry’) are inflected like regular Russian nouns of the first and second declension respectively. (1)

V sovetskoe vremja tot kto ne učil-sja in Soviet.N.ACC time.ACC that.M who NEG learn.PST.IPFV.3SG.M-REFL i byl avaroj krajnjak polučal and be.PST.3SG.M slacker.INS at_least receive.PST.IPFV.3SG.M prava i stanovilsja šofer-om driver’s_license and become.PST.IPFV.3SG.M driver-INS ‘In the Soviet times, whoever did not study and was a slacker would at least get a license and become a driver’

(2)

Nevesta s šüvän-om v-letaet bride with outcry-INS IN-fly.PRS.3SG ‘The bride dashes into the house, screaming’

v in

dom house.ACC

Some roots borrowed from Azeri demonstrate quite flexible derivation. Such is the root tapš which on its own is a noun with a very specific meaning of ‘patronage in the form of pulling strings in someone’s favor’ and is derived from the Azeri root tapşır- (‘entrust in’, ‘consign to’). The noun tapš in Baku Russian can produce verbs, both perfective (tapšanut’) and imperfective (tapševat’), that can, in turn, form reflexive verbs (tapšanut’sja, tapševat’sja) and participles (tapšujuščij, tapšanuvšij), including passive ones (tapšanutyj, tapšovannyj). In fact, growing symmetry of Azeri-Russian bilingualism in Baku allows an infinite number of borrowings that can be adapted to Russian morphology to various degrees. Hence let us consider other sources of vocabulary. Calquing is a common way to derive new expressions in language-contact areas, and Baku Russian does not constitute an exception with its abundant calques from Azeri.

Baku Russian: Colonial heritage in a post-colonial era | 41

(3)

Oni v Xyzy ostavalis’ they in Xızı stay.PST.IPFV.3PL ‘They were staying (at a hotel) in Xızı’

The verb ostavalis’ (inf. ostavat’sja) is a calque of the Azeri verb qalmaq meaning ‘to stay’, by extension also meaning ‘to live’. In standard Russian, when talking about a hotel stay, speakers are more likely to use the verb ostanovit’sja (‘to stop’). (4)

Načal mne gnat’ pro foton-y, bozon-y, begin-PST.PFV.3SG I.DAT chase.INF about photon-PL boson-PL ne znaju čto NEG know.PRS.1SG what ‘He began talking nonsense about photons, bosons, and so on’

The expression ne znaju čto is a calque of the Azeri nə bilim nə (‘I do not know what’), which in informal speech means ‘so on’, ‘et cetera’. Another source of Baku Russian-specific words is lexical extensions of existing lexical units from standard Russian. (5)

U nas za futbol-om sejčas xorošo smotrjat at we.GEN after football-INS now well look.PRS.3PL ‘They are now taking good care of football in our country’

While it is true that the expression ‘smotret’ za + noun’ means ‘to take care’, its use with an abstract noun sounds odd to first-language speakers of standard Russian, who would instead use the verbs zabotit’sja or udeljat’ vnimanie. (6)

Daže kogda nomer kuplen na čužoe even when number buy.PTCP.PSV.PST.3SG.M on someone_else’s.N UL, čerez-čerez naxodjat real'nogo vladel’ca ID indirectly find.PRS.3PL real.ACC.M owner.ACC ‘Even when the number is bought with someone else’s ID, they find the real owner indirectly (by passing from one person to another)’

The expression čerez-čerez is a redoubling of the Russian preposition čerez (‘through’). In informal Russian spoken in Russia, the expression čerez-čerez does occur in a similar sense, though as a preposition which requires an object in the accusative case, e.g. najti čerez-čerez znakomyx (‘to find through (a network of) acquaintances’). In Baku Russian, the expression is used as an adverb. Cases when a less frequent synonym in standard Russian enjoys a higher distribution in Baku Russian can sometimes be explained by the proximity to the word in Azeri. Thus, the word ‘market’ (i.e. ‘a place where traders set up

42 | Murad Suleymanov

stalls and buyers browse the merchandise’) has two translations in standard Russian: more common rynok (Slavic in origin) and less common bazar (Persian in origin, but exists in many Turkic languages, including Azeri). Baku Russian speakers, in contrast, prefer to use the latter word, undoubtedly being able to relate it to the Azeri word bazar. The word rynok, on the other hand, would be used mostly in the sense of ‘commercial demand’. Words inherited from non-standard Russian dialects present an interesting case for historical linguistics, potentially enabling researchers to determine trade and immigration patterns of the Russian-speaking population. The word ‘eggplant’ is often translated to Baku Russian as baklažan, as it is the case in standard Russian; however, some speakers would also translate it as dem’anka, a word widely understood in Baku, but unknown to most speakers in Russia. The word dem’anka is, in fact, known even nowadays in Astrakhan,5 a Russian city on the northern coast of the Caspian Sea, which historically has had very strong economic ties with Baku, going back to mediaeval times, long before Russian tsars’ first attempts to conquer the Caucasus. Not surprisingly, one of the earliest mentions of the word was made by Samuel Gottlieb Gmelin, an eighteenth-century German botanist and physician, who spent eight years of his life in Russia and visited Astrakhan, of which he wrote: “In the gardens of Astrakhan, instead [of European vegetables,] badendžans or dem’ankas and cayenne pepper are grown in abundance” (Gmelin 1936: 310). Another similar case is the word for ‘red beet’: the standard Russian word svekla is hardly used in Baku compared to the more popular word burak which is also typical for southern Russia and eastern Ukraine, where many pre-Soviet immigrants in Azerbaijan hailed from. Other ethnic communities living in Baku, notably the Armenians, have also made contributions to the vocabulary of Baku Russian. The most recognizable lexical unit of Armenian origin that is still in use is the expression asma eğala, which derives from asum ē eğel ē (‘they say it has happened so’). The expression is used in lieu of the Russian adverb jakoby or the particle mol, meaning ‘supposedly’, ‘allegedly’, ‘as if’, often sarcastically. It also corresponds to the Azeri adverb guya. (7)

V toj in that.F.PREP asma_eğala, supposedly

že armii same army.PREP i prokljatye and damned.PL

Ameriki gde rasizm, America.GEN where racism kapitalist-y za takoe capitalist-PL for such.N

|| 5 The word is attested in Russian periodicals printed in Astrakhan as late as 2006, cf. Perkina, Maria. “Podoždite polmesjaca i kupite naši, nastojaščie” (“Wait for Half a Month and Buy Ours, the Real Ones”). Komsomolec Kaspija. 26 May 2006.

Baku Russian: Colonial heritage in a post-colonial era | 43

mogut posadit’ be_able.PRS.3PL imprison.INF ‘Even in the army of the United States, supposedly a country of racism and damned capitalists, one can get imprisoned for such a deed’ (8)

Ja ponimaju čto Gruzija sejčas ob"ekt voždelenija I understand.PRS.1SG COMP Georgia now object longing oppozicii, asma_eğala oni vsjo dostigli, opposition.GEN supposedly they all achieve.PST.PFV.3PL i u nix tam raj and at they.GEN there paradise ‘I understand that Georgia nowadays is an object of admiration for the opposition, as if they have achieved everything and (as if) now there is paradise down there’

The expression is commonly used in the same sense with just the first part – asma. (9)

Stol-y, lomjaščie-sja ot odnotipnyx salat-ov table-PL crack.PTCP.PL-REFL from same_type.PL.GEN salad-PL.GEN i asma krasivo na-rezannyx and supposedly beautifully ON-cut.PTCP.PSV.PFV.PST.3PL.GEN frukt-ov fruit-PL.GEN ‘Tables overladen with salads of the same type and supposedly nicely cut fruit’

Another word of Armenian origin, albeit less common, but still known to speakers from the older generation, is dymbo meaning ‘fool’, ‘idiot’. The word dmbo has the same connotation in standard Armenian, but the original meaning of the word is ‘punch’. The word is used in the same sense in the Russian variety of Tbilisi, Georgia. Finally, Baku Azeri is characterized by a number of lexemes, whose Russian origin is undoubtable, but which appear to be cases of local invention. These include the words sobirun or sobiron (‘get-together’; from the transitive verb sobirat’ – ‘to gather’), domašnik (‘slipper’; from the word dom – ‘house’), as well as verbal expressions idti na šatal (‘to skip school’, from idti – ‘to go’ + presumably from the verb šatat’sja – ‘to come loose’, figuratively: ‘to idle about’), kidat’ svixu (‘to set a date’; from kidat’ – to ‘throw’ + diminutive from svidanie – ‘date’), kidat’ vyzov (‘to miscall’, ‘to make a telephone call, hanging up quickly after the first or second ring before the person answers, using the call itself as a signal that the person should call you’; from kidat’ + vyzov ‘evocation’).

44 | Murad Suleymanov

6 Syntax Baku Russian syntax does not display heavy deviation from standard Russian, given that by the time when this variety began forming, Azeri syntax itself had been greatly influenced by Iranian languages, prompting its shift toward the Indo-European sentence structure. In the decades after World War II, this shift was reinforced by the exposure to Russian, meaning that the influence of Azeri and Russian was, in fact, mutual. For modern spoken Azeri, both SOV and SVO orders are typical. Baku Russian follows the same pattern and displays almost the same ratio of distribution of each type of sentence order.

6.1 Noun phrase Standard Russian possessive constructions can be right-branched and leftbranched for declinable nouns (Ivan – second declension masculine, Sveta – first declension feminine) and only right-branched for non-declinable nouns (Nelli). Left-branching is extremely rare for inanimate possessors. (10)

kniga Ivan-a book Ivan-GEN or Ivan-ov-a kniga* Ivan-POSS.ADJ-F book ‘Ivan’s book’ (* archaic)

kniga Svety book Sveta.GEN or Svetin-a kniga Sveta.POSS.ADJ-F book ‘Sveta’s book’

kniga Nelli book Nelly – – ‘Nelly’s book’

In Baku Russian, in addition to the acceptability of the forms in the first row (which are the only option for inanimate possessors), there is a tendency of leftbranching of possessive constructions regardless of what kind of noun represents the possessor (note the change of the vowel in the feminine possessive adjective marker from to ). Left-branched nouns in the genitive do not occur in standard Russian. There are no semantic differences between the two options below. From the grammatical point of view, the possessive adjective option is more restricted (note that it is only used for first-declension nouns) and less likely to be extended by a modifier. (11) NGEN + N

2nd declension Il’gar-a žena İlqar-GEN wife

1st declension Faridy Fəridə.GEN or

mat’ mother

Non-declinable F Mehriban syn Mehriban son

Baku Russian: Colonial heritage in a post-colonial era | 45

POSS.ADJ + N –

– ‘İlqar’s wife’

Faridyn-a mat' Fəridə.POSS.ADJ-F mother ‘Fəridə’s mother’

– – ‘Mehriban’s son’

The rule does not apply exclusively to proper nouns: (12)

Brat-a kvartira brother-GEN apartment ‘The brother’s apartment’

(13)

Moej podrugi my.GEN female_friend.GEN ‘My friend’s husband’

muž husband

It is clear that this left-branching is triggered by possessive structures of Azeri (İlqar-ın arvad-ı, Fəridə-n-in ana-s-ı), which can never be right-branched. Finally, it is important to note that first-declension names stressed on the final syllable cannot derive possessive adjectives, hence: Lejly mat’ (‘Leyla’s mother’) and not *Lejlyn-a mat’; Gusejnagi mat’ (‘Hüseynağa’s mother’) and not *Gusejnagin-a mat’.

6.2 Discourse particles Among other key morphological features is the unique use of the particle dā (the bar above in this case marks vowel length). It should be noted that this particle in Baku Russian has been copied from Azeri, and hardly corresponds to any of the meanings of the homophonous Russian da. Broadly speaking, the Azeri particle expresses some form of required attention, but the study of the wide range of meanings conveyed by it should be undertaken as part of a semantic analysis of Azeri data. As for Baku Russian, it is worth pinpointing only selective meanings. Dā can be used for focalization (in this case, it is on the phrase pust’ snosjat’): (14)

Vse dumali äščšči gde-to tam na Bailovo all-PL think.PST.3PL ITJ somewhere there on Bayıl snosjat, pust' snosjat dā demolish.PRS-3PL let demolish.PRS-3PL FOC ‘Everyone was thinking, “Oh, whatever, they are demolishing [buildings] down in Bayıl; well, let them demolish them”’

It can also be used as a discourse marker.

46 | Murad Suleymanov

(15)

Ničego ne možem sopernik-u sdelat’ nothing NEG be_able.PRS-1PL rival-DAT there ‘There is nothing we can do to our rival, you see?’

dā DM

The use in (15) can indeed correspond to the Russian da (‘yes’), also used as a discourse marker to form tag questions. However, this is unlikely, given that questions in Baku Russian are marked by a very distinct intonation pattern inherited from Azeri, which was not the case in this or similar utterances. In addition, the discourse marker dā is used in Baku Russian much more frequently than da for tag questions in Russian spoken in Russia. This discourse marker is not necessarily used in the sentence-final position. (16)

Sejčas uže vremja ne to dā, naverno now already time NEG that.N DM perhaps ‘I guess the times are different now, you see?’

6.3 Complement and adverbial clauses The most interesting case is the use of the Russian complementizer čto, whose usage has been extended to fit the versatile usage of the Azeri subordinator ki. In the following examples from Baku Russian, čto is used as a quotative, which is not encountered in standard Russian: (17)

On skazal čto da konečno ja xoču igrat’ he say.PST.3SG.M QUOT yes of_course I want.PRS.1SG play.INF ‘He said: “Yes, of course, I want to play”’

(18)

Ja govorju čto možet ty ne xočeš’ I say.PRS.1SG QUOT maybe you.2SG NEG want.PRS.2SG ‘I tell him: “Maybe you don’t want to say it?”’

govorit’ say.INF

In standard Russian, čto would signal indirect speech. However, this is not the case in Baku Russian, as it becomes evident from the lack of tense or pronoun agreement. This usage reflects a similar areal phenomenon in Azeri, Persian, Armenian, Georgian, Ossetic, and East Caucasian languages. In line with this areal future, the quotative čto is used in cases where the sentence cannot even be reanalyzed as standard Russian indirect speech. (19)

On ves’ večer xodil obižennyj na I all.M evening walk.PST.IPFV.3SG.M upset on čto ty menja ne u-deržal QUOT you.2SG I.ACC NEG AW-hold.PST.3SG.M

menja I.ACC

Baku Russian: Colonial heritage in a post-colonial era | 47

‘All evening he was walking around upset with me (for the fact) that I had not held him’ Finally, the particle čto is used to introduce adverb clauses. (20)

Rolik eščjo o tom čto začem ne nado clip also about that.PREP.NF LW why NEG necessary pokupat’ produkciju [company name] buy.INF production.ACC ‘The clip is also about why one should not buy [company name] goods’

(21)

Do menja ne do-xodit to čto začem kogda until I.GEN NEG UP-go.PRS.IPFV.3SG that.N LW why when nužno so-brat’-sja dlja kakogo-to očerednogo necessary DF-take.INF-REFL for some.GEN another.GEN flešmob-a, to molodjož’ sobiraet-sja flashmob.GEN then youth with.take.PRS.3SG-REFL skoordinirovanno in_an_organized_manner ‘I fail to understand why, when there is a need to get together for some kind of flash mob, all the youths get together in an organized manner’

In the following example, čto can be treated either as a quotative or as a linking word signaling an adverb clause. (22)

Vsjo all.N

leto on summer he

vozmuščal-sja be_outraged.PST.IPFV.3SG.M-REFL

tem that.INS.NF

čto kak možno ljubit’ ēto how possible love.INF it.ACC ‘The entire summer he was outraged as to how one can like that’

LW

7 Conclusion The purpose of this article was to introduce the reader to a peculiar case of language contact that has been understudied despite constituting an inseparable and recognizable part of the everyday life and culture of a modern megalopolis. Baku Russian is not the only example of a colonial language variety that emerged in a multiethnic environment in the Soviet Union. This phenomenon was typical for urban areas where Russian was not an indigenous language.

48 | Murad Suleymanov

Other notable cases are Russian koinés of Odessa, Ukraine (relatively wellstudied) and Tashkent, Uzbekistan. Among the reasons for the neglect of Baku Russian in academic literature could be considered perhaps the fact that this language variety did not form until relatively late in the Soviet era, the fact that it arguably represents a case of creolization and not ‘authentic Russian’, and the fact that most speakers are not ethnic Russians, hence any variation can be written off as first-language influence or an imperfect mastery of Russian (in spite of this ‘influenced’ Russian often revealing itself, in fact, to be the speaker’s first language). Today Baku Russian continues to be an inherent cultural characteristic of the capital of Azerbaijan, albeit not to the same scale as before. With time, it is likely to see more and more influence from Azeri affecting its phonetics and grammar. Its social role seems rather stable and does not promise to undergo major change in the near future. Baku Russian represents a fruitful field for linguists, interested in language contact, creolistics, anthropology, as well as colonial and post-colonial studies.

Abbreviations6 1 2 3 ACC ADJ AW COMP DAT DF DM FOC GEN F IN INF INS IPFV ITJ

first person second person third person accusative case adjective ‘away’ complementizer dative case ‘down from’ discourse marker focalization marker genitive case feminine ‘into’ infinitive instrumental case imperfective aspect interjection

LW M N NEG NF ON PFV PL POSS PREP PRS PST PSV PTCP QUOT REFL SG UP

|| 6 Meanings of verbal prefixes in quotation marks.

linking word masculine neuter negation marker non-feminine ‘onto’ perfective aspect plural marker possessive prepositional case present tense past tense passive participle quotative reflexive singular ‘up to’

Baku Russian: Colonial heritage in a post-colonial era | 49

References Patriarch Alexius II of Moscow (ed.). 2000. Pravoslavnaja ēnciklopedija. [Orthodox encyclopædia]. Vol. 1. Moscow: Cerkovno-nauchnyj centr “Pravoslavnaja ēnciklopedija”. Altstadt, Audrey. 1992. The Azerbaijani Turks: Power and identity under Russian rule. Stanford: Hoover Press. Aslanov, Heydər. 1967. Zaimstvovanija iz azerbajdžanskogo jazyka v russkom ostrovnom govore [Azeri loanwords in the insular Russian variety]. Baku: ADK. Aşurbəyli, Sara. 1992. Istorija goroda Baku. [History of the city of Baku]. Baku: Azərnəşr. Bakinskij Rabočij. 24 August 1956, 198. 3. Bakıxanov, Abbasqulu. 1991. Golestân-e Iram. (translated into Russian by Ziya Bünyadov) Baku: Elm Baskakov, Alexander. 1977. O funkcionirovanii russkogo jazyka v Azerbajdžanskoj SSR [On the functioning of the Russian language in the Azerbaijan SSR]. In Ivan Bilodid & Fedot Filin (eds.), Russkij jazyk kak sredstvo mežnacional’nogo obščenija [Russian as a means of interethnic communication]. Moscow: Nauka. Bennigsen, Alexandre & S. Enders Wimbush. 1985. Muslims of the Soviet Empire: A guide. London: C. Hurst & Co. Publishers. Bérézine, Ilya. 2006. Voyage au Daghestan et en Transcaucasie [Travels to Daghestan and Transcaucasia]. Paris: Librairie orientaliste Paul Geuthner. Bromley, Julian. 1977. Sovremennye ētničeskie processy v SSSR [Contemporary ethnic processes in the USSR]. Moscow: Nauka. Budaqova, Zərifə, Şəmsəddin Sədiyev & Aydın Ələkbərov. 1977. Samoučitel’ azerbajdžanskogo jazyka [Self-instruction manual of Azeri]. Baku: Elm. Darieva, Tsypylma. 2011. A “remarkable gift” in a postcolonial city: The past and present of the Baku Promenade. In Tsypylma Darieva, Wolfgang Kascuba & Melanie Krebs (eds.), Urban spaces after socialism: Ethnographies of public places in Eurasian cities, 153–180. Frankfurt: Campus Verlag. Gmelin, Samuel Gottlieb. 1936. Putešestvie po Rossii dlja issledovanija trex carstv prirody [Journey through Russia to research the three kingdoms of nature]. In Vladimir Alexejev (ed.), Istoričeskie putešestvija. Izvlečenija iz memuarov i zapisok inostrannyx i russkix putešestvennikov po Volge v XV–XVIII vv. [Historical journeys. Excerpts from memoires and travel notes of foreign and Russian travelers along the Volga in the 15th–18th centuries]. Stalingrad: Kraevoe izdatel’stvo. Grenoble, Lenore A. 2003. Language policy in the Soviet Union. Dordrecht: Kluwer. Karavaev, Alexander. 2008. Russkaja reč’ i kul’tura v stranax SNG (na primere Azerbajdžana) [Russian speech and culture in the CIS countries, using Azerbaijan as a case study]. Research of the Eurasia Heritage Foundation. Information and Analytical Centre of the Lomonosov Moscow State University Department of History. Katz, Serafima. 1963. Leksika sela Slavjanki Kedabekskogo rajona Azerbajdžanskoj SSR [Vocabulary of the village of Slavyanka, Gədəbəy District, Azerbaijan SSR]. Candidate of Sciences dissertation. Kelner, Victor (ed.). 2003. Očerki po istorii russko-evrejskogo knižnogo dela [Articles on the history of Russian-Jewish book production]. Saint Petersburg: Rossijskaja nacional’naja biblioteka.

50 | Murad Suleymanov

Konstantinov, Vjačeslav. 2007. Evrejskoe naselenie byvšego SSSR v dvadcatom veke. [Jewish population of the former USSR in the twentieth century]. Jerusalem: Lira. Məmmədquluzadə, Cəlil. 1906. Bizim «obrazovannı»lar [Our ‘Obrazovanny’s]. Molla Nasraddin 2. Mikirtuni, Artur. 1952. Govory russkix pereselencev v Azerbajdžane. [Language varieties of Russian settlers in Azerbaijan]. Moscow: AKD. Miller, Boris. 1929. Taty, ix rasselenie i govory [Tats, their settlements and language varieties]. Izvestija obščestva obsledovanija i izučenija Azerbajdžana, #8. Baku: Izdatel’stvo obsledovanija i izučenija Azerbajdžana. “Naselenie Zakavkazskogo kraja” [Population of the Transcaucasian Krai] 1893. A set of statistics on the population of the Transcaucasian region, extracted from the 1886 family lists. Tiflis. Quliyeva, Lalə, Firavan Məmmədbəyli, Elvira Heydərova & Gülarə Kərimova. 2006. Slovar’ russkogo ostrovnogo govora Azerbajdžana [Dictionary of the insular Russian variety of Azerbaijan]. Baku: Baku State University. Shahnazarian, Nona. 2008. A gde že my svoi? Begstvo i migracii armjan Azerbajdžana [And where can we become ‘ours’? The flight and migrations of Armenians from Azerbaijan]. Vestnik Evrazii. Acta Eurasica. #2 (40). Moscow: Migrations and Diasporas in Eurasia. Shahnazarian, Nona. 17 November 2014. Internally Displaced Persons (IDPs) from Azerbaijan as Unwanted Co-Ethnics: Intra-group tensions, representations and integration of refugees in Armenia and de facto Nagorno-Karabakh Republic. Presentation. Ecole des Hautes Études en Sciences Sociales, Paris. “Soviet people”. 1983. In Alexei Rumjancev (ed.), Naučnyj kommunizm: slovar’ [Dictionary of scientific communism]. Moscow: Izdatel’stvo političeskoj literatury, 164. Turovsky, Rostislav. 1998. Kul’turnye landšafty Rossii [Cultural landscapes of Russia]. Moscow: Institut kul’turnogo i prirodnogo nasledija. Yagodynsky P. N. 1886. Ēkonomičeskij byt gosudarstvennyx krest’an južnoj časti Kubinskogo uezda [The economic life of state peasants in the Southern part of the Kuba Uyezd]. In Materialy dlja izučenija ēkonomičeskogo byta gosudarstvennyx krest’an Zakavkazskogo kraja [Materials for the study of the economic life of state peasants in the Transcaucasian Krai]. Vol. 2, Tiflis. Ždanko, Tatjana (ed.). 1990. Semejnyj byt narodov SSSR [Family life of the peoples of the USSR]. Moscow: Nauka.

Baku Russian: Colonial heritage in a post-colonial era | 51

Appendix Tab. 1: Ethnic composition of Baku in 1886–1979 Year

Azeris

Russians

Armenians

Ashkenazi Jews

Others

Total

1886

37,530

21,390

24,490

391

2,849

86,611

(53.0 %)

(24.7 %)

(28.3 %)

(0.5 %)

(3.3 %)

1897

40,341

37,399

19,099

3,369

11,696

(36.0 %)

(33.4 %)

(17.1 %)

(3.0 %)

(10.5 %)

1903 1913 1926 1939 1959 1970 1979

44,257

59,955

26,151

included

28,513

(28.4 %)

(38.5 %)

(16.8 %)

in ‘Others’

(18.3 %)

45,962

76,288

41,680

9,690

41,052

(21.4 %)

(35.5 %)

(19.4 %)

(4.5 %)

(19.1 %)

118,737

159,491

76,656

19,589

78,860

(26.2 %)

(35.2 %)

(16.9 %)

(1.0 %)

(19.3 %)

139,160

241,353

106,368

29,009

55,557

(24.4 %)

(42.2 %)

(18.6 %)

(5.1 %)

(9.7 %)

211,372

223,242

137,111

23,879

46,903

(32.9 %)

(34.7 %)

(21.3 %)

(3.7 %)

(7.3 %)

361,638

242,023

164,406

23,499

59,891

(42.5 %)

(28.4 %)

(19.3 %)

(2.8 %)

(7.0 %)

530,556

229,873

167,226

22,049

59,891

(52.4 %)

(22.7 %)

(16.5 %)

(2.2 %)

(7.0 %)

111,904 155,876 214,672 453,333 571,447 642,507 851,457 1,013,436

Eeva Sippola

Postcolonial language ideologies: Writing in Chabacano Abstract: This paper examines language ideologies in the Chabacano-speaking communities of Cavite City and Ternate (Philippines). Based on a corpus of written materials, we analyze how Chabacano speakers actively reshape their language and respond to changes in the postcolonial setting. The results show that Chabacano writing occurs in limited domains, and both etymological spellings and independent modifications are used. More generally, they show how the language endangerment situation of the varieties, the writers’ first language of literacy, and their language ideologies affect writing practices. The prestige of Spanish as the former colonial language is re-appropriated in the etymological spellings in order to express local identities and to challenge the endangerment situation. Keywords: Chabacano, writing, language ideologies, postcolonial

1 Introduction This paper examines language ideologies and practices in the written language of the Chabacano-speaking communities in Cavite City and Ternate, Philippines. Over the last decades, Chabacano has become an endangered minority language in these communities, mainly threatened by Tagalog and, to a lesser extent, by English (Lipski 1986, 1987, Fortuno-Genuino 2011, Lesho & Sippola 2013). Although a limited number of written texts have been produced during the history of the language, the Chabacano varieties are mostly used in oral communication in these communities. They are uncodified, and there is not a standardized writing system for them. The present study identifies and analyses writing practices in order to identify some of the key agents and linguistic forms shaping, and being shaped by, language ideologies in this postcolonial zone of contact (after Errington 2007: 165). Orthography is seen here as a set of linguistic practices that writers employ

|| Eeva Sippola: University of Bremen, Linguistics and Literary Studies, Postfach 330440, 28334 Bremen, Germany. E-mail: [email protected]

54 | Eeva Sippola

when representing a language without a writing standard (Sebba 1998: 19). Writing practices are always filled with social and cultural implications and tightly connected to the social and cultural contexts. By examining adapted spellings and orthographic choices, we can gain insights into the symbolic significance that is attached to alternative writing and literacy practices. A careful analysis of these choices and practices can offer valid social explanations for them at the individual, group, and societal levels in postcolonial situations. The materials under analysis include texts written in the two communities since 2000. They include language promotion and education materials, dictionaries, folk stories and personal stories, and rap lyrics that have been written by native Chabacano speakers.1 The materials are classified according to the text type and the spelling conventions adapted. This is done in order to reveal the different social meanings indexed by the choice and highlighting of particular features (see for example Mühleisen 2005: 6). This paper is structured as follows. First, the theoretical background section presents the connection between language ideologies, writing, and orthographic practices in postcolonial contact varieties. Second, we introduce the current sociolinguistic situation and examples of language shift in the Chabacano communities of the Manila Bay region. The main analysis focuses on the developments that are observed in the written use of the language. Special focus is given to the orthographic practices in these communities. The conclusions summarize the findings and connect it with the wider framework.

2 Language ideologies and writing in postcolonial contexts 2.1 Language ideologies and writing Language practices and ideologies, reflexive sensibilities about languages and language use, shape and are shaped by processes of historical and social change (Makihara & Schieffelin 2007: 4). In contact situations, speakers are often forced to actively reflect on their language use and practices and to negotiate their language ideologies. Their study is an ideal way to identify the conse|| 1 The total number of words in the sample is approximately 35,000 for Cavite Chabacano and 30,000 for Ternate Chabacano, excluding the dictionaries, which follow predefined orthographic choices.

Postcolonial language ideologies: Writing in Chabacano | 55

quences of cultural and linguistic contact in postcolonial situations. Here, language ideology is understood as a set of beliefs, values, and classifications of language use, values, and norms that have social meanings and are connected to moral and political interests (based on Silverstein 1979, Irvine 1989: 255). The ways we choose to make one way of speaking or writing representative is always shaped by broader factors and purposes, allowing for questions into what guides our strategies of selection, and whether or not these are conscious choices (see Errington 2007: 10). Language and linguistic ideologies are used as strategies for maintaining social power and domination. As with spoken language, ideologies about literacy and writing have social consequences. The colonial tradition saw civilization as founded in alphabetic literacy, and the definition of literacy in postcolonial societies is still a political matter today, closely linked to social control and domination (Errington 2007, Woolard 1998: 23, Woolard & Schieffelin 1994). The theoretical framework for our study is based to a great extent on the work of Mark Sebba, who has published extensively on the relationships between literary practices, orthography, and linguistic ideologies (for a summary, see Sebba 2007). In linguistics, orthography has been approached in two ways. An autonomous view of orthography sees the decisions made about it as ideologically neutral. Orthography is reduced to a mere technology, and the accuracy of spelling systems in representing the spoken sounds is central to this approach. Other important points to consider are the learnability and the supporting functions for transitional literacy in the communities that are choosing a spelling system to be adapted (Sebba 1998). In an ideological view of orthography, on the other hand, spelling is understood as a space to negotiate identity and power relations. In the ideological approach, the purpose of literacy and the status of languages are central. Moreover, characters or graphemes may have more relevant symbolic meaning than phonemic symbolic meaning (Sebba 1998: 20). Different types of variation, such as “wrong” spellings or spellings that represent non-standard pronunciations or archaisms, can be used as meaning creating entities (Sebba 2007: 32–34). More precisely, orthography is a social practice embedded in the social and cultural practices of the writers and speakers of the language: social and ideological issues regulate the choice of a particular written representation, and social meaning is created through orthographic choices in writing (Sebba 2000). Orthography itself is a symbol of normativization and representation of the linguistic norms that are closely related to linguistic policies and ideologies. Social and cultural factors shape the orthographic practices and the writing system of a language. These include literacy practices within the community, the nature of bilingualism among the literate members of the population, and

56 | Eeva Sippola

ideological beliefs, such as attitudes towards a language and its prestige in the community (Sebba 2000: 926). There is a great deal of variation in the types of written expressions or genres in different spaces around us. Some spaces of literacy are more regulated than others that allow a greater freedom of expression (Sebba 2007: 43–44). The fully regulated end of the continuum normally includes published texts and school writing, while advertisements, poetry, and personal letters represent more flexibility. Graffiti can be seen as the space of rebellion, where writing is performed as an act of resistance that breaks traditional classifications. In addition, the new media, such as the Internet and text messaging, also provide a large, unregulated space that is growing in importance in creole writing.

2.2 Creole orthographies Mühleisen (2005: 2–5) provides an overview of writing and codification in postcolonial contexts, with a special focus on creole languages. Creoles have a characteristic social history that is tightly connected to colonialism, or other unequal power relations. Most creole languages have no standard orthography and are not as extensively used in writing as their lexifiers or substrate languages. They often carry low prestige, as creoles typically exist in functional complementary distribution with the lexifier or other standard languages. They are commonly stigmatized for lacking the vocabulary for more formal registers or official uses; for this reason, the writing and codification of creoles is highly valued. These factors are naturally very heterogeneous and the use of a creole in writing ranges between standardized languages that are used in a variety of text types (e.g. Papiamentu, Bislama) and uncodified varieties that are only marginally used in writing (e.g. Papiá Kristang). According to Romaine (2005: 103), creole languages are often placed in a situation where autonomy or distance (Abstand) and development or elaboration (Ausbau)2 are in constant negotiation between the European colonial lexifier and the autonomous creole code. In situations where creoles are in contact with their lexifier, they can only achieve linguistic autonomy by becoming less like the European colonial language, but paradoxically, they often develop a standard that is modeled on the colonial language (Sebba 1998: 22). Standardization processes or orthographies adapted for creole languages do not follow a

|| 2 Notions from Kloss (1967).

Postcolonial language ideologies: Writing in Chabacano | 57

uniform pattern, but some tendencies can be observed. For English-lexifier creoles, phonemic orthographies are used only in the Pacific (e.g. Tok Pisin) and in Africa (e.g. Sierra Leone). Elsewhere, English-lexifier creoles are generally written using the conventions of the lexifier, although expert recommendations might be different (Romaine 2005: 135; for recent developments, see Deuber & Hinrichs 2007). For a large number of the French-lexifier creoles, on the other hand, phonemic orthographies are used. The most famous example is the Haitian orthography, which departs from the etymological model to express a creole identity distinct from French. However, the ideological background of these standardization choices is nonetheless complex, and the proposed orthography is not always accepted and used by the whole community (for the Haitian case, see Schieffelin & Doucet 1998: 306). Winer (1990: 243) suggests that the French creoles often distinguish more clearly between the lexifier and the creole variety, a difference also perceived by their speakers, which would make the adaptation of a distinct orthography easier than in the case of English-lexifier creoles, which often exist on a continuum with Standard English. As for Spanish- and Portuguese-lexifier creole languages, the case of Papiamentu is interesting. Papiamentu is the native language of the islands of the Netherlands Antilles, Aruba, Bonaire, and Curaçao. Papiamentu has a particularly high level of prestige for a creole language, as it is spoken by all social classes in various domains and it is widely used in the media. There are several daily newspapers and other regular publications in Papiamentu, including some published in the Netherlands for the West Indian community in Europe (Kouwenberg 2013). Papiamentu is also one of the few creoles that have a written form that is widely accepted and used by the speakers. It has two different spelling systems. One of these is used in Aruba, and it follows a more etymological tradition that reproduces many Spanish spelling conventions. The other system, which is more phonemic, is used on Bonaire and Curaçao. For example, in Aruba /k/ is for words of Dutch origin, but or for words of Spanish origin. In Bonaire and Curaçao /k/ is consistently . Sebba’s (1998: 24) explanation for these differences is based on the social and political context: Aruba is the closest island to Venezuela, where the language is standard Spanish, and the Papiamentu variety spoken in Aruba has more Spanish influence than the language of the other islands. Consequently, the speakers in Aruba have decided to emphasize the relationship with Spanish by adopting a distinct and characteristic spelling system for their variety. Regarding Chabacano and the situation in the Philippines, standardization and orthography have been addressed only in a limited number of articles, mostly concerning the southern variety spoken in Zamboanga. As the most vital

58 | Eeva Sippola

and largest community, without an immediate threat of endangerment, Zamboanga Chabacano orthography is not directly included in our study, but it provides an important example of spelling practices in the Philippine context. Forman (2001) discusses some problems stemming from favoring the Spanish orthography in Zamboanga Chabacano. Zamboanga Chabacano serves as a lingua franca in the Zamboanga region, and it is learned as a second language, mainly by immigrants from other parts of the Philippines. Favoring an etymological orthography presents a possible threat to the promotion and preservation of the language in its written form, as this makes it more difficult for the immigrants to learn. In addition, Forman raises an important question about colonial categorizations in labeling and analyzing Zamboanga Chabacano, using the example of distinguishing between “local”, Southern Philippine words and words “of Spanish origin”, which is a topic Zamboanga Chabacano speakers often mention. Logically, all Chabacano words used today in Zamboanga are local (Forman 2001: 101). Recently, a five million-word corpus of Zamboanga Chabacano was collected from both oral and written sources to produce the Chabacano Reader, language learning material directed primarily to second language learners (Tardo 2006, Miravite et al. 2009). The authors note that the spelling of Chabacano is inherently variant, and therefore problematic. For the language name itself, there are five different ways of spelling it: the hybrid and , Spanish , Filipino , and , which is often a result of a typo or the name of a fruit. Based on a Google search, the most common of these is Chabacano, closely followed by Chavacano. Naturally, a Google search is not limited to spellings by first language speakers, but it might also reflect academic preferences in spelling the language name. In the corpus collected in Zamboanga, these two forms occur in very similar numbers (Miravite et al. 2009: x). For the Chabacano Reader, which consists of a number of texts, their translations, and a vocabulary, the spelling of Chabacano words was modified and regularized to reflect Philippine orthographies (including Tagalog/Filipino, Cebuano, Hiligaynon, and Tausug). The authors explain their motivation for this choice with ideological and practical claims. First, Chabacano is characterized as “truly a Philippine language” (Miravite et al. 2009: ix). Second, the phoneme-grapheme inconsistencies of the Spanish spelling system are presented, such as for both /k/ and /s/, or and for /b/. These are treated as problematic for local authors who may not be fluent in Spanish. Finally, it is assumed that speakers do not have metalinguistic awareness of the etymological origin of the lexical items, which would make the adaptation of etymologically based spelling conventions highly challenging (Miravite et al. 2009: ix).

Postcolonial language ideologies: Writing in Chabacano | 59

3 Chabacano-speaking communities of the Manila Bay region The Philippine Creole Spanish dialects are known collectively as Chabacano. They are widely recognized as creole languages different from Spanish, not mutually intelligible with the lexifier, and containing particular grammatical structures and lexicons. Chabacano varieties have historically been spoken in Cavite and Ternate in the Manila Bay region, and in Zamboanga, Cotabato, and Davao in Mindanao in the southern Philippines.3 Even if the varieties are mutually intelligible to a great degree, there are substantial grammatical and lexical differences between them, due to the different processes in the settlement and formation of the communities. In addition, the languages are mostly used in local contexts, and there is little contact between the speakers of the different varieties, even in the Manila Bay region (Lesho & Sippola 2013: 2). Previous sociolinguistic work on Chabacano (Lipski 2001, 2010; Lesho & Sippola 2013, 2014) has unveiled the endangered situation of the Manila Bay varieties, as well as the complex relations to the former colonial languages, Spanish and English, and the Philippine languages that form the substrate and adstrate of these varieties. Furthermore, several historical and prestige dimensions are at play, as the lexifier language is no longer actively spoken alongside the creole, but the new prestige languages Filipino and English are present. In the following sections, we will offer a closer look into the situation and provide background information for the study of the writing practices in the Chabacanospeaking communities threatened by language endangerment.

3.1 The postcolonial context in the Philippines The Philippines is an ethnically diverse and highly multilingual country, where over a hundred local languages and two official languages, English and Filipino, are spoken. Filipino is a standard register of Tagalog, the main language of the Manila region, and it is used as the national language of the Philippines. English is mainly the language of higher education, business, and the media, while Filipino is generally employed for local communication, certain school

|| 3 There are also mentions of Chabacano spoken in the Ermita district of Manila (see Fernández 2012).

60 | Eeva Sippola

subjects, and entertainment. Local languages, such as Chabacano, are used for daily communication and in the home (Gonzalez; 1998: 503, 2003: 3). The Philippines never experienced a real mestizaje (Spa. ‘mixture’, which refers to biological and cultural blending) due to the low number of Spaniards on the islands, as the colony was far away from Spain. It was governed through Mexico, and local governors applied Spanish rule on the islands (Quilis & Casado-Fresnillo 2008: 55, 59). Spanish never had the same diffusion nor became as widely used by the Philippine population as it was in the Americas. During the colonial period, the language of administration was Spanish, but the crown and the religious orders, with the exception of the Jesuits, were against teaching Spanish to the natives and used local languages in their missions (Quilis & Casado-Fresnillo 2008: 62–66; Steinhauer 2005: 75). At the end of the 18th century, Spanish began to be used in the school system, but it is obvious that education was a privilege of the local elites, and Spanish was never widely used among the multi-ethnic population (Lipski 1987:123, Quilis & CasadoFresnillo 2008: 66). Despite the limited number of speakers, or perhaps due to it, Spanish had a function as a marker of social position among the speakers of Tagalog, and possibly also other groups. But as access to Spanish was so limited, it was probably used with influence and mixing from the other languages present in the multilingual, colonial society (cf. Wolff 2001: 234). However, the number of Spanish loanwords in the Philippine languages is relatively high, as Spanish was used as the lingua franca of the colonial powers in the region. Lesho & Sippola (2013: 03) explain that in the 20th century, local languages in the Philippines, including Chabacano around Manila Bay, began to lose speakers or become increasingly restricted to private domains. These changes were largely caused by the presence of the Americans and the implementation of English, as well as the rise of Philippine nationalism promoting Filipino. During most of the first half of the 20th century, English was the language of instruction in public schools. At the end of the Spanish period, the number of English speakers was almost non-existent, and only 2.6 % of the population spoke Spanish (Gonzalez 1998: 495). However, by World War II, already one fourth of the population spoke English. Initially, local languages were intended to be taught alongside English, but for practical reasons, this proved difficult. In addition, there was a colonial ideology that stigmatized local languages labeling them as unsuitable for knowledge (Thompson 2003: 20, Lesho & Sippola 2013: 4). Spanish influence gradually waned over the course of the century. Spanish was no longer required in Philippine schools after World War II, but students at the university level were required to take two to four years of the subject until 1987 (Lipski et al. 1996: 274). Today, Filipino and English continue to be the main languages of literacy and instruction in the school system, and they are also taught as subjects in their own right beginning in the early grades (Lesho & Sippola 2013: 4).

Postcolonial language ideologies: Writing in Chabacano | 61

3.2 Sociolinguistic situation of the Chabacano-speaking communities in Cavite City and Ternate Today, Chabacano is a minority language in Cavite City and Ternate, the two Chabacano-speaking communities in the Manila Bay region. The creole varieties are in competition with the official languages, Filipino and English, which both enjoy a high social status in the current climate and are instrumental for social advancement. There is very little contact with the largest Chabacano-speaking community in Zamboanga, in the southern Philippines. Cavite Chabacano is highly endangered, and according to recent estimations, there are fewer than 4,000 speakers in Cavite City (Lesho & Sippola 2013: 23). The community is undergoing a strong shift to Tagalog, and there are very few young speakers left. On the other hand, Ternate Chabacano is in a relatively safe situation. There are only around 3,000 speakers of Chabacano in Ternate, but it is spoken by all generations in several contexts with unbroken transmission in a large part of the community. Children learn to speak Chabacano as a first language at home, and in some cases Tagalog is only fully introduced when starting school. Despite this, the effects of language shift are noticeable, and a growing number of children are becoming dominant in Tagalog (see for example Lesho & Sippola 2013: 9). Language attitudes in Cavite City and Ternate have been shown to be twofold, mirroring and reproducing both negative and positive ideologies (Lesho & Sippola 2013, 2014). The Chabacano language has symbolic value in both communities. It is often used in ceremonial contexts and is considered part of the local identities, strongly associated with being an authentic resident of Cavite City or Ternate. The speakers also indicate pride in their Spanish heritage. At the same time, Chabacano is often seen as “broken Spanish”, and the language of the poor in Ternate, which contributes to the negative language attitudes. Lesho & Sippola (2014) have studied the folk perceptions of variation among the different Chabacano communities. Using methods from perceptual dialectology, they are engaged in exploring the complex sociolinguistic dynamics in these communities. Speakers distinguish between the Chabacano varieties based mostly on lexical and phonological differences. These linguistic differences are attributed to each variety’s perceived closeness to Spanish or to the adstrate Philippine languages. Linguistic variation is attested by the speakers in the areas of prosody, lexicon, personal pronouns, mid vowel raising, and generational differences, among others. Chabacano speakers do not comment on variation between /y/ and /ʎ/ or /lj/, or the occurrence of /h/, which are often mentioned by linguists as features that differentiate the varieties spoken in Cavite and Ternate. Some of this variation is also relevant for writing practices, as we shall see later. The lan-

62 | Eeva Sippola

guage ideologies that emerge when analyzing the prestige attribution and the relationships to the superstrate and adstrates show that Spanish is perceived as a prestige variety. Moreover, language shift to Philippine languages is perceived as a threat to the Chabacano varieties (Lesho & Sippola 2014).

3.3 Language shift and change in progress As mentioned in the previous section, there is a growing concern for the future of Chabacano, especially in Cavite City, and to some extent also in Ternate. Older speakers and language activists especially express this concern. As shown in Lesho & Sippola (2014: 26–27), perceptions of generational differences normally focus on the growing Tagalog influence in the speech of youngsters and children, or the language shift to Tagalog and English. There is an idea of “pure” uncontaminated speech, even if all the speakers are multilingual and code-switch extensively between two or more languages in their everyday communication. For example, language activists and specialists, who often represent local cultural elites, occasionally incorporate Spanish elements into their language, instead of the generally used English or Tagalog terms (for example, in school and political domains). Some linguistic elements are connected with the change in progress. This change is towards Tagalog, and it often mirrors similar processes or developments in it. It should be noted, however, that no sociolinguistic studies following the variationist paradigm have been conducted on Chabacano, and the processes described here are mostly areas of grammar that show significant variation and point towards particular tendencies. The list presented in the following paragraphs is by no means exhaustive, and due caution should be exercised with the interpretation. It should also be noted that several different codes could have existed during the history of Chabacano. Some of them might have born a closer resemblance to Spanish, while others might have contained more Philippine elements, depending on the context of use. This kind of variation has also been attested in other creole communities (see for example Hinrichs & Farquharson 2011). At the lexical level, the variation is mostly domain related. All generations use frequent loans from English and Tagalog when talking about school, sports, popular culture, and TV, but the influence from these prestige languages is especially noticeable in the speech of the younger generations. Examples include mayor, nurse, driver, writer; grade, high school, university; basketball, beach, cellphone, sé text ‘to text’, drugs … etc. Older speakers often incorporate

Postcolonial language ideologies: Writing in Chabacano | 63

Spanish terms, especially in formal domains, while younger speakers resort to Tagalog items. Another particular phenomenon in older generations is frequent self-correction when using Tagalog words. This happens especially when talking with other language activists or linguists, and is not as frequent in everyday conversations with peers. At the phonological level, the change in progress in Ternate can to some extent be seen in the palatalization of the stops /k/, /t/, and /d/, and the fricative /s/, before /j/. In the grammatical elements and structures, variation that might be an indication of change can be observed in the use of possessive pronouns and relative particles. Other areas of interest include word order, TMA-marking, topic and object marking, and reciprocal constructions. Overlapping with lexical variation, the variation in the use of conjunctions and indefinite terms is salient. This typically means that young speakers favor Tagalog elements over the Spanish ones. Unexplored areas include several grammatical elements, variation in the vowel system, and prosodic features (for Cavite Chabacano, see Lesho 2013). Since it is very difficult to find younger speakers of Cavite Chabacano, we will exemplify the grammatical variation in the use of possessive pronouns in Ternate Chabacano. Adnominal possessives in Ternate Chabacano have a simple form, as for example mi ‘my’ and a compound form, which is constructed in the same manner as the possessive construction with the preposition di ‘of’, as in di-mi ‘my’ (for a detailed description of the paradigm, see Sippola 2011: 126). The simple forms mi ‘my’, bos ‘your’, and su ‘his/her’ exist for the singular series. In the plural series, only compound forms with di are used. Both forms occur preposed to the noun, as in (1). Both the simple and the compound form can occur with the determiner (k)el and the plural marker manga. The compound form can also occur after the noun, as in (2), without a difference in meaning (Sippola 2011: 126). (1)

Mi nay din tasé tyénda kel di-éli my mother ENC IPFV.make shop DEF of-3SG ‘My mother sells that fish of hers/his.’

(2)

Koryósu tamyén el good also DEF ‘Their life is also good.’

bída life

kohída. catch

di-lótru. of-3PL

This free variation in possessive construction types is also well known in Spanish, where the genitive phrases occur in postnominal positions substituting for the possessive adjective (see for example Penny 1991: 143–144). In a similar manner, Tagalog has possessives modifying the noun in prenominal and postnominal positions (Schachter & Otanes 1972: 135). In Ternate Chabacano,

64 | Eeva Sippola

double use of the possessives is sometimes observed, as in (3). This kind of rare use of possessives shows how two different systems are interacting and can be combined in a language loss situation. (3)

Ta-yubá

lótru lús na di-lótru 3PL light LOC of-3PL ‘They carry light(s) on their boat(s).’ IPFV-carry

su bánka. their boat

It seems that the young speakers especially use more compound forms in prenominal positions, and older speakers mirror the lexifier construction type more closely. The variation can also be due to differences between more formal, and Spanish-like, registers and the more informal registers, which often have more incorporation of Tagalog elements and frequent code-switching between Tagalog and Chabacano. Both forms, except for the double use of possessives, occur in written samples (Nigoza 2007, 2013). The use of simple forms is, however, much more prominent. The use of relative particles seems to follow similar patterns. The use of a relativizer is optional in Ternate Chabacano, as in (4), but the relativizers kel, ke, ki, and kyen ‘who’ are occasionally observed. The relative clauses are normally positioned after the nucleus, but sometimes preposed clauses mirroring Tagalog constructions can also be observed, as in (5). (4)

Kel mi íha [ya-kába parmaséwtika] my daughter PFV-finish pharmacy ‘My daughter who finished Pharmacy [studies]’. DEF

(5)

Ésti~ésti kel

[ta-buská mótru] iglésya. IPFV-search 1PL church ‘This one is the church we are looking for.’ DEM.RD

DEM

As shown in Sippola (2009), the relativizer occurs more commonly in formal, written registers than in speech. The current published materials on Cavite Chabacano (Escalante 2005, 2010), which have been prepared mainly for language preservation purposes, do not take the observed variation into account. The choice of the use of the relativizer is stylistic, promoted by language activists, who prefer the Hispanic standard for language promotion materials.

Postcolonial language ideologies: Writing in Chabacano | 65

4 Responses to change in the written domain 4.1 Written materials In recent years, there have been more standardization and preservation efforts in Cavite City than in Ternate, as the community is facing a pressing need to act as a response to the urgent endangerment situation. Recent developments show that the speakers of Cavite and Ternate Chabacano are taking advantage of new social domains and local media, such as the Internet, text messaging, local newspapers, and rap lyrics, which further contributes to preservation efforts. Efforts to include Chabacano in the school curriculum as a mother tongue or a heritage language have also been undertaken (Lesho & Sippola 2013: 17–18). For Cavite Chabacano, more documentation and materials are available, including printed and online materials, such as newspaper columns, stories, poetry, language materials for educational purposes (Escalante 2005, 2010, Sippola 2010), and dictionaries. For Ternate Chabacano, such materials are not widely available. There are collections of stories (Ocampo 2007, Nigoza 2007), a language material leaflet for educational purposes (Nigoza 2013), and occasional texts, lyrics, and private letters. Most materials for educational purposes have been produced in Cavite City, and they have been used for heritage language learning. When looking at the response to new domains and media, there are several Internet pages by language activists and heritage language speakers, and there are Facebook groups for both communities. In addition, Ternate Chabacano speakers occasionally send text messages and chat online in Chabacano (Lesho & Sippola 2013: 16).

4.2 Orthographic practices Although the aim of this paper is not to study the phonemic accuracy of the orthographic practices, it is useful to be familiar with the phoneme inventories of the varieties before having a closer look at the different symbols and systems used in Chabacano writing in these communities. Chabacano varieties have a 5vowel system, including /i/, /e/, /a/, /o/, and /u/. Table 1 gives an overview of the consonant phonemes.

66 | Eeva Sippola

Tab. 1: Consonant phonemes in Cavite and Ternate Chabacano (Ramos 1963, Sippola 2011, Lesho 2013)4 Bilabial Stop p b Nasal m Trill Tap Fricative Affricate Approximant Lateral approx.

Dental/Alveolar Post-alveolar t

d n (r) ɾ

Palatal (ɲ)

Velar

Glottal

k ɡ ŋ

s

(ʔ)

h tʃ (ʤ) l

j (ʎ)

w

Both the Spanish and Tagalog orthographies can reproduce the majority of the sounds in Chabacano. The phonemic status of /r/, /ɲ/, /ʎ/, and /ʔ/, which have a very limited distribution in Chabacano, has received varying treatments by different authors. Lesho & Sippola (2014: 12) explain that this is due in part to the analysis, not actual differences in the production. For example, the Spanish palatals /ʎ/ and /ɲ/ are sometimes analyzed as phonemes (Lesho 2013 for Cavite Chabacano) and sometimes as the consonant clusters /lj/ and /nj/ (Ramos 1963 for Cavite, Sippola 2011 for Ternate), but their distribution is basically the same across varieties. Similarly, Cavite Chabacano is the only variety described as retaining the Spanish distinction between /ɾ/ and /r/ (Ramos 1963, Lesho 2013), but it is unclear that the number or distribution of rhotics actually differs across the Chabacano varieties. There is also inconsistency in the descriptions in how to classify sounds of non-Spanish origin, as for example the glottal /Ɂ/, which occurs in words of Tagalog origin.

4.2.1 Alphabet and grapheme choices Besides examples of spelling that can be analyzed directly, the materials produced for educational purposes offer metalinguistic comments about the orthographic choices in them. These kinds of comments provide us with information about the conscious choices of the writers with regard to orthography. In the preface of the Diccionario Chabacano (Asociacion Chabacano del Ciudad de Cavite 2008), the authors explain that Cavite Chabacano is “a vernacular that resulted from the contact of Spanish and Tagalog. It is a spoken language and

|| 4 The phonemic status of the symbols in parenthesis has not been established conclusively.

Postcolonial language ideologies: Writing in Chabacano | 67

has no established writing system.” The system adopted in the dictionary utilizes the Spanish alphabet with five vowels (, , , , ) and twenty-five consonants (including the graphemes , , , , which have no phonemic correspondences in Chabacano, and others, such as for /ɲ/ and for /r/, whose phonemic status is debatable). The authors have decided to include the Tagalog with the Chabacano alphabet since some Tagalog words with are found in the Chabacano lexicon. The pronunciation of each letter is given in the preface. Table 2 lists the items given in the dictionary that have an ambiguous grapheme-phoneme relation. Tab. 2: Grapheme-phoneme relations in the Diccionario Chabacano Del Ciudad de Cavite Grapheme

Pronunciation [IPA value]

C

/s/ before /e/, /i/ /k/ before /a/, /o/, /u/ /ch/ [/tʃ/] /g/ before /a/, /o/, /u/ /h/ before /e/, /i/ (exception /gisa/ given in the dictionary) /h/ or silent /ly/ [/lj/] or [/ʎ/] /ng/ [/ŋ/] /ks/ /y/ [/j/] /s/

Ch G

H Ll NG X Y Z

The educational materials in our sample have very similar explanations about Chabacano in Cavite City being an oral language, and that “Spanish and Tagalog-based words are spelled closely to the original, although the pronunciation might be different” (Escalante 2010: 1). For example, Escalante (2010: 2) has chosen to present the alphabet and the sounds of Chabacano using English as an intermediate language and a starting point. is mentioned as the Tagalog letter pertaining to the Chabacano alphabet, and , , , and as the four Spanish letters, which together with the 26 English letters, make the Chabacano alphabet. Variation is explained for the grapheme , which is pronounced as /e/ or /i/ depending on the position, as /o/ or /u/, as /k/ or /s/, as “hard g” /g/ or /h/, as mute in Spanish-origin words and as /h/ in Tagalog-origin words, in Tagalog words, and as /ku/. and , and and are presented as interchangeable consonant spellings, as are and in final positions (Escalante 2010: 11–14).

68 | Eeva Sippola

Quite differently, the materials produced in Ternate (Nigoza 2007, 2013) have only sporadic comments about the spelling and the symbols used. Nigoza (2007: 83–84) mentions that: “G is added to Bahra [Ternate Chabacano] words ending in n. However, some pronounce the word with the g ending. Others pronounce the word as it is.” The Spanish grapheme is said to be written as in Ternate Chabacano. Variation between and and between and , as well as the addition of syllable and word final , are also mentioned.

4.2.2 Writing in Chabacano The actual orthographic practices or ways of writing the creole can be studied directly from written samples, which give access to the conscious and unconscious choices of the writers. The lexicon in our sample can be classified into different groups, based on the use of spelling and grapheme selection as well as the modifications that have been adapted. Both varieties are presented together here, as similar classifications can be observed in them. Table 3 presents spellings that reproduce the Spanish or the Tagalog spellings, with the only exception of leaving out the Spanish accent marks.5 Naturally, we cannot always establish the origins of all the words, and some might be local innovations. For example, one of the most frequent items in Chabacano is the preposition /na/, which is written as or as by some writers in Ternate Chabacano. The origin of is unclear, but the final has a special function in the writing of Ternate Chabacano, as we shall see later. Tab. 3: Etymological spelling conventions Spelling

Pronunciation

Meaning

yo (Spanish) maestra (Spanish) mi (Spanish) miel (Spanish) lugar (Spanish) para (Spanish) aniversario (Spanish) ciudad (Spanish)

/jo/ /maestra/ /mi/ /miel/ /lugar/ /para/ /aniversario/ /sjudad/

1SG teacher (female) 1SG.POSS honey place for anniversary city

|| 5 The tables only include selected examples, and the words presented are only a small subgroup of all the possible words in the categories. Alternative spellings are very common across the lexicon, as shown for Zamboanga Chabacano in Miravite et al. (2009).

Postcolonial language ideologies: Writing in Chabacano | 69

Spelling

Pronunciation

Meaning

pati (Tagalog) handa (Tagalog) nga (Tagalog)

/pati/ /handa/ /ŋa/

and prepared food emphatic clitic particle

Table 4 presents spellings that differ from the original to some degree. This group includes mostly Spanish-origin words, but also some of Tagalog-origin. Interestingly we find a greater number of words in which the function and meaning differ from the original in this group. Probably the greater degree of adaptation is motivated by the frequency of the items as well as their functional load. Most of the alternative spellings reproduce variation in the pronunciation or etymological graphemes that have an ambiguous sound correspondence in Chabacano. Tab. 4: Adaptations in Chabacano spelling Orthographic variation

Spellings and meaning

Pronunciation

vowel variation in final positions

esti ‘this’ [esti] grande, grandi ‘big’ [grandi] qui, que ‘that’ [ki] debi ‘must’ [debi] [di] de, di ‘IRR’, ‘PP’ vowel variation in other positions beng, bung, ung, bien ‘very’ [bung] vowel variation, deletion of a pey, pwedi, puede ‘can’, ‘be able’ [pey], [pwedi], [pwede] syllable deletion, unstressed vowel, haci, hace, se ‘make’ [(a)se] / deletion, vowel variation nway, nuway, no hay ‘NEG.EX’ [n(u)way] /, vowel variation con, cun, cung, kung, ung ‘OBJ’, [kun] ‘with’ / cabando, cabandu, kabando, [kabando], [kabandu] kabandu ‘after’ (sequential marker) / chene, cheni, tiene, tieni, tyene, [tjene] tyeni ‘EX’ / / kyeri, kyere, quiere, quieri, cheri[cheri], [kjeri], [kjere], ‘want’ /chere/ / quatru, cuatro ‘four’ /kwatru/ / quel, kel, quiel ‘DEF’ /kel/ /beldaj/

velday ‘truth’ / buscavida, buskabida ‘a living’ /buskabida/

organiza ‘organize’ /organisa/

When comparing different material types and the orthographic choices in them, the results show a clear division in the writing practices. In the materials pro-

70 | Eeva Sippola

duced in Cavite City by Cavite Chabacano speakers, etymological orthography, as presented in Table 3, is clearly preferred in all types of written texts. The only notable exceptions to the preference of the Spanish etymological spelling are the English adaptations. This is the case for newly introduced items referring to modern technologies, and for items that are (partially) shared between Spanish and English, such as nation, association, difference/diferencia, etc. In addition to this clear tendency, some local adaptations are included. One of these is variation in unstressed vowels. Example (6) presents a contemporary text written in Cavite Chabacano. Here our attention focuses on the use of the Spanish graphemes , , , , and spellings that represent the local pronunciation of the vowels in Spanish origin words. These include, for example, (< Spa. con) and (< Spa. este). As can be seen in , words that closely resemble each other in English and Spanish are often spelled the English way, as speakers are more commonly bilingual or fluent in English than in Spanish. This might of course also be due to the English spelling programs of the text processors used in the Philippines. (6)

Extract from a newsletter message in Cavite Chabacano Mensaje Felicidades cun el grupo del Los Chabacanos alli na San Diego na occasion del pang-35 aniversario de ustedes. Saludo tambien cun Nina Anson, Presidente de ustedes, na su liderato y buen corazon. Velday qui debi niso ayuda cuno y otro para preserva y propaga el Chabacano de niso ciudad. Quieri tambien yo dale sabi cun ustedes cun el favor de Dios, ya publica ya niso asociacion el Diccionario Chabacano del Ciudad de Cavite, el primero proyecto qui ya pricura nisos haci cuando ya organiza niso esti asociacion. Otravez, felicidades cun todo ustedes! ‘Message Congratulations to the Los Chabacanos group in San Diego on the occasion of your 35-year anniversary. Greetings also to Nina Anson, your President, on her leadership and good heart. It is true that we have to help each other to preserve and promote the Chabacano language of our city. I also want to tell you that, with the help of God, our association published the Diccionario Chabacano del Ciudad de Cavite [Dictionary of Cavite City Chabacano], the first project we attempted to do when we organized this association. Again, congratulations to all of you!’

Postcolonial language ideologies: Writing in Chabacano | 71

In Ternate, on the other hand, we observe much more variation in the writing practices. Etymological orthography is the most common spelling adapted in published materials and in private messaging of people who are 40 and older. Younger writers employ the Tagalog spelling system (with the systematic inclusion of 6). A number of features that reflect the local pronunciation are reproduced in the written form as well. These include vowel variation, as in Cavite, the final , and syllable final . As in Cavite, there are adaptations from English in all age groups. In the published Ternate Chabacano materials, there is a tendency towards etymological orthography, but local particularities are accentuated, as can be seen in (7). In this sample, our attention is drawn to the use of the Spanish graphemes , , , and as in ‘person’. Local traits are the final in and the final nasal in ‘also’ and ‘harvest’. is also used preposed to in ‘to catch’. (7)

Extract of a story about local history in Ternate Chabacano (Nigoza 2007: 6) Cuando quel, esti cañaduci, una di sistema di buscavida aqui nah Bahra. Tiene plantaciong di cañaduci aqui nah Casuyang. Si tiempo di anihang, aqui nah Tarapichi ta mole para se asucar. Quel otro cañaduci ta yuva nah asucarera central nah Nasugbu nah Batangas. Ta se embarca esti nah grande trac para vende a ya. Bueno tamieng quel dueño di questi Tarapichi ta dali pruba cung manga jente quel miel y balicocha ta hace lotru a ya. Quel manga jente, como sabi lotru qui tiene ty mole cañaduci aya, di inda lotru para ingsangga. Pila-pila lotru ta ingsangga miel. ‘During those days, sugarcane planting was among the occupations here in Ternate. Here in Casuyang, there were sugarcane plantations. In harvest time, sugar was produced here in Tarapichi. Other harvests were brought to the sugar central in Nasugbu, Batangas. These were boarded on large trucks and sold there. The owner of this Tarapichi was so kind giving honey and candies that were produced there to the people. The people, as they knew that sugarcane was processed there, they went to get it in queues, to get the honey.’ [My translation. For a free translation in English, see Nigoza (2007: 18).]

|| 6 and are sometimes included in Tagalog alphabet, although they are used only in foreign origin words.

72 | Eeva Sippola

4.2.3 Age and identity in spelling Due to the age and education level of many writers, Spanish spelling is more accessible and more easily reproduced in writing in Cavite City. In Ternate, only older speakers use Spanish spellings. These speakers typically studied Spanish at school or in college, and grew up during the time when Spanish was an official language in the Philippines. In texts written by speakers who are younger than 40 years of age, phonemic spelling or Philippine spelling conventions are used. This is a practically motivated choice, as younger speakers in Ternate today have only limited access to texts written in Spanish. The young Chabacano speakers have Tagalog as their first language of literacy, accompanied by English, which is taught at school from an early age. An example of this is presented in (8). In this example, only the diverges from ordinary Tagalog writing. Other local features also observed are the mid vowel raising and final in ‘OBJ’ and ‘nation’. (8)

Extract from a personal story in Ternate Chabacano

‘I want to finish studies as teacher, to be able to help my mother to have an easy life, to be able to help the boys and girls to write and read, I want to make a big shop, to be able to build a big house, I want to go to another country to visit beautiful places.’ In addition to this writing sample, which reflects a typical production of a young Ternate Chabacano speaker, we find written use of the language attached to new language practices by young speakers. In recent years, young Ternate Chabacano speakers have started to expand the use of the language to new domains, including not only Internet and text messaging related uses but also more traditional domains, such as music. Rap is a popular music genre in the Philippines today, and its popularity has also reached Ternate. As a consequence of their own activity and language promotion programs at school and in the municipality, Chabacano is occasionally used in rap lyrics, especially in peer meetings, local music contests, and town fiestas. Rap lyrics are often draft-

Postcolonial language ideologies: Writing in Chabacano | 73

ed on paper or in an electronic format on mobile phones, before the performance. In the samples we have access to, there is a slightly higher number of etymological spellings than in the written texts by other younger Ternate Chabacano speakers. For example, there is an occasional use of , as in /kjen/ ‘who’ and /kel/ ‘DEF’, as well as as in /iho/ ‘son’. However, there are words where /k/ is written as and /h/ as , as in Tagalog. Local features such as in and /bahra/ ‘Ternate’ and the use of apostrophes in and is also interesting; the latter might be a way of marking glottal stops. The marking of local features can be explained by the fact that rappers see themselves as promoters of local language and history.

5 Discussion The analysis shows that new literary practices have been created in the Manila Bay Chabacano communities as a response to the language endangerment situation affecting the genres in which texts are produced (see also Mühleisen 2005: 6). In Cavite and Ternate, there has been a growing production of new written materials aimed at the preservation and promotion of the language. In Cavite these have been mostly language learning materials and dictionaries, in addition to other written texts, such as newsletters, blog texts, and newspaper columns. In Ternate, the text genres are more connected to the promotion of local history and culture. Folk stories and accounts of local history have been published in Chabacano. Likewise, rap lyrics by young Chabacano speakers in Ternate make use of Chabacano to express their social realities and local identities. As for other languages, the writing of previously uncodified varieties gains public visibility generally in different stages: first, writing happens in simple humorous texts, letters, riddles, and folksongs, and later it extends to newspapers, official documents, and other types of more formal writing (Mühleisen 2005: 4, Kloss 1952: 26–27). This is partly confirmed by our Chabacano data, especially for Ternate. Key roles in the development of new orthographies are often played by bilingual cultural elites such as language activists or linguistic experts, with or without formal linguistic training (Sebba 2007: 59, 79–80). In our study, this can be seen especially in Cavite City in the production of language materials and written texts in newspapers and other forums. This can be extended to Ternate as well, where printed publications use etymological spellings. These are also adapted by younger generations, especially by groups that are interested in

74 | Eeva Sippola

local language and identity promotion, as can be seen in the case of the rap lyrics. Regarding the spellings used for these varieties, two tendencies have been identified. First, there are some differences between the communities, as Ternate writers highlight the local phonological features more consistently than in Cavite. Writing in creole is strongly connected to the sociolinguistic situation and level of standardization processes and the production of written materials. Second, the age of the writers affects their orthographic choices. Most young speakers write according to the Philippine spelling tradition, while older writers prefer etymological spellings. Interestingly, young speakers involved in language promotion activities, such as the rappers in Ternate, introduce a higher number of Hispanic spellings in their texts, indexing a connection to the colonial past. Another contributing factor is the nature and degree of bilingualism, as noted by Sebba (2000) for Sranan. This can be observed especially in the generational differences. Young speakers’ first languages of literacy are Tagalog and English, and Spanish is clearly absent. Older speakers, on the other hand, have had access to Spanish, including exposure to the written form in school and for administrative purposes. The postcolonial situation of the multilingual Philippine society also shapes how Chabacano is written. Languages do not exist in a vacuum, and these writing practices have a clear connection to the social and historical context of the creole varieties and to the writers’ beliefs and attitudes about the different languages of their community. Lesho & Sippola (2014: 27, 36) have shown that language shift to Philippine languages is perceived as a threat to Chabacano. Tagalog influence on the lexicon and grammar are interpreted as problematic by older speakers and language activists in Cavite City and Ternate, which is also reflected in their literacy practices. For historical reasons, Spanish is perceived as a prestige variety, with several positive linguistic and ethnic connotations, which are also appropriated in writing. Moreover, some speakers still often consider Chabacano to be a dialect of Spanish, which serves as a further motivation for choosing an etymological orthography. The tendency to use the Spanish conventions taps into the prestige that Spanish had and still has as a colonial language. However, some speakers believe there is a ‘pure’ or ‘legitimate’ Chabacano that is distinct from Spanish (Lesho & Sippola 2014: 35), which echoes the desire to express independence and difference from the colonial language. When studying the writing of language activists, it becomes clear that orthography is used to create and express identities that are different from Tagalog, the main language of the communities, which is also the main threat for the endanger-

Postcolonial language ideologies: Writing in Chabacano | 75

ment of Chabacano (see also Lipski 2001: 157). The situation in the Philippine creole communities is thus significantly different from the orthography debate of many other postcolonial communities, where the colonial languages continue to be used actively in situations of diglossia. The prestige of English seems to protect it from negative perceptions. The speakers are less aware of the English influence on the lexicon (Lesho & Sippola 2014: 27–28), and English spellings seem to enter the Chabacano writing especially in cases where Spanish and English lexical items overlap. As noted in the case of rap lyrics, this is not only a question of writing practices; more generally, new indexes to American popular culture are entering the language practices of Chabacano-speaking communities.

6 Concluding remarks In this paper, we have shown that there is an active negotiation and reshaping of language practices in the Chabacano-speaking communities of the Manila Bay region in order to respond to changes in the social and historical context. Language endangerment has led to a situation in which the speakers, especially language activists, have created new, written uses for the creole. The speakers use multilingual and multicultural resources in their text production and orthographic choices, which are further shaped by the language endangerment situation, the nature of their bilingualism, their first language of literacy, and their language ideologies about Chabacano and the former colonial and current national languages Spanish, Filipino, and English. These postcolonial creole varieties show an interesting case of selecting elements from the earlier colonial language, Spanish, that can be used for language preservation purposes. The resources for writing and orthography are reinterpreted to index a historical connection to the former prestige language, while also indexing local authenticity and an identity distinct from Tagalog, the main language of the communities. The prestige of Spanish as the former colonial language is thus reinterpreted in the creation of new language ideologies in these postcolonial creole communities.

76 | Eeva Sippola

Abbreviations DEF

definite article

OBJ

object

DEM

demonstrative

PFV

perfective aspect

ENC

enclitic particle

PL

plural

EX

existential

POSS

possessive

IPFV

imperfective aspect

PP

preposition

IRR

irrealis

RD

reduplicated

LOC

locative

SG

singular

NEG

negation

References Asociacion Chabacano del Ciudad de Cavite. 2008. Diccionario Chabacano del Ciudad de Cavite. Cavite City: Office of the City Mayor. Deuber, Dagmar & Lars Hinrichs 2007. Dynamics of orthographic standardization in Jamaican Creole and Nigerian Pidgin. World Englishes 26(1). 22–47. Errington, Joseph. 2007. Linguistics in a Colonial world: A story of language, meaning and power. New York: Blackwell Publishing. Escalante, Enrique R. 2005. Chabacano... for everyone: A guide to the Chabacano language. Manila: Baby Dragon Printing Press. Escalante, Enrique R. 2010. Learning Chabacano: A handbook. Manila: Baby Dragon Printing Press. Fernández, Mauro. 2012. Leyenda e historia del Chabacano de Ermita (Manila). UniverSOS 9. 9–46. Forman, Michael. 2001. Confidence in Chabacano: Counterbalance to a western ideology of language. In Fernández, Mauro (ed.), Shedding light on the Chabacano language. Special issue of Estudios de Sociolingüística 2(2). 95–117 Fortuno-Genuino, Cecilia. 2011. Is Chabacano dying? In Ambeth R. Ocampo (ed.), Selected papers from Philippine-Spanish Friendship Day conferences, 2009, 1–25. Manila: National Historical Commission of the Philippines. Gonzalez, Andrew. 1998. The language planning situation in the Philippines. Journal of Multilingual and Multicultural Development 19(5). 487–525. Gonzalez, Andrew. 2003. Language planning in multilingual countries: The case of the Philippines. Paper presented at the Conference on Language Development, Language Revitalization, and Multilingual Education in Minority Communities in Asia, 6–8 November, Bangkok. http://www.sil.org/asia/ldc/plenary_papers/andrew_gonzales.pdf (checked 04/12/2014.) Hinrichs, Lars & Joseph T. Farquharson. 2011. Variation in the Caribbean. From creole continua to individual agency. Amsterdam: John Benjamins. Irvine, Judith T. 1989. When talk isn’t cheap. Language and political economy. American Ethnologist 16. 248–267 Kloss, Heinz. 1952. Die Entwicklung neuer Germanischer Kultursprachen von 1800–1950. München: Pohl & Co.

Postcolonial language ideologies: Writing in Chabacano | 77

Kloss, Heinz. 1967. ‘Abstand languages’ and ‘Ausbau languages’. Anthropological linguistics 9(7). 29–41. Kouwenberg, Silvia. 2013. Papiamentu structure dataset. In Susanne Maria Michaelis, Philippe Maurer, Martin Haspelmath & Magnus Huber (eds.), Atlas of Pidgin and Creole language structures online. Leipzig: Max Planck Institute for Evolutionary Anthropology. (Available online at http://apics-online.info/contributions/47, checked 05/31/2014) Lesho, Marivic. 2013. The sociophonetics and phonology of the Cavite Chabacano vowel system. Columbus, OH: The Ohio State University dissertation. Lesho, Marivic & Eeva Sippola. 2013. The sociolinguistic situation of the Manila Bay Chabacano-speaking communities. Language Documentation and Conservation 7. 1–33. Lesho, Marivic & Eeva Sippola. 2014. Folk perceptions of variation among the Chabacano creoles. Revista de Crioulos de Base Lexical Portuguesa e Espanhola 5. 1–46. Lipski, John. 1986. The reduction of /s/ in Philippine creole Spanish and its implications for historical Hispanic dialectology. Diachronica 3. 43–66. Lipski, John. 1987. Descriollización del criollo hispanofilipino: El caso de Zamboanga. Revista española de lingüística 17(1). 37–56. Lipski, John. 2001. The place of Chabacano in the Philippine linguistic profile. In Fernández, Mauro (ed.), Shedding Light on the Chabacano Language. Special issue of Estudios de Sociolingüística 2(2). 119–163. Lipski, John. 2010. Chabacano y español: resolviendo las ambiguedades. Lengua y migración 2(1). 5–41. Lipski, John, Peter Mühlhäusler & F. Duthin. 1996. Spanish in the Pacific. In Stephen Wurm, Peter Mühlhäusler & Darrell Tryon (eds.), Atlas of languages of intercultural communication in the Pacific, Asia, and the Americas, 271–298. Berlin/New York: Mouton de Gruyter. Makihara, Miki & Bambi B. Schieffelin. 2007. Cultural processes and linguistic mediations. Pacific explorations. In Miki Makihara & Bambi B. Schieffelin (eds.), Consequences of contact: Language ideologies and sociocultural transformations in Pacific societies. 3–29. Oxford: Oxford University Press. Miravite, Rommel M., Ulysses Clint N. Sanchez, Day S. Tardo, Shirven John B. Viloria & Fr. David John M. Delos Reyes, S. J. 2009. Chavacano Reader. Hyatsville: Dunwoody Press. Mühleisen, Susanne. 2005. Introduction: Creole languages in Creole literatures. Status and standardization. Creole languages in Creole literatures. Journal of Pidgin and Creole Languages (Special Issue) 20(1). 1–14. Nigoza, Evangelino. 2007. Bahra. The history, legends, customs and traditions of Ternate, Cavite = Bahra. Manga historia, alamat, custumbre y tradiciong di Bahra. Emmanuel Franco Calairo & M. Neriza Villanueva (eds.), Cavite: Cavite Historical Society. Nigoza, Evangelino. 2013. Platicada’y Bahra (For elementary grades). Ternate: Cavite West Point College. Ocampo, Esteban A. de. 2007. The Ternateños. Their history, languages, customs and traditions. Manila: National Historical Institute. Penny, Ralph. 1991. Gramática histórica del español. Barcelona: Ariel. Quilis, Antonio & Celia Casado-Fresnillo. 2008. La lengua española en Filipinas: Historia, situación actual, el chabacano, antología de textos. Madrid: Consejo Superior de Investigaciones Científicas. Ramos, Felicidad G. 1963. A contrastive analysis of the sound system of the Cavite Chavacano and the English languages. Manila: Abad Santos Educational Institution MA thesis. Romaine, Suzanne. 2005. Orthographic practices in the standardization of pidgins and creoles: Pidgin in Hawai’i. Journal of Pidgin and Creole Languages 20(1). 101–140.

78 | Eeva Sippola

Schachter, Paul & Fe T. Otanes. 1972. Tagalog reference grammar. Berkeley: University of California Press. Schieffelin, Bambi B. & Rachelle Charlier Doucet. 1998. The “real” Haitian Creole: Ideology, metalinguistics, and orthographic choice. In Bambi Schieffelin, Kathryn Woolard & Paul Kroskrity (eds.), Language ideologies: Practice and theory, 285–316. New York: Oxford University Press. Sebba, Mark. 1998. Phonology meets ideology: The meaning of orthographic practices in British Creole. Language Problems and Language Planning 22(1). 19–47. Sebba, Mark. 2000. Orthography and ideology: Issues in Sranan spelling. Linguistics 38(5). 925–948. Sebba, Mark. 2007. Spelling and society: The culture and politics of orthography around the world. Cambridge: Cambridge University Press. Silverstein, Michael. 1979. Language structure and linguistic ideology. In Paul Cline, William F. Hanks & Carol L. Hofbauer (eds.), The elements. A parasession on linguistic units and levels, 193–247. Chicago: Chicago Linguistic Society. Sippola, Eeva. 2009. La variación estilística en el chabacano: el caso de los relativizadores. In Lars Fant, Johan Falk, María Bernal & Fernando Ferrán Meliá (eds.), Actas del II Congreso de Hispanistas y Lusitanistas Nórdicos. 547–559. (Romanica Stockholmiensia 26) Sippola, Eeva. 2010. Chabacano for everyone? Chabacano language projects in Cavite City in comparison with other Chabacano communities. In Bettina Migge, Isabelle Léglise, & Angela Bartens (eds.), Creoles in education, 55–80. Amsterdam: John Benjamins. Sippola, Eeva. 2011. Una gramática descriptiva del chabacano de Ternate. Helsinki: University of Helsinki Ph.D. thesis. Steinhauer, Hein. 2005. Colonial history and language policy in Insular Southeast Asia and Madgascar. In Alexander Adelaar & Nikolaus P. Himmelmann (eds.), The Austronesian languages of Asia and Madagascar, 65–86. London/New York: Routledge. Tardo, Day S. 2006. Developing the Chavacano Reader Project from the Chavacano Corpus. Paper presented on the 10th International Conference on Austronesian Linguistics (10ICAL). 17–20 January, Puerto Princesa City, Palawan, Philippines. http://www.sil.org/asia/philippines/ical/papers.html (checked 12/05/2006) Thompson, Roger M. 2003. Filipino English and Taglish: Language switching from multiple perspectives. Amsterdam: John Benjamins. Winer, Lise. 1990. Orthographic standardization for Trinidad and Tobago: Linguistic and sociopolitical considerations in an English Creole community. Language Problems and Language Planning 14(3). 237–268. Wolff, John. 2001. The influence of Spanish on Tagalog. In Klaus Zimmerman & Thomas Stolz (eds.), Lo propio y lo ajeno en las lenguas austronésicas y amerindias. Procesos interculturales en el contacto de lenguas indégenas con el español en el Pacífico e Hispanoamérica, 233–252. Frankfurt/Madrid: Vervuert/Iberoamericana. Woolard, Kathryn. 1998. Introduction. Language ideology as a field of inquiry. In Bambi Schieffelin, Kathryn Woolard & Paul Kroskrity (eds.), Language ideologies: Practice and theory, 3–47. New York: Oxford University Press. Woolard, Kathryn A. & Bambi B. Schieffelin. 1994. Language ideology. Annual Review of Anthropology 23(1). 55–82.

Martina Drescher

Hybridized discourse markers in Cameroonian French? The example of déjà Abstract: This study, located at the junction of pragmatics, variational linguistics and contact linguistics, aims to contribute to a better knowledge of the pragmatics of Cameroonian French (CF). Based on the assumption that the multiple processes of appropriation and indigenization observable in postcolonial varieties of sub-Saharan French concern also its pragmatic and discursive level, I concentrate on the temporal and aspectual adverb déjà (‘already’, ‘even’, ‘yet’), which can also be used with primarily discourse structuring functions. The adverb déjà drew my interest because some of its uses in CF diverge from those known from so-called ‘hexagonal’ French (HF) spoken in France and thus support the hypothesis of hybridized discourse markers. Based on data coming from both media and face-to-face interactions different meanings of déjà in CF will be examined. The analysis shows that it is difficult to tell typically oral uses from those specific to a regional variety since the modifications may either be the result of production constraints that is the oral nature of the data, or represent hybridizations due to the specific ecology of French in Cameroon. Keywords: discourse marker, Cameroonian French, hybridization, language contact, language variation

1 Introduction This study, located at the junction of pragmatics, variational linguistics and contact linguistics, focuses on discourse markers in Cameroonian French (CF). Based on the assumption that the multiple processes of appropriation and indigenization one can observe in postcolonial varieties of sub-Saharan French concern also its pragmatic level, I concentrate on hybridized discourse markers. As one of the two official languages in multilingual Cameroon (the second, yet politically and demographically less important being English), French has experienced a great dynamic that resulted in the emergence of an endogenous – and

|| Martina Drescher: University of Bayreuth, Language and Literature Studies, Romance and General Linguistics, 95440 Bayreuth, Germany. E-Mail: [email protected]

80 | Martina Drescher

up to now predominantly spoken – variety of the language. This regional French differs in various aspects from the so-called ‘hexagonal’ French (HF) that is spoken in France. It is characterized both by influences of the contact languages and by more general features of reduction but also creativity that might have their origin in its acquisition as a chiefly second language L2. Until recently, the pragmatic and discursive level has been paid little attention in the analysis of such appropriation processes (Drescher 2012a, 2014). The aim of my study is thus to contribute to a better knowledge of CF’s pragmatics by concentrating on the temporal and aspectual adverb déjà (‘already’, ‘even’, ‘yet’), that can also be used with primarily discourse structuring functions. The adverb déjà drew my interest because some of its uses in CF seem to diverge from those known in HF. This backs the hypothesis of possible hybridizations at the semantic-pragmatic level, due to the specific ecology of CF. Diverging semantic, pragmatic and rhetoric conventions of the local contact languages – a field Manessy (1994) calls sémantaxe – might result in transfers and also trigger certain changes in the use of discourse markers which I refer to as ‘hybridized discourse markers’. Based on data coming from both media and face-to-face interactions, various uses of déjà in CF will be examined. As already mentioned, alongside its grammatical meanings, this item also carries pragmatic and discursive meanings since it functions as a discourse marker with modal, argumentative and interactional values. A closer look on the various occurrences of the adverb déjà in CF will not only contrast them with its uses in hexagonal French, but will also shed light on some of the pragmatic and discursive practices in postcolonial Cameroon. I first discuss briefly the theoretical background of my investigation by placing it in the context of research on discourse markers and language contact (section 2). Then I follow with some information on Cameroon’s linguistic landscape and the ecology in which French evolves (section 3). The next section is dedicated to a succinct overview of research on déjà in HF (section 4). Thereafter, I move on to the analysis preceded by a presentation of the data (section 5). A short conclusion will summarize the main findings and indicate some perspectives for further research (section 6).

Hybridized discourse markers in Cameroonian French? | 81

2 Theoretical background: discourse markers and language contact Since the 1970s there is a still growing interest in discourse markers which represent a primarily functional category composed of words and expressions belonging to different grammatical classes like adjectives, adverbs, conjunctions, interjections but integrating also some frozen verbal phrases. From a functional perspective, markers index the location of an utterance within its emerging local contexts. [...] markers propose the contextual coordinates within which an utterance is produced and designed to be interpreted. And, finally, [...] they contribute to the integration of discourse – to discourse coherence (Schiffrin 1996: 315).

In addition, discourse markers also hint to epistemic stances and have to do with the coordination between speaker and hearer, e.g. with regard to turntaking. According to Auer & Günthner (2003: 5), they are metapragmatic signs that allow for an interpretation of the verbal actions, by commenting on or steering them, by ensuring their formal requirements and by anchoring them in the context. Nevertheless, the different sub-classifications of discourse markers, based either on a narrow or a wide understanding of the concept, show that the outline of the category remains – at least with regard to French – still vague and controversial (Drescher & Frank-Job 2006). Discourse markers are subject to variation, especially to diatopic variation. Thus, some of the French discourse markers are specific to regional varieties.1 Whereas the French-Canadian forms and expressions carrying discursive functions have been studied extensively since the 1980s (Vincent 1993; Dostie 2004), the discourse markers in African varieties of French have received very little attention.2 Yet it is well known that discourse markers are quite vulnerable to language contact. Especially in those parts of Canada where French is a minori|| 1 One could mention here the discourse marker sti, a phonologically reduced form of the derived interjection hostie (‘host’), frequent in Canadian French and going back to a swear word drawn initially from the sacred and thus tabooed religious domain (Drescher 2000, 2009). 2 Only recently some pilot studies on discourse markers in African varieties of French were conducted. Yet most of them have been carried out on a rather small empirical basis or by using constructed examples. Abolou (2010) describes the items ke, borrowed from local languages (baoulé, dyula, bété), and non (‘no’, ‘not’) in popular Ivorian French. Klaeger (2010) examines the uses of même (‘even’) in Burkina Faso’s French and Skattum (2012) presents an analysis of bon (‘well’) based on a French corpus from Mali.

82 | Martina Drescher

ty language und thus in close and constant contact with English, it appears that with regard to discourse markers not only semantic and pragmatic interferences but also borrowing is common. As in the historical Canadian region of Acadia or in today’s mostly English-speaking Louisiana, this may lead to doubled discourse markers like tu sais/you know or ben/well stemming from the two languages in contact which Neumann-Holzschuh (2009: 137) considers a case of “contact induced grammatical replication”.3 From a more general, typological point of view, Matras (2000, 2007, 2010) has convincingly shown that discourse markers belong to the first elements borrowed in a contact situation where the L2 dominates the first language L1. Emphasizing the pragmatic dimension as well as the communicative and cognitive factors of language contact, he notes that discourse markers rank nearly on top of a scale of borrowability and thus seem to open the door for other, rather morpho-syntactic transfers.4 This may be due to their formal as well as their semantic and pragmatic characteristics: With regard to their structural properties relatively free and quite universal concerning their discursive functions, these items can be easily transferred from one language to another. Moreover, they are often emblematical and may thus contextualize – especially in multilingual societies – different linguistic identities. Thus, in Maschler’s (2000: 438) words, “studies of discourse markers in bilingual conversation […] are of prime importance to the study of language contact and change”. However, if discourse markers are easily borrowed from a foreign language, they seem rather difficult to master in a foreign language. Thus, their borrowability contrasts sharply with their late acquisition in a L2. As yet the use of discourse markers in situations where the L1 influences the L2, thus in contact situations characterized by imperfect learning in the sense of Thomason (2001) has hardly been analyzed. But there are a few studies in the related field of second language acquisition. Sankoff et al. (1997) for instance, in an investigation on Anglophone Montrealers’ fluency in French, note that discourse markers are reliable indicators of the general competency in a given L2. Whereas “the least fluent, least competent L2 speakers used almost no discourse markers, […] the more successful L2 speakers were those who could control native discourse markers in a nativelike fashion” (Sankoff et al. 1997: 213). Because of their various and sometimes very complex pragmatic and discursive functions, discourse markers are easily subject to interferences. In a

|| 3 See also Chevalier (2000, 2007), Neumann-Holzschuh (2008). 4 See Maschler (2000: 440) who raises the question whether “the borrowing of discourse markers is […] the result of codeswitching”.

Hybridized discourse markers in Cameroonian French? | 83

long lasting contact situation with a majority of L2 speakers, as it is typical for postcolonial French speaking countries in sub-Saharan Africa that are part of the so called francophonie seconde, massive transfers from the L1 might occur and trigger shifts in the meaning of certain discourse markers. These shifts will eventually become fossilized and enter into the endogenous variety of French.5 To sum up, discourse markers represent a particularly interesting field of research for studies on sub-Saharan varieties of French, which are in permanent and intense contact with indigenous African languages and therefore testify to a great dynamics and a complex ecology.

3 The ecology of French in Cameroon As already mentioned Cameroon’s linguistic landscape is very complex since there are more than 250 African languages covered by the two official languages French and English. The acquisition of French generally starts at elementary school, but as there are still high rates of illiteracy; the majority of the population has no knowledge of the official language. However alongside this guided acquisition, people, especially in urban areas where French functions as a language for interethnic communication, also learn the language outside the institution through ‘natural’ interactions in their respective environments. This mode of acquisition is less subject to normative rules. Thus it fosters appropriation and finally the formation of an endogenous variety of French. Compared to HF, this regional French has undergone major transformations on all levels of language. Whereas its specificities on the phonetic, morpho-syntactic and lexical level are well known (Mendo Zé 1992; Zang Zang 1998; Biloa 2003), this does not hold true for its pragmatic and discursive level. By concentrating on the use of discourse markers, the purpose of my contribution is precisely to investigate variation at this level. The adverb déjà that takes the center stage here has already been examined with regard to HF. Hence the following section is dedicated to a brief synthesis of previous research.

|| 5 As attested by their entries in the different lexicographic inventories of African varieties of French, some shifts in the meaning of discourse markers have already been noticed and roughly described. Going back to the 1960s, these inventories focus on lexical items with denotative meanings. Thus for the most part they consider nouns, verbs and adjectives and leave aside other grammatical classes inventoried less systematically. This also applies to discourse markers.

84 | Martina Drescher

4 A synthesis of research on déjà in hexagonal French There is an extensive body of research on déjà in HF that I can hint to only very briefly. Whereas older studies concentrate on the grammatical functions of déjà as a temporal and aspectual adverb6, the more recent research also considers its pragmatic functions by stressing its modal, argumentative and emphatic uses as a discourse marker. An important part of these early investigations is based on construed examples and reflects thus the linguist’s intuition. Other, more empirical studies draw on written sources, mainly fictional or media texts and examples from the Internet. Still, analyses based on oral data that illustrate the uses of déjà in spontaneous, spoken language are rare. And to my knowledge, no investigation has been dedicated to the uses of the adverb in a regional, or even African variety of French, with the exception of a few studies dealing with a possible influence of the German adverb schon – in certain contexts an equivalent of déjà – on some uses of déjà in Alsace or in French speaking parts of Switzerland.7 Most of the authors agree to distinguish grammatical and pragmatic meanings of déjà, but there is great variation in terminology and especially in the subclassification of the pragmatic uses. Buchi (2007), who contrasts grammème and pragmatème, proposes a diachronic analysis of déjà.8 Morency (2011), who

|| 6 See Martin (1980); Hoepelman & Rohrer (1980); Franckel (1989). 7 In this use, called “interactional”, déjà mitigates a request to recall information that the speaker should normally have available as in c’est quoi déjà son nom? (‘what’s his name again?’). Valikangas (1985) explains this interactional value of déjà by a pragmatic transfer from German to French through varieties of French which are in direct contact with German (e.g. Alsace or French speaking Switzerland). See also Valikangas (2004). Mosegaard Hansen (2003: 46) by contrast interprets it as an “extension of phasal déjà to the speech act level”. Franckel (1989: 282) relates it to “degenerated knowledge” and Morency (2011) calls it “reminder déjà“. Buchi (2007), on the other hand, draws the attention to a different, but similar use, also observable in Alsace and probably copied from the German modal particle schon, where déjà mitigates an assertive utterance connoting that the speaker is enervated as in Je le ferai déjà (‘Ok, I will do it‘). 8 Buchi (2007: 1) concentrates on pragmatic meanings (enunciative, discursive) of déjà and discusses issues related to a possible pragmaticalization of the adverb from a historical perspective. She distinguishes first a “scalar” use where déjà is a synonym of d’abord (‘first’), second an “ordinal” use where déjà, detached at the beginning of a proposition, marks order or succession and finally an “adversative” use, typical of spontaneous, oral speech. This use of

Hybridized discourse markers in Cameroonian French? | 85

argues along the lines of the relevance theory, focuses on the argumentative and discursive uses of déjà. According to this author, déjà is a procedural expression with a minimal conceptual kernel, filled up by elements from the context. Its minimal semantics triggers an instructional cognitive treatment that guides the recipient’s interpretation of the utterance via interferences. For Morency (2011: 20), it is not temporality, but a subjective or even modal effect that is primordial in the basic meaning of déjà. Mosegaard Hansen (2003, 2008) and Mosegaard Hansen & Strudsholm (2008) present a series of publications on déjà which represent the most extensive and thorough study of the adverb’s pragmatic and discursive functions. Mosegaard Hansen (2008) opposes content-level uses to context-level uses, a distinction that comes close to Buchi’s grammème and pragmatème. The content-level uses embrace phasal and iterative uses as well as the use as a temporal focus particle (Mosegaard Hansen 2008: 169). Phasal déjà expresses precocity in the occurrence of a process (Les artichauts fleurissent déjà ‘The artichokes flower already’) and it is negated by ne … pas encore (‘not yet’) whereas the iterative déjà indicates a past experience (Tu as déjà mangé des artichauts ‘Have you eaten artichokes before?’) and is negated by ne … jamais (‘never’). As a temporal focus particle, déjà does not belong to the verbal constituent, but to the adverbial phrase of the sentence (Je l’attends déjà depuis 2 heures ‘I have been waiting for him for two hours already’). In this case, the negation is realized with ne … que/seulement (‘I have been waiting for him for two hours only’). These grammatical uses contrast with the context-level uses where déjà functions as a pragmatème.9 In modern French, these pragmatic and discursive uses seem particularly abundant. Mosegaard Hansen (2008: 218) distinguishes here between a scalar use, a categorizing use and déjà as scalar focus particle.10 With the scalar use the idea of an objective movement on the time line gives place to the movement of a conceptualizing subject who goes through a scale of values. Categorizing déjà (a new term for a meaning that Mosegaard Hansen 2003 referred to as “comparative déjà” and that was called “déjà de marginalité” in Mosegaard Hansen & Strudsholm 2008) no longer points to a scale of values, but to a conceptual category conceived by the conceptualizing subject

|| déjà seems frequent in direct speech and it has as an equivalent après tout (‘after all’). Here the argumentative meaning is prevailing. 9 Mosegaard Hansen (2003) notes that, from a typological point of view, phasal adverbs seem to easily develop supplementary modal and discursive meanings. 10 See Mosegaard Hansen (2003) where these uses are covered by the more general category of “comparative” meanings.

86 | Martina Drescher

as an entity with a center and a periphery. Here déjà indicates that this entity constitutes an instance situated at the margins of the targeted conceptual category. The scalar déjà which was called “non temporal focus particle” in earlier studies operates on a pragmatic scale of arguments applicable and non applicable to the predicate: déjà emphasizes an argument that is more informative compared to others. Mosegaard Hansen (2008) also enumerates discourse-marking11 and interactional functions with their respective sub-categories: thematic/listing and conjunction with QUE and interrogative and imperative functions that specify the interactional uses of the adverb. The thematic déjà appears detached at the beginning of a sentence and it clearly has an argumentative value. However from a rhetorical point of view it is neutral since it does not confer a positive or negative orientation to the arguments it introduces.12 From a functional point of view the conjunctional use (déjà que) seems quite similar to the thematic one. The interactional meanings are both marginal as they refer to very specific and restricted pragmatic conditions of use (see fn. 7). Since both the terminology and the subclassification have been modified several times, this systematization became quite confusing. The synopsis below, taken from Mosegaard Hansen & Strudsholm (2008: 472) who compare French déjà to its Italian equivalent gia, summarizes the different grammatical and pragmatic-discursive meanings of déjà in HF. However, the terminology has been slightly modified by taking into account the new categories introduced in Mosegaard Hansen (2008) whereas the older designations are put in parentheses.13 Tab. 1: Uses of déjà (adapted from Mosegaard Hansen & Strudsholm 2008: 472) Uses of déjà 1. 1.1

content-level uses phasal

1.2

iterative

Les artichauts fleurissent déjà ‘The artichokes flower already’ Tu as déjà mangé des artichauts? ‘Have you eaten artichokes before?’

|| 11 This category was called “connective” in Mosegaard Hansen (2003). Comparative and connective uses both formerly figured under the heading of the more global category of modal uses. 12 This argumentative use has already been emphasized by Cadiot et al. (1985) who compare déjà to toujours (‘always’) which in some contexts fulfils similar functions. 13 The English translations of the examples are mine.

Hybridized discourse markers in Cameroonian French? | 87

Uses of déjà 1.3

temporal focus particle

2. context-level uses 2.1 modal 2.1.1 scalar 2.1.2 categorizing (of marginality)

Je l’attends déjà depuis deux heures ‘I have been waiting for him for two hours already’

10.000 euros, c’est déjà une somme! ‘10.000 euros, that is already a sum!’ Menton, c’est déjà la France ‘Menton is already in France’ Déjà son visage me déplaît ‘First his face displeases me’

2.1.3 scalar (non temporal) focus particle 2.2 discourse-marking (connective) 2.2.1 thematic/listing J’ai bien aimé ce film: déjà, c’est original, et puis il y a de très belles photos ‘I liked this film: first it is original and then there are very nice photos’ 2.2.2 conjunctional (with que) Je n’aime pas Max: déjà qu’il fume comme un pompier, mais en plus il est agressif ‘I don’t like Max: first, he smokes like a fireman, and in addition he is aggressive’ 2.3 interactional 2.3.1 interrogative Quel est ton nom, déjà? ‘What’s your name again?’ 2.3.2 imperative Montre moi [sic] déjà un faisceau de sources qui disent que pour qu’un état soit un vrai état, il faut la phrase « peuple du machinchose » sur internet ‘Show me first a bundle of sources which say that for a state in order to be a real state there has to be the sentence ‘people of the thingsmabob’ on the internet’

As the fluctuating categorizations of the different pragmatic and discursive meanings of the adverb déjà show, the distinctions seem difficult to make and they are probably less clear-cut than the above table suggests. In spite of these problems they provide a first classificatory grid for an empirical analysis of the data, even if, as we will see in a moment, some of them have to be modified in order to catch the specificities of spoken Cameroonian French.

5 Data and analysis: déjà in Cameroonian French Even though in the introductory chapter I evoked the hypothesis of transfers at the semantic-pragmatic level that may result in the formation of hybridized discourse markers, my purpose here is not to carry out a comparative analysis

88 | Martina Drescher

contrasting déjà with its equivalents in at least one of its contact languages.14 Rather I focus on its uses in spoken CF and analyze them empirically. This means that I am leaving the important, but from a methodological point of view also very challenging comparative issue and thus the verification of the underlying assumption for subsequent studies. Yet a synchronic description of déjà in CF is an indispensable condition for future comparative research. As stated before the empirical observation that some uses of the adverb in CF diverge from those in HF serves as a starting-point. Some evidence for this assumption can also be found in the inventories of African varieties of French. Indeed these lexicographic works embrace a differential perspective, but focus mainly on lexical items, so that there is only scarce information on divergent uses of discursive or pragmatic devices.15 The data consists of 150 occurrences of déjà collected in different CF corpora. First, I draw on a media corpus, composed of various shows from the phonein Un autre monde (‘Another world’), featuring a traditional healer who offers consultations on health issues (Drescher 2012a). In addition, I resort to some other phone-ins where the auditors intervene on current political and/or social topics (Drescher & Calamaro in press). Second, I draw on audio taped interactions between traditional healers, so called docteurs ambulants (‘travelling doctors’), who promote and sell medical products to the passengers on board country buses.16 100 occurrences of déjà come from the media corpus and 50 examples are taken from the face-to-face interactions in the buses. In some cases it has been very difficult or even impossible to assign one of the functions established on the basis of HF data to the examples found in the Cameroonian corpus. This is not always due to the regional origin of the data, but rather to their spontaneous and oral character. It also became clear that the context plays a crucial role in disambiguating the different uses of déjà.

|| 14 Whereas descriptions of discourse markers for relevant Cameroonian contact languages are still lacking, Dimmendaal (2014) presents a cross-linguistic analysis of so-called “attitude markers” in Nilotic by stressing their various interpersonal and communicative functions. Some of these uses – e. g. information structuring or modality marking – are quite similar to those discussed with regard to déjà in CF. 15 The IPLFAN (2004: 11) signals for instance that the adverb déjà is used in the French variety spoken in the Central African Republic with a meaning close to enfin (‘finally’, ‘at last’). 16 These data have been collected by Liliane Ngawa (Bayreuth) to whom I am deeply indebted for giving me access to the corpus of her doctoral thesis.

Hybridized discourse markers in Cameroonian French? | 89

5.1 Content-level uses of déjà In the docteurs ambulants face-to-face corpus the content-level or grammatical uses of déjà are largely prevailing. 48 of a total of 50 occurrences belong to this category whereas in the media corpus only 45 of the 100 occurrences have a grammatical meaning. All in all 93 of the 150 occurrences, that is more than half of my examples, are content-level uses.

5.1.1 Phasal DÉJÀ All in all, there are 56 occurrences where déjà signals the precocity of an event, 25 from the face-to-face corpus (docteurs ambulants) and 31 from the media corpus.17 (1)

tu go au guichet demander si le car part déjà ‘you go to the counter ask if the bus leaves already’

(2)

en tout cas il fait déjà chaud dans la voiture ‘anyway it is already hot in the car’

(3)

elle a déjà pris 10 paquets de bobolos ‘she has already taken 10 packs of bobolos (cassava sticks)’

(4)

elle espère que j’ai déjà acheté un terrain (.) je lui ai dit non pas encore ‘she hopes that I have already bought a lot (.) I told her no not yet’

5.1.2 Iterative DÉJÀ 26 occurrences belong to the category of iterative déjà where the adverb refers to a past experience. 21 instances stem from the docteurs ambulants-corpus and only five examples from the media corpus. (5)

vous avez déjà entendu parler de la filariose? ‘did you hear talk of filariasis before?’

|| 17 The transcription of the oral data follows orthographic conventions with the exception of the signs “(.)” and “=”. The first indicates short pauses, the second a quick latching. The English translations are only rough glosses of the CF utterances, especially regarding the very complex meanings of the adverb déjà that takes the centre stage here. In addition the translation does not preserve the typical features of CF.

90 | Martina Drescher

(6)

qui a déjà goûté ça auparavant? ‘who has ever tasted this before?’

(7)

vous avez déjà travaillé euh avec si rémy semengue? ‘did you work uhm with si rémy semengue before?’

5.1.3 Temporal focus particle DÉJÀ In this case the scope of déjà is a temporal adverb or an adverbial phrase. In face-to-face interactions, there are only two occurrences belonging to this category and eight more examples come from the media corpus. In total there are ten occurrences in the whole corpus. (8)

là on est déjà vers 12h30 ‘now we are already about 12h30’

(9)

en fait cela dure déjà depuis disons deux ans ‘in fact this lasts already since let’s say two years’

(10)

il souffre de ce mal depuis un an déjà ‘he suffers from this illness since one year already’

I will turn now to the context-level uses where déjà functions as a discourse marker.

5.2 Context-level uses of déjà According to Mosegaard Hansen & Strudsholm (2008) this global category includes the modal, connective and interactional uses of déjà. No examples of interactional and conjunctional (déjà que see 2.2.2 of the table above) déjà could be detected in the data. Besides, a clear distinction between modal and discourse-marking uses has not always been possible. This is probably due to the oral nature of the data and their rather aggregative structure. They often lack explicit coherence so that it is quite difficult to identify an argumentative structure articulated by the discourse marker déjà. Moreover, prosody – for instance the short pause following the listing déjà – seems to play a vital part in the distinction of the adverb’s different meanings in oral language. If déjà contributes to a large extent to the articulation of discourse by structuring the speech, it is not necessarily an argumentative device. In spoken French it rather is a structural marker.

Hybridized discourse markers in Cameroonian French? | 91

First I discuss some examples of modal uses which on the whole are not very frequent in the Cameroonian data.

5.2.1 Modal uses There are no examples for modal uses in the docteurs ambulants-corpus and in total only eight occurrences in the media corpus, which contains five scalar, one categorizing use of déjà, and one case of a scalar focus particle.

5.2.1.1 Scalar DÉJÀ Remember that the scalar déjà indicates the cognitive movement of a conceptualizing subject on a scale of values. (11)

pour arriver à la guérite et déposer son courrier (.) c’est déjà assez pénible ‘to arrive at the booth and deposit one’s mail (.) it is already very hard’

(12)

avec l’huile de bois seulement c’est déjà très efficace ‘with wood oil only it is already very effective’

(13)

la spatule ne coûte pas cher il y a les petites spatules qu’on trouve déjà à cent cinquante francs ‘the spatula isn’t expensive there are the small spatula one can already find for one hundred and fifty francs’

5.2.1.2 Categorizing DÉJÀ In the case of a categorizing use the speaker construes a cognitive category made up of a center and a periphery. There is only one example that seems to match this category. Here déjà figures in a rhetorical question. (14)

mais est-ce que lorsqu’on veut à tout prix l’amour de quelqu’un sans que la personne ne consente c’est pas déjà faire du mal à cette personne? ‘but if one wants at any price the love of somebody without the person’s consent doesn’t it already hurt this person?’

92 | Martina Drescher

5.2.1.3 DÉJÀ as a scalar focus particle In this pragmatic meaning, déjà highlights an argument by presenting it as more informative than others which were also possible in the given context. Here too, the data contain only one example. (15)

lorsqu’on connaît la complexité de (.) la complexité de la maladie euh (.) parce qu’il faut également dire que l’être hum=l’organisme humain déjà a ce phénomène de complexité ‘when one knows the complexity of (.) the complexity of the illness uhm (.) because one should also to say that the human being=the human organism already has this phenomenon of complexity’

5.2.2 Discourse-marking uses Regarding genuinely discourse-marking uses Mosegaard Hansen (2008) and Mosegaard Hansen & Strudsholm (2008) propose two sub-categories of which one – a combination of déjà with the conjunction que – is not represented in the data. The remaining thematic (or listing) use where déjà appears detached at the beginning of a sentence seems to have argumentative values as well. This quite narrow understanding of the discourse-marking uses of déjà might have its origin in a concentration on primarily written data where coherence is marked in a more explicit way and with – at least partly – different means. However, such a restricted comprehension of the adverb’s discourse-marking functions excludes most of the uses one can observe in primarily oral data. In particular, it does not allow to account for a large amount of occurrences in our CF corpus. Here the discourse-marking, more precisely the connective and structuring meanings of déjà are abundant. But one can barely trace them back to what Mosegaard Hansen & Strudsholm call thematic or listing use. I therefore adopt here a broader understanding of ‘discourse-marking’ which considers the various structuring uses of déjà where the adverb contributes primarily to process the emerging discourse. As a structural marker, déjà functions as an ‘anchor-word’ in the process of verbal production. It hints to cognitive processes during the verbalization of an utterance ‘under construction’ by linking and articulating its different parts. This aggregative technique of tying up ‘verbal fragments’ reflects the linearity of discourse, its emergence in time, and it is a typical feature of spontaneous spoken language where the speaker needs to lean on stable elements in order to build up its speech (Ong 1982). I therefore agree with Paillard (2004) who notes that déjà allows the speaker to stabilize his speech by indicating that the elements in its scope are fixed compared to previ-

Hybridized discourse markers in Cameroonian French? | 93

ous, non-stabilized expressions. In oral language one of the main functions of the adverb consists thus to signal a dire stabilisé. This understanding of discourse-marking that focuses on the structural meanings of déjà seems more appropriate to grasp the adverb’s uses in spontaneous spoken speech and it allows also to account for many of the occurrences in the Cameroonian data, even if, due to the sometimes rather incoherent character of the sequences containing déjà, it might be difficult to reconstruct exactly the speaker’s stabilized speech. In the media corpus, the discourse-marking uses of déjà in the just discussed, structural sense are very frequent: 47 of 100 occurrences belong to this category that might need a further differentiation on a broader empirical basis. By contrast, there are only two examples in the docteurs ambulants-corpus. (16)

nous allons nous intéresser déjà à ces personnes qui vont pouvoir nous appeler ‘and we will [first/already] care about these persons who will be able to call us’

(17)

oui déjà euh je voulais simplement dire que le numéro de téléphone c’est le ‘yes [first/already] uhm I just wanted to say that the phone number is the’

(18)

ce qui est sûr (.) la personne à qui on a fait l’envoûtement euh elle le sait pas déjà (.) et puis maintenant l’envoûtement continue à faire son effet ‘what is certain (.) the person to whom one has made the witchery uhm s/he doesn’t know it at first (.) and then now the witchery continues to be effective’

Whereas déjà in (16) confirms the thematic or listing meaning, the two following excerpts illustrate a more structural use where the adverb appears to fix the emergent speech. The hesitation marker and the short pause in the immediate context signal also that the speaker is probably busy with planning activities. In the following section I discuss some uses of déjà that seem specific to CF or, at least, salient since they are difficult to accept for native speakers of HF.

5.3 Some specific uses of déjà in Cameroonian French To start with, there is the co-occurrence of the adverb with même that is quite frequent in CF and – as an acceptability test in a pilot study has shown – unanimously rejected by native speakers of HF (Calamaro 2012). Obviously même functions as an intensifier that reinforces the adverb déjà.

94 | Martina Drescher

(19)

le car est même déjà plein hein, il reste 3 places ‘the bus is even already full isn’t it? 3 seats are remaining’

(20)

j’ai même déjà mal aux fesses mola ‘I have even already pain in my buttocks mola (friend)’

(21)

il faut même déjà savoir ce que c’est un repas complet ‘one has even already to know what it is a complete meal’

Although the most frequent combination is même déjà, there are also some occurrences where the order is inversed: déjà même produces a different effect, as in the following example where it introduces an additional, but nevertheless important argument in a discussion on traditional beliefs about marriage and conception. (22)

oui une jeune femme qui n’est pas mariée qui s’assied sur la porte si l’homme entre même il va ressortir (.) parce qu’elle ne peut pas s’asseoir sur la porte en regardant l’intérieur c’est toujours en regardant l’extérieur (.) et déjà même c’est dur parce qu’elle ferme la porte ‘yes a young woman that isn’t married and that is sitting on the door step if the man enters he will leave again (.) because she may not sit on the door step looking inside it must always be looking outside (.) and even so it is hard because she closes the door’

These quite frozen co-occurrences with même belong not only to CF, but seem today specific to the African varieties of French.18

|| 18 However from a diachronic perspective, one can also find occurrences of déjà même and même déjà in hexagonal French as research within the Frantext corpus shows. The first example goes back to the 17th century: “Vous avez vu ci-devant tout l’extérieur des quatre premières années de la régence, et je vous ai déjà même expliqué l’effet que la prison de M. de Beaufort fit d'abord dans les esprits” (de Retz, Jean-François (1613–1648/1679): Mémoires, tome 1, p. 286– 288) (‘You have seen ahead the whole exterior of the first four years of the regency, and I have explained to you already even the effect that the prison of M. de Beaufort had first on the spirits’). The second excerpt that exemplifies the combination même déjà, dates also back in the 17th century since it is taken from the well-known French drama writer Corneille: “Léontine, ne croyant pas pouvoir cacher longtemps cet enfant que Maurice avoit commis à sa fidélité, vu la recherche exacte que Phocas en faisoit faire, et se voyant même déjà soupçonnée et prête à être découverte voulut mettre dans les bonnes grâces de ce tyran ...“ (Corneille, Pierre (1647): Héraclius, empereur d’Orient: Au lecteur, p. 144–145). (‘Léontine [...] seeing herself even already suspected and ready to be discovered, aimed to put this despot in a merciful mood ...’). The English translations are mine.

Hybridized discourse markers in Cameroonian French? | 95

The negation of iterative déjà is striking too, as some occurrences diverge clearly from the norm of standard French that requires jamais (‘never’) in the following two excerpts. (23)

qui a déjà utilisé? personne n’a encore utilisé? ‘who has used before? Nobody has never used?’

(24)

qui a déjà utilisé? personne n’a déjà utilisé? ‘who has used before? Nobody has used before?’

Among the recurrent constructions, there is also the combination of déjà with the deontic expression il faut. (‘one has to’, ‘one must’) which occurs 12 times in my data. Because of the deontic value of the impersonal verb falloir, the function of déjà might come close to its imperative uses subsumed by Mosegaard Hansen (2008) to the category of interactional meanings. (25)

mais pour construire la région il faut déjà construire le pays ‘but to construe the region one has first to construe the country’

(26)

euh pour le cas du mal d’estomac il faut déterminer déjà le mal d’estomac hein ‘uhm regarding the stomach ache one has first to determine the stomach ache hasn’t he’

(27)

il faudrait déjà soigner cette hémorroïde ‘one should first treat this haemorrhoid’

Finally, there are rather frequent combinations of déjà with generic verbs like faire ‘to make’ or vouloir ‘to want’ where the adverb brings in an additional semantic aspect. In these expressions typical of CF (and perhaps more generally of African varieties of French), the meaning of déjà contributes to narrow down the sense of constructions that replace more specific verbs in HF (Drescher 2012b).19 (28)

la plaie là a déjà fait combien d’années maman? ‘this wound it has made already how many years mom?’

(29)

dès que vous allez arriver là-bas tu attends quand elle veut déjà accoucher avant de lui donner

|| 19 Camille Abolou (University of Bouaké, Ivory Coast) evokes also the possibility of a grammatical transfer of aspectual meanings from local contact languages. According to him, déjà indicates the perfective aspect that is the accomplishment of the activity expressed by the verb (p.c.).

96 | Martina Drescher

‘as soon as you will arrive there you wait when she wants already to give birth before you give it to her’ Vouloir déjà is equivalent to the locution ‘to be ready’, whereas déjà faire seems to be a paraphrase of the verb ‘to last’. These innovations attest the speaker’s lexical creativity in a context where French as a second language (L2) is constantly in contact with other languages. The new expressions may become routinized through recurrent usage and finally end up as parts of CF’s endogenous linguistic norm. The outcome of the analysis of the Cameroonian data is that the phasal and iterative uses of déjà are abundant in the face-to-face interactions whereas the pragmatic and discursive uses are surprisingly rare. In fact, only two occurrences have been identified, by contrast to the media corpus where context-level uses are rather frequent. This is especially astonishing as one would expect a greater use of déjà as a discourse marker in direct interactions. Obviously, the frequency and distribution of the different uses is affected by interaction types, communicative genres or the formality of the situation. Moreover, it is possible that speaker’s fluency in French plays a role here since the pragmatic and discursive uses seem more abstract and subtle than the temporal and aspectual ones and therefore perhaps more difficult to master. The speakers in the media corpus altogether seem to have a better command of French. Here the discoursemarking uses – and particularly the structuring device in relation with the verbalization process – are nearly as frequent as the temporal and aspectual meanings since 47 of 100 occurrences belong to this category while the so-called interactional uses are completely absent in my data. In addition, the adverb becomes part of certain constructions where it contributes to specify the meaning of the whole unit.

6 Conclusion The analysis of déjà underlines the need for empirical approaches to the study of pragmatic and discursive contact phenomena. For only a data based investigation is capable of discovering the uses typical for spontaneous, spoken language that otherwise run the risk of being considered deviant or simply overlooked by the linguist working with construed examples. An empirical basis is all the more important since the intuition in this field is fragile and sometimes hardly reliable. In addition, one has to cope with the complex, context-sensitive meaning of the adverb déjà that is difficult to grasp as previous studies with

Hybridized discourse markers in Cameroonian French? | 97

their various classifications attest. In my opinion, the vulnerability of discourse markers in contact situations is at least partly due to the context-sensitivity of their meaning. The presence of other languages fosters not only borrowing, but also semantic and pragmatic shifts that may both reduce or complicate the basic meaning. In the case of the adverb déjà especially its pragmatic uses seem to differ from those in HF. First, the interactional uses are not attested in the corpus. Second, the discourse-marking meanings seem rather different given that there is no conjunctional use of the adverb. In addition, the thematic or listing use does not cover the whole range of the adverb’s discursive functions since the majority of the occurrences points to a structural meaning in relation with speech processing and discourse production. Hence, there is some evidence for a hybridization of the adverb in CF and it seems that déjà participates in the appropriation process that characterizes the dynamics of French in Cameroon. From the results of the previous analysis one can assume that the differences concern not only the adverb’s meaning, but also affect its distribution – the different collocations exemplify this aspect – and its frequency as testifies the case of the structural use. But it also shows the need for additional analyses based on comparable data from HF. For the time being, studies dedicated to the uses of déjà in spontaneous spoken French, where the adverb might have similar functions as in CF, are lacking. Therefore, it is rather difficult to tell typically oral uses from those specific to a regional variety, that is CF. In other words: Are the differences we noticed the result of production constraints, that is the oral “conception” of the data in the sense of Koch & Oesterreicher (1990/2011)? Or are they the result of the specific linguistic environment of Cameroonian French that is its ecology? Only in that second case do we deal with hybridized discourse markers. Even though the previous analysis suggests that at least some of the pragmatic and discursive meanings of déjà might be related to its ecology, additional studies both on comparable data from the main contact languages and on HF are necessary in order to validate or invalidate this hypothesis.

References Abolou, Camille Roger. 2010. Des marqueurs ke et non en français populaire d’Abidjan: stratégies discursives et modélisation. Le français en Afrique 25. 325–342. Auer, Peter & Susanne Günthner. 2003. Die Entstehung von Diskursmarkern im Deutschen – ein Fall von Grammatikalisierung? InLiSt 38. Biloa, Edmond. 2003. La langue française au Cameroun. Bern: Lang. Buchi, Eva. 2007. Approche diacronique de la (poly)pragmaticalisation de fr. déjà (« Quand le grammème est-il devenu pragmatème, déjà ? »). In David Trotter (ed.), Actes du XXIVe

98 | Martina Drescher

Congrès International de Linguistique et de Philologie Romanes (Aberystwyth 2004), 251– 264. Tübingen: Niemeyer, vol. 3 Cadiot, Anne, Oswald Ducrot, Than-Binh Nguyen & Anne Vicher. 1985. Sous un mot, une controverse: Les emplois pragmatiques de toujours. Modèles linguistiques 7. 105–124. Calamaro, Francesca. 2012. L’adverbe « déjà » en français camerounais: Analyse d’un corpus radiophonique. Université de Bayreuth: Mémoire de maitrise. Chevalier, Gisèle. 2000. Description lexicographique de l’emprunt well dans une variété de français parlé du sud-est du Nouveau-Brunswick. In Danièle Latin & Claude Poirier (eds.), Contacts de langue et identités culturelles. Perspectives lexicographiques, 85–97. Québec: Les Presses de l’Université Laval. Chevalier, Gisèle. 2007. Les marqueurs discursifs réactifs dans une variété de français en contact intense avec l’anglais. Langue française 154. 61–77. Dimmendaal, Gerrit J. 2014. Attitude markers in Nilotic: A cross-linguistic perspective. Studies in Nilotic Linguistics 8. 1–14. Dostie, Gaétane. 2004. Pragmaticalisation et marqueurs discursifs. Analyse sémantique et traitement lexicographique. Bruxelles: De Boeck/Duculot. Drescher, Martina. 2000. Eh tabarnouche! c’était bon. Pour une approche communicative des jurons en français québécois. Cahiers de Praxématique 34. 133–160. Drescher, Martina. 2009. Sacres québécois et jurons français: Vers une pragmaticalisation des fonctions communicatives? In Béatrice Bagola (ed.), Français du Canada – Français de France VIII. Actes du Colloque international de Trèves du 12 au 15 avril 2007, 177–185. Tübingen: Niemeyer. Drescher, Martina. 2012a. Crosscultural perspectives on advice. The case of French and Cameroonian radio phone-ins. In Stefan Hauser & Martin Luginbühl (eds.), Contrastive media analysis – approaches to linguistic and cultural aspects of mass media communication, 11–45. Amsterdam/Philadelphia: Benjamins. Drescher, Martina. 2012b. Le français burkinabé entre divergences et convergences. L’exemple des locutions avec faire. Le français en Afrique 27. 169–188. Drescher, Martina. 2014. La dimension pragmatico-discursive du français en contact. L’exemple des consultations à la radio camerounaise. Journal of Language Contact 7. 62– 92. [http://booksandjournals.brillonline.com/content/journals/19552629/7/1] Drescher, Martina & Barbara Frank-Job (eds.). 2006. Les marqueurs discursifs dans les langues romanes: approches théoriques et méthodologiques. Frankfurt/Main etc.: Peter Lang. Drescher, Martina & Francesca Calamaro. in print. e:t nou:s allo:ns nous intéresser déjà à:: – Quelques emplois de déjà en français camerounais. In Alain Berrendonner, Maj-Britt Mosegaard Hansen & Rodica Zafiu (eds.), Actes du XXVIIe Congrès international de linguistique et de philologie romanes (Nancy, 15–20 juillet 2013). Section 10: Linguistique textuelle et analyse du discours. Franckel, Jean-Jacques. 1989. Étude de quelques marqueurs aspectuels du français. Genève: Droz. Frantext http://www.frantext.fr/ Hoepelman, Jaap & Christian Rohrer. 1980. Déjà et encore et les temps du passé du français. In Jean David & Robert Martin (eds.), La notion d’aspect, 119–143. Paris: Klincksieck. http://hal.archives-ouvertes.fr/docs/00/14/92/75/PDF/BuchiACILPR24Deja.pdf]. Inventaire des particularités lexicales du français en Afrique Noire (32004). Vanves, EDICEF/ AUF. [= IPLFAN]

Hybridized discourse markers in Cameroonian French? | 99

Klaeger, Sabine. 2010. Ce n’est même pas possible. Syntaxe et fonctions de même en français burkinabé. In Martina Drescher & Ingrid Neumann-Holzschuh (eds.), La syntaxe de l’oral dans les variétés non-hexagonales du français, 37–51. Tübingen: Stauffenburg. Koch, Peter & Wulf Oesterreicher. 1990/2011. Gesprochene Sprache in der Romania. Französisch, Italienisch, Spanisch. Berlin/New York: de Gruyter. Manessy, Gabriel. 1994. Le français en Afrique Noire. Mythe, stratégies, pratiques. Paris: L’Harmattan. Martin, Robert. 1980. Déjà et encore: de la présupposition à l’aspect. In Jean David & Robert Martin (eds.), La notion d’aspect, 168–180. Paris: Klincksieck. Maschler, Yael. 2000. What can bilingual conversation tell us about discourse markers? Introduction. International Journal of Bilingualism 4(4). 437–445. Matras, Yaron. 2000. Fusion and the cognitive basis for bilingual discourse markers. International Journal of Bilingualism 4(4). 505–528. Matras, Yaron. 2007. The borrowability of structural categories. In Yaron Matras & Jeanette Sakel (eds.), Grammatical borrowing in cross-linguistic perspective, 31–73. Berlin/New York: de Gruyter. Matras, Yaron. 2010. Language contact. In Mirijam Fried, Jan-Ola Östman & Jef Verschueren, (eds.), Variation and change. Pragmatic perspectives, 203–214. Amsterdam/Philadelphia: John Benjamins. Mendo Zé, Gervais. 1992. Une crise dans les crises. Le français en Afrique noire francophone. Le cas du Cameroun. Paris: ABC. Mendo Zé, Gervais. 2009. Insécurité linguistique et appropriation du français en contexte plurilingue. Paris: L’Harmattan. Morency, Patrick. 2011. Déjà: un marqueur procédural de subjectivisation. Tranel 51, 19–43. Mosegaard Hansen, Maj-Britt. 2003. From aspectuality to discourse marking: The case of French déjà and encore. Belgian Journal of Linguistics 16, 23–51. Mosegaard Hansen, Maj-Britt. 2008. Particles at the semantics/pragmatics interface: Synchronic and diachronic issues. A study with special reference to the French phasal adverbs. Oxford: Bingley. Mosegaard Hansen, Maj-Britt & Erling Strudsholm. 2008. The semantics of particles: Advantages of a contrastive and panchronic approach. A study of the polysemy of French déjà and Italian già. Linguistics 46(3). 471–505. Neumann-Holzschuh, Ingrid. 2008. Oui YEAH! Zu Syntax und Pragmatik ‘gedoppelter’ Diskursmarker im Louisiana-Französischen. In Elisabeth Stark, Roland Schmidt-Riese & Eva Stoll (eds.), Romanische Syntax im Wandel, 469– 485. Tübingen: Narr. Neumann-Holzschuh, Ingrid. 2009. Les marqueurs discursifs ‘redoublés’ dans les variétés du français acadien. In Beatrice Bagola (ed.), Français du Canada – Français de France VIII. Actes du Colloque international de Trèves du 12 au 15 avril 2007, 137–155. Tübingen: Niemeyer. Ong, Walter J. 1982. Orality and literacy. The technologizing of the word. London/New York: Methuen & Co. Ltd. Paillard, Denis. 2004. « Déjà ». Communication at the Chronos VI conference at Université de Genève, 1–12. http://llf.linguist.jussieu.fr/llf/Gens/Paillard/dejaPaillardChronos6.pdf (checked 10/06/14). Sankoff, Gilian, Pierrette Thibault, Naomi Nagy, Hélène Blondeau, Marie-Odile Fonollosa & Lucie Gagnon. 1997. Variation in the use of discourse markers in a language contact situation. Language Variation and Change 9. 191–217.

100 | Martina Drescher

Schiffrin, Deborah. 1987/1996. Discourse markers. Cambridge: Cambridge University Press. Skattum, Ingse. 2012. Bon, marqueur discursif en français parlé au Mali. Le Français en Afrique 27. 201–217. Thomason, Sarah Grey. 2001. Language contact. An introduction. Edinburgh: Edinburgh University Press. Valikangas, Olli. 1985. Paradigmes logiques et contacts de langues: A propos de déjà, encore et schon, noch. In Jean-Claude Bouvier (ed.), Contacts de langues. Discours oral, 15–26. Aix-en-Provence: Université de Provence. Vol. 7. Valikangas, Olli. 2004. Wie heißt er schon ? – Comment s’appelle-t-il déjà? Zur Problematik der Erinnerungsfragen. In Jorma Koivulehto, Irma Hyvärinen, Petri Kallio, Jarmo Korhonen & Leena Kolehmainen (eds.), Etymologie, Entlehnungen und Entwicklungen. Festschrift für Jorma Koivulehto zum 70. Geburtstag, 423–437. Helsinki: Société néophilologique. Vincent, Diane. 1993. Les ponctuants de la langue et autres mots du discours. Québec: Nuit Blanche Éditeur. Zang Zang, Paul. 1998. Le français en Afrique. Normes, tendances évaluatives, dialectalisation. München/Newcastle: Lincom.

Andrei A. Avram

Kolonial-Deutsch and Français Tirailleur A comparative overview Abstract: The present paper looks at two pidginized varieties of German and French, respectively. Kolonial-Deutsch is a planned pidgin, which has never come into being, while Français Tirailleur is at least in part the outcome of deliberate language policies. Kolonial-Deutsch and Français Tirailleur share a number of structural features, which are comparatively analyzed. The paper focuses on the analysis of the morphology, syntax and vocabulary of KolonialDeutsch and Français Tirailleur. The findings are discussed within the wider framework of pidgin and creole linguistics as well as of the German and French colonial language policies. It is shown that the structural similarities between Kolonial-Deutsch and Français Tirailleur are, to some extent, a rather expected outcome, given that both varieties reflect the racist stereotypes typical of the prevailing colonialist ideology of the time. Keywords: language policies, planned pidgin, word formation, syntax, vocabulary

1 Introduction The present paper aims to compare the structures of two pidgins, KolonialDeutsch and Français Tirailleur, with a focus on their morpho-syntax and vocabulary, and on their ideological underpinnings. A major difference between the two pidgins at issue should be mentioned from the very outset. Kolonial-Deutsch was an artificially constructed pidgin (Gerhardt 1985, Mühlhäusler 1997: 260–261, Perl 2001, Mühleisen 2005, 2009), which never got to the point of being actually used. Français Tirailleur was a “patois évolutif” (Clayton 1994: 414), i.e. subject to depidginization via a more or less permanent contact with French, its lexifier language. This has made it necessary to operate a selection in the examples used to illustrate its structure. Thus, all examples are exclusively from “non-depidginized” Français Tirailleur, i.e. from the variety exhibiting in the least the influence of French.

|| Andrei A. Avram: University of Bucharest, Department of English, 7–13 Pitar Moş Str., 010451 Bucharest, Romania. E-mail: [email protected]

102 | Andrei A. Avram

In the case of Kolonial-Deutsch all examples are taken from the only available source, namely Schwörer (1916). As for Français Tirailleur, the examples are not limited to the data in Delafosse (1904) and Anon. a. (1916), on which most previous work (e.g. Manessy 1978, Houis 1984, Corne 1999) almost exclusively relies. The corpus of Français Tirailleur includes letters (Dupratz 1864), travel accounts (Barret 1888, Béchet 1889, Mandat-Grancey 1900, Colrat de Montrozier 1902), diaries (Augouard 1905, Desjardins 1925), memoirs (Marceau 1911, Baratier 1912), articles (Leymaire 1898, Cantilly 1909, Marie-Victoria 1921), a monograph on the languages spoken in Côte d’Ivoire (Delafosse 1904), and two textbooks (Anon. a. 1916, Anon. b. 1918). All examples appear in the orthography used in the sources; relevant portions are highlighted in boldface. The examples from Français Tirailleur include the date of the attestation whenever there are discrepancies between the year of publication of the source and the year of the attestation: the latter is a date explicitly mentioned in the source. The length of the quotations has been kept to a reasonable minimum. Relevant items appear in boldface. The paper is organized as follows. Section 2 outlines the socio-historical context in which Kolonial-Deutsch was constructed and in which Français Tirailleur was formed. Particular attention is paid to the prevailing colonialist ideology of the time, to the widely popular artificial languages as well as to the then recently introduced direct methods of teaching foreign languages. The following sections, 3 through 5, are concerned, in turn, with the main characteristics of word formation, the syntax, and the vocabulary of Kolonial-Deutsch1 and Français Tirailleur2 respectively. The findings are summarized and discussed in section 6.

2 Socio-historical context 2.1 Colonialism and language issues As shown by e.g. Van den Avenne (2005), Mühleisen (2005, 2009), the second half of the 19th century and the first decades of the 20th century witness, among others, four strands which prove relevant to the topic at issue: the numerous attempts at theorizing the justification of European colonialism in terms of the

|| 1 See also Perl (2001) and Mühleisen (2005: 34–36). 2 See also Manessy (1984), Houis (1985), and Corne (1999: 199–201).

Kolonial-Deutsch and Français Tirailleur | 103

so-called “mission civilisatrice” of the European colonial powers of the time; the popularity of artificial languages, such as Esperanto, Ido, Volapük, etc.; the newly introduced direct methods of teaching foreign languages, in particular the Berlitz method; the different language policies pursued by the European colonial powers, including in “language wars”.3 Obviously, there are varying degrees in which these particular strands mix as ingredients, depending on the particular socio-historical context in particular countries.

2.2 Germany Colonial administrators as well as missionaries were among the first to raise the issue of the necessity of a means of interethnic communication in Germany’s colonies. They were quick to realize that German did not really stand a chance of being reasonably well learned by a significant segment of the indigenous populations in Germany’s colonies in West Africa, South-West Africa, East Africa, and Asia. Several potential solutions were envisaged. Significantly, they all shared the assumption that an artificially constructed language would be the best possible choice, be it for a transitional period only. As early as 1891, i.e. only several years after the colonial arrangements at the Conference in Berlin, “a propaedeutical transition language, a constructed Papuan or ‘Pigeondeutsch’”4 was proposed for German New Guinea (Voeste 2005: 167). This language, whatever the final decision, was to “be easy to learn […] and later replaced by German” (Voeste 2005: 167). A report by Governor Albert Hahl, German New Guinea, dated 10 August 1903 (quoted in Voeste 2005: 167), also stressed the fact that “an easy to learn transition language is needed with the natives” Another solution to the problem of communication in a multilingual setting was to teach and subsequently use an artificial language. Thus, according to Voeste (2005: 168), a Catholic mission in German New Guinea considered using the artificial language Volapük as a medium of instruction in primary school. While the above proposals never materialized, during World War I such ideas did find their way into concrete and fleshed-out projects. As already noted

|| 3 In the sense of Calvet (1987). 4 Where Pigeon probably reflects English Pigeon, which was at the time a frequent misnomer or a misspelling of the term “pidgin”.

104 | Andrei A. Avram

by Mühlhäusler (1997: 260), “two attempts to artificially create a simplified German […] Baumann’s Weltdeutsch (1916) and Schwörer’s Kolonial-Deutsch (1916)”5 are known, both “written during the First World War in expectation of a German victory resulting in large-scale colonial expansion”. Mühlhäusler (1997: 260) writes that “Weltdeutsch was designed primarily for the use of allies and friends, particularly those in Eastern Europe”, i.e. “for the use of ‘civilized’ people, the primary concern [...] was to keep the language as close as possible to High German”. According to Mühlhäusler (1997: 260) and Werkmeister (2010), the main defining characteristics of Weltdeutsch are a quasi-phonemic orthography and the elimination of non-functional variation in the grammar and the vocabulary. Kolonial-Deutsch, on the other hand, was specifically designed as a version of simplified and reduced German, to be taught to the indigenous population in Germany’s colonies (Mühlhäusler 1997: 260, Perl 2001, Mühleisen 2005, 2009, Orosz 2011). Its author, Emil Schwörer, builds his case by briefly considering and subsequently dismissing a number of alternative solutions. These include the choice of a local language spoken by the “natives”, such as Swahili, French, Pidgin English, artificial languages, German, and Pidgin German. Swahili is dismissed given that it is not even known by all “natives” in East Africa, not to mention the other German colonies (Schwörer 1916: 10). Further, according to Schwörer (1916: 11), “it would be absolutely out of the question to introduce French or […] Pidgin English” since this “would be unthinkable for national and […] language-esthetic reasons”. Pidgin English, varieties of which were spoken in German colonies such as Togo and New Guinea, is an enemy and a target of choice for derogatory and dismissive remarks. For Schwörer (1916: 26), Pidgin English is “an ugly language” and, by way of consequence, “any similarity between the corrupted Pidgin English and the systematically constructed KolonialDeutsch should be rejected outright”. Finally, “modern artificial languages such as Volapük and Esperanto would be even more impossible in an African context” (Schwörer 1916: 11). German itself, for all its merits, will not do. In a blatantly racist passage Schwörer (1916: 31) argues that German is “such an exceedingly difficult language” and that it “shall never qualify as a lingua franca of the natives, whose limited spiritual energy is hindered by […] their helplessness and often also by external circumstances”. Ever the pragmatically motivated one, Schwörer (1916: 31) goes on and adds: “take into consideration […] the fact that

|| 5 Mühlhäusler (1997: 260) does not mention Baumann (1915), his first description of Weltdeutsch as what he calls a “language for mutual comprehension” (Verständigungssprache) of the Central Powers, and more generally, a “world auxiliary language” (Welthilfsssprache).

Kolonial-Deutsch and Français Tirailleur | 105

we have all the reasons to spare them [= the ‘natives’] excessive and superfluous spiritual work”. Finally, Pidgin German is not an option either since this would be the result “in short, of a degeneration of [the German] language, not of a language reform” (Schwörer 1916: 62). Having eliminated all potential competitors, Schwörer (1916: 15) also discusses some of the advantages which his Kolonial-Deutsch presents, in the wider context of Germany’s colonial policies and of its national and colonial ideologies: it is a working language for the German colonizers; it can be used between Germans and “natives” as well as among different linguistic groups of “natives”; it makes it possible to transfer unreliable “natives” from one colony to the other, thus effectively neutralizing what would be now termed potential security risks; it is a powerful symbol of German authority.

2.3 France The issue of an adequate means of interethnic communication in France’s colonial empire was addressed by a large number of colonial administrators, army officers, and educationalists. Many of the proposals put forth shared the core idea that a variety of simplified and reduced French would be the best possible solution. However, a controversial point was whether this would be a variety or several varieties emerging naturally in the multilingual settings of France’s colonies, e.g. the outcome of simplifications and reductions by the “natives” or rather a systematically constructed one. Consider, for instance, the views articulated by Étienne Aymonier at the end of the 19th century. Basically, Aymonier (1890: 35) promotes “a very reduced” French which “will be sufficient for expressing concrete ideas”. According to Aymonier (1890: 38), “teaching reduced and simplified French is the most delicate undertaking, which requires all our efforts and all our hard work”. Adopting a cost-benefit approach, Aymonier (1890: 65) suggests that “we organize systematically the teaching of what I call sabir, the nègre speech […] which offers the capital advantage that it can be spread all over at a small cost and in a very short period of time”. Moreover, a number of concrete suggestions as to the nature of this simplified and reduced variety of French are made. Thus, Aymonier (1890: 35) proposes that “we provisionally eliminate in the French language […] the grammatical difficulties, most of the synonyms and abstractions, almost all the conjugations (except, for example, for some of the third persons in the singular and of impersonal verbs)”. Aymonier (1890: 67–68) insists in particular on verbs, specifying “the grammar would require elimination of tenses and moods, while largely preserving the verb in the infinitive,

106 | Andrei A. Avram

with the third persons in the singular of usual expressions: il peut, il faut, etc.”. Also, Aymonier (1890: 35) explicitly advocates categorial multifunctionality6: “Amour […], which doubles aimer, would have to be eliminated; parler shall also mean parole”. Aymonier (1890: 68) aptly summarizes in a nutshell the result of such simplifications and reductions: “to sum up, sentences such […] moi aller promener demain”. Similar views were expressed by e.g. Irénée Carré, an educationalist and school inspector (quoted by Lehmil 2007: 215), who advocated teaching rudimentary French to the “natives”, which “would be nothing else than the French language as simplified as possible, i.e. stripped of all variations in gender and number, preserving only the infinitive of the verb, and consisting of only simple, independent and isolated sentences: no spelling, and consequently, no grammar”. On the other hand, at the Congress of the Alliance Française in 1893, one of the participants even proposed the diffusion of several varieties of reduced French, i.e. one for each colony, since “the inhabitants of these colonies […] have sufficiently simplified French so that we do not have to help them” (Lehmil 2007: 213). This controversy regarding agentivity, i.e. who should be or who are the actors responsible for the emergence and development of simplified and reduced French surrounds Français Tirailleur as well. Français Tirailleur, also known as petit-nègre, petit-français (Valdman 1978: 39), Français Tiraillou and moi-y’a-dit, was used in the French army units of West African tirailleurs, and its use ceased officially in 1926.7 The origin of Français Tirailleur is a matter of some debate. Valdman (1978: 39–40), for instance, claims that Français Tirailleur “can be undoubtedly be traced back directly to Sabir (Petit Mauresque) in North Africa”, but provides absolutely no evidence in support of this claim. Another view, quite frequently voiced, is that Français Tirailleur was created exclusively by the “natives”. According to Delafosse (1904: 263), petit-nègre “is spoken by our indigenous tirailleurs and our servants, almost identically in Tonkin and in West Africa, which would tend to prove that it is the natural and rational simplification of our language”. Delafosse (1904: 264) is adamant in his conclusions: “it is truly the Negro – or, more generally speaking, the primitive –

|| 6 In the sense of Mühlhäusler (1997: 137). 7 See Règlement provisoire du 7 juillet 1926 pour l’enseignement du français aux militaires indigènes (apud Van den Avenne 2005: 134): “it is strictly forbidden to speak sabir or petit-nègre”, qualified as “old deviations”. See also Poulot (2011).

Kolonial-Deutsch and Français Tirailleur | 107

who has forged petit-nègre, by adapting French to his state of mind”.8 The author of a textbook of Français Tirailleur (Anon. a. 1916: 5) writes that “our black tirailleurs […] have created a language which is called ‘petit-nègre’”. A similar opinion is expressed by general Mazillier, in a letter dated 27 April 1918 (quoted in Van den Avenne 2005: 133): the natives “make up the ‘petit-nègre’ which they use with us and among themselves”. Some modern scholars have also adopted this point of view. According to Gyasi (2012: 226), it was the tirailleurs who were at the origin of the spread of “français-tirailleur” or “français petit nègre”. The opposite view, according to which the “natives” had no access to French and that it was simplified and reduced French that was actually used with and even taught to them, is found in a considerable number of sources. As early as the 1880s, Barret (1888) explicitly states that “as we usually speak to them in ‘nègre’, nègre they give us back in a mixture which is neither French nor African”. A French officer (Marceau 1911: 23) writes that “after having for a while heard approximate French spoken, the Sudanese recruit manages, in this Berlitz in action, to understand certain concrete ideas and even abstractions”. Teaching is explicitly mentioned by the same author (Marceau 1911: 24), “if you add to this the adjective ‘bon’ you would have taught the tirailleur half of the French language”. There are also descriptions of Français Tirailleur being taught. Consider first a quotation (Baratier 1912: 22) illustrating the teaching of vocabulary, as witnessed by an officer of the French army: the young pupils were absorbing the gestures and words of their NCO. Pointing to his eye, old Dialiké said: “Ça, y a œil”. And everybody, putting their finger on their eye would repeat: “Ça, y a œil”. A minute later, after having explored all the details of the face, Dialikés finger went down and stopped under his chin and all repeated after him: “Ça, y a cou”.

Grammar also appears to have been taught: the NCO “was explaining the use of possessive pronouns: mon, ton, son, notre, votre, leur” (Baratier 1912: 22). The teaching of Français Tirailleur is also explicitly mentioned or recommended in a number of documents issued by military authorities. Thus, Anon. b. (1918: 3) refers to the necessity “to teach as quickly as possible our young tirailleurs coming from our West Africa the words and expressions used by their NCOs”. In a note dated 11 March 1918 (quoted by Van den Avenne 2005: 132), general Mangin concludes that “the truth is that unity of language must be pursued via the teaching of simplified French which already constitutes the language of the

|| 8 Delafosse (1904) is considered by many to be the first description of Français Tirailleur. In fact, as will be seen in various quotations, it is equally a prescriptive work.

108 | Andrei A. Avram

tirailleur” and also specifies that “this teaching is carried out by the direct method (the Berlitz method)”. A similar conclusion is reached by general Mordacq, in a letter dated 23 March 1918 (quoted by Van den Avenne 2005: 132): “unity of language can be usefully and easily pursued only by teaching simplified French, already spoken by our indigenous NCOs and by a large number of tirailleurs”. Accounts by civilians also point in the same direction. According to Cousturier (1920: 103), the tirailleurs could have learned French “if only they had the opportunity to hear it and practice it [but] they never had this opportunity in France”, given that “their instructors knew how to generalize an Esperanto”. Moreover, French civilians themselves appear to have contributed to the use of Français Tirailleur: I witnessed the arrival of the first tirailleurs in the region of Saint-Raphaël. The population immediately understood that if ‘petit-nègre’ was good enough for military life, it was good enough for commercial life and immediately substituted it for French (Cousturier 1920: 105).

It would seem that this view was also espoused by the tirailleurs who “found out, faced with laughter, that their language ridicules them: ‘this is French for tirailleurs only’ they sadly realize” (Cousturier 1920: 105); one tirailleur is quoted as concluding that “these are words devised by Europeans to poke fun at the Senegalese” (Cousturier 1920: 105). Finally, the same conclusion is formulated by a number of modern scholars. These include linguists such as Houis (1984: 15), for whom Français Tirailleur was systematized into an autonomous variety, more of “an ideological variety of French born out of a situation of domination” than a real pidgin, and Corne (1999: 200), who describes “the français tirailleur used by officers and French NCOs” as “a language of command, deliberately made up of easily learned formulae”. Fogarty (2008: 161), a military historian, also writes that “officers taught soldiers a simplified version of French [= Français Tirailleur]”. One last view, also endorsed in the present paper, is that both parties, i.e. the “natives” and the French, contributed to the formation of Français Tirailleur. Chaudenson (2003: 54) defines Français Tirailleur as “a slightly artificial pidgin, which has its origins in the approximate varieties produced by speakers in an exolingual situation and in didactic generalizations said to facilitate and accelerate the teaching of this minimal French”. To sum up, unlike Kolonial-Deutsch, which is an entirely artificial language, Français Tirailleur is only in part a planned language.

Kolonial-Deutsch and Français Tirailleur | 109

3 Aspects of word formation 3.1 Kolonial-Deutsch Schwörer (1916) does not discuss any means of word formation. The only relevant cases are ambiguous in that they allow for an interpretation in terms of compounding or, alternatively, of affixation. Consider the following examples: (1a)

(1b)

Arbeit-Mann work man ‘worker’ Last-Mann burden man ‘bearer’

[Schwörer 1916: 45, fn. 5]

[Schwörer 1916: 45, fn. 5]

While Mann appears to be the second member of a compound, it could equally well be analyzed as the equivalent of the German agentive suffix -er.9

3.2 Français Tirailleur Affixation is generally absent from Français Tirailleur. The absence of affixation is possibly the consequence of an ideological tenet: “one obviously must use only the simplest forms of words” (Delafosse 1904: 264). The following is one of the very few instances of suffix use identified in the sources: (2)

attaqu-ation 1889 attack (V)-ation ‘attack (N)’

[Béchet 1889: 78]

Note that although the derived form obtained is not attested in French, which might suggest that, for some speakers at least, -ation was a suffix. Corne (1999: 201) claims that total “reduplication marking intensity or continuity is a frequent feature”. However, his single example illustrates in fact repetition: (3)

tirailleur ya besoin tirer, tirer, tirer toujours 1916 tirailleur must fire fire fire always ‘the soldiers must keep on firing’

[Corne 1999: 2011]

|| 9 Circumstantial evidence for the second interpretation can be adduced from several Englishlexifier pidgins and creoles in which the reflex of English man functions as an agentive suffix.

110 | Andrei A. Avram

Moreover, the meanings of reduplicated forms in Anon a. (1916) – Corne (1999)’s only source – are identical to those conveyed by repetitions: (4a)

(4b)

courir vite, vite run fast fast ‘run very quickly’ charger vite, vite, vite load fast fast fast ‘load [his gun] very quickly’

[Anon a. 1916: 22]

[Anon a. 1916: 26]

It appears, then, that Français Tirailleur does not exhibit productive reduplication, but rather it resorts to repetition. Such repetitions, including instances of “triplication” and “quadruplication”, are found in other sources: (5a)

(5b)

(5c)

Français, grands, grands, grands! 1878 [Augouard 1905: 52] French great great great ‘The French are really great!’ infirmier sale, sale, sale 1918 [Cousturier 1920: 272] male nurse dirty dirty dirty ‘the male nurse was very dirty’ mourir tout seul! tout seul! tout seul! tout seul 1918 die all alone all alone all alone all alone ‘die completely alone’ [Cousturier 1920: 27])

In light of the above, it can be concluded that total reduplication is not a productive word formation means in Français Tirailleur.

4 Syntax 4.1 Kolonial-Deutsch Schwörer (1916: 47, fn. 1) explicitly advocates categorial multifunctionality: “it is enough if concepts are expressed in one of the word classes, be it as a noun, verb, adjective or also as an adverb”. (6a) (6b)

umsonst sein ‘to be in vain’ and ‘lack of perspective’ [Schwörer 1916: 47, fn. 1] zufrieden sein ‘be happy’ and ‘happiness’ [Schwörer 1916: 47, fn. 1]

However, Schwörer (1916) does proceed with very precise specifications regarding the characteristics of the various word-classes and their uses.

Kolonial-Deutsch and Français Tirailleur | 111

Nouns are characterized by the absence of grammatical gender and of case inflections. The genitive and dative case markers are replaced by analytical structures with von and an respectively: von de Mensch ‘the man’s’, von die Menschen ‘the men’s’; an de Mensch ‘to the man’, an die Menschen ‘to the men’. Plurality is expressed by the invariant marker -en. Determiners include the definite and indefinite articles as well as demonstrative adjectives. The definite and indefinite articles are invariant in form: de in the singular, die in the plural ‘the’, and eine ‘a, an’. Demonstratives are somewhat better represented by a larger number of invariant forms: diese ‘this’, de eine ‘this one’, de andere ‘the other one’, de da ‘the one over here’, de dort ‘the one over there’, die da ‘the ones over here’, die dort ‘the ones over there’. Note that in the last four forms, the co-occurrence of the deictics da ‘over here’ and dort ‘over there’ increases their transparency. Attributive possession is expressed by analytical structures: (7a)

de

Name von de Kaiser name of DEF.PL emperor ‘the emperor’s name’ die Feinden von de Kaiser DEF.PL enemy-PL of DEF.SG emperor ‘the emperor’s enemies’

[Schwörer 1916: 58]

DEF.SG

(7b)

[Schwörer 1916: 58]

Invariant possessive adjectives are also listed: meine ‘my’, deine ‘your (SG)’, seine ‘his’ and ihre ‘her’, unsere ‘our’, euere ‘your (PL)’, ihre ‘their’. However, Schwörer (1916: 33) writes that the third person singular and plural possessive adjectives “are to be omitted or replaced with a similar analytical expression” and that the analytical structure “can moreover be often used instead of the possessive”, as in (8b) below: (8a)

(8b)

unsere Sprache 1PL.POSS language ‘our language’

[Schwörer 1916: 33, fn. 3]

de

[Schwörer 1916: 33, fn. 3]

Sprache von uns language of 1PL ‘our language’

DEF

Subject pronouns are as in German, ich ‘I’, du ‘you (SG)’, er ‘he’ and sie ‘she, wir ‘we’, ihr ‘you (PL)’, sie ‘they’, whereas object pronouns are invariant in form and etymologically derived from German dative or accusative forms: mir, dir, ihm and sie in the singular, uns, euch, and sie in the plural.

112 | Andrei A. Avram

Other pronominal forms include the invariant demonstrative das ‘this’ and the relative pronouns de ‘who (SG)’, die ‘who (PL)’, and was ‘which’. Adjectives do not inflect either for number or for case. As for the degrees of comparison, only the absolute superlative occurs, whereas the comparative of superiority is restricted to two basic adjectives: groß, größer, größte ‘big, bigger, biggest’, and gut, besser, best ‘good, better, best’. Only cardinal and ordinal numerals are listed, whose form is that of their German etyma. However, Schwörer (1916: 38) recommends that the number for tens be placed before that for units “since it corresponds to the order of numerals in the Bantu languages”. Verbs, which generally have an invariant form etymologically derived from that of the German infinitive, are preceded by the auxiliary verb tun ‘to do’: (9a)

(9b)

Er tut wohnen in sehr große […] Stadt 3SG do-PRS.3SG live in very big city ‘He lives in a very big […] city’ Niemand tat verstehen mir nobody do-PST.3SG understand 1SG.OBJ ‘Nobody understood me in the beginning’

[Schwörer 1916: 58]

[Schwörer 1916: 56]

Only six auxiliary verbs inflect as in German for person, number and tense: sein ‘to be’10, haben ‘to have’11, tun ‘to do’ (see the examples under (9)), können ‘can’, müssen ‘need, must, have to’, wollen ‘to want’. There are only two tenses: the Present and the Past Tense, for these auxiliary verbs – the Present and the Perfect (with the auxiliaries sein and haben) – for the other verbs. As for the future, it “can be expressed by means of adequate paraphrases” (Schwörer 1916: 36) or by “a clarifying addition”, such as an adverbial of time. There are only two moods, the indicative and the imperative. The imperative consists of the bare form of the verb. Schwörer (1916: 36) specifies a constraint on the use of the imperative to the effect that “the native has to place the word ‘please’ in front of every imperative when addressing a White”. That is, distinct imperative structures should directly reflect strict status hierarchies and hegemonic relations. Finally, only the active voice is preserved while “the passive is eliminated” (Schwörer 1916: 35) or “is replaced by adequate paraphrases” (Schwörer 1916: 36).

|| 10 Schwörer (1916: 35, fn. 1) suggests that, since they would be easier to pronounce, bis and is could replace bist ‘you (SG) are’ and ist ‘he/she is’ respectively. 11 According to Schwörer (1916: 35, fn. 1) haben, which is easier to pronounce, could replace habt ‘you (PL) have’.

Kolonial-Deutsch and Français Tirailleur | 113

The negator nicht12 (Schwörer 1916: 37) is placed as in German: (10a) deine Sprache ist noch nicht gut 2SG.POSS language is still NEG good ‘your language is not good yet’ (10b) ig wissen nit 1SG know NEG ‘I don’t know’

[Schwörer 1916: 579

[Schwörer 1916: 57]

There are a small number of adverbs, mostly of place and of time. The question words are: wer ‘who’, was/was für eine ‘what’, wo ‘where (location)’, woher ‘from where (direction)’, wann ‘when’, wie ‘how’, and warum ‘why’. A rather large number of prepositions are retained, which are “generally used as in German” (Schwörer 1916: 38): an ‘at, to’, auf ‘on’, aus ‘from’, bei ‘at’, bis ‘until’, für ‘for’, gegen ‘against’, in ‘in’, mit ‘with’, neben ‘near, next to’, ohne ‘without’, seit ‘since’, um ‘at’, über ‘over’, and zu ‘to’. The conjunctions and complementizers consist of: aber ‘but’, auch ‘also’, daß ‘that’, denn ‘since’, oder ‘or’, und ‘and’, vor ‘before’, weil ‘because’, and wenn ‘when; if’. Word order is dealt with in some more detail, since it needs to satisfy a variety of criteria. First, “the style of K.D. [= Kolonial-Deutsch] could be compared to a certain extent to the military way of speaking, just like the relation of the White to the native, who is always subordinate to him” (Schwörer (1916: 40). Word order, then, must reflect the power differential. Second, “word order is to be […] as simple as possible, clear and natural” (Schwörer 1916: 40). However, Schwörer (1916: 40, fn. 1) hastens to add that “the German in the colonies cannot be supposed to speak K.D. in the same simple, primitive forms, which are usual and adequate for the native”. Third, “words which belong together should not be separated by a number of intervening words” Schwörer (1916: 41). The racist assumption is, obviously, that the “natives” would not be able to cope with an intervening adjunct or a heavy NP, which should therefore be shifted, as in (11a) and (11b) respectively: (11a) Das ist gewesen nit gut für meine DEM is been NEG good for 1SG.POSS ‘This wasn’t good for my work.’

Arbeit work

[Schwörer 1916: 56]

|| 12 Schwörer (1916: 37) also considers the use of an alternative, phonetically simpler form, nit.

114 | Andrei A. Avram

(11b) Er hat gehabt eine sehr gute und feste Regierung 3SG has had INDEF very good and strong government ‘He has had a very good and strong government’ [Schwörer 1916: 58] Fourth, word order must generally be SVO, “even more regularly than in German” (Schwörer 1916: 41), including in subordinate clauses: (12a) wir müssen haben, was de Kaiser tat sagen 1PL must have what DEF emperor did say ‘we must have what the emperor said’ (12b) Wenn ich habe genug Geld if 1SG have enough money ‘If I have enough money’

[Schwörer 1916: 59]

[Schwörer 1916: 57]

Subject-verb inversion is restricted to interrogative sentences and to sentences starting with an adverb, in particular when the subject is a personal pronoun: (13a) Tust du nun wissen? now know do 2SG ‘Do you know now?’ (13b) dann ist de Nutzen viel größer DEF profit much higher then is ‘the profit is then much higher’

[Schwörer 1916: 57]

[Schwörer 1916: 59]

Clauses are coordinated with und ‘and’ or aber ‘but’: (14a) angreifen ist dann schwer möglich und man hat mehr Respekt attack is then difficult possible and one has more respect vor uns [Schwörer 1916: 59] for 1PL ‘to attack [us] is rather difficult and one has more respect for us’ (14b) ist recht, aber deine Sprache ist noch nicht gut is right but 2SG.POSS language is still NEG good ‘that’s right, but your language is not good yet’ [Schwörer 1916: 57] Several types of subordinate clauses are attested: (15a) direct object clause: ich hab nit verstehen können, was er tat befehlen 1SG have NEG understand can what 3SG do-PST.3SG order ‘I couldn’t understand what he ordered’ [Schwörer 1916: 56] (15b) relative clause: wenn wir tun bauen Alles, was wir müssen haben all what 1PL must have when 2PL do-PRS.1PL grow ‘when we have built all that we must have’ [Schwörer 1916: 59]

Kolonial-Deutsch and Français Tirailleur | 115

(15c) conditional clause: viel besser ist, wenn wir tun bauen Alles [Schwörer 1916: 59] much better is if 1PL do-PRS.1PL build all ‘it is much better if we grow everything’ (15d) adverbial clause of reason: Ich will nun wieder halten Schule für euch, 1SG want now again hold school for 2PL weil ich habe Zeit [Schwörer 1916: 57] because 1SG have time ‘Now, I want again to hold classes for you, because I have time’ However, Schwörer (1916: 41, fn. 1) also recommends parataxis/juxtaposition as an alternative to the use of conjunctions and complementizers.

4.2 Français Tirailleur To start with, mentioned should be made of the occurrence of categorial multifunctionality. The examples below illustrate the use of: content both as an adjective, in (16a), and as a verb, in (16b); saleté as an adjective, in (16c); lourd as an adjective and as a verb, in (16d): (16a) nous, ya contents de toi [Marie-Victoria 1921: 488] 1PL COP pleased of 2SG ‘we are pleased with you’ (16b) moi content partir 1899 [Colrat de Montrozier 1902: 155] 1SG want leave ‘I want to leave’ (16c) Mon chemise y en a déjà saleté beaucoup! 1920 1SG.POSS shirt COP already dirt a lot ‘My shirt is very dirty’ [Desjardins 1925: 262] (16d) lourd ‘heavy; to press’ [Anon. a. 1916: 22] Nouns have no gender and no plural. There is neither a definite nor an indefinite article. Some nouns exhibit agglutination13 of the etymological article le/la/l’, un, du, des: (17a) laroute ‘road’ (17b) lebras ‘arm’ (17c) lécole ‘school’ 1918

[Anon. a. 1916: 7] [Anon. a. 1916: 7] [Cousturier 1920: 179]

|| 13 The term is generally used in the literature on pidgin and creole languages.

116 | Andrei A. Avram

(17d) (17e) (17f) (17g)

n’homme ‘man’ doulé ‘milk’ 1918 z’enfants ‘children’ zhommes ‘men’ 1917

[Baratier 1912: 31] [Cousturier 1920: 209] [Marceau 1911: 64] [Cousturier 1920: 65]

Both the absence of the article and its occasional agglutination are commented upon in various sources. Delafosse (1904: 265), for instance, mentions the “suppression of the article or its perpetual preservation, turning into a sort of prefix of the noun: son maison or son la-maison”. Anon. a. (1916: 7) is very explicit with regard to the forms with an agglutinated article: “our indigenous noncommissioned officers […] consider the article and the noun as forming a single word”, and that “they will say […] mon latête, which proves that for them the word is […] latête”. The only demonstrative adjectives are preposed ça ‘this/these’ and postposed y en a là ‘this/these’. The expression of attributive possession appears to have exhibited considerable variation. (18a) possessee + possessor fusil mon camarade gun 1SG.POSS comrade ‘my comrade’s gun’ (18b) possesse + pour + possessor fusil pour mon camarade gun for 1SG.POSS comrade ‘my comrade’s gun’ (18c) possessor + possessee tirailleur fusil tirailleur gun ‘the tirailleur’s gun’ (18d) possessor + son + possessee mon camarade son fusil 1SG.POSS comrade 3SG.POSS gun ‘my comrade’s gun’

[Delafosse 1904: 265]

[Delafosse 1904: 265]

[Anon a. 1916: 14]

[Delafosse 1904: 265]

Note that, although Anon. a. (1916: 14) recommends the structure in (18c), the only structures attested in the texts are those in (18a, b), which can also cooccur, as in (19): (19)

chef son section pour lui chief 3SG.POSS section for 3SG ‘the commanding officer of his section’

[Anon. a. 1916: 19]

Kolonial-Deutsch and Français Tirailleur | 117

Anon. a. (1916: 9) states that “the only possessive adjective which is often used is the adjective mon” and that ton ‘yours’ “is used more rarely”, and recommends the use of the structure possessee + pour + personal pronoun for the remaining persons. However, forms etymologically derived from French possessive adjectives, e.g. son ‘his/her’, are also found, as in (18d) and (19). The personal pronouns are derived etymologically from the French stressed forms: moi ‘I’, toi ‘you (SG)’, lui ‘he/she’, nous ‘we’, vous ‘you (PL), eux ‘they’. However, forms etymologically derived from the unstressed subject pronouns are also found: (20a) je metter lui sale police 3SG hall police 1SG put ‘I will take him to the police station’ (20b) il parti pas 3SG leave NEG ‘He did not leave’

[Delafosse 1904: 264]

[Delafosse 1904: 265]

Other pronominal forms include the possessive pronoun with the structure ça y en a + pour + personal pronoun, and the demonstrative ça ‘this/these’. Adjectives are invariant in form. They are generally placed as in French even though Anon. a. (1916: 9) recommends the structure noun + copula + adjective: (21)

Tirailleur y en a bon tirailleur COP good ‘The good tirailleur’ [lit. ‘tirailleur who is good’]

[Anon. a. 1916: 9]

The only degree of comparison amply documented is the absolute superlative, formed with beaucoup ‘much’ or trop ‘very’, either preceding or following the adjective: (22a) Beaucoup fort! 1878 much strong ‘very strong’ (22b) Moi trop vieux very old 1SG ‘I am very old’ (22c) sauvages beaucoup 1898 savage much ‘very savage’ (22d) y en a beaucoup trop COP much very ‘There are very many’

[Augouard 1905: 71]

[Barret 1888: 142]

[Colrat de Montrozier 1902: 155]

[Anon. b. 1918: lesson 100]

118 | Andrei A. Avram

Only cardinal numerals are attested in the corpus. Anon. a. (1916: 10) writes that they are “always placed after the noun”. In fact, preposed cardinal numerals occur much more frequently, including in the samples in Anon. a. (1916), as in (23b): (23a) Y en a

deux trains 1919 [Desjardins 1925: 109] COP two train ‘There are two trains’ (23b) tirailleur y a besoin tirer trois cartouches [Anon. a. 1916: 27] tirailleur must shoot three bullets ‘the tirailleur must shoot three bullets’ Verbs are invariant in form. According to Delafosse (1904: 265), verbs are etymologically derived from the French infinitive form of verbs of the first conjugation class, from the French past participle, imperative or infinitive changed by analogy with that of the first conjugation class for verbs of the second, third and fourth conjugation classes: parler ‘to speak’, fini ‘to finish’, vouler ‘to want’, défender ‘to defend’. Other sources (not mentioned by Delafosse 1904, Anon. a. 1916, and Anon. b. 1918) from which verbs are etymologically derived include: nouns besoin ‘must’, moyen ‘can’; adjectives: content ‘to want’. Delafosse (1904: 265) and Anon a. (1916: 12) state that: the Present Tense is expressed by bare, uninflected verbs. Unmarked verbs in the Present Tense are, indeed, attested: (24a) toi dire 1898 2SG say ‘you say’ (24b) toi connais 1878 2SG know ‘you know’ (24c) Betsileos bien connaître ça Betsileos well know DEM ‘The Betsileos know that very well’

[Mandat-Grancey 1900: 134]

[Augouard 1905: 94]

[Cantilly 1909: 245]

However, verbs in the Present Tense are very frequently preceded by y a or y en a, and also by y, contra Corne (1999: 200), who only mentions y a: (25a) moi y parler vrai 1SG speak true ‘I am telling the truth’ (25b) Lui, y a bien connaître 1907 3SG well know ‘He knows very well’

[Leymaire 1898: 122]

[Baratier 1912: 27]

Kolonial-Deutsch and Français Tirailleur | 119

(25c) Tirailleur y en a commencé tirer. tirailleur start shoot ‘The tirailleur starts shooting.’

[Anon b. 1918, lesson 36]

The only relevant example in Delafosse (1904) suggests that the Past Tense is expressed by unmarked verbs. However, verbs in the Past Tense may be preceded by y a or y en a, and less frequently by y: (26a) toi y a venir voir mon 2SG COP come see 1SG.POSS ‘you came to see my father’ (26b) Mabiala lui y tuer blanc Mabiala 3SG COP kill white ‘Mabiala killed the white man’ (26c) moi y en a frappe 1898 1SG COP hit ‘I hit [them]’

papa father

[Béchet 1889: 869]

[Leymaire 1898: 122]

[Colrat de Montrozier 1902: 155]

Moreover, Anon. a. (1916: 12–13) includes the following recommendation: “to avoid any confusion, place in the sentence a word which indicates that it is the past”, illustrated by the example below: (27)

Moi y a parti hier. ‘I left yesterday.’

[Anon a. 1916: 13]

Anon a. (1916: 13) states that “the future is formed in the same way as the present”. However, here again y, y a or y en a are also attested: (28a) nous demain y faire parler bon fusil [Leymaire 1898: 122] 1PL tomorrow COP make speak good gun ‘tomorrow we will make our guns talk’ (28b) Ça y a jamais faire soldats 1899 [Baratier 1912: 104] DEM COP never make soldiers ‘They will never become real soldiers’ (28c) France y en a oublier jamais ça. [Anon b. 1918, lesson 100] forget never DEM France COP ‘France will never forget this.’ Examples such as (27) and (28a–c) show that contextual clues or adverbials of time can disambiguate the meaning of the sentence. More generally, since y, y a and y en a are compatible with present, past and future interpretations of the sentences in which they occur, it would appear that a sensible conclusion is that none of the competing variants has been grammaticalized into a tense marker. The two moods attested, the indicative and the imperative are both

120 | Andrei A. Avram

expressed by unmarked verbs. Finally, only the active voice is represented in the available samples. Three types of copulas are frequently found, Ø, y a, and y en a, with various meanings (e.g. predicative, equative, existential, locative): (29a) Ça

rien nothing ‘This is nothing’ (29b) Et ça, Mme Dahomey and DEM woman Dahomey ‘And these are women from Dahomey’ (29c) Ici y en a crapile! 1919 jerk here COP ‘There are jerks here! (29d) mon case y a près de l’eau water 1SG.POSS house COP near ‘my house is near the river’

[Dupratz 1864: 394]

DEM

[Mandat-Grancey 1900: 134]

[Desjardins 1925: 20]

[Baratier 1912: 31]

Predicative possession is generally expressed by the locative copula y a or y en a, and by the reflex of French gagner ‘to win’: (30a) moi ya pantalon 1SG COP trousers ‘I have trousers’ (30b) moi y en a cochons les femmes 1881 pigs women 1SG COP ‘I have sows’ (30c) toi, ya gagné deux garçons two boys 2SG COP have ‘you have two boys’

[Marie-Victoria 1921: 488]

[Augouard 1905: 231]

[Marie-Victoria 1921: 488]

The negator is generally pas. According to Delafosse (1904: 265), “negation is simply expressed by the word ‘pas’ placed after the verb”. However, in all other sources, without exception, pas is always preposed, which presumably reflects already established usage. Consider the following selection of examples: (31a) y en a

connaître faire pitit! 1881 know make child ‘[they] do not know how to make children’ (31b) moi, pas mirer lui NEG see 3SG 1SG ‘I have [never] seen him’ COP

pas

[Augouard 1905: 94]

NEG

[Barret 1888: 324]

Kolonial-Deutsch and Français Tirailleur | 121

(31c) nous y a pas réclamé 1PL COP NEG complain ‘we did not complain’ (31d) Moi y a pas savoir COP NEG know 1SG ‘I did not know’

[Béchet 1889: 172]

[Baratier 1912: 31]

Only an extremely small number of place and time adverbs are found. With very few exceptions, no prepositions are used. In particular, locative and directional prepositions normally do not occur. As for conjunctions and complementizers, only quand ‘when’ and si ‘if’ are relatively frequently attested. Both Anon. a. (1916) and Anon. b (1918) include recommendations regarding word order. For Anon. a. (1916: 17), “to be able to be quickly understood by the Negroes, it is necessary […] to let one’s thought flow in the very simple mold of the primitive sentence, subject, verb, object”. Anon. b. (1918: 3) insists on “very simple sentences and connected […] in a logical manner”. Whether overtly expressed or rather an overtone, racist prejudice surfaces again, in confirmation of Fogarty’s (2008: 57) conclusions that such rules “clearly assumed the essential intellectual simplicity of West Africans, whose languages were correspondingly ‘primitive’”. The only word order pattern attested in the sources is SVO, including in interrogative sentences, which presumably rely on intonation. Clause coordination is almost exclusively achieved by parataxis, but the use of mais ‘but’ is relatively frequent: (32a) manger bon; dormer bon; s’amuser bon; travailler pas bon NEG good eat good sleep good have fun good work [Dupratz 1864: 397] ‘To eat is fine, to sleep is fine, to have fun is fine, to work is not fine’ (32b) y a croire ça, mais […] lui y a crevé [Béchet 1889: 86] COP believe DEM but he COP die ‘[he] believed in that, but he would have died’ The following are the most frequently attested types of subordinate clauses: (33a) direct object clause, no overt complementizer lui croire y en a moi 1899 [Colrat de Montrozier 1902: 206] 3SG believe COP 1SG ‘he will believe it was me’ (33b) conditional clause, no conjunction: Tirailleur « content cabinet », pas descendre quand même 1919 NEG go down however tirailleur want toilet [Desjardins 1925: 18]

122 | Andrei A. Avram

‘If a tirailleur wants to go to the toilet, [he] should not get off’ (33c) conditional clause: Si fusil y a chargé tirailleur y a besoin décharger [Anon. a. 1916: 22] If gun COP load tirailleur COP must unload ‘If the gun is loaded the tirailleur must unload it’ (33d) adverbial clause of reason: Moi voulait cochons les hommes, pasque moi y en a 1SG want pig men because 1SG COP cochons les femmes 1881 [Augouard 1905: 231] pig women ‘I want boars, because I have sows’ (33e) adverbial clause of time: Quand tirailleur monter wagon, fini descendre 1919 when tirailleur go up carriage finish go down ‘After the tirailleurs have got on the carriage, [they] are not allowed to get off’ [Desjardins 1925: 18]

5 Vocabulary 5.1 Kolonial-Deutsch Schwörer (1916: 44–49) discusses in detail the proposed vocabulary of KolonialDeutsch, which is supposed to meet both quantitative and qualitative criteria. Racist arguments are adduced in support of both types of criteria. Consider first the quantitative criteria. According to Schwörer (1916: 44), “it is possible to reduce the gigantic German vocabulary […] to the number required for African linguistic needs and manage with 500-600 words”. This drastic reduction is called for since the vocabulary of German “causes supplementary difficulties to the natives” (Schwörer 1916: 44). The selected vocabulary is divided into two groups: the first group (some 150 words) for beginners, “also for the clumsy”; the second one for advanced learners (Schwörer 1916: 48). Consider next the qualitative criteria. Essentially, the vocabulary of Kolonial-Deutsch should consist of “only those words […] which cannot be easily replaced by others” and of “only those which are useful […] first of all for the intercourse of the Whites with the Negroes, namely for their own colonial purposes” Schwörer (1916: 45). In the selection of lexical items “the interests of the German colonizers and their comprehension needs are [a] decisive [criterion]” (Schwörer 1916: 45, fn. 1). Recommendations are made regarding both the inclu-

Kolonial-Deutsch and Français Tirailleur | 123

sion and the exclusion of specific classes of lexical items. Thus, Schwörer (1916: 45) states that it is necessary “to select concepts which can be used as generally as possible and which therefore replace many others”, e.g. gut ‘good’, Sache ‘thing’, machen ‘to make, to do’. On the other hand, “abstract concepts (such as principle, morals, authority, science), which the Negro only rarely understands, are superfluous and are to be avoided” (Schwörer 1916: 47). The lexical items to be included also need to satisfy a formal criterion. According to Schwörer (1916: 47), it is necessary to avoid using phrases “with a complicated structure”, e.g. Rücksicht nehmen ‘to take into consideration’, sich Mühe geben ‘to do one’s best’, den Beweis antreten ‘to produce evidence’”. Finally, various strategies are suggested in order to compensate for the small size of the vocabulary. One such strategy is lexical polysemy: (34a) fest ‘solid; strong’ (34b) Junge ‘young man, child, son, servant’ (34c) zuverläßig ‘reliable; faithful’

[Schwörer 1916: 53] [Schwörer 1916: 50] [Schwörer 1916: 53]

The negator nicht is used to form antonyms: (35a) nicht recht for Unrecht NEG right ‘injustice’ (35b) nicht sauber for schmutzig NEG clean ‘dirty’ (35c) nicht schlafen for Schlaflosigkeit NEG sleep ‘sleeplessness’

[Schwörer 1916: 45, fn. 3]

[Schwörer 1916: 45, fn. 3]

[Schwörer 1916: 45, fn. 3]

Another strategy is the use of machen ‘to do, to make’ as a light verb: (36a) groß machen big make ‘to increase’ (36b) sauber machen clean make ‘to clean’ (36c) Spiel machen game make ‘to play’

[Schwörer 1916: 45, fn. 3]

[Schwörer 1916: 45, fn. 2]

[Schwörer 1916: 47, fn. 1]

Finally, circumlocutions are recommended as substitutes of German words not selected for inclusion in the vocabulary of Kolonial-Deutsch:

124 | Andrei A. Avram

(37a) de

alte Mann for Greis old man ‘old man’ Frau für waschen for Wäscherin woman for wash ‘washerwoman’ Sache für messen for Maßstab thing for measure ‘measuring instrument’ schnell gehen for laufen/rennen, springen quick go ‘to run; to jump’ Vater und Mutter for Eltern father and mother ‘parents’

[Schwörer 1916: 45, fn. 2]

DEF

(37b)

(37c)

(37d)

(37e)

[Schwörer 1916: 46, fn. 1]

[Schwörer 1916: 46, fn.1]

[Schwörer 1916: 45, fn. 2]

[Schwörer 1916: 45, fn. 2]

5.2 Français Tirailleur Racial prejudice also explains the various suggestions and comments regarding the size and nature of the vocabulary of Français Tirailleur. In a letter dated 27 April 1918 (quoted in Van den Avenne 2005: 134), General Mazillier expresses the widely held view that “the Sudanese dialects […] only have a very limited number of words which are sufficient for the natives”. Not surprisingly, such opinions are also directly reflected in proposals regarding the size of the vocabulary of Français Tirailleur as well. According to Anon. a. (1916: 17), for instance, it is necessary that “the number of words used be reduced as much as possible”, while Anon. b. (1918: 3) is more explicit, in suggesting that the number should be limited to “six hundred words”. Several specific criteria, also rooted in racial prejudice, must be met by the lexical items to be included. For Delafosse (1904: 264), “above all one must use only words expressing ideas that Negroes can understand”. In a more pragmatic vein, Anon. b. (1918: 3) writes that the words to be used “are the most common ones and occur constantly in the language of the indigenous tirailleurs”. Issues such as synonymy and homophony are also addressed. Thus, Anon. a. (1916: 17) states that it is necessary “to avoid expressing the same idea by several different words [because] this would confound the native”. As noted by Fogarty (2008: 157), this recommendation betrays a subtle racial prejudice, since it is “predicated upon the unspoken assumption that the soldiers’ limited minds would be unable to grasp variation in vocabulary”. Marceau (1911: 25) adds the exhorta-

Kolonial-Deutsch and Français Tirailleur | 125

tion “beware of homonyms”14, reiterated by Anon. a. (1916: 17), who writes that it is advisable “to avoid using words which are pronounced in the same way, to indicate different things”. In addition to the above general characterization of the make-up of the vocabulary, a number of strategies are attested and/or recommended, which all aim at compensating for the drastically reduce size of the vocabulary. Many words are polysemous, as shown in the following examples: (38a) faire manière ‘to try, to do one’s best; to use, to make use of’ [Anon. a. 1916: 21] (38b) gagner “the idea of acquiring, of receiving (more rarely), of taking, of catching (more rarely)” [Marceau 1911: 24] (38c) mirer ‘see; look at’ [Anon. a.: 1916: 19] (38d) moitié ‘middle, in the middle; half; centre’ [Anon. a. 1916: 22] A possible means of forming antonyms is the use of the negator pas: (39a) pas

obéir obey ‘to disobey’ (39b) pas vite NEG quick

[Anon. a. 1916: 19]

NEG

[Anon. a. 1916: 17]

A large number of forms attest to the extensive use of faire ‘to do, to make’ as a light verb: (40a) faire mauvais make bad ‘to commit a crime’ (40b) faire couvert make covered ‘to protect’ (40c) faire cabinet 1919 make toilet ‘to relieve oneself’

[Cantilly 1909: 245]

[Anon. a. 1916: 29]

[Desjardins 1925: 18]

Also used are circumlocutions, some of which are quite long and convoluted: (41a) y en a

manière way ‘omniscient’ COP

tout! all

[Marceau 1911: 32]

|| 14 In support of this recommendation, Marceau (1911: 25) reports that a French colonel having asked for sel [sεl] ‘salt was brought a selle [sεl] ‘saddle’.

126 | Andrei A. Avram

(41b) faire même make same

chose y a froid, même thing COP cold same

chose y a peur 1916 thing COP fear [Anon. a. 1916: 26]

‘to tremble’ (41c) si lui y a debout, si lui y a pausé, si lui y a genou, if 3SG COP kneel if 3SG COP standing if 3SG COP rest si lui y a couché [Anon. a. 1916: 28] if 3SG COP lie ‘in all positions’ Finally, note that in spite of specific recommendations that synonyms are to be avoided, examples of lexical synonymy do occasional occur: (42a) faire salué/faire présenter armes ‘to salute’ [Anon. a. 1916: 19] (42b) faire même chose (ou même manière) quand y en a genou kneel do same thing or same way when COP ‘do as when you kneel’ [Anon. a. 1916: 23] (42c) tenir, garder ‘to hold’ [Anon. a. 1916: 30]

6 Conclusions According to Mühlhäusler (1997: 260), “the only case where a Pidgin was deliberately invented as a means of social control is that of Pidgin German”. As shown in the present paper, however, it would seem that Français Tirailleur is also, at least in part, the outcome of deliberate language policies and of language planning. Kolonial-Deutsch and Français Tirailleur exhibit a number of similarities, particularly in their word order and the characteristics of the vocabulary. Many of these directly reflect stereotypes of the colonialist ideology of the time.15 The structures of German and French are simplified and reduced, given the firmly rooted belief of Europeans that the “natives” cannot possibly learn and use the “proper”, full-fledged versions of these languages. Particularly worth underscoring is what might be called the “perverse circularity” underlying the planning and evaluation of Français Tirailleur. This pidginized French, the outcome of what is at least in part deliberate restructuring premised on the aforementioned conviction, is regarded as the very confirmation of the alleged inabil-

|| 15 For Kolonial-Deutsch, see also Heiler (2010: 63–64), Orosz (2011), and Avram (2014); for Français Tirailleur, see also Houis (1984) and especially Fogarty (2008: 156–157 and 161).

Kolonial-Deutsch and Français Tirailleur | 127

ity of the “natives” to learn and use anything but a “primitive” language. Other characteristics shared by Kolonial-Deutsch and Français Tirailleur are reminiscent of features of artificial languages, e.g. the use of invariant forms or the use of the invariant negators nicht and pas respectively to create antonyms as in Esperanto, which makes use of the prefix mal- to derive such forms, cf. rapida ‘quick’ vs malrapida ‘slow’. Although they are planned or partly planned languages, both KolonialDeutsch and Français Tirailleur have been shown to display categorial multifunctionality, lexical polysemy, lexical synonymy. As in other pidgins, categorial multifunctionality is a consequence of the lack of inflections, in particular in Français Tirailleur, and of the small size of the vocabulary.16 The latter also accounts for the occurrence of lexical polysemy. The occurrence of lexical synonymy in Français Tirailleur, in spite of the small size of the vocabulary, is only apparently surprising. Indeed, other naturally formed pidgins with a similarly reduced vocabulary do include some pairs of synonyms and even entire synonymic series (from the lexifier language or from different source languages). More generally, the various instances of variation in Français Tirailleur confirm Corne’s (1999: 200) conclusion that “it is almost certainly the case that the language as used by the African NCOs and other ranks displayed considerably more variation than is evident in the manual17 itself”. As for Kolonial-Deutsch, since it has never been actually used, it is characterized by absence of variation, a feature generally typical of artificially constructed languages. In a well-known typology, Mühlhäusler (1997: 5–6) distinguishes the following developmental stages in the so-called “life cycle” of pidgins and creoles: jargon;18 stable pidgin; expanded pidgin;19 creole. Both Français Tirailleur and, to a lesser extent, Kolonial-Deutsch exhibit some of the morpho-syntactic and lexical features diagnostic of jargons, discussed and illustrated in Mühlhäusler (1997: 128–138). In addition, the absence of reduplication in Français Tirailleur confirms the fact that the occurrence of productive morphological reduplication correlates with the developmental stage of the variety at issue. As shown by Bakker (2003: 44) and Bakker & Parkvall (2005: 514, 519), reduplication in jargons and stable pidgins is unproductive. The absence of reduplication in Français Tirailleur is therefore consistent with its classification on independent grounds as a jargon.

|| 16 See Mühlhäusler (1997: 137). 17 The manual referred to is Anon a. (1916). 18 Also known as “minimal pidgin” or “pre-pidgin”. 19 Also known as “extended pidgin”.

128 | Andrei A. Avram

While both Kolonial-Deutsch and Français Tirailleur resemble other jargons, the former has been shown to be more complex in its morpho-syntax. This reflects Schwörer’s professed disdain for Pidgin English, on the one hand, and his insistence that his Kolonial-Deutsch should on no account be equated with Pidgin German, on the other. The more radical restructuring evinced by Français Tirailleur is hardly surprising given that this pidgin has actually been used. It therefore reflects not only planned simplification and reduction, but also the outcomes of such strategies in actual use.

Abbreviations COP DEF DEM INDEF NEG OBJ

copula definite article demonstrative indefinite article negator object

PL POSS PRS PST SG

plural possessive person past singular

References Anon. a. 1916. Le français tel que le parlent nos tirailleurs sénégalais. Paris: Imprimerie Militaire Universelle. Anon. b. 1918. Méthode d’enseignement du français tel que le parlent les sénégalais. Paris: Imprimerie Militaire Universelle. Augouard, Mgr. 1905. 28 années au Congo. Lettres de Mgr Augouard, tome I. Poitiers: Société Française d’Imprimerie de Librairie. Avram, Andrei A. 2014. Colonialist ideology and language planning: The case of KolonialDeutsch. Paper presented at “On Form and Pattern”, 29–31 May. Bucharest: Humboldt Kolleg. Aymonier, Étienne. 1890. La langue française et l’enseignement en Indo-Chine. Paris: Armand Colin. Bakker, Peter. 2003. The absence of reduplication in pidgins. In Silvia Kouwenberg (ed.), Twice as meaningful. Reduplication in Pidgins, Creoles and other contact languages, 37–46. London: Battlebridge. Bakker, Peter & Mikael Parkvall. 2005. Reduplication in Pidgins and Creoles. In Bernhard Hurch (ed.), Studies in reduplication, 511–532. Berlin/New York: Mouton de Gruyter. Baratier, Colonel. 1912. Épopées africaines. Paris: Fayard. Barret, Paul. 1888. L’Afrique Occidentale. La nature et l’homme noir, tome II. Paris: Challamel. Baumann, Adalbert. 1915. Wede, die Verständigungssprache der Zentralmächte und ihrer Freunde, die neue Welthilfssprache. Diessen vor München: Huber.

Kolonial-Deutsch and Français Tirailleur | 129

Baumann, Adalbert. 1916. Das neue, leichte Weltdeutsch für unsere Bundesgenossen und Freunde! Seine Notwendigkeit und seine wirtschaftliche Bedeutung von Prof. Dr. Adalbert Baumann. Vortrag, geh. 1915 . In laut-shrift geshriben!. Diessen vor München: Huber. Béchet, Eugène. 1889. Cinq ans au Soudan Français. Paris: Librairie Plon. Calvet, Louis-Jean. 1987. La guerre des langues et les politiques linguistiques. Paris: Payot. Cantilly, Louis de. 1909. En pleine brousse Malgache. Revue du Madagascar 11(1). 243–249. Chaudenson, Robert. 2003. La créolisation: théorie, applications, implications. Paris: L’Harmattan. Clayton, Anthony. 1994. Histoire de l’armée française en Afrique, 1830–1962. Paris: Albin Michel. Colrat de Montrozier, Raymond. 1902. Deux ans chez les anthropophages et les sultans du centre africain. Paris: Librairie Plon. Corne, Chris. 1999. From French to Creole. The development of new vernaculars in the French colonial world. London: University of Westminster Press. Cousturier, Lucie. 1920. Des inconnus chez moi. Paris: La Sirène. Delafosse, Maurice. 1904. Vocabulaires comparatifs de plus de 60 langues ou dialectes parlés à la Côte d’Ivoire et dans les régions limitrophes aves des notes linguistiques et ethnologiques, une bibliographie et une carte. Paris: Ernest Leroux. Desjardins, René. 1925. Avec les sénégalais par delà l’Euphrate. Paris: Calmann-Lévy. Dupratz, Père. 1864. Lettre du Père Dupratz. Annales de l’Œuvre pontificale de la Sainte Enfance 6. 393–401. Fogarty, Richard S. 2008. Race & war in France: Colonial subjects in the French Army, 1914– 1918. Baltimore: The Johns Hopkins University Press. Gerhardt, Ludwig. 1985. “Kolonial-Deutsch” und “Newspeak”: sprachpolitische Utopien zwischen Phantasie und Terror. Afrikanische Arbeitspapiere 1. 85–96. Gyasi, Kwaku A. 2012. “Le français petit nègre” and the construction of social identity in colonial and postcolonial Africa. International Journal of Humanities and Social Science 2(19). 224–231. Heiler, Timo. 2010. Laboratorien für die Moderne? Die deutschen Kolonien als Versuchsfeld einer urbanen und gesellschaftlichen Neugestaltung. Journal of New Frontiers in Spatial Concepts 2. 57–68. http://ejournal.uvka.de/spatialconcepts/archives/1151 (checked 05/10/2014). Houis, Maurice. 1984. Une variété idéologique du français: le langage tirailleur. Afrique et Langage 21. 5–17. Lehmil, Linda S. 2007. À l’école française: politique coloniale de la langue 1830–1944. PhD dissertation, Tulane University. Leymaire, Henri. 1889. La mort de Mabiala le féticheur. À travers le monde 4. 121–124. Mandat-Grancey, Edmond de. 1900. Au Congo (1898). Impressions d’un touriste. Paris: Librairie Plon. Manessy, Gabriel. 1984. Français-tirailleur et français d’Afrique. Cahiers de l’Institut Linguistique de Louvain 9(3/4). 113–126. Marceau, Le capitaine. 1911. Le tirailleur soudanais. Paris/Nancy: Berger-Levrault. Marie-Victoria, Sœur. 1921. Les petits sénégalais à Biskra. Les missions catholiques: Bulletin hebdomadaire de l’Œuvre de la propagation de la foi 53. 438. Mühleisen, Susanne. 2005. Emil Schwörers Kolonial Deutsch (1916). Sprachliche und historische Anmerkungen zu einem “geplanten Pidgin” im kolonialen Deutsch Südwest Afrika. Philologie im Netz 31. 30–48.

130 | Andrei A. Avram

Mühleisen, Susanne. 2009. Zwischen Sprachideologie und Sprachforschung. Kolonial Deutsch als Verkehrssprache für die Kolonien. In Ingo H. Warnke (ed.), Deutsche Sprache und Kolonialismus. Aspekte der nationalen Kommunikation 1884–1919, 97–118. Berlin/New York: De Gruyter. Mühlhäusler, Peter. 1997. Pidgin and Creole linguistics. Expanded and revised edition. London: University of Westminster Press. Orosz, Kenneth J. 2011. Colonialism and the simplification of language: Germany’s KolonialDeutsch experiment. In Michael Perraudin & Jürgen Zimmerer (eds.), German colonialism and national identity, 101–114. New York: Routledge. Perl, Matthias. 2001. Kolonial-Deutsch as restructured German. In Birgit Igla & Thomas Stolz (eds.), “Was ich noch sagen wollte…”. A multilingual festschrift für Norbert Boretzky on occasion of his 65th birthday, 237–247. Berlin: Akademie Verlag. Poulot, Brice. 2011. L’enseignement du français aux troupes coloniales en Afrique. Revue historique des armées 265. http://rha.revues.org/index7336.html (checked 04/10/13). Schwörer, Emil. 1916. Kolonial-Deutsch. Vorschläge einer künftigen deutschen Kolonialsprache in systematisch-grammatikalischer Darstellung und Begründung. Diessen vor München: Huber. Valdman, Albert. 1978. Le créole: structure, statut et origine. Paris: Éditions Klincksieck. Van den Avenne, Cécile. 2005. Bambara et français-tirailleur. Une analyse de la politique linguistique de l’armée coloniale française: la Grande Guerre et après. In Documents pour l’histoire du français langue étrangère ou seconde, SIHFLES 35, 123–150. Voeste, Anja. 2005. “Die Neger heben”? Die Sprachenfrage in Deutsch-Neuguinea (1884– 1914). In Elisabeth Berner, Manuela Böhm & Anja Voeste (eds.), Ein großs vnnd narhafft haffen. Festschrift für Joachim Gessinger, 163–174. Golm: Universität Potsdam, Institut für Germanistik. Werkmeister, Sven. 2010. El alemán como lengua mundial? Anotaciones históricas acerca de un proyecto fallido. Matices en lenguas extranjeras 4. 1–15. http://www.revistamatices.unal.edu.co (checked 05/23/14).

Janneke Diepeveen and Matthias Hüning

The status of Dutch in post-colonial Suriname Abstract: Dutch is an official language not only in the Netherlands and Belgium, but also in Suriname, a country in South-America. Before its independence, Suriname was a colony of the Netherlands, starting as early as 1667. After its independence in 1975, the multilingual Republic of Suriname maintained Dutch as its official language, the language of education and public life. In this paper, we shall address two seemingly conflicting developments which take place in this former Dutch colony: on the one hand, the growing use of the creole language Sranantongo as a lingua franca across Suriname and on the other hand, the persistence of Dutch. We shall argue that the linguistic developments in Suriname must be understood against the background of a young nation which is constructing its own post-colonial national identity. Keywords: Suriname, Dutch, Sranantongo, diglossia, standardization, postcolonialism

1 Introduction: Suriname and the Dutch language area1 Dutch is a West-Germanic language and the mother tongue of about 23 million people. It is an official language in six countries – two in Europe and four in South-America. In Europe, Dutch is an official language in the Netherlands and in Belgium, more specifically in the region of Flanders in the north of Belgium, but also in the capital of Brussels, where both Dutch and French are official languages. In the republic of Suriname in South-America, Dutch is the one and

|| 1 We want to thank Philipp Krämer for his valuable comments on an earlier version of this paper. || Janneke Diepeveen: FU Berlin, Institute of German and Dutch Languages and Literatures, Habelschwerdter Allee 45, 14195 Berlin, Germany. E-Mail: [email protected] & Matthias Hüning: FU Berlin, Institute of German and Dutch Languages and Literatures, Habelschwerdter Allee 45, 14195 Berlin, Germany. E-Mail: [email protected]

132 | Janneke Diepeveen and Matthias Hüning

only official language.2 It is the language for education, in public life and in official documents. Finally, Dutch is the official language for a group of islands in the Caribbean, the former Netherlands Antilles which include Aruba, Bonaire, Curaçao, Saba, St. Maarten and St. Eustatius (altogether ca. 256,000 inhabitants). The islands Aruba, Curaçao and St. Maarten are independent countries within the Kingdom of the Netherlands. Curaçao and St. Maarten became autonomous in 2010 whereas Bonaire, St. Eustatius and Saba (the ‘BES’-islands) became special communities within the country of the Netherlands. However, Dutch is not their only official language. Since 2007, Aruba and Curaçao have Dutch and Papiamento (Papiamentu) as their official languages. The latter is an Iberian creole and it is the dominant language for daily speech, the media and primary education. St. Maarten has Dutch and English as its official languages and the latter is used for daily communication. For the BES-islands there is a special language policy: Dutch is the official language, but the use of English and Papiamentu is officially recognized in certain domains of education and in the communication with government offices.3 Whereas the Caribbean islands are officially part of the Kingdom of the Netherlands, the Republic of Suriname has been fully independent since 1975. Autonomy was prepared in 1954, when both Suriname and the Netherlands Antilles attained a certain degree of self-determination. Before that, they were Dutch colonies. The Dutch established several settlements and colonies in the 17th century: in the north of South-America, in South-East Asia and in the South of Africa. We shall briefly introduce Suriname, its history and its ethno-linguistic characteristics in section 2. In sections 3, 4 and 5 of this paper we concentrate on the linguistic situation of post-colonial Suriname. Even though Dutch is the official language, a special situation exists where Sranantongo, an English-based creole, finds widespread use as the lingua franca. This situation seems stable, but it is controversial. We provide evidence that Dutch has a strong position in post-colonial Suriname and we investigate why this may be the case, giving an explanation on the basis of eight factors. Our discussion is based on recent literature and on data collected primarily through personal communication and interviews with various people in

|| 2 By the time we received the proofs of this paper there has been a surprising new development. On 15 February 2015 Suriname’s news site Waterkant.net announced that the Minister of Education, Ashwin Adhin, is planning to recognize at least 20 languages spoken in Suriname as official languages. We cannot foresee the consequences of this decision. 3 For information on these regulations, we refer to the following website: http://www. rijksoverheid.nl/onderwerpen/erkende-talen/documenten-en-publicaties/brochures/2012/04/ 16/taalregelingen-voor-de-bes-eilanden-in-de-nederlandse-wet-en-regelgeving.html.

The status of Dutch in post-colonial Suriname | 133

and from Suriname.4 This exploratory investigation is part of the project Dutch++. Examples and new models for learning and teaching pluricentric languages. In the e-learning project Dutch++ Surinamese Dutch is presented as one of the national varieties of Dutch.5 In the present paper we shall argue that the linguistic situation of Suriname, particularly the status of Dutch vis-à-vis Sranantongo, must be understood against the background of a young nation in the midst of constructing its own post-colonial identity.

2 Multi-ethnic and multilingual Suriname Let us briefly introduce Suriname before we move on to discuss its ethnic composition (2.1), its linguistic landscape (2.2) and multilingual practices (2.3). The Republic of Suriname is situated in South-America, to the north of Brazil, and between Guyana and French Guiana. Although the country has four times the size of the Netherlands, it has a relatively small population of only half a million people. Most of them live in the capital of Paramaribo. Nearly 80 % of Suriname’s surface is covered by tropical forests. Suriname has plenty of natural resources including wood, gold and bauxite. Nevertheless a large portion of its population lives in poverty. Linguistically, Suriname has an isolated position in South-America with Dutch as its official language. Dutch is the only language for education, for the government, official documents and regulations, for science, etc. Suriname is the only former Dutch colony in which the Dutch language is still prominent (Bies 1997: 9).

|| 4 We owe many thanks to Henning Radke, a former student of the Freie Universität Berlin who prepared his MA thesis in Suriname and helped us establish local contacts and collect additional data. 5 Dutch++ is a multilateral project which is funded by the Lifelong Learning Programme of the European Commission (Grant Agreement number: 2011-4037/001-001). Project coordinator is the University of Vienna (Austria). Further partners are Freie Universität Berlin (Germany), Tilburg University (Netherlands) and Thomas More (Belgium). Dutch++ website: http://dutch plusplus.ned.univie.ac.at/; website of the Lifelong Learning Programme: http://eacea.ec.europa.eu/llp/about_llp/about_llp_en.php.

134 | Janneke Diepeveen and Matthias Hüning

2.1 Multi-ethnic society Suriname is a multi-ethnic society. This is due to its history of colonialism with long-lasting slavery and indentured contract laborers. Carlin & Arends (2002) and Borges (2014) give an overview of the complex origins of Suriname’s ethnic and ethno-linguistic composition with references for further reading. We will limit ourselves to a short summary. From 1600 onwards, Europeans colonized the coast of the Guianas in the north of South-America. They set up commercial centers and plantations. Initially the Caribbean region was particularly important for slave trade. In 1651 the English started sugar plantations with experienced planters and African slaves. They laid the foundation for large scale sugar cultivation. The Dutch acquired Suriname in 1667 through an arrangement with the English, who took over New Amsterdam. From 1667 onwards, Suriname was a Dutch colony and would remain so until its full autonomy in 1975.6 In early colonial times this stretch of land was called Dutch-Guiana and it was originally inhabited by Amerindian civilizations. Today there are about 12,000 Amerindians left, and they live in the forest, striving to keep their culture intact. Suriname was a plantation colony which produced for instance sugar, coffee and cotton. As the plantation colony grew and the Amerindian slaves could no longer meet labor demands, the colonists brought in thousands of slaves from the west coast of Africa. In the early 18th century many slaves managed to escape from the plantations into the forest and established their own free societies there. As in many other (former) colonies, the community of escaped slaves was known as Maroons. In Suriname they are also referred to as ‘bush Creoles’. After the abolition of slavery in 1863 the Dutch were in need of new people to work the plantations. This was the start of a period of contract laborers from Asian countries. First of all Chinese workers immigrated from Hong Kong. From 1873 onwards, there were Indian immigrants (Hindustani) from the colony of British-India, followed by Javanese immigrants from the colony of Dutch-India (Indonesia) in 1890. Many Asian immigrants stayed in Suriname. Today the Hindustani even make up the largest ethnic group in Suriname. The second largest group are the Creoles. This term refers to persons of African descent who may often show some admixture with other ethnic groups. In addition to the groups discussed in this section Suriname still attracts new immigrants. Immigrants from Brazil are particularly involved in artisanal

|| 6 The English briefly reoccupied Suriname from 1799 to 1816.

The status of Dutch in post-colonial Suriname | 135

gold mining (Borges 2014: 29), whereas many Chinese immigrants choose the Paramaribo area to work as shopkeepers.

2.2 Multilingual society Suriname is a multilingual society. First of all there is multilingualism on a national level: over twenty languages are spoken in Suriname. According to their origin, they can be classified into four main groups: Amerindian languages, European languages, creoles and Asian languages. For more information on each of these groups we refer to Carlin & Arends (2002). a) The Amerindian languages are the original languages of Suriname, i.e., they are spoken by the indigenous people. They can be linguistically divided in Arawakan languages (e.g., Arawak), Carib languages (e.g., Carib, Trió) and Warao languages (e.g., Warao). b) European languages were imported by the colonists, including the English and the Dutch. Today, Dutch is the most important European language: it is the official language; many people have adopted it as their main language and it is associated with upward social mobility. c) However, English becomes more important, particularly since Suriname joined the CARICOM (Caribbean Community and Common Market), an originally English-speaking organization of Caribbean communities, in 1995. Not only does Suriname have a commercial relation with the Caribbean, but it is also culturally much more related to the Caribbean than to its geographically closer South-American neighbors. In addition, people in Suriname are regularly confronted with English in their everyday life. For instance, they are used to watching American TV-programs and films without subtitling (in the Netherlands and Flanders all English-spoken films and programs are subtitled). d) Other European languages in Suriname include Portuguese, which played a significant role in the 17th century as many sugar estates were owned by prosperous Portuguese Jews. The Sephardic Jews kept using Portuguese until far into the 18th century (Borges 2014: 17). Today, Brazilian Portuguese plays a considerable role in Suriname due to a recent immigration wave from Brazil (see e.g. Romero 2008). e) Creoles originated in the contact of European colonial languages with those of (descendants of) African slaves who worked at the plantations. The slaves brought along their own West-African languages while the European settlers usually spoke various varieties of their home regions. At the plantations in Suriname, the slaves had limited access to the European languages but nev-

136 | Janneke Diepeveen and Matthias Hüning

ertheless needed to communicate with their masters, who were (initially) often English-speaking and with their foremen. In this context, various Surinamese creoles were created. Sranantongo or Sranan, formerly called NegroEnglish (Negerengels), Taki-taki, or Nengre, is an English-based creole with influences by West-African languages, Portuguese and Dutch. The Dutch influences were only added at a later stage. Sranantongo became the mother tongue of the slaves who passed it on to their children. In other words, Sranantongo is historically associated with the Creole population. Today it serves as the main language for interethnic communication in Suriname. f) Other Surinamese creoles include Saramaccan (Saamaka), Aukaans (Ndyuka), Paramaccan (Paamaka) and Kwinti. They originated from the languages used by the Maroons, the slaves who had left the plantations. These creoles are mostly used for communication within Maroon communities. g) The Asian contract laborers imported various Asian languages into Suriname. In addition, Surinamese varieties of these Asian languages came into existence. The Hindustani brought several Indic languages, which over time were influenced by Sranantongo and by Dutch. This gave rise to Sarnami Hindustani, a Surinamese variety which is not like any present Indic language found in India (Borges 2014: 27). Javanese laborers brought along their Indonesian mother tongue, which was over time influenced by Sranantongo, Dutch and Sarnami Hindustani. This gave rise to a Surinamese variety of Javanese called Surinamese Javanese. The Chinese contract laborers brought different Chinese varieties to Suriname, particularly Hakka or Keija (a southern Chinese language). Today, following a new wave of Chinese immigrants, mainly Mandarin and Cantonese are used in Suriname (Borges 2014: 25). Chinese, Sarnami and Surinamese Javanese are predominantly used for communication within the respective ethnic group.

2.3 Individual multilingualism Besides societal multilingualism, there is individual multilingualism in Suriname: individuals use two or three languages regularly. Each ethnic group in Suriname has its own language for communication within the own group. For further communicative situations, people in Suriname use Dutch or Sranantongo. Thus, a child may use Sarnami Hindustani with his relatives at home, Dutch in the classroom and Sranantongo with his friends. Code-switching is common, both inter- and intrasententially: people may switch to another language within the same conversation (intersentential code-switching) but even within one sentence (intrasentential code-switching) (Kleine 2002: 218). This

The status of Dutch in post-colonial Suriname | 137

phenomenon can also be observed on the internet. For a study on codeswitching in Suriname with a special focus on Dutch and Sranantongo, see Radke (2013). The preceding observations on individual multilingualism in Suriname are in line with current sociolinguistic research into social networks and communities of practice (see for instance Wardhaugh 2010 for references). While the older notion of speech community was too restrictive, these new concepts offer an adequate way to explain linguistic practice in Suriname. Thus, each language user participates in various social networks or communities of practice (e.g., home, the classroom, a club), each of which is associated with one or more languages. When participating in a certain community of practice, speakers foreground a certain social identity, which is marked by the use of a certain language. When moving on to the next occasion, they foreground a different identity and switch to another language. The linguistic practices mentioned above give rise to various types of contact phenomena between the languages of Suriname. We refer to Borges (2014) for an extensive study. In the following sections we restrict our attention to Dutch and Sranantongo.

3 Dutch and Sranantongo in Suriname We will now focus on the two languages that play a central role in post-colonial Suriname: the European language Dutch (3.1) and the Surinamese creole Sranantongo (3.2). We provide some information on their history in Suriname and we briefly introduce their present-day status. For more extensive information on the history and status of Dutch in Suriname, with references for further reading, see for instance Gobardhan-Rambocus (1997), Kleine (2002) and (2013). Suriname’s so-called ‘linguistic schizophrenia’ (3.3.) will be investigated more closely in the remainder of the paper.

3.1 Dutch Dutch has been the only official language in Suriname since its territory was colonized by the Dutch in 1667. During early colonial times, Dutch held the prestige position in Suriname, although the Dutch colonial government never propagated it (Gobardhan-Rambocus 1997: 246). In the beginning of Dutch rule, use of the language was limited to a small group consisting particularly of

138 | Janneke Diepeveen and Matthias Hüning

planters who did not allow the ‘lower classes’, particularly black slaves, to use it, let alone learn it. Although the Dutch planters wanted their language to be exclusive, they could not stop it from spreading. A generation of children was born as a result of relationships of Dutch planters with black female slaves which would constitute the core of Suriname’s Dutch-speaking population (Gobardhan-Rambocus 1997: 227). They would give rise to a new class of selfdeclared free people who tried to distinguish themselves from the slaves. The Dutch language was one means by which to do this. From 1876 onwards school attendance was made compulsory in Suriname for all children from 7 to 12 years old. This was the starting point of a language policy by the Dutch colonial government (Gobardhan-Rambocus 1997: 246). It was determined by law that the language to be used in education was Dutch. The Dutch authorities strongly discouraged the use of Sranantongo, which they considered the language of the slaves. It was forbidden to speak Sranantongo at school and if children did, they were told to go and wash their mouths. For some time there were exceptions for children of Asian immigrants, but by 1910 they too were educated only in Dutch (Eersel 1997: 216). This was a clear sign from the Dutch colonial government that they wished to turn Suriname into a Dutch-speaking colony, although they did not intend to replace Sranantongo entirely (Gobardhan-Rambocus 1997: 229, 2008: 193). Schools for ‘free’ black children and children born of mixed couples were established in 1760; children of slaves were not educated until 1844 (Gobardhan-Rambocus 1997: 228). Until the first half of the 20th century the Surinamese society underwent a strong assimilation to Dutch (christian) culture or ‘dutchification’ (GobardhanRambocus 1997: 232–233). Important factors enabling this degree of assimilation were education, missionary work, mass media and the fact that the Surinamese society was organized according to the Dutch model. Another factor were activities by the Sticusa (foundation for cultural cooperation), an influential organization strengthening cultural ties between the Netherlands, Suriname and the Netherlands Antilles. Assimilation to Dutch culture and society was not equally strong for all ethnic and social groups in Suriname. It was particularly the elite which was ‘dutchified’ and this was mostly the case for the Creole population. It is important to note that during the process of assimilation there have been no direct attempts to wipe out the other languages spoken in Suriname (compare Gobardhan-Rambocus 1997: 246). Commissions and institutions concerned with spreading Dutch did not aim at replacing all other languages by it. After the Second World War there was a growing self-confidence of Suriname (Gobardhan-Rambocus 1997: 231). This resulted in a semi-autonomous status within the Kingdom of the Netherlands in 1954 in which Suriname was still financed by the Netherlands but for the most part made its own decisions.

The status of Dutch in post-colonial Suriname | 139

During this period the Dutch language maintained its prestige position. This was partly due to the Surinamese elite of Dutch-speaking Creoles in the Paramaribo area. This prominent group, consisting of public servants, teachers and so on, chose Dutch as the language for progress in society (GobardhanRambocus 1997: 232). The Creole community was quite self-confident and demanded full independence from the Netherlands. Finally, in 1975, the fully autonomous Republic of Suriname was installed with President Johan Ferrier. Due to the burden of colonial history, the relation between Suriname and the Netherlands has always been of a complicated nature and it remains tense today. In a way this also holds for their cooperation with respect to the Dutch language. In 2003/2004, the position of Dutch in Suriname was officially confirmed as Suriname became an associate member of the Dutch Language Union. This is an intergovernmental organization in which the Netherlands and Flanders (the Dutch-speaking region of Belgium) collaborate with respect to the Dutch language in a very wide sense, including literature, spelling and education.7 Critics of the accession of Suriname to the Dutch Language Union have uttered that Suriname has a European-oriented language policy rather than focusing on itself (Brandon et al. 2007 and Kroon & Kurvers 2009). Other critics point out that from a Surinamese perspective, the position of the members of the Dutch Language Union is not equal: Suriname is an associate member only. It does not have its own financial budget, hence it cannot start its own projects or initiatives but it is merely asked to cooperate in activities and research projects started by the Netherlands and Flanders. Due to this inequality, the Dutch Language Union is perceived by certain critics in Suriname as a neocolonial organization. They are convinced that it is a strategy of the Dutch to keep controlling Suriname. Their view may be strengthened indirectly by the current government of Suriname, which does not actively support the Dutch Language Union (Helen Chang, p.c.).8 Today, Dutch is the prestige language used for all official and formal situations: written documents, government offices, court proceedings, and also in the domains of work, media and education. Dutch is the language of instruction and it is taught in schools. Dutch is also the biggest language in Suriname in terms of the number of speakers. 200,000 consider it their first language and another 200,000 their second. It should be stressed that there are immense differences concerning the use of Dutch in the capital city of Paramaribo and in

|| 7 In the period of 2007–2011, the Dutch Caribbean joined the Dutch Language Union. 8 Helen Chang is a Surinamese staff member of the Dutch Language Union. She represents the Dutch Language Union in Suriname.

140 | Janneke Diepeveen and Matthias Hüning

the remote districts in the interior parts of Suriname. In the latter areas, Dutch mostly is a foreign language (L2 or L3) and children usually first encounter Dutch at school (see 3.2.). By contrast, in the Paramaribo area, Dutch is more and more frequent as a mother tongue. School boys and school girls use Dutch more than adults (see Asin et al. 2004). People are more and more willing to pass the Dutch language on to their children. Many parents who themselves first encountered Dutch at school, now make sure their own children are already confronted with Dutch at home. In the Paramaribo area Dutch is commonly used in informal circumstances. This is shown in research, e.g., Fierens (2006), who interviewed young Surinamese in the capital and found that they consider it normal to use Dutch in informal situations like the cafeteria.

3.2 Sranantongo Sranantongo is an English-based creole which is, historically at least, the first language for the Creole population in Suriname. The history of this language probably starts in 1651, when the English established sugar plantations on the territory of Suriname. After the Dutch had occupied the colony in 1668, the English linguistic base was still maintained (Essed 1983: 47). From then on, Dutch vocabulary was integrated as well. From 1844 onwards, children of slaves were educated in Sranantongo, which by that time was already an important lingua franca (Eersel 1997: 213). As explained in 3.1., Dutch became the official language of education in 1876. Nevertheless, Sranantongo remained an important lingua franca and today, it has even become the lingua franca of Suriname which is used for interethnic daily communication. People with different backgrounds, who speak different languages at home, typically use it for their daily communication and new immigrants use it, for instance new Chinese immigrants working as shopkeepers. To them, Sranantongo is easier to learn and use than Dutch (see Paul Middellijn in Romero 2008). Sranantongo is particularly strong as a lingua franca in the interior parts of Suriname. People there do understand Dutch but they use other languages at home, for instance Maroon languages. As mentioned in 3.1, only a quarter of families in the interior parts of Suriname uses Dutch at home (Van Maele 2013). Children from different ethnic backgrounds tend to communicate in Sranantongo, which they often learn on the street. Most of them first encounter Dutch at primary school, where it is the language of instruction. For these children Dutch is a foreign language which may be difficult to learn, considering that is linguistically remote both from their L1 and from the English-based creole Sranantongo they use as a lingua franca. In this respect, Suriname differs from

The status of Dutch in post-colonial Suriname | 141

other creole-speaking countries in which the language of instruction has served as the base language of the creole (e.g., French and a French-based creole) and may thus be perceived as more familiar. Sranantongo is not officially allowed as a language for education and it is not even taught as a school subject. The same educational methods are used and the same goals are to be attained all over Suriname. There are no different methods in the city and in the interior parts of Suriname. In the end, all pupils have to pass the same tests, whether Dutch is their native language or not and no matter how much or how little support they get at home. This is a wellknown problem. In the interior parts of Suriname half of the children have to repeat the first year at primary school, many children skip school regularly or they do not go there at all. Most pupils have to go to secondary school in the Paramaribo area. When pupils with different backgrounds meet here, the ones coming from the interior parts of Suriname are clearly disadvantaged. For all of these reasons it has been suggested that the native languages or at least Sranantongo should be officially established as an ‘auxiliary language’ at school (Van Maele 2013). Teachers in primary school already use Sranantongo from time to time, but there is no official language policy to back them up. Sranantongo is mainly an oral language: it is uncommon for people to write it down. As a spoken language, Sranantongo has particular social functions in Surinamese society. People switch to Sranantongo when they express emotion, when they want to tell a story, gossip, discuss something or tell jokes. Thus, at the workplace, colleagues may go through the work schedule of the day in Dutch language and switch to Sranantongo when they want to add something personal or emotional. Fierens (2006) found in her survey that 45 % of young Surinamese (of the age of 15–16) find a joke in Dutch less funny than a joke in Sranantongo. In the documentary Ver weg of dichtbij (Far away or nearby) produced by the Dutch Language Union a boy explains that he and his schoolmates talk about ‘boring things’ in Dutch, and that they use Sranantongo when things get exciting (Visser 2005). Sranantongo is perceived as short, snappy and to the point. No wonder it is popular in rap music and with young people. For the same reason, Sranantongo is used in public campaigns, in both oral and written form (slogans), for instance against smoking or against domestic violence. People’s associations with Sranantongo are ambiguous. There is a social stigma attached to it; it is often considered to be vulgar. For example, Surinamese men will not speak Sranantongo when they try to seduce women (Farid Khodabaks, p.c.). At the same time, Surinamese tend to feel a strong sense of national identification with Sranantongo. In addition, Sranantongo is used in political campaigns: president Bouterse introduced Sranantongo phrases in his

142 | Janneke Diepeveen and Matthias Hüning

political speeches in the 1980s to reach the people. This contributed to an association of the use of Sranantongo with nationalist politics (Romero 2008).

3.3 Linguistic schizophrenia? On the basis of 3.1 and 3.2 we may conclude that we are dealing with two seemingly conflicting developments taking place in the former Dutch colony. Carew (1982) refers to this situation as Suriname’s “linguistic schizophrenia”. Sranantongo is strongly present in everyday life and in the minds of people, it is the second largest language in Suriname in terms of the number of speakers and it is steadily growing as a lingua franca across Suriname. Nevertheless, there is no official recognition of Sranantongo. On the other hand, Dutch is maintained as the official language, the language of education and public life in Suriname. In addition, there is a growing willingness, particularly in the Paramaribo area, to speak Dutch and of parents to pass it on to their children.

4 Explaining Suriname’s linguistic situation How can we explain the strong position of Dutch (in terms of social prestige, actual language use and official recognition) in post-colonial, multilingual Suriname – in comparison with Sranantongo as the country’s lingua franca? In this section we identify eight possible explanatory factors. We start with objective factors and end with two central attitudinal factors. We compare the position of Dutch with the situation of Sranantongo.

4.1 Dutch grants access to further education and better jobs Dutch holds the prestige position in Surinamese society. It is the language one has to learn in order to function well in society (Gobardhan-Rambocus 1997: 246). In Suriname, a good command of Dutch is perceived as a ticket to professional success. People realize that this starts with higher education and this, like primary and secondary education in Suriname, takes place entirely in Dutch. Dutch is the main language at Anton de Kom university of Suriname, in other institutions and of academics in general. Higher education in Suriname has improved greatly in recent years and new study programs are being developed in the Paramaribo area. More and more people are willing to learn, they

The status of Dutch in post-colonial Suriname | 143

aim for a better job and choose university education as a way to achieve that. The movement in favor of Dutch in Suriname (for instance, people’s willingness to learn Dutch and pass it on to their children) can be directly related to the progress in education (Helen Chang, p.c.). Dutch is required in most job vacancies. Good jobs are often situated in some kind of government office. Whether you want to work in a government office, in the media or in a private company, you are expected to have an excellent command of Dutch (see for instance Van Maele 2013). People in the districts, too, are convinced that a good command of Dutch is necessary to get by in the Surinamese society. They themselves feel they would be completely isolated without Dutch; it is their bridge to the city, to Paramaribo (Van Maele 2011). For this reason they will not give up on Dutch and they do not think highly of organizing primary school education only in the native languages.

4.2 Dutch constitutes a bridge to the Netherlands For the people in Suriname, Dutch constitutes a bridge to the Netherlands in several ways. There is an important personal motivation for people in Suriname to have a thorough command of the Dutch language. Due to frequent migration, just about every individual living in Suriname has at least one relative in the Netherlands (Essed-Fruin & Gobardhan-Rambocus 1992: 12). After Suriname’s independence in 1975, its society faced great changes. In 1980, Desi Bouterse initiated a military dictatorship in which human rights were damaged and drug crime or murder were no exception. The sad climax was the event now known as the ‘December Murders’ in 1982. Many Surinamese migrated to the Netherlands to escape Bouterse’s military regime, to leave a society which they felt was hopeless, or to follow university education in the Netherlands. The final outcome of this wave of migration is that about 345,000 people of Surinamese descent are living in the Netherlands today (2010, Dutch Central Bureau of Statistics; see Borges 2014: 28).9 A thorough command of Dutch enables people in Suriname to be in touch with their relatives and the social world of these rela-

|| 9 Within the scope of this paper we cannot go into detail but it may be interesting to add that there have been some recent political changes in Suriname. There was a return to democracy by the end of the eighties. In 1991, Ronald Venetiaan was president; he was elected again in 2000 and 2005. In 2010, the former military dictator Desi Bouterse was chosen for president, which was quite controversial from the point of view of the Netherlands. By many people in Suriname, however, Bouterse is seen as a politician who is close to the people.

144 | Janneke Diepeveen and Matthias Hüning

tives. For financial reasons, the Netherlands is often the only destination where Surinamese can afford to spend their holiday (Essed-Fruin & GobardhanRambocus 1992: 18). Having contacts with the Netherlands is not only necessary for private reasons but also for professional reasons. A good command of Dutch allows Surinamese to complete university studies in the Netherlands. People in Suriname may aim for better job opportunities in the Netherlands, and an excellent command of Dutch is essential to pursue these goals.

4.3 Dutch is internationally and politically important On the one hand, one may argue that Suriname is rather isolated as a Dutchspeaking country within South-America. However, for Suriname, maintaining Dutch is actually a way to be less isolated. Dutch does not only provide a bridge to relatives in the Netherlands, but in a more global perspective, it gives the Surinamese access to the world arena. After all, Dutch is spoken outside Suriname, first of all nearby, in the Dutch Caribbean, and secondly – and most importantly – overseas, in the Netherlands and Belgium. Compared to other European countries, the Low Countries are small but they are well experienced in the world arena and they have plenty of connections in world trade. There are indications (cf. Carew 1982) that Surinamese consider Dutch an internationally and politically important language. To them, Dutch is a window to the world that for instance Sranantongo can never be, since it is only spoken in Suriname and does not seem at all relevant outside the territory of Suriname. However, the choice of Dutch may still be considered remarkable compared to other European languages which are internationally and politically far more significant, such as English, Spanish and Portuguese. These languages are not only relevant for Suriname’s relations with its neighbors but they are particularly important in the areas of business and trade. Indeed, it has repeatedly been proposed that English should be an, or even the only, official language of Suriname. Various politicians in Suriname, but also Surinamese linguists (e.g., Hein Eersel) and writers (e.g., Paul Middellijn) defended the opinion that the official language of Dutch should be replaced by English (see e.g. Bies 1997, Brandon et al. 2007: 1, Romero 2008). The motivations to introduce English as the official language of Suriname are partly of a commercial nature, particularly to facilitate Suriname’s participation in the international organization CARICOM but also more generally, to strengthen Suriname’s links to the Caribbean and North America.

The status of Dutch in post-colonial Suriname | 145

4.4 Literary tradition in Dutch Dutch is not only politically and economically important but also culturally. Through Dutch, Suriname has access to a wide literary tradition, including literary work by Dutch and Flemish authors. With authors like Clark Accord, Albert Helman, Cynthia McLeod, Ismene Krishnadath, and several others, Suriname itself can boast a literary tradition in the Dutch language, also respected in the Low Countries. The literary and cultural role of Sranantongo is somewhat different. Audience and readership are mostly limited to Suriname. It should be added that Sranantongo is mainly an oral language, the effect of which is strongly felt by people in Suriname. An excellent example is Suriname’s national anthem. In 1960 there was an official recognition of a strophe in the national anthem in Sranantongo, which used to be entirely in Dutch. The famous author Cynthia McLeod notes that the Sranantongo part is the only strophe that people in Suriname actually sing. This illustrates a certain feeling of ‘national’ identification attached to Sranantongo.

4.5 Dutch is standardized and taught in institutions Since 1876, people in Suriname have been and continue to be taught full literacy in Dutch. For children, learning to express themselves in Dutch is the central goal of language teaching in primary school (Nationaal leerplan voor het basisonderwijs in Suriname, by MINOV 2004; see Asin et al. 2004). However, it should be noted that primary education takes place entirely in Dutch: not only do pupils with different mother tongues have to learn the Dutch language, they also have to learn other school subjects like mathematics and history in this foreign language (Van Maele 2013). This is difficult and teachers sometimes have to use Sranantongo as a final resort when they notice that pupils are unable to understand the subject matter. Although Suriname today is looking at much greater tolerance towards Sranantongo than in previous generations, Sranantongo is still not taught at school. Since it is no school subject, people do not learn to write it and are not familiar with its official orthography. There are two recent official orthographies in use: Sranantongo has had an official orthography since 1960 which was revised in 1986 (Sebba 2000: 927). In the domain of edited publications, there are attempts to follow the 1986 norms, but outside this domain (e.g., names, slogans, advertisements), there is a range of orthographic practices (Sebba 2000:

146 | Janneke Diepeveen and Matthias Hüning

927–928). Written Sranantongo is only found in seldom occurrences outside any ‘official’ domain where two norms may even be used alongside each other. Thus, for many speakers of Sranantongo, writing the language down is an ad hoc process (Sebba 2000: 929). A similar lack of standardization holds for Sarnami Hindustani and Suriname Javanese, although committees for their standardization do exist (particularly for establishing a uniform spelling). As long as there is a lack of standardization, there is little chance for these languages to become school subjects in Suriname (and vice versa: if a language becomes a school subject, this may contribute to its standardization). Children in Suriname are educated in Dutch at primary and secondary school using Dutch schoolbooks and literature, they learn Dutch spelling, consult Dutch dictionaries and grammar books and skim newspapers in Dutch language. ‘Dutch’ is synonymous to ‘Netherlandic’ in some respects, e.g., Suriname uses many schoolbooks published in the Netherlands which are not adapted to Surinamese society. With respect to standard Dutch, Suriname is in an ambiguous position. On the one hand, education builds upon official Dutch spelling and grammar regulations laid out in the Low Countries. As long as this is the case, teachers in Suriname will mostly refer to the Netherlandic Dutch standard. On the other hand, education has to come to terms with Suriname’s own variety of Dutch, which is influenced by the other languages spoken in Suriname (see 4.8).

4.6 Maintaining Dutch seems the most economical There is a very practical reason for Suriname to retain Dutch as the main language in society: Suriname cannot afford the costs of introducing other official language(s). Introducing a different official language for the country would establish a need for new schoolbooks, new street signs, translations of official documents, etc. The costs would be too high for Suriname to bear. Retaining Dutch as the official language has the advantage of no extra costs (in addition to the current existing costs for educational material).10 It should be noted, however, that Suriname faces high (invisible) costs because many children are educat-

|| 10 Suriname gets some financial support for Dutch-language educational material, for instance from the Belgian non-profit organisation VVOB (Flemish Association for Development Cooperation and Technical Assistance) (Helen Chang, p.c.).

The status of Dutch in post-colonial Suriname | 147

ed in Dutch instead of their L1: there is a considerable educational inequality and social disparity. From the point of view of the population, there is no real concern for language regulations. Many Surinamese live in poverty. Estimations on how many Surinamese live below the national poverty line differ but the highest figure is as high as 70 %.11 The first concern of these people is obviously a very basic one: ‘surviving’. A large number of Surinamese are too busy struggling to survive to actually worry about language policy.

4.7 Dutch has a unifying effect since it is not tied to a particular ethnic group Besides objective and practical factors there are highly important attitudinal factors we need to refer to when we attempt to understand why Dutch as the ‘colonial language’ can persist in post-colonial Suriname. Dutch is perceived as everybody’s language and as such it is not tied to a particular ethnic group (Bies 1997: 14, Essed-Fruin & Gobardhan-Rambocus 1992: 18, and see 4.8.). This makes it attractive in the multi-ethnic society. It could be argued that Sranantongo too is a unifying language since it is used by different groups for interethnic communication. Sranantongo is also the lingua franca learnt by new immigrants. However, contrary to Dutch, Sranantongo is not entirely free of any associations with a particular ethnic group. Sranantongo may be associated by certain groups with the Creole population in and near Paramaribo; this association may particularly be felt by the Hindustani community (EssedFruin & Gobardhan-Rambocus 1992: 18) and by the Maroons (Romero 2008). Part of the Creole population itself considers Sranantongo as the language of their ethnic group, since it is their native language and an exponent of Creole culture (Gobardhan-Rambocus 2001: 510). We should repeat in this context that there are also some associations between Sranantongo and nationalist politics. Notwithstanding such associations, various ethnic groups across Suriname freely use Sranantongo as their lingua franca. Another language which could have a unifying effect is English. As indicated in 4.3, English has been proposed by some as a possible official language for

|| 11 Figure for 2002; see “Suriname moet armoedecijfer zelf vaststellen”, NoSpang.com (02/ 20/2014) http://www.nospang.com/index.php?option=com_content&view=article&id=38638: suriname-moet-armoedecijfer-zelf-vaststellen&catid=73:binnenland&Itemid=65 (checked 05/ 27/2014).

148 | Janneke Diepeveen and Matthias Hüning

Suriname. The main arguments in favor of English pertain to Suriname’s external relations, particularly with its Caribbean neighbors. However, when it comes to the integration of ethnic groups within Suriname, English does not seem like a natural choice. Even though Suriname used to be occupied by the English, the English language has not been maintained as a community language by any significant group of Surinamese (Borges 2014: 15). Thus, we should not overestimate the unifying effect of English in comparison with Sranantongo and Dutch.

4.8 Dutch is no longer the ‘colonial language’ but Suriname has its own national variety of Dutch During colonial times, Dutch in Suriname was expected to be as linguistically close to Dutch in the Netherlands as possible. Influences from Sranantongo were not accepted and any mistakes caused by Sranantongo were corrected. People in Suriname were more or less forced to speak Netherlandic Dutch, i.e., Dutch the way it was spoken in the Netherlands. Since (Netherlandic) Dutch was obligatory, it was perceived as the language of the oppressor and a language people in Suriname could and would hardly identify with. Older generations may still have this perception. However, there appear to have been changes in the younger generations. Fierens (2006) shows on the basis of a written survey that 64 % of young Surinamese do not consider Dutch as the colonial language. Only a minority of 16 % does. It has been suggested that Suriname’s independence plays a role in the changed perception of Dutch. Carew (1982: 1) notes that a sharper national awareness accompanied Suriname’s independence (1954/1975) and that this has inspired a new look at Dutch. In other words, together with Suriname’s independence from the Netherlands, the Dutch language has obtained a different position in society. However, changing perceptions can probably be situated earlier. In his dissertation from 1950 Johan Ferrier (later Suriname’s first president) stated the principles of how education in Suriname should be organized (Gobardhan-Rambocus 1997: 235). At that time, he was very clear in demanding that there should be only one single official language for education and that this ought to be the most prestigious one. In his view, this was standard Dutch as spoken in the Netherlands. However, during the 1960s Suriname experienced a growing tolerance towards Surinamese variants in standard Dutch. Dutch in Suriname was influenced by other languages used there, particularly Sranantongo. Some scholars claimed that codification of Surinamese-Dutch would

The status of Dutch in post-colonial Suriname | 149

have a positive effect in education since it would make Surinamese pupils more confident of their language skills (Gobardhan-Rambocus 1997: 238–239). During the following decades, tolerance towards Surinamese variants has continued to grow and Suriname stopped focusing only on European (i.e., Netherlandic) Dutch. The changing view is reflected in the gradual recognition and standardization of Surinamese Dutch as a separate national variety of Dutch (e.g. Bies 1994, 2009). Influences from Sranantongo are no longer banned from Dutch, but they are accepted as part of the Surinamese variety. This is reflected in Renata de Bies’ dictionary of Surinamese Dutch (prepared in 2008; commercial edition in 2009) which incorporates many words of Sranantongo origin. However, since lexicographers have to base description particularly on informal spoken language, they have the difficult task of determining in which cases a word is a Surinamese Dutch variant and in which cases it is actually an instance of interference from Sranantongo. In a more general sense, codeswitching is one of the main difficulties encountered in attempts to standardize Surinamese Dutch. Surinamese Dutch may be viewed as a continuum of (spoken) varieties situated on a scale ranging between 100 % Sranantongo and 100 % Dutch. Which words or phrases ought to be considered ‘standard Surinamese Dutch’ and which ones ought to be left out? This is particularly hard to determine, especially while the use of Sranantongo elements in speech appears to be linked to social factors. Standardization of Surinamese Dutch requires further linguistic investigation, which is difficult to carry out since Suriname lacks financial means for long-term research activities. This lack of research funding is particularly unfortunate considering that there are well-educated linguists in Suriname who are motivated to contribute to their variety’s standardization. Their number is likely to increase thanks to the fact that the Anton de Kom University in Paramaribo recently started a master program of Dutch coordinated by Renata de Bies. This program and the publication of Bies’ dictionary are important achievements for Surinamese people to find support and to feel more confident in using their own variety. Particularly teachers are expected to profit from this since they often deal with linguistic uncertainty. A Surinamese Dutch dictionary is one step for them to gain a stronger awareness of lexical variants, their (social) meaning and use (GobardhanRambocus 2008: 193). A Surinamese Dutch grammar could be another important step. As long as there exists no full academic description of Surinamese Dutch grammar, teachers have to decide individually and intuitively whether a certain grammatical item used by their pupils can be accepted as a Surinamese Dutch variant or should be rejected as an interferential element – which relates to the question: how much interference should be tolerated anyway?

150 | Janneke Diepeveen and Matthias Hüning

Dutch in Suriname obviously originates as the ‘colonial language’, the language of the oppressor. For many people in Suriname today, Dutch has become their own language: Dutch is no longer perceived as the language of the oppressor since Suriname is confident about its own national realization of Dutch (compare Gobardhan-Rambocus 2001: 511). There may still be differences between generations, however. The older generation was brought up with the idea that Dutch in Suriname contains mistakes that have to be corrected, particularly influences from Sranantongo. The younger generation is brought up differently. They firmly agree that they do not speak like the Dutchmen. In fact, any Surinamese speaking like a Dutchman is most likely to be laughed at (Essed-Fruin & Gobardhan-Rambocus 1992: 16). Instead, Suriname holds on to its own pronunciation and vocabulary. Thus, the Surinamese realization of Dutch seems to be used as a social marker for speakers in Suriname to distinguish themselves from speakers from the Netherlands. This does not mean that all language users are aware of, or ready to accept, linguistic variation in Dutch.12 The concept of ‘Surinamese Dutch’ as a linguistic variety in its own right has not (yet) established itself in the minds of people, whereas in Belgium, the concept of ‘Belgian Dutch’ as opposed to Dutch in the Netherlands is already more widely known and used. Apart from a growing recognition of the Surinamese variety from within (in Suriname), there is also recognition from outside. The most significant example probably is the Dutch Language Union. When it was established, the Union was very eager to keep the Dutch language ‘united’. However, its point of view has developed over the years. Today the Dutch Language Union leaves room for a pluricentric view with differing norms. It represents the idea of Dutch as a pluricentric language with different standard realizations of Dutch in different nations. The Union accepts three different national standard realizations of Dutch: Belgian, Netherlandic and Surinamese Dutch; the status of Carribean Dutch remains implicit.13 In a pluricentric view Surinamese variants are not seen as deviations from the (Netherlandic) standard but they are explicitly accepted as characteristics of equal national varieties of one and the same language.

|| 12 It is an aim of the project Dutch++ to help establish this kind of linguistic awareness among speakers, learners and teachers of Dutch all over the globe. 13 The Dutch standard language is described by the Dutch Language Union on its website as ‘Eén standaardtaal met drie poten’ (one standard language with three legs). See: http://taalunieversum.org/inhoud/feiten-en-cijfers#standaardtaal (checked 05/27/2014).

The status of Dutch in post-colonial Suriname | 151

5 Conclusion Carew (1982) described Suriname’s situation as ‘linguistic schizophrenia’. However, the factors we discussed in section 4 suggest that the linguistic situation of Suriname may be organized in terms of a relatively stable diglossia (extended diglossia in the sense of Fishman 1967). There is an organization on a sociocultural level with two functionally distinct languages – that is, if we ignore Suriname’s other twenty languages, which are mainly used for communication within the ethnic group. In Suriname, Dutch and Sranantongo are seen as being not in conflict with one another. Each fulfils relatively clearly defined roles in society: Dutch is used in one set of circumstances and Sranantongo in an entirely different set. The two languages are kept quite apart in their functions, which is a key defining characteristic of diglossia (Wardhaugh 2010: 85). In this view, Dutch can be described as the ‘high variety’, the powerful variety which holds the prestige position in Surinamese society. It is the language one has to learn in order to function well in society (Gobardhan-Rambocus 1997: 246). Whereas Dutch is standardized and taught in institutions, Sranantongo is not even a school subject. Sranantongo can be described as the ‘low variety’: It has a social stigma attached to it and it is considered less beautiful and less expressive than Dutch, which is used to seek prestige. However, Sranantongo has covert prestige (in the sense of Labov 1966): by using it, speakers choose to differ from the standard. People prefer to use Sranantongo when they express emotion, when they want to tell a story or a joke. Whereas Dutch is the language for official and formal functions, for written documents, written media and fine literature, Sranantongo is first and foremost an oral language which is used in informal and spoken situations. This view of the language situation in Suriname as being organized along diglossic lines is, however, challenged by the actual language use of the speakers. While Dutch and Sranantongo are ascribed different functions in Surinamese society by their speakers, these speakers do not always keep them apart in daily practice. As mentioned above (in 4.8.), the multilingual speakers make use of their linguistic repertoires in different ways. We need a continuum view of the spoken varieties in order to describe the language(s) used in actual discourse. The social expectations with regard to the functions and the use of Dutch and Sranantongo do not (or at least not necessarily) correspond to communicative reality. We conclude that the current linguistic situation in Suriname must be understood against the background of a nation which is constructing its own national identity. Suriname is a young nation which, on the one hand, is still deal-

152 | Janneke Diepeveen and Matthias Hüning

ing with its colonial past and struggling to define itself and its relation to the former ‘mother country’, the Netherlands. On the other hand, Suriname has to position itself vis-à-vis the Caribbean islands and the neighboring SouthAmerican nations. Suriname’s current linguistic situation reflects the identity it is constructing. How stable is the socially accepted diglossia, i.e., which macrolevel changes may occur in the following generations? Will Dutch keep its strong position? Will Dutch and Sranantongo continue to exist side by side? Will English eventually play a role as well? Macro-level changes in Suriname’s linguistic situation may be instigated from above and from below. Changes from above pertain to measures taken by the Surinamese government towards a language policy. The final memorable measure taken was the participation of Suriname in the Dutch Language Union, which was considered as an indication that Suriname actively strengthened the position of Dutch as its official language. However, the debate on which should be the official language(s) in Suriname continued. In an article in The New York Times (Romero 2008), the Surinamese writer Paul Middellijn made the following statement: “We shook off the chains of Dutch colonialism in the 1970s, but our consciousness remains colonized by the Dutch language”. Therefore, he proposed that English should be made Suriname’s official language instead of Dutch. Other voices urge Suriname to reconsider its policy with respect to the language of education (e.g. the Flemish journalist Pieter Van Maele 2011/2013). So far, there is no indication that the government is taking any efforts to introduce Sranantongo as a school subject or as (additional) language of instruction. As we already pointed out, Sranantongo will remain an oral language as long as it is not taught in institutions and not standardized. There are, however, recent attempts to codify Sranantongo: Eddy van der Hilst (2013) recently published a grammar of Sranantongo. Codification is a significant step towards standardization (compare Bies 2009: 18). Changes in Suriname’s linguistic situation from below pertain to language attitude and language use of the Surinamese population. As long as Dutch is perceived by Surinamese people as the prestige language, they will continue passing it on to their children and it will probably persist in its current functional domains. Brandon (2006) conducted a written survey with 315 Surinamese respondents (both in Suriname and in the Netherlands) on attitudes towards Dutch. The survey shows that Surinamese people agree that Dutch is part of the Surinamese identity (particularly in view of the historical connections with the Netherlands). The respondents agree that Dutch should remain the main language of education in Suriname. All in all they have a positive attitude towards Suriname’s accession to the Language Union, although the Surinamese in the Netherlands are slightly more critical. Despite the overall positive

The status of Dutch in post-colonial Suriname | 153

attitude towards Dutch, the Surinamese, particularly those in Suriname, believe that a good command of English is important for the socio-economical development of their country (membership of the CARICOM). Brandon’s survey shows that Surinamese are in favor of two official languages for Suriname and they prefer the combination Dutch/English over Dutch/Sranantongo. Interestingly, there is a larger preference for the latter combination among Surinamese in the Netherlands. This is in line with Borges’ (2014: 29) observation that Sranantongo is established “as an important and highly visible heritage language in the Netherlands”. Apparently, the use of Sranantongo is an essential marker for Surinamese to achieve group identity in the Netherlands. At the same time, in Suriname, the use of English is on the rise in certain contexts. Not only is English the language of the internet, the younger generation also likes to introduce English phrases in daily conversation (which may be facilitated by the fact that Sranantongo is an English-lexicon creole). If this generation starts passing on English to their children, this may strongly affect the linguistic situation of Suriname. Since we are concerned with the status of Dutch as a pluricentric language (see introduction), we are also interested in which way macro-level changes in Suriname’s linguistic situation might affect the Dutch language structurally. As pointed out above there has already been a rise in tolerance towards Sranantongo elements in Dutch. The time that Dutch needed to be as close to Netherlandic Dutch as possible is definitely over. It is not uncommon to introduce Sranantongo vocabulary and expressions in both spoken and written Dutch. Also, several transfer phenomena can be observed (see Borges 2014 for various contact phenomena and Radke 2013 for examples from chat language). Increasing influences from Sranantongo may lead to modifications of Surinamese Dutch. This might have additional consequences for Dutch as a pluricentric language: it could lead to further divergence from European Dutch or even a separation between Surinamese and European Dutch. Two decades ago, Geerts (1992: 73) reflected upon the status of Dutch as a pluricentric language by raising the question whether “[...] the independence of Surinam could lead to an increasing differentiation between European and Surinamese Dutch so that at a given moment the Surinamese might want to declare the Dutch in Surinam emancipated, just as happened in South Africa”. At the moment we do not see any direct indications that Surinamese Dutch is developing into a language of its own. In fact, it would be disadvantageous for Suriname should Surinamese Dutch diverge too much from European Dutch, since Suriname still benefits from smooth communication with the Low Countries. Suriname finds itself in an ongoing process of constructing its own national, post-colonial, identity. Within this process, both Sranantongo and Suri-

154 | Janneke Diepeveen and Matthias Hüning

namese Dutch are awarded central roles: each contributes in its own way to the Surinamese identity. Whereas the use of Sranantongo is a solidarity marker for achieving a Surinamese group identity, the use of Surinamese Dutch may be a marker for post-colonial Suriname to differentiate itself from the Netherlands. Whether they can establish themselves jointly as Suriname’s national languages will be up to the next generations.

References Asin, Linda, Lila Gobardhan & Ismene Krishnadath. 2004. Onderwijs in en van het Nederlands in Suriname. Startdocument. Den Haag: Nederlandse Taalunie. Bies, Renata de. 1994. Op weg naar standaardisatie van het Surinaams-Nederlands: Een lexicale beschrijving van vijftien jaar administratietaal van de Republiek Suriname. Dissertatie, Universitaire Instelling Antwerpen. Bies, Renata de. 1997. Het Nederlands in Suriname. Vlaanderen 46. 9–15. Bies, Renata de. 2009. Prisma woordenboek Surinaams Nederlands. Houten: Het Spectrum. Borges, Robert. 2014. The life of language: Dynamics of language contact in Suriname. Utrecht: LOT (LOT Dissertation Series 348). Brandon, Ruth. 2006. Suriname in de Nederlandse Taalunie. Een beleidsanalyse en een onderzoek naar opvattingen van Surinamers over de toetreding van Suriname tot de Nederlandse Taalunie. Tilburg: Universiteit van Tilburg [unpublished master thesis]. Brandon, Ruth; Sjaak Kroon & Jeanne Kurvers. 2007. Nederlands in Suriname. Opvattingen van Surinamers over de toetreding van Suriname tot de Nederlandse Taalunie. OSO Tijdschrift voor Surinamistiek 26(2). 257–273. Carew, Joy Gleason. 1982. Language and survival: Will Sranan Tongo, Suriname’s lingua franca, become the official language? Caribbean Quarterly 28(4). 1–16. Carlin, Eithne B. & Jacques Arends (eds.). 2002. Atlas of the languages of Suriname. Leiden: KITLV Press. Eersel, Hein. 1997. De Surinaamse taalpolitiek: een historisch overzicht. In Kees Groeneboer (ed.), Koloniale Taalpolitiek in Oost en West. Nederlands-Indië, Suriname, Nederlandse Antillen en Aruba, 207–223. Amsterdam: Amsterdam University Press. Essed, Eva D. 1983. De veranderende status van Sranan. In Eddy Charry, Geert Koefoed & Pieter Muysken (eds.), De talen van Suriname; Achtergronden en ontwikkelingen, 47–52. Muiderberg: Coutinho. Essed-Fruin, Eva D. & Lila Gobardhan-Rambocus. 1992. Het Nederlands in Suriname. Neerlandica Extra Muros 39(3). 10–21. Fierens, Laura. 2006. Taalattitudes van Surinaamse jongeren ten opzichte van het Nederlands. Leuven: Katholieke Universiteit Leuven [unpublished master thesis]. Fishman, Joshua A. 1967. Bilingualism with and without diglossia; diglossia with and without bilingualism. Journal of Social Issues 23(2). 29–38. Geerts, Guido. 1992. Is Dutch a pluricentric language? In Michael Clyne (ed.), Pluricentric languages. Differing norms in different nations, 71–91. Berlin/New York: Mouton de Gruyter.

The status of Dutch in post-colonial Suriname | 155

Gobardhan-Rambocus, Lila. 1997. Suriname en het Nederlands. In Kees Groeneboer (ed.), Koloniale Taalpolitiek in Oost en West. Nederlands-Indië, Suriname, Nederlandse Antillen en Aruba, 225–249. Amsterdam: Amsterdam University Press. Gobardhan-Rambocus, Sabitrie Lilawatie. 2001. Onderwijs als sleutel tot maatschappelijke vooruitgang. Een taal- en onderwijsgeschiedenis von Suriname, 1615-1975. Zutphen: Walburg Pers. Gobardhan-Rambocus, Lila. 2008. Nieuw woordenboek van de Surinaamse bijdrage aan het Nederlands [Bespreking]. Ons Erfdeel 51. 191–193. Hilst, Eddy van der. 2013. Taki Sranantongo bun. De grammatica van het Sranan deel 1. Paramaribo: Suriprint N.V. Kleine, Christa de. 2002. Surinamese Dutch. In Eithne B. Carlin & Jacques Arends (eds.), Atlas of the languages of Suriname, 209–230. Leiden: KITLV Press. Kleine, Christa de. 2013. Dutch in Suriname. In Frans Hinskens & Johan Taeldeman (eds.), Language and space. An international handbook of linguistic variation. Vol. 3: Dutch, 841–858. Berlin/Boston: De Gruyter. Kroon, Sjaak & Jeanne Kurvers. 2009. Opvattingen over Nederlands en andere talen als instructietaal op Aruba en in Suriname. Toegepaste Taalwetenschap in Artikelen 82. 57–68. Labov, William. 1966. The social stratification of English in New York City. Washington DC: Center for Applied Linguistics. Radke, Henning. 2013. Mehrsprachigkeit und Code-Switching: Zur Interaktion zwischen Niederländisch und Sranantongo in surinamischer Onlinekommunikation. Berlin: Freie Universität Berlin. [unpublished master thesis] Romero, Simon. 2008. In debate over official language, Suriname seeks itself. The New York Times 03/23/08. http://www.nytimes.com/2008/03/23/world/americas/23iht23suriname.11335008.html (checked 05/23/14). Sebba, Mark. 2000. Orthography and ideology: Issues in Sranan spelling. Linguistics 38(5). 925–948. Van Maele, Pieter. 2011. Zonder het Nederlands waren we pas echt geïsoleerd. Taalschrift 81. http://taalschrift.org/editie/81/zonder-het-nederlands-waren-we-pas-echt-ge-soleerd (checked 05/23/14). Van Maele, Pieter. 2013. Suriname moet Sranantongo in de klas omarmen. Taalschrift 98. http://taalschrift.org/editie/98/suriname-moet-sranantongo-de-klas-omarmen (checked 05/23/14) Visser, Marieke. 2005. Suriname is geen koekoeksjong. Taalschrift 02/17/2005. http://taalschrift.org/reportage/000773.html (checked 05/27/2014). Wardhaugh, Ronald. 2010. An introduction to sociolinguistics. 6th ed. Chichester, West Sussex/Malden, MA: Wiley-Blackwell.

Susanne Mohr

From Accra to Nairobi – The use of pluralized mass nouns in East and West African postcolonial Englishes Abstract: The present article investigates the use of pluralized mass nouns such as furnitures in two East African (Kenyan, Tanzanian) and two West African (Nigerian, Ghanaian) Englishes. Using corpus data from the ICE and GloWbE corpora, the analysis has several levels. Firstly, similarities and differences concerning the use of 22 selected mass nouns are scrutinized. This includes dynamics and contexts of use, potentially providing reasons for the use of mass nouns in postcolonial African Englishes. Finally, the use of plural mass nouns is discussed in the context of areoversals of African Englishes. Keywords: English in Africa, language variation, morphosyntax, semantics, number, corpus linguistics

1 Introduction The field of World Englishes is dedicated to the study of the varieties of English worldwide, their sociolinguistic settings and structural features. Especially the postcolonial or “New” Englishes spoken in former Asian and African colonies, have received increased attention in the past decades (e.g. Platt et al. 1984; Fishman et al. 1996; Schneider 2003; Kachru et al. 2006). They have a shared socio-historical background as former colonies of the British Empire and are often characterized by situations of intense language contact. Platt et al. (1984: 2f.) established four sociolinguistic criteria in order to determine a variety’s status as a “New” English, resulting in three distinct types of postcolonial Englishes. These types occur in the following sociolinguistic situations: Type 1: one or more local languages and a non-English language of wider communication are present besides English

|| Susanne Mohr: Rheinische Friedrich-Wilhelms-Universität Bonn, Department of English American and Celtic Studies, Regina-Pacis-Weg 5, 53113 Bonn, Germany. E-mail: [email protected]

158 | Susanne Mohr

Type 2: one or more local languages and an English-based pidgin as language of wider communication are present besides English Type 3: an English-based creole, usually in a (post)creole continuum is present besides English Pertaining to the African situation, the first two types are of particular importance. Examples of the first type are Kenya, Uganda and Tanzania with Kiswahili as the non-English language of wider communication, also called the “African lingua franca” (Schmied 2008, 2012).1 Ghana and Nigeria constitute examples of the second type, with Ghanaian and Nigerian Pidgin as regional linguae francae (Huber 1999; Faraclas 1996). Further, a distinction according to acquisition is reported by Schmied (2012). He claims that English in Africa can be spoken as 1) a native language (L1) by African-born whites and expats, 2) a native language by locally born Africans, 3) a non-native language spoken fluently as a second language (L2), 4) a non-native language spoken imperfectly as a foreign language. This distinction takes up the classification made by many researchers differentiating between English as a native language (ENL), English as a second language (ESL) and English as a foreign language (EFL) (e.g. Moag 1992) and emphasizes the function of English within a society. However, an acquisition-based classification of varieties is rather problematic for the context under investigation, as for many other postcolonial contexts for that matter. It applies European analytical frameworks to complex linguistic situations that are completely different from Western multilingual settings, and cannot account for the sociolinguistic reality of these situations (Anchimbe 2007). There, it is often not clear which language in the repertoire of a speaker constitutes his “first”, “second” or “third” language (Webb & Kembo-Sure 2001, De Bot et al. 2005) so that a clear-cut distinction in terms of order of acquisition as presented above is not possible. Differences between a speaker’s languages should rather be defined according to settings of use, just as language competence should take into account domains or sociolinguistic spaces (Blommaert et al. 2005) and speakers’ communicative goals in specific situations. The traditional measurement of linguistic competence according to writing, speaking, reading and understanding is not always applicable in these contexts either (Anchimbe 2007). Accordingly, the concept of the “native” speaker is rather problematic and hardly definable considering

|| 1 The status of Swahili differs considerably on the African continent and between the individual countries of English-speaking East Africa though. Hence, it has a much stronger status in Tanzania as an official language than in Uganda, for example (Schmied 2012).

The use of pluralized mass nouns in East and West African postcolonial Englishes | 159

the notion of fluency or competency as mentioned in types 3) and 4) of the acquisition-based model above needs to be redefined in these contexts. “Truncated multilingualism” that can often be found in these contexts is domain-specific (Blommaert et al. 2005). Given the situation described here, the type of English that this paper is concerned with is difficult to classify. This can only be achieved with reference to the type of corpus data used, which will be described in more detail in section 3. For the moment it should suffice to say that the Englishes analyzed for the respective countries constitute educated language, often of the written type. Typologically, New Englishes in Africa and Asia have shown similar traits (e.g. Kortmann & Szmrecsanyi 2004; Szmrecsanyi & Kortmann 2009; Szmrecsanyi & Kortmann 2011; Kortmann 2013; Kortmann & Lunkenheimer 2013; Kortmann & Schröter forthcoming) which have been proposed to constitute varioversals of indigenized L22 varieties (Lunkenheimer 2012). The African varieties of English remain scarcely researched and only recently have studies started to compare African postcolonial varieties, referring to national varieties that have developed in former British colonies especially after independence, from a typological point of view. While examinations such as Schmied (2006, 2008, 2012) compare selected features of East African Englishes, they are not typologically oriented. Brato & Huber (2012) and Huber (2012) are the first to investigate in this direction. Huber (2012) found that there are four typically African morphosyntactic features. These are the use of double determiners, come-based future/ingressive markers, very as qualifier and the nonstandard use of plural mass nouns. Kortmann & Schröter (forthcoming), however, identify five features, the use of double determiners, no number distinction in demonstratives, come-based future/ingressive-markers, conjunction doubling and the use of postpositions as diagnostic features for Africa. The present analysis of plural mass nouns mentioned by Huber (2012) can help shed further light on these diverging descriptions. Both analyses are based on data from the Mouton World Atlas of Variation in English (WAVE) (Kortmann & Lunkenheimer 2012), which constitutes the most important tool for the typological study of World Englishes today and is also available as an electronic database online (Kortmann & Lunkenheimer 2013, http://ewave-atlas.org/). Altogether, the atlas contains 235 morphosyntactic

|| 2 The problems related to the classification of Englishes according to acquisition have just outlined. The term “L2 variety” was used by Lunkenheimer (2012) and was hence re-used here with reference to her text. The same holds true for other usages of this terminology throughout the paper.

160 | Susanne Mohr

features whose frequency is rated for each variety of English. These ratings range from “A – feature is pervasive or obligatory” over “B – feature is neither pervasive nor extremely rare” over “C – feature exists, but is extremely rare” to “D – attested absence of feature”. The ratings are based on judgments provided by experts on a certain variety. However, Lunkenheimer (2012) and Kortmann (2015) mention that these judgments are often based on introspection on the part of the expert, rather than on empirical investigation. Further, these ratings disregard stylistic or genre-specific differences of language use and emanate from general linguistic tendencies. This is especially problematic in the context of postcolonial linguistics which has shown the importance of different text types in postcolonial societies. It is hence imperative to verify the frequency ratings by empirical studies and analyze the usage constraints of individual features (Kortmann 2013). This paper focuses on feature 55 of WAVE, the non-standard use of plural mass nouns, in New African Englishes, comparing two East African (Kenyan and Tanzanian English), with two West African (Ghanaian and Nigerian English) varieties. The main questions to be scrutinized are first of all whether the frequencies of the feature provided in WAVE can be verified for the respective varieties, and secondly whether on the basis of these results the use of plural mass nouns shows tendencies towards areoversality or Africa as a geographical region. In order to answer these research questions, the paper draws on corpus data from the International Corpus of English (East African, Ghanaian and Nigerian components), and the Corpus of Global Web-Based English (GloWbE) (Davies 2013). In section 2 of this paper, an account of the count-mass distinction in language, contrasting English and African languages, is provided. Section 3 elaborates on the data and methodology applied, while section 4 presents the results of the corpus analysis. These results are discussed in detail in section 5. Section 6 concludes the paper with a summary of the results and an outlook on future research.

2 Number marking, mass and countability The linguistic means to express number and countability vary considerably cross-linguistically (Storch & Dimmendaal 2014).This is also due to the fact that the distinction into count and mass nouns is multidimensional, not only referring to grammatical, but also to semantic and ontological distinctions. On the semantic-ontological level, countability refers to individuated entities that can

The use of pluralized mass nouns in East and West African postcolonial Englishes | 161

be counted, while mass concepts denote unindividuated entities that cannot be separated into countable units (Massam 2012). Sorting and individuation are sub-sections of research in this context (Storch & Dimmendaal 2014). Although the general concepts of mass and countability seem to be available to all humans, their lexicalization and grammaticalization differ considerably crosslinguistically. A general controversy between typologists (Chierchia 1998, 2010) and universalists (Borer 2005) concerning the count-mass dichotomy has existed for a long time. Originally, the Quinian theory that conceptually, count nouns identify countable units, while mass nouns refer to entities that come in mass form and cannot be separated into individual elements (Quine 1960) was the general consensus. Gillon (1996) or Chierchia (1998) challenged this assumption in claiming that a conversion of mass nouns into count readings according to sorter and portion coercion is only possible if the denotation of a mass noun already comprises minimal parts into which the noun can be subdivided. This refers to Pelletier (1975) and Bunt (1985) who stated that all nouns, irrespective of count or mass, can be coerced into the other category by the Universal Grinder and the Universal Sorter. An example of the grinder is shown in (1), an example of the sorter in (2) (modified from Wiese 2012:69): (1)

There is an eggplant/are eggplants on the plate. I want eggplant on my pizza.

[count] [mass]

(2)

She drinks beer. The best beers are from Germany.

[mass] [count]

Moreover, Wiese (2012) mentions portion coercion as in (3). (3)

She drinks water. We’ll have two waters please.

[mass] [count]

Barner & Snedeker (2005) showed that both children and adult native speakers of English show a considerable influence of mass vs. count syntax for quantity judgments in conversion environments of flexible terms such as stone or chocolate. With regard to conversion, Corbett (2001: 82) emphasizes that countability is not so much a feature of a noun with respect to its semantics but rather that it is determined by noun phrase structures. This, he argues, is the reason why many nouns can undergo coercion as exemplified in examples (1)–(3). Another approach to the count-mass distinction that does not take the possibility of conversion into consideration since it is supposed to be frequently inherent in a noun, is proposed by Rijkhoff (1991), who suggests that countability is marked as nominal aspect, which is defined as “the way a property is

162 | Susanne Mohr

represented in the spatial dimension” (Rijkhoff 1991: 291). This is based on the importance of spatial orientation for human cognition and takes the features SHAPE and STRUCTURE into account. SHAPE refers to referents with a definite outline, STRUCTURE to referents that are divisible. Taking these features as a starting point, there are not two but four nominal aspects which are often inherent in nouns: conceptual, mass, individual and collective (Rijkhoff 1991: 294). Conceptual nouns often occur in South East Asian languages such as Mandarin Chinese. They are unmarked for shape and structure and hence are nondivisible and have no definite outline. Consequently, they cannot be pluralized as can be seen in example (4) below. (4)

Mandarin yi pi ma one CL horse ‘one horse’

In these languages, a classifier, such as pi above, is necessary to combine the nouns with a numeral.3 This holds irrespective of the semantic count-mass distinction. The only peculiarity of ontological mass nouns in Mandarin is that they are preceded by a massifier if combined with a numeral. Massifiers differ from classifiers such as ben in (4), in that they also name the unit of individuation, as ge which is the massifier for people in particular. Conceptual differences between speakers of languages with many conceptual nouns and languages such as English that have fewer members of this category were shown in studies which found that speakers of plural languages like English are less likely to construe an entity as unindividuated than speakers of classifier languages like Mandarin (Imai & Gentner 1997; Lucy & Gaskins 2001, 2003; Imai & Mazuka 2003, 2007). Individual nouns, in contrast, often occur in large numbers in IndoEuropean languages. They are marked for SHAPE as they have a definite outline but not for STRUCTURE and are hence non-divisible. An example is English car. Nouns with mass aspect are marked for STRUCTURE and hence divisible but not marked for SHAPE and thus have no definite outline (Rijkhoff 1991: 295). A typical example is water which, if divided consists of smaller “portions” of water.

|| 3 It has been proposed that there is a general distinction between languages like Chinese, which are also called transnumeral or classifier languages since they do not mark nouns for plural, and plural languages like English that do mark plural on nouns (Chierchia 2010). A third type of language is also mentioned in which neither numeral classifiers not number marking are used. These are extremely rare though.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 163

Finally, collective aspect nouns are marked for both SHAPE and STRUCTURE. Thus, “the property designated by [that kind of] noun […] defines a number of distinct individuals, which all share the same property […]” (Rijkhoff 1991: 296). An example is bunch which, if divided consists of many similar parts, e.g. flowers. Grammatically, the realization of mass and countability is diverse. While it may be lexically inherent and not marked at all, tripartite number systems, distinguishing between singulative, collective and plural are not rare crosslingustically (Storch & Dimmendaal 2014). Hence, many languages for example of the Nilo-Saharan language phylum show these distinctions. This reflects a conceptual differentiation into entities that naturally occur in mass and plural form or as individuals, on the grammatical level. This conceptual differentiation is crucial for the distinction of count and mass nouns in English as well. The following sections delineate the grammatical and conceptual differences between the count-mass distinction in English and African languages.

2.1 Mass nouns in English As opposed to classifier languages outlined before, English is a plural language that morphologically marks nouns for the plural. Mass nouns are an exception from this rule, as mass nouns such as rice, water, furniture cannot be marked for plural and cannot be combined with determiners or numerals (cf. examples (5)– (7)). Count nouns such as cat, book, chair can be (Quirk et al. 1985). (5)

*There are silverwares on the table. There is silverware on the table.

(6)

*I took a new luggage on the train with me. I took new luggage on the train with me.

(7)

*We have seven furnitures in our living room. We have seven pieces of furniture in our living room.

Cowper & Currie Hall (2012) claim that English nouns are inherently unspecified for mass or count and only the morphosyntactic filling in of features determines their category membership. This contradicts Pelletier’s (1975), Bunt’s (1985) and Wiese’s (2012) theories mentioned above. For English (and other languages), another class of nouns that grammatically behaves like mass nouns but cognitively belongs to the category of count nouns has been claimed (e.g. Gillon 1996; Chierchia 1998; Barner & Snedeker

164 | Susanne Mohr

2005; Schmied 2006; Li et al. 2009; Wiese 2012; Wiltschko 2012). Examples are furniture, cattle or silverware. They have either been termed “collective nouns” or “object-mass nouns”. In English, these are grammatically mass nouns in that they cannot be inflected for plural, occur without determiners and cannot be combined directly with a numeral as in examples (5)–(7). However, conceptually these nouns quantify over individual objects instead of substances like other mass nouns such as water, and are divisible into minimal parts (Huntley-Fenner et al. 2002; Barner & Snedeker 2005; Li et al. 2009; Wiese 2012). Mohr (2014) showed that the distinction between object-mass and substance-mass nouns is decisive in the use of non-standard plural mass nouns by Kenyan and Tanzanian speakers of English. Generally, it is object-mass nouns that are used in non-standard plural form by L2 speakers. Further, the problems African New English speakers have with these mass nouns were shown to be culturally determined, due to different conceptualizations of many of these nouns (cf. section 2.3).

2.2 Mass nouns in African languages The indigenous languages of Africa can be categorized into four large language phyla: Niger-Congo, Nilo-Saharan, Afro-Asiatic and Khoisan (Heine & Nurse 2000). In Kenya, there are 67 living languages which belong to the Bantu (NigerCongo), Cushitic (Afro-Asiatic) and Nilotic (Nilo-Saharan) sub-families. Tanzania’s 126 living languages mostly belong to the Bantu language family, although Nilotic, a few Cushitic and a few Khoisan languages can also be found (Lewis et al. 2014). Kiswahili (Bantu) is the current regional lingua franca of Eastern Africa (Schmied 2006). In West Africa, pidgin and creole languages are wide spread, and indigenous languages like Hausa function as lingua franca in parts of Nigeria and Ghana (Huber 2008a, 2008b). There are several aspects of African number systems that are interesting in the context of this paper. The first one is the noun class system of Bantu languages. The semantics of the noun classes for Bantu have often been attempted to be determined (Richardson 1967; Dixon 1968; Givón 1971; Creider 1975; Denny & Creider 1976) but the grouping of nouns remains controversial. Although the noun classes might indicate number, they are very different from numeral classifiers in classifier languages and could rather be compared to the gender systems of Indo-European languages (Aikhenvald 2003). Kiswahili, for example, has 18 noun classes, most of which are paired into a singular and a plural class. Class membership is indicated by a prefix.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 165

(8) (8a)

(8b)

Kiswahili m-toto CL1-child ‘child’ wa-toto CL2-child ‘children’

As can be seen, the prefix m- in (8a) indicates class membership and singular, while the prefix wa- in (8b) indicates class membership as well as plural number. The number value especially of many Kiswahili non-count nouns is not clear though, as they are not paired (Cheng & Schadeberg 2007). Examples are joto ‘heat’ (class 5) and usingizi ‘sleep’ (class 11). However, according to Cheng & Schadeberg (2007), the only “true” mass nouns of Kiswahili are mchele ‘rice’ (class 3), ngano ‘wheat’ (class 9), mahindi ‘maize’ (class 6) and ulezi ‘millet’ (class 11). In this context, the literature does not mention words like serikali ‘government’ and jumuiya ‘community’. Semantically, they are similar to objectmass nouns like cattle. The other uncountable nouns mentioned above are also categorized according to syntactic behavior: mass nouns cannot be combined with numerals and the quantifiers kila ‘each’, zima ‘entire’ and kadhaa ‘a few’ as count nouns. Instead, non-count nouns are combined with the quantifiers -ote ‘all’, -ingi4 ‘much’, kidogo ‘a little’ or kadhri ‘some’. As for the quantifier -chache (‘few’), differing descriptions can be found. While Cheng & Schadeberg (2007) mention that is cannot be combined with non-countables, Zerbian & Krifka (2008: 390) provide a counter example: (9)

Kiswahili Ma-ji ma-chache teremka kwa CL6-water CL6-little get.off PREP ‘Little water is flowing in the river.’

mto.5 CL3.river

However, in certain circumstances mass nouns can be accompanied by one of the count noun quantifiers as in kila mchele ‘each kind of rice’. Apparently, there is a caveat among native speakers as some of them accept these kinds of forms, while others do not (Schadeberg, p.c.).

|| 4 Zerbian & Krifka (2008) mention -engi as the Swahili quantifier meaning ‘many’. 5 Emphasis in the original.

166 | Susanne Mohr

Accordingly, Bantu languages like Kiswahili show a morphosyntactic and semantic distinction into count and mass nouns, which has not been exhaustively described as yet. Number assignment and the count-mass distinction in Atlantic (NigerCongo) languages, which are the native languages of West African speakers of English6, is the second intriguing issue in the context of this paper. Certain mass nouns can be pluralized regularly like the word for ‘information’ in Wolof: xibaar (SG) – xibaar yi (PL). This emphasizes the arbitrariness of the count-mass distinction cross-linguistically. A concept like ‘information’ can for example also be pluralized in French (informations) or German (Informationen) but not in English. Other mass nouns in Atlantic languages occur in the uninflected form in the plural and have to be marked by a diminutive marker indicating singular if they occur in small quantities, like the word for ‘water’ in Jóola: mal (PL or ‘water in general’) – jal (SG), or the word for ‘salt’ mutassaen (PL or ‘salt in general’) – jitassen (SG). Most interestingly, Jóola also has a standard distinction into singular and plural for one class of nouns and a tripartite classification into singular, plural and a second plural (probably collective) for another class of nouns. Example (10a) shows the simple singular-plural distinction, (10b) the tripartite distinction. (10) Jóola (10a) é-luup CL3-house ‘house’ (SG) (10b) é-lonloŋ CL3-earring ‘earring’ (SG)

sí-luup CL4-house

‘houses’ (PL) sí-lonloŋ CL4-earring ‘earrings’ (PL)

bá-lonloŋ CL5-earring

‘many earrings’

This different morphological treatment of certain nouns hints at a different conceptualization of these which might ultimately impact the pluralization of mass nouns in English. The concepts of count and mass in African languages, as well as their grammatical manifestations are intriguing and their implications for African Englishes manifold. However, it is extremely difficult to determine the exact impact of a certain language and feature on African New Englishes, due to the large number

|| 6 All information presented in the remainder of this section is based on interviews with native speakers and data from the questionnaires distributed for the study presented (Mohr 2014). So far, this issue has not been empirically researched. I am indebted to M. Sambou, J. Mansaly, S. Wade and F. Cissé for their explanations on this issue.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 167

of indigenous languages and the omnipresent multilingualism in African countries, as mentioned above. Hence, the topics of “native language”7 interference or substratal influence are only touched upon here. This is in line with Ellis (1985), stating that L1 interference has been overemphasized in second language acquisition research. Instead, cognitive explanations for the pluralization of mass nouns that have been explored in Mohr (2014) are focused on.

2.3 Mass nouns cross-linguistically and cross-culturally To date, studies on the cross-cultural conceptualization (outside of acquisitional contexts) of mass and countability remain rare. Mohr (2014) treated exactly this issue with respect to different learners of English. The study investigated the perception and conceptualization of mass nouns across different African cultures, in comparison with European ones. This was based on theories concerning object-mass and substance-mass nouns as claimed by Wiese (2012) mentioned above. In a rating task, participants were asked to rate substance-mass, object-mass and countable items according to their individuality. Differences between these two classes could be shown in that countable items were generally rated to be more individualized than mass terms. What is more, differences between speakers of African languages (many Bantu languages like Kiembu or Kikuyu in Eastern Africa, or Twi in Western Africa, as well as Volta-Congo or Western-Nilotic languages) and speakers of European languages (German in this case), could be proven. Object-mass terms such as staff or furniture were rated very differently by the speakers of African as compared to those of German origin. Interestingly, flexible terms like string were also rated with considerable variation (Mohr 2014). The results mirror the cross-linguistic differences and arbitrariness of the grammatical manifestation of count and mass elaborated on above. The ratings reflect the grammatical treatment of nouns in African and Indo-European languages, which varies considerably too (cf. example (9), in which water is countable in Kiswahili, or French informations which is countable vs. English information as a mass noun although both are Indo-European languages (cf. also Inagaki (2014) on the variation of countability cross-linguistically). The results also proved that the grammatical and semantic-ontological mismatch of features in object-mass terms seems to play an important part in || 7 Note that the problem of categorization into native and non-native languages was already discussed in the introduction.

168 | Susanne Mohr

the pluralization of mass nouns in speakers of postcolonial Englishes. This is further proven by Bock et al. (2006) who showed that certain collectives are even constructed differently among speakers of different varieties of English. In their study on collectives (object-mass nouns) in British and American English, they showed that the nouns are more often used with singular verbs and pronouns in American English than in British English. In their experiments, phrases like these staff were fairly acceptable for British speakers, but not at all for Americans. This could, according to them, be traced back to differences in experience with collectives that might ultimately create differences in lexical specifications of the nouns. This, then, leads to dialectal differences in the use of collectives. Ultimately, this is an interesting finding that might have an influence on the use of, especially, object-mass nouns in African Englishes, as well.

3 Data and method In order to examine the frequencies and dynamics of plural mass nouns in the four African varieties under investigation here, a corpus analysis of different corpora was executed. These were the African components of the International Corpus of English (ICE) and the African components of the Corpus of Global WebBased English (GLoWbE) (Davies 2013). The British National Corpus as provided by Brigham Young University (BNC-BYU) served as a reference corpus for British English. These corpora largely consist of written texts or computer mediated (written) communication which is extremely different from the contexts of use of the African languages mentioned in 2.2 and 2.3.8 Blommaert et al. (2005) and Anchimbe (2007) emphasize that this is a crucial problem in the determination of multilingual spaces and language competency, especially with respect to African languages. Many of them are not written and/or standardized. Generally, standard languages are often artificial constructs and rather European ideas (Romaine 2000). As “ideological symbols of nationhood” (Romaine 2000: 88), they are often oriented towards linguistic centers and disregard language in peripheral areas (Blommaert et al. 2005). The contexts of use of the mass nouns under investigation here are hence different in the African languages and the New Englishes. With respect to the following study, this might also be of influence, especially since the varieties in the ICE components usually represent educated varieties of postcolonial Englishes (cf. Skandera 2003 for Kenyan English).

|| 8 Kiswahili is an exception here, since it also has a strong written tradition.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 169

As mentioned above, ICE-East Africa (Schmied et al. 1999), ICE-Nigeria (Gut 2013), ICE-Ghana (Huber forthcoming) and the equivalent African components of GloWbE, as well as the BNC were utilized for the present corpus analysis. The corpora are often-used for research into the varieties of English. Due to the circumstances of data collection, this is usually “educated English” as used in the respective countries. The ICE components are standardized in that each of them contains roughly 1 million words from different text types (written and spoken). ICE-EA is special in that it contains data from two varieties of English, namely Kenyan and Tanzanian English. Further, written to be spoken texts such as parliamentary debates or speeches can be found in the corpus.9 The software tool that was used to search the components of the corpus was AntConc (http://www.antlab.sci.waseda.ac.jp/software.html), an electronic concordancing tool. The GloWbE corpus contains 1.9 billion words altogether. Although this might seem considerably more than the individual ICE components, it has to be borne in mind that this comprises texts from 20 different countries altogether. The number of words of the individual GloWbE components as compared to the ICE components and the BNC is shown in Table 1. Tab. 1: Total number of words in the corpora used for the study Corpus

Total number of words

ICE-EA ICE-Nigeria ICE-Ghana GloWbE Kenya GloWbE Tanzania GloWbE Nigeria GloWbE Ghana BNC-BYU

1,005,345 1,010,382 421,48610 41,069,085 35,169,042 42,646,098 38,768,231 98,363,78311

A comparison of the ICE and the GloWbE data is especially interesting as the GloWbE data complement the ICE data in constituting very informal data from web pages. In contrast, written language is more formal since it is more deliber-

|| 9 This category of texts was included because not enough spoken data was available for the sub corpus. 10 Only the written part of ICE-Ghana (version from April 2014) could be considered as it has been properly transcribed and tagged. The spoken data is not yet ready for analysis. 11 This refers to the number of POS-tagged items in the corpus.

170 | Susanne Mohr

ate. Thus, this kind of data is probably more comparable to the data from the African languages discussed above. Further, the GloWbE data are more recent. As the GloWbE data are freely available and searchable upon registration on the website, no software tool was needed for the analysis of the corpus. Altogether, 22 words were chosen to be investigated, either mentioned in studies on New Englishes in general (e.g. Baskaran 2008) or on African Englishes in particular (e.g. Skandera 2003). The procedure included a pilot study and the subsequent main study. In the pilot study, 12 nouns were chosen: research, information, equipment, evidence, advice, knowledge, music, traffic, homework, cattle, luggage, furniture. Based on the results of the pilot study, the list of items was modified for the subsequent investigation, excluding words that were marginal in frequency. The words for the main study were: silverware, cattle, advice, luggage, ketchup, beer, stone, furniture, research, toothpaste, information, tea, equipment, music, string, traffic, cheese, luck, evidence, homework, staff, offspring.12 As can be seen, apart from object and substance-mass nouns, flexible mass terms such as string and stone were included as well. The frequency ratings of F 55 in WAVE are A for Kenya, B for Tanzania, B for Ghana, and A for Nigeria (Kortmann & Lunkenheimer 2012). It will be interesting to investigate whether these ratings can be confirmed by the current analysis.

4 Results In the following, the results of the corpus analyses are presented. Firstly, the frequencies of the terms in the individual corpora are provided, and subsequently analyzed in context.

4.1 Frequencies of the mass terms Frequencies of all the mass terms that were ultimately investigated are indicated in total, in comparison to the total number of instances of each mass noun in the respective corpus. The frequencies are indicated in percent. Table 2 shows the frequencies of the plural forms in the ICE components.

|| 12 Offspring is sometimes indicated as a noun with an irregular plural instead of a mass noun. It was included here since it falls into the category of object-mass nouns.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 171

Tab. 2: Frequencies of the pluralized mass terms in the African ICE components

researches informations equipments evidences advices furnitures knowledges musics traffics homeworks cattles luggages silverwares ketchups beers stones toothpastes teas strings staffs offsprings cheeses

ICE-EA

ICE-Nigeria

ICE-Ghana

2.2 % 0.3 % 10.7 % 1.0 % 1.7 % 0% 0% 0% 0% 0% 0.6 % 0% 0% 0% 7.81 % 36.7 % 0% 1.0 % 50 % 0.29 % 30.8 % 0%

3.2 % 1.1 % 4.1 % 6.5 % 1.8 % 0% 0% 0% 0% 0% 0% 33.3 % 0% 0% 0% 22.7 % 0% 0% 22.7 % 2.8 % 100 % 0%

0% 0.7 % 4.8 % 3.4 % 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 21.4 % 35.5 % 0% 0% 28.6 % 0% 0% 0%

As can be seen in Table 2, both differences and similarities can be made out concerning the frequencies of mass nouns in the three corpora. Hence, there are only three items that occur relatively frequently in plural form in all ICE components, namely equipments, strings and stones. Only the form equipments is nonstandard, as string and stone are flexible nouns that can be used in singular and plural form. Further, furnitures, knowledges, musics, traffics, homeworks, silverwares, ketchups, toothpastes and cheeses do not occur at all in any of the ICE corpora. This is less surprising for words like ketchups and toothpastes which are classical substance-mass nouns that cannot be pluralized, and are in general very non-African concepts, but more so for cheeses, which can be used in a “type-of reading”, or homeworks, which is often mentioned to be a deviant form in non-native speakers of English. Cheese, however, is often bought in large blocks in Africa, which could make it more prone for substance-mass usage. The non-usage of silverwares can be explained by the low occurrence of the term in the corpora in general: only in ICE-Nigeria it is used once. In the context of substance and object-mass nouns, only the non-occurrence of furnitures is surprising in that it is an object-mass term that should pose serious problems to non-

172 | Susanne Mohr

native speakers. However, furniture shows very low frequencies in the West African corpora and is not very frequent (19 tokens) in ICE-EA either. A similar case is cattles, which would have been expected to be frequent as it is an objectmass term; however it only occurs in ICE-EA (1 token). In the West African corpora cattle is an infrequent term in general, but in ICE-EA it occurs fairly frequently (178 tokens). This is an interesting case and remains to be analyzed further in the GloWbE corpora. Finally, teas does not occur in the West African corpora. This is also unexpected, given that it is a grammatical form as tea is a flexible noun. In terms of differences between the corpora, there are several nouns that are frequent in one or two varieties but not in the third one. The first one is beers, which is frequent in ICE-EA and ICE-Ghana but does not occur in ICE-Nigeria. Both differences are statistically significant at the 95 % confidence level in a chisquare test. A high frequency of beers was to be expected as it is a flexible term that can be pluralized. It is thus rather interesting that it does not occur in Nigeria. Offsprings is frequent in the East African and Nigerian data but does not occur in Ghana. The fact that it does not occur in ICE-Ghana at all while it can be found in the other ICE components and shows a considerable frequency in Nigeria has to be taken with a grain of salt as only the written component of ICE-Ghana, and of this only a preliminary version could be used. It remains to be investigated whether the results will remain the same once the spoken data are available. Statistically, the differences shown here are not significant. However, the differences between the frequencies of researches, equipments, stones and staffs in the individual corpora are significant, with equipments and stones showing very high statistical significances. This shows that although the frequency of the plural form of equipments is high in general, there are considerable differences in the use of the form between the different African varieties. Further, it emphasizes the differences in the use of flexible terms such as stone(s). In a second step, the GloWbE data were analyzed. The frequencies, again displaying the occurrence of the plural form as compared to the total number of occurrences of the noun in the respective corpus are shown in Table 3. Tab. 3: Frequencies of the pluralized mass terms in the African GloWbE components

researches informations equipments evidences advices

GloWbE Kenya

GloWbE Tanzania

GloWbE Nigeria GloWbE Ghana

0.6 % 0.1 % 3.8 % 0.9 % 1.7 %

1.3 % 0.2 % 5.4 % 2.1 % 1.6 %

1.8 % 0.3 % 7.2 % 3.4 % 1.7 %

1.0 % 0.1 % 5.2 % 1.8 % 1.2 %

The use of pluralized mass nouns in East and West African postcolonial Englishes | 173

furnitures knowledges musics traffics homeworks cattles luggages silverwares ketchups beers stones toothpastes teas strings staffs offsprings cheeses

GloWbE Kenya

GloWbE Tanzania

GloWbE Nigeria GloWbE Ghana

0.5 % 0 %13 0.1 % 0.2 % 1.0 % 0.03 %15 0.8 % 1.8 % 0 % 15.1 % 36.3 % 1.5 % 2.0 % 30.6 % 1.1 % 3.5 % 7.9 %

0.5 % 0.1 % 0.2 % 0.2 % 0.6 % 0.1 % 0.6 % 0% 0% 21.5 % 17.0 % 4.3 % 3.2 % 25.4 % 1.5 % 6.5 % 8.4 %

1.4 % 0.1 % 0.3 % 0.5 % 0% 1.0 % 2.6 % 1.5 % 0% 6.4 % 29.1 % 4.6 % 5.3 % 35.5 % 2.2 % 13.3 % 5.6 %

0.9 % 0 %14 1.0 % 0% 0.8 % 0.3 % 1.3 % 4.3 % 4.4 % 15.7 % 30.2 % 6.9 % 3.2 % 38.5 % 1.9 % 2.3 % 15.1 %

As shown in Table 3, the results for the individual GloWbE components seem relatively similar to each other. In general, almost all forms are attested, as opposed to the ICE components where many forms could not be found (cf. Table 2). However, there are certain forms that are very rare even in GloWbE, such as informations, furnitures, musics, traffics, homeworks, cattles and ketchups. Ketchups seems to form an exception though as it does not occur at all in the Kenyan, Tanzanian and Nigerian component, but in the Ghanaian one.16 This, as mentioned above, is probably due to the fact that it is a very non-African concept. Stones and strings show the highest frequencies in all components; it has to be borne in mind though that these are flexible nouns. Equipments is the only form that is relatively frequent (percentages of roughly 4 % or more in all varieties) in all components. This is similar to the results obtained for the ICE data. Toothpastes is also frequent in all components but the Kenyan one, as is beers in all corpora but the Nigerian one. The latter is again similar to the situation observed in the ICE corpora (cf. Table 2), and, similarly, the difference is not statistically significant. Three items are exceptional in the Ghanaian component in that they are more frequent than in any of the other varieties. These are cheeses, silverwares and ketchups. || 13 The actual frequency is 0.02 % (2 in 8,2014) which was rounded to 0 %. 14 The actual frequency is 0.03 % (2 in 7,718) which was rounded to 0 %. 15 The actual frequency is 0.03 % (1 in3,773) which was rounded to 0 %. 16 However, it only occurs once (which accounts for 4.4 % of the tokens).

174 | Susanne Mohr

In comparison, most differences between the individual GloWbE components were statistically significant at the 95 % level, except for furnitures, homeworks, luggages, silverwares, ketchups and toothpastes. This is not surprising given that these terms are (equally) infrequent in the different components. Three of the terms that showed interesting results for the ICE corpora deserve a second look in GloWbE. These were cattles, offsprings and beers. While cattles did not occur at all in ICE-Nigeria and ICE-Ghana, it could be found in the East African component. In the GloWbE data the results are more consistent: the noun shows similarly low frequencies for all components. Offsprings did not occur at all in ICE-Ghana but was reported to be frequent in the East African and Nigerian data. In GloWbE, it is found in Ghana with a relatively low frequency that is similar to the frequencies of the term in Kenya and Tanzania. Hence, the non-occurrence of the term in ICE-Ghana might be due to the preliminary nature of the corpus. Finally, beers was peculiar in ICE in that it did not occur in the Nigerian data. In the GloWbe data it can be found in Nigerian English, however, it is noteworthy that it is much less frequent than in the other varieties (cf. Table 3). There seems to be no straightforward reason for this fact, the only factor that might have an influence in this respect is that in Hausa, the term giya ‘beer’ is transnumeral.17

4.2 Mass terms in context After the analysis of the frequencies of the mass terms, a few examples of the contexts in which they were found seem in order. First, examples of the use of equipments, a noun that was frequent in both the ICE and the GloWbE corpora are given. (11)

We have pleasure in introducing ourselves as one of the leading suppliers of Sports Equipments, Sports Wear and Sports Trophies. ICE-EA, Kenya (W1B-BK24)

(12)

For example cooking utensils and even agricultural equipments such as plough instead of tractor. ICE-EA, Tanzania (W1A019T)

(13)

[…] for us to engage with the process through which these equipments get recovered from the ports then match them to project areas where there is demand. GloWbE Nigeria

|| 17 I would like to thank my reviewer for this suggestion.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 175

(14)

[…] give him a contract in million [sic!] of dollars and then tell him to keep all these equipments, still worth millions of dollars after five years for free? GloWbE Ghana

The above examples are interesting in many respects. Example (11) shows that the pluralised form is even used in advertising. This could be an indication of the general acceptance of the form as it is usually avoided to use nongrammatical language in official documents. Sentence (12) is interesting as the minimal parts of the equipment, namely a plough or tractor are listed. This clearly shows that the concept of equipment is perceived to be individualized with different parts that form the whole. Example (13) is intriguing in that it depicts the use of a demonstrative pronoun with the plural form. The pronoun is also a plural form (these) and the verb is also plural, which emphasizes the conceptualization of the term as a plural referent. This is proven further by (14), in which the plural form is also preceded by a plural pronoun, and the quantifier all. This also indicates an individualized (countable) conceptualisation of the noun, as only countable nouns can be modified by all. Two further examples of the noun offsprings that was frequent in the ICE and GloWbE components alike are provided below. (15)

May be Tet as all those offsprings of Mumbi has been susy chopping the real boy’s money with some lover boy so the boy wanted to get even! GloWbE Kenya

(16)

On the other hand, if you (the State) do not want to be disturbed further by this 5 Naira offspring of yours for resource allocation, you declare that henceforth, all of your six offsprings should go and manage themselves with whatever they have in their domains. GloWbE Nigeria

As can be seen in (15), the plural form of offspring is also used with a plural demonstrative pronoun as is equipments in (13) and (14). Interestingly, it is followed by a singular verb form. This is in line with Bock et al. (2006) who found that verb agreement in collective mass nouns is less affected than pronoun number. Example (15) shows not only one token of the mass noun offspring but two. However, the first instance of the noun is a standard singular form, while the second one is pluralized. This is interesting as it might indicate a not (yet) stabilized number system for this noun. A similar example can be found in ICEEA for cattle: (17)

Yesterday, the President said that some cattles were slashed all over their bodies. However, can you compare cattle with people? If cattle into your garden […] ICE-EA, Kenya (S1B051HK)

176 | Susanne Mohr

Further instances of nouns that were used in singular and plural form by one speaker/writer could be found in all of the other corpora. This is further evidence of a non-stabilized number system. Finally, two further uses should be exemplified. The first one is luggages in ICE-Nigeria, which is the only ICE component that contained any instances of the plural form. An interesting case is shown in (18): (18)

Another humour can be seen in the case of nicola (the man servant) who was embarrased by Catherine when he brought the luggages of bluntschili and it was seen by Mayor Petkoff, that was why Catherine was angry and she embarrassed him. ICE-Nigeria (W1A, ex_49)

The sentence provides an example of the pluralized luggages that probably refers to several suitcases. Apart from that, the example is interesting because it starts with another mass noun, humor. Although this was not part of the items under analysis in this paper, it should be considered here. Although the noun is not pluralized, it is preceded by another, which normally cannot be combined with mass nouns, especially if abstract. A sentence like *he acquired another knowledge is not grammatical, he gave me another gold containing a concrete noun is odd. Only certain mass terms can be combined with another in certain contexts, such as I would like another water which, however, entails a portion reading. Hence, in (18) above, humor is clearly conceptualized as a countable noun which it is not in Standard English. An example of silverwares in Ghanaian English is (19). (19)

Its a season where we will have to defend our UEFA Champions League title, and also the FA Cup as we also eye to win other Silverwares, something which has never been easy for any club as far as European football is concerned. GloWbE Ghana

This sentence is also interesting in two respects: first of all, it shows an instance of silverwares in GloWbE Ghana, where it was most frequent. Secondly, it shows a different meaning of the term as compared to the singular silverware (meaning ‘cutlery’) in Standard English. This meaning is along the lines of ‘sports trophy’, especially in a football context. This is the same as the meaning of the term that has been reported for Kenyan English by Mohr (2014). In GloWbE Ghana this is, however, not as consistent as in Kenyan English. Another example of the use of silverwares in Ghanaian English is (20): (20)

Fortunately, it is proved that Tiffany jewelry can express the voices of lovers and its original silverwares and tableware are fascinated. GloWbE Ghana

The use of pluralized mass nouns in East and West African postcolonial Englishes | 177

In the above context the noun is probably used in its standard British English meaning. Informal conversation with participants from Ghana during data collection confirmed that the new meaning of the noun (‘sports trophy’) is by no means established in Ghanaian English. After having elaborated on the results of the corpus analysis with respect to frequencies of the plural mass nouns and contexts of use, they are further discussed in the next section.

5 Discussion As was shown in the previous section, many of the 22 mass nouns that were investigated here are indeed used in corpora of Kenyan, Tanzanian, Nigerian and Ghanaian English. Some of the plural forms, such as equipments or offsprings, could in fact be described as “frequent”. These are forms that fall into the category of object-mass nouns and are subject to the aforementioned mismatch of grammar and semantics (cf. sections 2.2 and 2.3). Classical substancemass nouns were rarely used in plural form, which is further evidence of the importance of the substance-mass/object-mass distinction for the pluralization of mass terms. Sample sentences from the corpora also showed that several of the object-mass nouns are conceptualized as consisting of individual items, as in (12) where different items of agricultural equipment are mentioned. This is in line with Mohr (2014) who showed that the perception and conceptualization of mass terms as individuated among African speakers of English has an impact on the frequency of the use of pluralized mass nouns. This is further confirmed by the use of plural demonstratives with these nouns as in examples (13), (14) and (15), showing the conceptual perception of these nouns as (collective) individuated and thus countable entities. However, the examples also showed several instances of the use of both the singular and plural form in one utterance or by one speaker (examples (16) and (17)). This might be an indication of a speakers’ insecurity concerning the correct form of these nouns and the non-conventionality of the plural forms. This is especially evident given that almost all plural forms are used less often than the singular (standard) forms of the mass terms. The forms that showed a frequency of 10 % or higher were very often grammatical in Standard English as they are flexible terms like stones or strings. These results entail the question of the applicability of the frequency ratings in WAVE. As mentioned above, the use of plural mass nouns is rated A (“pervasive or obligatory”) in Kenyan and Nigerian English and B (“neither pervasive nor extremely rare”) in Tanzanian and Ghana-

178 | Susanne Mohr

ian English. Mohr (2014) already indicated that the A rating for Kenya seems questionable. This is confirmed by the present analysis. Further, the A rating of the feature for Nigeria seems controversial given the current results. A feature that shows small frequencies like the ones outlined in section 4 does not seem to be “pervasive or obligatory”. A rating as “neither pervasive nor extremely rare” (B) could be more fitting. Another factor that has to be kept in mind though is the relative frequency of these forms as compared to Standard English. For this, a comparison with the BNC data was revealing. Table 4 below shows the frequencies of the terms in the BNC-BYU. Tab. 4: Frequencies of the pluralized mass terms in the BNC-BYU BNC-BYU researches informations equipments evidences advices furnitures knowledges musics traffics homeworks cattles luggages silverwares ketchups beers stones toothpastes teas strings staffs offsprings cheeses

0.9 % 0.04 %18 0.3 % 0.2 % 0.1 % 0.03 % 0.2 % 0.2 % 0.1 % 1.1 % 0% 0% 0% 2.4 % 12.6 % 28.9 % 2.9 % 3.0 % 34.2 % 1.0 % 0.4 % 13.3 %

The fact that seems most evident from the above table is that although the flexible terms stones, strings and beers are frequent in that they show percentages higher than 10 %, other (object) mass terms are much less often pluralized. Further, teas is much less frequent than pluralized forms of the other flexible || 18 The actual frequency was 0.04 % (15 in 37,877) which was rounded to 0 %.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 179

nouns. Instead, cheeses is used unexpectedly often. Mostly, the plural is used with a “type-of” meaning, as in (21). (21)

More than half the 12.5 billion kilograms of milk produced annually in the Netherlands today is sold abroad and, of course, Dutch cheeses are renowned as well. BNC-BYU

Interestingly, cheeses is only equally frequent in GloWbE Ghana (cf. Table 3) and even less frequent in the other corpora. Disregarding the flexible terms and focusing on the mass nouns that are non-grammatical in the plural in Standard English, it is evident that some of them occur less frequently in the BNC data than in the African data. Regarding the ICE components, only evidences shows statistically significant results for all three components at the 95 % confidence level though. Equipments and offsprings are significantly more frequent in ICE-EA and ICE-Nigeria than in the BNC, while researches and advices are in ICE-EA, and luggages and staffs in ICENigeria. Concerning the GloWbE corpora, equipments is also significantly more frequent in all components than in the BNC data, as are evidences, advices and luggages. Furnitures and staffs are significantly more frequent in all components but the Kenyan one, as are silverwares in all but the Tanzanian one, and offsprings in all but the Ghanaian component. Researches is significantly more frequent than in the BNC in Tanzania and Nigeria, as is cattles in Nigeria. Interestingly, homeworks is statistically significant in Nigeria too, only it is less frequent in the Nigerian data as compared to the BNC. The non-standard form is hence more frequent in the British data, which might be due to the fact that homeworks is a common “mistake” mentioned to many English language learners. The Nigerian speakers might hence be very attentive of that particular form. Altogether, the forms that are significantly more frequent in the African data as compared to the British corpus are all object-mass terms (with advice as a borderline case). Substance-mass nouns like toothpastes and abstract terms like knowledges do not show significant differences to the British data. This further proves the hypothesis that it is mainly object-mass terms that are problematic for non-native speakers, which might in turn be caused by the aforementioned mismatch of grammar and semantics. Finally, a comparison of the use of the plural mass nouns in the different African components is important for the question of whether this feature might ultimately qualify as an areoversal of African Englishes as mentioned by Huber (2012). Kortmann (2013) mentions geography can be expected to play a role for the degree to which a certain aspect of a variety’s morphosyntax is borne out. It has to be kept in mind though that this paper only considers West and East

180 | Susanne Mohr

African Englishes, while South African data are not analyzed. Thus, all conclusions remain tentative. As shown in section 4, there are certain similarities between the different corpora such as the high frequency of use of equipments or the different semantics of silverwares in Ghanaian and Kenyan English, referring to ‘sports trophies’. Moreover, forms like luggages or evidences are all significantly more frequent in the African data as compared to the BNC. However, considerable (statistical) differences were also found between the different African corpora. In the ICE data, four nouns showed statistically significant differences between the three components, for GloWbE, 16 of the 22 investigated terms were significantly different in use. This shows that on the one hand, there are certain similar tendencies concerning the use of plural mass nouns in the West and East African data. On the other hand, the results are rather divergent across the individual varieties, which hint at the individual “character” of each of these varieties. The use of plural mass nouns seems to be borne out differently in the different African varieties – a claim in line with Schneider (2007), who emphasizes pragmatic and stylistic constraints of a feature play an important role in the formation of a postcolonial English. Irrespective of the similarities and differences between the individual African varieties, it can be confirmed that plural mass nouns do occur in Africa. However, according to WAVE, this is also the case in many other New Englishes, such as varieties spoken in former Asian colonies. In order to qualify as a feature that is diagnostic of a certain world region, its use should be rare or unattested in other world regions (Kortmann & Schröter forthcoming). Hence, it seems that F 55 could rather qualify as a varioversal of postcolonial Englishes.

6 Conclusion The present paper provided an insight into the frequency and use of pluralized mass nouns in four African English varieties (Kenyan, Tanzanian, Nigerian and Ghanaian English) as compared to Standard (British) English. The use of these plural mass nouns has been cited as a common feature in postcolonial Englishes in general and in African Englishes in particular. The corpus analysis showed that the frequency of these nouns in the African data from the ICE and GloWbE components is in fact higher than in British English as exemplified by the BNCBYU. However, there are considerable differences in the use of several of the mass nouns in the individual African varieties. Furthermore, the frequency ratings indicated in WAVE seem inappropriate for Kenyan and Nigerian English, as the pluralization of mass nouns seems

The use of pluralized mass nouns in East and West African postcolonial Englishes | 181

neither pervasive nor obligatory for the nouns investigated. A change of the frequency ratings could hence be considered. A more detailed analysis into the dynamics of the usage of the plural forms revealed that it is mainly object-mass terms that are used differently by nonnative speakers, a result confirming the hypothesis that it is a mismatch of grammar and semantics that is problematic in this respect. Results of the study on mass nouns in East African Englishes (Mohr 2014) could be further verified. As Kortmann & Schröter (forthcoming) have stressed, areality only seems to be of secondary importance when it comes to the distribution of features in different varieties of English. What is more important in this respect is variety type, i.e. L2 English in the cases under scrutiny here. Although the difficulties with a categorization of African Englishes according to order of acquisition as implied by the terminology “L2” have been discussed here, a summary of all African Englishes into one variety type could be supported. The sociolinguistics of the development of the varieties, as well as their contexts of use, i.e. multilingual spaces, are similar. Hence, a category of “African postcolonial Englishes” is valid. Possibly, the factor of variety type influences the use of plural mass nouns as well. Another unresolved question in this regard is the minimum frequency of a certain feature in the varieties from one world region in order to qualify for areoversality. Kortmann & Schröter (forthcoming) ignored frequencies and only considered the presence of a feature in a certain variety, while Brato & Huber (2012) and Huber (2012) did consider frequency as a criterion in their analysis. The current paper cannot answer this question conclusively either, but it can and has provided further information on the actual frequency of F 55 in four varieties of African English. This should definitely be taken into account for future typological studies. The present analysis of mass nouns in East and West African Englishes is far from comprehensive in that no South African Englishes were included, which would be an interesting future endeavor. Moreover, a list of only 22 mass nouns was considered, which is obviously not complete either. Many of them refer to non-African concepts which should be reconsidered. In the future it would be interesting to investigate more or possibly all mass nouns included in a corpus. Schneider & Schreier (2014) introduced a preliminary version of a possible method to extract all mass nouns from a corpus. Once this method is refined it would be ideal to apply it to the African data of ICE and GloWbE. Finally, the dynamics and contexts of the plural forms remain an interesting issue for research. The use of plural demonstratives with the forms could be shown here and it would be interesting to investigate whether certain patterns of use

182 | Susanne Mohr

can be made out, also with respect to singular and plural verbs as has been studied by Bock et al. (2006) for British and American English. Ultimately, it would be intriguing to see whether the use of plural mass nouns is similar in other New Englishes in Asia, which seems likely given that variety type plays a more important role for the distribution of features in Englishes around the world. Many evidences and informations remain to be collected and analyzed.

References Aikhenvald, Alexandra Y. 2003. Classifiers: A typology of noun categorization devices. Oxford: Oxford University Press. Anchimbe, Eric A. 2007. Multilingualism, postcolonialism and linguistic identity: Towards a new vision of postcolonial spaces. In Anchimbe, Eric A. (ed.) Linguistic identity in postcolonial multilingual spaces, 1–19. Newcastle: Cambridge Scholars Publishing. Barner, David & Jesse Snedeker. 2005. Quantity judgements and individuation: Evidence that mass nouns count. Cognition 97. 41–66. Baskaran, Loga. 2008. Malaysian English: Morphology and syntax. In Mesthrie, Rajend (ed.) Varieties of English 4: Africa, South and Southeast Asia, 610–623. Berlin & New York: De Gruyter Mouton. Blommaert, Jan, James Collins & Stef Slembrouk. 2005. Spaces of multilingualism. Language & Communication 25. 197–216. Bock, Kathryn, Anne Cutler, Kathleen M. Eberhard, Sally Butterfield, J. Cooper Cutting & Karin R. Humphreys. 2006. Number agreement in British and American English: Disagreeing to agree collectively. Language 82(1). 64–113. Borer, Hagit. 2005. In name only. Structuring sense. Vol. 1. Oxford: Oxford University Press. Brato, Thorsten & Magnus Huber. 2012. English in Africa. In Raymond Hickey (ed.), Areal features of the Anglophone world, 161–185. Berlin/New York: De Gruyter Mouton. Bunt, Harry C. 1985. Mass terms and model-theoretic semantics. Cambridge: Cambridge University Press. Cheng, Lisa L. & Tilo Schadeberg. 2007. On count and mass in Bantu. Paper presented at the first meeting on Bantu, Chinese and Romance nouns and noun phrases. SOAS, London. Chierchia, Gennaro. 1998. Reference to kinds across languages. Natural Language Semantics 6. 339–405. Chierchia, Gennaro. 2010. Mass nouns, vagueness and semantic variation. Synthese 174. 99– 149. Corbett, Greville G. 2001. Number. Cambridge: Cambridge University Press. Cowper, Elizabeth & Daniel Currie Hall. 2012. Aspects of individuation. In Diane Massam (ed.), Count and mass across languages, 27–53, Oxford: Oxford University Press. Creider, Chet A. 1975. The semantic system of noun classes in Proto-Bantu. Anthropological Linguistics 17. 127–138. Davies, Mark. 2013. Corpus of Global Web-Based English: 1.9 billion words from speakers in 20 countries. Available online at http://corpus2.byu.edu/glowbe, last accessed 30-12-2014.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 183

De Bot, Kees, Wander Lowie & Marjolijn Verspoor. 2005. Second language acquisition. Abingdon: Routledge. Denny, J. Peter & Chet A. Creider. 1976. The semantics of noun classes in Proto-Bantu. Studies in African Linguistics 7(1). 1–30. Dixon, Robert M. W. 1968. Noun classes. Lingua 21. 104–125. Ellis, Rod. 1985. Understanding second language acquisition. Oxford: Oxford University Press. Faraclas, Nicholas F. 1996. Nigerian Pidgin. London: Routledge. Fishman, Joshua A., Andrew W. Conrad & Alma Rubal-Lopez (eds.). 1996. Post-imperial English. Status change in former British and American colonies, 1940–1990. Berlin: Mouton de Gruyter. Gillon, Brendan S. 1996. The lexical semantics of English count and mass nouns. Papers presented at the workshop on the breadth and depth of Semantic lexicons, Santa Cruz. Givón, Talmy. 1971. Some historical changes in the noun class system of Bantu, their possible causes and wider applications. In Chin-Wu Kim & Herbert Stahlke (eds.), Papers in African Linguistics, 33–54. Edmonton: Linguistic Research. Gut, Ulrike. 2013. The international corpus of English – Nigeria. Münster. Available online at http://sourceforge.net/projects/ice-nigeria/. Heine, Bernd & Derek Nurse. 2000. African languages. An introduction. Cambridge: Cambridge University Press. Huber, Magnus. 1999. Ghanaian Pidgin English in its West African context. Amsterdam: John Benjamins. Huber, Magnus. 2008a. Ghanaian English: Phonology. In Rajend Mesthrie (ed.), Varieties of English. Africa, South and Southeast Asia, 67–92. Berlin/New York: Mouton de Gruyter. Huber, Magnus. 2008b. Ghanaian Pidgin English: Phonology. In Rajend Mesthrie (ed.), Varieties of English. Africa, South and Southeast Asia. 93–101. Berlin/New York: Mouton de Gruyter. Huber, Magnus. 2012. Regional profile: Africa. In Bernd Kortmann (ed.), The Mouton atlas of variation in English: Grammar, 806–823. Berlin/New York: Mouton de Gruyter. Huber, Magnus. forthcoming. The international corpus of English – Ghana (Preliminary version of April 2014). Gießen. Huntley-Fenner, Gavin, Susan Carey & Andrea Solimando. 2002. Objects are individuals but stuff doesn’t count: Perceived rigidity and cohesiveness influence infants’ representations of small groups of discrete entities. Cognition 8. 203–221. Imai, Mutsumi & Dedre Gentner. 1997. A cross-linguistic study of early word meaning: Universal ontology and linguistic influence. Cognition 62. 169–200. Imai, Mutsumi & Reiko Mazuka. 2003. Re-evaluation of linguistic relativity: Language-specific categories and the role of universal ontological knowledge in the construal of individuation. In Dedre Gentner & Susan Goldin-Meadow (eds.), Language in mind: Advances in the issues of language and thought, 430–464, Cambridge, MA: MIT Press. Imai, Mutsumi & Reiko Mazuka. 2007. Revisiting language universals and linguistic relativity: Language-relative construal of individuation constrained by universal ontology. Cognitive Science 31. 385–414. Inagaki, Shunji. 2014. Syntax-semantics mappings as a source of difficulty in Japanese speakers’ acquisition of the mass-count distinction in English. Bilingualism: Language and Cognition 17. 464–477. Kachru, Braj B., Yamuna Kachru & Cecil Nelson (eds.). 2006. The handbook of world Englishes. Basingstoke: Blackwell.

184 | Susanne Mohr

Kortmann, Bernd. 2013. How powerful is geography as an explanatory factor in morphosyntactic variation? Areal features in the Anglophone world. In Peter Auer, Martin Hilpert, Anja Stukenbrok & Benedikt Szmrecsanyi (eds.), Space in language and linguistics: Geographical, interactional and cognitive perspectives, 165–194, Berlin/New York: Mouton de Gruyter. Kortmann, Bernd. 2015. Looking up how grammars vary across the world: The WAVE perspective. Paper presented at the University of Bonn. Bonn. Kortmann, Bernd & Benedikt Szmrecsanyi. 2004. Global synopsis – morphological and syntactic variation in English. In Bernd Kortmann, Edgar W. Schneider, Kate Burridge, Rajend Mesthrie & Clive Upton (eds.), A handbook of varieties of English, vol. 2: Morphology and syntax, 1122–1182. Berlin/New York: Mouton de Gruyter. Kortmann, Bernd & Kerstin Lunkenheimer (eds.). 2012. The Mouton atlas of variation in English. Berlin/New York: Mouton de Gruyter. Kortmann, Bernd & Kerstin Lunkenheimer (eds.). 2013. The electronic world atlas of varieties of English. Leipzig: Max Planck Institute for Evolutionary Anthropology. Available online at http://ewave-atlas.org, last accessed 10-10-2014. Kortmann, Bernd & Verena Schröter. forthcoming. Varieties of English. In Raymond Hickey (ed.), The Cambridge handbook of areal linguistics. Cambridge: Cambridge University Press. Kortmann, Bernd, Edgar W. Schneider, Kate Burridge, Rajend Mesthrie & Clive Upton (eds.). 2004. A handbook of varieties of English, Vol. 2: Morphology and syntax. Berlin/New York: Mouton de Gruyter. Lewis, M. Paul, Gary F. Simons & Charles D. Fennig (eds.). 2014. Ethnologue: Languages of the world. Eighteenth edition. Dallas, Texas: SIL International. Available online at http://www.ethnologue.com, last accessed 30-12-2014. Li, Peggy, Yarrow Dunham & Susan Carey. 2009. Of substance: The nature of language effects on entity construal. Cognitive Psychology 58. 487–524. Lucy, John A. & Suzanne Gaskins. 2001. Grammatical categories and the development of classification preferences: A comparative approach. In Melissa Bowerman & Stephen C. Levinson (eds.), Language acquisition and conceptual development, 257–283, Cambridge: Cambridge University Press. Lucy, John A. & Suzanne Gaskins. 2003. Interaction of language type and referent in the development of nonverbal classification preferences. In Dedre Gentner & Susan GoldinMeadow (eds.), Language in mind: Advances in the issues of language and thought, 465– 492, Cambridge, MA: MIT Press. Lunkenheimer, Kerstin. 2012. Typological profile: L2 varieties. In Bernd Kortmann & Kerstin Lunkenheimer (eds.), The Mouton world atlas of variation in English: Grammar, 845–872, Berlin/New York: Mouton de Gruyter. Massam, Diane (ed.). 2012. Count and mass across languages. Oxford: Oxford University Press. Moag, Rodney F. 1992. The life-cycle of non-native Englishes: A case study. In Braj B. Kachru (ed.), The other tongue: English across cultures, 270–288, Oxford: Pergamon. Mohr, Susanne. 2014. Informations on researches and knowledges – Mass nouns in East African Englishes. Paper presented at the 3rd Bonn Applied English Linguistics Conference (BAELc3), Bonn. Pelletier, Francis J. 1975. Non-singular reference: Some preliminaries. Philosophia 5. 451–465. Platt, John T., Heidi Weber & Ho Mian Lian. 1984. The New Englishes. London: Routledge & Kegan Paul. Quine, Willard. 1960. Word and object. Cambridge, MA: MIT Press.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 185

Quirk, Randolph, Signey Greenbaum, Geoffrey Leech & Jan Svartvik (eds.). 1985. A comprehensive grammar of the English language. London: Longman. Richardson, Irvine. 1967. Linguistic evolution and Bantu noun class systems. Paris: Centre Nationale de Recherche Scientifique. Rijkhoff, Jan. 1991. Nominal aspect. Journal of Semantics 8. 291–309. Romaine, Suzanne. 2000. Language in society. An introduction to sociolinguistics. Oxford: Oxford University Press. Schmied, Josef. 2006. East African Englishes. In Braj B. Kachru, Yamuna Kachru & Cecil Nelson (eds.), The handbook of world Englishes, 188–202. Basingstoke: Blackwell. Schmied, Josef. 2008. East African English (Kenya, Uganda, Tanzania): Morphology and syntax. In Rajend R. Mesthrie (ed.), Varieties of English. Africa, South and Southeast Asia, 451– 471. Berlin/New York: Mouton de Gruyter. Schmied, Josef. 2012. Standards of English in East Africa. In Raymond Hickey (ed.), Standards of English, 229–255. Cambridge: Cambridge University Press. Schmied, Josef, Diana Hudson-Ettle & Barbara Krohne. 1999. The international corpus of English – East Africa. Bayreuth/Chemnitz. Schneider, Edgar W. 2003. The dynamics of New Englishes: From identity construction to dialect birth. Language 79(2). 233–281. Schneider, Edgar W. 2007. Postcolonial English: Varieties around the world. Cambridge: Cambridge University Press. Schneider, Gerold & Daniel Schreier. 2014. Mass nouns in English: Automatic detection and regional variation. Paper presented at the ICAME 35 conference, Nottingham. Skandera, Paul. 2003. Drawing a map of Africa. Idiom in Kenyan English. Tübingen: Gunter Narr. Storch, Anne & Gerrit J. Dimmendaal (eds.). 2014. Number – Constructions and semantics. Amsterdam/Philadelphia: John Benjamins. Szmrecsanyi, Benedikt & Bernd Kortmann. 2009. Vernacular universals and angloversals in a typological perspective. In Markku Filppula, Juhani Klemola & Heli Paulasto (eds.), Vernacular universals and language contacts, 33–53, New York/London: Routledge. Szmrecsanyi, Benedikt & Bernd Kortmann. 2011. Typological profiling. In Joybrato Mukherjee & Marianne Hundt (eds.), Exploring second-language varieties of English and Learner Englishes, 167–187. Amsterdam/Philadelphia: John Benjamins. Webb, Vic & Kembo-Sure. 2001. African voices. An introduction to the languages and linguistics of Africa. Oxford: Oxford University Press. Wiese, Heike. 2012. Collectives in the intersection of mass and count nouns: A cross-linguistic account. In Diane Massam (ed.), Count and mass across languages, 54–74, Oxford: Oxford University Press. Wiltschko, Martina. 2012. Decomposing the mass/count distinction: Evidence from languages that lack it. In Diane Massam (eds.), Count and mass across languages, 146–171, Oxford: Oxford University Press. Zerbian, Sabine & Manfred Krifka. 2008. Quantification across Bantu languages. In Misa Matthewson (ed.), Quantification: A cross-linguistic perspective, 383–415. Bingley: Emerald.

186 | Susanne Mohr

Appendix Tab. 5: Frequencies of the pluralized mass terms in the African ICE components, normalized to 100,000 words

researches informations equipments evidences advices furnitures knowledges musics traffics homeworks cattles luggages silverwares ketchups beers stones toothpastes teas strings staffs offsprings cheeses

ICE-EA 1.09/58.69 0.20/67.74 2.19/18.20 0.50/48.94 0.10/11.64 0/1.90 0/37.4 0/55.90 0/8.55 0/0.11 0.10/17.61 0/1.39 0/0 0/0 0.80/10.24 2.88/4.97 0/0.20 0.10/9.95 0.50/1.00 0.10/34.61 0.40/0.90 0/0.80

ICE-Nigeria 0.59/18.11 0.40/34.94 0.30/7.03 0.59/8.51 0.10/5.54 0/0.30 0/21.38 0/8.12 0/7.22 0/0.10 0/0.10 0.20/0.40 0/0.10 0/0.69 0/1.19 0.50/1.68 0/0.10 0/3.86 0.49/1.68 0.59/20.68 0.10/0.10 0/0.50

ICE-Ghana 0/48.16 0.47/69.75 0.24/4.98 0.47/13.76 0/12.57 0/1.90 0/36.54 0/20.17 0/8.78 0/1.66 0/2.13 0/0.24 0/0 0/0 0.71/3.32 2.61/7.35 0/0.71 0/1.19 0.95/3.32 0/20.17 0/1.19 0/0.71

Tab. 6: Frequencies of the pluralized mass terms in the African GloWbE components, normalized to 100,000 words

researches informations equipments evidences advices furnitures knowledges

GloWbE Kenya 0.22/36.25 0.07/56.35 0.32/8.32 0.09/12.03 0.15/9.00 0/1.6019 0/20.00

GloWbE Tanzania 0.50/37.58 0.14/59.26 0.63/11.66 0.30/14.15 0.14/9.20 0.01/2.70 0.04/28.14

GloWbE Nigeria 0.44/24.29 0.14/42.25 0.44/6.13 0.36/10.56 0.20/11.52 0.01/1.02 0.02/21.47

GloWbE Ghana 0.28/28.31 0.06/48.85 0.44/8.51 0.27/12.17 0.11/8.86 0.01/1.19 0/19.91

|| 19 The actual frequency of 0 numbers in italics was either 1 or 2 tokens. However, when normalized and rounded, these result in 0.

The use of pluralized mass nouns in East and West African postcolonial Englishes | 187

musics traffics homeworks cattles luggages silverwares ketchups beers stones toothpastes teas strings staffs offsprings cheeses

GloWbE Kenya 0.02/20.18 0.02/9.21 0.01/1.00 0/9.19 0.01/1.21 0.01/0.28 0/0.06 0.44/2.91 2.13/5.86 0/0.16 0.17/8.27 0.64/2.08 0.22/19.79 0.02/0.70 0.07/0.92

GloWbE Tanzania 0.04/21.94 0.02/7.49 0/1.02 0/4.61 0.01/2.57 0/0.12 0/0.06 1.05/4.90 2.77/16.28 0/0.13 0.19/5.97 0.46/1.81 0.41/26.88 0.04/0.70 0.09/1.08

GloWbE Nigeria 0.07/26.47 0.04/9.30 0/0.60 0.02/1.59 0.02/0.63 0/0.27 0/0.03 0.09/1.47 1.37/4.72 0.01/0.15 0.11/2.13 0.63/1.78 0.33/15.18 0.07/0.55 0.03/0.51

GloWbE Ghana 0.03/35.67 0/8.05 0/0.68 0/1.02 0.01/0.79 0.01/0.30 0/0.06 0.29/1.83 1.46/4.83 0.01/0.19 0.08/2.42 0.64/1.68 0.33/17.47 0.01/0.45 0.11/0.73

Tab. 7: Frequencies of the pluralized mass terms in the BNC-BYU, normalized to 100,000 words

researches informations equipments evidences advices furnitures knowledges musics traffics homeworks cattles luggages silverwares ketchups beers stones toothpastes teas strings staffs offsprings cheeses

BNC-BYU 0.25/27.26 0.01/38.51 0.03/8.89 0.04/21.38 0.01/10.38 0/3.4820 0.02/14.45 0.03/14.70 0.01/6.54 0.01/0.82 0/2.56 0/0.63 0/0.06 0/0.09 0.46/3.65 3.22/11.13 0.01/0.21 0.25/8.32 1.36/3.98 0.23/22.88 0/0.95 0.39/2.94

|| 20 The actual frequency of 0 numbers was either 1 or 2 tokens. However, when normalized and rounded, these result in 0.

Astrid Alexander-Bakkerus

A colonial grammar or arte Abstract: The manuscripts Add. 25,323 and Add. 25,324 from the British Library concern a description of Jebero, a moribund language spoken in North Peru. The former manuscript contains a Spanish-Quechua-Jebero vocabulary, the latter a Christian doctrine in Jebero and in Quechua and a grammar of Jebero. The codices date from the 18th century. They were not meant to be published. The author, whose name is not given, wrote them for his fellow missionaries. In this paper I give a number of features of colonial grammars/artes, and show why we can consider the manuscripts Add. 25,323 and Add. 25,324 as a whole and why we can assume that, as such, they constitute a colonial grammar. Subsequently we pay attention to the Christian doctrine, the most important piece of both codices and the reason why the other parts, i.e. the trilingual vocabulary and the grammar of Jebero, were written. Keywords: Jebero, colonial grammars, Christian doctrine

1 Introduction Notwithstanding the fact that the name of the author of the codices Add. 25,323 and Add. 25,324 was omitted, there are reasons to believe that the manuscripts were written by the Jesuit missionary, Samuel Fritz (1654–1726). Samuel Fritz explored the basin of the Amazon River, made a chart of the Marañón territory and contacted the Jebero people in 1694. The manuscript Add. 25,323 comprises a word list, called Vocabulario dela Lengua Castellana, la del Ynga, y Xebera ‘Vocabulary of the Spanish language, that of the Ynga, and Jebero’. The concept Ynga in the title refers to the Inca, the ruling population group in Peru at that time, and the language of the Inca is Quechua, a lingua franca. Xebera is the eighteenth century spelling of the word Jebera. The vocabulary has some three thousand entries. Add. 25,324 includes a Christian doctrine and a grammar of the Jebero language. The language is still spoken by some elderly people in the district of Jeberos in North Peru. Pilar Valenzuela, who did fieldwork in Jeberos, started a revitalization project with the Jebero people.

|| Astrid Alexander-Bakkerus: University of Amsterdam, Faculty of Humanities, Spuitstraat 134, 1012 VB Amsterdam, Netherlands. E-mail: [email protected]

190 | Astrid Alexander-Bakkerus

The Jebero language belongs to the Kawapana language family, a small family consisting of two members of which Chayawita is the other member. As so many other Amerindian languages, Jebero is an agglutinative language and nouns and verbs are the most important word classes. Noun and verb stems can have both prefixes and suffixes.1 Jebero does not have an open, distinctive class of adjectives, but nominalized forms and nouns expressing a quality, a property, or a color, may function as such. The language further has a small, closed class of non-derived adverbs and a few conjunctions only. Subordinate clauses are formed by means of nominalizations. The order of the constituents is predominantly SOV. Interestingly, the language has a dual first person plural and it does not make a distinction between a past, present and future tense, but between a nonfuture and a future tense. The Jebero verbal paradigm further comprises an imperative mood and a great number of nominalized forms.

2 Colonial grammars, characteristics2 Soon after the ‘discovery’ of new continents by western explorers, Europe colonized a great number of the ‘newly discovered’ lands, and a good many grammars in all sorts of indigenous languages were produced for reasons of communication. Most of the grammars were written by missionaries. They were sent by the church to these countries in order to propagate the catholic faith and to convert as many people as possible. When Europe discovered and colonized the ‘world’, it already had a long history of language description, a tradition that dated back to the works of art of the Greek and Latin philosophers.3 (In the Antiquity and in the Middle Ages, grammar was considered to be one of the seven liberal arts.) Moreover, at that time, Latin was not only the language of science, it was also – and still is – the language of the church. The grammarians and missionaries who wrote those grammars thus were trained in Latin, and they knew the classics. It is therefore no surprise that, considering the long tradition of language description, the classical training of the

|| 1 In Quechua, a neighboring language, stems do not have prefixes. They are followed by suffixes. 2 For a discussion in great detail concerning colonial grammars see Zimmermann (1997), and for more information about colonial linguistics see Stolz et al. (2011). 3 For more information about the Greco-Latin tradition of language description, the purpose of a Greek or Latin grammar, and the contents of it, see Alexander-Bakkerus (2005: 45–47).

A colonial grammar or arte | 191

grammarians and the status of Latin, the authors of an arte fell back on the Greek and Latin grammars when they had to describe an indigenous language and that their language descriptions were based on a Greco-Latin model, all the more so because this model appeared to be very suitable for the description of other foreign, i.e. European, languages. This Latin model used by the European grammarians contained the following characteristics:  a nominal declension with six cases;  a verbal conjugation with five tenses (present, imperfect, perfect, pluperfect, future), five moods (indicative, imperative, subjunctive, optative, infinitive) and a number of participles and gerunds/supines;  a sentence divided into in eight parts (noun, verb, pronoun, adjective, adverb, preposition, conjunction, interjection);  a dichotomy between categorematic parts, i.e. declinable, significant parts (nouns, pronouns, verbs, participles) and syncategorematic parts, i.e. nondeclinable, co-significant parts (adverbs, conjunctions, prepositions, interjections) introduced by Priscian (6th century) and by the Modistae (Middle Ages);4  a grammar beginning with sounds, followed by chapters about syllables, words and sentences, and ending with texts, the pièce de résistance and the reason why the grammar was written. A comment should be given on the last item. In Greek and Latin grammars, the pièce de résistance consisted of pieces of poetry, for the grammarians did not wrote their work to teach a language or to convert people, but with a view of promoting the art of poetry and of showing the beauty of it. Therefore they preceded the poetic texts by philosophical treatises about ‘voice’ or ‘elements’ (sounds), syllables, words and sentences, features used by poets to achieve a work of art. Grammar thus was for the benefit of poetry. In colonial grammars written by missionaries, however, the pièce de résistance consisted of a Christian doctrine or parts from the Christian doctrine. The fact is, the grammars were written for the benefit of the church, and they could be considered as un instrumento de la evangelización ‘un instrument of evangelization’ (Zimmermann 1997: 15). Since the missionaries wrote the grammars to transmit their faith, they often used words and phrases from the Bible and the Christian doctrine to exemplify a linguistic phenomenon. Therefore, their grammars, besides including a Christian doctrine or parts of it, are full of biblical verses and doctrinal phrases, and of concepts referring to the catholic faith.

|| 4 See Alexander-Bakkerus (2008: 219–220).

192 | Astrid Alexander-Bakkerus

3 Ms. Add. 25,323 The additional manuscript 25,323 has 69 pages (folio 1 recto – 35 recto). It comprises a trilingual Spanish-Quechua-Jebero vocabulary and notes that precede (folio 1 recto – 2 verso) and follow (folio 35 recto) the vocabulary. The notes concern, amongst other things, the pronunciation of problematic Jebero sounds. These sounds are troublesome, because they do not occur in Spanish or have no phonemic value in Spanish, so that they cannot be symbolized by means of a letter of the Spanish alphabet. The lexicographer then uses German words to describe these problematic sounds. He refers for instance to the German words Mörser ‘mortar’ and Schreibpapier ‘writing paper’ to explain that the Jebero sound system has a front mid vowel [ɶ] and a palatal fricative [ʃ], and he uses the German grapheme ö and the German trigraph sch, respectively, to represent these sounds that sound so strange to a speaker of Spanish. The fact that the author knows German also makes it likely that the manuscripts were written by Samuel Fritz, who was born in Bohemia and knew German. The dictionary does not only contain ca. three thousand words or entrees (see also section 1), it also contains many sentences in which concepts are explained and are used in context.

4 Ms. Add. 25,324 The Ms. Add. 25,324 counts 29 folios. It comprises a part of a Doctrina Christiana ‘Christian doctrine’ and a grammar of Jebero, Gramatica dela Lengua Lengua Xebera (henceforth GLX). The Christian doctrine, as contained in the manuscript (folio 1 recto – 18 verso), is treated in section 5. In this section we pay attention to the GLX. The grammar (folio 19 recto –29 verso) can be divided into two segments. In the first segment, folio 19 recto-24 verso, the different parts of the sentence are treated. The second segment, folio 25 recto-29 verso, is devoted to the conjugation of the verb. Segment 1 has the following contents: Del nombre ‘About the noun’, Del plural ‘About the plural’, Del substantivo y adjectivo ‘About the substantive and adjective’, Delos grados de comparación ‘About the degrees of comparison’, Delos numerales ‘About the numerals’, De los pronombres ‘About the pronouns’, Delos relativos ‘About the relative pronouns’, Delas preposiciones ò postposiciones ‘About the prepositions or postpositions’, Delos adverbios ‘About the

A colonial grammar or arte | 193

adverbs’, Delas conjunciones ‘About the conjunctions’, Delas ynterjecciones ‘About the interjections’. The GLX thus seems to divide the sentence into ten parts, instead of into eight parts, as was customary. Nine sentence parts are discussed in the first segment: nouns/substantives (+ plural), adjectives (+ degrees of comparison), numerals, pronouns, relative pronouns, prepositions/postpositions, adverbs, conjunctions, interjections; and one sentence part, the verb, the most important part of the sentence, is discussed in the last segment. However, a closer look at the data reveals that the grammarian also pays attention to the verb in the first segment. In the section about the plurals, for instance, the author also discusses verbal plurality, and what he classifies as substantives and adjectives actually are nominalized verb forms. We thus may introduce the following rearrangement in the first segment: nouns, verbs, adjectives, numerals, pronouns, relative pronouns, prepositions/postpositions, adverbs, conjunctions, interjections. In addition, when we range the numerals under the class of the nouns and the relative pronouns under that of the pronouns, a classical division into eight parts emerges: nouns (+ numerals), verbs, adjectives, (relative) pronouns, prepositions + postpositions, adverbs, conjunctions, interjections. Interestingly, the author traditionally analyses the declinable, significant parts of the sentence (noun, verb, adjective, pronoun) first, and the non-declinable, co-significant ones (prepositions/postpositions, adverbs, conjunctions, interjections) afterwards, so that the former neatly precede the latter in his arrangement. The author, by separating both groups, thus makes a distinction between the significant and the co-significant parts. So, we may consider the first segment of the GLX as a traditional grammar, in which the eight parts of sentence are treated, and the declinable parts are analyzed separately from the non-declinable parts. The second segment, the one with the conjugation of the verb + the verbal paradigms, can then be considered as a practical appendix to the first segment.

5 A Christian doctrine As said, the manuscript Additional 25,324 also contains a part of a Christian doctrine. It counts 36 pages: folio 1 recto – 18 verso, and it can be divided into three sections: 1. texts in Quechua + a translation into Jebero, folio 1 recto – 12 verso; 2. texts in Jebero, folio 13 recto – 14 recto; 3. texts in Jebero + a translation into Quechua, folio 15 recto – 18 verso.

194 | Astrid Alexander-Bakkerus

Folio 14 verso is not listed in the sections mentioned above. The text on folio 14 verso is in Spanish and it does not belong to the Christian doctrine. On the folio, the writer addresses himself to his successors, advising them, amongst other things, to explain to the Indians, over and over again, “lo que rezan, lo que niegan, lo que creen, lo que han de hazer ò dexar” ‘what they are praying, what they are denying, what they are believing, what they have to do or to omit’, so that they are not “como papageyo” ‘like a parrot’. The advices given by the writer and the fact that he addresses himself to his fellow missionaries give evidence that he wrote the manuscript for the church, so that it could be left for a successor when the priest was transferred to another mission post. Section 1 begins with two loose folios: folio 1 recto and verso. The text of the folios, a part of the daily catechism, is repeated later on, and we encounter it again on folio 4 recto and verso. This means that the Christian doctrine actually begins on folio 2 recto, containing the Act of Contrition. The fact that the Christian doctrine part begins with loose folios suggests that, originally, the folios were not put on order, that the arrangement had not been determined yet, and that the book could have been arranged otherwise. The incoherent arrangement makes it likely that the manuscript was not meant to be published. The doctrine thus really begins with the Act of Contrition on folio 2 recto. The Act is followed by a summary of the daily catechism and by questions, answers and formulas concerning the sacraments of baptism, confession, and matrimony. Section 2 contains four prayers in Jebero: the Lord’s Prayer, Hail Mary, Creed, and Salve. The prayers are followed by an enumeration of the Ten Commandments, the Commandments of the Church, the Sacraments of the Church, and by a general confession. The prayers date from the time of Father Lucas de la Cueva, a Jesuit priest, who in 1638, together with another Jesuit priest, Father Caspar de Cuxia, established a series of missions on the northern bank of the Marañón River. Section 3 is a repetition of the preceding section. It contains the same prayers, the enumeration of the commandments and the sacraments, and a general confession. The difference between both sections is that the Jebero text in section 3 is accompanied by a translation in Quechua. However, a far more interesting, difference between section 2 and section 3 is that their texts, although concerning the same subjects and having the same standardized use of words, nevertheless represent different language strata. The Jebero text in the former section dates from the 17th century, whereas the one in the latter section dates from the end of the 18th century. The difference between both language strata is discernible for instance in the text of the prayer Ave Maria ‘Hail Mary’ below:

A colonial grammar or arte | 195

17th century version

end 18th century version

mucha-i-q’n Maria honour-3sA.IMP-2sO Mary ‘Hail Mary!’ (lit. ‘Let one honor you Mary!’)

muchaa-pa-ti-q’n Maria honour-DUR-3sA-2sA Mary ‘Hail Mary!’ (lit. ‘Mary, being honored by one!’)

gracia-lecn motîo-la, grace-COM be.filled-2sS ‘(You are) full of grace.’

gracia-neng-quec motîo-la, grace-3sPOS-IN be.filled-2sS ‘(You are) full of (his) grace.’

Dios queñma-lecn ñapa-li God 2s-COM be-3sS ‘(God) [The Lord] is with thee’.

Dios apu-mapoa queñma-lecn napa-li God father-1pPOS 2s-COM be-3sS ‘(God) [The Lord] (our father) is with thee’.

cuap’r-losa-quegla queñma woman-PL-SEP 2s mointin-la be.blessed-2sS ‘Blessed art thou amongst women’

cuapr-losa-quegla quenma woman-PL-SEP 2s mointin-la be.blessed-2sS ‘Blessed art thou amongst women’

du-palén-quegla huahua-paln womb-2sPOS-SEP child-2sPOS Jesus mointi-n vila Jesus be.blessed-3sS son ‘[and] blessed is (the son) Jesus, child) [the fruit] of thy womb’.

du-palén-quecla oclinando-li womb-2sPOS-SEP be.born-3sS Jesus huahua-palen mointi vila Jesus child-2sPOS 3sS.be.blessed son ‘[and] blessed is (the son) Jesus, (your (your child, born from) [the fruit of] thy womb’

Sa. Maria inilad, St. Mary virgin ‘Holy (virgin) Mary,’

Sa. Maria virgin St. Mary virgin ‘Holy (virgin) Mary,’

Dios achin, God mother-3sPOS ‘mother of God,’

Dios achin, God mother-3sPOS ‘mother of God,’

cuda hucha-pi-döc-maleg 1pEXC sin-PP-PL-CAUS mucha-vana-o pray-BEN-1s ‘pray for us sinners.’

queñmoa hucha-van-losa-maleg 1p sin-POSS-PL-CAUS mucha-vana-o-döc pray-BEN-1sO-PL ‘pray for us sinners.’

epa-quec now-LOC

epala queñmoa now 1p

timin-pa-ting-unda. die-DUR-3sS-COR

timin-a-cova-su die-NOM-1p-NOM

196 | Astrid Alexander-Bakkerus

‘Now and in the hour of our death’. (lit. ‘Now and when one is dying’.)

Amen ‘Amen’.

ugli-quec-unda. day-LOC-COR ‘Now and in the hour of our death.’ (lit. ‘Now and on the day that we are dying’.) Amen. ‘Amen’.

A comparison between both versions shows that the 17th century text is direct and shorter than that of the 18th century, and that the usage of words in the former is less elaborated than that in the later version. The reason why the newer version appears to be more elaborated than the older one could be that Samuel Fritz, the putative writer of the 18th century doctrine, could spend more time on the translation of the doctrine than Lucas de la Cueva, or, that the 17th century Jebero informants with whom Lucas de la Cueva worked in order to translate the prayer into their language were less informed in this subject than the ones who worked with the author of the manuscripts in the 18th century. Fact is that the usage of words in the 18th century prayer seems to be more developed than the usage in the 17th century version. The extensions and semantic changes in the prayer Hail Mary of the 18th century, vis-à-vis the text of the prayer of 17th century, are discussed in the analysis below: line 1: in the 17th century text, the verb mucha(a)- ‘to honor’ is in the imperative mood. It is an exhortation to honor Mary; in the 18th century text, the verb mucha(a)- ‘to honor’ has a nominalized, durative form functioning as a relative clause of which Mary is the head. It no longer incites mankind to honor Mary. The honoring of Mary is now a fact: Mary is honored by men; line 2: in the 17th century version, the word ‘grace’ is not specified, so that it has a general meaning. It could mean ‘benevolence’, ‘charm’, ‘honesty’, ‘mercy’, ‘virtuousness’, etc. in the18th century version, the word ‘grace’ is specified, and it is a very important and powerful ‘grace’, because it is the mercy of God. By means of this mercy, God absolves us of sin, so that we, free of sins, can ascend into heaven, instead of descending full of sin into the underworld. Since Mary is full of God’s grace, she is almost equal to God, and she can be considered as his female counterpart. Now she can also act as a direct intermediary between God and us. So, instead of addressing God directly, we can address Mary and ask her, a) for forgiveness of our sins, and, b) to plead God to do the same. A double plea will

A colonial grammar or arte | 197

benefit our cause and God will certainly hear the pleas when they come from his homologue, Mary; line 3: in the 18th century version, the name ‘God’ is followed by the ornamental epithet ‘our father’. The adjunction of the possessive pronoun ‘our’, indicates that the father is not only the God of the Spaniards, but also the God of both Spaniards and Indians. This reference to an all-embracing god is missing in the 17th century text; line 5: the specification oclinando-li ‘born’ in the 18th century version emphasizes that Jesus is born naturally, like any other child, i.e. that he is born as a human being and that he is human. The specification refers to one of the Mysteries of the Roman Catholic Church, namely, the nature of Jesus, i.e. that Jesus, being God’s son, but born as a human being, is both divine and human; line 6: both texts, the 17th century version as well as the 18th century one, refer to another Mystery of the church: the virginity of Mary. Remarkably, in the older text the writer tries to translates the concept ‘virgin’ into Jebero, and he uses the word inilad, meaning ‘intact’, as the most appropriate translation. In the 18th century, the loan word virgen ‘virgin’ seems to have become so current, that the translator does not need to search for an equivalent in Jebero and that he can employ the borrowing without any further explanation; line 7: here we have a case of narrowness on the one hand, against broadness on the other hand: the 17th century author employs a first person exclusive marker cuda ‘we’, i.e. ‘you and me’, excluding all other (third) persons, whereas the 18th century writer uses a first person plural pronoun queñmoa ‘we’, including all the persons of the community. There is also an inconsistence in the 17th century version: the person who benefits from the action is a first person plural ‘we’, but the benefactive person marked on the verb is a first person singular ‘I’. Plurality is here indicated by means of the personal pronoun cuda ‘1p exclusive’, preceding the verb form. The 18th century verb form has a regular first person plural benefactive marking on the verb. line 8: the sentence ‘now and in the hour of our death’ is difficult to translate into Jebero, because the notion of ‘hour’ does not exist in Jebero. Nevertheless, it is again the 18th century version that is more specific. The 17th century sentence

198 | Astrid Alexander-Bakkerus

epa-quec timin-pa-ting-unda ‘now and when one is dying’ is rather vague. In this version, the determination of time (‘hour’) is not translated. In the 18th century version, on the other hand, the word ‘hour’ is translated as ugli ‘day’, a concept that is very well known by the Jebero people.

6 Concluding remarks In section 2 we have seen that most of the colonial grammars were written by priests. They contacted the indigenous people with the purpose to spread the Roman Catholic faith, and they learned and described the language of the inhabitants in order to be able to proclaim the faith. We have also seen that such grammars more or less share the same features:  like the Greco-Latin grammars, the colonial grammars traditionally start with sounds, followed by words and sentences, and a number of them end with texts, i.e. with a Christian doctrine or parts of the doctrine;  they divide the sentence into eight parts: nouns, verbs, adjectives, pronouns, prepositions and postpositions, adverbs, conjunctions, interjections;  they make a distinction between the categorematic parts and the syncategorematic parts;  and, the most striking resemblance, the grammatical description is based on the structure of Greek and Latin. The Jebero grammar meets all the criteria mentioned above: (i), it is a language description composed in colonial times for the benefit of the church; (ii), it distinguishes eight parts of the sentence; (iii), it separates the categorematic parts from the syncategorematic ones; (iv), the structure of the indigenous language is modeled on that of Greek and Latin. The only difference between the Jebero grammar and other colonial grammars is the arrangement of its contents. The classical arrangement into sounds, words, sentences, and texts is not applicable to the grammar as contained in the Ms. Add. 25,324. The manuscript does not begin with sounds. It begins with texts, viz. parts of the Christian doctrine, instead of ending with them, and a section about sounds is missing. However, in section 4 we have seen that the manuscript has some loose folios, and that the order of the folios seems to be undetermined. Therefore, it is not unthinkable that the book could have been arranged otherwise and that the GLX preceded the doctrine texts, so that the manuscript ended with texts. The same counts with regard to the manuscript Add. 25,323, which, as we have stated, contains a Spanish-Quechua-Xebero dictionary, and opens and

A colonial grammar or arte | 199

closes with a number of notes concerning the pronunciation of Jebero. The deletions and insertions in both the notes and the dictionary show that the manuscript was in ‘status nascendi’. It is likely then that the author would have joined the notes about phonetics together when making a fair copy of the draft, so that the manuscript would start with a complete section about sounds. Furthermore, it is also not unthinkable that both manuscripts originally consisted of a bunch of loose folios and occasional notes, which were written in stages, and which, in due course, formed a whole – all the more so because the priest wrote them either for personal use or for the general use of the church. In all probability, splitting up the document into two separate manuscripts took place afterwards, when Samuel Fritz’ papers were sold and passed into the hands of the British Library. So, when we join the manuscripts together again and rearrange the draft contents, which were written gradually over the course of time, we may obtain a traditional, classical ordering in sounds (the initial section in the additional manuscript 25,323), words and sentences (the vocabulary in Ms. Add. 25,323 + the grammar in Ms. Add. 25,324, the following manuscript), and texts (the Christian doctrine in Ms. Add. 25,324). In conclusion we can say that the additional manuscripts 25,323 and 25,324 from the British Library jointly constitute a characteristic colonial grammar, and that it is a useful grammar. The grammar is a worthy object of study, not only because it concerns a moribund language of which little is known, but also because it has an exceptional Christian doctrine. The doctrine as contained in the additional manuscript 25,324 is extraordinary, because it reveals two different language layers: a Jebero stratum dating from the 17th century and a stratum from the 18th century, as is shown in section 5.

Abbreviations 1, 2, 3 A BEN CAUS COM COR DUR EXC IN IMP

first, second, third person agent benefactive causative comitative coordinator durative exclusive inessive imperative

LOC NOM

O PL POS POSS PP

S S, P SEP

locative nominalizer object plural marker possessive possession marker past participle subject singular, plural separative

200 | Astrid Alexander-Bakkerus

References Alexander-Bakkerus, Astrid. 2005. Eighteenth-century Cholón. Utrecht: LOT. Alexander-Bakkerus, Astrid. 2008. Two colonial grammars: Tradition and innovation. Asian Jounal of Latin American Studies 21(1). 215–255. Stolz, Thomas, Christina Vossmann & Barbara Dewein (eds.). 2011. Kolonialzeitliche Sprachforschung. Berlin: Akademie Verlag. Zimmermann, Klaus (ed.). 1997. La descripción de las lenguas amerindias en la época colonial. Frankfurt am Main/Madrid: Vervuert/Iberoamericana.

Kazuko Matsumoto

The role of domain and face-to-face contact in borrowing The case from the postcolonial multilingual island of Palau* Abstract: With great attention to the domain in which borrowing first occurs, is maintained, or becomes obsolete, this paper explores borrowing in Palauan from two perspectives: investigating to what extent different colonial source languages have influenced Palauan by means of borrowing by using the contact-induced borrowing scale as the analytical framework; and to what extent borrowing has been linguistically assimilated into Palauan and socially accepted by the speakers by employing such concepts as Lehnwort, Fremdwort and Gastwort. This paper concludes with an emphasis upon the importance of both the domain and face-to-face contact in order to understand the mechanism underlying the process of adopting and retaining borrowing in postcolonial multilingual communities. Keywords: language contact, contact-induced language change, borrowing scale, loanword, domain, Palauan, Spanish borrowing, German borrowing, Japanese borrowing

1 Introduction Borrowing is one of the possible linguistic outcomes when languages come into contact with each other. Amongst a number of approaches to language contact

|| * This research has been supported by a JSPS KAKENHI grant (Ref. No. 25580085). I would like to thank all those in Palau who kindly co-operated with my research, as well as David Britain, Shuichi Yatabe, Mizuho Hidaka, Alexander Gilmore and Caleb Otto for their useful comments and advice, and my research assistant, Akiko Okumura, for her invaluable contribution to the progress of this research. All remaining mistakes and inadequacies are my own. This paper represents a development of ideas on borrowing in Palauan, first explored in Matsumoto (2010). || Kazuko Matsumoto: University of Tokyo, Department of Language and Information Sciences, 38-1, Komaba, Meguro, Tokyo 153-8902, Japan. E-mail: [email protected]

202 | Kazuko Matsumoto

research, many sociolinguistic studies have demonstrated that the “domain of language use” (Fishman 1972) is a useful analytical tool for explaining patterns of language use in various multilingual communities. However, very few sociolinguistic investigations have applied the concept of “domain” itself to examine the mechanisms that underlie the process of borrowing (e.g., Lev-Ari et al. 2014). It is not clear, therefore, whether and how the domains can help us account for the adoption, retention and extent of borrowing in multilingual communities. As we will see later, Palau provides an ideal community in which to observe the changing borrowing patterns associated with varying language use in a range of domains. As a result of successive waves of colonization throughout the twentieth century, Palauan has come into contact with such major world languages as Spanish, German, Japanese and English to varying degrees. Thus, different generations have experienced different language use in a range of domains depending upon administrative policy, and this is clearly reflected in the extent and change of borrowing. Therefore, this study will pay great attention to the domain in which borrowing first takes place, is retained or becomes obsolete, while exploring borrowing in Palauan from two perspectives: investigating to what extent different source languages have influenced Palauan by means of borrowing from the viewpoint of the “contact-induced borrowing scale” (Thomason & Kaufman 1988); and to what extent borrowing has been linguistically assimilated into Palauan and socially accepted by the speakers by deploying such terms as Lehnwort, Fremdwort and Gastwort (McArthur 1992). In other words, I assess not only linguistic properties (i.e., borrowability of different grammatical categories and the linguistic assimilation of borrowing) but also the social characteristics of borrowing (i.e., the degree and domain of contact where borrowing initially occurs, is maintained or becomes obsolete, as well as the perception of borrowing as vernacular or foreign by speakers). The first section summarizes the history of language contact in Palau dating back over a century, explaining the sociolinguistic repercussions of political, economic and social developments through the course of Palauan history, together with the consequent multilingualism in contemporary Palau. This overview is essential to portray the different types of “contact” that each colonial power brought about and the consequent borrowing seen in the present day. The second section explains the two frameworks on which the following analyses of borrowing are based, while the third section presents the analyses of borrowing in Palauan both qualitatively and quantitatively on the basis of a wide range of data. I demonstrate that both the domain and face-to-face contact are crucial factors in understanding the

The role of domain and face-to-face contact in borrowing | 203

social and linguistic mechanisms that underlie the process of adopting and maintaining borrowing in a postcolonial multilingual community.

2 Palauan language contact history Tab. 1: Language contact history in Palau (Source: Based on Matsumoto 2001) Period

Language in contact

Factors engendering contact

Administration

1885 – 1899 (14 years)

Spanish

Missionary work

Spanish administration

1899 – 1914 (15 years)

German

Missionary work, Education Mining industry, Militarism

German administration

1914 – 1945 (31 years)

Japanese

Japanese administration as Japan’s Mandate authorized until 1933 by the League of Nations

1945 – 1994 (49 years)

American English

Immigration, Education Cultural hegemony Commercial activities Industrial development Imperialism, Militarism Politics, Education Cultural hegemony

1994 to Present Day

Filipino English

Labor migration

American English

Politics, Education Cultural hegemony

Filipino English Japanese

Labor migration, Education Tourism, Education Cultural hegemony

American administration as the US Trust Territories of Pacific Islands authorized by the United Nations The Republic of Palau (Compact of Free Association with the US)

The Palau Islands are an archipelago located in the Western Caroline region of the Pacific, with a population of 20,300 (Office of Planning and Statistics 2005). The Austronesian indigenous language, Palauan, has, as a result of a century of successive colonial domination by Spain, Germany, Japan and the US, come into prolonged contact with a number of non-local languages. Table 1 summarizes, in chronological order, the relationship between the colonial contact languages and the specific factors during their reign that engendered language contact. The history of Palauan language contact may be roughly classified into the following three stages on the basis of causes of the contact: a classic stage where the propagation of religion urged Palauans to learn missionary words (Spanish and German); a pre-modern stage where mainly political domination

204 | Kazuko Matsumoto

by foreign nations obliged the Palauans to learn the dominant language at school (German, Japanese, and then American English); and a contemporary stage where scientific and technological advances, cultural hegemony, economy relying upon the tourism industry from Japan and other Asian nations, labor migration from the Philippines and globalization have led the Palauans to have contact with American and Filipino English as well as Japanese. I will now look closely at the history of Palauan language contact with those four former colonial languages.

2.1 Palauan language contact with Spanish and German Palauan language contact with Spanish began in 18851 when Spanish Roman Catholic missionaries landed on Palau for the purpose of converting the island’s population to Christianity (Hezel & Berg 1980: 373). However, no systematic social, political or economic policies were put into practice, nor were formal education programs established (Shuster 1978: 149). In 1891, however, the Spanish Capuchins set up a mission station and school in Palau (Engelberg 2006: 9), and the priests occasionally delivered catechism classes; these classes were attended by 48 youths, but most of them were not Palauans but Chamorros from the Northern Mariana Islands (McKinney 1947: 83). The day-to-day pattern of Spanish language use, the teaching methods and the attitudes toward Spanish in Palau are unknown. Nevertheless, it is most likely that Palauan language contact with Spanish was restricted to religious domains only and that their attitudes toward Spanish were respectful since it was considered to be the language of “God”. The German era began in 1899 when Germany purchased almost all of Spain’s former territories in Micronesia as a result of the German-Spanish Treaty that followed Spain’s defeat in the Spanish-American War of 1898 (Rechebei & McPhetres 1997: 121–122). The rationale for German colonization in Micronesia was to exploit its natural resources in order to contribute to economic development in Germany (Aoyagi 1977: 45). The early period of German domination was spent only in converting the islands’ populations to Christianity, while seeking potential resources. From 1903, when phosphate deposits were discovered in Angaur as well as when copra production increased, Germany started promoting and greatly expanding commercial and military enterprises in Palau (Hezel

|| 1 For contact before 1885, see Engelberg (2006) and Britain & Matsumoto (2015).

The role of domain and face-to-face contact in borrowing | 205

& Berg 1980: 397; 421–423). In 1909, workers in the phosphate business included not only hundreds of Palauans, but also many imported laborers: 23 Europeans, 55 Chinese, 98 Yapese and 126 from the central Carolines (Hezel & Berg 1980: 421–423). In 1912, this number increased to more than 800, nearly 90 % of whom were from across Micronesia (Firth 1973 in Engelberg 2006: 9–10). However, no information is available as to what language became the principal lingua franca in the phosphate related business, a simple (possibly pidgin) German or English-based Pacific pidgin.2 With regard to education, in 1902, firstly a small vocational school was established for the training of policemen, where mainly German and math were taught (Palau Community Action Agency 1977: 195) to between 20 and 30 students (Hezel 1984). Then, in 1907, the first mission school opened in Koror with 54 students, followed by four more mission schools in Koror and another in Melekeok in 1908, and then by three more in the following years (Engelberg 2006: 10–11). The school program lasted for three years and was devoted mainly to catechism and German language studies in addition to basic math, geography, art and music (Shuster 1982: 150–152; Aoyagi 1977: 45). In 1914, the final year of German administration, 361 children out of the total population of 4,200 attended schools, most of whom were from children of the meteet – the top of the hereditary caste hierarchy (Shuster 1982: 150–152).3 Positive attitudes toward German at that time have been reported: Palauan pupils were anxious to study it, while the Palauans not only asked for the Fathers to open more schools, but built some at their own expense in Melekeok and Ngatmel (Engelberg 2006: 14). Nevertheless, a teaching staff shortage appears to have resulted in less effective teaching methods, such as learning German songs by heart in some schools (Engelberg 2006: 12). The German population in Micronesia was never large; in 1913, only 259 Germans lived in the whole of Micronesia (Aoyagi 1977: 45). To sum up, German and Spanish control of Palau was mostly “symbolic rather than strategic” (Peattie 1988: 36). The Spanish East Indies located their base in the Philippines, while the German New Guinea established their head-

|| 2 Mühlhäusler (1996) and Mühlhäusler & Baker (1996) claim that the use of English-based pidgins was still prevalent in Micronesia at the very start of the 20th century, since the Pacific in the 19th century was busy with European and American whalers, trading seamen, and a motley array of beachcombers, absconders and victims of shipwreck (see Matsumoto and Britain forthcoming b for details). 3 Only one German public school offered sixth-year advanced education in Saipan, and only one Palauan finished it (Shuster 1982: 151; Imaizumi 1990: 4).

206 | Kazuko Matsumoto

quarters in New Guinea; hence, Micronesia including Palau was merely peripheral, never their main focus. In addition, their activities would have affected a rather restricted and small area of Palau – Koror. Communication and transportation facilities were not developed in Palau at that time, with residents of the largest jungle-covered island of Babeldaob and the outlying islands and atoll remained relatively unaffected by the turmoil and cultural upheaval of these colonial incursions (Hezel & Berg 1980). Thus, although Germany embarked on economic and educational reforms on a larger scale than Spain had, it was not, on the whole, sufficient enough to have a strong effect on Palauan social structure and language.

2.2 Palauan language contact with Japanese and American English Nearly thirty years of Japanese occupation began in 1914 as a result of the Anglo-Japan alliance during the First World War (Palau Community Action Agency 1978: 280–281). Micronesia was considered to be valuable not only for a vital strategic advance in any future conflict with the US (Peattie 1988: 42), but also as an outlet for Japan’s growing population (Shuster 1978: 9, 42). Palau developed, to use Schneider’s (2007: 25) sociolinguistically-driven classification, as a “settlement colony”, one formed with a demographic dominance of speakers of the colonial language. Japan established the headquarters of the Nan’yō-chō (the South Seas Bureau) in Koror, Palau, made it the capital of the Nan’yō-guntō (the South Sea Islands), and subsequently launched radical reforms in the demographic, economic and educational arenas as well as in the infrastructure of the island (see below for details). As a result, Palau became the political, economic and educational center of Micronesia for the first time in its history. First and foremost, there was a massive influx of Japanese civilian immigrants into Palau. In 1935, there was approximately a one to one ratio (6,553 and 6,230 respectively) of Japanese and Palauans, but the Japanese outnumbered Palauans by three to one (17,006 and 6,509) in 1937, and by four to one (23,980 and 6,514) in 1941 (Nan’yō-chō 1928, 1939, 1941, 1942). Importantly, these immigrants in the early period consisted of not only the ruling elites and military personnel but also carpenters, miners, textile industrial workers, drivers and farmers (Rinji Nan’yō Guntō Bōbitai 1922: 54–65) who had been recruited from Japan for their laboring skills and who worked with the islanders in Japanese enterprises in Palau. These Japanese civilian immigrants in Koror lived with indigenous Palauan residents in the same neighborhoods, rather than establish-

The role of domain and face-to-face contact in borrowing | 207

ing exclusive Japanese communities.4 As a result of this integrated settlement pattern, Palauan children in Koror interacted daily with Japanese children. Even before they went to school, they acquired Japanese ‘moral codes’, aspects of Japanese culture, such as traditional Japanese tales and songs, the Japanese language, jingles used for memorizing multiplication in mathematics and so forth. Ultimately this contact led to a large number of marriages between the Japanese and Palauans, which resulted in the emergence of a considerable Japanese Palauan population on the islands. Thus, through their everyday contact with Japanese in the neighborhood as well as home domains, as Hezel & Berg (1980: 474) describes, the Palauan norms of living, eating and drinking had come to vigorously incorporate Japanese practices (see also Matsumoto & Britain 2003a, forthcoming a). The introduction of the first ever nationwide education system also seems to have supported the diffusion of the Japanese language. Although there were some policy changes (e.g., subjects taught) in schools over time, almost all Palauan children received three years of compulsory education and two years of supplementary education, during which all subjects were taught in Japanese by Japanese teachers and the use of Palauan was forbidden. Advanced schools providing a two-year training in carpentry, agriculture, dress-making and nursing were also present in Koror, at which selected students from other Micronesian islands as well as Palauans communicated in Japanese as a lingua franca.5 Moreover, several out-of-school educational organizations were set up in order to enable Palauan graduates to further incorporate the Japanese social system as well as Japanese thought into their daily lives (see Matsumoto 2011 for details). Furthermore, a money-oriented economy significantly affected the social life and language use of the people. The appearance of Koror changed dramatically; it was described as a “suburb of Yokohama” (Shuster 1978: 13), a “handsome tropical city” (Kluge 1991: 5) and “Little Tokyo” (Leibowitz 1996: 14) by journalists, missionaries and scholars. Numerous restaurants, cafés, bars, a variety of shops as well as commercial and government offices lined the main street, while factories, laboratories and brothels were built on the back streets. In particular, due to an excessive number of private houses, it was said that the small island of Koror appeared overcrowded. Palauans routinely used Japanese as their second lan-

|| 4 The only exceptions were some Japanese settlement villages that were established to accommodate settlers on Babeldaob Island during the latter period. By contrast, however, early Japanese migrants concentrated mainly on Koror, Malakal, Arakabesang and Angaur Islands where no such exclusive Japanese settlements were built (Rinji Nan’yō Guntō Bōbitai 1922: 6–7). 5 Several Palauans went to mainland Japan for high school or vocational school.

208 | Kazuko Matsumoto

guage. The degree and frequency of everyday interaction between the Japanese and Palauans was great enough to have brought about a local variety of Japanese in Palau (Matsumoto & Britain 2003b and forthcoming b; Matsumoto 2011 and 2013). Half a century of US domination started at the end of the Pacific War, in which the Japanese strategy of using military bases in Micronesia as steppingstones to Hawaii made the US aware of the strategic military value that Micronesia had for the US. For that reason, the US wished to maintain permanent access to Micronesia, but, at the same time, it needed to appear to be providing democratic free choice and advancing self-government in Micronesia, in order not to lose its own international credibility (Anglim 1988: 1). This dilemma led to a rather different form of administration and contact to what Palau had previously experienced. The confidential US Solomon Report (1963 in Anglim 1988: 10) shows that having no economic development plan was, in fact, the intentional US policy, making the Palauan economy totally dependent on the US and leaving no operative industry. Meanwhile, the US encouraged the growth of the Palauan public sector by providing generous salaries for government employment and by privileging its workers to buy American luxury items and imported food, hence making the poorly-paid private sector unattractive and underdeveloped (Anglim 1988: 9). Furthermore, in order to produce a local elite that would be loyal to the US, the US manipulated official appointments and provided American funds to do so (US Solomon Report 1963 in Anglim 1988: 10). Thus, this left little possibility for self-sufficiency in Micronesia. As a result, Micronesia struggled to achieve independence from the US, and of all the UN Trust Territories in the world it took Palau the longest time to gain independence. And despite independence, all of the “independent” countries that emerged from the Trust Territories of Pacific Islands are in a “Compact of Free Association” with the US, and are heavily reliant on American funding in turn for providing access, potential or real, to the US military. In terms of education, the school system and the availability of American teachers varied over time depending upon changes in the foreign political climate and the US federal budget. Up until 1962, Micronesian educational issues were “to a great extent forgotten by the American government” (Shuster 1982: 179); there were no qualified teachers or textbooks, and only a limited budget (Palau Community Action Agency 1978: 487). As a consequence, initially the new schools employed some well-educated Palauans who had attended the Japanese schools and hence took on a “Japanese flavor” (e.g., shūshin education, the Japanese moral education), although the school system itself appears to have been Americanized (i.e., elementary from grades 1 to 6; and intermedi-

The role of domain and face-to-face contact in borrowing | 209

ate schools from grades 7, 8 at first and 9 in 1950) (Shuster 1982: 183). Later the civil administration initiated a “Bilingual Education Program for Micronesia” (Trifonovitch 1971), allowing Palauans to produce their own teacher-training curricula and textbooks (Shuster 1982: 183, 212). From 1962 until 1980, “tens of millions of aid dollars” together with American personnel poured into Palau (Shuster 1982: 199). The Bay of Pigs invasion and the Cuba Missile Crisis intensified the Cold War confrontation between the US and the Soviet Union, reminding the US once again of the strategic importance of Micronesia (Shuster 1982: 197). As a result, support for educational programs expanded enormously, introducing American-style education similar to that in the US mainland, including American textbooks and the American standard of school grades (i.e., 1–8 elementary; 9–12 high) to Palauan schools (Abe 1986: 205; Shuster 1982: 200). The Vocational School was re-established as the Micronesian Occupational College in 1969, while the American Dependent School ceased to exist6 (Shuster 1982: 203–206). Initially, twenty-four American contract teachers arrived, but were then replaced between 1966 and 1970 by many young Peace Corps Volunteers (Shuster 1982: 199, 208). By the end of 1966, the first 323 volunteers had arrived in Micronesia, and in 1968 there was one volunteer to every 100 Micronesians (Lingenfelter 1974: 59 in Abe 1986: 206). Furthermore, during the 1970s and 1980s, a large proportion of the US funds7 were used to send more than 75 % of Palau’s high school graduates to US colleges each year (Shuster 1982: 216). This all led to the creation of a PalauanEnglish bilingual elite, educated in the US and working for high salaries in public sector government offices. Regarding demography, overall, only a small number of Americans were stationed in Palau, such as military and administrative personnel, missionaries, schoolteachers, members of the Peace Corps, judges and attorneys, and then only temporarily – no systematic Anglophone immigration took place. It appears that the volunteers, unlike the American officials, were rather idealistic and sympathetic to the islanders – there was some interaction between the volunteers and Palauans, but their short contracts and small numbers in comparison to the islanders suggest that they made relatively little linguistic impact.

|| 6 In 1964, six Palauan students were integrated among the seventeen Americans, but in the early 1970s, such an integrated class disappeared, since “there were too few American children any longer to warrant such special classes” (Shuster 1982: 204, 206). 7 For example, Palau received $357,200 in 1978, $841, 300 in 1979 and over one million dollars in 1980 (Shuster 1982: 213), for a population of just over 12,000 in 1980 (Office of Planning and Statistics 2006: 23).

210 | Kazuko Matsumoto

Perhaps the biggest demographic changes that the US period brought to Palau were: (a) the expatriation of all earlier Japanese settlers; (b) the importation of Filipino labor employed by Palauans, thanks to the so-called “compact money” from the US; and (c) the migration of young Palauans to the US for educational and employment purposes, thanks to US grants as well as their status as “habitual residents”. Overall, while the Japanese era saw a rather interventionist and integrationist approach at engendering consent, the US period was characterized by a distant hands-off stance. Thus, it is most likely that the intensive Japanese administrative strategy in Palau was far more influential upon Palauan society and language, with respect to infrastructure, demography, economy, education, belief systems and lifestyle, than either the earlier European or the later American domination. However, given America’s longer period of control (i.e., for half a century) and on-going financial aid, there are, of course, a number of similarities in the sociolinguistic consequences of these two most recent colonial periods in the history of Palau. During each administration, recognition of Japanese or English as the high status language as opposed to Palauan as the low language (i.e., diglossia) was established, while the usual linguistic compartmentalization was reinforced so that the colonial languages were used in the school, legal, administrative and written domains, while Palauan was mainly spoken in the home and traditional domains. Moreover, Palauan-Japanese or Palauan-English bilingualism became the norm, while the use of Japanese or English borrowing and code-switching in Palauan conversation has come to function as typical ‘ingroup’ language behavior among the different generations. However, the crucial difference is that the use of Japanese was not restricted to those official domains; on the contrary, face-to-face interaction in Japanese was commonplace in everyday life – in the neighborhood, at work and in the marketplace.

2.3 Multilingualism in contemporary Palau Palau finally became a technically independent nation in 1994 after approving and signing the agreement of the Compact Free Association with the US. Today, just a few very elderly Palauans survive to remind us of the Palauan Japanese speech community. Nevertheless, given the lengthier and more intensive contact with Japan and the US, it seems reasonable that the impact of both countries should still appear to be of significance for Palau. English has remained as the official language along with the indigenous language, Palauan, while the

The role of domain and face-to-face contact in borrowing | 211

teaching of Japanese as an additional language has been widely adopted in Palauan high schools and college. Palauan as the only national language has succeeded to some extent in expanding its use in wider domains than before; on radio, at work, in shops as well as in the neighborhoods, Palauan is predominantly spoken. Nevertheless, the sociolinguistic situation in Palau still appears to be diglossic, since official documents, school textbooks and newspapers are still written in English. Furthermore, English monolingual children have begun to emerge in Koror, worrying many that it may be an early symptom of Palauan language death. Thus, over the last century, Palau has experienced dramatic socio-political, economic, educational and demographic changes, creating Palauan language contact with Spanish, German, Japanese and English to differing degrees. It will be explored later how those social changes and contact have contributed to the Palauan language structure in the form of borrowing.

3 Analytical framework and methodology This section provides the two frameworks on which the following analyses of borrowing are based, explaining the data and methods used for each analysis.

3.1 Linguistic assimilation and social acceptance of borrowing Three German terms are often adopted by philologists to describe the process of linguistic assimilation and social acceptance of borrowing. McArthur (1992: 623) defines each of them as shown below: A Gastwort (guest-word) is an unassimilated borrowing that has kept its pronunciation, orthography, grammar, and meaning, but is not used widely… A Fremdwort (foreignword)… has been adapted into the native system, with a stable spelling and pronunciation (native or exotic)… A Lehnwort (loanword) proper is a word that has become indistinguishable from the rest of the lexicon and is open to normal rules of word use and word formation (bold emphasis added).

Important here is that those terms not only capture the different degrees of linguistic assimilation of borrowing into the recipient language, but also indicate the difference in social acceptance or awareness of borrowing as vernacular or foreign by speakers.

212 | Kazuko Matsumoto

Practical difficulties, however, in using this conceptual framework should be pointed out. Firstly, Gastwort seems to be difficult to distinguish from codeswitching owing to its linguistic properties as well as its non-widespread use. It is noteworthy, however, that when more than one language is juxtaposed in a bilingual person’s utterances, in practice the distinction between borrowing and code-switching becomes too fuzzy to draw a line. Romaine (1995: 124), for instance, supports the view that borrowings in a bilingual community may “start off as codeswitches” and gain acceptance as borrowings “by recurring over time in the speech of more and more individuals”. After denying both of extreme positions, one of which believes that “code-switched elements will never turn into borrowings” and the other of which claims that “code-switching is the only mechanism through which foreign morphemes are incorporated into a language”, Thomason (2001: 132–133) explains that code-switches and borrowings are “two different things”, but “they lie at opposite ends of a continuum; the difference between them is not categorical”. In this study, therefore, some examples of Gastwort may include potential code-switches recorded or observed during my on-going fieldwork in Palau since 1997. The second difficulty is that the stability of spelling and orthography of loanwords cannot be an indicator to distinguish these three terms in this study. Since Palau is traditionally an oral society, ‘stable spelling’ for Palauan, even today, is still disputable.8 Furthermore, Palauans do not seem to feel comfortable with Palauan as a written language, given the fact that the local newspaper Tia Belau’s attempt to issue Palauan translations of articles originally written in English was given up within a couple of years due to unpopularity. Therefore, my analysis of this conceptual framework will exclude the issue of whether the spelling of borrowing is stable or not, but focus on the difference in social acceptance or awareness of them as vernacular or foreign by Palauans as well as the different degrees of linguistic assimilation of borrowing depending upon source languages. A qualitative analysis will be conducted on the basis of my long-term participant observation, ethnographic interviews and conversational data recorded from different generations of Palauans.

|| 8 Moreover, different generations are literate in different orthographies (i.e., Japanese katakana and the Roman alphabet) (see Matsumoto 2001).

The role of domain and face-to-face contact in borrowing | 213

3.2 Contact-induced borrowing scale Thomason & Kaufman (1988) postulate that “degree of contact” determines whether or not different types (i.e., lexical, phonological, phonetic and morphosyntactic) of borrowing are likely to take place. That is, degree of borrowing is mostly dependent on the “intensity of contact” between the two peoples speaking the languages, and that given high-intensity contact, anything could be borrowed. Here, intensity of contact involves factors of “socioeconomic dominance”, “the duration of the contact period”, “the level and proportion of bilingualism”, “speakers’ attitudes” and so forth (Thomason 2001). As Table 2 illustrates, a hierarchy called the “borrowing scale” at four different levels was then proposed, which takes contact intensity as its main variable and lists the likely sorts of borrowing for scenarios of ever increasing intensity of contact. Tab. 2: Borrowing scale (Thomason 2001: 70–71)9 Category

Intensity of contact

Type of borrowing

1 Casual contact

Borrowers need not be fluent in the source language, and /or few bilinguals among borrowing-language speakers Borrowers must be reasonably fluent bilinguals, but they are probably a minority among borrowing-language speakers

Only non-basic vocabulary borrowed (most often nouns, but also verbs, adjective and adverbs)

2 Slightly more intense contact 3 More intense contact 4 Intense contact

More bilinguals, attitudes and other social factors favoring borrowing Very extensive bilingualism among borrowing-language speakers, social factors strongly favoring borrowing

Still non-basic vocabulary (content/ function words) borrowed; slight structural borrowing (new phonemes realized by new phones, but in loanwords only) Basic vocabulary also borrowed; moderate structural borrowing (addition of new phonemes even in native vocabulary; syntax; morphology). Continuing heavy lexical borrowing in all sections of the lexicon; heavy structural borrowing.

In terms of ease of borrowing, the borrowing scale argues that it moves from lexical items through phonology to syntax, with morphology the most difficult to borrow (Thomason 2001: 96–97) (cf. Romaine 1995). Important to my analysis

|| 9 Due to limited space, the scale is simplified here and mainly shows the aspects relevant to Palauan.

214 | Kazuko Matsumoto

is that: (a) among lexical items, nouns are most easily borrowed, while verbs, adjectives and adverbs are next easily borrowed; (b) distinctions between categories 2 and 3 can be whether new phonemes appear in borrowings or whether phonemicization of previously allophonic alternations occur in the recipient language as well as whether lexical borrowing affects non-basic vocabulary only or includes basic vocabulary as well (see below for details). In the following analysis section, therefore, I will apply the contact-induced borrowing scale to borrowing in Palauan by investigating both the type of borrowing in Palauan and the intensity of contact in Palau over a century, paying particular attention to key social factors, such as “socioeconomic dominance”, “the duration of the contact period” and “the level and proportion of bilingualism” and “speakers’ attitudes”. Both quantitative and qualitative analyses will be conducted on the basis of loanwords listed in the Palauan-English Dictionary (Joseph 1990); more examples of borrowing and potential code-switches recorded or observed; Palauan language contact history provided in section 2; as well as results of an ethnographic questionnaire survey on attitudes toward Japanese, English and use of borrowing and code-switching (Matsumoto 2001).

4 Analysis of borrowing in Palauan 4.1 Overall picture of borrowing in Palauan To begin with, it would be useful to have an overall picture of borrowing in Palauan. Table 3 illustrates to what extent each source language has contributed borrowings to Palauan on the basis of two Palauan-English dictionaries published in 1977 (McManus) and 1990 (Joseph). The older dictionary by McManus (1977) was analyzed by Engelberg (2006), while the more recent one by Joseph (1990) was quantified by us. Tab. 3: Number and proportion of borrowing in Palauan according to source language (Sources: McManus 1977 in Engelberg 2006 and Joseph 1990) Source language English Japanese German Spanish Others Total

1977 128 345 51 23 1 548

1990 23.4 % 62.9 % 9.3 % 4.2 % 0.2 % 100 %

252 639 39 89 4 1023

24.6 % 62.5 % 3.8 % 8.7 % 0.4 % 100 %

The role of domain and face-to-face contact in borrowing | 215

It may be worth clarifying how the origin of each borrowing was defined in this study. Although both dictionaries note which source language each loanword originates from, there were some cases which require careful examination on where the borrowing directly comes from. For instance, these dictionaries indicate that the Palauan kohei comes from the English verbal phrase ‘go ahead’. There are some records (e.g., Nihon kokugo dai-jiten 2001), however, that the English phrase go ahead had already been adopted into Japanese in Japan as gōhē particularly in naval language as well as the fishing industry before Palauan had contact with Japanese.10 Furthermore, the phonological adaptation of kohei reflects similarity with Japanese pronunciation of gōhē, since both /eː/ and /eɪ/ are often possible phonological variants of in Japanese. Therefore, the source language of kohei is categorized as Japanese.11 Overall, it indicates that the greatest proportion of borrowing, at over 60 %, comes from Japanese. The second highest contributor is English at around 25 %, while both Spanish and German loanwords amount to less than 10 %. In a strict sense, the two dictionaries may not be comparable; notably, the 1990 dictionary has greatly expanded the number of lexical items, implying that the 1977 dictionary may have rather limited coverage of the language. Therefore, the difference in the proportion of borrowing with different origins between 1977 and 1990 should not automatically be interpreted as an increase or decrease of borrowing from each source language. For instance, the rise in the proportion of Spanish loanwords in 1990 does not mean that new Spanish words have been recently borrowed, but rather that some Spanish words may have been missed by the 1977 dictionary. However, it also seems to be true that there have been some changes in borrowing in Palauan. For instance, the number of German borrowings appears to have declined. It is likely that the domains of borrowing depending upon the source language help us to explain some of the reasons for those changes. Spain brought Christianity to Palau for the first time, together with numerous missionary words (e.g. kerus and misang from Spanish cruz and Misa, meaning

|| 10 Interestingly, the borrowing gōhē is obsolete in modern Japanese. 11 Similarly, many English terms related to baseball had already been adopted into Japanese in Japan before baseball was introduced by the Japanese to Palau during the Japanese administration. The phonological adaptation of these terms clearly shows that they come from Japanese not directly from American English. For instance, the original English words containing /l/ were adapted as /r/ in Japanese due to a lack of /l/ in Japanese, so Palauan imported words with /r/ rather than with the original /l/. Note that strictly speaking the Japanese sound of is neither /r/ nor /l/, but the common belief held by Palauans is that Japanese speakers cannot pronounce /l/, so they use /r/ instead.

216 | Kazuko Matsumoto

‘cross’ and ‘Mass’), and Roman Catholicism remains the dominant religion in Palau. Therefore, it makes sense that those Spanish missionary words have been preserved in the religious domain. German borrowing, on the other hand, is associated with not only missionary words in the religious domain but also with the names for modern products and concepts in military, mining and education domains. German words in the latter domains have tended to be replaced by the subsequent Japanese loanwords, since the Japanese colonial government implemented far more thorough social reforms in Palau in these domains. For instance, the words for ‘grammar’ and ‘letter’ were first adopted from German Grammatik as kramatik and Brief as berib, but they came to coexist with or even appear to be supplanted by the Japanese bunpoo and tegami (Pal. bumpo and tengami). However, the subsequent half-century of US domination sometimes led to the obsolescence of some Japanese borrowings in the domains where German and Japanese words had been in competition. Good illustrations of this are that the words for ‘company’ and ‘school’ were first adopted from Japanese kaisha as kaisia and gakkō as kakko but came to coexist with or even seem to be superseded by the English borrowings kombalii and skuul. Nevertheless, the exception seems that the so-called ‘core’ or ‘basic’ vocabulary – “the kinds of words that tend to be present in all languages” (Thomason 2001: 70) such as body parts, numbers, words that express feelings and senses and baby talk – have remained Japanese. For instance, body parts include chi12, chude, sinzo and kosi (Jp. i, ude, shinzo and koshi meaning ‘stomach’, ‘arm’, ‘heart’ and ‘hip’), whilst numbers include niziu and mang (Jp. nijū and man meaning ‘twenty’ and ‘ten thousand’). Regarding words that express feelings and senses, the Japanese words meaning ‘funny’, ‘tasty’ and ‘lonely’ were borrowed as omosiroi (Jp. omoshiroi), choisii (oishī) and sabisi (sabishī). Common expressions such as ‘No worry!’, ‘(You’ll be) fine or (Are you) OK?’, ‘Be careful!’ and ‘(I’m) sorry!’ were also adopted from Japanese as simbainai (Jp. shinpai nai), daiziob (daijōbu), chabunai (abunai) and komeng (gomen).13 Japanese words and expressions which are used to address babies or small children (i.e., baby talk) were also adopted; e.g., nenneng (Jp. nenne, meaning ‘sleep well’), chensi (en shī14, meaning ‘sit down’) and sisi (shishi, meaning ‘wee-wee’). Other core vo-

|| 12 Note that is the Palauan orthographic representation of a glottal stop. 13 In both Japanese and Palauan, null subjects are possible and common. 14 The nursery word chensi might be a case where an interdialect form has been adopted as borrowing. Various nursery words meaning ‘sitting’ seem to have been used in different regions in Japan; enko, ento, encho etc. in Kantō (Tokyo, Kanagawa, Saitama, Chiba, Gunma,

The role of domain and face-to-face contact in borrowing | 217

cabulary includes basic concepts used in everyday life, such as ‘sound’ (Jp. oto as choto in Palauan), ‘place’ (basho as basio), ‘meaning’ (imi as chimi), ‘color’ (iro as chiro), ‘problem’ (mondai as mondai), ‘promise’ (yakusoku as iaksok), ‘vegetable’ (yasai as iasai) and many others. As Holmes (2008: 28) pointed out, in diglossic situations, there is a tendency that “the H vocabulary includes many more formal and technical terms such as conservation and psychometric, while the L variety has words for everyday objects such as saucepan and shoe”. In the case of Japanese borrowing in Palauan, however, not only formal words, but also those for everyday items and expressions were transferred from the high language through face-to-face interaction with neighboring Japanese settlers and co-workers (Matsumoto & Britain 2003a and forthcoming a). Thus, the reasons for all of those changes in borrowing depending on source language highlight the importance of the domain and intensity of contact as well as everyday face-to-face interaction in order to understand the change in borrowing.15

4.2 Linguistic assimilation and social acceptance of borrowing in Palauan I now turn to applying the three philological terms to borrowing in Palauan. Firstly, Spanish and German loanwords are mostly modified to fit Palauan phonological rules and morphological constructions (see Joseph 1984 for details). In terms of linguistic assimilation, therefore, they may be categorized as Lehnwort.

|| Ibaragi, Tochigi), ento in Tōhoku and Hokkaidō (Fukushima and Sapporo) and en in Kyūshū (Fukuoka) (Fujiwara 1996; Hirayama 1992). So, when addressing babies in an imperative sentence, the light verb do (Jp. suru) is conjugated to be attached to them, and hence en(cho)shite, for example, would have been used in Eastern and Kyūshū dialect speaking areas. What is interesting is that the use of the light verb shī (rather than shite) in an imperative sentence is one of the characteristics of Western dialect (e.g., Kinki). That is, en from the Eastern and Kyūshū dialect speaking regions is combined with si from the Western dialect speaking region, supporting the claim by Matsumoto & Britain (2003b) and Matsumoto (2013) that different migrant dialects of Japanese were in contact in Palau, resulting in koineization. 15 I am aware that a synchronic comparison cannot offer a fully comparable picture for different source languages adopted at different times, since there are cases where some borrowings have fallen out of use due to their irrelevance today (e.g., Japanese loanwords related to war have fallen out of use in Pohnpeian; Miyagi 2000) or replaced by new borrowings from another source language and so forth.

218 | Kazuko Matsumoto

However, the awareness of people recognizing those words as foreign-originated words seems to depend upon both the word itself and the generation of the speakers. Because these words are chronologically the oldest and the degree of their linguistic nativization is high, the younger generation does not tend to realize that they are loanwords. However, since many borrowings with Spanish and German origins are related to religion, the older generation still recognizes all of them as non-native but do not necessarily know which language they originally came from. As time goes by, however, it is likely that they will eventually lose their foreignness in Palauans’ awareness. Hence, they are in a transitional stage from Fremdwort to Lehnwort. In the case of Japanese and English borrowing, however, the distinctions proposed by the three terminologies might be difficult to apply neatly, since Palauan contact with Japanese and English has resulted in both societal and individual bilingualism in Palauan-Japanese and Palauan-English. That is, the elderly, who are bilingual in Palauan and Japanese, use Japanese borrowings that are linguistically unassimilated into Palauan, keeping their pronunciation16, grammar and meaning; i.e., Gastwort. By contrast, youngsters, who are Palauan monolinguals or Palauan-English bilinguals, use Japanese loanwords that are nativized into Palauan; viz. Fremdwort. The reverse is true for English borrowings. The young Palauan-English bilinguals tend to preserve the original English pronunciation, grammar and meaning when English words are borrowed into Palauan conversation (Joseph 1984: 82); hence, Gastwort. In contrast, the elderly Palauan-Japanese bilinguals, who have lower or no command of English, often use English borrowings that are linguistically Palauanized; i.e., Fremdwort. Thus, the linguistic assimilation of Japanese and English loanwords depends upon the generation of speakers who are bilingual in different combinations of languages. However, the social acceptance and awareness of people recognizing those words as vernacular or foreign turn out to be more complex. Although Japanese borrowings are chronologically newer than Spanish and German, Japanese loanwords include the so-called ‘core’ or ‘basic’ vocabulary, such as body parts, numbers, words that express feelings and senses and baby talk, as described earlier. Therefore, some in the young generation in fact do not recognize that they are of foreign origin; hence Lehnwort. That is, variations have emerged in the recognition of Japanese borrowing in Palau depending on the generation of

|| 16 Although Joseph (1984: 82) explains that Japanese loanwords are totally assimilated into the Palauan phonological system, it is not very likely that he intended to examine the different use of borrowing across different generations.

The role of domain and face-to-face contact in borrowing | 219

speaker as well as the domain of borrowing. Thus, overall, Japanese borrowings appear to show the widest range of linguistic and social assimilation, from Gastwort to Fremdwort and further to Lehnwort. The chronologically newest English borrowings seem to fall into both Gastwort and Fremdwort, but since most of them remain distinguishable from the indigenous lexicon, they may not be yet categorized as Lehnwort. In summary, it has been shown that the level of bilingualism is a more important factor than the age of loanwords in understanding the linguistic assimilation of borrowing in multilingual Palau (i.e., Japanese Gastwort), while the domain where borrowing is adopted also seems to be a more crucial element than the recentness of the borrowing in gaining the social acceptance of borrowing as vernacular (i.e., core vocabulary from Japanese vs. missionary words from European languages).

4.3 Contact-induced borrowing scale in Palauan I now turn to applying the contact-induced borrowing scale to the Palauan data. I will first examine the degree of contact by looking at each social factor involved in the intensity of contact, such as “socioeconomic dominance”, “the duration of the contact period”, “the level and proportion of bilingualism” and “speakers’ attitudes” during each administration period. I will then see if the predicted type of borrowing is observable. As mentioned in section 2, the Spanish and German administrations are characterized as examples of symbolic domination. Hence, socioeconomic dominance would have been weak for three principle reasons: as a money economy had not been fully introduced; contact was short in time and less intensive in degree; and even individual bilinguals scarcely existed, although a respectful attitude toward Spanish as well as a positive attitude toward learning German were likely to have existed. Since the degree of contact appears to have been minimal, Palauan contact with Spanish and German seems to fit best into category one, “casual contact” in the borrowing scale (see Table 2). The predicted type of borrowing in that category neatly corresponds to Spanish and German borrowing in Palauan, i.e., “lexical borrowing only”. Table 4 shows the number and proportion by grammatical category of original words and phrases listed in the Palauan-English Dictionary (Joseph 1990) depending upon the source language. As the contact-induced borrowing scale (Table 2) predicts, nouns are most frequently borrowed amongst the grammatical categories in any language. As far as Spanish and German loan-

220 | Kazuko Matsumoto

words are concerned, their entire original grammatical category turned out to be nouns with only a few exceptions; one Spanish adjective (Sp. cristiano, Pal. keristiano, meaning ‘Christian’), one German adverb (Ger. auswendig, meaning ‘by heart’; Pal. chausbengdik (v.t.), meaning ‘know/learn thoroughly, memorize’) and one German verb (Ger. schenken17, ‘give as gift’; Pal. ousengk18 (v.i.), ‘give a gift to; thank’). It is interesting that this particular German adverb borrowed as a verb in Palauan corresponds to the German teaching method adopted during their regime (i.e., learning German songs by heart). Thus, in the case of Spanish and German borrowing in Palauan, the predictions that the borrowing scale makes seem to be robust. Tab. 4: Grammatical category of original words and phrases borrowed in Palauan (Source: Joseph 1990) Noun & Noun Phrase N Spanish German Japanese English

88 35 541 242

% 98.9 89.7 84.7 96.0

Adjective & Adj. Phrase N

% 1 0 33 2

1.1 0.0 5.2 0.8

Adverb N 0 1 4 0

% 0.0 2.6 0.6 0.0

Verb & V. Phrase N 0 2 57 8

Others

Total

% N % N % 0.0 0 0.0 89 100 5.1 1 2.6 39 100 8.9 4 0.6 639 100 3.2 0 0.0 252 100

By contrast, as section 2 describes, far more intensive contact with Japanese has led to bilingualism in a wide section of the population of Palau. During the Japanese administration, Japanese was used not only in the educational and administrative domains, but also in the work, commercial and neighborhood domains, since it was a settlement colony. Socioeconomic dominance of Japanese was strongly marked. Positive attitudes toward Japanese as well as Japanese borrowing have also been expressed by contemporary Palauans (see Matsumoto 2001 for details). Hence, Palauan contact with Japanese might appear to be applicable to category four, “intense contact”. However, given a rather short duration of the contact period (i.e., three decades), category three, “more intense contact” seems to be more appropriate. According to Thomason (2001), the predicted type of borrowing in category three appears to correspond approximately to Japanese borrowing in Palauan from the following three perspectives.

|| 17 As pointed out by Engelberg (2006: 6), it may also be derived from the German noun Geschenk meaning ‘gift’. 18 The Palauan prefix ou- “derives verbs that designate ownership or use of, control over, or participation in the entity designated by the stem itself” (Joseph 1984: 110).

The role of domain and face-to-face contact in borrowing | 221

Firstly, Table 4 highlights that Japanese borrowing shows a clearly different pattern from the others in that of all the non-nouns, only 15 out of 113 are not of Japanese origin. That is, Japanese borrowing includes a considerable number of adjectives, adverbs, verbs and interjections in addition to over 500 nouns and nominal phrases.19 Secondly, as described earlier, Japanese borrowing include numerous basic vocabulary items (e.g., body parts, numbers, words that express feelings and senses and baby talk) as well, which suggests that it falls into category three. Thirdly, as explained earlier, the important differences in linguistic consequence between categories two and three are whether new phonemes appear in borrowings or whether phonemicization of previously allophonic alternations occur in the recipient language. Joseph (1997) states that the influx of Japanese loanwords has in fact resulted in the establishment of five new phonemes in Palauan (i.e., /f/, /h/, /n/, /p/ and /z/). Their use was originally restricted to Japanese and then extended to Spanish and English loanwords. Taking /h/ as an example, among the three allophones that the Japanese phoneme /h/ has, the voiceless glottal fricative [h] has been established as the single allophone for the new Palauan phoneme /h/ (Joseph 1984: 85). Some Spanish loanwords with orthographic g or j (e.g., birhen, Sp. Virgen, ‘Virgin’) are currently pronounced either as the voiceless velar fricative [x] or [h], while English loanwords with /h/ have been mostly adopted in Palauan with /h/ (e.g., hos, Eng. hose) (Joseph 1984: 86). Thus, although a closer examination of their impact upon the pronunciation of Palauan lexical items is needed, the extended use of the new phonemes brought by Japanese into non-Japanese-originated words suggests that Palauan contact with Japanese falls into category three, which again supports the validity of predictions that the borrowing scale makes.

|| 19 The adverbs include daitai (Jp. daitai, meaning ‘about’), dosei (Jp. doose, ‘after all’), tokuni (Jp. toku ni, ‘in particular’) and sekkak (Jp. sekkaku, ‘with considerable effort’). Other interesting examples are both adjective phrases (negative) and verb phrases from Japanese. Both the negation -nai and the aspect form -teru are blended into Japanese loanwords in Palauan. Chauanai (Jp. awa-nai, meaning ‘does not match’), kikanai (Jp. kika-nai, ‘not effective’) and otsuriganai (Jp. o-tsuri ga nai, ‘have no change’) are examples of the former, while komatter (Jp. koma-tteiru, ‘be in trouble’), skareter (Jp. tsukare-teiru, ‘be tired’), and kangkeister (Jp. kankeishi-teiru, ‘be related’) are examples of the latter. Interesting is that Palauan has adopted the vernacular form -teru as -ter, rather than the standard form -teiru, suggesting that face-toface contact with Japanese settlers was a stronger impetus to absorb Japanese borrowing than the textbooks used in schools (see Matsumoto and Britain forthcoming b for a view on the vernacularity of Palauan Japanese).

222 | Kazuko Matsumoto

Neither syntax nor morphology alone seems to be borrowed from Japanese into Palauan. It may be worthwhile to note, however, that some Japanese morphemes were adopted as a part of the loanword, although they were not attached to Palauan words. For example, the Japanese prefix o- and suffix -san, both of which express politeness, are incorporated into many Japanese loanwords. As shown by okane (Jp. o-kane, meaning ‘money’), okiak (Jp. o-kyaku, ‘customer’) and chosoroi (Jp. o-soroi, ‘matching, identical’), o- is always attached to them. The latter, -san, is adopted as -sang in Palauan, often being attached to a personal name, occupation etc.; e.g., daikusang (Jp. daiku-san, meaning ‘carpenter’) and oningiosang (Jp. o-ningyō-san, ‘doll’). Thus, a bi-/multi-morphemic item in Japanese seems to be integrated as a mono-morpheme in Palauan. The older Palauans, who are fluent speakers or even rememberers of Japanese – they had in the past heard Japanese spoken but had never really learnt it – tend to be aware that both affixes signify politeness, while the young Palauans with little Japanese knowledge usually do not. Young Palauan gangsters would, therefore, demand okane rather than kane, although they are not asking for money politely! Older Palauans, on the other hand, used to make a joke when they saw me drinking mango juice, by saying “Enjoy o-mangko juice” to me. They pretended to be superficially polite by adding o- to the fruit mango, but also to be humorous as they knew the Japanese borrowing omangko (Jp. (o)manko) refers to ‘vagina’… Finally, in the case of contact with American English, the longest institutional bilingual policy in the educational and administrative domains led to Palauan-English bilingualism in the widest section of the population and positive attitudes toward English as well as Palauan-English code-switching (see Matsumoto 2001 for details), recently resulting in some English monolinguals among the young population in Koror. However, a crucial point is that everyday face-to-face contact with Americans hardly existed in Palau, although US grants and Palauans’ special status as US habitual residents brought some elite Palauans to the English language in the US. Thus, although it does not seem to be straightforward to measure the degree of contact with American English, given the prolonged contact period and the largest proportion of bilinguals, intensity of contact with American English may be category 3, “more intense contact”. Table 4 illustrates that English shows a similar pattern to the earlier European languages. Overall, almost all of the English lexicon is nouns, with the exception being eight verbs and two adjectives; e.g., saing (Eng. ‘sign’), stob (‘stop’) and chautomatik (‘automatic’). Examples of English borrowing at the sentence level, such as ikkamuu (Eng. ‘It can move’), which is used when one

The role of domain and face-to-face contact in borrowing | 223

tries to move something heavy, are not common. No basic vocabulary nor function words have been borrowed from English, suggesting that the type of borrowing is category one, “casual contact”. Thus, in the case of English borrowing in Palauan, there is a discrepancy between what the intensity of contact suggests (i.e., category three) and what the type of borrowing shows (i.e., category one), indicating the limits to the predictions that the contact-induced borrowing scale makes. Some cautions in applying the borrowing scale should be noted here. As Thomason (2001: 85) admits, “the fact that social conditions make a particular change possible does not mean that that change will inevitably happen: appropriate social settings … can never guarantee that particular changes, or any changes at all, will occur”. In particular, Thomason (2001: 61) calls speakers’ attitudes “the wild card”, as they “can be either barriers to change or promoters of change” (Thomason 2001: 85), causing “violations of most of the generally valid predictions about contactinduced change” (Thomason 2001: 61). Therefore, a further study of speakers’ attitudes may help us to gain an insight into English borrowing in Palauan. Moreover, as discussed earlier (see 3.1), this study supports Thomason’s (2001: 132–133) neutral position that, on the one hand, borrowing and code-switches are “two different things”, but on the other that they “lie at opposite ends of a continuum; the difference between them is not categorical”. Consequently, we included and analyzed potential code-switches that are closer to borrowing as examples of Gastwort but excluded code-switches furthest away from borrowing. A further exploration of the latter type of code-switch may reveal more about the contribution of English to Palauan. To sum up, the contact-induced borrowing scale proposed by Thomason & Kaufman (1988) has been found to be useful in understanding Palauan language contact with Spanish, German and Japanese, but less so with American English. Furthermore, despite the fact that positive attitudes seem to have existed towards each of the four source languages and that English has had a greater duration of contact and more widespread bilingualism than Japanese, it turns out that Japanese falls into the highest category in the borrowing scale among the four source languages. Thus, the overriding factor affecting the adoption and maintenance of borrowing in the case of Palau appears to be whether the islanders had face-to-face interaction in everyday domains.

224 | Kazuko Matsumoto

5 Conclusion and discussion This study highlighted the importance of the domains and face-to-face contact in order to understand the mechanism underlying the process of adopting and retaining borrowing from three perspectives. First, through examining the contact-induced borrowing scale in Palauan, we observed that Japanese has had the greatest influence upon Palauan. Japanese borrowing was found in a wider range of grammatical categories, included much larger chunks and established five new phonemes in Palauan. Second, through analysis of the linguistic assimilation of borrowing into Palauan, we are able to spotlight that in the case of a fluid multilingual speech community, the speakers’ ability in the source language affects the degree of linguistic assimilation, so that Gastwort is observable among Palauan-Japanese bilinguals, despite the fact that it was nearly a century ago that Japanese began being transferred to Palauan. Thirdly and perhaps most interestingly, through examining the social acceptance of borrowing by Palauans, we found that the perception of borrowing varies depending upon the domain in which the borrowing is adopted. The influx of civilian Japanese settlers led to everyday face-to-face interaction between Japanese and Palauans in the neighborhood, work and commercial domains, so that even Japanese core vocabulary, words expressing feelings and senses and nursery words have become part of Palauan, losing their foreignness. In short, although this is still speculative, borrowing may begin through indirect contact, but ultimately may need to reach local everyday interaction before core vocabulary and new phonemes are put in place in postcolonial multilingual communities. Thus, the crucial role of social infiltration and face-to-face contact in informal domains is highlighted.

The role of domain and face-to-face contact in borrowing | 225

References Abe, Goe. 1986. An ethnohistory of Palau under the Japanese colonial administration. Unpublished PhD Thesis, Kansas: University of Kansas. Anglim, John. 1988. Palau’s strategic position places Palauan democracy at risk. Working Paper 40. 1–42. Peace Research Centre, Research School of Pacific Studies, Australian National University, Canberra. Aoyagi, Machiko. 1977. Bunka hen’yō to shakai soshiki: Mikuroneshia, yappu to parao no baai [Culture changes and social organization: The case of Yap and Palau in Micronesia]. Shien 37(2). 43–65. Britain, David & Kazuko Matsumoto. 2015. Palauan English. In Jeffrey P. Williams, Edgar W. Schneider, Peter Trudgill & Daniel Schreier (eds.), Further studies in the lesser-known varieties of English, 305–343. Cambridge: Cambridge University Press. Engelberg, Stefan. 2006. The influence of German on the lexicon of Palauan and Kosraean. In Keith Allan (ed.), Selected Papers from the 2005 Conference of the Australian Linguistic Society. http://www.als.asn.au (checked February 2010). Fishman, Joshua A. 1972. Domains and the relationship between micro- and macrosociolinguistics. In John J. Gumperz & Dell Hymes (eds.), Directions in sociolinguistics: The ethnography of communication, 435–453. Oxford/New York: Basil Blackwell. Fujiwara, Yoichi. 1996. Nihon hogen jisho: Shōwa & Heisei no seikatsugo [Dictionary of Japanese dialects: Everyday words]. Vol. 1. Tokyo: Tōkyōdōshuppan. Hezel, Francis X. 1984. Schools in Micronesia prior to American administration. Pacific Studies 8. 95–111. Hezel, Francis X. & Mark L. Berg. 1980. Micronesia: Winds of change. Saipan: Trust Territory of the Pacific Islands. Hirayama, Teruo (ed.). 1992. Gendai nihon hōgen jiten [Dictionary of modern Japanese dialects]. Vol. 3. Tokyo: Meijishoin. Holmes, Janet. 2008. An introduction to sociolinguistics (3rd edn). London: Longman. Imaizumi, Yumiko. 1990. Nihon no gunseiki nan’yō guntō tōchi 1914–22 [Japanese naval administration in Micronesia 1914–22]. Kokusaikankeigaku Kenkyū [Studies on international relations] 17. 1–18. Department of Advanced Social and International Relations, University of Tokyo, Tokyo. Joseph, Lewis S. 1984. The impact of borrowing on Palauan. In Byron Wilbur Bender (ed.), Studies in Micronesian linguistics C-80. 81–123. Joseph, Lewis S. 1990. New Palauan-English dictionary. Honolulu: University of Hawaii Press. Joseph, Lewis S. 1997. Handbook of Palauan grammar. Vol. 1. Koror: Bureau of Curriculum & Instruction, Ministry of Education. Kluge, P. F. 1991. The edge of paradise: Americans in Micronesia. New York: Random House. Leibowitz, Arnold. 1996. Embattled island: Palau’s struggle for independence. Westport: Greenwood Publishing Group. Lev-Ari, Shiri, Marcela San Giacomo & Sharon Peperkamp. 2014. The effect of domain prestige and interlocutors’ bilingualism on loanword adaptations. Journal of Sociolinguistics 18(5). 658–684. Matsumoto, Kazuko. 2001. Multilingualism in Palau: Language contact with Japanese and English. In Thomas E. McAuley (ed.), Language change in East Asia, 84–142. London/New York: Routledge.

226 | Kazuko Matsumoto

Matsumoto, Kazuko. 2010. Palauan language contact and change: A sociolinguistic analysis of borrowing in Palauan. In Jelisava Dobovsek-Sethna, Frances Fister-Stoga & Cary Duval (eds.), Linguapax Asia: A retrospective edition of language and human rights issues, collected proceedings of Linguapax Asia Symposia 2004–2009, 36–53. Tokyo: Linguapax Asia. Matsumoto, Kazuko. 2011. Contact linguistics of diaspora Japanese: Linguistic innovation and attrition in the Western Pacific. Language, Information, Text 18. 33–61. Tokyo: University of Tokyo. Matsumoto, Kazuko. 2013. Cake baking in a postcolonial Japanese speech community: Koineization of Palauan Japanese in the Pacific. Language, Information, Text 20. 63–85. Tokyo: University of Tokyo. Matsumoto, Kazuko & David Britain. 2003a. Language choice and cultural hegemony: Linguistic symbols of domination and resistance in Palau. In Mirjana Dedaic & Dan N. Nelson (eds.), At war with words (Language, Power and Social Process 10), 315–358. Berlin: Mouton de Gruyter. Matsumoto, Kazuko & David Britain. 2003b. Contact and obsolescence in a diaspora variety of Japanese: The case of Palau in Micronesia. Essex Research Reports in Linguistics 44.38–75. Matsumoto, Kazuko & David Britain. 2006. Palau: Language situation. In Keith Brown (ed.), Encyclopedia of Language and Linguistics (2nd edn.) Vol. 9, 129–130. Elsevier: Oxford. Matsumoto, Kazuko & David Britain. forthcoming a. Food-related lexical borrowings as indicators of the intensity of language contact in the Pacific. In Siria Guzzo & Giuseppe Balirano (eds.), Food and diasporic communities of practice. London: Palgrave Macmillan. Matsumoto, Kazuko & David Britain. forthcoming b. The vernacularity of Palauan Japanese. In Kazuko Matsumoto & David Britain (eds.), The sociolinguistics of the Japanese diaspora, Special Issue of International Journal of the Sociology of Language. McArthur, Tom. 1992. The Oxford companion to the English language. Oxford: Oxford University Press. McKinney, Robert. 1947. Micronesia under German rule 1885–1914. Unpublished MA Dissertation, Stanford University, California. McManus, Edwin G. 1977. Palauan-English dictionary. Honolulu: University of Hawaii Press. Miyagi, Kimi. 2000. Japanese loanwords in Pohnpeian: Adaptation and attrition. Japanese Linguistics 7. 114–132. Mühlhäusler, Peter. 1996. German in the Pacific area. In Stephen A. Wurm, Peter Mühlhäusler & Darrell T. Tryon (eds.), Atlas of languages of intercultural communication in the Pacific, Asia and the Americas. Volume II: 1: Texts, 345–352. Berlin: Mouton de Gruyter. Mühlhäusler, Peter & Philip Baker. 1996. English-derived contact languages in the Pacific in the 20th century. In Stephen A. Wurm, Peter Mühlhäusler & Darrell T. Tryon (eds.), Atlas of languages of intercultural communication in the Pacific, Asia and the Americas. Volume II: 1: Texts, 497–522. Berlin: Mouton de Gruyter. Nan’yō-chō [South Seas Bureau]. 1928. 1939. 1941. 1942. Nan’yō guntō yōran [Annual report on Micronesia]. Tokyo. Nihon kokugo dai-jiten [Dictionary of Japanese]. 2001. (2nd ed.) Tokyo: Shōgakkan. Office of Planning and Statistics. 2005. 2002–2003 Census of population and housing of the Republic of Palau. Koror. Office of Planning and Statistics. 2006. 2005 Census, Vol. II: Census monograph population and housing Profile. Koror.

The role of domain and face-to-face contact in borrowing | 227

Palau Community Action Agency. 1977. A history of Palau: Traders and whalers, Spanish administration, German administration. Vol. 2. Koror. Palau Community Action Agency. 1978. A history of Palau: Japanese administration, US Naval Government. Vol. 3. Koror. Peattie, Mark R. 1988. Nan’yō: The rise and fall of the Japanese in Micronesia 1885–1945. Honolulu: University of Hawaii Press. Rechebei, Elizabeth D. & Samuel F. McPhetres. 1997. History of Palau: Heritage of an emerging nation. Koror: Ministry of Education. Rinji Nan’yō Guntō Bōbitai [Provisional South Seas Defence Corps]. 1922. Nan’yō Guntō Tōsei Chōsa Hōkoku [Investigation Report on Micronesia]. Tokyo. Romaine, Suzanne. 1995. Bilingualism (2nd edn.). Oxford: Blackwell. Schneider, Edgar. 2007. Postcolonial Englishes. Cambridge: Cambridge University Press. Shuster, Donald R. 1978. Major patterns of social change instituted in Micronesia during Japanese colonial rule, 1914–1940. Unpublished Term Paper, University of Hawaii, Manoa. Shuster, Donald R. 1982. Islands of change in Palau: Church, school, and elected government, 1891–1981. Unpublished PhD Thesis, University of Hawaii, Manoa. Thomason, Sarah G. 2001. Language contact: An introduction. Edinburgh: Edinburgh University Press. Thomason, Sarah G. & Terrence Kaufman. 1988. Language contact, creolization, and genetic linguistics. Oxford: University of California Press. Trifonovitch, Gregory J. 1971. Trust territory of the Pacific Islands. In Thomas Sebeok (ed.), Current Trends in Linguistics, vol. 8: Linguistics in Oceania, 1063–1087. The Hague: Mouton.

Norbert Schaffeld

Representing bicultural knowledge systems and epistemic decolonization The role of place, language, and metaphor in Yothu Yindi’s “Treaty” Abstract: By taking the song “Treaty” as a characteristic example, this article explores the ways in which the crucial link between place, language, and metaphor has exerted a profound impact on Yothu Yindi, the famous Australian rock band which combines Aboriginal and non-Aboriginal members. In order to formulate a political vision, the band’s bilingual narrations of place contain mythological concepts and metaphors whose tenor underscores the intended co-existence of different, but equal cultures and knowledge systems. To fully understand the corresponding process of epistemic decolonization, the mythical significance of the land and its decentered rhizomatic network will be discussed in greater detail. Keywords: colonial and post-colonial place, toponomic representation, rhizomatic knowledge, bicultural knowledge, epistemic decolonization, Yothu Yindi

1 Colonial and post-colonial places In one of his more recent books, Caliban’s Voice, published in 2009, Bill Ashcroft freely borrows Saussurean terms to conceptualize the difference between space and place, a distinction that seems to be a specific trademark of the English language in that it sets apart “space as unbounded extension” from “place as a location” (Ashcroft 2009: 75). It is with an informed sense of the role of language in spatial discourse that Ashcroft ventures to draw the following parallel: Place is the equivalent of parole – utterance itself rather than the potentiality of utterance that is space. Place, we might say, is uttered into being and maintained by narrative. (Ashcroft 2009: 75)

|| Norbert Schaffeld: University of Bremen, Linguistics and Literary Studies, English Literature, Postfach 330440, 28334 Bremen, Germany. E-Mail: [email protected]

230 | Norbert Schaffeld

With this ontological relation between place and narrative in mind, this paper will address three major phenomena of a spatial knowledge system. Firstly, the Europeans’ notion of Australia as empty space will be juxtaposed with the mythical significance the land had and still has for its Aboriginal peoples. Focusing on the textual nature of the land, different readings of a decentered rhizomatic system will here be discussed in view of the spatial knowledge system of the indigenous population. Secondly, the totemic sites will be classified as meaningful parts of a rhizomatic network that goes well beyond Deleuze and Guattari’s model. And thirdly, the immensely popular song “Treaty” from the rock band Yothu Yindi will be taken as a helpful example to illustrate the use of a rhizomatic, i.e. a non-hierarchical, key metaphor for bicultural knowledge and epistemic decolonization. When in 1770 Captain James Cook landed in Botany Bay, the territory of the Eora people, he immediately took possession of the entire east coast of New Holland (Australia) by declaring British sovereignty over the land. Later on the fundamental legal principle applied here came to be known as the doctrine of terra nullius. While the term appears not to have been in use during the first decades of Australian occupation, its validity nevertheless remains unchallenged, as it is quite common to denote a historical phenomenon with a lexical item not contemporaneous with the period under scrutiny. Here I borrow a line of argumentation put forward by James Warden who goes on to remind us that “[t]he fact that later historians adopt or invent a term to express an ensemble of beliefs, practices and events that can be discerned in the past is one of the organizing principles of history writing.” (Warden 2006, 49.3–49.4) Indeed, the term terra nullius appears to embrace denotative as well as connotative meanings whose scope should be taken into account when referring to its historiographical, legal and political currency in Australian culture. It is from an Aboriginal point of view that Colin Bourke and Helen Cox specify the conditions under which the doctrine was adopted even before, one might want to add, the word entered the legal discourse of 19th century Australia: Under the doctrine [of terra nullius] colonising powers such as England could apply their own law to land which they peacefully occupied, if the land was uninhabited, or was occupied by a people without settled laws or customs. In conquered or ceded (surrendered) countries the pre-existing laws of that country were applied until they were displaced or altered by the new sovereign. (Bourke & Cox 2009: 59)

Since the Aboriginal inhabitants of Australia were not seen as a people with settled laws, the whole continent was soon classified as belonging to no one. In the process of colonialism what was taken to be empty space from a European

Representing bicultural knowledge systems and epistemic decolonization | 231

or Lockean perspective (Frost 1981 & Connor 2002) was rapidly transformed into place by the three linguistic operations of erasing, naming and narrating (Ashcroft 2009: 75). In this context mapping or cartography can be seen as a strong instrument of discursive control which turns what was readily classified as empty space without native title into colonial place, that is, into a textualized entity, capable of both signifying difference and constructing identity (Ashcroft 2009: 76). Here, the textual nature of place involves the dialogue between the different utterances relating to the site as well as a pattern of meaning which first and foremost constitutes its genuine textuality. But there is an immediate problem. What the Europeans were rather keen to define as empty space was in fact already invested with ancestral or mythical significance. As far as Australia is concerned, this involves a deep-rooted pre-contact tradition of 40,000 or maybe 60,000 years, in the course of which the place became rich in indigenous meaning. In the past, post-colonial theory has often made full use of the metaphor of the palimpsest to illustrate a stratified textuality of place and its sequence of consecutive erasures (Carter 2010). Whenever and wherever Aboriginal narratives of place were able to question a dominant layer, it was simply seen as an act of indigenous re-inscription. Yet, Ashcroft pointedly asks, how can one describe such a phenomenon when the metaphor being used establishes the process of erasure as a prerequisite of constituting place? In other words: Can what has been erased in the colonial endeavor come to the fore later on, and if so what are the terminological options to give it metaphorical shape? For Ashcroft the concept of the palimpsest denotes the period of colonization, while it is the metaphor of the rhizome, or, more precisely, the idea to see place as a rhizomatic or subterranean text which allows for post-colonial transformations “vertically through time as well as laterally in space.” (Ashcroft 2009: 77) By way of leading up to this thesis, Ashcroft notes that the metaphor of the palimpsest has been extraordinarily successful in elaborating the textuality of place – a text on which previous inscriptions have been erased but remain as traces in the present. But the palimpsest suggests that place can be brought into the world only be erasure. It visualizes the text as a flat plane, a misleading view of the temporal continuity of the struggle out of which an experience of place emerges. The insistence on place as a rhizome rather than palimpsestic text reveals that representations of place are always a potential product of the dis-articulated resistances and transformations of the inhabitants, the province of the hybrid, the mundane, the quotidian. (Ashcroft 2009: 77)

A decade before, Ashcroft had already shown a preference for the term rhizome when he used the metaphor to provide a critique of “the centre/margin binarism which imperialism constructs.” (Ashcroft 1999: 116) For him the “[r]hizome is a botanical term for a root system which spreads across the ground (as in bam-

232 | Norbert Schaffeld

boo) rather than downwards, and grows from several points rather than a single tap root” (Ashcroft 1999: 116). Much of this article is then devoted to revealing the rhizomatic, and by no means monolithic operation of power; the latter he considers to be a myth and hence by necessity ineffectively addressed by the long-cherished categories of marginality that critical reading abounds with (Ashcroft 1999: 116–117). According to Ashcroft this myth wrongly asserts the existence of one tap root and it is this metaphorical marker of centrality that elsewhere might be called the standard or the canon (Ashcroft 1999: 117). It follows that imperialism only appears to be monolithic, precisely because its rhizomatic or interlinked centers have agreed to propagate a “consensus about its vertical structure” (Ashcroft 1999: 118). Once an imperial discourse is established as a seemingly monolithic entity, every form of resistance is both targeted and polarized, as it eventually becomes one element of a very basic binarism – the colonized and the colonizers, black and white. Indeed, it is the logic of this argument that if post-colonial identity is solely determined by race, settlercolonies such as Australia will be excluded from the status of post-coloniality. Against the grain of such a view Ashcroft asserts the existence of a social discourse whose diversity is hidden in the face of its imperial counterpart so that there are at least two rhizomes, “the imperial and the repressed, the colonial and the post-colonial” (Ashcroft 1999: 119). Within this theoretical framework he finally favors a poetics and a politics of transformation which presuppose a colonized societies’ appropriation of imperial discourses and replace a simplified imagery, for example the trunk and branch myth, with attributing a rhizomatic quality and diversity to both colonial power and post-colonial resistance (Ashcroft 1999: 121). While Ashcroft locates the metaphor of the rhizome at the centre of a theoretical debate that seeks to establish the post-colonial status of the former settler colony Australia, its use follows a different track the moment the term is applied in the field of Aboriginal Studies. But before entering into the complexities of that realm, a few thoughts on Deleuze & Guattari (2011), who popularized the concept, seem to be in order. To begin with, both philosophers identify the rhizome as a concept that refers to “a process of networked, relational and transversal thought” (Colman 2007: 231) which due to its moving matrix can neither be seen as a closed entity nor as a closed system of thought. By definition the rhizomatic network is decentered, it does not allow for hierarchies or those binary approaches that Deleuze & Guattari associate with arborescent Western thought, i.e. with tree-like structures and their inbuilt epistemological taxonomy. In their seminal book A thousand plateaus, Deleuze & Guattari conceptualize the rhizome as an open subterranean stem system whose principal features

Representing bicultural knowledge systems and epistemic decolonization | 233

encompass its potential connection to anything other as well as its capacity of asignifying [sic] or non-signifying rupture, that is to say, any cut in a given line will lead to a restart at old or new spots (Deleuze & Guattari 2011: 7–10). Rhizomes appear to accord to semiotic chains and they resemble a performative map with multiple entryways and connections (Deleuze & Guattari 2011: 14). While the arborescent cultures of the West are likened to “hierarchical systems with centers of significance and subjectification” (Deleuze & Guattari 2011: 18), the two philosophers regard Oceania as the model region of rhizomorphous cultures when they pose the rhetorical question: Does not the East, Oceania in particular, offer something like a rhizomatic model opposed in every respect to the Western model of the tree? André Haudricourt even sees this as the basis for the opposition between the moralities or philosophies of transcendence dear to the West and the immanent ones to the East […]. (Deleuze & Guattari 2011: 20)

It is immediately obvious that the regional localization of Western arborescent thought and Oceanian rhizomatic culture can only operate as a mere theoretical device of demarcation and dissociation. Any colonial encounter in this part of the world and elsewhere in the settler-colonies was accompanied by the transplantation of a tree-like epistemological hierarchy that confronted rhizomatic knowledge with the vigor and supplanting power of an aggressive neophyte. As we know the uneven clash of the two knowledge systems found many sites of asymmetric negotiation, yet one stands out as the most prominent issue because it is inextricably related to the significance of the land. At the end of this chapter and before dealing with Aboriginal spatial cosmology in some more detail, the argument leads to the preliminary conclusion that the Aboriginal places and the knowledge attached to them in some respects resemble a rhizomatic system that is still facing a discursive struggle over ownership and toponomic representation. Although arguing from a slightly different angle, Ashcroft likewise concludes that “Post-colonized place is therefore a site of struggle on which the values and beliefs of indigenous and colonizer contend for possession. Ultimately this possession occurs in language” (Ashcroft 2009: 77).

2 Aboriginal spatial cosmology It is with an informed sense of the practice of restricted access to indigenous knowledge, my own hermeneutic limits therein and the pitfalls of cultural appropriation that I would like to recall the fact that I am not a member of an Aboriginal community. Thus, the only view to be followed here is admittedly a par-

234 | Norbert Schaffeld

tial one from outside and I am fully aware of its limitations. My perspective comes close to what the American anthropologist Clifford Geertz (2000: 57–58) called the experience-distant concept, an approach to be substantially complemented by more authentic or experience-near accounts should they be open to the non-initiated and part of a published or oral discourse. But since the indigenous agents of the gamma-discourse, which is discussed in more detail later on, advocate a bicultural form of exchange, the corresponding design of this paper might ensure an appropriate academic positioning. For the Aboriginal people of Australia the mythical Dreaming or Alchera claims the status of irrefutable significance as a unity of living myths with an eternal or unfading quality.1 While it is true that in the past some efforts have been made to define the Australian Aborigine as homo religiosus (Thiel 1984: 13), classifications such as this one are misleading in that they imply a belief in a central avataric figure (Cowan 1989: 40). Indeed, more useful in this respect is James Cowan’s approach, which is based on the Dreaming’s sharing of permanent truth: But what the Dreaming does provide is knowledge of the Primordial Event [ … ] in a way that denotes the possibility of full participation in it by the individual. In this sense the Dreaming provides Aborigines with a metaphysical discipline without the need to assume for itself a spiritual primacy as is the case with most other revelationary religions […]. (Cowan 1989: 40)

Although the Dreaming constitutes a common mythical property of all Aborigines, it does involve different terminologies and manifestations depending on the tribe’s linguistic and cultural stock. Howsoever the concept is named, its metaphysical contents express “primordial truths that trace the birth of the world and man’s place in it” (Cowan 1989: 2). The myths of creation do not describe a process that started in a huge void, rather they evoke a significant transformation of a plain, or, more simply, of a featureless and infinite space. In the cosmology of the first Australians, uncreated primordial beings took on the final formation of the landscape by creating and naming places linked among themselves (Rumsey 2001: 11). This landscape of totemic reference is always seen in its totality comprising geological features, flora and fauna as well as man himself. In the course of this transformation, which terminates the Dreamtime proper, the totemic ancestors left dreaming tracks that bear their personal

|| 1 Fiona Magowan, however, claims that in practice “the concepts of the Dreamtime are adapted for each circumstance and context, constantly being redefined as necessity to external pressure dictates.” (Magowan 1997: 137)

Representing bicultural knowledge systems and epistemic decolonization | 235

signature “in the guise of topographic landmarks, contour variations, trees, animals – in fact, all manifestations of life on earth.” (Cowan 1989: 25–26) It is, however, important to note that Aboriginal people, for example the Aranda of Central Australia, see the landmarks as totemic signatures at the same time as they identify them as the actual bodies of the primordial beings (Strehlow 1947: 28). In an earlier Aranda ethnography, Spencer and Gillen had already given information about the movements of primordial beings, however according to them the topographic features only mark the body of the totemic ancestor who was believed to have actually died on the specific site (Rumsey 2001: 36). The Yolŋu from Arnhem Land, who will be central for this article, see their ancestral progenitors (waŋarr) as sentient and permanently present parts of the topographic land (Corn 2010: 85). Though ethnographic views from outside tend to deviate from each other, then only because of the different indigenous tribes they focus on; Strehlow’s seminal work on Aranda traditions stands out in that even decades after its first publication, it has not lost relevance in revealing the complex phenomenon of a mythological landscape that is capable of simultaneously hosting the spatial equation of the sign and the body, of signifier and signified. The quality of a Golden Age or that of a paradise, which an occasionally sympathetic Western culture attributes to the Dreamtime, are nothing but hetero-stereotypical interpretations that do not find any parallel in the cosmology of Aboriginal people. Indeed, the crucial part of the concept remains the act of creation, the initial event which has been inscribed into the landscape by means of a life-essence called kurunba. This is the work of the primordial beings, who combine totemic, human and animal characteristics. They have left the lifeessence in the topographical formations as a kind of cultural texture. By way of illustrating this view, Cowan notes that the spiritual landscape of Aboriginal people appears to be endowed with an iconicity, in which each sign relates to a metaphysical referent: In other words, the landform has become iconic in essence, fulfilling a role of containment, not only of physical attributes (shape, texture, mineral content etc.), but of metaphysical significations. It is this quality that Aborigines term as kurunba – that is, the power that gives a landmark its inherent form over and above that of its mere physical presence. (Cowan 1989: 26)

This is where Cowan’s understanding of the mythical landscape at least partly coincides with a broadly post-modern concept of the text, inasmuch as the latter

236 | Norbert Schaffeld

can be regarded as one cultural practice of signification.2 Likewise, Ronald and Catherine Berndt have compared the “physiographic sites and places” of the Aboriginal landscape to the “chapter headings of a book” (Berndt & Berndt 1989: 6), which – one could add – mark a complicated network of metaphysical meaning. At the same time, however, this mnemonic land facilitates the dialogue with primordial beings, which, assisted by ritual acts, ceremonies and performances, perpetuates the actual Dreamtime into the here and now. To take the main argument of this paper further, I would like to borrow and extend Rumsey’s thoughts on topographic objectification (Rumsey 2001: 19) by classifying the totemic sites as linked places invested with topographic, toponymic and mythological qualities and thus as meaningful parts of a rhizomatic network. But to what extent does this indigenous rhizomatic web correspond to Deleuze and Guattari’s model? Their writings chart a course of clear-cut opposition between rhizome and arborescence, between nomadic socioterritorial and centralized, state-based regimes (Rumsey 2001: 19–20). With these binary oppositions Deleuze & Guattari propose what amounts to a discursive philosophical agency, which proves itself incapable of redressing arborescent thought, or, more precisely, the dominant mode of Western knowledge systems. Yet, they seem to be fully aware of this epistemic pitfall when they somewhat surprisingly concede: We invoke one dualism only in order to challenge another. We employ a dualism of models only in order to arrive at a process that challenges all models. Each time, mental correctives are necessary to undo the dualisms we had no wish to construct but through which we pass. (Deleuze & Guattari 2011: 22–23)

This admission by no means disarms the critics who, like Rumsey, pointedly ask “whether such simplistic dualism as theirs was really necessary in order to found their critique” (Rumsey 2001: 41). While it is widely accepted that the sets of knowledge attached to the sacred sites as well as their spatial distribution are non-hierarchical, rhizome-like entities whose narratives are based on the textual nature of the land, Deleuze & Guattari’s placing of the tree as the exclusive model of Western thought fails to account for a wider use of the metaphor. Focusing on the life cycle of the banyan tree and the strangler figs, Rumsey explains to what extent the rhizomatic mode can indeed become a feature of these

|| 2 Broadening this theoretical supposition, the editors of the book Writing Worlds argue that “places are intertextual sites because various texts and discursive practices based on previous texts are deeply inscribed in their land­scapes and institutions” (Barnes & Duncan 1992: 7–8).

Representing bicultural knowledge systems and epistemic decolonization | 237

specific trees so that the thought of an oppositional pair turns out to be misleading. He argues as follows: In the developmental cycle of both the banyan tree and the strangler figs, there is a confounding of the one-way “hierarchical” relation between trunk and branch in that the aerial roots that grow from branches can give rise to new trunks, which themselves give rise to new branches, and so on. Furthermore, the aerial roots, after reaching the ground, can reconnect with the root system that is already in place and nourishing the branch from which they are growing. (Rumsey 2001: 33)

This description stands in stark contrast to Deleuze & Guattari’s thought as it describes the rhizome as the defining property of one of the key-trees in Aboriginal mythology. Quite often the banyan tree marks a sacred site, a mythological place that objectifies a collective identity and ownership in a natural site. In this sacred nature, the banyan tree is closely connected to the early phase of the land rights movements in Australia, especially those of Arnhem Land.

3 Epistemic decolonization In a memorial lecture held in 1998 in the town of Yirrkala on the Gove Peninsula, located at the north-eastern corner of Arnhem Land, Galarrwuy Yunupiŋu, an elder of the Gumatj group points out the significance of the banyan tree as a sacred site while linking the future of this specific landmark to an important legal issue. It was in 1963 that the Australian government granted a bauxite lease over the Aboriginal land close to Yirrkala without asking permission from the traditional owners, the Yolŋu people. On the occasion of the lecture Galarrwuy Yunupiŋu said: I suppose I was 16 when the news travelled urgently throughout Yirrkala that the sacred banyan tree at Nhulunbuy was going to be damaged by the mining company. There was a rippling anger that shot through the old people. People called out in the high upset tones of my people when they are under threat. Old men I knew to be strong and courageous and afraid of nothing were shaken. […] That tree is a special place – inside it are important things. It’s like heart of the country – our beliefs about our land reside in that tree and at the site of the tree, they reside in the rocks, in the water and in our minds. We know these things to be true. When I was very young my father would take me to this place of the banyan tree. We would always stop here as we walked across Gumatj land. The spirit marked by that tree was respected and it was felt truly in our hearts. (Yunupiŋu 1998: 3)

238 | Norbert Schaffeld

From the early 1960s onwards protest and resistance against the land policy of the Australian government took many different shapes. Among the most prominent was the so called “Yirrkala Bark Petition”, composed of two bark panels with a bilingual statement in Yolŋu-Matha and English that asked the government to spare Aboriginal land. Both panels are now permanent exhibits in Parliament House in Canberra (Corn 2010: 96). Since Galarrwuy Yunupiŋu’s father was an active member of Aboriginal resistance, it comes as no surprise that his children were familiar with the land rights movement from an early age on. Thus, Yunupiŋu’s younger brother, the late Dr Yunupiŋu,3 who eventually became the front man of the Aboriginal rock band Yothu Yindi, must have been heavily influenced by paternal setbacks and disappointments when the clan’s rights were altogether ignored.4 Today, the Yolŋu encompass about 7,000 inhabitants who speak YolŋuMatha5 [human speech], itself a group of seven mutually unintelligible languages, which can also be differentiated by a discrete lexicon of esoteric yäku [names] (Corn 2010: 84). In exposing the crucial link between language, land and identity, Corn identifies the lexicon of esoteric yäku as “the most valuable traditional asset any mala [group] can possess other than country itself” and he further explains that “[e]ach mala owns at least one wäŋa (‘country’, ‘homeland’) incorporating all lands, waters and natural resources therein, in addition to an incumbent canon of hereditary yäku (‘names’), manikay (‘songs’), buŋgul (‘dances’) and miny’tji (‘designs’). (Corn 2010: 85) In Yolŋu epistemological knowledge is subdivided into classes to which access is open (garma), sheltered (dhuni’) or restricted (ŋärra’), and it is again the tree that is used as a convenient metaphor when referring to public and restricted spheres of knowledge. While the tree as such represents public knowledge, its subterranean roots stand for forms of restricted knowledge, and it is their covert embeddedness that they

|| 3 Dr Yunupiŋu died in June 2013. Because of cultural sensitivities of the Aboriginal community his first name will not be used in this paper. 4 In a legal case held in 1992, the High Court of Australia delivered the so-called Mabo decision, which was regarded as a breakthrough in the land rights debate as it rejected the established doctrine of terra nullius for the first time. Alan Rumsey explains the legal consequences: “Under the new disposition, now formalized by the Native Title Act of 1993 (amended in 1998), the potential existence of residual ‘native title’ is recognized as a matter of law, but its actual existence remains to be demonstrated as a matter of fact in any given case that comes before the court.” (Rumsey 2001: 39) 5 The text of this paper observes the conventions for Yolŋu-Matha spelling as they are given in R. Zorc’s Yolŋu-Matha Dictionary (1996). All quotes throughout the text are of course printed in their original version.

Representing bicultural knowledge systems and epistemic decolonization | 239

share with the waŋarr, the ancestral progenitors of the Dreaming who are present in the here and now (Corn 2010: 85). Towards the end of the twentieth century and in the immediate context of the Australian Bicentenary in 1988 and the Sydney 2000 Summer Olympics, a trend within contemporary Aboriginal culture emerged with an aim at bridging the divide between European and Aboriginal Australians. It challenged segregation practices and advocated the rhetoric of biculturalism and public events of performative convergence. Perhaps it may come as a surprise that here popular culture is given pride of place, in this case the songs of the ethno-rock band Yothu Yindi from the Northern Territory of Australia.6 There are, however, at least two reasons for doing so. First, the band’s productions at that time, for example the two albums Tribal Voice (1992) and Freedom (1993), carry on the tradition of song and performance dominant in Yolŋu culture by using innovative stylistic devices. And second, Yothu Yindi7 not only integrate narrations of place in their bilingual texts, they have also formulated a political Aboriginal vision with a world-wide reception, platinum sales and numerous awards – something never achieved before. The band’s multiple instrumentation brings electronic technology, e.g. electric guitars, keyboards and drumkits, together with traditional instruments, such as the didjeridu (yidaki) and ironwood clapsticks (bilma) to reflect the mixed Aboriginal and non-Aboriginal composition of the band. It also augments the effect of the bilinguality of the songs, English and the tribal Gumatj (Rirratjingu), that is, the bilingual lyrics come to represent a bicultural approach. At this point, it is worth mentioning that the band’s lead singer and songwriter, Mr Yunupiŋu, was among the first Yolŋu educators to have achieved an academic degree. He also became Australian of the Year in 1992 for his achievements in advancing and intensifying understanding between Aboriginal and non-Aboriginal people. Indeed, part of his song writing is devoted to “shaping and theorizing his ideas about new ways of schooling Yolŋu children biculturally” (Corn 2010: 88). The objective of this curriculum should not be confused with the policy of assimilation as it is meant to establish equal cultural and political status. In the face of the political content of his song-poetry, criticism of a possible commercial orientation towards a Western popular taste re-

|| 6 Yothu Yindi were formed in Darwin in 1986 by three Yolŋu men from Yirrkala and the Gove Peninsula (Arnhem Land). The trio was joined by three rock musicians. 7 Yothu Yindi means child and its mother. See Stubington & Dunbar-Hall (1994: 251). Fiona Magowan further differentiates between egocentric and sociocentric dimensions of the dyad. See Magowan (1997: 138).

240 | Norbert Schaffeld

mains beside the point. In the songs “Treaty”, “Tribal Voice”, “Mainstream”, “Timeless Land”, “Gunitjpirr Man”, and finally “Mabo”, Yothu Yindi support the Aboriginal civil and land rights movement, which was able to attract international attention during the 1988 Bicentenary, the two hundredth anniversary of the founding of the British colony, or, to put it differently, the invasion of Australia. The declarations of intent made by the Australian government on the occasion of the Bicentenary in 1988 are the subject of the song “Treaty” (Corn 2011: 22–23). In that year Prime Minister Robert Hawke promised within his administration a treaty would be signed, confirming prior indigenous ownership, land sovereignty, and the affirmation of human rights, among them linguistic rights (Bourke 2009: 48–49). But as this promise was soon forgotten, Yothu Yindi got together with Peter Garrett8 of the Australian rock band Midnight Oil, the indigenous musician Bart Willoughby, and the musician Paul Kelly to create the song “Treaty”, which legitimately asks for the fulfillment of the promised legal arrangement. The lyrics run as follows: Well I heard it on the radio – And I saw it on the television – Back in 1988 – All those talking politicians – Words are easy, words are cheap – Much cheaper than our priceless land – But promises can disappear – Just like writing in the sand – Treaty Yeh Treaty Now Treaty Yeh Treaty Now – Nhima Djatpangarri nhima walangwalang (You dance djatpangarri, that’s better) – Nhe Djatpayatpa nhima gaya nhe Matjini … Yakarray (You’re dancing, you improvise, you keep going, wow) – Nhe Djat’pa nhe walang (You dance djatpangarri, that’s good) – Gumurrtjararrk Gutjuk (My dear paternal grandson) – This land was never given up – This land was never bought and sold – The planting of the union jack – Never changed our law at all – Now two rivers run their course – Separated for so long – I’m dreaming of a brighter day – When the waters will be one – Treaty Yeh Treaty Now Treaty Yeh Treaty Now – Treaty Yeh Treaty Now Treaty Yeh Treaty Now – Nhema Gayakaya nhe gaya nhe (You improvise, you improvise) – Nhe gaya nhe matjini walangwalang nhe ya (You improvise, you keep going, you’re better) – Nhima djatpa nhe walang (You dance djatpangarri, that’s good) – gumurrtjararrk yawirriny (My dear young men) – Nhe gaya nhe matjini gaya nhe matjini (You improvise, you keep improvising, you keep going) – Gaya gaya nhe gaya nhe matjini walangwalang (Improvise, you improvise, you keep going, that’s better) – Nhema djatpa nhe walang (You dance djatpangarri, that’s good) – Nhe gumurrtjarrk nhe ya (You dear things) – Promises – Disappear – Priceless land – Destiny – Well I heard it on the radio – And I saw it on the television – But promises can disappear – Just like writing in the sand – Treaty Yeh Treaty Now Treaty Yeh Treaty Now – Treaty Yeh

|| 8 Peter Garrett was a member of the House of Representatives from 2004 to 2013. In 2007 he was appointed Minister for the Environment, Heritage and the Arts and in 2010 he became Minister for School Education, Early Childhood and Youth.

Representing bicultural knowledge systems and epistemic decolonization | 241

Treaty Now Treaty Yeh Treaty Now – Treaty Yeah Treaty Ma Treaty Yeah Treaty Ma – Treaty Yeh Treaty Ma Treaty Yeh Treaty Ma9

Any spatialized reading of the song’s lyrics would probably foreground three major references whose order of appearance reflects spatial concepts of the precontact, the contact and the post-colonial phase. In particular, they relate to the spiritual landscape of the Aboriginal people, to the European doctrine of terra nullius and, finally, to a dreamlike location of metaphorical biculturalism. To begin with, Yothu Yindi appraise the land to be priceless, precisely because as a sacredly inscribed text it is beyond economic valuation. The Aboriginal people do not consider themselves to be private owners of the land in the Western sense of the word possession, but as a people collectively responsible for the integrity of the landscape. For group owners of a land that is composed of a rhizomatic text of meta-physical significations (Cowan 1989: 26), the very idea of giving consent to its destruction would, even from a neutral perspective, clearly correspond to a scheme, which interferes with the work and the presence of the ancestral progenitors. Yet, for the British colonizers, the Nomadic society’s lack of a sovereign and their apparent disinterest in landed property were the two major criteria to declare this land to be terra nullius, an empty space without property rights that otherwise might have been respected as was the case, for instance, in New Zealand.10 Today, the nullification of pre-colonized place, practiced by the British, would still be evident in data sets of historical documents where corpus linguists will come across the following words significantly prefixed with “un-“, i.e. unawakened, uncleared, undescribed, undiscovered, unexplored, unfamiliar, ungrazed, uninhabited, unknown, unnamed, unoccupied, unpeopled, unproductive, unsettled, untamed.11 In terms of a future almost utopian vision, surely the most important spatial reference is the metaphor of the two rivers that after a long time of separation will flow together in a dreamlike time to come. In this figure of speech the vehicle of the two rivers reflects a Yolŋu metaphor called ganma in the north-eastern Arnhem Land languages. Traditionally, ganma was coterminous with a lagoon, a site where fresh and salt water meet. Stubington and Dunbar-Hall give the following definition:

|| 9 See Yothu Yindi (1992): “Treaty”. Tribal Voice. Mushroom Records International, 845.556 and the homepage of the band http://www.yothuyindi.com/music/treaty.html. 10 The Treaty of Waitangi, signed in 1840, guaranteed Māori ownership of their land as well as endowing the indigenous people with the rights of British subjects. 11 This list is taken from Arthur (1999: 66).

242 | Norbert Schaffeld

ganma is a word used to describe the place where the fresh water from a river meets the salt water of the sea. For the Yolngu [Aboriginal person], fresh water and salt water are significant opposites, and the turbulence and fertile potency of their meeting place is a powerful metaphor (Stubington & Dunbar-Hall 1994: 243)

To fully grasp the meaning of ganma one has to bear in mind that it also denotes the bond between sister clans (yapapulu), which observe “a diplomatic relationship of respectful distance and recognition of each other’s political independence” (Corn 2010: 94). The exchange between these clans is compared to the convergence of two currents, freshwater and saltwater, that will not merge in a process of assimilation but will produce sites of discernable as well as collaborative interplay. (Corn 201: 94–95) From here it is only a small step to take in Yothu Yindi’s bicultural reading of the ganma-metaphor in terms of a creative co-existence of at least two cultures. Given the fact that some elements of primordial knowledge become popularized reference points of a bicultural dialogue, Fiona Magowan sees the Yolŋu people’s need to maintain discursive control over the shared bodies of knowledge to prevent their devaluation or debasement (Magowan 1997: 145). She has a valid point here, yet only as far as ganma applies to the sociocentric clan structure of the Yolŋu. Indeed, in a noteworthy contrast to the aqua-aesthetic use of the ganma-metaphor, its sister clan reference would represent an implicit invitation to Euro-Australians (balanda) to share components of the sacred law.12 In the Aboriginal aqua-aesthetics, water is to be seen as a symbol of knowledge, especially when waters that represent mother/child or grandmother/granddaughter relationships “come together with the possibility of reproducing ancestral knowledge and lineages in their flow.”13 According to Helen Watson (quoted from Neuenfeldt 1993: 1), the ganma-metaphor describes “the situation where a river of water from the sea (Western knowledge) and a river of water from

|| 12 See Howard Morphy’s definition of sister clans as groups which share sacred knowledge and ritual. The clans have a similar position in the regional system of marriage alliance (Morphy 1991: 54, 311). 13 Cf. http://www.culturalsurvival.org/ourpublications/csq/article/ganma-negotiating-indigenous-water-knowledge-a-global-water-crisis. The relevant paragraph is here quoted in full: “If flowing water carries ‘feelingful’ emotion, it is because the aqua-aesthetics of Yolngu ancestral waters embody identities and personalities. Where these waters come together, an interaction of different personalities is implied in their ebb and flow. And a conjunction of personalities is also a conjunction of groups and kinship relations. Each water has its own flavor, design, and temperament held in its names, which are ritually intoned. These colors and tastes change as one water meets and mixes with the next. Waters that relate as mother/child or grandmother/granddaughter come together with the possibility of reproducing ancestral knowledge and lineages in their flow.”

Representing bicultural knowledge systems and epistemic decolonization | 243

the land (Yolngu knowledge) mutually engulf each other on flowing into a common lagoon and becoming one.” While from a strictly terminological point of view Bhabha’s (1994) concept of the third space and Ashcroft’s notion of post-colonial place seem to sit rather uneasily together, both are alike in addressing a spatial process of negotiation, which is virtually indispensable for genuine, post-colonial encounters (Teverson & Upstone 2011: 10). The turbulence as expressed in the estuarine ganma-metaphor proves to be a convenient image to illustrate the discursive process of negotiation (Magowan 1997: 149). To locate the ganma-vehicle at the centre of this ideological contestation inspires two further readings. First, it marks the terrain of cultural and political exchange as an Aboriginal site, and, second, it promotes a process of epistemic decolonization when indigenous knowledge and its Western counterpart become powers of equal rank.14 The turbulence of the political rock song is initially accompanied by the latter’s bilingualism and the stylistic opposition of the English and Aboriginal modes. For a world audience the Yolŋu parts might simply signify Aboriginality, for the lead singer Yunupiŋu it evoked reminiscences of his youth when the Aboriginal music of the djatpangarri style was highly popular (Corn 2011: 23). This specific linguistic and musical confluence is, of course, not coincidental as Stubington & Dunbar-Hall have shown: “Rock songs and djatpangarri are essentially vocal forms with rhythmic accompaniment; both have a link to dancing, and are repetitive and formulaic. Ideologically, both are adaptable to topics of current interest” (Stubington & Dunbar-Hall 1994: 257). Given the pronunciation of Australian Aboriginal English (AAE) (Lawe Davies 2001: 253), and Yunupiŋu’s vocal quality, the way to the bicultural encounter is already paved linguistically and metaphorically, before the repeated hook “Treaty Now” opts for the bilingual collocation “Treaty Ma”. And it seems that at last the components have finally formed an appropriate political pair.

|| 14 In 1990 Mr Yunupiŋu, his band, and Yolŋu leaders established the Yothu Yindi Foundation to set up the Garma Cultural Studies Institute to both promote the education of the Yolŋu and invite non-Aboriginal people to benefit from indigenous knowledge. From 2007 onwards the Foundation supported the Dilthan Yolngunha (Healing Place), which combined hereditary healing practices with conventional medicine. http://www.culturalsurvival.org/publications/cultural-survival-quarterly/australia/yothu-yindi-foundation-new-direction-arnhem-land.

244 | Norbert Schaffeld

4 Conclusion In summary, one can argue that Yothu Yindi verbalize a pre- and post-colonial place into popular being by making full use of a traditional metaphor whose vehicle, the confluence of two rivers, represents the tenor of a bicultural encounter of different cultures and knowledge systems. Indeed, the rhizomatic text of a lagoon provides significations that characterize the place as a site of cultural and political negotiations. And their agenda includes epistemological questions whose articulation is destined to mark one important aspect of the slow process of decolonization. To arrive at this conclusion, this paper has shown the way in which the terra nullius doctrine was embraced avant la lettre when an allegedly empty Australian space that seemed to be without native title was quickly transformed into colonial place by a discursive toponomy. The notion of colonial place is contrasted with an indigenous understanding of place that is endowed with ancestral or mythical significance and which can be discussed by applying Deleuze and Guattari’s theory of a decentered rhizomatic network. In the case of Aboriginal Australia the rhizomatic web can for instance include landscape formations, trees, rivers and currents, which – unlike metaphoric traditions of the West – all embody epistemological places whose physical contours and narrative distribution do not adhere to principles of hierarchy or taxonomy. Issues of toponomic representation and ownership overlap in the spatial cosmology of the Yolŋu people from Arnhem Land who see their ancestral progenitors as permanently present parts of the topographic land. Given the direct nexus between mythological topography and Aboriginal ownership of the land, or even their conclusive equation, it is no surprise that the land policy of the Australian government has provoked diverse forms of protest and resistance. Yet in spite of their relative prominence, none of the petitions became so popular as Yothu Yindi’s political vision. This article has revealed to what extent “Treaty” reflects the hereditary song and performance traditions of the Yolŋu while integrating Western stylistic devices. Singing mythological place into being, Yothu Yindi not only claim native title over the land, the band also envisions a co-existence between cultures of equal rank. The estuarine ganma-metaphor, which is a vital component of Yolŋu aquaaesthetics, refers to the convergence of two currents, freshwater and saltwater, in a way that emphasizes the cooperative over the conflicting, the still identifiable over the unrecognizable. It should thus be evident that water, a Yolŋu symbol of knowledge, and its turbulence become a powerful image of cultural and political negotiation. In Yothu Yindi’s view the contract is one to be signed by

Representing bicultural knowledge systems and epistemic decolonization | 245

partners of equal status and is thus paralleled by a political, cultural and epistemic decolonization that culminates in diverse balances, among them the still visionary equilibrium between indigenous and Western knowledge.

References Arthur, Jay M. 1999. The eighth day of creation. Journal of Australian Studies 23(61). 65–74. Ashcroft, Bill. 1999. The rhizome of post-colonial discourse. In Roger Luckhurst & Peter Marks (eds.), Literature and the contemporary. Fictions and theories of the present, 111–125. Harlow: Pearson Education. Ashcroft, Bill. 2009. Caliban’s voice. The transformation of English in post-colonial literatures. London/New York: Routledge. Barnes, Trevor J. & James S. Duncan. 1992. Introduction. Writing worlds. In Trevor J. Barnes & James S. Duncan (eds.), Writing worlds. Discourse, text and metaphor in the representation of landscape, 1–17. London/New York: Routledge. Berndt, Ronald M. & Catherine H. Berndt. 1964/1988. The world of the first Australians. Aboriginal traditional life. Past and present. Canberra: Aboriginal Studies Press. Berndt, Ronald M. & Catherine H. Berndt. 1989. The speaking land. Myth and story in Aboriginal Australia. Ringwood, Vic.: Penguin Books Australia. Bhabha, Homi. 1994. The location of culture. London: Routledge. Bourke, Colin & Helen Cox. 2009. Two laws. One land. In Colin Bourke, Eleanor Bourke & Bill Edwards (eds.), Aboriginal Australia. An introductory reader in Aboriginal Studies, 56–76. St Lucia: University of Queensland Press. Bourke, Eleanor. 2009. Australia’s first peoples. Identity and population. In Colin Bourke, Eleanor Bourke & Bill Edwards (eds.), Aboriginal Australia. An introductory reader in Aboriginal Studies, 38–55. St Lucia: University of Queensland Press. Carter, Paul. 2010. The road to Botany Bay. An exploration of landscape and history. Minneapolis: The University of Minnesota Press. Colman, Felicity J. 2007. Rhizome. In Adrian Parr (ed.), The Deleuze dictionary, 231–233. Edinburgh: Edinburgh University Press. Connor, John. 2002. The Australian frontier wars 1788–1938. Sydney: Allen & Unwin. Corn, Aaron. 2010. Land, song, constitution. Exploring expressions of ancestral agency, intercultural diplomacy and family legacy in the music of Yothu Yindi with Mandawuy Yunupiŋu. Popular Music 29(1). 81–102. Corn, Aaron. 2011. Treaty now. Popular music and the indigenous struggle for justice in contemporary Australia. In Ian Peddie (ed.), Popular music and human rights. Vol. II: World music, 17–26. London: Ashgate. Cowan, James. 1989. Mysteries of the dream-time. The spiritual life of Australian Aborigines. Bridport: Prism Press. Cultural Survival. n.d. Yothu Yindi Foundation. A New Direction in Arnhem Land. http://www.culturalsurvival.org/publications/cultural-survival-quarterly/australia/ yothu-yindi-foundation-new-direction-arnhem-land (checked 02/14/15) Deleuze, Gilles & Felix Guattari. 2011. A thousand plateaus. London: Continuum. Easthope, Antony 1983. Poetry as discourse. London/New York: Methuen. Frost, Alan. 1981. New South Wales as Terra Nullius. The British Denial of Aboriginal Land Rights. Historical Studies 19(77). 513–523.

246 | Norbert Schaffeld

Geertz, Clifford. 2000. Local knowledge. Further essays in interpretive anthropology. New York: Basic Books. Hodge, Bob & Vijay Mishra. 1991. Dark side of the dream. Australian literature and the postcolonial mind. Sydney: Allen & Unwin. Keupp, Heiner. 1988. Riskante Chancen. Das Subjekt zwischen Psychokultur und Selbstorganisation. Heidelberg: Asanger. Lawe Davies, Chris. 2001. Aboriginal rock music. Space and place. In Tony Bennett, Simon Frith, Lawrence Grossberg, John Shepherd & Graeme Turner (eds.), Rock and popular music. Politics, policies, institutions, 249–265. London/New York: Routledge. Magowan, Fiona. 1997. ‘The land is our märr (essence), it stays forever’. The Yothu-Yindi relationship in Australian Aboriginal traditional and popular musics. In Martin Stokes (ed.), Ethnicity, identity and music. The musical construction of place, 135–155. Oxford/New York: Berg Publishers. Magowan, Fiona. 2002. Ganma. Negotiating indigenous water knowledge in a global water crisis. Cultural Survival Quarterly 26(2). http://www.culturalsurvival.org/ourpublications/csq/article/ganma-negotiating-indigenous-water-knowledge-a-global-water-crisis (checked 01/19/15) Mandawuy Yunupingue [sic]. 1990. Mandawuy Yunupingue. Yothu Yindi Band. In L. Thompson (comp.), Aboriginal voices. Contemporary Aboriginal artists, writers and performers, 100– 103. Brookvale, N.S.W.: Simon & Schuster. Morphy, Howard. 1991. Ancestral connections. Art and an Aboriginal system of knowledge. Chicago: University of Chicago Press. Neuenfeldt, Karl W. M. 1993. Yothu Yindi and Ganma. The cultural transposition of Aboriginal agenda through metaphor and music. Journal of Australian Studies 17(38). 1–11. Oodgeroo Noonuccal [Kath Walker]. 1970/1981. My people. A Kath Walker collection. Milton, Q.: The Jacaranda Press. Rumsey, Alan. 2001. Tracks, traces, and links to land in Aboriginal Australia, New Guinea, and beyond. In Alan Rumsey & James F. Weiner (eds.), Space, narrative, and knowledge in Aboriginal Australia and Papua New Guinea, 19–42. Honolulu: University of Hawai’i Press. Shoemaker, Adam. 1989/1992. Black words white page. Aboriginal literature 1929–1988. St Lucia, Q.: University of Queensland Press. Strehlow, Theodore G. H. 1947. Aranda traditions. Melbourne: Melbourne University Press. Stubington, Jill & Peter Dunbar-Hall. 1994. Yothu Yindi’s ‘Treaty’. Ganma in Music. Popular Music 13(3). 243–259. Teverson, Andrew & Sara Upstone. 2011. Introduction. In Andrew Teverson & Sara Upstone (eds.), Postcolonial spaces. The politics of place in contemporary culture, 1–13. Houndmills: Palgrave Macmillan. Thiel, Josef Franz. 1984. Religionsethnologie. Grundbegriffe der Religionen schriftloser Völker. Berlin: Reimer. Warden, James. 2006. Atramentous history. History Australia 3(2). 49.1–49.8. Yothu Yindi. 1992. Treaty. Tribal voice. Mainstream. In Yothu Yindi, Tribal Voice. Mushroom Records International, 845.556. Yothu Yindi. 1993. Timeless land. Freedom. Mabo. Gunitjpirr Man. Our generation. In Yothu Yindi, Freedom. Mushroom Records International, 845.194. Yothu Yindi. n.d. http://www.yothuyindi.com/music/treaty.html. (checked: 02/16/2015) Yunupiŋu, Galarrwuy. 1998. We know these things to be true. The third Vincent Lingiari memorial lecture. https://app.lms.unimelb.edu.au/.../xid-2413046_2 (checked: 02/05/2015) Zorc, R. David. 1996. Yolŋu-Matha dictionary. Batchelor: Education Technology Unit.

Marina Wienberg

Moanus or Titan? On Ludwig Cohn’s amateur linguistics Abstract: There is a plethora of linguistically interesting documents which date back to the German colonial period. Many of these documents are stored in archives or private collections. Numerous language descriptions, word lists and other linguistically furled-notes have remained unpublished to this day. In this paper, the focus is on Ludwig Cohn, a typical amateur linguist of the colonial period, and on his linguistic notes and language-related ideas which deserve scrutiny in want of a better understanding of the zeitgeist of the second German Empire. Keywords: German colonies, Ludwig Cohn, languages of the Pacific, archives, manuscripts

1 Introduction The role played by philologically (more or less) untrained individuals forms a particularly interesting subject matter. The success (or lack of success) of contributions from non-linguists to the shaping of the linguistic disciplines in general might shed light on the origins of at least some of the preconceived ideas that still exist regarding characteristics of putatively “exotic” languages spoken outside Europe. In the context of the project Koloniallinguistik, the impact nonphilologists have had on the linguistic activities in the German colonies can be gathered from the recent reports on the related work by Gertrude Hornbostel (Vossmann 2011), Laurentius Bollig (Käser 2011), Philipp Adam Delaporte and Alois Kayser (Stolberg 2011), and Hermann Costenoble (Dewein 2013). These five names are representative of a much more sizeable group of linguistic nonprofessionals whose work needs to be studied thoroughly. In this sketch, my focus lies on yet another amateur-linguist who has hitherto gone unnoticed in the historiography of Austronesian linguistics, viz. Ludwig Cohn.

|| Marina Wienberg: University of Bremen, Linguistics and Literary Studies, Postfach 330440, 28334 Bremen, Germany. E-Mail: [email protected]

248 | Marina Wienberg

Many Germans went to the later colonies often several decades before the establishment of the so called Schutzgebiete or protectorates. They were traders, researchers and missionaries of different denominations. Adolf Lüderitz, for example, was a merchant from Bremen who became the founder of German South-West Africa, Germany’s first colony (Gründer 2001: 42). Augustin Krämer, Friedrich Fülleborn, Otto Reche among others were researchers who participated in the Hamburger Südsee-Expedition (i.e. Hamburg South Seas Expedition) in 1910 and in the frame of this expedition wrote documentations about Melanesia and Micronesia (Schindlbeck 2001: 132ff.). Of the many missionaries we should also mention e.g. Carl Hugo Hahn, a missionary and linguist who worked in South and South-West Africa and who published a dictionary of the Herero language in 1875 that included a grammar sketch. Another important figure is Father Laurentius Bollig, a missionary who worked on the Chuuk islands and who, in 1927, was the first to publish a much acclaimed work about the Chuuk islanders, describing their life and their language (Käser 2011). Frequently, these pioneers were also the first to discover the unbelievable richness of linguistic diversity in the colonies. Their lack of linguistic training notwithstanding, many amateurs made an effort to describe the languages they came into contact with. Many of the earliest (and sometimes also the only) texts ever written in the autochthonous languages of the former German colonial empire are credited to members of this heterogeneous group of people. Besides the above mentioned persons, members of the military and colonial administrative staff belonged to this group as well. Here we can mention e.g. Albert Hahl who was a colonial servant in German New Guinea and who implemented an administrative system considered revolutionary for the colonial administration of New Guinea, as he employed indigenous people as administrative servants (Hiery 2001b: 301). These people were in charge of creating German-style administrative structures and regulating the relationship of the German colonizers with the autochthonous people. Furthermore, several scientific expeditions explored the newly acquired territories for a variety of purposes. At this point we should mention the Hamburger Südsee-Expedition (i.e. Hamburg South Seas Expedition) which was undertaken under the leadership of Georg Thilenius in 1908–1910. This expedition was Germany’s largest ethnological and ethnographical endeavor in the South Seas region (Schindlbeck 2001: 199). Another example is the series of research expeditions of the Deutsche Inner-Afrikanische Forschungsexpedition (i.e. German inner African research expedition), which were undertaken from 1904 to 1935 under the leadership of Leo Frobenius. The results of his research travels have been published in many books and numerous articles. Most of his field notes and the visual documentation he created on his expeditions have been

Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 249

preserved and are kept e.g. in the Frobenius-Institut’s1 archives in Frankfurt-amMain. Obviously, these expeditions and their findings were well documented in print. However, it has to be emphasized that numerous linguistically relevant documents, notes, or descriptions have never been published. Many of them have been hidden away and still lie undiscovered, for example, in private collections. Much of this unpublished material is locked away in archives. In this paper only a small fragment of these unpublished and sometimes forgotten documents will be shown. This article is structured as follows. The first section is a sample of my research in the project EdArKoS (i.e. Edition von Archivalien zur kolonialzeitlichen Sprachforschung – Edition of archival language descriptions and documentations of the German colonial era). This sample highlights some of the linguistic notes and ideas of Ludwig Cohn, who worked at the Übersee-Museum2 in Bremen during this crucial historical period. Originally, his linguistic notes refer to the linguistic experiences Cohn made on occasion of his expeditions to the Admiralty Islands at the turn of the 20th century. In the second section, some excerpts from Cohn’s notebook will be compared with the structures of Titan, one of the languages still spoken on Manus Island today.

2 An interesting discovery in the archive of the Übersee-Museum in Bremen It was by no means the rule that the languages spoken in the previous German colonies were described by linguists. As already mentioned, it was usually a missionary, a trader or a traveler who was in charge of noting down and collecting things seen and heard. During my research in the archive of the ÜberseeMuseum in Bremen, I discovered some travel and research notes written by Ludwig Cohn who worked from 1904 to 1934 at the Übersee-Museum. A great number of his notebooks and travel notes have been stored in this archive. The-

|| 1 http://www.frobenius-institut.de. 2 This Museum was founded in 1875 as Städtische Museum für Natur-, Völker- und Handelskunde (i.e. Municipal Collection of Natural History and Ethnography) in Bremen. In 1935 it was renamed to Deutsches Kolonial- und Übersee-Museum (i.e. German Colonial- and OverseasMuseum). Today, it is called Übersee-Museum (i.e. Overseas Museum) and can be classified as Natural History and Ethnographic museum (http://www.uebersee-museum.de).

250 | Marina Wienberg

se notes tell of the research he did during explorations of the Admiralty Islands in 1908–09 and 1912–13. In this context it should be mentioned that German New Guinea including the Admiralty Islands was one of the first acquisitions of the German colonial empire. It consisted of the north-eastern part of New Guinea and several nearby island groups. The mainland part of German New Guinea and the nearby islands of the Bismarck Archipelago (previously Kaiser-Wilhelmsland) are now part of Papua New Guinea. The Admiralty Islands are a group of eighteen islands in the Bismarck Archipelago, to the north of New Guinea in the South Pacific Ocean, sometimes called the Manus Islands, after the largest island. Here, on Manus Island, Ludwig Cohn conducted field research for three months.

Fig. 1: Ludwig Cohn3

As to Ludwig Cohn himself: only very little is known about him. From the few documents that exist we do know that he was born on the 4th of February 1873 in St. Petersburg in Russia. His family moved there from Treuburg in Masuria, which was then a part of German East Prussia. In 1889, he began to study Natural Sciences in St. Petersburg. In 1893 he transferred to the Albertus-University

|| 3 http://www.suedseetraeumereien.de/Ludwig_Cohn_und_seine_Sammelreisen_fuer_das_ Bremer_Museum.html.

Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 251

in Konigsberg in East Prussia where he began his doctorate in Zoology. After that, he worked as a zoologist in Konigsberg (today Kaliningrad, Russia), where his scientific career began. He later also worked in Berlin and in Greifswald in Germany. As a Jew, he believed that he had little chance to develop his career as a university lecturer. Thus in 1904 he applied for the vacant position of Zoology Assistant at the so-called Museum of Nature, Ethnology and Commerce in Bremen (today Übersee-Museum). Because of his excellent references and publications, he was appointed to the position of Zoology Assistant in 1904 and, later, in 1920 he became the head of this same department. At the museum, he first worked as a zoologist and explorer. From 1904 to 1934, he was a travelling museum curator. His interests included not only zoological but also ethnological collections. The objects he collected on two separate expeditions to the South Pacific Islands covered a great part of the Oceania exhibition of the Übersee-Museum of Bremen up until the late 1970s. It is important to know that Ludwig Cohn did not have any linguistic or philological education, but he did have a natural talent for learning languages (Abel 1969: 96). In addition to Russian and German, he had some knowledge of Arabic and of some Melanesian languages. He even dared to demonstrate his knowledge of languages in the linguistic field when he wrote about his observations and his belief that a lot of Arabic borrowings could be found in the Melanesian languages. His thesis titled Spuren der Araber in der Südsee (i.e. Arab traces in the South Seas) was published in 1921 in the journal Deutsche geographische Blätter (i.e. German Geographical Gazette) edited by the Geographical Society in Bremen. Thus he wrote (Cohn 1921: 55ff.): Beim lesen der von Hambruch herausgegebenen ‘Südseemärchen’ kam ich auf eine unerwartete Spur: viele Personennamen, Ortsnamen etc. sind aus dem Arabischen abzuleiten. [...] Meine Beweisführung geht von sprachlicher Grundlage aus: einige Worte daher über den Erhaltungszustand des Arabischen in der mündlichen Überlieferung sowie über die Umwandlungen, welche es erfahren. Der erstere ist z.T. geradezu wunderbar, wenn man die vielen Jahrhunderte der Überlieferung berücksichtigt; doch finden sich daneben auch, wie selbstverständlich, erhebliche Lautumbildungen, falsche Silbentrennungen, allgemeine Korrumpierungen etc. Wurden doch hier Worte einer Sprache übernommen, deren Laute den Eingeborenen z.T. ganz unaussprechbar waren, und auch das spätere Bestreben, unverständliche Fremdworte ähnlichklingenden Worten der eigenen Sprache anzupassen, hat vielfach zur Korrumpierung beigetragen. [...] Hier nur einige der häufigsten und wichtigsten: das ghain, der harte Gutturallaut, wird zu einem k, ch oder g oder zu einem Konsonanten den die Autoren mit ng, ṅ umschreiben;

252 | Marina Wienberg

das ‘ain des Arabers wird meist durch eine Wortunterbrechung oder durch ein eingeschobenes o wiedergegeben; b und f werden nicht selten zu einem p; [...] das weiche arabische th (englisch ausgesprochen) wird oft zu einem s oder auch zu t; r wird vielfach (insbesondere wenn in der nächsten Silbe ein l folgt) zu l; [...] Vielfach werden die hartklingelnden Konsonantenverbindungen des Arabischen durch eingeschobene Vokale und Diphthonge mundgerecht gemacht; [When reading the ‘South myths’ published by Hambruch, I encountered something quite unexpected: a great deal of personal names, toponyms etc. would seem to be derived from Arabic. [...] My argumentation refers to the linguistic base: therefore some words on the conservational status of the Arabic in the oral tradition and through the transformation they underwent. The first is in parts quite marvellously when one considers the many centuries of tradition; nonetheless, considerable sound changes, false syllabication, common corruptions etc. are also to be found. However, the words of one language adopted here, whose sounds were partly unpronounceable and the effort to adapt incomprehensible foreign words to similar sounding expressions in the one’s own language, contributed greatly to corruption. [...] Here, only some of the frequent and the most important: The ghain (a guttural sound) changes to k, ch or g or to a consonant denoted by the authors [Hambruch] as ng or ṅ; The ‘ain of Arabs is mostly reproduced through a word interruption or through an inserted o; b and f sometimes change to p; [...] The soft Arabic th (English pronounced) converts frequently into s or t; r is frequently replaced by l (especially if an l follows in the next syllable); [...] The harder Arabic consonant clusters are frequently adapted by the insertion of vowels or diphthongs;]4

Cohn attempted to prove this untenable explication with numerous examples taken from a variety of Pacific myths. Thus, based on the following two examples, one for a proper name and one for a toponym, one recognizes which principle Ludwig Cohn thought to show that Arabic borrowings existed in the languages of the South Seas: (1)

Extract from a Palau fable [Cohn 1921: 58–59] Im Dorfe Ngaramesgang5 wohnt der Häuptling Gobakerai bedagal, ein Schlangenungetüm mit Menschenkopf ‘In the village Ngaramesgang lives the chief Gobakerai bedagal, a snake monster with human head’

Cohn segments these Palau words as follows:

|| 4 English translation by M.W. 5 Highlighted in bold type by M.W.

Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 253

Ngaramesgang: ghar6 means a ‘depression on a mountain (see Appendix, no. 2)7; meskan (see Appendix, no. 83) means ‘domicile, house’. Gobakerai bedagal: goba is derived from ghubar (see Appendix, no. 23) ‘dust’; kerai is from kara (see Appendix, no. 24) ‘to stir up’, bedagul should be divided into be-dagal: be is Arabic bi ‘with, through’; dagal is derived from thakalet (see Appendix, no. 25) ‘weight’. Based on this segmentation, Cohn proposes a different transcription from Hambruch: Er wühlte den Staub durch seine Schwere auf (beim Kriechen) – ‘He stirred up the dust with his weight (while crawling)’. Thus, Cohn sees in his translation “also eine Beschreibung statt eines Namens” (Cohn 1921: 59) – meaning: this is a description and not a name. (2)

The name of an island close to Nauru [Cohn 1921: 77] Juinareren – kleines sagenhaftes Nachbareiland, ‘wo es keine Menschen und keine Wesen gab, sondern nur niedriges Gestrüpp’. jui = dschuh257 = Hunger; nareren = ghararun258 = mit dem Tode bedrohende Gefahr. Das wäre etwa eine Insel des ‘drohenden Hungertodes’. ‘Juinareren is a small, fabled neighboring island where there are no people and no creatures, only brushwood’

Cohn’s segmentation: Jui is from Arabic dschuh (see Appendix, no. 257) which means ‘hunger’; nareren is derived from Arabic ghararun (see Appendix, no. 258) with the meaning ‘life-threatening danger’. Thus, the name of the island could mean an island with ‘death by starvation’. This speculative segmentation suffices to give an impression of Cohn’s idea concerning the Arabic borrowings in the South Seas languages. These examples make it quite understandable why this attempt did not meet with the approval of the linguists who heavily criticized his amateurish proposal (Abel 1969: 96). Nevertheless, this academic failure does not mean that Cohn’s linguistic field notes are of no interest.

|| 6 For ghar and the other Arabic words in the Arabic transcription see Appendix attached at the end of this paper. 7 These and the other Arabic words were listed by Cohn (1921: 101–104) and attached to the text Spuren der Araber in der Südsee (Arab traces in the South Seas).

254 | Marina Wienberg

3 Comparison of Cohn’s and Bowern’s grammar notes In this section, I compare Cohn’s notes with the equivalent paradigms of the Titan language. These paradigms form part of the Sivisa Titan sketch grammar which was summarized by Claire Bowern and published in 2011. As already mentioned, I found the notes which are being presented here in Cohn’s diaries while doing my research in the archive of the Übersee-Museum of Bremen. These diaries tell us that Cohn collected lists of words used by the people on the island of Manus at the time of his visit. He also tried to describe the basic language structure. Unfortunately, we do not know exactly which language Cohn tried to describe. Among his notes, I found no indication of the name of the language. Today, we can only speculate: either he did not know the name, or he considered it to be too unimportant. In addition, it should be said that about two dozen languages and their dialects are still being spoken on the Island of Manus and the nearby offshore islands in the Province of Manus in Papua New Guinea today. At the beginning of the 20th century, due to the obviously lack of knowledge pertaining to the linguistic situation in this region, the language(s) of Manus Island were simply named “moanus language” (Bowern 2011: 3). In his notes, Cohn mentioned pronouns, and numerals. And he drew up a brief list of basic vocabulary words including about 1100 items for the German part of the list. Unfortunately, we do not know how he came by his data initially. What we do know is the approximate geographical location of the region where this language was spoken: according to Cohn’s notes, it was on Manus Island in Nbunai (the present name could be Bunai), a village which was on the southern coast in the vicinity of Kelaua harbor. This location seems to coincide with the area where Titan, one of the Manus languages, is spoken today. According to the Ethnologue, this language is also known as Manus, M’bunai, Moanus or Tito.8 It might in fact be that Ludwig Cohn took notes of Titan during his stay in the Admiralty Islands. Until today, there have been very few published descriptions of Manus languages. One of these descriptions was recently published. It refers to a sketch of Titan grammar, which was summed up by Claire Bowern. For the most part, her work is based on the published and unpublished material collected by Father Josef Meier, who was a Roman Catholic missionary of the Order of the Sacred Heart on the Bismarck Archipelago at the beginning of the 20th century. Thanks

|| 8 http://www.ethnologue.com/language/ttv.

Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 255

to the help of his autochthonous apprentice Po Minis, Meier could record myths and stories of the Manus people, which he then translated into German. These texts were published between 1906 and 1909 in Anthropos and they are considered to be one of the earliest documentations of and in the Manus language. Whether Ludwig Cohn made use of Meier’s texts and translations remains unknown because Cohn did not mention Meier in his notes. Furthermore, Bowern mentions other published and unpublished linguistic and ethnographic notes and studies concerning Melanesia in general and the Manus Islands in particular which have been accrued by diverse researchers from 1901 up to our time.

3.1 Forms of pronouns Because of the limited scale of this text, it is not possible to cover all of Cohn’s notes here. Nevertheless, we can scrutinize some excerpts from his diary9 in this paper. For a start, we look at the paradigms of several pronouns which are shown in Figure 2.

Fig. 2: A page from the notebook of Ludwig Cohn with paradigms of personal and demonstrative pronouns, negative and interrogative pronouns

|| 9 The photocopies done by M.W. in the archive of Übersee-Museum of Bremen.

256 | Marina Wienberg

The transcription of these paradigms will be shown and explained in the following tables and compared with the relevant pronouns summarized by Bowern. In Table 1 the personal pronouns are presented as taken down by Cohn. Table 2 shows the personal pronouns summarized by Bowern. When one compares both tables, one can see that, except for some orthographic divergences, the paradigms are almost identical. Tab. 1: Personal pronouns by Cohn (compare Figure 2) Singular Eng I you he

Dual

Ger ich du er

Eng Ger jo we two (you and me) wir zwei (du und ich) oi we two (you and he) wir zwei (du und er) i we two wir zwei they two sie zwei Trial Eng Ger we three (I and you two) wir drei (ich und ihr beide) we three (I and they two) wir drei (ich und sie beide) you three ihr drei they three sie drei Plural Eng Ger we (I and you) wir – ich und ihr we (I and they) wir – ich und sie you Ihr they Sie

toru, taru rumo, jotaru joru uoru aru

to, joto, joto, tulumo joito uoto ato

ta, jota joja aua ala

The difference between these two tables is that Cohn assumed a trial number. In contrast to Cohn, Bowern (see Table 2) recognized the existence of the paucal number in Titan. She proves this with examples from Meier’s texts, although Meier himself placed these pronouns in the trial-category. Tab. 2: Personal pronouns by Bowern (2011: 43) Singular

Dual

Paucal

Plural

1st (Incl) (Excl)

yo

(yo)táru(rúmo) yṓru

(yṓ)to (tulúmo) yoίto

(yṓ)ta yṓya

2nd 3rd 3rd oblique

oi i wei

wṓru/waru aru

wṓto/wato āto

awa āla

Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 257

The sentences Bowern found in Meier’s texts demonstrate the existence of the paucal number. As you can see in example (3) below, the pronoun āto refers to more than ‘three’ whereas Cohn (Table 1) noted this pronoun as a trial pronoun for sie drei or ‘they three’: (3)

Paucal pronoun āto [Bowern 2011: 45] Poŋopoŋ akou, ato ti cucumoaato. Kor I pia, Pangaus ten 3PC be.on pod-3PC land 3SG be.dark ato ra. 3PC spring.out ‘There were ten pandanus fruits in pods. The land was dark. They sprang out of the pod.’

In Cohn’s grammatical notes, there is no indication of the paucal number. That is why we cannot be certain whether or not Cohn was aware of the category under discussion.

3.2 Demonstrative pronouns The items for the demonstrative pronouns given by Cohn (see Figure 2) and by Bowern appear to be completely identical. However Cohn seems to have grouped the pronouns corresponding to German grammatical gender distinction (Table 3). While Bowern, with reference to Meier’s glosses, differentiates between the ‘independent’ and ‘attributive’ forms of demonstrative pronouns and speaks of “three levels of distance: proximal, distal and far distal” (Bowern 2011: 52). She also mentions a form which Meier noted as ‘“der (mit Nachdrück), die, das”, i.e., ‘the, with stress/emphasis’’ (Bowern 2011: 52). Tab. 3: Demonstrative pronouns by Cohn (compare Figure 2) Ger

Eng

M

dieser tita

F

diese dieses

N

ita ta

Ger

Eng

Ger

titan der (emphatic) this die itan this das tan

The

jener (near) jene jenes

this

The The

Eng

Ger

tito

that

ito to

that that

jener (far) jene jenes

Eng ilatu/ latu latu tu

that that that

Bowern (2011: 52) argues that, according to Meier’s notes, each demonstrative pronoun can feature three forms: a basic syllable, which is glossed as ‘negative’ in Table 4, and two augmented forms with ti- and i-. That way the augmented forms with ti- should be independent demonstrative pronouns, without modify-

258 | Marina Wienberg

ing a noun (example 4), and the forms beginning with i- are attributive ‘adjectival’ demonstratives which modify a noun (example 5). Tab. 4: Demonstrative pronouns in Titan (Bowern 2011: 52) Independent

Attributive

(Negative)

English

tita titan tito ilatu

ita itan ito latu

ta tan to tu

this, here the (emphatic) that (near) that (far)

(4)

The independent distal demonstrative tito Esol i va : “Tito kile ca?” snake 3SG say this for what ‘Snake said, “What’s this for?”’

[Bowern 2011: 52]

(5)

The attributive demonstrative pronoun ito Kanian ito kulikon-kei. feast that (near) bark ‘That feast [of ours][will be] tree-bark.’

[Bowern 2011: 52]

Owing to the fact that there are no texts published by Cohn, it is impossible to assess Cohn’s interpretation of the demonstratives empirically. Bowern did not find any evidence in Meier’s texts for the short negative form ta, tan and to in Table 4, but according to Meier, these pronouns should only be used preceding the negator poen (Bowern 2011: 54). In Cohn’s notes, a negation is expressed with the following items: titá poen; titán poen; titó poen; ilatu poen (compare Figure 2). Cohn did not provide a translation for these items but in his word list we find the translation for poen: poen or (ne)-poen for German nicht or ‘not’. (6)

Negative in Cohn’s notes (compare Figure 2) tan poen, latu. this not that Nicht dieser, sondern jener ‘Not this but that’

According to Cohn’s Table 3 the German translation should be nicht das, sondern jene(r), which is only slightly different from that in example 6.

Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 259

3.3 Comparison of numerals Figure 3 and the tables below show that the numerals noted by Cohn and those which are shown by Bowern nearly correspond. However, Bowern says that “[t]he numeral system of Titan is decimal and subtractive, with seven, eight, and nine formed as ‘ten less three’, ‘ten less two’, ‘ten less one’” (Bowern 2011: 66). Seen from this point of view, the initial morpheme ada in the adatalo for ‘seven’ should mean ‘ten’. The similar structure of numerals is found in Cohn’s notes. Thus, we can deduce in the Figure 3 the followings items: ‘andratalo (noch 3)’, ‘andraruo ([noch] 2)’ and ‘andrasi ([noch] 1)’ which could be rendered in English as ‘ten still 3’, ‘ten still 2’ and ‘ten still one’. Because of lack of explication in Cohn’s notes, I could only interpret these items as “seven still three until ten”, “eight still two until ten”, and “nine still one until ten”. Unfortunately, deciphering what Cohn penciled in as additional notes proved to be impossible.

Fig. 3: Numerals noted down by Ludwig Cohn

260 | Marina Wienberg

Tab. 5: Comparison of the basic numerals by Cohn and by Bowern (2011: 67) Basic

Cohn

Bowern

1 2 3 4 5 6 7 8 9 10

si, esi ruo, eruo talo, etalo ea elima, lima euono, uono andratalo ( still 3) andraruo (“2) andrasi (“1) ákou, sángol

(e)si (e)ruo (e)talo ea (e)lima (e)wono adatalo adaruo adasi akou

The numerals listed in Table 5 show some orthographic differences. Additionally, Cohn gives the form sángol for ‘ten’ as an alternative form for ákou. Meier mentioned this form – saŋol – in his notes as well (Bowern 2011: 67). Tab. 6: The higher numerals by Cohn and by Bowern (2011: 69) Cohn 10 akou 11

akou pe si

12 akou pe ruo

Cohn

Bowern

akou

Bowern 100

sandat

saŋat

1,000

poesi (boisi) poesi

akou pe si

200

rungat

ruŋat

2,000

poeruo

poeruo

tulungat

tuluŋat

3,000

poetalo

poentalo

angat

aŋat

4,000

poena

poena

akou pe ruo 300

20 rukou, rungol rukou ~ ruŋol 400

Cohn

Bowern

30 tulungol

tuluŋ

500

limangat

limaŋat

5,000

poelima

poelima

40 angol

aŋol

600

uonongat

wonoŋat

6,000

poeuono

poewono

50 tówal

toval

700

andratulungat adatuluŋat 7,000

poeandratalo poenadatalo

60 uonongol

wonoŋol

70 andratulungol adatuulŋol 80 andrarungol

adarukou~ad

800

andrarungat

adaruŋat

8,000

poeandraruo poenadaruro

900

andrasangat

adasaŋat

9,000

poeandrasi

poenadasi

10,000 poenakou

poenakou

20,000 motjorumo

mocorumo

aruŋol 90 andrasangol

adaakou adasanŋol

For the most part, the higher numerals are structured as follows: The composite numbers are formed with the conjunction pe ‘and’: akou pe ruo ‘ten and two’ for ‘twelve’. The multiples of ten are formed with a basic numeral, or with a part of the segmental chain of the basic numeral, and -ngol (Cohn) or -ŋol (Bowern 2011: 68) which is the last syllable of sangol/saŋol ‘ten’: rungol (Cohn)/ruŋol Bowern (2011: 69) ‘two ten’ for ‘twenty’. The multiples of hundred are similar to the multiples of ten with a basic numeral, or with a part of the segmental chain of the basic numeral, and -nat (Cohn) or -ŋat (Bowern 2011: 68) which is the last

Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 261

syllable of sandat/saŋat ‘hundred’, f. ex.: rungat (Cohn) – ruŋat (Bowern 2011: 28) ‘two hundred’. The multiples of thousand show a different structure. They are formed with a marker for thousand poe(n) which was preceded by the basic numeral, or with a part of the segmental chain of the basic numeral: poeruo (Cohn; Bowern 2011: 69) literally ‘thousand two’ for ‘two thousand’. In Table 6 the numeral ‘fifty’ tówal (Cohn)/toval (Bowern 2011: 69) is an exception. According to the numeral’s structure shown by Cohn and Bowern, ‘fifty’ should have the form *‘(e)limangol’ following Cohn or *‘(e)limaŋol’ following Bowern. The reason for this exception was not commented upon by Cohn and Bowern.

4 Conclusion The comparison of Cohn’s and Bowern’s notes shows that there are striking similarities. I cannot in all certainty answer the question which language Ludwig Cohn thought to describe, but it could quite conceivably be a dialect of the Titan language. Even though Cohn’s notes are not complete and do not give us a comprehensive description of the language, we can see that Cohn was very interested in the language which was spoken in the region he researched. Whether or not and to which degree Cohn made use of Meier’s or some other person’s language knowledge, still remains to be answered. Cohn’s notes are only a small part of a great quantity of linguistically interesting documents stored in archives worldwide. This paper shows only a small fragment of what I have found within the framework of my project. The examples given in this paper show that some texts could contain far too daring ideas such as for example the supposed Arabic loanwords in the Melanesian languages. Nonetheless, the remaining texts could supply some really useful notes hidden in the archive’s shelves. Thus, they could give us an impression on ideas and contribute to our knowledge of language description in the German colonial period.

Abbreviations Eng EXCL F

Ger INCL

English exclusive feminine German inclusive

M N PC SG

masculine neuter paucal singular

262 | Marina Wienberg

Acknowledgments I would like to give my warmest thanks to the colleagues of the Archive of the Übersee-Museum of Bremen who allowed me to access the manuscripts and helped me to work with the documents. Furthermore, I would like to thank my colleagues in the project Koloniallinguistik – Languages in colonial context, Barbara Dewein, Daniel Schmidt-Brücken, Susanne Schuster, Thomas Stolz and Ingo H. Warnke for the fruitful and interesting discussions as well as valuable advice. All remaining errors are mine.

Primary sources Cohn, Ludwig: Tagebücher. The original is stored in the archive of the Übersee-Museum in Bremen.

References Abel, Herbert. 1969. Ludwig Cohn. In Wilhelm Lührs (ed.), Bremische Biographie 1912–1962, 96. Bremen: H. M. Hauschild. Bade, Klaus J. (ed.). 1982. Imperialismus und Kolonialmission Kaiserliches Deutschland und koloniales Imperium. Wiesbaden: Steiner. Bollig, P. Laurentius. 1927. Die Bewohner der Truk Inseln. (Anthropos ethnologische Bibliothek 3.1). Münster: Aschendorffsche Verlagsbuchhandlung. Bowern, Claire. 2011. Sivisa Titan: Sketch grammar, texts, vocabulary based on material collected by P. Josef Meier and Po Minis. Honolulu: University of Hawai’i Press. Cohn, Ludwig. 1921. Spuren der Araber in der Südsee. Deutsche Geographische Blätter 39(2). 56–104. Dewein, Barbara. 2013. H. Costenoble’s work on Chamorro (re-)edited. In Steven Roger Fischer (ed.), Oceanic voices – European quills. The early documents on and in Chamorro and Rapanui, 177–199. Berlin: Akademie Verlag. Gründer, Horst. 2001. Die historischen und politischen Voraussetzungen des deutschen Kolonialismus. In Hermann J. Hiery (ed.), 27–58. Hahn, Carl Hugo. 1875. Grundzüge einer Grammatik des Herero (im westlichen Afrika) nebst einem Wörterbuch. Berlin. Hiery, Hermann J. (ed.). 2001a. Die Deutsche Südsee 1884–1914. Ein Handbuch. Paderborn: Schöningh. Hiery, Hermann J. 2001b. Die deutsche Verwaltung Neuguineas 1884–1914. In Hermann J. Hiery (ed.), 277–311.

Moanus or Titan? On Ludwig Cohn’s amateur linguistics | 263

Käser, Lothar. 2011. ‘Den Buchstaben h können die Eingeborenen nicht aussprechen’. Pater Laurentius Bollig und die Sprache von Chuuk. In Thomas Stolz, Christina Vossmann & Barbara Dewein (eds.), 263–285. Krämer, Augustin. 1932. Truk: Ergebnisse der Südsee-Expedition, 1908–1910. (Ethnographie, Mikronesien, Bd. 5). Hamburg: Friedrichesen, de Gruyter. Kröger, Rüdiger. 2011. Dokumentation afrikanischer Sprachen durch Herrnhuter Missionare in Deutsch-Ostafrika. In Thomas Stolz, Christina Vossmann & Barbara Dewein (eds.), 161–187. Meier P. Josef. 1906. Berichtigungen zu Dr. Schnee’s Mitteilungen über die Sprache des Moánus ( Admiralitätsinseln). Anthropos 1(2). 210–228. http://www.jstor.org/stable/ 40442122 Meier, P. Josef. 1907. Mythen und Sagen der Admiralitätsinsulaner. Einleitung. Anthropos 2. 646–667. http://www.jstor.org/stable/40442222 Mückler, Hermann. 2010. Mission in Ozeanien. Bd. 2. Wien: facultas.wuv. Rentrop, Gundula. 2004. Ludwig Cohn – Zoologe und Forschungsreisender des Bremer Städtischen Museums 1904–1935. In Dorothea Deterts (ed.), Auf Spurensuche. Forschungsberichte aus und um Ozeanien zum 65. Geburtstag von Dieter Heintze, 79–86. Bremen: Übersee-Museum. Schindlbeck, Markus. 2001. Deutsche wissenschaftliche Expeditionen und Forschungen in der Südsee bis 1914. In Hermann J. Hiery (ed.). 132–155. Schmidt-Brücken, Daniel, Susanne Schuster, Thomas Stolz, Ingo H. Warnke & Marina Wienberg (eds.). 2015. Koloniallinguistik. Sprache in kolonialen Kontexten. Berlin: de Gruyter. Stolberg, Doris. 2011. Sprachkontakt und Konfession. Lexikalische Sprachkontaktphänomene Deutsch-Nauruisch bei den Missionaren Delaporte und Kayser. In Thomas Stolz, Christina Vossmann & Barbara Dewein (eds.), Kolonialzeitliche Sprachforschung. Die Beschreibung afrikanischer und ozeanischer Sprachen zur Zeit der deutschen Kolonialherrschaft, 285– 304. Berlin: Akademie Verlag. Stolberg, Doris & Stephan Engelberg. 2012. Sprachwissenschaft und kolonialzeitlicher Sprachkontakt. Sprachliche Begegnungen und Auseinandersetzungen. Berlin: Akademie Verlag. Stolz, Thomas, Christina Vossmann & Barbara Dewein (eds.). 2011. Kolonialzeitliche Sprachforschung. Die Beschreibung afrikanischer und ozeanischer Sprachen zur Zeit der deutschen Kolonialherrschaft. Berlin: Akademie Verlag. Vossmann, Christina. 2011. Gertrude Hornbostels Aufzeichnungen im Lichte zweier Klassiker der Chamorroforschung. In Thomas Stolz, Christina Vossmann & Barbara Dewein (eds.), 231–247. Zimmermann Klaus & Birte Kellermeier-Rehbein (eds.). 2015. Colonialism and missionary linguistics. Berlin: de Gruyter.

264 | Marina Wienberg

Appendix

Index of Authors Abe, Goe 209 Abel, Herbert 251, 253 Abolou, Camille Roger 81 Aikhenvald, Alexandra Y. 164 Alexander-Bakkerus, Astrid 190f. Alexius II of Moscow 28 Al-Sayyid Marsot, Afaf Lutfi 3 Altstadt, Audrey 30ff. Al-Ziriklī, Khayr al-Dīn 15 Amery, H. F. S. 14, 22 Anchimbe, Eric A. 158, 168 Anglim, John 208 Aoyagi, Machiko 204f. Arends, Jacques 134f. Arthur, Jay M. 241 Ashcroft, Bill 229, 231, 233, 243 Asin, Linda 140, 145 Aslanov, Heydər 38 Aşurbəyli, Sara 28 Auer, Peter 81 Augouard, Mgr. 102, 110, 117f., 120, 122 Avram, Andrei A. 126 Aymonier, Étienne 105 Baker, Philip 205 Bakıxanov, Abbasqulu 29 Bakker, Peter 127 Baratier, Colonel 102, 107, 116, 118ff. Baring, Evelyn 3, 22 Barner, David 161, 163 Barnes, Trevor J. 236 Barret, Paul 102, 107, 117, 120 Baskakov, Alexander 32f. Baskaran, Loga 170 Baumann, Adalbert 104 Béchet, Eugène 102, 109, 119, 121 Behnstedt, Peter 14 Bennigsen, Alexandre 31 Bérézine, Ilya 29 Berg, Mark L. 204ff. Bergman, Elizabeth M. 14 Berndt, Catherine H. 236 Berndt, Ronald M. 236 Besant, W. H. 4 Bhabha, Homi 243 Bies, Renata de 133, 144, 147, 149, 152 Biloa, Edmond 83

Blommaert, Jan 158f., 168 Bock, Kathryn 168, 175, 182 Boggis, Robert James Edmund 6 Bollig, P. Laurentius 247f. Bonola, F. 11 Borer, Hagit 161 Borges, Robert 134ff., 143, 148, 153 Bourke, Colin 230 Bourke, Eleanor 240 Bowern, Claire 253f., 256ff., 260 Brandon, Ruth 139, 144, 152f. Brato, Thorsten 159, 181 Britain, David 201, 204f., 207f., 217, 221 Bromley, Julian 35 Buchi, Eva 84f. Budaqova, Zərifə 38 Bunt, Harry C. 161, 163 Burkitt, F. C. 2 Cadiot, Anne 86 Calamaro, Francesca 88, 93 Calvet, Louis-Jean 103 Cantilly, Louis de 102, 118, 125 Carew, Joy Gleason 142, 144, 148, 151 Carlin, Eithne B. 134f. Carstens, Patrick Richard 11 Carter, Paul 231 Casado-Fresnillo, Celia 60 Chaudenson, Robert 108 Cheng, Lisa L. 165 Chevalier, Gisèle 82 Chierchia, Gennaro 161f., 163 Clayton, Anthony 101 Cohn, Ludwig 253, 256ff., 260 Colman, Felicity J. 232 Colrat de Montrozier, Raymond 102, 115, 117, 119, 121 Connor, John 231 Corbett, Greville G. 161 Corn, Aaron 235, 238f., 242f. Corne, Chris 102, 108f., 118, 127 Cousturier, Lucie 108, 110, 116 Cowan, James 234f., 241 Cowper, Elizabeth 163 Cox, Helen 230 Creider, Chet A. 164 Currie Hall, Daniel 163

266 | Index of Authors

Daly, Martin W. 4 Darieva, Tsypylma 34, 37 Davies, Mark 168 de Bies, Renata (see Bies, Renata de) De Bot, Kees 158 de Cantilly, Louis (see Cantilly, Louis de) de Kleine, Christa (see Kleine, Christa de) de Mandat-Grancey, Edmond (see MandatGrancey, Edmond de) Delafosse, Maurice 102, 106f., 109, 116, 118ff., 124 Delamaire, Marie-Stéphanie 11, 14 Deleuze, Gilles 230, 232f., 236f., 244 Denny, J. Peter 164 Deny, J. Mus̲h̲īr 22 Desjardins, René 102, 115, 118, 120ff., 125 Deuber, Dagmar 57 Dewein, Barbara 247 Dimmendaal, Gerrit J. 88, 160, 163 Dixon, Robert M. W. 164 Dostie, Gaétane 81 Doucet, Rachelle Charlier 57 Drescher, Martina 80f., 88, 95 Dunbar-Hall, Peter 239, 242f. Duncan, James S. 236 Dupratz, Père 102, 120f. Eersel, Hein 138, 140, 144 Ellis, Rod 167 Engelberg, Stefan 204f., 214, 220 Errington, Joseph 53, 55 Escalante, Enrique R. 64f., 67 Essed, Eva D. 140, 143 Essed-Fruin, Eva D. 143, 147, 150 Faraclas, Nicholas F. 158 Farquharson, Joseph T. 63 Fernández, Mauro 59 Fierens, Laura 140f., 148 Finch-Hatton, Murray 2 Fishman, Joshua A. 151, 157, 202 Fogarty, Richard S. 108, 121, 124, 126 Forbes, Duncan 9 Forman, Michael 58 Fortuno-Genuino, Cecilia 53 Franckel, Jean-Jacques 84 Frank-Job, Barbara 81 Fritz, Samuel 189, 192, 196, 199 Frost, Alan 231 Fujiwara, Yoichi 217

Gaskins, Suzanne 162 Geerts, Guido 153 Geertz, Clifford 234 Gentner, Dedre 162 Gerhardt, Ludwig 101 Gillon, Brendan S. 161, 163 Givón, Talmy 164 Gmelin, Samuel Gottlieb 42 Gobardhan-Rambocus, Lila 137ff., 142ff., 147ff. Gonzalez, Andrew 60 Green, Arthur Octavius 5, 8ff., 20, 22f. Grenoble, Lenore A. 32, 36 Gründer, Horst 248 Guattari, Felix 230, 232f., 236f., 244 Günthner, Susanne 81 Gut, Ulrike 169 Gyasi, Kwaku A. 107 Haggenmacher, Karl 2 Hahn, Carl Hugo 248 Heiler, Timo 126 Heine, Bernd 164 Hezel, Francis X. 204ff. Hiery, Hermann Joseph 248 Hilst, Eddy van der 152 Hinrichs, Lars 57, 63 Hirayama, Teruo 217 Hoepelman, Jaap 84 Holmes, Janet 217 Houis, Maurice 102, 108, 126 Huber, Magnus 158f., 164, 169, 180f. Hulme-Beaman, Ardern G. 4 Huntley-Fenner, Gavin 164 Imai, Mutsumi 162 Imaizumi, Yumiko 205 Inagaki, Shunji 167 Irvine, Judith T. 55 Jerrold, Walter 4 Joseph, Lewis S. 214, 217ff. Kachru, Braj B. 157 Karavaev, Alexander 37 Käser, Lothar 247f. Katz, Serafima 38 Kaufman, Terrence 202, 213, 223 Kelner, Victor 33 Kembo-Sure 158 King, L. W. 14

Index of Authors | 267

Klaeger, Sabine 81 Kleine, Christa de 136f. Kloss, Heinz 56, 73 Kluge, P. F. 207 Koch, Peter 97 Konstantinov, Vjačeslav 33 Kortmann, Bernd 159, 170, 180f. Kouwenberg, Silvia 57 Krämer, Augustin 248 Krifka, Manfred 165 Kroon, Sjaak 139 Kurvers, Jeanne 139

Mikirtuni, Artur 37 Miller, Boris 29 Miravite, Rommel M. 58, 68 Miyagi, Kimi 217 Moag, Rodney F. 158 Mohr, Susanne 164, 166f., 176ff., 181 Morency, Patrick 84 Morphy, Howard 242 Mosconas, Demetrius 5, 11, 13f., 16f., 20, 22 Mosegaard Hansen, Maj-Britt 84f., 90, 92, 95 Mühleisen, Susanne 54, 56, 73, 101f., 104 Mühlhäusler, Peter 101, 104, 106, 126f., 205

Labov, William 151 Lamothe, Ronald M. 5, 22 Lane-Poole, Stanley 14f. Lawe Davies, Chris 243 Lehmil, Linda S. 106 Leibowitz, Arnold 207 Lesho, Marivic 53, 60ff., 65f., 74 Lev-Ari, Shiri 202 Lewis, B. Bāb-i Serʿaskeri 22 Lewis, M. Paul 164 Leymaire, Henri 102, 118 Li, Peggy 164 Lipski, John 53, 60f., 75 Lucy, John A. 162 Lunkenheimer, Kerstin 159, 170

Nakao, Shuichiro 17, 21 Nakhlah, Yacoub 9, 12 Nallino, Carlo Alfonso 2 Neuenfeldt, Karl W. M. 242 Neumann-Holzschuh, Ingrid 82 Nigoza, Evangelino 64f., 68 Nurse, Derek 164

Magowan, Fiona 234, 239, 242f. Mak, Lanver 5 Makihara, Miki 54 Mandat-Grancey, Edmond de 102, 118, 120 Manessy, Gabriel 80, 102 Marceau, Le capitaine 102, 107, 116, 124f. Marie-Victoria, Soeur 102, 115, 120 Martin, Robert 84 Maschler, Yael 82 Massam, Diane 161 Matras, Yaron 82 Matsumoto, Kazuko 201, 203ff., 207f., 212, 214, 217, 220ff. Mazuka, Reiko 162 McArthur, Tom 202, 211 McKinney, Robert 204 McManus, Edwin G. 214 McPhetres, Samule F. 204 Məmmədquluzadə, Cəlil 30 Meier, P. Josef 254, 256ff., 260f. Mendo Zé, Gervais 83

Ocampo, Esteban A. de 65 Oesterreicher, Wulf 97 Ong, Walter J. 92 Orosz, Kenneth J. 104, 126 Otanes, Fe T. 63 Paillard, Denis 92 Parkvall, Mikael 127 Peattie, Mark R. 205f. Pelletier, Francis J. 161, 163 Penny, Ralph 63 Perl, Matthias 101f., 104 Perry, C. R. 18 Platt, John 157 Porter, Whitworth 8 Poulot, Brice 106 Quilis, Antonio 60 Quine, Willard 161 Quirk, Randolph 163 Quliyeva, Lalə 38 Radke, Henning 133, 137, 153 Ramos, Felicidad G. 66 Rechebei, Elizabeth D. 204 Redhouse, James W. 21f. Richardson, Irvine 164 Rijkhoff, Jan 161f. Rohrer, Christian 84

268 | Index of Authors

Romaine, Suzanne 56f., 168, 212f. Romero, Simon 135, 140f., 144, 147, 152 Roth, Mitchel P. 18 Rumsey, Alan 234ff.

Tien, Anton 5ff., 11, 20ff. Toledano, Ehud R. 2 Trifonovitch, Gregory J. 209 Turovsky, Rostislav 32

Sankoff, Gilian 82 Schachter, Paul 63 Schadeberg, Tilo 165 Schieffelin, Bambi B. 54f., 57 Schiffrin, Deborah 81 Schindlbeck, Markus 248 Schmied, Josef 158f., 164, 169 Schneemann, Christina 247 Schneider, Edgar W. 157, 180, 206 Schneider, Gerold 181 Schreier, Daniel 181 Schröter, Verena 159, 180f. Schwörer, Emil 102, 104, 109ff., 122f., 128 Scudamore, Frank 4f., 18ff., 22f. Sebba, Mark 54ff., 73f., 145f. Shahnazarian, Nona 32 Shaw, Stanford J. 22 Shidyāq, Aḥmad Fāris 9 Shuster, Donald R. 204ff. Silverstein, Michael 55 Sippola, Eeva 53, 50ff., 74 Skandera, Paul 168f. Skattum, Ingse 81 Snedeker, Jesse 161, 163 Spitta, Wilhelm 2 Spitta-Bey, Guillaume 8 Spitta-Bey, Wilhelm 8 Steingass, Francis J. 22 Steinhauer, Hein 60 Stolberg, Doris 247 Stolz, Thomas 190 Storch, Anne 160, 163 Strehlow, Theodore G. H. 235 Strudsholm, Erling 85, 90, 92 Stubington, Jill 239, 242f. Szmrecsanyi, Benedikt 159

Upstone, Sara 243

Tardo, Day S. 58 Teverson, Andrew 243 Thiel, Josef Franz 234 Thimm, C. A. 2, 5 Thomason, Sarah G. 82, 202, 212f., 216, 220, 223 Thompson, Roger M. 60

Valdman, Albert 106 Valikangas, Olli 84 Van den Avenne, Cécile 102, 107, 124 van der Hilst, Eddy (see Hilst, Eddy van der) Van Maele, Pieter 140f., 143, 145, 152 Versteegh, Kees 6 Vincent, Diane 81 Visser, Marieke 141 Voeste, Anja 103 Vollers, Karl 2 Vossmann, Christina 247 Wahrmund, Adolf 9 Warden, James 230 Wardhaugh, Ronald 137, 151 Watson, Charles Moore 5, 14ff., 20, 22 Webb, Vic 158 Werkmeister, Sven 104 White, Arthur Silva 3 Wiese, Heike 161, 163f., 167 Williams, Josiah 11 Willmore, John Selden 2 Wiltschko, Martina 164 Wimbush, S. Enders 31 Winer, Lise 57 Wingate, Francis Reginald 4, 22 Woidich, Manfred 7, 14 Wolff, John 60 Woolard, Kathryn 55 Wright, William 9 Yagodynsky, P. N. 40 Yunupiŋu, Galarrwuy 237 Zang Zang, Paul 83 Ždanko, Tatjana 35 Zerbian, Sabine 165 Zimmermann, Klaus 190f. Zorc, R. David 238

Index of Languages Arabic 1ff., 30, 251f., 261 – Bimbashi 21 – Classical 2, 9 – colloquial 10 – Egyptian 2, 4, 6, 9f., 12, 14f., 18, 23 – Juba 21 – Modern Standard 2, 4 – Standard 7, 9f., 15ff., 23 – Sundanese 14 – Syrian 7, 9 Aramaic 7 Arawak 135 Armenian 27, 31f., 35, 47 Aukaans/Ndyuka 136 Azeri 27, 29ff., 34, 36, 38ff., 42ff. – Baku 40, 44 Bislama 56 Carib 135 Cebuano 58 Chabacano/Chavacano 53f., 57, 59ff., 70ff. Chayawita 190 Chinese 162 – Cantonese 136 – Hakka/Keija 136 – Mandarin 136, 162

Français Tirailleur 101, 106ff., 124, 126ff. French 2, 5, 7, 18, 57, 79, 81f., 84f., 94, 96f., 101, 104ff., 117, 120, 126, 131, 141, 166f. – Cameroonian 79f., 87ff., 92ff. – Canadian 81 – Ivorian 81 – Burkina Faso's 81 – Mali's 81 Georgian 31, 47 German 4, 18, 84, 101, 103f., 111ff., 122f., 126, 166f., 192, 202f., 205, 215ff., 223, 251, 254 Greek 6, 190f., 198 Gumatj 239 Hausa 164 Hebrew 33 Herero 248 Hiligaynon 58 Ido 103 Italian 2, 6f. Japanese 202ff., 206f., 210f., 214ff., 220f., 223f. Jebero 189f., 192ff., 196ff. Jóola 166

Dutch 57, 131ff., 135ff., 148ff. English 2, 4ff., 12, 19ff., 26, 53, 57, 59ff., 67, 70, 74f., 79, 82f., 109, 135, 140, 144, 147, 152f., 157ff., 161ff., 166ff., 177, 202, 209f., 212, 214f., 218, 221f., 229, 238, 243 – American 203, 215, 222f. – Filipino 203 – Ghanaian 157, 160, 176ff., 180 – Kenyan 157, 160, 169, 176ff., 180f. – Nigerian 157, 160, 174, 177f., 180f. – Tanzanian 157, 160, 169, 177f., 180 Esperanto 103f., 108, 127 Filipino 58ff., 75 Flemish 145

Kembo-Sure 158 Kiembu 167 Kikuyu 167 Kiswahili 158, 164f., 167f. Kolonial-Deutsch 101f., 104f., 108, 113, 122f., 126ff. Kwinti 136 Latin 190f., 198 Ossetic 47 Ottoman 22 Palauan 201ff., 207, 209f., 212ff., 223 Papiá Kristang 56 Papiamento/Papiamentu 56f., 132

270 | Index of Languages

Paramaccan/Paamaka 136 Persian 22, 29, 42, 47 Pidgin English 104 Pidgin German 104f., 126 Pohnpeian 217 Portuguese 57, 135f., 144 – Brazilian 135 Quechua 189f., 192ff., 198 Russian 27, 30ff., 41f., 44, 46, 48, 251 – Baku 28f., 39ff., 44ff. Sabir – Petit Mauresque 106 Saramaccan/Saamaka 136 Sarnami Hindustani 136, 146 Spanish 53, 57ff., 64, 66ff., 70ff., 74f., 144, 189, 192, 194, 198, 202ff., 215, 217ff., 221, 223 Sranan/Sranantongo 74, 131f., 136ff., 140ff., 144f., 147f., 150ff. Suriname Javanese 136, 146 Surinamese Dutch 133, 149f., 154 Swahili 104, 158, 165

Tagalog 53, 58ff., 64, 66ff., 71f., 74 Taki-taki/Nengre 136 Tat 29 Tausug 58 Titan 253, 256 Tok Pisin 57 Trió 135 Turkish 1f., 4ff., 18, 22f. – Ottoman 2, 21, 23 Twi 167 Volapük 103f. Warao 135 Weltdeutsch 104 Wolof 166 Yiddish 33 Yolŋu-Matha 238

Index of Subjects Aborigines 234f. Admiralty Islands 249f., 254 Afghanistan 28 Arabic dialects 1, 13, 18, 23 archives 247, 249, 261 Armenia 28, 31ff., 36f. army 1ff., 10f., 14f., 20, 22f., 28, 105ff. Arnhem Land 235, 237, 239, 241, 244 Aruba 57, 132 assimilation 138, 202, 211f., 217ff., 224, 239, 242 Azerbaijan 27ff., 35, 37f., 42, 48 Belgium 131, 139, 144, 150 bicultural knowledge 230, 234, 239, 241ff. bilingualism 30, 35f., 41, 56, 74f., 210, 213f., 218ff., 222f., 243 Bismarck Archipelago 250, 254 Bonaire 57, 132 borrowing 39ff., 82, 97, 197, 201ff., 210ff., 251ff. borrowing scale 201f., 213f., 219ff., 223f. Brazil 133ff. British occupation 1ff., 23 British-India 134 calque/calquing 40f. Cameroon 79f., 83, 97 Central African Republic 88 Christian doctrine 189, 191ff., 198f. Chuuk islands 248 code-switching 40, 64, 136f., 210, 212, 214, 222 codification 56, 148, 152 colonial and postcolonial place 229, 231, 243f. colonial language 48, 53, 56, 59, 74f., 101, 135, 147f., 150, 204, 206, 210 colonial period 60, 210, 247, 261 colonialism 56, 102, 134, 152, 230 (de)colonization 202, 204, 230f., 237, 243f. colonizers 105, 122, 232, 241, 248 contact language 80, 88, 95, 97, 203 contact situation 54, 82f., 97

corpus 28, 38, 57f., 81, 88ff., 96f., 102, 118, 157, 159f., 168ff., 172, 174, 177, 179, 181, 241 corpus analysis 160, 168f., 177, 180 creole 21, 56f., 59, 61, 63, 68, 74f., 109, 115, 127, 132, 134ff., 147, 153, 158, 164 Curaçao 57, 132 diglossia 2, 75, 151f., 210 discourse marker 46, 79ff., 87f., 90, 96f. domain 31, 57, 60, 62f., 65, 72, 81, 132, 139, 145f., 152, 158f., 202, 204, 207, 210f., 215ff., 219f., 222ff. domination 55, 108, 203f., 208, 210, 216, 219 Egypt 1ff., 18ff. foreign language 82, 102f., 140, 145, 158 France 3, 79f., 105, 108 Fremdwort 202, 211, 218f. French Guiana 133 Gastwort 202, 211f., 218f., 223f. Georgia 31, 43 German colonies 104, 247, 249 German New Guinea 103, 205, 248, 250 Germany 103ff., 203f., 206, 248 Ghana 158, 164, 169f., 172, 174, 176f., 179 grammar 1f., 6, 8f., 12, 15, 17f., 20, 23, 48, 62, 74, 104ff., 146, 149, 152, 177, 179, 181, 189ff., 198, 211, 216, 218, 248, 253f. Haiti 57 high/low variety 2, 62, 74, 151 Hong Kong 134 hybridization 80, 97 identity 33ff., 37, 55, 57, 72, 75, 133, 137, 151ff., 231f., 237f. immigration 27, 29, 31, 33, 36, 42, 135, 203 India 4, 136 Indonesia (Dutch-India) 134, 136 interference 82, 85, 149, 167 intonation 38, 46, 121 Iran 28f., 33, 40 Iraq 28

272 | Index of Subjects

Kenya 158, 160, 164, 169f., 172ff., 178 language ~ contact 29, 41, 48, 80ff., 157, 201ff., 206, 211, 214, 223 ~ endangerment 59, 73, 75 ~ guides 1ff., 5ff. ~ ideologies 53ff., 62, 75 ~ policy 31, 37, 132, 138f., 141, 147, 152 ~ shift 29, 34, 54, 61f., 74 Lebanon 6 Lehnwort 202, 211, 217ff. lingua franca 58, 60, 104, 132, 140, 142, 147, 158, 164, 189, 205, 207 linguistic/language attitudes 30, 36, 56, 61, 74, 152, 204f., 213f., 219f., 222f. linguistic diversity 29, 248 linguistic policy (see language policy) linguistic schizophrenia 137, 142, 151 linguistic/language situation 1f., 11, 23, 30, 54, 61, 74, 132f., 142, 151ff., 157f., 211, 254 linguistic variation 61, 150 loanword 38, 60, 211ff., 221f., 261 Manus Island(s) 250, 254 manuscripts 189, 192, 196, 199, 261 Maroon(s) 134, 136, 140, 147 mass noun 157, 159ff., 170f., 175ff., 179ff. Melanesia 248, 251, 254, 261 Micronesia 204ff., 248 minority language 38, 53, 61 missionaries 103, 190f., 194, 204, 207, 209, 248 multilingualism 135ff., 159, 167, 202, 210 native language 2, 57, 141, 143, 147, 158, 166f., 190 Netherlands 57, 131ff., 135, 138f., 143f., 146, 148, 150, 152ff. Netherlands Antilles 57, 132, 138 New Guinea 103f., 206, 249, 250 New Holland 230 New Zealand 241 Nigeria 158, 164, 169ff. number system 163f., 175f.

official language 2, 31, 59, 61, 72, 79, 83, 135, 137, 140ff., 144, 146ff., 152f., 158, 210 oral communication/discourse/language 53, 67, 90, 93, 97, 141, 145, 151f., 234, 252 oral data/sources 58, 84, 88ff., 92 orthography 53, 55ff., 66, 70f., 74f., 102, 104, 145, 211f. Palau 201ff., 252 Palestine 4, 6 paucal 256f. Peru 189 Philippines 53, 57ff., 70, 72, 204f. phonology/phonetics 37f., 48, 199, 213 pidgin 38, 101, 103, 108f., 115, 126ff., 158, 164, 205 plantations 71, 134ff., 140 postcolonial ~ communities/societies 55, 75, 160 ~ context 54, 56, 59, 158 ~ languages/varieties 79, 159 ~ situation 54f., 74 postcolonialism prestige 37, 56f., 59, 62, 74f., 137ff., 142, 151f. reduplication 109f., 127 Russian Empire 28f., 34 Russification 32f., 35f. Saba 132 second language (L2) 32, 36, 38, 58, 80, 82, 96, 158, 167 semantics 85, 161, 164, 177, 179ff. Sierra Leone 57 social meaning 54f., 149 Soviet Union 27, 33ff., 47, 209 spelling 14, 19, 54f., 57f., 66ff., 106, 139, 146, 189, 211f., 238 St. Eustatius 132 St. Maarten 132 standardization 56f., 65, 74, 146, 149, 152 Sudan 1, 8, 11, 14, 21 Suriname 131ff. Switzerland 84 syntax 44, 101f., 110, 128, 161, 179, 213, 222 Syria 6f.

Index of Subjects | 273

Tanzania 158f. 164, 169f., 172ff., 179 terra nullius 230, 238, 241, 244 Togo 104 toponomic representation 233, 244 transcription 6f., 9f., 19f., 22, 89, 252f., 255 Uganda 158 Uzbekistan 48 Venezuela 57 vocabulary 6, 8, 10, 14f., 18, 23, 34, 36, 38, 40, 42, 56, 58, 101f., 104, 107, 122ff., 140, 150, 153, 189, 192, 199, 213f., 216ff., 221, 223f., 254

word formation 102, 109f., 211 writing 2, 53ff., 59, 61, 65, 67ff., 71ff., 146, 158, 236 Yothu Yindi 230, 238ff.