Hispanisation: The Impact of Spanish on the Lexicon and Grammar of the Indigenous Languages of Austronesia and the Americas 9783110207231, 9783110197938

Citation preview

Hispanisation

≥

Empirical Approaches to Language Typology 39

Editors Georg Bossong Bernard Comrie Yaron Matras

Mouton de Gruyter Berlin · New York

Hispanisation The Impact of Spanish on the Lexicon and Grammar of the Indigenous Languages of Austronesia and the Americas edited by Thomas Stolz Dik Bakker Rosa Salas Palomo

Mouton de Gruyter Berlin · New York

Mouton de Gruyter (formerly Mouton, The Hague) is a Division of Walter de Gruyter GmbH & Co. KG, Berlin.

앝 Printed on acid-free paper which falls within the guidelines of the 앪 ANSI to ensure permanence and durability.

Library of Congress Cataloging-in-Publication Data Stolz, Thomas. Hispanisation : the impact of Spanish on the lexicon and grammar of the indigenous languages of Austronesia and the Americas / Thomas Stolz, Dik Bakker, Rosa Salas Palomo. p. cm. ⫺ (Empirical approaches to language typology ; 39) Includes bibliographical references and index. ISBN 978-3-11-019793-8 (cloth : alk. paper) 1. Indians ⫺ Languages ⫺ Foreign elements ⫺ Spanish 2. Indians ⫺ Languages ⫺ Foreign words and phrases ⫺ Spanish. 3. Austronesian languages ⫺ Foreign elements ⫺ Spanish. 4. Austronesian languages ⫺ Foreign words and phrases ⫺ Spanish. 5. Spanish language ⫺ Influence on foreign languages. I. Bakker, Dik. II. Title. III. Series. PM128.S86 2008 497⫺dc22 2008008019

Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available in the Internet at http://dnb.d-nb.de.

ISBN 978-3-11-019793-8 ISSN 0933-761X © Copyright 2008 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin. All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording or any information storage and retrieval system, without permission in writing from the publisher. Printed in Germany.

Preface On 5–8 May, 2005 the first international conference devoted to “Romancisation world-wide: The impact of French, Italian, Portuguese and Spanish on the autochthonous languages of Africa, the Americas, Asia and Austronesia (with special focus on Hispanisation)”1 was held at the University of Bremen organised by the head of the local linguistics department, Thomas Stolz. Thirty-one linguists from seventeen countries world-wide gathered in Bremen to discuss the contact-induced phenomena which can be observed in the indigenous languages outside Europe which have been exposed to the influence exerted by one of the major Romance languages. For the first time in the history of our discipline(s), Romance Philologists, general linguists and students of Afro-Asiatic, Niger-Kordofanian, Amerindian and Austronesian languages as well as experts of Pidgins and Creoles approached jointly the issue of linguistic Romancisation. Soon, it became clear that the general topic of our conference defines a promising new field for researchers. The participants recognised many commonalities characteristic of the changes triggered by contacts with Romance languages – commonalities which seem to be largely independent of the individual Romance donor language and the genetic affiliation of the recipient languages. For many of the latter, the original talks delivered on occasion of the conference were a first in the sense that language contact phenomena in that particular language had never been addressed before. Moreover, with the notable exception of Spanish, the Romance languages had been studied at best sparingly in their function as donors in extra-European contact situations. The extant studies of Romance contacts with non-European languages mostly refer to individual constellations. Our conference marks the starting-point for largescale comparative work on Romancisation processes which will also have positive repercussions in the field of language contact studies in general because besides previously unknown fresh data the papers contribute to developing an over-arching theory of language contact. There is already a well-established tradition of studies dealing with the impact of Spanish on the indigenous languages of the Americas, Austronesia and Africa. Together with Portuguese, Spanish is the European language with the longest history of overseas-expansion. Estimates as to the number of languages which have come into contact with Spanish outside Europe during the last five hundred years are not available. Nevertheless, it is a fair guess that Spanish was and has been a partner in contact of literally

vi Preface hundreds of autochthonous languages around the globe. This fact gives Spanish a special status within language contact studies because investigating Hispanisation on a large scale allows the linguist to compare the outcome of language contact situations with one variable (= the recipient languages) and one constant (= the donor language) in the formula. This constellation is helpful if one strives to find out about recurrent patterns, extrapolate on their basis and formulate generalisations which go beyond the special case of Hispanisation. The Linguistics Department at the university of Bremen, for instance, hosted two previous conferences (both financed by the Volkswagen-Foundation) in 1996 on the Hispanisation of Mayan languages (documented in Koechert and Stolz [1998]) and in 1999 on Hispanisation processes in Amerindian and Austronesian languages (documented in Zimmermann and Stolz [2001]). In 2004, the Linguistics Department at the university of Amsterdam organised an international conference whose focus was upon the contacts between Spanish and Amerindian languages in Latin America. Owing to the quantity of speakers who, on occasion of our conference in 2005, addressed situations in which Spanish is the donor language, we editors decided to publish a volume dedicated exclusively to the phenomenon of Hispanisation. (Three of the papers which focus on Hispanisation processes are published elsewhere, namely Dik Bakker, Ewald Hekking, and Jorge Gómez Rendón on Hispanisation in Otomí and Quechua is included in Stolz et al. [2008a] devoted to theory-related and methodology-related issues whereas Gloria Nistal’s contribution on Spanish in Equatorial Guinea and Mohand Tilmatine’s article on Hispanisation of Tarifit occur in Stolz et al. [2008b] which focuses upon Romancisation in Africa). This volume contains ten contributions to aspects of Hispanisation in the indigenous languages of the Americas and Austronesia. They are ordered according to geographic principles starting with Mexico. On the basis of twelve selected typological features, Anne Jensen looks at older stages of Nahuatl and tries to connect idiosyncrasies of Nahuatl as opposed to other Uto-Aztecan languages to Hispanisation. José Antonio Flores Farfán studies modern varieties of Nahuatl which are characterised by massive Hispanisation. The author takes the high degree of disintegration of the inherited Nahuatl structures as evidence for a growing neglect on the part of the native speakers who seem to be distancing themselves from Nahuatl as a means of self-identification. Zapotec is in the focus of Martina Schrader-Kniffki’s article. The author reviews the various contact varieties of Zapotec resulting from different degrees of exposure to Spanish. Her leading question is whether or not purist attitudes might constitute an intermediate stage which eventually gives way to the genesis of a mixed lan-

Preface

vii

guage. With the contribution by Luciano Giannelli and Raoul Zamponi on Kuna the reader is taken to Central America (Panama). The authors give a detailed account of Spanish features and items which have made it into Kuna. However, they emphasise that the recipient language has proved relatively resistant as to the integration of Hispanisms. Two contributions address contact situations in South America. Jorge Gómez Rendón investigates the limits of language contact with specific reference to Imbambura Quichua. In spite of the high number of Hispanisms in Imbambura Quichua, it is claimed that the language is not experiencing a typological shift. Media Lengua is one of the paradigm cases of the mixed languages debate. Suzanne Dikker devotes her article to the question of what happens to Spanish prepositions in Media Lengua which is based on Quichua morpho-syntax and thus would normally prefer postpositions. Since Spanish prepositions remain postpositions in Media Lengua, Suzanne Dikker is able to challenge a number of assumptions about relexification and the supposed irrelevance of structural aspects in language contact. Four contributions study Hispanisation in Austronesia. Closest to the Americas is Easter Island (Rapa Nui) whose bilingual setting is analysed by Steven Roger Fischer. He states that the seemingly impossible has happened on this Chilean island, namely: the formerly rapid progress made by Spanish has come to a halt such that one can speak of a reversal of Hispanisation. Steve Pagel compares Hispanisation processes in Rapanui and Chamoru (Marianas). Both languages are heavily influenced by Spanish. However, the author demonstrates that the usual labels of contact linguistics such as Creoles, mixed languages etc do not apply to Hispanised Rapanui and Chamoru. Patrick O. Steinkrüger argues that the Spanish linguistic influence is still rather important in the linguistic landscape of the Philippines. The results of partial Hispanisation can be found in many segments of phonology, lexicon and – to a lesser degree also of – grammar of the major Philippinian languages (Tagalog, Cebuano etc.). In addition, there are the Spanish-based Chabacano varieties. In the final paper, Rosa Salas Palomo and Thomas Stolz try to determine the attitudes of speakers of modern Chamoru towards Hispanisms. The reaction of their respondents to the experimental tasks paints an intricate picture which inter alia shows that the usual socio-linguistic parameters alone do not help the linguist to understand what motivates the speakers preferences for or against Hispanisms. From all these contributions, we expect that follow-up studies will develop which further our understanding of the mechanisms of language contact processes in general. The conference was made possible by a generous grant by the Volkswagen-Foundation (Hannover). The editors are also grateful to Sonja Kettler

viii Preface and Cornelia Stroh whose technical expertise and editorial skills made this volume become a reality. Rachel Montague helped us to nativise the English of the contributions written by linguists with a first language other than English. A word of thanks is due to Mouton de Gruyter – especially Ursula Kleinhenz – for their non-bureaucratic way of handling things. The editors of Empirical Approaches to Language Typology deserve our gratefulness for accepting the book in their well renowned series. Georg Bossong painstakingly reviewed each and every contribution to this volume and made very helpful comments. Thomas Stolz, Dik Bakker, Rosa Salas Palomo Bremen (Germany), Amsterdam (The Netherlands) and Lancaster (United Kingdom), Mangilao (Guam) November 2007

Notes 1. We have replaced the original terms “Romanicisation” and “Hispanicisation” with “Romancisation” and “Hispanisation” after a longish discussion with linguistically well-informed native speakers of English who assured us that the new terms are more suited as the older ones – especially “Romanicisation” – invite readings which would be off the mark.

References Koechert, Andreas and Thomas Stolz (eds.) 1998 Convergencia e individualidad. Las lenguas mayas entre hispanización e indigenismo. Hannover: Verlag für Ethnologie. Stolz, Thomas, Dik Bakker, and Rosa Salas Palomo (eds.) 2008a Aspects of language contact. New theoretical, methodological and empirical findings with special focus on Romancisation processes. Berlin /New York: Mouton de Gruyter. 2008b Romancisation in Africa. The influence of French, Italian, Portuguese and Spanish on the indigenous languages of Africa. Bochum: Brockmeyer. Zimmermann, Klaus and Thomas Stolz (eds.) 2001 Lo propio y lo ajeno en las lenguas austronésicas y amerindias. Procesos interculturales en el contacto de lenguas indígenas con el español en el Pacífico e Hispanoamérica. Frankfurt/M.: Vervuert.

Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v

Section I – Latin America Hispanisation in Colonial Nahuatl?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anne Jensen

3

The Hispanisation of modern Nahuatl varieties . . . . . . . . . . . . . . . . . . . . . . José Antonio Flores Farfán

27

From language mixing to mixed language via purism? Spanish in contact with Zapotec (Oaxaca/Mexico) . . . . . . . . . . . . . . . . . . . Martina Schrader-Kniffki Hispanisms in Kuna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Luciano Giannelli and Raoul Zamponi Spanish lexical borrowing in Imbabura Quichua: In search of constraints on language contact . . . . . . . . . . . . . . . . . . . . . . . . Jorge Gómez Rendón

49 77

95

Spanish prepositions in Media Lengua: Redefining relexification . . . . . 121 Suzanne Dikker Section II – Austronesia Reversing Hispanisation on Rapa Nui (Easter Island) . . . . . . . . . . . . . . . . 149 Steven Roger Fischer The old, the new and the in-between: Comparative aspects of Hispanisation on the Marianas and Easter Island (Rapa Nui) . . . . . . . . . . 167 Steve Pagel

x

Contents

Hispanisation processes in the Philippines . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Patrick O. Steinkrüger Pro or contra Hispanisms: Attitudes of native speakers of modern Chamoru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237 Rosa Salas Palomo and Thomas Stolz Index of subjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Index of languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 Index of toponyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280 Index of authors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285

Section I Latin America

Hispanisation in Colonial Nahuatl? Anne Jensen

1.

Introduction

Nahuatl is a southern Uto-Aztecan language spoken in Central Mexico. Shortly after the Spanish invasion, Spanish friars invented an orthography for Nahuatl resulting in an abundance of sources written in Nahuatl, which give us access to the information needed to identify the characteristic features of colonial Nahuatl. Colonial Nahuatl – as well as modern Nahuatl – is a polysynthetic language with agglutinating morphology. Four stages of Nahuatl are postulated, three of which are the result of processes that have led to typological changes in Nahuatl: Nahuatl as a UtoAztecan language, Nahuatl as Mesoamerican language, colonial Nahuatl and modern Nahuatl. The two first stages will be examined in order to make clear how Colonial Nahuatl differs from them, while modern Nahuatl is only briefly touched upon. The point of departure for discussing the posited changes is twelve typological parameters. 2.

Four stages of Nahuatl

The four stages of Nahuatl and the three processes leading to typological change are: 1. 2. 3. 4.

Nahuatl as a Uto-Aztecan language process 1  Nahuatl as a Mesoamerican language process 2  Colonial Nahuatl process 3  Modern Nahuatl

2.1. Typological features of Uto-Aztecan languages Since Nahuatl is a Uto-Aztecan language, the first stage is Proto UtoAztecan, which is characterized by the following eleven of the twelve typological features taken into consideration.1

4 Anne Jensen 2.1.1. Constituent order The constituent order in the Uto-Aztecan languages is verb-final, and the reconstructed order for Proto Uto-Aztecan is SOV. 2.1.2. Adpositions At least two different syntactic constructions can be reconstructed for Proto Uto-Aztecan: a. A noun is followed by a postposition attached to a pronominal form: the tree he/his-under ‘under the tree’. b. A postposition is attached directly to a noun, usually an inanimate noun: tree-under ‘under the tree’. 2.1.3. Relative clauses Three types of relative constructions can be reconstructed for Proto UtoAztecan: a. prehead relative clauses, b. posthead relative clauses, c. extraposed relative clauses occurring at the end of the matrix clause. 2.1.4. Possessive constructions Two possessive constructions are reconstructed for Proto Uto-Aztecan: a. A possessive pronominal element is prefixed to the possessum, which in turn is succeeded by the possessor: his-dog the man ‘the man’s dog’. b. An existential construction: his children were/existed ‘there were his children’. 2.1.5. Copula constructions No copulative verb can be reconstructed for Proto Uto-Aztecan. Instead a pronominal affix is attached to a nominal predicate: he-warrior ‘he is a warrior’.

Hispanisation in Colonial Nahuatl?

5

2.1.6. Adjectives In Uto-Aztecan languages adjectives are derived synchronically or diachronically from other parts of speech; there are very few underived adjectives. 2.1.7. Impersonal constructions Some Uto-Aztecan languages have an impersonal construction, and many languages do not allow any overt expression of the agent similar to a byphrase in English. Therefore, the impersonal construction is analyzed as having an unspecified, non-referential agent: someone saw him/he was seen or there was seeing of him. 2.1.8. Incorporation In some Uto-Aztecan languages, we see incorporations of nouns encoding the instrument used to perform the activity encoded by the verb, e.g., footpush ‘push with foot’. 2.1.9. Causative and applicative A causative or applicative suffix can be reconstructed for Proto UtoAztecan, and an applicative suffix can be reconstructed for Proto Southern Uto-Aztecan. 2.1.10. Number The number category for nouns in most Uto-Aztecan languages includes singular and plural, whereby plural marking is restricted to animate nouns. These markings can be reconstructed for the proto-language. 2.1.11. Numerals Low numerals may be reconstructed for the proto-language, but no numeral system as such.

6 Anne Jensen 2.2.

Nahuatl as a Mesoamerican language

It is impossible to distinguish between the stage termed “Nahuatl as a Mesoamerican language” and the subsequent stage termed “Colonial Nahuatl” because the written sources we have access to were written after the Spanish invasion in 1521. There are, however, some sources that may have been written at some point within the period between 1531 and the 1540’s and other sources from the 1550–1570’s (TLA, CF2), that is, at a time shortly after the invasion. The Nahuatl encountered in the early sources differs from other southern Uto-Aztecan languages, and this difference may be due to the Nahuatl speakers’ contact with speakers of other languages in Mesoamerica. Evidently, Nahuatl was not a homogeneous entity. At all stages, the Nahuatl speech community has encompassed dialectal, sociolectal and other lectal variation – as applies to all other speech communities. For colonial and modern Nahuatl we have evidence for dialectal variation, but I will not take this into consideration in the following survey of the typological features. 2.2.1. Constituent order It is still under discussion as to whether the unmarked order is VSO or VOS. One of the rare occurrences in which all syntactic arguments of a transitive verb are expressed is shown in (1). The verb is -ilwia ‘say something to someone’, and the constituent order is VSO: (1)

ye no: sep-pa ø-ki-ø-wa:l-ilwi
now also one-time SUB.3-OBJ.3SG-OBJ.3SG-DIR-say.PRET in wi:cil-o:po: -i in kwawi--i
kak a: DEF proper name DEF proper name PART -on-a- iye IMP.SUB-DIR-OBJ.INAN-see ‘Now Huitzilopochtli once more said to Cuahuitlìcac: “Please be on the look out!”’ (CF III, folio 2 verso)

The unmarked constituent order is today verb-initial.

Hispanisation in Colonial Nahuatl?

7

2.2.2. Adpositions The only adposition is the postposition -ko ‘in’, ‘at’ which attaches to inanimate noun roots. In (2), the postposition is attached to the stem a:-kal- ‘boat’: (2)

niman ye ik a:-kal-ko ø-mo-a:li
-ke
then already hence water-house-in SUB.3-REFL-put-PRET.SUB.PL ‘then immediately after they put themselves into (the) boat’ (TLA, document V, folio 16 verso)

Apart from this postposition, a number of relational nouns derived from body parts are also found. In (3), the root i
ti- ‘belly’ is combined with the postposition: (3)

ø-k-ø-ilwia-
in a
to
-ke
SUB.3-OBJ.3SG-OBJ.3SG-say-SUB.PL PART ruler-PL ka:n o:-ø-wa:l-ma:-ki:s-ke
where AUG-SUB.3-DIR-hand-go out-PRET.SUB.PL ø-k-im-on-ilwia-
ka SUB.3-OBJ.3SG-OBJ.3PL-DIR-say-SUB.PL PART a:-i
ti-k onka:n a:kokolko water-belly-at there toponym ‘The rulers say to him [sic!]: “Where did they escape?” They say to them: “In the water there at Acocolco”’ (TLA, document V: folio 7 recto)

In (4), the relational noun -pan ‘on’, ‘(on the) surface (of)’ is attached to two noun roots, a:ka- ‘person’ is animate, while naka- ‘flesh’ is inanimate: (4)

aw in sa ø-a-cakw-ti-ya
-ke
and PART just SUB.3-OBJ.INAN-shut-LIG-go-PRET.SUB.PL sa a:ka-pan sa naka-pan in just person-on just flesh-on PART o:-ø-on-ki:s-ke
in AUG-SUB.3-DIR-go out-PRET.SUB.PL PART ø-om-pano:-ke
SUB.3-DIR-cross over-PRET.SUB.PL ‘And those who just went being the last ones just went out on people, just crossed over on bodies’ (CF XII, folio 42 verso)

8 Anne Jensen In (5), -pan occurs in a discontinuous possessive construction. A possessor prefix is attached to the relational noun, and the possessor is a
to
ka:ti:an-i ‘ruler-messenger’: (5)

inin a
to:l-li i:-pan ø-m-i
toa:-ya DEM word-ABS.SG POSS.3SG-on SUB.3-REFL-say-IMPF in a
to
ka:-ti:an-i PART ruler-messenger-ABS.SG ‘These words were said about the ruler-messenger’ (CF IV, folio 203 recto)

The possessed relational noun in (5) precedes the possessor. 2.2.3. Relative clauses At this stage, prehead relative clauses are found. The relative clause in (6) is inside the square brackets. The head is the noun phrase in am-i
i:yo:-cin in amo-a
to:l-cin ‘your honoured breath, your honoured words’: (6)

ma: ti-k-to-mak-toka-tin PART SUB.1PL-OBJ.3SG-REFL-give-follow-VET.PL [in nika:n ø-wa:l- ki:sa] in am-i
i:yo:-cin PART here SUB.3-DIR-go out PART POSS.2PL-breath-HON in amo-a
to:l-cin PART POSS.2PL-word-HON ‘Let us avoid pretending that your (PL) honoured breath, your (PL) honoured words that are coming out here are given to us’ (CF III, folio 36 verso)

However, posthead relative clauses are the most frequent in the sources. The head in (7) is i
wi- ‘feather’, it is modified by a relative clause: (7)

aw in o:-ø-a- pan niman and PART AUG-SUB.3-OBJ.INAN-sweep.PRET then in i
wi- [in ø-k-on-kwi:-skiya SUB.3-OBJ.3SG-DIR-take-IRR PART feather-ABS.SG PART i:-iilla:n ø-k-on-a:li
] POSS.3SG-stomach SUB.3-OBJ.3SG-DIR-put.PRET ‘And when she had swept, then she wanted to take out the feather that she had put on her stomach’ (CF III, folio 1 verso)

Hispanisation in Colonial Nahuatl?

9

2.2.4. Possessive constructions Two kinds of possessive constructions are found at this stage. One of them is shown in (8), a possessor prefix is attached to two noun roots: (8)

in

i:-siwa:-w i:-to:ka: ø-tekapa PART POSS.3SG-woman-POSS.SG POSS.3SG-name SUB.3-proper name ‘Tecapa is his wife’s name’ ~ ‘Tecapa is the name of his wife’ (Hinz et al. 1983, Vol. 1, folio 21 recto)

The other construction is shown in (9), it is an existential construction: (9)

ø-is-ka


i:-siwa:-w

SUB.3-here-exist POSS.3SG-woman-POSS.SG

‘There his woman exists’/’there is his women’ (Hinz et al. 1983, Vol. 1, folio 7 recto) Apart from those two constructions, the derivational suffix -e
/-wa
derives possessive nouns from other nouns; as in (10) from the nounroot siwa:‘woman’: (10) a-ya:-mo ø-siwa:-wa
NEG-already-NEG SUB.3-woman-DERI ‘He is still not a woman-owner’ ~ ‘he does not have a wife yet’ (Hinz et al. 1983, Vol. 1, folio 1 verso) 2.2.5. Copulative constructions There is no copulative verb at this stage; a pronominal subject prefix is attached to a nominal predicate. The subject of the nominal predicate ako
[i] ‘slave’ in (11) is third person singular, which has no overt expression (11): (11) ya ka to:l- iwi- ø-ako
-[i] already 15-year-ABS.SG SUB.3-slave-ABS.SG ‘She is a slave since 15 years’ ~ ‘She has been a slave for 15 years’ (Hinz et al. 1983, Vol. 2, folio 80 verso)

10 Anne Jensen 2.2.6. Adjectives The only adjective at this stage of Nahuatl is we:yi ‘big’. In (12), it modifies the noun iwi- ‘day’: (12) onka:n ø-m-i
toa in ke:nin me: i
ka-
SUB.3-REFL-say PART how Mexican-ABS.PL there we:yi iwi- ø-ki- i:w-ke
in big day-ABS.SG SUB.3-OBJ.3SG-do-PRET.SUB.PL PART i
kwa:k o:-ø-ki:s-ke
españoles when AUG-SUB.3-go out-PRET.SUB.PL Spaniards onka:n me: i
ko there Mexico ‘Here it is told how the Mexicans made a big day (= a feast) when the Spaniards went out here from Mexico’ (CF XII, folio 51 verso) Nouns may modify other nouns. An example of this is shown in (13); the noun kwal-li ‘good thing’ modifies the compounded noun to: omi ‘rabbitbones’: to: -omi- (13) ø-ki-namaka in kwal-li SUB.3-OBJ.3SG-sell PART good thing-ABS.SG rabbit-bone-ABS.SG Apart from that, the preterit form of verbs may also occur as modifiers of nouns. The constituent order shown in (12)–(13) is modifier-head, occurs most frequently. However, as seen in (14), the order modifier-head also occurs. The adjective we:yi-ntin ‘big.PL’ modifies the preceding noun ko:ko:wa
‘snakes’: (14) i:n-te pa ø-a-
toa in i: ki -tin POSS.3PL-by SUB.3-OBJ.INAN-say PART all-PL ko:-ko:wa-
we:yi-ntin RED-snake-ABS.PL big-PL ‘It tells about all the big snakes’ (CF XI, folio 87 verso)

Hispanisation in Colonial Nahuatl?

11

2.2.7. Impersonal constructions There is an impersonal construction at this stage of Nahuatl. In (15), the impersonal form of the ditransitive verb te:mo:lia ‘question somebody about/ for something’ is given: (15) niman ye ik ø-mo-te:moa teo:-kwia- then already thus SUB.3-REFL-search god-excrement-ABS.SG te:-a-te:mo:-li:-lo OBJ.ANIM-OBJ.INAN-search-APPL-IMPS ‘Then immediately after gold is searched for, people are questioned (about/for something)’ ~ ‘Then immediately after there is searching for gold, there is questioning (about/for something)’ (TLA, document V: folio 18 verso) 2.2.8. Incorporation Nahuatl at this stage allows for incorporation of different types. One type is the incorporation of an object argument. Two examples are shown in (16): the compounded noun stem o: ikwal- ‘fruit’ is incorporated into two different verbs, and the noun root k waw- ‘tree’ is incorporated into the verb -to:ka ‘plant’, ‘bury’: ø- o: -kwal-e
o:i-kwal-namaka-k PART flower-good.thing-sell-PTC SUB.3-fruit-DERI ø-o:ikwal-ma:ma ø-k waw-to:ka SUB.3-fruit-carry on one’s back SUB.3-tree-plant ‘The fruit-seller, he is an owner of fruit, he carries fruit on his back… he plants trees’ (CF X, folio 57 recto)

(16) in

Another type is the incorporation of noun roots encoding the instrument. In (17), the twisting action is carried out with teeth, the noun root an- ‘tooth’ is incorporated: (17) ø-a-an-ko:liwka:-wia SUB.3-OBJ.INAN-tooth-twisted thing-DERI ‘He twists (something) with (his) teeth’ (about the sandal maker) (CF X, folio 53 verso)

12 Anne Jensen The incorporated noun may also encode the location at which the state of affairs takes place. The incorporated noun root in (18) is kwa:- ‘head’, which encodes the location for the water-putting: (18) a
mo ø-mo-kwa:-a:tekia NEG SUB.3-REFL-head-put water on ‘He does not put water on himself on the head’ ~ ‘He is not baptized’ (Hinz et al. 1983, Vol. 2, folio 54 verso) 2.2.9. Causative and applicative Causative and applicative forms occur rather frequently in the early sources. One example of a causative is given in (19); the causative suffix -tia attaches to the verb -nemi ‘live’: (19) aw i(n) Miguel i(n) ø-k(i)-nemi:-tia and PART proper name PART SUB.3-OBJ.3SG-live-CAUS ikomoteka- a
-e ø-k(i)- i:wa mi:l-li proper name NEG-something SUB.3-OBJ.3SG-do field-ABS.SG ‘And Miguel that Xicomotecatl supports (makes live) does not cultivate any fields’ (Hinz et al. 1983, Vol. 1, folio 27 verso) However, the majority of the causatives found are entries in Molina’s dictionary from 1571, that is, they may already have been lexicalized at an earlier stage of Nahuatl. An applicative form is given in (20), the applicative suffix -lia attaches to the verb piya ‘guard’: (20) ø-is-katki amo-koko:kA-w in AwIs-i SUB.3-here-exist POSS.2PL-property PART arm-ABS.SG in o:-ø-ame: -ø-piya-lia:-ya PART AUG-SUB.3-OBJ.2PL-OBJ.3SG-guard-APPL-IMPF in amo-a
to
ka-w PART POSS.2PL-ruler-POSS ‘Here is your (PL) property, the arms that your (PL) ruler guarded for (TLA, document V, folio 15 verso) you (PL)’ The verb piya is monotransitive, and the direct object in (20) in Awis-i ‘the arms’, is the NP-relative. By means of the applicative, the possessor

Hispanisation in Colonial Nahuatl?

13

amo- ‘your (PL)’ is promoted to a syntactic argument of the verb (an indirect object). 2.2.10. The number category of nouns At this stage, the number category includes singular and plural for animate nouns: (21) ko:wa:- snake-ABS.SG ‘a snake’

ko:-ko:wa-
RED-snake-ABS.PL ‘snakes’

Inanimate nouns, however, are not marked for plural. The inanimate noun kakawa- ‘grain of cacao’ in (22) is marked for singular, but the numeral means ‘200’: (22) ma
ak-po:wal-li kakawa-  10-counting-ABS.SG grain of cacao-ABS.SG ‘200 grains of cacao’ (Hinz et al. 1983, Vol. 1, folio 10 recto) 2.2.11. Numerals Nahuatl at this stage has a vigesimal system, composed of combinations of twenty, which allows for counting indefinitely. In (22) above the stem -po:wal- means ‘a counting’, i.e., ‘twenty’. 2.2.12. Classifiers At this stage, Nahuatl has classifiers, the most frequent one is te- ‘stone’: (23) ma:kwil-te- to:tol-te- ‘five-stone-ABS.SG turkey/bird-stone-ABS.SG ‘five-stone eggs’ (Hinz et al. 1983, Vol. 2, folio 70 verso) 2.2.13. Summary Nahuatl at this stage has become a member of the Mesoamerican Linguistic Area since it shares the five areal features that are common to nearly all Mesoamerican languages:

14 Anne Jensen 1. Nominal possession is of the type illustrated in (8) – a possessor prefix is attached to the possessed, which precedes the possessor. 2. Relational nouns are used to encode spatial and other relations [see (3)– (5)]. 3. The unmarked constituent order is verb-initial. 4. Nahuatl has a vigesimal numeral system. 5. Nahuatl has shared semantic loan translations, that include the word for ‘egg’, which is ‘bird-stone’ – as was shown in (23). Furthermore, numeral classifiers and directional morphemes are found in many Mesoamerican languages, but cannot be reconstructed for Proto UtoAztecan. Nahuatl at this stage has numeral classifiers and directional morphemes [as is shown in (1) and (3) above]. Nahuatl has undergone typological changes that may be ascribed to contact with speakers of other Mesoamerican languages. 2.3.

Colonial Nahuatl

It is not possible to set up any clear delimitation of a stage of Nahuatl called “Colonial Nahuatl”. First, because the degrees of contact between Nahuatl speakers and Spaniards clearly differed greatly along geographical, social and many other parameters. Second, because the sources only give us access to written Nahuatl, which may have differed from speech. Third, because the sources cover various genres – chronicles, letters, testaments and other legal documents. In order to trace some of the changes, I draw upon some of the sources from 1611 and onwards. 2.3.1. Constituent order There are instances of SV-constituent order in the later sources. In (24) the first sentence has SV-order: the subject a:ka- ‘person’, ‘man’ precedes the verb ø-mo-kwepa-k ‘returned’, but the succeeding sentence has VSorder: the verb o:-ø-wa:l-la
‘came’ precedes the subject. The document is from 1611:

Hispanisation in Colonial Nahuatl?

15

mitik (24) niman a:ka- ø-mo-kwepa-k then man-ABS.SG SUB.3-REFL-turn-PRET toponym niman o:-ø-wa:l-la
to-ta
-cin mitik then AUG-SUB.3-DIR-go.PRET POSS1PL-father-HON toponym ‘Then a/the man returned to Mitic, then our honoured father came to Mitic’ (COD 170) This may indicate a change – at least in some dialects – from VS-/VO-order towards a pattern equivalent to the Spanish one, which allows for both orders. This is also the case in some modern dialects in which both orders are used. 2.3.2. Adpositions With respect to the encoding of spatial and other relations, the postposition -ko ‘on’, ‘at’ is rare in the sources; an occurrence from 1720 is shown in (25). Here it attaches to the noun root te:n- ‘edge’ and forms part of a possessive construction: (25) onka:n ø-pe:wa in i:-te:n-ko there SUB.3-begin PART POSS.3SG-edge-at in a:-pami- PART water-ditch-ABS.SG ‘It starts there at the edge of the irrigation ditch’ (Lockhart 1991: 103) Relational nouns also occur; in (26), which is from the same document as (25); the relational noun -te ‘by’, ‘at’ is used: (26) ø-on-a
si ok se:-pa in i:-te

SUB.3-DIR-arrive still one-time PART POSS.3SG-by/at in a:-pami- PART water-ditch-ABS.SG ‘It arrives again at the irrigation ditch’ (Lockhart 1991: 103) Another example is from 1795, the relational noun -na:wak ‘next to’ is attached to the noun root mi:l- ‘field’:

16 Anne Jensen (27) onka:n ø-on ok tohloca o
-i there SUB.3-DIR-lie toponym road-ABS.SG i:-mi:l-na:wak dionisio cosme POSS.3SG-field-next to proper name ‘It (a piece of land, AJ) lies there at (the) Toluca road next to Dionisio Cosme’s field(s)’ (COD 74) In some modern dialects, the relational nouns have been reanalysed as prepositions. As can be seen from the examples in (25) and (26), the relational nouns in possessive constructions precede the noun encoding the ground, so the constituent order is equivalent to the order of a preposition and a noun encoding the ground in Spanish. 2.3.3. Relative clauses There are no pre-head relative clauses in the documents examined. That does not exclude the possibility that prehead relative clauses may appear in other sources. It suggests though, that the prehead relative clause was disappearing in Colonial Nahuatl – Spanish had and still only has posthead relative clauses. In the modern data I have access to, only posthead relative clauses have been observed. 2.3.4. Possessive constructions The possessive constructions shown in (8)–(9) in section 2.2 are not found in the sources. They have been substituted by a construction similar to the Spanish possessive construction. In the early sources, the verb piya means ‘guard’, ‘take care of’. You can see this from (28), which is a repetition of (20) in section 2.2. The ruler guarded the arms for his people, he did not own the arms, the possessor prefix amo- is second person plural, showing that the arms belong to the people: (28) ø-is-katki amo-koko:kA-w in AwIs-i SUB.3-here-exist POSS.2PL-property PART arm-ABS.SG in o:- ø-ame: -ø-piya-lia: ya PART AUG-SUB.3-OBJ.2PL-OBJ.3SG-guard-APPL-IMPF in amo-a
to
ka-w PART POSS.2PL-ruler-POSS

Hispanisation in Colonial Nahuatl?

17

‘Here is your (PL) property, the arms that your (PL) ruler guarded for (TLA, document V, folio 15 verso) you (PL)’ However, in early source from 1583, piya clearly means ‘have’, the woman in (29) does not guard her few children, she has or owns them: (29) ka

a
mo miyak-tin ni-ki(m)-piya no-pil-wa:n much-PL SUB.1SG-OBJ.3PL-have POSS1SG-child-PL ‘I have not many my children’ (Lockhart 1991: 70) PART NEG

The construction is a mixture of a Spanish one and the one seen in (8) above. This may be ascribed to the fact that kinship terms very rarely occur in the non-possessed form. Across languages, a change of meaning from ‘guard’ to include ‘have’ is not rare. In the case of Nahuatl, the change could have taken place independently of the contact with Spanish, but it seems likely that the change is contact-induced. The most recent instance of a possessive noun is from 1653. The possessive noun in (30) is comparable to the possessive noun in (10) in section 2.2: (30) a:lte:pe-wa
-ke
town-DERI-PL lit. ‘owners of (a) town’ ~ ‘inhabitants’

(COD 176)

The possessive noun is derived from the noun root a:lte:pe- ‘town’ by means of the derivational suffix -wa
. 2.3.5. Copulative constructions The zero-copulative construction is found in the sources. The one given in (31) occurs in a source from 1653. The subject prefix for second person singular ti- is attached to the possessed noun root -I Ipa3 ‘representative’: (31) ka

tewa-cin

t-i:-I Ipa

PART PRON.2-HON SUB.2SG-POSS.3SG-representative

noestro padre our father

santo padre holy father ‘YOU (emphatic) are the representative of our father, (the) holy father’ (COD 174) Such a copulative construction is still found in modern dialects.

18 Anne Jensen 2.3.6. Adjectives The modifying items shown in (12)–(13) in section 2.2 are also found in colonial Nahuatl. A few Spanish adjectives appear in the sources, especially within the Christian domain. One of these – santo ‘holy’ – occurs in (31). 2.3.7. Impersonal constructions No impersonal constructions have been observed in the later sources. However, this does not suggest much since we do not know whether all dialects had an impersonal construction, and we do not know how frequent the construction has been. 2.3.8. Incorporation Only a few instances of incorporation have been found in the later sources. In the example shown in (32) from 1795, the noun stem a a:wil- ‘payment’ has been incorporated into the verb -maka ‘give something to someone’: (32) ka

Dios ø-kin-mo-aa:wil-maki-li-s PART God SUB.3-OBJ.3PL-REFL-payment-give-HON-FUT ‘God will give (HON) them (a/his) reward’ (COD 76)

The scarce number of incorporations may suggest that this formation of verbs is not productive in Colonial Nahuatl. The same is true of modern Nahuatl, and this may be due to contact with Spanish, which does not allow for incorporation of this type. 2.3.9. Causative and applicative The only causative forms are those found in the early sources, which at that time may have already been lexicalized. There are, however, applicative forms in later sources. The one in (33) is from 1795, the applicative form of the monotransitive verb -ka:wa ‘leave’ allows the beneficent to be encoded as a syntactic argument: (33) ka

ni-k-ø-ka:wi-li:-ti:w

no- wi:w

PART SUB.1SG-OBJ.3SG-OBJ.3SG-leave-APPL-DIRE POSS.1SG-grandchild

‘I will go and leave it (a piece of land, AJ) to my grandson’ (COD 74)

Hispanisation in Colonial Nahuatl?

19

2.3.10. Number At this stage of Nahuatl, some inanimate nouns are inflected for plural. The inflected nouns in (34) occur in a source from 1746: (34) kal-tin seki kwaw-tin a:ma-me
house-PL some tree-PL paper-PL ‘houses…some trees…papers’ (Karttunen & Lockhart 1978: 170–174) In many modern dialects all nouns are inflected for number. This change may haven taken place without any contact with Spanish, but it may also be contact-induced since Spanish inanimate nouns are marked for plural. 2.3.11. Numerals In the later sources, the Nahuatl numerals occur, as can be seen in (35): (35) se: a:l-li na:w-po:wal one land-ABS.SG four-counting ( =20) ‘…a (piece of) land (of) 80 (rods)’

(COD 74)

However, Spanish numerals also occur, so the Nahuatl vigesimal numeral system is being substituted by the Spanish one, or by a mixture of the Spanish and the Nahuatl one as we see in some modern dialects. 2.3.12. Classifiers Only very few occurrences of classifiers have been found in the most recent documents; the example in (36) is from 1795: (36) sen-te- misa one-stone-ABS.SG mass ‘one-stone Mass’

(COD 74)

This may suggest that classifiers are in the process of being abandoned at this stage of Nahuatl.

20 Anne Jensen 2.3.13. Other features of colonial Nahuatl The contact with Spanish is evident in the early and the later sources with respect to neologisms and loanwords. Neologisms were formed for the new items introduced by the Spaniards, for instance the word for gun – which is a compounded noun consisting of the root e- ‘fire’ and the stem kikis‘trumpet from a sea shell’: (37) e-kikis-i fire-trumpet-ABS.SG ‘gun’

(CF XII, folio 8 verso)

From this new noun a verb was derived by means of the suffix -wia, meaning of the derived verb is ‘make use of N’, so the verb in (38) means ‘make use of a gun’, that is ‘shoot’: (38) e-kikis-wi:lo:-k fire-trumpet-DERI IMPS-PRET ‘People were shut by guns’

(CF XII, folio 38 recto)

Furthermore, the semantic domain of some words was extended by the Nahuatl speakers – one example may be the verb -piya ‘guard’, that acquired the meaning ‘have’. Another example is the verb -i
akoa ‘damage’ or ‘hurt’, which was ascribed the meaning ‘sin’. A range of Spanish nouns were already borrowed at an early stage. Many of the first loanwords are names for officials or professions. From these nouns, verbs and other items were derived in the way verbs were derived from nouns in Nahuatl. Spanish infinitives were borrowed at a later stage – one example from 1746 is seen in (39). The Spanish infinitive is notificar ‘notify’. It is treated like a noun root – the Nahuatl suffix -oa derives verbs from nouns: (39) ø-mo-notificar-o:-s SUB.3-REFL-notify-DERI-FUT ‘He shall notify’

(Karttunen and Lockhart 1976: 174)

It is not clear at what time the borrowing of Spanish infinitives starts.

Hispanisation in Colonial Nahuatl?

21

2.3.14. Summary Colonial Nahuatl differs from Nahuatl – Nahuatl as a Mesoamerican language: some relational nouns may already have started their change into prepositions at this stage; the possessive constructions seem to be substituted by a construction with a have-verb; the possessive nouns have become rare; and the numeral system is being substituted by the Spanish decimal system. Still other changes may have taken place, e.g. in the tense system and the valency patterns, but there has been little to no research done on these topics. 3. Conclusion The 12 typological features discussed and the three stages of Nahuatl touched upon are summarized in Table 1.4 By comparing the second and the third columns, it can be seen that Nahuatl has quite a few features in common with the other Uto-Aztecan languages. The grey cells highlight the features common to Nahuatl and the other Mesoamerican languages, which make Nahuatl a part of the Mesoamerican Linguistic Area. This typological change is likely contact-induced since the Nahuatl speakers supposedly immigrated to Central Mexico and further south thereby coming into contact with speakers of Mesoamerican languages. A comparison of the third and fourth columns reveals that still more changes have taken place in the change from the stage termed “Nahuatl as a Mesoamerican language” to the stage called “Colonial Nahuatl”. These – or some of these – changes may be ascribed to the contact with the Spaniards and the Spanish language. However, the present work has not been carried out on a quantitative base, and a quantitative study may be the first step in order to gain more substantive evidence for the changes observed. In order to provide more evidence for the hypothesis that the changes observed in colonial Nahuatl may be termed “contact-induced changes” (cf. Thomason 2001), we have to compare the changes observed in Nahuatl with changes that have taken place in other Mesoamerican languages, given that all Mesoamerican languages have been exposed to contact with Spanish. Of particular interest are languages that – like Nahuatl – are polysynthetic with an agglutinative morphology. By means of such a comparison we may be able to set up more substantive hypotheses with regard to the kind of contact-induced changes we may expect for polysynthetic languages which came into contact with a more isolating language.

22 Anne Jensen Appendix Table 1. Three stages of Nahuatl Typological features

Uto-Aztecan

Nahuatl

Colonial Nahuatl Spanish

1. Constituent order

SOV

V (O) S (O)

S V and V S

2. Adpositions

Postpositions

postposition

postposition

relational nouns

relational nouns

S V and V S

prepositions

prepositions

3. Relative clauses

pre- and posthead

pre- and posthead

pre-head

pre-head

4. Possessive constructions

he/his-dog the man

he/his-dog the man

piya ‘have’

tener ‘have’

there is his dog

there is his dog possessive nouns possessive nouns

5. Copula constructions

zero-copula

zero-copula

zero-copula

ser ‘be’

6. Adjectives

no

no

no + Spanish loans

adjectives

7. Impersonal constructions

yes

yes

no?

no

8. Incorporation

yes

yes

yes?

no

9a. Causative

yes

yes

yes/nonproductive?

no

9b. Applicative

yes

yes

yes

no

10. Plural of inanimate no nouns

no

for some yes

yes

11. Numeral system

no?

vigesimal

vigesimal / decimal

decimal

12. Classifiers

no

yes

yes?

no

Notes 1. This section is mainly based on Langacker’s (1976, 1979 [ed.]) work on UtoAztecan. 2. ABS = absolutive = non-possessed, ANIM = animate, APPL = applicative, AUG = augment, present in the preterite and the pluperfect , CAUS = causative, DEF = definite, dem = demonstrative, DERI = derivational morpheme, DIR = directional

Hispanisation in Colonial Nahuatl?

23

prefix, DIRE = directional inflection with a final meaning, FUT = future, HON = honorific, IMP = imperative, IMPF = imperfective inflection – which may not have strictly imperfective meaning, IMPS = impersonal, INAN(IM) = inanimate, IRR = irrealis, LIG = ligature, NEG = negation, OBJ = object, PART = particle, PL = plural, POSS = possessor prefix showing the person and number of the possessor, or possessor suffix showing the number of the possessed, PRET = preterit, PRON = pronoun, PTC = participle, RED = reduplication, REFL = reflexive prefix, SG = singular, SUB = subject, VET = vetative. c = dental affricate with dentalalveolar release,  = dental affricate with palato-alveolar release,  = dental affricate with lateral release,  = post-alveolar fricative,
= glottal stop. CF = Sahagún, Bernardino de (1979): Codice Florentino, COD = Anderson, Arthur J. et al. (1976): Beyond the Codices, TLA = Klaus, Susanne (ed.) (1999): Análes de Tlatelolco. 3. The capital letters are used for indicating that we have no evidence as to the length of the vowels. 4. The question marks in the column for colonial Nahuatl indicates that there is no empirical base for proposing a firm yes or no, or that the construction in question seems to be rather marginal.

References Anderson, Arthur J.O., Frances Berdan, and James Lockhart (eds.) 1976 Beyond the Codices. The Nahua view of Colonial Mexico. Berkeley: University of California Press. Anderson, Arthur J. O. and Susan Schroeder (eds.) 1997 Codex Chimalpahin.Volume 1 & 2. Oklahoma: University of Oklahoma Press. Campbell, Lyle 1987 Syntactic change in Pipil. International Journal of American Linguistics 53 (3): 253–280. 1997 American Indian languages. The historical linguistics of native America. New York /Oxford: Oxford University Press. Campbell, Lyle and Alice Harrison 1995 Historical syntax in cross-linguistic perspective. Cambridge: Cambridge University Press. Campbell Lyle, Terrence Kaufman, and Thomas C. Smith-Stark 1986 Meso-America as a linguistic area. Language 62.3: 530–558. Canger, Una 1976 Field notes. 1980 Five studies inspired by Nahuatl verbs in -oa. Copenhagen: The Linguistic Circle of Copenhagen.

24 Anne Jensen 1988

Nahuatl dialectology – a survey and some suggestions. International Journal of American Linguistics 54 (1): 28–72. 1996 Is there a passive in Nahuatl? In Content, expression and structure. Studies in Danish functional grammar, Elisabeth Engberg-Pedersen, Lisbeth Falster Jacobsen, Michael Fortescue, Peter Harder, and Lars Heltoft (eds.), 1–15. Amsterdam /Philadelphia: John Benjamins. Carochi, Horacio, S. J. 1983 [1645] Arte de la lengua mexicana con la declaracion de los adverbios della. Facsimile of 1645 edition. México: Universidad Nacional Autónoma de México. Cline, S. L. (ed.) 1993 The book of tributes. Early sixteenth-century Morelos Censuses from Morelos. Los Angeles: UCLA Latin American Center Publications. Hinz, Eike, Claudine Hartau and Marie-Luise Heimann-Koenen (eds.) 1983 Aztekischer Zensus. Zur indianischen Wirtschaft und Gesellschaft im Marquesado um 1540. Band I & II. Hannover: Verlag für Ethnologie. Jensen, Anne 1995/97 Field notes. forthc. a The emergence of progressive aspect in Modern Nahuatl. forthc. b Syntactic changes in Nahuatl – contact-induced or not? Karttunen, Frances and James Lockhart 1976 Nahuatl in the Middle Years. Berkeley: University of California Press. 1978 Textos en náhuatl del siglo XVIII: un documento de Amecameca, 1746. Estudios de Cultura Náhuatl 3: 153–175. Klaus, Susanne (ed.) 1999 Anales de Tlatelolco. Markt Schwaben: Verlag Anton Saurwein. Langacker, Ronald W. 1976 Non-distinct arguments in Uto-Aztecan. Berkeley and Los Angeles: University of California Press. Langacker, Ronald W. (ed.) 1977 Studies in Uto-Aztecan grammar. Vol. 1: an overview of Uto-Aztecan grammar. Dallas: The Summer Institute of Linguistics and The University of Texas at Arlington. 1979 Studies in Uto-Aztecan grammar. Vol. 2: Modern Aztec grammatical sketches. Dallas: The Summer Institute of Linguistics/The University of Texas at Arlington. 1984 Studies in Uto-Aztecan grammar. Vol. 4: Southern Uto-Aztecan grammatical sketches. Dallas: The Summer Institute of Linguistics / The University of Texas at Arlington. Launey, Michel 1986 Catégories et opérations dans la grammaire nahuatl. Thèse présentée à l’Université de Paris-IV pour l’obtention du Doctorat d’Etat (spécialité: Linguistique). Paris.

Hispanisation in Colonial Nahuatl? 1992

25

Introducción a la lengua y a la literatura náhuatl. México: Universidad Nacional Autónoma de México. Lockhart, James 1991 Nahuas and Spaniards. Postconquest Central Mexican history and philology. Stanford: Stanford University Press. 1992 The Nahuas. After the Conquest. Stanford: Stanford University Press. Molina, Fray Alonso de 1970 [1571] Vocabolario en lengua castellana y mexicana y mexicana y castellana. México: Editiorial Porrua, S.A. Sahagún, Bernardino de 1979 Codice Florentino: el manuscrito 218–20 de la collección Palatina de la Biblioteca Medicea Laurenziana. Florencia: Biblioteca Medicea Laurenziana. Smith-Stark, Thomas C. 1994 Mesoamerican calques. In Investigaciones lingüisticas en Mesoamérica, Carolyn J. MacKay and Verónica Vázquez (eds.), 15–50. Mexico: Universidad Nacional Autónoma de México. Suárez, Jorge A. 1977 La influencia del español en la estructura gramatical del náhuatl. Anuario de Letras, Vol. XV. México. Sullivan, Thelma D. 1987 Documentos tlaxcaltecas del siglo XVI en lengua náhuatl. México: Universidad Nacional Autónoma de México. Thomason, Sarah G. 2001 Language contact – an introduction. Edinburgh: Edinburgh University Press.

The Hispanisation of modern Nahuatl varieties1 José Antonio Flores Farfán

1. Introduction Historically speaking, along with Maya (Yucatec), Nahuatl has been the most investigated language of the whole American hemisphere. It has been researched from many different perspectives including the description of linguistic and, to a lesser extent, sociolinguistic structures (for a good example cf. Hill and Hill 1986). Nevertheless, the cumulative principle advocated by Labov (1972) is a case in point for these and any other Mesoamerican languages, since the more we know about them, the more questions are opened. For instance, although extensive research on Nahuatl dialectology has been carried out (cf. Canger 1988), it is not clear whether the postulated dialects are reaching the point where they can be considered separate languages, a discussion the surface of which has hardly been scratched (but see Suárez 1983). In the same vein, consider that although a number of Nahuatl grammars are today available in published form, including grammars on so called Classical Nahuatl (cf. Andrews 1975; Launey 1995)2, these are meant only as introductory books or sketches of the language; they do not to mention studies of the pragmatics of Nahuatl (e.g. the structure of conversational Nahuatl), the use and acquisition of different (e.g. honorific) structures in everyday usage, or, alongside the investigation of different Nahuatl genres, the study of Nahuatl variability, and particularly of Nahuatl variability due to contact with Spanish – the topic advanced with the present contribution. Given this context, this paper provides a review of the lesser known facts related to the effects of Spanish contact on Nahuatl, a topic on which even though we have a good amount of studies (cf. for example Hill and Hill 1986; Karttunen and Lockhart 1976a; Lockhart 1992; Flores Farfán 1999), there are still a number of issues waiting to be investigated, or that call for convincing explanations. These relate to the study of different contact-influenced varieties, especially those that are reaching the brink of extinction. In this sense, I will specifically analyze the contact phenomena linked to heavily obsolescent varieties;3 documenting and analyzing the extreme effects of Spanish on Nahuatl that correspond to a social situation

28 José Antonio Flores Farfán in which the language is swiftly fading away. For this purpose, it will prove useful to refer to non- or minimally Hispanised varieties, as represented by Classical, and, to a lesser extent colonial and even monolingual contemporary Nahuatl, or, in short, conservative Nahuatl varieties. An excellent, immediate reference to the most conservative varieties that can be reconstructed by means of an analysis of early colonial sources is the chapter by Jensen (this volume), which refers to Colonial Nahuatl, included in this volume. Several qualifications must be made before proceeding to the analysis of the Nahuatl materials proper. On the one hand, even though I am aware of the importance of understanding the interface of different layers related to linguistic analysis, for the sake of clarity different levels of linguistic organization will be distinguished as if they where separate entities, which of course they are not. On the contrary, one of the basic tenets of the present contribution is to demonstrate the close relationship between language and society, as evidenced in the inextricable link between extreme linguistic simplification and obsolescence, as presented here, and the social context which sustains and nurtures language shift. On the other hand, many of the points dealt with relate directly to language change. Since I concentrate on Nahuatl change resulting from contact with Spanish, my approach may give the impression that no attention is being paid to the possibility of explaining at least some of the phenomena depicted here on the basis of internal motivations, rather than as induced contact phenomena. Let me just stress that even if my interest is to understand the impact of Spanish on Nahuatl, I do not consider it useful to offer one sided explanations to understand phenomena which by nature are complex, and which probably require concurrent analyses both “internal”, and “external”. This will become clear while presenting, for instance, the morphophonemic shift of Nahuatl plural forms (see below). 2.

Exploring new facts and interpretations of Nahuatl Hispanisation

2.1. Phonology Historically, Nahuatl does not have the voicing contrasts which Spanish productively exploits. Yet it has appeared in Nahuatl as a phonemic innovation, if only in one fossilized pair. Compare:

The Hispanisation of modern Nahuatl varieties

(1)

(2)

ki-maka 5 OB -give ‘(S)he gives it to him/her’

29

(e.g. in Balsas Nahuatl: BN)4

ki-maga

(BN)

OB-hit

‘(S)he hits him/her’ It seems that such a distinction has slowly displaced or is displacing the more conservative Nahuatl form: (3)

ki-mak-ilia OB-give-APL ‘(S)he applies hand to him/her’; i.e., he hits him/her’6

The only explanation I can think of for the displacement of the native form is a sociolinguistic one, / / has a much more Spanish “feel”7. As stated above, the introduction of / / is nevertheless limited to this pair, although it is well established in most Nahuatl dialects. Voicing of Nahuatl phonemes is also attested in varieties of the language that are close to extinction, such as Chilacachapa Nahuatl (CHN): (4)

Chilacachapa Nahuatl a. malin-dika in koal crook-PROG DET snake ‘The snake crooks’ b. tewan-deh instead of tewan-teh -PL -PL ‘We’

(Godínez 2002)

(Notice the lack of contrastive vowel length in these examples, another index of the extremely endangered status of this Nahuatl variety.) In less investigated Nahuatl varieties, such as those spoken in Acatlan Guerrero (AG), the introduction of other Spanish phonemes is reported, particularly the use of the bilabial fricative / / as in (5), (5a, c) and the labiodental fricative /f/ as in (5d), which substitutes for /w/, an “original” Nahuatl sound. Over twenty years ago, when I did fieldwork in Acatlan, these changes were only slightly attested in speakers of the generation that by then were in their early fifties. Today the use of Spanish phonemes has been thoroughly extended to those in their twenties, probably the last generation

30 José Antonio Flores Farfán that will speak Nahuatl in AG, unless a dramatic change reversing ongoing language shift occurs (cf. Palemón Arcos 2005): (5)

Acatlan Guerrero a. otibel totlahtol ‘You became able to speak our language (i.e. Nahuatl)’ (CN: tihuel) b. weeweentsiin (AG, 1979): bebentsin (AG, 2004) ‘Elder, person of respect’ c. weeye (AG, 1979): beye (AG, 2004) ‘Big’ d. nafa, tafa, yafa (AG) ‘I, you, (s)he’ (In 1979 I registered naWa, ‘I’, taWA, ‘you’, etc.) cf. CN: nehuatl, tehuatl, yehuatl.

As suggested in 5d [W] represents the unvoicing of the bilabial glide /w/, probably reflecting a transitional stage in the process of shifting towards the Spanish fricative, which, as can be seen, is today almost complete.8 With regard to the vowel system, in highly Hispanised and extremely endangered Nahuatl varieties, vowel quantity has lost productivity or is rapidly losing ground, as is also attested in (5). Since Spanish distinguishes between /o/ and /u/ it is also possible that /u/, which in Nahuatl is an allophone of /o/, could become a new vowel due to Spanish influence. Yet there is not enough convincing evidence to support this suggestion. The example provided by Amith (1991), who claims that a historical change has occurred giving rise to /u/ in Nahuatl, which he does not attribute to the Hispanisation of Nahuatl, is not at all conclusive and indeed controversial, since as can be seen the reduplicated form in (6) is compared to a bound morpheme in (7): (6)

mu-muwi RED-fear, fright ‘She is scared’

(BN)

(7)

mo-wi 2POS-road ‘Your way’

(BN)

A phonological simplification, due at least partially to Spanish influence, can be seen in the fact that /¢/, which is not part of the Spanish repertoire,

The Hispanisation of modern Nahuatl varieties

31

has become or is becoming /s/, present in both systems, which facilitate phonological convergence. This requires much more investigation, since /¢/ may be subject to a series of internal assimilations (cf. Dakin 1979: 9). Nevertheless, compare (8) and (9), in which no phonetic assimilation could explain the change from /¢/ to /s/: (8)

mitz-maca-z (CN)9 2OB-give-FUT ‘(S)he gives it to you’

vs.

mis-maka-s (BN) 2OB-give-FUT

(9)

m tztli ‘moon’

vs.

meestli

(CN)

(BN)

Another fact that can at least partially be attributed to the impact of Spanish on Nahuatl which already connects to the morphological level is the substitution of the glottal stop by -n: (10) nemi-h live-PL ‘They live’

(CN)

vs.

nemi-n live-PL

(BN)

Nasality in Nahuatl is in deed linked to plurality, as becomes evident when one enumerates Nahuatl plural suffixes belonging to different word classes: -meh, chichi-meh ‘dogs’; -kaan, xtlakwaa-kaan! Eat! (PL); -waan: nokalwaan ‘my houses’. In sources from the colonial period studied by Karttunen and Lockhart (e.g. 1976a: 12) -n substitutes the glottal stop in a wide geographical area and in a considerable span of time in the colonial era. Indications of an early written identification of both segments in turn may also be referencing an oral phenomenon which anteceded the inception of -n as the plural verbal marker in modern times. Such a colonial background includes the intrusion of final -n as hypercorrection along the lines of a formal stylistic register (Karttunen and Lockhart, 1976b, 1976a: 8–12, 1977) probably also indexing social variability, as shown elsewhere for CN (cf. Flores Farfán 2004; in press). However, in Karttunen and Lockhart (1976a: 10–11) we do not encounter instances of -n as a verbal plural marker, supporting the idea of final -n as an innovation, at least in modern BN. A purist, monolingual, Nahuatl perspective would tend to deny that the plural marker stems from Spanish, appealing exclusively to the possibility of the internal evolution of the nasal as a plural marker, which as shown in fact exists. From my point of view both explanations are valid; i.e., they do not exclude but rather complement each other. The convergence of both forces probably provides a more complete and realistic explanation of the

32 José Antonio Flores Farfán phenomenon, even when [contact] induced change results are crucial in the emergence of such an innovation. 2.2. Morphology The glottal stop substituted by -n is also attested in imperative forms: (11) xia-n! ‘Go!’ (PL)

(BN)

Given that this is only attested with the irregular verb yaw ‘to go’, which allows the reduction of the plural optative suffix -cn, this is more likely an internally motivated phenomenon – the reduction of a longer suffix. Yet in the bilinguals’ mind this internally motivated fact may be reinforcing, or historically has reinforced, the identification of the Nahuatl-Spanish plural form(s), facilitating the emergence of -n as an innovation. Moreover, the glottal stop represented by /h/ in CN, as in -tlazohtla ‘to love’, is apparently lost in BN, at least in the generation in their late 40s and early 50s in Xalitla, probably the last generation that today, if at all, still speaks, or at least understands, Nahuatl. Another interesting aspect, which also presents itself at the interface of the phonological and the morphological levels, is the change of the Nahuatl absolutive marker -tl to -l, a phenomenon actually present in double loans in CHN (cf. 13d, 16), as in: (12) o-ki-miki in kone-l PER-OB-die DET child-ABS ‘She killed the child [in an abortion]’

(CHN)

Notice that konel resembles analogous lexical items that have become part of the Mexican Spanish repertoire, as xochi-tl : xochi-l ‘flower’, a very common popular name in Mexico these days.10 We are here witnessing a case of double loan that apparently expands the Nahuatl (singular) absolutive repertoire. At least in CHN, coexistent forms such as those in (13) are in deed attested: (13) a. a-tl a-l -ABS -ABS ‘water’

(CHN)

The Hispanisation of modern Nahuatl varieties

33

Again notice the lack of contrastive vowel length in al, which in conservative and even less Hispanised varieties is /a:/. In Malinche Mexicano, as represented by San Miguel Xaltipan (SMX: cf. Nava Nava 2003) in the state of Tlaxcala, -l may also be functioning as an absolutive suffix (cf. Hill and Hill 1986, passim), a fact at least indirectly induced by Spanish influence: (13) b. tlahz-l -ABS ‘rubbish, filth’

(SMX)

Apparently this possibility is limited to nouns of which the stem also ends in -l, such as: (13) c. melaak ookichiih tliilmoo-l iic ye kaxtoo-l? -ABS -ABS ‘Didn’t she prepare black mole 15 days ago?’ (adapted from Nava Nava 2003: 91)

(SMX)

In most other Nahuatl dialectal varieties, including CN, these forms present the absolutive /-li/, as in BN tlatooh-li, ‘word, discourse, utterance’, a variant derived from the regressive assimilation of /-i/. An alternative analysis of -l in final position would state that such nouns belong to the noun class to which the stem belongs to -ø. This possibility is strongly supported by historical evidence (Karttunen and Lockhart 1976a; Lockhart 1992), and today attested in the preference to ascribe all loanwords or neologisms such as the fully integrated pitso ‘pig’, to a -ø suffix. Both interpretations conform to the possibility of blurring the Nahuatl clear-cut distinction between possessive and absolutive nominal themes in terms of allowing simplification of possessive and absolutive forms. Notice that both in conservative and Hispanised varieties such as BN such clear-cut distinction is retained. This implies separate sets of number suffixes for the absolutive and the possessive, roughly -ø, -tl, -tli, -in (SG); -tin (CN), -meh (BN) (PL), for the absolutive and -w,-ø (SG), -waan (PL). The existence of a single form -l for the singular absolutive and the singular possessive in CHN and SMX probably contributes to the blurring of the absolutive and possessive markers for separate nominal themes, just as in Spanish (see below). Yet at least in CHN a conflict could arise from the assumption that one usage is at stake in the face of double loans as konel in (12) and (16). Compare (13d) and (13e):

34 José Antonio Flores Farfán (13) d. lo vinieron a tirar on konel ‘They came to throw the child away’

(Godínez 2002: 126)

e. ay nokonel, ya tiak timonamiktia! (Godínez 2002: 75) ‘Oh my son, you are already going to get married’ As suggested here, highly simplified varieties favour all types of convergence in both directions, yet favouring the phonological and lexical convergence towards more Spanish items as shown in (13a) and (14), even when there are originally Nahuatl. These cases indicate how Spanish is gaining ground towards almost total Nahuatl extinction. Thus it is possible that the absolutive marker linked directly to the Spanish phonology might tend to prevail numerically. This is indeed the case with other Spanish phonological segments, as pinpointed above in (8) and (9) for the change from /¢/ to /s/. Moreover, consider the change from /¢/ to //, a typical substitution made by a speaker whose mother tongue is Spanish: (14) tepichinto (CHN): tepitz ntn ‘tiny’ (cf. Spanish chiquitito)11

(CN)

Phonological convergence is not limited to one segment, but rather presents itself in all those phonemes that Spanish speakers have reinterpreted by means of Spanish phonology.12 Moreover, double loans are not only used in CHN, but are also an index to how near these varieties are to extinction. The use of Nahuatlisms in local Spanish is probably more viable communicatively speaking than the use of equivalent Nahuatl forms, which are limited to a handful of elder speakers. Thus we find infinitive forms as in (15), where the Nahuatlism in fact has the original Nahuatl meaning: (15) tepachear (CHN) ‘to smash with a rock’ (cf. Mexican Spanish tepache, an alcoholic beverage made from fermented pineapple, derived from tepatz n) In this sense we find lexical items that are identical in CHN and local Spanish, as in: (16) so konel ‘his son’ (cf. 12, 13d, 13e)

(local Chilacachapa Spanish)

Thus the process of substituting the absolutive marker -tl with -l can be interpreted as a clear index of the Hispanisation of extremely endangered

The Hispanisation of modern Nahuatl varieties

35

Nahuatl varieties, as is the case of CHN. In the long run, if this variety would survive at all (a highly unlikely possibility), it is indeed not only possible that the -l absolutive marker would tend to prevail, but also that the clear-cut distinction between absolutive versus possessive forms would be completely blurred, then finally disappear, as a transitional stage to total displacement, as is already attested in: (17) no-kone-tl 1POS-child-ABS ‘It is my son’

(CHN)

(18) no-yolo-tl 1POS-heart-ABS ‘It is my heart’

(CHN)

In 1955 Weitlaner and Barlow had already registered a similar form: (19) a. i-kokone-l 3POS-doll-ABS ‘She is his doll’

(CHN)

50 years later, such an innovation is also appearing in similar situations where Spanish is emerging as the primary, or even the mother tongue, of the communities, as is the case of SMX, in the state of Tlaxcala (cf. Hill and Hill 1986): (19) b. mo-soowaltsiin-tle 2POS-female-ABS ‘She is your wife’

(SMX) (adapted from Nava Nava 2003: 94, 104)

These last four examples turn out to be ungrammatical not only in CN and conservative Nahuatl varieties, but even in other Hispanised and endangered modalities of the language, such as Xalitla Nahuatl (where most of my BN examples come from). Since in Spanish, unlike Nahuatl, one does not distinguish between possessed and absolutive morphology (cf. Flores Farfán 1999: 92–101), this is the type of simplification that a native Spanishspeaking person would produce while learning Nahuatl. This phenomenon eloquently indexes the inception of Spanish as the primary tongue in several of these communities. Other examples that in CHN can be attributed to the impact of Spanish on Nahuatl are a series of deletions of otherwise Nahuatl obligatory morphology, such as what occurs in (20a), where a simplified Spanish-like Nahuatl

36 José Antonio Flores Farfán form is produced and also presents the case of konel ‘child’, which as we have seen, is in turn borrowed back to Nahuatl, becoming a double loan: (20) a. o-ki-miki in kone-l PER-OB-die DET child-ABS ‘She killed the child’

(CHN)

The canonical Nahuatl form creating a transitive verb derived from an intransitive form, in this case miki ‘to die’, would require, apart from the object prefix (which is the only trace of transitivity in 20a), a transitive suffix, specifically the causative -tia. Thus the full-fledged Nahuatl form would be: (20) b. oo-ki-mik-ti in konee-tl PER-OB-die-CAUS DET child-ABS ‘She killed the child’

(BN)

Notice how, canonically, the creation of the transitive verb also requires the deletion of -i from the stem, a fact that actually constitutes part of the criteria to classify a verb in a specific class (those which delete a final vowel in the perfect tense). This is not the case in the simplified version of okimiki in (20a), which actually only resembles the Spanish requirement to produce a transitive form from an intransitive verb: the presence of a clitic (as in duerme ‘he sleeps’ > lo duerme ‘he puts him/her to sleep; e.g. a baby).13 As shown, the canonical Nahuatl grammatical form requires a causative suffix -tia. The deletion of obligatory morphology probably due to Spanish influence is also the case in: (21)

tleko! ‘Get up!’

(CHN)

Compare with CN: (22) xi-tleco! IMP-ascend ‘Get up!’ Or compare the following expressions in CHN and Chilacachapa local Spanish: (23) mitote kwali! ‘Dance properly!’

(CHN)

The Hispanisation of modern Nahuatl varieties

37

(24) mitote como se debe! ‘Dance properly!’ Such extreme convergence with Spanish has produced the potential for the eradication of the canonically Nahuatl clear cut distinction between indicative and optative forms, as in: (25) amo chiwa-s pendejo! NEG do-FUT stupid ‘Don’t play stupid!’

(CHN)

Deletion of morphological prefixes is not limited to imperative forms such as (21) and (23) but can extend to the whole personal paradigm, as witnessed in (25), as well as in (26), even when this last example presents the emphatic pronoun nal ‘I’, which in CN is nehuatl [nehwatl]: (26) tlamachichitol nal neki-s watch I want-FUT ‘I need a clock’

(CHN)

(27) amo mati NEG know ‘I don’t know’

(CHN)

As an overall index of strong Nahuatl Hispanisation notice the deletion of the obligatory specific subject and object prefixes in (23), (26) and (27). Other innovations in highly Hispanised varieties include the identification of the future marker -s, among other usages, with the Spanish infinitive (cf. Flores Farfán 1999), facilitating the emergence of forms such as CHN: (28) nal amo tleno ki-toti-s NEG REL OB-say-FUT I ‘I don’t know what to say’

(CHN)

The direct identification of -s with the Spanish infinitive clearly manifested itself in the speech of a 40 year-old speaker during elicitation, in which Spanish finite forms were provided in order to obtain Nahuatl equivalents:14 (29) mati-s -FUT ‘to know’

(CHN)

38 José Antonio Flores Farfán (30) kochi-s -FUT ‘to sleep’

(CHN)

(31) miki-s -FUT ‘to die’

(CHN)

(32) anelo-s -FUT ‘to swim’

(CHN)

As suggested, subject prefix deletion is attested not only with intransitive roots, but also with transitive forms, as in the case of the object prefix in (29) and (32a)15: (32a) para yaw-e tlali-s (CHN) PREP go-PL put-FUT ‘In order to go and put it [e.g. the cement to construct a house]’

2.3. Syntax Typologically, Hispanised Nahuatl varieties are becoming more analytic than synthetic (cf. Jensen this volume); i.e., closer to the colonial language. Historically this fact has also had a strong impact on the productivity of Nahuatl new word coinages, considerably limiting processes of composition and incorporation (cf. Whorf 1946). This trend is labelled here “syntagmatisation”, which of course favours lexical borrowing, the favourite target of common purist quizzing and shibboleths, such as numerals or neologisms in local Nahuatl ideologies (cf. Hill and Hill 1986; Flores Farfán 2001, 2005). Compare (33a) and (33b): (33) a. ø-tlaxkal-chiiwa-ø-ø 3SG-tortilla-make-PRES-SG ‘She makes tortillas’ b. ø-ki-chiiwa-ø-ø tlaxkahli 3SG-3OB-make-PRES-SG tortilla ‘She makes tortillas’

(BN)

(BN)

While (33a) is characteristic of more conservative modern Nahuatl varieties, (33b) tends to be superimposed as the general usage in highly Hispanised

The Hispanisation of modern Nahuatl varieties

39

modalities, at least in bilingual usage. Differences actually seem to conform to generational and historical differentials linked to more Hispanised varieties. Thus, while an elder in BN would still produce forms as (33a), the younger generation would, if at all, prefer more analytical forms such as (33b). Strong purist reactions to more Hispanised varieties could focus on similar analytical expressions and even on forms such as (33a) and (33b), ironically suggesting that both are due to Spanish influence. Even when both forms are attested in CN, such monolingual (purist) perspective would attribute a Hispanic origin to (33a) and (33b), probably pointing to the existence of tlaxkaloowa ‘she makes tortillas’, considering it the authentic, pure, or real Nahuatl word, derived form tlaxkahli ‘tortilla’, plus the verbalizer -oowa. Compare the following texts, which actually present the same poem, one in CN16, and the other in modern Amanalco, Tezcoco, Nahuatl (ATN): (34) a. Nonantzin, ihcuac nimikiz, motlecuilpan xinechtoka, ompa tiaz titlaxcalchihuaz ompa nopampa xichoca. Itla acah mitztlahtlaniz, ‘Nonatzin, tle ica tichoca?’ Xiquilhui, ‘Ca xoxouhqui in cuahuitl ihuan nechchochoctia ica cehcencah popoca’ (Peralta Ramírez 1994/1995: 342) b. Tinonakon, kwak nimikis, ixne:chtoka ikxitla motlekwil wan kwak tihchiwas motlaxkal ompa ixne:chcho:kili. Wan kwak aka mitstla?tlanilis, ‘Tinonankon, tleka timocho:ki:tia? Ixkilwi te?wa, ‘cosa xoxohki in kwawitl wan miyak poktli nechcho:ktia’ (Peralta Ramírez 1994/1995: 343) Holly mother, when I perish, burry me upon your oven. While you go to make tortillas weep there for me. And if someone asks you: ‘my lady, what makes you cry? Tell them that the wood is green and it produces a lot of smoke, ‘is all that smoke that makes me cry’ (my translation) The CN and the ATN version of this text present a number of differences, the most striking of which is the well-known contact induced change from a postpositional to a prepositional language. This is evidenced in the systematic dropping of the third person possessive prefix ii- in all relational forms in the ATN text (cf. for example CN i-hcuac nimiquiz vs. ATN kwak nimikis), which also illustrates the case of syntagmatisation in point. Other facts include the substitution of emphatic CN ca by the Spanish loan cosa and the general simplification of the Nahuatl general quantifying structure (e.g. in the ATN version of the poem, the numeral [distributive]

40 José Antonio Flores Farfán reduplication cehcenca ‘whole of a lot’ does not appear). It is also probable that several syntactic innovations, as depicted here and elsewhere, originated in calques from Spanish structures, including fixed expressions such as cosa xoxohki ‘green stuff’, as will become clear in the next section. A final syntactic example is: (35) para tiawe koa cemento buy ‘in order to go and buy cement’

(CHN)

(35) is based on a double infinitive Spanish construction, a verb of movement + verb = para ir a + verb ‘in order to go + verb’: i.e., para ir a comprar ‘in order to go and buy’. Notice that the second transitive verb is completely bare; the obligatory specific object prefix is missing. This resembles the process of formation of a quasi-infinitive in (32a), and approximates even more closely the Spanish infinitive formation (the less marked verbal form of the colonial language; a category historically absent from Nahuatl yet overwhelmingly imposed by most Spanish grammarians in their colonial descriptions, a fact that probably also favoured its identification with Nahuatl future marker -s). 2.4. Semantics and pragmatics Among many others, several cases of semantic calques that have become part of the Nahuatl repertoire are manifested in forms that are today fossilized as lexical items, especially in courtesy rules: (36) a. tlaxtlaawi ‘Thank you’ (derived from ø-ki-tlaxtlaawa-ø-ø) 3SG-OB-to pay-PRES-SG ‘(S)he pays her/him’

(BN)

The full-fledged form calqued from the Spanish courtesy formula que Dios te lo pague ‘May god pay you back’ is (36b), which, in turn, is probably the source of (36a). (36) b. maa dios ø-mits-tlaxtlaawa-ø-ø EX god 3SG-2OB-to pay-PRES-SG ‘May god pay you back’

(BN)

41

The Hispanisation of modern Nahuatl varieties

Comparing (36a) to (36b) it becomes evident that in (36a) the transitive verb is treated as an intransitive form. This fact is shown by the presence of a final -i, which is becoming part of the Nahuatl fossilized lexicon. This is actually an innovation which allows the deletion of the whole derivational morphology, giving rise to a highly unmarked form. At the same time, it also resembles Spanish gracias ‘thank you’. The reply to (36a) is: (36) c. x-tlah tliinoon NEG-something REL ‘You’re welcome’

(BN)

This is clearly modelled on Spanish no hay de que ‘nothing to thank for’. Similarly, in other varieties we encounter: (37) a. tlasokamati (Hueyapan, Morelos: HUEY, SAT) ‘Thank you’ (derived from ø-ki-tlasokamati-ø-ø) 3SG-OB-to be grateful-PRES-SG ‘(S)he thanks him /her’ Notice how similar calques that are fossilized as fixed expressions in turn reinforce the trend to “syntagmatise” Nahuatl typology and reduce the coinage potential of the language, as has been suggested since Whorf’s (1946) pioneering work. The fixed reply-formula equivalent to (36b) in these varieties is: (37) b. amoo tleen NEG

(HUEY)

REL

‘You’re welcome’

(cf. Spanish no hay de que)

Other calques that probably originated in speakers of Nahuatl as a second language include: (38) a. kenon t-unka-ø-ø? (BN) INT 2SG-to be-PRES-SG ‘How are you?’ (cf. Spanish ¿cómo estás? How are you?) Compare these with greetings in more conservative Nahuatl varieties, in which more specific forms are utilized. For instance, in much more vital varieties of the language one pinpoints the time of the day when one greets, as in:

42 José Antonio Flores Farfán (38) b. keenoon oo-ti-mo-tlaaihuii-ti-ø-ø? (HUEY) INT PER-2SG-REF-light-CAUS-PER-SG ‘How did you cause the light on yourself; i.e., how did you wake up?’ Nevertheless, even in conservative varieties a series of innovations modelled on Spanish greetings are encountered, such as a Hispanised equivalent of (38b): (38) c. tlaneextili ‘Good morning’

(BN) 17

Even when at first glance one could think that tlaneextili could be glossed as derived from ø-tla-neex-ti-lia-ø-ø : 3SG-appear-CAUS-APL-PRES-SG, which would be translated as ‘(s)he appears something to him/her’, in fact the analysis is far form being that transparent. Again, even if this lexical item can be reconstructed in terms of (transitive) Nahuatl derivational morphology, it is unlikely that it is transparent for speakers themselves, who treat it as a fossilized item. Over a decade ago, in Tepoztlan, Morelos (TEP), where Nahuatl is no longer spoken, I registered with the last speakers of the language an innovation in saying keemah ‘yes’, which seems to be modelled on Spanish cómo no ‘why not’: (39) keen aamo REL

NEG 18

‘Yes’

Other greeting courtesy rules based on Spanish formulas used by the same speakers in TEP included: (40) kwaali toonali ‘Good day; i.e., good morning’ (cf. Spanish buenos días ‘good morning’)

(TEP)

Or leave-taking courtesy rules as (41): (41) kwali otli ‘Good road; i.e., good bye’ (cf. Spanish buen camino ‘good bye’)

(TEP)

All these examples constitute part of the history of extreme Nahuatl convergence with Spanish. Although one could be inclined to interpret them as

The Hispanisation of modern Nahuatl varieties

43

speakers’ language survival strategies, in most cases, at least as depicted here, they represent clear linguistic indexes of a social situation in which the indigenous language is reaching the brink of extinction. 3. Concluding remarks In this contribution a series of examples of profoundly Hispanised Nahuatl varieties were provided, showing extreme processes of simplification, attrition and obsolescence that are leading to the extinction of Nahuatl in several communities in which the ancestral language is hardly (or by the time this chapter is published, not at all) spoken. However discouraging it may be, it is interesting that the processes depicted here resemble what in the sociolinguistic literature is depicted as the emergence of an incipient pidgin or an initial process of pidginisation. Yet in our case such a process is not giving rise to a separate new language or even a simplified functional variety of a language, such as a foreigner or baby talk register (cf. Ferguson 1969), at least not in the long run. Rather, extreme processes of convergence and simplification index a social situation in which speakers of Nahuatl no longer consider it important or even necessary to speak their ancestral tongue; giving rise to the establishment of the colonial language as the mother tongue of the community. Although ironically, they speak varieties of Spanish powerfully influenced by Nahuatl (cf. Flores Farfán 1998, 1999) which are also highly stigmatized, especially in the face of the standard superimposition of Spanish as the only national, hegemonic language. Notes 1.

2. 3. 4.

I gratefully acknowledge Michael Harrington for his help with the English text. I also received a great deal of generous support from Frances Karttunen, to whom I am particularly grateful. I also benefited from comments and suggestions from Dik Bakker and Georg Bossong, who are also gratefully acknowledged. Any shortcomings are of course of my own responsibility. For a discussion of the notion Classical Nahuatl from a sociolinguistic perspective see Flores Farfán (2004, 2005, in press). For the study of obsolescence see Dorian (1992). Since to understand the contact phenomena at the phonological level it is not necessary to describe the phonetic details, the script utilized, unless otherwise indicated, is a phonemic one, although it is roughly more phonetic than orthographic. Such a script is considered self-comprehensive (for a description of the phonological equivalences see Flores Farfán 1999).

44 José Antonio Flores Farfán absolutive, APL = applicative, CAUS = causative, DET = determinant, EX = exhortative, FUT = future, HON = honorific, IMP = imperative, INT = interrogative, OB = object, NEG = negative, P = person, PER = perfect, PL = plural, POS = possessive, PREP = preposition, PRES = present, PROG = progressive, RED = reduplication, REF = reflexive, REL = relationiser, SG = singular. 6. Frances Karttunen has called my attention on the similarity of this form to the English idiom ‘She really gave it to him’, meaning she assaulted him verbally. According to the same author, “…maybe maka was originally derived from mah/ mak ‘hand’, but I think that would have become opaque centuries ago.” (p.c.). 7. The idea of / / having such a Spanish flavour is also suggested in Hill and Hill (1986: 203, 347). 8. Following Karttunen (p. c.): “Another possibility is that rather than moving toward Spanish intervocalic bilabial fricatives, maybe naWa is an example of regressive devoicing of /w/ after /h/, [/h/ represents the glottal stop, as in CN /nehwatl, tehwatl/, yehwatl/.” Yet, this is a concurrent rather than an exclusive possibility. Moreover, as can be seen in (5d), /w/ is not only moving toward the bilabial fricative, but becoming the labio-dental fricative /f/, a much more Spanish sound. Although marginally, /f/ has been present for a long time in Nahuatl in the fully integrated loans fiero ‘dangerous, bad’ vs. fierro ‘iron’ a fact that may have reinforced or is reinforcing the substitution of /w/ by /f/. 9. Since there is a written tradition in CN, in this case Nahuatl is written following the normalized orthography proposed by Andrews (1975). 10. In Luces del Otomí we find one of the first attestations of the phonological substitution of -tl by -l recorded by early grammarians: “El [nombre] Otomí, que dan los españoles, parece ser el mismo que dan los mexicanos, aunque diminuto o mutilado. Es la razón que los españoles no pronuncian con perfección todos los términos mexicanos, principalmente los que tiene la partícula tl… Y así se advierte…en la palabra flor…xochitl…xuchil. Y esto puede haber acaecido [con la palabra otomí] que diciendo el mexicano Otomitl, el castellano haya dicho Otomí.” (Anónimo 1893: 6). 11. One might be tempted to say that chiquito is a fully integrated Nahuatl loan in Mexican Spanish, linked to Nahuatl tzitz quitn ‘tiny’ (CN). In the face of the existence of the Latin-derived form chico ( Head of > Modifier of > Modifier of Predicate Referential Referential Predicate

Phrase

Phrase

Phrases

Phrase

As other implicational hierarchies proposed in the literature, the parts-ofspeech hierarchy not only implies a sequence but also specifies a direction (expressed by >). Directionality determines the scope of lexicalization and grammaticalization processes and is particularly relevant to explaining nonattested cases. For the sake of our study, directionality may be considered a constraint to flexibility. Directionality may explain why, for instance, IQ speakers do not use Spanish adverbs in adjective or noun positions although this may well be permitted by the existence of a non-specialized lexical class of non-verbs. 3.3.1. Implications of the parts-of-speech theory to lexical borrowing The point of departure for the present research is that parts-of-speech systems of the languages involved in the borrowing process are relevant to determining the type and the number of borrowed lexical classes. To be specific, the parts-of-speech system of the target language co-determines the borrowing of lexical classes from the source language and their functional adaptation in the target system. The implications of the theory of parts-ofspeech systems to lexical borrowing include several hypotheses and subhypotheses that may be read from two standpoints: the perspective of the

Spanish lexical borrowing in Imbabura Quichua

101

source language (hypothesis 1 below) with emphasis on the recognition of lexical classes and their frequency; and the perspective of the target language with emphasis on the use and function of borrowed lexemes (hypotheses 2–4 below). In what follows I develop all the possible hypotheses in detail, although in this study only two of them (H2 and H3) will be tested on an empirical basis. The remaining two hypotheses are irrelevant for the analysis pursued here either because a cross-linguistic comparison is required for meaningful results (H1), or because the target language of this study (IQ) is not a rigid but a flexible language (H4). The testing of the hypotheses across typologically different languages with flexible and rigid parts of speech has been carried out elsewhere (Bakker et al. 2008). H1.

Contiguous borrowing hypothesis: the typological distance between the source language and the target language is bridged in the order given by the parts-of-speech hierarchy. To be specific:

a. The more to the left a lexical class, the greater the number of lexemes borrowed. Therefore, heads are borrowed more often than modifiers, just like modifiers of referential phrases are borrowed more often than those of predicate phrases. b. Languages that borrow lexemes from one lexical class of the source language, more often than not borrow lexemes from the previous lexical classes. Thus, a language that borrows modifiers of referential phrases, usually borrows heads of referential and predicate phrases but not necessarily modifiers of predicate phrases, unless it is a flexible language filling this syntactic slot with a specialized lexical class (cf. 1c). c. Flexible and rigid languages may borrow lexemes from the lexical class immediately following the last lexical class attested in their own system. Type-2 languages would borrow nouns in larger numbers since nouns are the lexical class following the last class in their system (verbs). Hypothesis 1 establishes the general conditions for lexical borrowing on the basis of a typological parameter (the parts-of-speech hierarchy). In principle it does not exclude borrowings from lexical classes outside the parts-ofspeech system of the target language, provided they belong to contiguous classes. In other words, it constrains lexical borrowing in contiguity terms but allows for differential uses of borrowed lexemes inside and outside the parts-of-speech system of the target language.

102 Jorge Gómez Rendón However, as the parameters for comparison are not absolute but relative with respect to the target language, the predictions encompassed by hypothesis 1 can be assessed only cross-linguistically, that is, when at least two target languages are compared. Therefore, the results here will not deal specifically with this hypothesis. The hypotheses presented here concern only language pairs in which the source is always a differentiated language and the targets either flexible or rigid languages. Other hypotheses concerning different language pairs are possible though not considered here for reasons of space. H2.

Functional adaptation hypothesis: the syntactic distribution of borrowed lexemes in the target language follows the distribution of native lexical classes. For instance, a language that borrows adjectives may use them as modifiers of referential and predicate phrases provided it has a lexical class of modifiers that fulfill both syntactic functions. This prediction is relevant not only to languages with differentiated parts-of-speech systems but also to languages with flexible and rigid systems. In all cases, the typological configuration of the target language does not undergo changes.

Table 4. An example of functional adaptation Diff-4: Source

Verb

Noun

Flex-3: Target

Verb

Noun

Syntactic slot

Head of predicate phrase

Head of referential phrase

Flex-3: Target

Verb

Noun

H3.

Adjective

Adverb

Modifier Modifier of referntial phrase

Modifier of predicate phrase

Modifier

Functional specialization hypothesis: borrowed lexemes fill only the syntactic slots they originally occupied in the source language, i.e. the borrowed members of one lexical class specialize in their original syntactic slot. This means, for instance, that borrowed adjectives and adverbs will be used only in their respective positions of modifiers of referential and predicate phrases but not interchangeably as if they formed one lexical class. This hypothesis applies only to flexible languages, which become gradually differentiated in the process.

Spanish lexical borrowing in Imbabura Quichua

103

Table 5. An example of functional specialization Diff-4: Source

Verb

Flex-2: Target

Verb

Syntactic slot

Head predicate phrase

Noun

Adverb

No-Verb Head referential phrase

Flex-2: Target

H4.

Adjective

Modifier referential phrase

Modifier predicate phrase

No-verb

Lexicalization hypothesis: borrowed lexemes fill empty slots in the parts-of-speech system of the target language. Thus, if a language does not have a lexical class to fill certain syntactic slot, borrowings of one lexical class come to fill that slot. This hypothesis is applicable only to rigid languages, which become gradually differentiated in the process.

Table 6. An example of lexicalization Diff 4: Source

Verb

Noun

Adjective

Adverb

Rig 6: Target

Verb

Noun

—

—

Syntactic slot

Head predicate phrase

Head referential phrase

Modifier referential phrase

Modifier predicate phrase

Rig 6: Target

Verb

Noun

Adjective

Adverb

All hypotheses imply and exclude each other in several ways. Hypotheses 2–4 have to do with the use of borrowed lexemes. Hypotheses 3–4 are applicable only if hypothesis 2 is not fulfilled because the former imply a change in the parts- of-speech system of the target language while the latter implies no modification. In order to make predictions about Spanish lexical borrowing in IQ, a typological classification of both languages according to their parts of speech is required. This will be done in the next section.

104 Jorge Gómez Rendón 3.4.

A typological classification of Spanish and IQ according to their parts of speech

Imbabura Quechua is a good example of a language that has one lexical class for heads of predicate phrases and one for other syntactic slots. Therefore, IQ is a verb-nonverb language and ranks second in the scale of flexibility of parts-of speech systems. Schachter (1985: 17) gives the following examples to prove this claim. (1)

rikashaka: alkalde+ta see:PST:1SG mayor+ACC12 ‘I saw the mayor’

(2)

chay alkalde runa DEM mayor man ‘that man who is mayor’

(3)

rikashaka: hatun+ta see:PST:1SG big+ACC ‘I saw the big one’

(4)

chay hatun runa DEM big man ‘that big man’

In these examples, hatun is both a referential phrase modifier (4) and a referential phrase head (3); similarly, alkalde is both a modifier of the noun runa (2) and a noun head itself (1). Evidence against this classification has been presented by Beck (2002: 144ff.) and is discussed extensively in Bakker et al. (2008). The arguments presented are insufficient however, and I assume in this paper that Quichua is a language that makes no distinction between nouns, adjectives and adverbs. The functional adaptation of Spanish borrowings as explained in the following sections supports this classification (cf. Tables 10a–b). 3.4.1. The source language: Spanish In broad terms Spanish is a language with a differentiated parts-of-speech system (type 4). There is however some evidence of flexibility with respect to modifiers, though this is not enough to make Spanish a flexible language. The following examples illustrate such claim:

Spanish lexical borrowing in Imbabura Quichua

(5)

105

no

pude tomar el tren rápido could take ART train fast ‘I could not take the fast train’/‘I could not take the train rapidly’ NEG

(6)

el

tiempo pasa rápidamente/rápido ART time pass. by: PRES swiftly /swift ‘Time passes by swiftly’

(7)

dime la verdad sinceramente/sincero* honestly /honest tell: 1DAT ART truth ‘Tell me the truth honestly’

(8)

a. casa-cuna crib-house ‘nursery’

b. pinta-labios paint.lips ‘lipstick’

(9)

a. la

b. la

casa piedra* house stone ‘the stone house’

ART

casa de piedra house of stone ‘the house of stone’

ART

While some Spanish adjectives may be used as predicate phrase modifiers (5–6), others cannot (7) and require the mente ending to this purpose. The first group of adjectives is in fact a closed class. On the other hand, virtually all adjectives can be used as heads of referential phrases but not all nouns may be used as adjectives as illustrated by the ungrammaticality of (9a), where a prepositional connective between both nouns is required to make it a well-formed phrase (9b). Unlike Quechua, Spanish notional nouns cannot be used as modifiers, which could also explain the fact that noun-noun compounds are exceptional in Spanish (8a) as opposed to verb-noun compounds (8b). On these grounds, we claim that Spanish is a type-4 language with specific lexical classes for all syntactic slots, though some degree of flexibility is evidenced in modifiers. 3.5. Expected borrowings from Spanish in Quichua according to parts of speech Spanish and IQ do not have contiguous systems of parts of speech, which has relevant consequences for the kind of borrowings expected and their syntactic use. Excepting hypotheses 1 – applicable only cross-linguistically – and 4 – not relevant because IQ is not a rigid language – the following predictions can be made:

106 Jorge Gómez Rendón a)

Functional adaptation hypothesis: since IQ has only two lexical classes (verbs and non-verbs), nouns, adjectives and adverbs borrowed from Spanish will be used in more than one syntactic position provided leftto-right directionality is observed. b) Functional specialization hypothesis: although IQ has one lexical class of non-verbs, lexemes borrowed from Spanish will be used according to the differentiated system of the source language. Thus verbs and nouns will be distinguished from modifiers. At the same time, some adjectives and adverbs will still be used indistinctively as modifiers. 4.

Phonetic and morphological accommodation of Spanish borrowings in IQ

The changes that Spanish loanwords undergo to be incorporated into IQ are various. The most visible change is the adaptation of Spanish lexemes into the phonological patterns of IQ. There are also few morphological changes, including the maintenance of certain Spanish morphemes. A third process is frozen borrowing, by which complete Spanish phrases are incorporated as unanalyzable units in IQ. 4.1. Phonological accommodation of Spanish borrowings in IQ The main sound changes experienced by Spanish borrowings in IQ have to do with the vowel system. Since Spanish and IQ have different vowel systems, a compromise between both seems to be the solution. (10) a. b. c. d. e.

IQ [mísa] IQ [mísa] / [míza] IQ [pátrun] IQ [últimu] / [último] IQ [prizidínti] / [prisidínte] / [prisidénte]

< Sp. /mésa/ < Sp. /mísa/ < Sp. /patrón/ < Sp. /último/ < Sp. /presidénte/

Spanish medial vowels are raised (/e/ >/i/, /o/ >/u/) or otherwise pronounced as close as possible to their Quichua equivalents (10a, c). Partial assimilation is more common for words with several medial vowels (10e), though nonassimilated Spanish words are not uncommon. Words with more than one medial vowel (10d–e) have different phonetic realizations depending on the environment, the speaker’s level of bilingualism, or the frequency of the word. In fact, the less frequent a word is used in everyday speech, the less

Spanish lexical borrowing in Imbabura Quichua

107

assimilated it is to IQ phonology. For Spanish words with several medial vowels, raising results in various ways to pronounce the same item. Homophones like (10a) and (10b) are usually disambiguated by the voicing of the intermediate sibilant. Nevertheless, phonological accommodation of borrowings is not always rule-governed and may be considered idiosyncratic inasmuch as it depends on the speaker’s level of bilingualism. Another phonological phenomenon found in borrowings across lexical classes is metathesis. While in some cases the syllabic order changes, in others, syllables are deleted or replaced. (11) tempora sañora almotrador

< temporada < zanahoria < administrador

(season, time) (carrot) (administrador)

Finally, a few cases include nouns borrowed in the guise of other nouns but with different meaning. (12) rifuirso (effort) kontrarina (to meet)

< refuerzo < encontrar

(reinforcement) (to meet with)

4.2. Morphological changes in Spanish borrowings in IQ Morphological changes in Spanish borrowings are relatively rare in comparison to phonological ones. As far as parts of speech are concerned, verbs are particularly prone to morphological changes while nouns, adjectives and adverbs are less affected. The most common morphological change is the drop of Spanish verbal endings. Since the stem is the base form to which IQ verbal morphology is added, infinitive endings are dropped systematically. This is not always the case however and one may find a few instances where finite or even cliticized forms serve as the stem for IQ verbal inflection. (13) tuka-ni play-1: PRES ‘I play (music)’


social factors > linguistic factors

This scale implies that Media Lengua-speakers play a decisive role in its maintenance, i.e., whether or not it forms an intermediate stage in the process of language-shift from Quechua to Spanish. The fact that it has disappeared in some communities, like Salcedo, where Muysken discovered it, suggests that, at least in those communities, Media Lengua did indeed form a transitional stage toward language shift. In other communities, however, it is unlikely that Media Lengua will disappear in the near future: it is used in schools5 and by influential members of the community. It thus seems that social prestige is a very important factor in the maintenance or loss of Media Lengua. Campbell (1993) also emphasizes the role of prestige in the process of borrowing, which they illustrate with examples of lexical borrowings as well as borrowed particles into Spanish and Amerindian languages respectively. The extent of bilingualism within a community has been shown to accelerate the process of CiCh (e.g., Brody 1987). For example, it has been argued that in the case of fully-fledged bilingualism, features of one of the language systems can ‘leak’ into the other language system. Stolz and Stolz (1996) argue that bilingualism is not a necessary requirement for CiCh, as they observed instances of incorporated Spanish structures in the language of monolingual (elderly) speakers. However, these two views are not necessarily incompatible: it is not unlikely that bilingual members of the community represent the initiators of CiCh, subsequently ‘infecting’ other, monolingual speakers of the community. This scenario is compatible with the

Spanish prepositions in Media Lengua: Redefining relexification

125

emergence of Media Lengua: the bilingual obreros created Media Lengua, after which it spread to the rest of the community as a function of expressing a mixed identity in changing times.

3.2. Linguistic factors In contrast to Thomason (2001), who argues that in CiCh “purely linguistic considerations are relevant but strictly secondary overall”, several scholars, amongst whom formalists like Lightfoot (e.g., 1991) state that syntactic change operates autonomously, i.e., independent of any possible semantic, pragmatic, discourse or sociolinguistic factors. This point of view correlates with the assumption of a strictly independent language faculty, like Universal Grammar. Because of the independence and unconsciousness of this language faculty, the course of CiCh could hardly be argued to depend on language-external factors. It is difficult to capture the existence of Media Lengua without taking into account language-external factors. However, it is not unthinkable that specific internal factors have played a role not so much in the mere existence of a contact language in the communities at hand, but considerably so in the final form Media Lengua has taken. Muysken (1999) proposes three linguistic universals of CiCh6: 1. CiCh occurs in function of filling functional gaps, 2. CiCh in tendency follows the innovatory possibilities of the changing language(s) involved, and 3. CiCh can only occur at those modules of syntax where the languages in contact are structurally compatible. The first universal is considered an important motivator for CiCh by a number of scholars (e.g., Campbell 1993). Campbell argues that this explains the extensive amount of Spanish coordinators and subordinators that are incorporated into Pipil (an Amerindian language from El Salvador), since this language was originally paratactic and therefore lacked these categories. Both Bakker and Hekking (1999) and Stolz and Stolz (1996) discuss the incorporation of prepositions in several Amerindian languages7. While Bakker and Hekking (1999) conclude that the presence of Spanish prepositions in Otomí should be understood as the result of a functional gap, Stolz and Stolz (1996) discard this type of analysis for their data, citing the ample occurrence of double constructions, where an original (native) element co-

126 Suzanne Dikker occurs with a Spanish element, as exemplified in (4b) for Nahuatl. (4a) through (4c) constitute alternative ways to convey identical propositions: (3) a. in b. in c. in

ii-

tlapachih-ca tlapachih-ca de tlapachih-ca de

DET POS.3-cover-NML

in in in

i-cal i-cal i-cal

o-huertz o-huertz o-huertz

POS DET POS.3-house ANT-fall.PERF

‘the roof of his house has come/fallen down’ Bakker and Hekking (1999) argue that in these cases the original element and the borrowed element show only partial synonymy, and that therefore a functional explanation still holds. However, the definition of a functional gap is highly problematic. First, the (generally vague) definition of ‘synonymy’ becomes crucial in the consideration of functional gaps. Second, it is not entirely clear what should be understood as functional as opposed to grammatical. For example, Campbell (1993) labels the borrowing of conjunctions, subordinating devices and discourse particles as the result of a grammatical gap in the RL, whereas Heath (1978) calls it a functional constraint on borrowing. The second universal is adopted by scholars who argue that language contact yields an acceleration of processes of internally motivated language. It coincides primarily with views in which there are no qualitative differences between diachronic development or ‘asexual’ reproduction (DeGraff 1999) on the one hand, and results of ‘non-genetic’, i.e., contact-induced paths of development (Thomason and Kaufman 1988). Although one can name a number of strong theoretical motivations to assume this universal, it is difficult to test its validity empirically, since it may be hard to define what the next logical step in a diachronic process would be. Therefore, it is not tested in the present study. The third universal is related to the second: languages in contact will only change at those modules where the native system allows it. Such notions of structural compatibility have been discussed extensively in the literature from the early days on (e.g., Weinreich 1953; Heath 1978; Moravcsik 1978). As a result of the different ways in which ‘compatability’ can be interpreted, Campbell (1993) calls for a unified and concrete definition, which is empirically testable. It does not fall within the scope of the present study to discuss this point in detail. However, the following concrete universal (Moravcsik 1978) is of particular importance to the discussion of the incorporation of Spanish preposistions into Media Lengua:

Spanish prepositions in Media Lengua: Redefining relexification

(4)

127

Prepositions are always borrowed as prepositions, and never as postpositions.

To his own surprise, Campbell (1998) has not been able to identify any counter-examples to this universal.8 Moreover, it appears that it is not only the case that prepositions are never borrowed as postpositions, but that, in addition, the word order typology of the borrowing language is of influence on the borrowability of adpositions. For example, Bakker and Hekking (1999) found twice as many instances of Spanish prepositions in Otomí than in Quechua. Otomí is a head-modifier language, while Quechua is a predominantly head-final language in terms of word order typology (cf., Greenberg 1966; Hawkins 1983). Tsunoda, Ueda and Itoh (1995) studied 19 of the word order parameters defined by Greenberg (VO/OV, NA/AN, VAux/ AuxV, NDem/DemN, VAdv/AdvV, NRelCl/RelClN, etc.) for 130 languages of ample genetic and geographic variety. They found that the position of the adposition relative to its complement is the strongest predictor for the overall head-modifier/modifier-head character of a language.9 This suggests that adpositions play a substantive role in the syntactic profile of a language as a whole. Thus, the results of Bakker and Hekking can tentatively be attributed to the fact that Quechua is a (postpositional) modifierhead language, which forms an impeding factor in the borrowing of prepositions from Spanish. Otomí, on the other hand, shows no typological restrictions to the borrowing of prepositions. These facts suggest that syntactic features like word order characteristics are not detachable from the lexical entry of a word in the process of borrowing, contrary to what Muysken (1997, 2000) predicts. The presence of prenominally (instead of postnominally) placed Spanish adpositions in Media Lengua, as well as the hampering of the incorporation of adpositions, would challenge Muysken’s relexification account. 3.3. Stages in contact-induced language change Many proposals have been formulated with regard to the order by which borrowing occurs, some more controversial than others. Scholars generally agree that lexical elements are borrowed before grammatical elements (e.g., Moravcsik 1978; Campbell 1993; Bakker and Hekking 1999). First, grammatical elements carry no extra-linguistic reference (e.g., Muysken 1997, 2000, see above), while lexical elements do. As a result, the latter are more salient in communicative contexts. Second, Heath (1978), as well as Tho-

128 Suzanne Dikker mason (2001) and others argue that the borrowability of an element depends on the degree of syntactic integration of the element in question. For example, evidence suggests that nouns are borrowed before verbs, since the latter carry an argument structure frame, while the former do not. This Principle of Local Functional Value (Heath 1978) obviously goes against Muysken’s (1997, 2000) view on relexification, since it suggests that phonological forms are not easily detachable from other properties of the lexical entry. Other relatively uncontroversial universal tendencies of the order of borrowing include that free morphemes are more easily borrowed than bound morphemes (e.g., Weinreich 1953), and that derivational affixes are borrowed before inflectional affixes (e.g., Moravcik 1978; Campbell 1993). As was shown in (3) above, some languages in contact show double constructions, in which the native item co-occurs with the borrowed item. According to Stolz and Stolz (1996) this is a wide-spread phenomenon in the Centro-American Sprachbund. Examples (5a)–(5d) are taken from Mixtec: (5) a. b. c. d. *

ni ni ni ni

candonda i siqui gobierno candonda i contra gobierno candonda i contra siqui gobierno candonda i siqui contra gobierno PRET rebel.PL 3 against against government ‘They rebelled against the government.’

These examples illustrate the pattern of the incorporation of Spanish prepositions into Mixtec: native and Spanish prepositions alternate and combine, but the borrowed preposition is always placed in a peripheral position with respect to the Mixtec preposition. Hence the ungrammaticality of (5d). The occurrence of double-constructions is neatly accounted for by the Overlap Model (Heine et al. 1991), in which innovations first coexist with original structures, until the native structures are pushed out of the system. Thomason’s (2001) Borrowing Scale incorporates most of the observed tendencies in the order of borrowing, plotting these against both the intensity of contact and the fluency of the speakers in the SL. She makes no reference to the phenomenon of double constructions, but it could easily be incorporated into the scale, since the idea of successive substitution is fully compatible with the step-wise description of Thomason. The Borrowing Scale can be schematized as follows:

Spanish prepositions in Media Lengua: Redefining relexification

129

Table 1. The Borrowing Scale (adapted from Thomason 2001) less
Contact Intensity  more

Borrowed features

Lexicon

Phonology

Morphology

Syntax

No fluency in SL

Partial fluency SL

Bilingualism RL-dominant

Full/Balanced Bilingualism

Non-basic vocabulary

+ mainly N’s

+

+

+

Function words

–

+ conj/adv.part

+ adp./pron’s

+

Basic vocabulary

–

–

+

+

Loan words

–

+

+

+

Native words

–

–

+

+

Prosodic features

–

–

+

+

Derivational affixes

–

–

+

+

Inflectional affixes

–

–

+

+

Morph. categories

–

–

+

+

New functions

–

(+)

(+)

+

Word order

–

+

+

+

Other

–

–

(+)

+

–

–

–

+

Major typological change

In a situation of casual contact, only non-basic vocabulary is borrowed from the SL. The borrowers have no fluency in the SL. In the second stage, with slightly more contact and fluency in the SL, borrowing occurs at different levels: apart from non-basic vocabulary, function words, some phonological features of the SL (like new phonemes), and some syntactic features (word order preferences and some new functions or restrictions) are adopted by the RL. If contact further intensifies, a situation of bilingualism arises, but the RL remains the dominant language. In this stage, most features can be borrowed, and structure is affected, including actual changes in the word

130 Suzanne Dikker order pattern. The fourth and last stage is characterized by full bilingualism. The dominant language is either the SL, or the SL and RL are equally important. As in the third stage, all features of the SL can be transferred into the RL, with the addition of possible major typological changes due to the influence of the SL. Thomason observes that the order of borrowing is more flexible when the languages in contact are typologically similar. Since Spanish and Quechua are typologically divergent, we would hypothesize that Media Lengua fits into the Borrowing Scale, with Quechua as the RL and Spanish forming the SL of the borrowed elements. Thomason (2001) argues that there are some qualitative as well as quantitative differences between ‘normal’ borrowing and Media Lengua. However, she observes that Bilingual Mixed Languages are likely to follow the borrowing scale closely. These two statements need not contradict one another if one understands qualitative differences in terms of social conditions. It is clear that the conditions for the genesis of Media Lengua have been qualitatively distinct from most of the other contact situations found. Media Lengua is an abrupt Bilingual Mixed Language10, and the motivation for the creation of Media Lengua was not communicative need, but rather the need for ethnic auto-identification. As far as quantity goes, Media Lengua borrowed more than twice as many lexical elements (90%, including basic vocabulary) as the most extreme case of lexical borrowing from Spanish into Quechua attested (40%). Muysken (1997) states that this fact proves that in the emergence of Media Lengua other mechanisms are involved than regular borrowing, but maybe this difference can be explained in terms of the specific social conditions under which Media Lengua arose, rather than assuming linguistic differences between relexification and more gradual processes of borrowing. Let us test if the characteristics of Media Lengua are explainable within the Borrowing Scale, as Thomason argues, or if we do in fact need a different explanation, like Muysken’s relexification account suggests. As far as the lexicon goes, Media Lengua would need to be situated in stage three of the Borrowing Scale, since also basic vocabulary, and arguably some affixes are borrowed from Spanish (see below). As for phonology, only in some words the original syllabic structure CV of Quechua is violated, like in the Spanish word sol, which is borrowed as sol as well as solo. This example also shows that the vowels are not always adapted to the Quechua system, since this would predict the transformation of the [o] into a [u]. These features are compatible with stage three. Even though strict relexification would imply that no suffixes are borrowed from Spanish, some Spanish morphology is in fact found in Media

Spanish prepositions in Media Lengua: Redefining relexification

131

Lengua. According to Muysken (1997), these instances do not form counterexamples to the claim that affixes are not relexified in Media Lengua. We find three Spanish affixes in Media Lengua: 1) The Spanish gerund suffix -ndu (-ndo in Spanish). This suffix is also found in earlier Indian-Spanish contact languages, and could have been inherited from those contact vernaculars; 2) The diminutive -itu/-ita. Muysken argues that, since these diminutive suffixes are also extensively borrowed in ‘normal’ Quechua, this would not form a clear counterexample either; and 3) The past participle suffix -do. This suffix is only borrowed in an adjectival use of the participle, so it could be argued that these adjectives are borrowed as unanalyzed chunks, therefore excluding -do as an independent suffix carrying a verbal function. However, even though I am unaware of any clear-cut counterexamples that would lead to a total rejection of Muysken’s analysis of these affixes, his is not necessarily the most straightforward conclusion. If one argues that these affixes are indeed productive, then this would lead to a placement of Media Lengua in stage three or four. With respect to syntax, certain changes in the direction of Spanish are observed (Ocampo and Klee 1995): while Quechua is an OV language, Media Lengua shows some VO order, which is proper to Spanish. Moreover, coordinating as well as subordinating conjunctions have been borrowed from Spanish. Compare Spanish, Quechua, and Media Lengua in Table 2: Table 2. Spanish, Media Lengua, and Quechua Language module Lexicon

Syntax

Spanish

Media Lengua

Quechua

Non-basic

S/(Q)

S (90%)

Q/(S < 40%)

Basic

S

S/(Q)

Q

Word Order

H-M

M-H/H-M

M-H

Sub /Coord

S

S

– /(S)

S = Spanish; Q = Quechua; H = Head; M = Modifier

These data suggest that Media Lengua is situated in the third stage of the Borrowing Scale. Finally, although Media Lengua clearly shows some word order shifts, it cannot be concluded that the language portrays major typological differences with respect Quechua. Therefore, it goes too far to say that Media Lengua is situated in the fourth stage of the Borrowing Scale.

132 Suzanne Dikker The social conditions suggest that Media Lengua finds itself in the third stage as well: the Media Lengua community has a relatively intense contact with Spanish (through school and work), and the youngest generation primarily speaks Spanish, while the oldest generation is mainly Quechuaspeaking. This last fact suggests that Media Lengua may form a stage between Quechua dominance and Spanish dominance within the community. It seems that Media Lengua can be captured by the Borrowing Scale, since all the features documented fit in with the third stage. Most importantly, Media Lengua does not contradict the Borrowing Scale in any of the language modules discussed. For example, it is not the case that major word order changes are documented without the borrowing of basic vocabulary items. Media Lengua deviates from the model only to the extent that the vast majority of the lexical elements are borrowed from Spanish, while relatively few structural facets have been introduced. This, however, can largely be attributed to the agglutinating character of Quechua, facilitating the incorporation of lexical roots without affecting the surrounding morphosyntactic material. Therefore, I adopt the null hypothesis that Media Lengua is no different from other language contact situations and adheres to the same principles and restrictions. 3.4. The Principle of Functional Explanation Bakker and Hekking (1999) argue that the interrelation between social, functional, and formal factors is one of dichotomy, in the sense that nonlinguistic and functional linguistic factors primarily form the motivators for CiCh, while formal linguistic factors mainly function as constraints on the freedom of CiCh. The borrowing of Spanish prepositions into Otomí as opposed to Quechua is an example of the interaction between these factors. Importantly, syntax itself can never be the leading factor in CiCh. In order to formalize this principle, Bakker and Hekking (1999) formulate an interpretation of the Principle of Functional Explanation, as proposed by Dik (1986). This principle “orders the factors that are thought to explain the shape of languages in a hierarchy”, with non-linguistic factors in a higher position than linguistic factors. This hierarchy can be depicted as follows: (6) The Principle of Functional Explanation (Bakker and Hekking 1999) (N-LING: SOC > AREAL > DISC) > (LING: (FUNC: PRAG > SEM) > (FORM: SYNT > MORPH > PHON))







MOTIVATORS









CONSTRAINTS




Spanish prepositions in Media Lengua: Redefining relexification

133

If this principle applies to the Quechua case studied by Bakker and Hekking, one would expect a similar restriction to the incorporation of prepositions into Media Lengua, since it essentially possesses the morpho-syntactic frame of Quechua. However, if Media Lengua constitutes a special case of language contact, these restrictions may not apply. In other words: if Muysken is right in suggesting that the process of relexification as it has occurred in Media Lengua is qualitatively different from other instances of language contact, then we may not expect Media Lengua to obey the rules of normal CiCh. In this section, we have seen that Media Lengua adheres to most of the general principles and constraints on CiCh that have been proposed in the literature. In the next section, we first discuss the linguistic status of adpositions in more detail, before dedicating the remainder of the present paper to a case study, focusing on the behavior of Spanish prepositions and spatiotemporal predicates in Media Lengua. As has become clear, adpositions form an interesting testing ground for Muyken’s claim that only the phonological form of an element is transferred in the process of relexification. This lies in stark contrast with Moravcik’s generalization that prepositions are never borrowed as postpositions (see (4) above), and challenges the observation that the placement of adpositions forms a strong word order typological indicator (Tsunoda et al. 1995). 4. The linguistic status of adpositions Before formulating specific hypotheses with regard to the incorporation of Spanish prepositions into Media Lengua, it is important to define the linguistic status of the category of adpositions, especially in view of the strict division between lexicon and grammar observed in Media Lengua. We have already seen that adpositions form a strong indicator for word order characteristics. But how to define this ‘Cinderella Category’ (Mackenzie 1992) in linguistic terms? Do they form a (closed) lexical class, or should they rather be analyzed as functional elements? Different linguistic theories take divergent standpoints with regard to the analysis of adpositions. While within Generative Grammar ‘P’ is traditionally considered as one of the four lexical categories, other theories have not analyzed adpositions as one of the primary lexical categories. For example, they play no role in the parts-ofspeech typology defined by Hengeveld et al. (2004). Others consider the class of adpositions to be hybrid. Mackenzie (1992, 2001) for example, working within the framework of Functional Grammar (e.g., Dik 1997)

134 Suzanne Dikker subdivides the category of English prepositions into two classes: grammatical prepositions on the one hand, which form a direct instantiation of some basic semantic function, and lexical prepositions on the other, which are lexical predicates and are part of a separate lexical class together with spatio-temporal adverbs. According to Mackenzie, as well as e.g., Fernández López (1999), the core function of the adpositional category is spatiotemporal in nature. Notional and formal uses have emerged in a diachronic process of semantic extension and bleaching. For example, the preposition por (‘by’) in Spanish carries the core spatio-temporal function of Path, but it is also used notionally (e.g., Cause) as well as formally (introducing the Agent in the passive voice). Mackenzie (2001) defines five basic semantic functions, each with one corresponding preposition in English: locative (at), source (from), path (via), allative (to), and approximative (toward(s)). The other prepositions are predicative in nature. Example (7a) shows the underlying representation for a PP with a basic preposition and (7b) shows the underlying representation for a predicative PP11: (7) a. (d1p1: houseN) loc



‘at the house’

b. (d1p1:underP (d1x1:tableN) Ref) Source (d1p2:behindP (d1x2:doorN) Ref) All from

under

to

behind

 ‘from under the table to behind the door’ An empirical diagnostics for the validity of this sub-classification is that two basic prepositions cannot form a complex preposition, while the basic prepositions can be combined with predicative prepositions (i.e., *from at vs. from under). There are two further sources of evidence that support Mackenzie’s analysis of the adpositional category. First, the classification is reflected by the binary aspectual primitives ([+/– telic] and [+/– inchoative]), proposed by van Riemsdijk and Huijbregts (2002) to account for the internal functional subdivision of prepositional and case systems in a variety of languages. For example, to is the instantiation of [-inchoative / +telic]. Second, (Imbabura) Quechua possesses five spatio-temporal suffixes (Cole 1982), which denote exactly the functions described by Mackenzie. Further specifications of spatio-temporal functions are expressed through independent predicates (here: washa):

Spanish prepositions in Media Lengua: Redefining relexification

135

(8) (p1:washa P (d1x1: wasi N) ref) Dir  wasi washa-man house behind-DIR ‘to behind the house.’ Although for Spanish the empirical diagnostics are less clear-cut, it does not seem to pose a problem for Mackenzie’s analysis. Therefore, by analogy, the following categorization is assumed to hold across all three languages: Table 3. The basic prepositions: English, Quechua and Spanish Semantic function

English

Quechua

Spanish

locative source path direction allative

at from via to toward(s)

-pi -manta -ta -man -kaman

en de por a para

Mackenzie characterizes these adpositional elements as grammatical. Bakker and Siewierska (2002), on the other hand, argue that all elements with extralinguistic reference are lexical in nature. In this definition, the basic adpositions are lexical items, since they carry a clear extra-linguistic reference. The difference between the basic and predicative prepositions is that the ones presented in Table 3 are selected at a later stage in the derivation and they carry more abstract lexical information. This analysis fits within a view of the grammar-lexicon division as gradual rather than absolute12. Note that a lexical versus grammatical account of these instantiations of semantic functions is of particular relevance to the research question at hand. In fact, we may go as far as stating that the behavior of these elements in Media Lengua can be used as a diagnostics in this respect: if Media Lengua makes use of basic Spanish prepositions, this is an indication that they carry lexical features (or at least some independently interpretable semantic properties), given that Media Lengua shows a neat division between lexicon (Spanish) and morpho-syntax (Quechua). With respect to the predicative adpositions, Mackenzie argues that they form a separate lexical class, together with spatio-temporal adverbs like back. Within this lexical category ‘Ad’, the difference between the adpositional and adverbial predicates can be reduced to valency: while the adpositions are monovalent, selecting a single argument, the adverbs are avalent,

136 Suzanne Dikker not taking any arguments. Some predicates can be used both monovalently and avalently. For example, an Ad like behind can be used both as a preposition (behind the door), and as an adverb (he was left behind). In Spanish, valency reduction (or enhancement) is also observed, although it seems to work slightly different from English: many predicative prepositions can be analyzed as compounds of a spatio-temporal adverb and a basic preposition (e.g., detrás de ‘behind (of)’). In colloquial Spanish, some of these monovalent Ad’s may be used nominally (e.g., detrás suyo, ‘behind him’, lit.: ‘behind his’). In Ecuadorian Spanish, the use of preposed possessives has also been observed (e.g., en mi detrás, ‘behind me’, lit.: ‘in my behind’; cf., Bosque and Demonte 1999: ch. 9). As far as I am aware, this does not occur with predicative prepositions like entre, which cannot undergo valency reduction (e.g., *él caminó entre Ø, ‘he walked between Ø’; *en su entre, ‘in his/her/their between’). We have been able to identify Quechua equivalents of the group of basic prepositions (see Table 3), but it is unclear whether there exists a one-to-one correspondence between the Ad-category and the lexical elements that fulfill similar functions in Quechua. The lexemes used in similar contexts (cf., example 8) can best be analyzed as relational nouns13. When taking seriously Muysken’s observation that a relexified language possesses the same lexical categories as the RL, we may thus hypothesize that no Spanish predicative prepositions are incorporated into Media Lengua, except those that can be used nominally. In other words: we expect lexemes like detrás to occur in Media Lengua, but not prepositions like desde.

5. Hypotheses In summary, we have so far established the following. First, so far no empirical linguistic facts indicate that Media Lengua constitutes a qualitatively distinct case of language contact. In quantitative terms, we can argue that the social situation in which Media Lengua arose has accelerated the process of CiCh, such that the language reached stage three of Thomason’s Borrowing Scale at a faster rate than most other contact languages. On the other extreme of the Principle of Functional Explanation (see 6 above), we can identify both formal linguistic factors that impede CiCh, and linguistic characteristics that do not hamper CiCh. For example, the agglutinating nature of Quechua allows for a straightforward incorporation of Spanish lexical roots into Quechua slots. On the other hand, we may expect that the typological differences between Quechua and Spanish impede the incorpo-

Spanish prepositions in Media Lengua: Redefining relexification

137

ration of certain elements, more specifically: prepositions. First of all, it has been found that the nature of a language as either prepositional or postpositional is indicative of its head-modifier/modifier-head properties in domains other than the word order within the PP (Tsunoda, Ueda, and Itoh 1995). Following Bakker and Hekking’s (1999) findings that the incorporation of Spanish prepositions into Quechua is restricted when compared to Otomí (a head-initial language), we expect that the presence of prepositions in Media Lengua is limited. Following Moravcik’s (1978) generalization that prepositions are never borrowed as postpositions, we do not expect a postnominal substitution of the Quechua suffixes by Spanish prepositions. Second, the fact that Quechua does not possess a lexical category equivalent to Mackenzie’s Ad class should hamper the incorporation of predicative prepositions. Thus, we observe a conflict of factors: on the left side of the scale of the Principle of Functional Explanation we find ample motivations to incorporate all lexical roots from Spanish, but there are formal linguistic factors that possibly halt this process. It thus becomes an interesting question whether it is true that linguistic considerations are invariably secondary to social factors that motivate CiCh, as Thomason (2001) argues. If this is the case, then we should find Spanish lexical roots of all types, possibly at the cost of introducing an entirely new lexical category into the Quechua inventory, in addition to a rupture of the core word order typological characteristics. If, alternatively, Media Lengua does not form a normal instance of CiCh, but is rather the result of the unique process of relexification as defined by Muysken (1981, 1997), we may expect either of the following options: 1) if prepositions are lexical elements that can be stripped of all features (specifically those encoding word order properties), leaving a bare phonological form and a meaning correspondence, then there should be no (word order) restrictions on the incorporation of Spanish prepositions in Media Lengua and we expect Spanish prepositions to have substituted the Quechua spatiotemporal suffixes (as postpositions, i.e., in congruence with Quechua morpho-syntax); 2) Media Lengua shows no occurrences of Spanish prepositions, because they should be analyzed as grammatical items. We have seen that the presence or absence of basic Spanish prepositions in Media Lengua can provide evidence in favor or against analyzing them as lexical elements. If they are not incorporated into Media Lengua, this may indicate that Mackenzie (2001) is correct in considering the five basic prepositions as grammatical. If, on the other hand, they are present in Media Lengua, this strongly suggests that they should be analyzed as lexical elements, in accordance with the tendency that incorporated Spanish items are lexical in nature.

138 Suzanne Dikker 6.

The sample

For the purposes of this study, I analyzed the transcription of an interview (6000 words) that was carried out by Jorge Gómez Rendón (University of Amsterdam) with a leading figure of the Comunidad de Casco Valenzuela, Ecuador, in the year 2000. Only prepositional and adverbial predicates were included for analysis, excluding those of which the lexical classification has been under debate (durante, excepto, incluso, mediante, salvo and según, DeBruyne 1999; Pavón Lucero 1999). In addition, so-called argument marking prepositions (VanValin and LaPolla 1997), like con in soñar con (‘dream of’) were excluded from the analysis, since they are not selected in a productive way, but rather form part of the verbal predicate. Prepositions that are part of an idiom (e.g., a in a veces, ‘sometimes’) or a compound (de in coche de madera, ‘wooden car’), as well as strings of code-switching to Spanish, were also left out. 7.

Results and discussion

When excluding the cases mentioned above, we observe an interesting pattern: only the basic prepositions (i.e., en, de, por, a, and para, see Table 3), as well as avalent spatio-temporal predicates are used productively14. No predicative prepositions are found. Let us first take a closer look at the behavior of Spanish basic prepositions in Media Lengua, before discussing the contexts in which the spatiotemporal predicates were found. 7.1. Prepositions As was discussed above, Bakker and Hekking (1999) argue that the borrowing of Spanish prepositions into Quechua can be explained by assuming that Quechua portrays a functional gap, while Stolz and Stolz (1996) reject this analysis, in view of the ample occurrence of double constructions. In the sample under investigation some double constructions were found, where a basic Spanish preposition and a Quechua suffix, both denoting the same semantic function, are expressed simultaneously: (9) a. en

año 1986-pi entra-rca –ni magisterio-man year 1986-LOCQue enter-PRET-1.SG teaching-DIR ‘In the year 1986 I started teaching.’ LOCSp

Spanish prepositions in Media Lengua: Redefining relexification

139

b. pero en colegio o escuela-pi -ca but LOCSp high.school or school-LOCQue -TOP ‘but in high school or school…’ These double constructions alternate with sentences with only a Quechua suffix (10) or only a Spanish preposition (11): (10) a. yo mi naci-rca-ni C.C.V.-pi, 1959 año-pi I REFL be.born-PRET-1.SG C.C.V-LOCQue 1959 year-LOCQue ‘I was born in the C.C.V., in the year 1959.’ b. pero lei-shpa inglés-manta castellano-man but read-GER English-SOURCE Que Spanish-DIR Que ‘but I can translate from English to Spanish…’ (11) después ya más mayor-lla ca-shpa anda-rca-ni a Quito later already more older-LIM be-GER walk-PRET-1.SG DIRSp Quito ‘Later, when I was older, I went to Quito.’ These examples are incompatible with the idea that borrowing is primarily motivated by the existence of a (functional) gap in the RL: the Spanish and Quechua forms alternate as true synonyms (e.g., compare [9a] and [10a]) Moreover, even though one could argue the lack of Ad predicates in Quechua to constitute a gap, this is not filled in Media Lengua, since no Spanish predicative prepositions are found. The findings do fit in with the Overlap Model discussed above (Heine et al. 1991), to which Stolz and Stolz (1996) adhere. Importantly, all prepositions are found in pre-nominal position: no preposition has been incorporated as a postposition, which we would have expected if only the phonological form of the Spanish preposition were transferred to Media Lengua, leaving the Quechua morpho-syntactic frame intact. It thus appears that a strict relexification account cannot explain the present data, while a general account of Media Lengua as a result of extensive borrowing does hold. In fact, the findings are exactly how we would have expected them to be on the basis of Moravcik’s universal (see [4]). Moreover, the findings tie in with the word order patterns observed in other domains (see Table 3). It appears that Media Lengua occupies an intermediate position between the purely head-final language Quechua and the primarily head-initial language Spanish, and can be categorized as ‘predicate-medial’ (Hengeveld et al. 2004). Tsunoda et al. (1995)’s observed relationship between the position of adpositions and other word order parameters is strengthened by this correlation. In contrast, Muyken’s relexification

140 Suzanne Dikker account is weakened, since at least some morpho-syntactic subcategorization features do appear to be carried over with a borrowed element. This suggests the validity of the Principle of Local Functional Value (Heath 1978). Finally, the apparent ease by which Spanish basic prepositions are incorporated into Media Lengua suggests that they are lexical in nature, supporting Bakker and Siewierska’s (2002) account of the lexicon, and contra Mackenzie (1992, 2001). 7.2. Spatio-temporal predicates As was mentioned above, no predicative prepositions were found in the sample under investigation. However, we do observe avalent spatio-temporal predicates, which occupy the slot of a relational noun and are treated as nouns morphologically: (12) a. y un hueco arriba-pi mano izquierda-pa and a hole aboveSp-LOCQue hand left-GEN ‘and a hole above his left hand’ b. otro-ca abajo-manta corta-shpa another-TOP belowSp-SOURCEQue cut-GER ‘another was cutting from below’ c. cama dentro-manta ali-lla jaicata sali-shpa-ca bed insideSp-SOURCEQue fairly-LIM quickly leave-GER-TOP ‘getting out from (inside) the bed fairly quickly’ Like with the basic prepositions, the Spanish spatio-temporal predicates alternate with Quechua forms (13a), and they occur in combination with double constructions of the type observed above (13b): (13) a. todito jahua-pi, encima-pi cahora all aboveQue-LOCQue aboveSp-LOCQue while ‘(we were) all on top, while we were above…’ b. ya vuelta para arriba-man already returning DIRSp aboveSp-DIRQue ‘… once returning up (the slope)’ No combinations of a Spanish spatio-temporal predicate and a Quechua relational noun were encountered.

Spanish prepositions in Media Lengua: Redefining relexification

141

The number of occurrences of these types of elements is too small to establish with confidence whether they should be analyzed as avalent Ad’s or relational nouns. However, the null hypothesis would be that the latter is the case, i.e., that the Media Lengua speakers employed the limited nominal use of these types of predicates in colloquial Spanish to establish an analogy between the Quechua spatio-temporal elements and the Spanish ones. This hypothesis is tentatively corroborated by the observation that Ecuadorian Spanish shows a wider use of these predicates in nominal environments than most other colloquial dialects (see above). The absence of predicative adpositions, in contrast to the incorporation of ‘relational nouns’, confirms Muysken’s (1997) observation that a relexified language possesses the same lexical categories as its base language (Quechua). Note however, that the present findings do not challenge an analysis of Media Lengua as a ‘normal’ consequence of CiCh: Thomason’s Borrowing Scale suggests that major typological changes, e.g., in the form of introducing new lexical categories, only occur in the very last stage of borrowing, when a community finds itself at the verge of suffering language shift. Thus, the present findings merely discard an ‘anything goes’ scenario. In other words: they confirm that, even though there are socio-linguistic factors that form a drive to borrow all lexical items from Spanish, these factors cannot overrule formal linguistic constraints entirely. At least not at the stage Media Lengua finds itself at the moment.

8. Conclusions The first, general conclusion of the present study is that the characteristics of Media Lengua can not only be accounted for by adhering to general principles and constraints on contact-induced language change, but are even incompatible with Muysken’s (1981) model of relexification: if only the phonological shapes of the lexical roots were taken from Spanish, we would have expected Spanish prepositions to surface as postpositions in Media Lengua. However, the opposite was found, and Moravcik’s (1978) universal that prepositions are never borrowed as postpositions was not disconfirmed. Let us now look at Muysken’s universals of CiCh as introduced in section 3.2 above:15 1. CiCh occurs in function of filling functional gaps. 3. CiCh can only occur at those modules of syntax where the languages in contact are structurally compatible.

142 Suzanne Dikker With regard to the existence of functional gaps, we have seen that Media Lengua does not appear to have borrowed any prepositions from Spanish as a result of such a principle. Only those prepositions are borrowed which have a synonymous correlate in Quechua: the Spanish basic prepositions alternate with the Quechua suffixes, and combine with them in double constructions. Moreover, there are no indications that Media Lengua is inclined to introduce the Spanish lexical category Ad, whereas it does portray a functional gap in this respect. We may then conclude that the extensive borrowing of Spanish elements into Media Lengua is the result of the social need to express an in-between identity by means of language. Functional gaps may have provided an additional motivational factor, but the present data show that these do not in itself constitute either a necessary or sufficient condition for borrowing. The current findings support the third universal: even though the sociolinguistic pressure to incorporate all Spanish lexical items, some formal linguistic restrictions still hold. In particular, we have seen that the introduction of predicative prepositions is hampered because Quechua lacks a lexical category Ad. However, the universal of structural compatibility does not hold absolutely. Given that Spanish is a head-modifier prepositional language, while Quechua is an originally strict modifier-head postpositional language, the introduction of Spanish prepositions would imply a clash between two opposite systems, and we would have expected that the incorporation of Spanish preposition were blocked entirely. However, this is not what we find. Instead, it appears that Media Lengua has loosened its rigidity of word order patterns towards a Spanish system, including the presence of prepositions. Thus, although “purely linguistic considerations are relevant but strictly secondary overall” (Thomason 2001), they do play an important role in allowing the incorporation of linguistic elements, which suggests a balance between the right and left extremes of the Principle of Functional Explanation (see 6 above). In conclusion, the behavior of Spanish spatio-temporal elements in Media Lengua provides evidence against strict relexification, at least in the case of adpositions, but are compatible with more general accounts of contactinduced language change. Second, the findings support Mackenzie’s linguistic classification of spatio-temporal elements, but challenge his claim that the basic prepositions are grammatical in nature.

Spanish prepositions in Media Lengua: Redefining relexification

143

Notes 1.

2. 3.

4. 5. 6. 7.

8. 9. 10.

11.

12. 13. 14. 15.

Imbabura Quechua is a Quechua variety, spoken by approximately 30,000– 50,000 indigenous inhabitants of the province of Imbabura, Ecuador. According to Cole (1982), 15–30% of its speakers are monolingual, a percentage that has decreased in the past 25 years as a result of the acceleration of the influence of Spanish in the daily life of the indigenous population. Taken from Muysken (1997). ACC = accusative, ANT = anterior, BEN = beneficiary, DET = determiner, DIR = direction, GEN = genitive, GER = gerund, LIM = limitative, LOC = location, NML = nominalizer, NOM = nominative, POS = possessive, PROG = progressive, PERF = perfect, PL = plural, PRET = preterite, REFL = reflexive, SG = singular, TOP = topic vengo is the irregular first person singular form of the verb venir (‘to come’). The official language of instruction is Quechua, but teachers often switch to Media Lengua. The relevance of these universals for Media Lengua is discussed in the last section. See e.g., Bakker and Hekking (1999), Brody (1995), Campbell (1993), van Coetsem (2000), Garzón (1993), Muysken (1981, 1997, 2000), Ocampo and Klee (1995), and Thomason (2001) for other examples and accounts of Spanish borrowings into a variety of Amerindian languages. But see Clements (1996) and Smith (1977) for possible exceptions. Tsunoda et al. (1995) distinguish two basic language types: 1. prepositional; and 2. postpositional/adpositionless. Thomason (2001) distinguishes abrupt genesis of bilingual mixed languages from gradual genesis, a distinction that is reflected in both the social context of genesis and the linguistic outcome. For the purposes of the present study representations are simplified. See Dik (1997) and Mackenzie (1992, 2001) for an elaboration on the structure of the Underlying Representation within the theory of Functional Grammar. See Bakker and Siewierska (2002) for a detailed discussion of the lexicon within the framework of Functional Grammar. See Hengeveld et al. (2004) for a discussion of parts-of-speech typology, including the PoS system of Quechua. Because of the small text sample, no (descriptive) statistics are given. As was mentioned above, the second universal, claiming that CiCh follows the innovatory possibilities of the RL, is difficult to test empirically, and will therefore not be discussed.

144 Suzanne Dikker References Bakker, Dik and Ewald Hekking 1999 A functional approach to linguistic change through language contact: The case of Spanish and Otomí. Working Papers in Functional Grammar 71: 1–32. Bakker, Dik and Anna Siewierska 2002 Adpositions, the lexicon and expression rules. In New perspectives on argument structure in functional grammar, Ricardo Mairal Usón and María Jesús Pérez Quintero (eds.), 125–178. Berlin /New York: Mouton de Gruyter. Bosque, Ignacio and Violeta Demonte (eds.) 1999 Gramática descriptiva de la lengua española. Madrid: Espasa. Brody, Jill 1987 Particles borrowed from Spanish as discourse markers in Mayan languages. Anthropological Linguistics 29 (4): 507–521. 1995 Lending the ‘unborrowable’: Spanish discourse markers in indigenous American languages. In Spanish in four continents: Studies in language contact and bilingualism, Carmen Silva Corvalán (ed.), 132– 147. Washington D.C.: Georgetown University Press. Campbell, Lyle 1993 On proposed universals of grammatical borrowing. In Historical Linguistics 1989: Papers from the 9th International Conference on Historical Linguistics, Henk Aertsen and Robert J. Jeffers (eds.), 91– 109. Amsterdam: John Benjamins. 1998 Historical linguistics, an introduction. Edinburgh: Edinburgh University Press. Clements, Joseph C. 1996 The genesis of a language: The formation and development of Korlai Portuguese. Amsterdam: John Benjamins. Cole, Peter 1982 Imbabura Quechua. Amsterdam: North-Holland Publishing Company. DeBruyne, Jacques 1999 Las preposiciones. In Gramática descriptiva de la lengua española, Ignacio Bosque and Violeta Demonte (eds.), 657–704. Madrid: Espasa. DeGraff, Michel 1999 Language creation and language change: Creolization, diachrony, and development. Cambridge: MIT Press. Dik, Simon C. 1986 On the notion ‘Functional Explanation’. Working Papers in Functional Grammar (11). 1997 The theory of functional grammar. Part 1: The structure of the clause. Berlin /New York: Mouton de Gruyter.

Spanish prepositions in Media Lengua: Redefining relexification

145

Fernández López, María del Carmen 1999 Las preposiciones en Español. Valores y usos; construcciones preposicionales. Salamanca: Ediciones Colegio de España. Garzon, Susan 1993 The process of language death in a Mayan community in Southern Mexico. International Journal of the Sociology of Language 93: 53–66. Greenberg, Joseph H. 1966 Some universals of grammar with particular reference to the order if meaningful elements. In Universals of language (2nd edition), Joseph H. Greenberg (ed.), 73–113. Cambridge: MIT Press. Hawkins, John A. 1983 Word order universals. New York /London: Academic Press. Heath, Jeffrey 1978 Linguistic diffusion in Arnhem Land. Canberra: Australian Institute of Aboriginal Studies. Heine, Bernd, Ulrike Claudi, and Friederike Hünnemeyer 1991 Grammaticalization: a conceptual framework. Chicago: University of Chicago Press. Hengeveld, Kees, Jan Rijkhoff, and Anna Siewierska 2004 Parts-of-Speech systems and word order. Journal of Linguistics 40: 527–570. Lightfoot, David 1991 How to set parameters: arguments from language change. Cambridge: The MIT Press. Mackenzie, J. Lachlan 1992 English spatial prepositions in Functional Grammar. Working Papers in Functional Grammar 46. 2001 Adverbs and adpositions: The Cinderella categories of Functional Grammar. Revista Canaria de Estudios Ingleses 42: 119–135. Milroy, James 1993 On the social origins of language change. In Historical linguistics, problems and perspectives, Charles Jones (ed.), 215–236. London: Longman. Moravcsik, Edith A. 1978 Language contact. In Universals of Human Language, Joseph H. Greenberg, Charles A. Ferguson, and Edith A. Moravcsik (eds.), 93– 123. Stanford: Stanford University Press. Muysken, Pieter 1981 Halfway between Quechua and Spanish: The case for relexification. In Historicity and variation in Creole studies, Arnold Highfield and Albert Valdman (eds.), 52–78. Ann Arbor: Karoma. 1997 Media Lengua. In Contact languages: a wider perspective, Sarah G. Thomason (ed.), 365–426. Amsterdam: John Benjamins.

146 Suzanne Dikker 1999

Three processes of borrowing; borrowability revisited. In Bilingualism and migration, Guus Extra and Ludo Verhoeven (eds.), 229–246. Berlin /New York: Mouton de Gruyter. 2000 Bilingual speech: a typology of code-mixing. Cambridge: Cambridge University Press. Ocampo, Alicia M. and Carol A. Klee 1995 Spanish OV/VO word-order variation in Spanish-Quechua bilingual speakers. In Spanish in four continents: studies in language contact and bilingualism, Carmen Silva Corvalán (ed.), 71–82. Washington D.C.: Georgetown University Press. Pavón Lucero, María Victoria 1999 Clases de partículas: preposición, conjunción y adverbios. In Gramática descriptiva de la lengua española, Ignacio Bosque and Violeta Demonte (eds.), 565–656. Madrid: Espasa. Smith, Ian 1977 Sri Lanka Creole Portuguese phonology. Ph. D. diss., Cornell University, Ithaca, N.Y. Stolz, Christel and Thomas Stolz 1996 Funktionswortentlehnung in Mesoamerika. Spanisch-amerindischer Sprachkontakt. Sprachtypologie und Universalienforschung 49 (1): 86–123. Thomason, Sarah G. 2001 Language contact, an introduction. Edinburgh: Edinburgh University Press. Thomason, Sarah G. and Terrence Kaufman 1988 Language contact, creolization, and genetic linguistics. Berkeley: University of California Press. Tsunoda, Tasaku, Sumie Ueda and Yoshiaki Itoh 1995 Adpositions in Word-Order Typology. Linguistics 33: 741–761. van Coetsem, Frans 2000 A general and unified theory of the transmission process in language contact. Heidelberg: Winter. van Riemsdijk, Henk and Riny Huijbregts 2002 Location and locality. In Progress in grammar: articles at the 20th anniversary of the comparison of grammatical models group in Tilburg, Marc van Oostendorp and Elena Anagnostopoulou (eds.), 1– 23. Amsterdam: Meertens Instituut. Van Valin, Robert D. Jr., and Randy J. LaPolla 1997 Syntax; structure, meaning and function. Cambridge: Cambridge University Press. Weinreich, Uriel 1953 Language in contact: findings and problems. Mouton: New York.

Section II Austronesia

Reversing Hispanisation on Rapa Nui1 (Easter Island) Steven Roger Fischer

1. Introduction Rapanui is the language of the indigenous Southeast Polynesians of the island of Rapa Nui (Easter Island, Isla de Pascua, Ile de Pâques, Osterinsel, etc.) in the South Pacific, located at 27º S, 109º W, whose ancestors arrived around AD 600.2 These ancestors were speaking perhaps Mangareva’s dialect of the Eastern Polynesian subgroup of languages, whose ultimate origin lay in the great Austronesian family of languages (Fischer 1997: 359 and 2001a: 122, and Du Feu and Fischer 1993: 165). Representing Austronesian’s eastern frontier, the Rapanui language reveals no evidence of a “nonPolynesian substrate” (such as that alleged by Thor Heyerdahl, among others) nor a “second Polynesian wave of settlers” following initial settlement (Fischer 1992: 187). Evidently Easter Island’s population remained exceptionally isolated for the greater part of their history, until the Dutch navigator Roggeveen’s “discovery” of the island for Europe in 1722. Today, the Rapanui language is spoken by about 800 out of approximately 4,000 Easter Islanders who claim Polynesian descent (including around 2,200 exiles). There are now approximately 2,200 Chilean Continentals also living on Rapa Nui, most of whom speak only Chilean Spanish. The total island population – comprising Rapanui, Chilean Continentals, and a small number of resident Europeans and Americans – currently stands at around 4,000. 2. A politico-pedagogical proposal of linguistic “equilibrium” Several years ago, in an astute study treating of Spanish’s influence in Hispanoamerica and Austronesia, Klaus Zimmermann (2001) highlighted a new politico-pedagogical concept of cultural communication. This innovative proposal – targeting initially a dominant Spanish and its respective “subject” languages – would see responsible societies cultivating respect for indigenous tongues and cultures; creating alphabets for these languages;

150 Steven Roger Fischer teaching these in local schools; and modernizing the indigenous tongues by introducing new terminologies and discursive strategies in order to encourage daily active use. The entire process should, as Zimmermann conceived it, create a socio-political “equilibrium” of both donor language and indigenous tongues, one that would allow the latter not only to survive, but to grow.3 This proposal, I believe, relates immediately to the Rapanui language’s present situation. 3. Rapanui’s sociolinguistic profile Until recently, I had experienced only Spanish’s domination on Easter Island, which I visit each year. And so, like many others, I had feared that within one generation the indigenous Rapanui language of Easter Island could well become extinct (Fischer 2001b: 331). However, circumstances change swiftly on the island. Within Easter Island’s own unique linguistic laboratory I believe I am beginning to witness Zimmermann’s described dynamics at work. That is, at present a greater “equilibrium of mutual influences” appears indeed to be targeted on Easter Island, one that might even be currently reversing Hispanisation. How and why this is happening is the focus of this article. But first, a brief sociolinguistic profile of the Rapanui language is warranted, one that should allow the reader to better appreciate latter-day developments on this famed, distant island. Over many centuries of near-total solitude on Earth’s most isolated inhabited island (until the British settled the South Atlantic’s Tristan da Cunha in the nineteenth century), as many as 12,000 Easter Islanders were speaking Rapanui at any one given time. The language’s earliest stage, Proto-Rapanui, still awaits reconstruction. Old Rapanui is the name I have given to that stage of the language spoken by the last survivors of the island’s pre-European culture, several of whom lived up into the second decade of the twentieth century (Du Feu and Fischer 1993: 165, and Fischer 1997: 360). This language is still preserved in many documented chants, recorded and still sung songs and kaikai (cat’s cradle recitations), and a small number of written narratives from the nineteenth and early twentieth centuries. Old Rapanui also underlies Easter Island’s rongorongo script, Oceania’s only indigenous writing before the twentieth century (Fischer 1997). The label Modern Rapanui embraces that rapidly changing language that, already by the end of the nineteenth century, had been greatly influenced by English, French, Tuamotuan, Mangarevan, and, above all, Tahitian.

Reversing Hispanisation on Rapa Nui (Easter Island)

151

A product of complex contact processes, Modern Rapanui is half Modern Tahitian. Indeed, the Tahitian component is so strong in Modern Rapanui that there is no doubt that we are dealing here with language intertwining.4 There is no Rapanui code-switching and /or code-mixing with Tahitian. Nearly all Rapanui speakers regard the Tahitian used in Modern Rapanui to be “native”, as already their grandparents and even great-grandparents had used this particular intertwined language at the end of the nineteenth century. At least in part, such a perception stems from the fact that each of the world’s languages is also a vehicle for “group identification”, a psycholinguistic phenomenon that necessarily determines a language’s use and fate. Unlike the Maori of New Zealand, Hawaiians, Tahitians, Tongans, Samoans, and many other Polynesians, however, most Rapanui people perceive their group identification to lie more generally within the greater assemblage of Polynesian languages as a whole. That is to say, perhaps because of their great isolation, as well as because of their long history of oppression from Europeans, North Americans, and South Americans, the Rapanui people have come to comprehend a much larger ethnic periphery. In this way, most Easter Islanders of Polynesian pedigree sense the Tahitian, Tuamotuan, and Mangarevan elements in their language – though not these same languages as spoken today, it must be stressed – to be “native” rather than foreign. Yet the actual situation is even more complex. Some Rapanui speakers discern within the local language a “layering” of the Polynesian patrimony. Without knowing why, some Easter Islanders have voiced to me that they simply “feel” that their legendary founding father, Hotu Matu’a, for example, might not really be indigenous after all (he actually derives from Hatu Motua, a Mangarevan mythical figure introduced into Easter Island settlement narratives in the 1870s); that matahiti (Tahitian ‘year’) feels somehow more awkward than ta’u (Rapanui ‘year’); that mahana (Tahitian ‘day’) should perhaps really be ra’ (Rapanui ‘day’); or that vahine (Tahitian ‘woman’) is not as “comfortable” for some reason as v ’e (Rapanui ‘woman’). Untutored as they are in the formal processes of linguistic borrowing from closely related East Polynesian languages, most Rapanui are vague about their reasons for feeling this way. Yet all would contend nonetheless that these introduced lexemes in Modern Rapanui, uncomfortable though they feel to some, would still lie well within their “group identification”. In contrast, all non-Polynesian elements in Modern Rapanui, being highly intrusive, would comprise “alien identification”. Most egregiously, the past two generations of indigenous Easter Islanders in particular have introduced a substantial Spanish admixture to Modern Rapanui. Spanish had

152 Steven Roger Fischer begun to affect Rapanui only after Chile’s political annexation of the island in 1888 when the local population – ravaged by labour raids, then disease, then exodus – stood at only 178. Spanish’s impact long remained minimal. Well into the twentieth century, Easter Islanders regarded English to be as important to them as Spanish. This is because the few foreigners of status on the island, the managers and foremen of the Company sheepfarm, spoke mostly Scots English. Until 1953, these outsiders, who could also speak Spanish but rarely spoke Rapanui, ruled with an iron hand. During this time, immigrant Spanish-speaking ranchworkers, who usually numbered only about twelve at most in any given year, on the whole kept to themselves and to their Rapanui mistresses, with whom they communicated in Spanish. Spanish-language instruction at the one-room school at the island’s sole community of Hangaroa was haphazard until 1934, when stricter Spanishlanguage instruction commenced in a newly constructed schoolhouse. However, formal education on the island had only minimal success, and all Easter Islanders remained fluent Rapanui speakers though children in the Chilean government’s school were generally forbidden to speak Rapanui, under threat of corporal punishment. In this way, Chilean Spanish – that is, the dialects of Santiago, Valparaíso, and Viña del Mar – was for many decades only an infrequent second language on Easter Island. All indigenous Rapanui people saw Spanish as the language of the primary school; of the Chilean Naval administration that, as of 1953, was controlling the island in place of the Company; of the local Roman Catholic church, whose German priest also used liturgical Latin and some Rapanui in song and prayer; and of the very infrequent dignitaries, government and military officials, and scientists who called. Suddenly everything changed in the 1960s. Easter Island’s civil rights movement of 1964 led to Chilean citizenship in 1966 for the then over 1,300 Easter Islanders, nearly all of whom were still indigenous Polynesians. By 1967, Islanders had their own Spanish-speaking municipalidad or municipal form of governance (Fischer 2005: 209–218). Virtually overnight, the island’s linguistic situation also changed dramatically. “To be Chilean” was the maxim of the day, and in order to be Chilean most Easter Islanders started speaking Spanish in communal contexts. That is, Spanish became the public, Rapanui the private tongue. With the new local municipality, Spanish was the obligatory passport to any form of employment on the island and also, something which was perhaps even more important to such an isolated population at the end of the world, to escape from the island’s often crushing remoteness. In the following years, up to half the island’s population might be absent from the island at any given time, living

Reversing Hispanisation on Rapa Nui (Easter Island)

153

on the Continent and speaking – at secondary school, teachers’ training college, office, shop, or in military uniform – only Spanish. On Easter Island itself, the new municipal government created many clerical jobs which required fluent Spanish, not Rapanui. Chile ensured that Spanish remained the sole written language on the island as well. In this way the adopted use of writing in the new local municipality determined, and daily reinforced, Spanish’s supremacy. The upshot of this historical development was an abrupt language shift – towards Spanish and away from Rapanui – by many adults in Spanishspeaking situations who then also demanded Spanish from their children. Civil freedom on Easter Island was bought at the price of ethnic identity. Throughout the 1970s and 1980s, chiefly because parents were convinced that future success lay in Spanish’s mastery, children were growing up answering their parents only in Spanish, even if addressed in Rapanui. Yet, in 1975, Chilean law made Rapanui language instruction obligatory in the primary school and so, by 1979, Rapanui was being taught there as well: but as a separate subject, like a foreign language. In consequence of such developments, bilingualism quickly emerged, promoting a sociolinguistic hierarchy – what linguist Miki Makihara has termed “colonial diglossia” (Makihara 2004). Soon, indigenous Islanders were increasingly using Spanish in all situations, even when speaking among themselves, despite the fact that still only a small number of Chilean Continentals were residing on the island. An entire generation of Easter Islanders grew up in this way, as native Chilean Spanish speakers. This also produced a highly complex phenomenon of code-switching among Rapanui speakers, as Spanish was nevertheless still regarded as foreign and intrusive. Unlike with Tahitian, however, it is clear that Rapanui will never intertwine with Spanish. Despite (or perhaps specifically because of) more than a century of aggressive Chilean patronage, the criterion of linguistic identification still lies very much with the Polynesian patrimony. Ethnic allegiance is far more robust here than national allegiance. This predisposition generally appears worldwide to be particularly cogent in cases of foreign abuse and exploitation, whereby linguistic preference also signals a social symptom. By the 1990s, however, most indigenous Islanders were no longer displaying even a passive knowledge of Rapanui, and the language spoken by many of those older than 25 had become greatly Hispanised. Spanish use within Modern Rapanui always has comprised almost exclusively codeswitching.5 That is, fluent Rapanui speakers know Spanish is not their only available option; their Spanish use, when speaking Rapanui, is irregular, with great speaker variation; there is hardly any morpho-phonological inte-

154 Steven Roger Fischer gration when using Spanish; and the intrusion is dependent on, or triggered by, neighbouring elements from Spanish. Perhaps most strongly evident of code-switching on Easter Island: fluent Rapanui speakers make Spanish a conscious choice for stylistic and/or situational effect, and always maintain a clear mental distinction between their Polynesian inheritance and the “alien identification”. Complicating the mental picture of linguistic indigenousness on Rapa Nui, however, is the fact that, because of extreme isolation and a minimal number of speakers (around 800), Modern Rapanui remains an “idiolectal language”. That is, each speaker maintains his or her own personal language, displaying considerable variety. For want of a written tradition here one misses, in addition, a strong pedagogical basis for transmitting the received lexicon and grammar. Language resistance – in the case of Easter Island this means resistance to Spanish – then becomes in itself a defining component which helps to supply the missing tradition. The greater psycholinguistic complex involved in such a highly charged situation has been perceptively addressed by Zimmermann: Es ist eben nicht so, daß hier zwei Systeme aufeinandertreffen, sondern daß das konstruierende Gehirn mit dem input aus zwei Sprachen je nach Konstruktionslage, d.h. der kognitiv-konstruktiven Verarbeitung der Außenweltfaktoren und der bereits etablierten kognitiven Systeme unterschiedlich umgeht. Zu dieser Konstruktionslage gehören auch Attitüden der Sprachreinheit, der Sprachschönheit, der Gruppenidentifikation mit der eigenen Sprache usw. … Das Individuum, das die Attitüde der Sprachreinheit ausgebildet hat (die es eventuell in diskursiv vermittelten Konstruktionen nachvollzogen hat), geht kognitiv strategisch bei der Konstruktion und Aneignung anders mit L1 [Language 1] und L2 [Language 2] um, als ein Individuum, das diese Attitüde nicht ausgebildet hat. Ja, es kann sogar, wie wir dies aus den puristischen Bewegungen kennen, bereits vollzogene Einflüsse wieder bewußt eliminieren und als Ersatz dafür autochthone Neologismen bilden. [It is not precisely a case of two systems meeting here, but of the construing brain differentially handling the input from two languages each according to componential circumstances, that is to say, according to the cognitive-constructive processing of outside factors and to the cognitive systems already established. To these componential circumstances belong as well attitudes of linguistic purity, beauty of language, group identification with one’s own tongue etc. … The individual who has formed an attitude of linguistic purity (duplicated in time in discursively conveyed constructions) cognitively manages in cases of construction and adoption in a strategically different fashion L1 and L2 than an individual who has not formed this attitude. Indeed, he can

Reversing Hispanisation on Rapa Nui (Easter Island)

155

even, as we know from puristic movements, consciously eliminate already effected influences once again and, as replacement for these, form autochthonous neologisms.] 6

The Modern Rapanui I have summarised above is the language for whose relationship with Spanish I provided a discriptive summary – in contrast to this sociolinguistic summary – at the first Bremen conference in 1999.7 At that time, in view of a perceived threat to Rapanui I also predicted that Easter Island’s original language might well become extinct within one generation. I feared this because the indigenous Easter Islanders, rather than speaking Spanish with resident outsiders and Rapanui among themselves, as would be the typical situation world-wide, had by the end of the twentieth century elaborated syncretic Spanish with many Rapanui elements as well as syncretic Rapanui with many Spanish elements, and were even mixing these two hybrids together, creating a highly complex linguacape (Fischer 2001b; Makihara 1999, 2001a,b, and 2004). Church bells were tolling over Rapanui’s open grave. 4. An apparent reversal of Hispanisation Now, in 2005, six years after that first Bremen conference, over half of Easter Island’s semi-permanent population of approximately 4,000 are immigrant Chilean Continentals. Everyone on the island is Spanish-speaking, with nearly everyone also claiming Spanish as their first language. Nearly eighty per cent of the 600 primary-school children – one-third of whom are Chilean Continentals – are monolingual Spanish speakers. As mentioned, about 800 of the current population of approximately 4,000 Easter Islanders (residents and exiles) of at least partial Polynesian ancestry speak Rapanui, and this with mixed competence. Nearly all fluent speakers of Rapanui, perhaps 500 in total, are over 35 years of age (Makihara 2005b). Despite such statistics which, anywhere else, might be viewed as alarming, events of the past few years are actually challenging those earlier predictions of Rapanui’s imminent demise. Indigenous Easter Islanders’ vigorous campaign for political autonomy has reverberated throughout the social web. That lesser half of the island’s population – those who were born there of mixed Polynesian and foreign ancestry (there have been no “pure” Polynesians on the island for some decades now) – are increasingly turning to Rapanui as the emblem of ethnic choice. “To be Chilean” is no longer important. “To be Rapanui” is now the maxim. And this apparently involves, before everything else, the Rapanui language.

156 Steven Roger Fischer The conversion has not occurred overnight. As we have already seen, since 1979 a succession of educational programmes has promoted the Rapanui language’s acquisition and use among schoolchildren in particular, even if presented as a “foreign tongue”. These programmes have had mixed results; at the least they have helped to maintain many children’s passive understanding of the language. Evening classes for adults were also offered. As of the mid-1980s Rapanui began to be spoken again in communal situations. Once again, Rapanui was a legitimate language choice in the public arena, contending with Spanish and, over time, eventually distilling those two new types of island speech: Spanish Rapanui and Rapanui Spanish (Makihara 1999). In 1990 a series of graded schoolbooks in Rapanui were published for those children who already spoke fluent Rapanui; however, this project failed when few children could read them and insufficient time was given to them in the classroom. (A common complaint at the time was that owning the schoolbooks and simply displaying them for prestige were seemingly more important to Rapanui families than reading them.) Once Chile’s Pinochet regime ended, also in 1990, a new political activism – one modelled after that of the Tahitians, Hawaiians, and Maori of New Zealand – inspired younger Easter Islanders, in particular, to loudly voice specific political grievances. Since then, with increasing vehemence, most indigenous Easter Islanders (but not Chilean Continentals, who now comprise the island’s resident majority) have been dynamically calling for a new body politic on the island, as these “ethnic Rapanui” (many of considerable foreign ancestry) have undertaken to redefine their status within the Chilean nation-state. Almost immediately after Pinochet’s removal from office, in fact, a strong Easter Island cultural revival, paired with demands for autonomy or even full independence from Chile, also commenced which highlighted nearly all indigenous Islanders’ perceived loss of their cultural and linguistic heritage. An expansive cultural centre even opened on the island in 1993 to promote the Rapanui language, offer classes, hold seminars, and stage cultural events. Yet nearly each “top-down approach” towards language reinvigoration has failed here, as each was resisted by individualistic Easter Islanders who commonly disdain authority of any kind, an historic trait based on decades of foreign abuse and exploitation. It is almost exclusively the interpersonal, private, and “bottom-up” approaches to reinvigorate the Rapanui language that are now showing first signs of success. That is, indigenous Easter Islanders (island residents who perceive themselves, for whatever reason, to be ethnically more “Rapanui” than “Chilean”) are now reclaiming, and in accelerating fashion, language domains – both

Reversing Hispanisation on Rapa Nui (Easter Island)

157

public and private – that, for the past 40 years, have mainly been the prerogative of Spanish.

5. Modes of politico-pedagogical innovation Easter Island’s new pedagogical foundation has been laid on earth that is well prepared for it. Although the island television station, which began with recorded Spanish broadcasting in 1976, still transmits all mainland programmes in Spanish, there are finally limited hours of local Rapanuilanguage broadcasting as well: announcing community news, discussing topics of immediate interest, conducting interviews, and playing Polynesian music often with lyrics sung in Rapanui by local performing groups. Competing with the Chilean Navy’s Spanish-language FM channel, Hangaroa’s “Radio Manukena” now broadcasts in Rapanui in both AM and FM. Within the past six years an increasing number of independent publications – poems, short stories, even political tracts and literary competitions – have begun making Rapanui a communal literary language. New descriptive Rapanui dictionaries, grammars, textbooks, and complementary literacy materials testify to a greater professional engagement. Despite such threshold commitments, however, a “fully literate Polynesian community” such as those in Tahiti, Tonga, and Samoa – that is, one that daily enjoys and depends on a written Polynesian tongue – has yet to develop on Easter Island. Nonetheless, this may happen soon, because the number of children displaying at least a passive competence in spoken and written Rapanui is now on the rise (Makihara 2005b). Such enthusiastic sowing produces a prompt harvest. At Hangaroa’s central market and three small supermarkets one now hears Rapanui being spoken publicly again, even by young people. Towards children, adults will now sometimes speak in syncretic Rapanui, no longer solely in Spanish; the children will understand, too, though their reply will commonly still be in Spanish. Perhaps more encouragingly, young people are using Rapanui among themselves more frequently. All public events now feature, alongside Spanish, a “perceived puristic” Rapanui: a Rapanui-Tahitian hybrid devoid of Spanish lexemes and grammatical borrowings. Easter Island’s annual “T pati Rapa Nui” festival, the year’s major cultural event, not only takes place in Rapanui and Spanish (and often English, too, in recent years, for the foreign tourists) on its large, painted, laser-illuminated stage, but also includes competition in Rapanui song, enacted legends featuring characters speaking a perceived “archaic” Rapanui, and kaikai or cat’s-cradle

158 Steven Roger Fischer recitations in Old Rapanui. Many of these “T pati” events are for schoolage children, who appreciate that the old island tongue of their ancestors can win them valuable prizes and bring prestige. Several Easter Island performance groups who celebrate Rapanui song and dance in costume figure among Polynesia’s best; their CDs are now heard everywhere on the island, day and night. Such activities have ostensibly led to the next stage of cultural evolution on the island, one according well with the “Zimmermann Doctrine” in that the new cultural communication, hitherto unimaginable on Easter Island, suggests the process might be able to achieve soon an advantageous equilibrium between Spanish and Rapanui. In 2000, several Easter Island teachers, mostly women, led by María Virginia (Viki) Haoa Cardinali, President of the Rapa Nui Language Academy, inaugurated a language-immersion educational programme – one similar to those of New Zealand, Hawai’i, and Tahiti. By 2005, it has included 134 children from families who ethnically identify first with their Polynesian ancestry. Currently involving the first to fourth grades, with each new year a grade is added to the programme; it will eventually culminate in high-school graduation. By 2013, therefore, a complete class should be graduating from the Easter Island High School having had all their instruction entirely in the Rapanui language, apart from compulsory Spanish and foreign language lessons according to prescribed Chilean curriculum. In a related programme, Marta Hotus, working together with Hangaroa’s Women’s Society, beginning in 2005 has inaugurated an annual pta ’uta’u or poetry contest promoting Rapanui literacy and pride in the local language; winning contributions are subsequently displayed throughout the community in prominent locations, for all to read. Further cultural manifestations are becoming evident, similarly testifying to a newly awakened pride in the Rapanui language that might target (if only intuitively at present) that “equilibrium” between Spanish and Rapanui. Since 2003, the Santiago and Pape’ete (Tahiti) flights to and from Easter Island on LAN-Chile, the only carrier allowed to land there, include on-board welcome and instructions by flight attendants also in Rapanui, besides the internationally required Spanish and English for these legs; the cabin crew are mostly indigenous Easter Islanders, the cockpit crew Chilean Continentals. At the time of writing, an “Easter Island Cultural and Linguistic Heritage Project” is also being planned, with the intention of establishing a digital archive, in the Rapanui language, of oral and video histories (Makihara 2005a). These modes of politico-pedagogical and further cultural innovation signal a dynamic change in Rapa Nui’s use of language. Of late a distinct

Reversing Hispanisation on Rapa Nui (Easter Island)

159

attitude shift has occurred once more, one that has accompanied a movement towards political autonomy, separate from but allied with Chile, and a renewed pride in the Easter Island patrimony. It is a positive reinvention of what it means to be an Easter Islander. And the process places language at the focal point of a new island identity, whose precise details are still being negotiated. Such a process is converting a language that had been in danger of extinction into one that is perceived to be indispensable, at least by the local minority community identifying primarily with an ever-diluting Polynesian genetic component. Noteworthy in this regard is what Zimmermann has written of analogous phenomena elsewhere: Partiendo de una teoría de funciones y ámbitos se puede graduar la vitalidad según el número y la importancia social de las funciones y los ámbitos en los que una comunidad hace uso de una lengua. Ya que la vitalidad depende de las acciones de los hablantes, sobre todo los de forma concertada, es decir en forma de política, hablar sobre la vitalidad siempre significa hablar sobre la creación de un medio ambiente favorable para el uso y la creación de posibilidades para su uso en todos los ámbitos y para todas las funciones. De no ser posible la realización total de estos postulados se tiene que intentar la realización parcial. Política lingüística y planificación lingüística en el contexto de la ecología lingüística son entonces un conjunto de medidas para evitar que las lenguas pierdan su vitalidad, y en el caso de haberla perdido, de convertir las lenguas en peligro de extinción en lenguas vitales. [Proceeding from a theory of functions and domains one can grade the vitality according to the number and the social importance of the functions and the domains in which a community makes use of a language. As the vitality depends on the actions of the speakers, above all those of concerted form, that is to say in form of politics, speaking about vitality signifies speaking about the creation of a milieu favourable to the use and the creation of possibilities for its use in all the domains and for all the functions. As the total realisation of these demands is not possible, one must aim at partial realisation. Linguistic politics and linguistic planning in the context of linguistic ecology are, then, a combination of measures to avoid languages losing their vitality and, in the case of having lost it, of converting languages in danger of extinction into robust languages.] 8

6. Modern Rapanui in light of the “Zimmermann Doctrine” What is the effect of all this? Zimmermann’s postulated dynamics, as described above, do appear to be operative – fortuitously, as those who champion endangered languages would certainly claim – on Easter Island, where

160 Steven Roger Fischer a perceived “equilibrium of mutual influences” (or at least an intuitively targeted equilibrium) seems presently to be reversing Hispanisation. At this point in time no fewer than five resident languages engage Polynesianoriented Islanders here: Old Rapanui, in public performance only (this language is largely unintelligible); Modern Rapanui, in both public and private domains; Spanish Rapanui, the local syncretism which uses much codeswitching and also relies on a greater number of Spanish borrowings than Modern Rapanui; Rapanui Spanish, the local syncretic compromise which is also used as an ethnic emblem; and Chilean Spanish, the language of the still dominant power and now prevalent resident population. Thus, the active language continuum here in 2005 would be: Modern Rapanui > Spanish Rapanui > Rapanui Spanish > Spanish. The first and second of these, Modern Rapanui and Spanish Rapanui, are not distinctly delineated languages per se, but robust speech varieties that interact with one another daily. Modern Rapanui, that blanket designation for the local Rapanui-Tahitian intertwining that is continuously changing, is at the very least not moribund, but apparently enjoying an evident, though still small and tentative, renaissance. With the presentation of evidence for a possible reversal of Hispanisation on Easter Island, however, one has recourse chiefly to sociolinguistic generalisations, not to descriptive specifications. This is because Spanish, when used within Modern Rapanui, is wielded as a tool: normally, but certainly not exclusively, for code-switching. Of course Modern Rapanui has also borrowed a considerable amount of Spanish (Fischer 2001b); but borrowings are limited mainly to introduced technology as well as to hitherto unknown concepts and perceptions deriving from non-Polynesian spheres of activity, with a few exceptions. (For example, some grammatical borrowing from Spanish figures in Rapanui as well.)9 Owing to the extremely small number of Rapanui speakers, there has been no attempt as yet to translate Spanish borrowings systematically and/or formally into some kind of standardised, “received” Rapanui; this might conceivably be achieved through expanding the semantic domain of the Old Rapanui lexicon. But under present circumstances there would be little advantage in doing this. All evidence for a reversal of Hispanisation will therefore have to remain quantitative: that is, either a Spanish word or phrase will be used or will not be used. In this event, a reversal of Hispanisation will simply be the demonstration of fewer Spanish intrusions and the gradual return to a Polynesian exclusivity. For this there are definite sociolinguistic indicators, which evidence several modes of active reversal. Fluent speakers of Modern Rapanui are, on

Reversing Hispanisation on Rapa Nui (Easter Island)

161

the whole, conscious of an unmistakable diminishing of code-switching particularly in more formal – that is, more “Rapanui” or locally ethnic – contexts. More children are now being formally integrated into indigenous language use. The hitherto private domain of Rapanui use is becoming once again increasingly public. New language projects are raising local awareness of the island’s original language, and promoting more active use in the larger community in general. And the island’s imminent political autonomy will certainly open a new domain of Rapanui use, which must increase not only its status, but also its application as both a spoken and written language. However, is this a “true” reversal of Hispanisation on Easter Island? That is, will this quantitative evidence for fewer instances of Spanish codeswitching – for this is what we are talking about for the most part – continue for some time to come, perhaps even accelerate? Only time will tell. The Hispanisation process that overwhelmed the island in the 1960s, an inundation that almost extinguished Rapanui’s flame, has decidedly receded, that much is clear. No professional observer would deny that Spanish is no longer as “aggressive” on Easter Island as it had been only one generation ago. A new Rapanui pride is heralding a new Rapanui identity, its exact features by and large still undefined. For the first time on the island, a small number of resident Continentals – the so-called “contis” who now make up over half the island’s population – are also learning Rapanui; several contis are even performing in singing and dancing groups, in their own way working towards that social “equilibrium” that Zimmermann proposed. In perceived value within the solitary community of Hangaroa, the Rapanui language is certainly no longer “below” Spanish. However, Spanish remains the island’s literate language, a fact that hinders the reversal process. Very few Easter Islanders yet regard Modern Rapanui as a written language, only an oral language, an entrenched perception that must be changed before Modern Rapanui achieves “true” equilibrium with Spanish. In addition, Chilean Spanish still offers things Rapanui cannot: higher education, a professional career, a passport to the Continent. And English, which is spoken with increasing frequency on the island, offers tourist dollars, foreign contacts, a passport to the world at large. Both currently compete with learning fluent Modern Rapanui. Linguists can perhaps help here. At present, several foreign linguists on the island are actively engaged in assisting the Rapanui people to understand, revive, and conserve their original language. Purists might deem this “interference”; others will of course assert that endangered languages cannot be rescued, only described.10 However, it is incumbent on all who pro-

162 Steven Roger Fischer fess a sincere attachment to a people to assist them in every possible way. For a linguist, this means applying one’s training and expertise to instruct, revive, and conserve. Anything less would be a betrayal – of one’s profession and one’s friendship. While this might not be regarded as scientific by many, it is eminently humanitarian. If applied effectively on Easter Island, such “interference” might hasten the reversal of Hispanisation. Descriptive duty might better share here, in other words, with prescriptive compassion. 7. Conclusion One might put forward that two fundamental lessons have been learnt in this Easter Island test case. First, the “Zimmermann Doctrine” of cultural communication achieving a socio-political “equilibrium” of both dominant and subject languages is indeed worthy of greater scientific scrutiny, even at the far reaches of the South Pacific. Second, ethnic perceptions are nothing if not robust. As we have seen, the indigenous Easter Islanders’ newly awakened pride, together with a perceived need for political autonomy, might indeed have rescued the Modern Rapanui language from extinction. By retarding or even reversing the aggressive Hispanisation which shaped and characterised their small island society from the 1960s up into the 1990s, the Rapanui people might now be able to allow their language, in its various syncretic yet increasingly puristic forms, the possibility of survival, perhaps even growth. But for how long might Rapanui survive and grow, one might ask? – Apparently so long as a perceived Polynesian identity can offer something of genuine personal value to those fortunate enough to live on this fascinating island at the end of the world. Notes 1.

2. 3.

Since the early 1990s it has been standard practice in Easter Island scholarship to write the island’s name as Rapa Nui in Chilean fashion, but the island’s indigenous people and their language as Rapanui in pan-Polynesian fashion. This custom is observed in the present article. For the postulated date and circumstances of settlement, see Fischer (2005: 17–20). Zimmermann (2001: 31–32): “Todo esto significa un cambio radical de las estructuras socio-políticas y culturales así como sociolingüísticas, que deberían a su vez tener un efecto también en los procesos de contacto futuro fomentando un mayor equilibrio sociopolítico de las lenguas. Éste, a su vez, podría contribuir a una mayor inmunidad de las lenguas indígenas evitando una parte de la

Reversing Hispanisation on Rapa Nui (Easter Island)

163

gran cantidad de influencias. También puede imaginarse el desarrollo de una actitud purista y con ella una acción eliminadora de influencias existentes, produciendo así un mayor equilibrio de influencias mutuas.” [All this signifies a radical change of the socio-political and cultural, as well as sociolinguistic, structures which, as far as they are concerned, should have an effect also on the future contact processes, encouraging a greater socio-political equilibrium of the languages. This, for its part, could contribute to a greater immunity of the indigenous languages, avoiding a part of the great quantity of influences. Also, one can imagine the development of a purist attitude and, with it, a purging operation of existent influences, producing as well a greater equilibrium of mutual influences.] 4. For details about intertwining, see Zimmermann forthcoming. 5. For the distinction between code-switching and borrowing, see Sakel and Matras forthcoming. 6. Zimmermann (2005: 3–24). 7. Fischer (2001b). 8. Zimmermann (2004: 97–98). 9. See Fischer (2001b: 322–326) and Makihara (2001a: 204–212). 10. Dixon (1997: 112): “Something can be done. Writing can be put to good use in producing primers for use in local schools (although children will never speak a local language from just having it taught to them in school, if it is not also used in the home). And writing can also be used to record traditional stories and songs. The language can be documented, before it vanishes. The loss of a language can be slowed a little; but it cannot be halted or reversed.”

References Dixon, R. M. W. 1997 The rise and fall of languages. Cambridge: Cambridge University Press. Du Feu, Veronica and Steven Roger Fischer 1993 The Rapanui language. In Easter Island Studies: Contributions to the history of Rapanui in memory of William T. Mulloy, Steven Roger Fischer (ed.), 165–168. (Oxbow Monograph 32.) Oxford: Oxbow Books. Fischer, Steven Roger 1992 Homogeneity in Old Rapanui. Oceanic Linguistics 31: 181–190. 1997 Rongorongo: The Easter Island Script. History, traditions, texts. (Oxford Studies in Anthropological Linguistics 14.) Oxford /New York: Clarendon Press. 2001a Mangarevan doublets: Preliminary evidence for Proto-Southeastern Polynesian. Oceanic Linguistics 40: 112–124.

164 Steven Roger Fischer 2001b

Hispanicization in the Rapanui language of Easter Island. In Lo Propio y lo Ajeno en las Lenguas Austronésicas y Amerindias, Klaus Zimmermann and Thomas Stolz (eds.), 313–332. (Lengua y Sociedad en el Mundo Hispánico 8.) Frankfurt: Vervuert / Madrid: Iberoamericana. 2005 Island at the end of the world: The turbulent history of Easter Island. London: Reaktion Books. Makihara, Miki 1999 Bilingualism, social change, and the politics of ethnicity on Rapanui (Easter Island), Chile. Ph.D. diss., Department of Linguistics, Yale University. 2001a Modern Rapanui adaptation of Spanish elements. Oceanic Linguistics 40: 191–222. 2001b Rapanui-Spanish bilingualism. Rongorongo Studies 11: 27–42. 2004 Linguistic syncretism and language ideologies: Transforming sociolinguistic hierarchy on Rapa Nui (Easter Island). American Anthropologist 106: 529–40. 2005a Easter Island cultural and linguistic heritage project. Paper presented at the VI International Conference on Easter Island and the Pacific, Reñaca, Chile, August 2005. 2005b Rapa Nui ways of speaking Spanish: language shift and sozialization on Easter Island. Language in Society 34 (5): 727–762. Sakel, Jeanette and Yaron Matras forthc. Modelling contact-induced change in grammar. In Aspects of language contact. New theoretical, methodological and empirical findings with special focus on Romancisation processes, Thomas Stolz, Dik Bakker, and Rosa Salas Palomo (eds.), Berlin /New York: Mouton de Gruyter. Zimmermann, Klaus 2001 Interculturalidad y contacto de lenguas: Condiciones de la influencia mutua de las lenguas amerindias con el español. In Lo Propio y lo Ajeno en las Lenguas Austronésicas y Amerindias, Klaus Zimmermann and Thomas Stolz (eds.), 17–34. (Lengua y Sociedad en el Mundo Hispánico 8.) Frankfurt: Vervuert, Madrid: Iberoamericana. 2004 Ecología lingüística y planificación lingüística. In Lenguas Vivas en América Latina, IV Jornadas Internacionales sobre Indigenismo Americano (Universidad Autónoma de Madrid), Ariadna Lluis í Vidal-Folch and Azucena Palacios (eds.), 93–109. Madrid: Universidad Autónoma de Madrid. 2005 Interferenz, Transferenz und Sprachmischung: Prolegomena zu einer konstruktivistischen Theorie des Sprachkontaktes. In Das Zimbrische zwischen Germanisch und Romanisch, Ermenegildo Bidese, James R. Dow, and Thomas Stolz (eds.). Bochum: Brockmeyer.

Reversing Hispanisation on Rapa Nui (Easter Island) forthc.

165

Constructivist theory of language contact and the Romancisation of indigenous languages. In Aspects of language contact. New theoretical, methodological and empirical findings with special focus on Romancisation processes, Thomas Stolz, Dik Bakker, and Rosa Salas Palomo (eds.), Berlin /New York: Mouton de Gruyter.

The old, the new and the in-between: Comparative aspects of Hispanisation on the Marianas and Easter Island (Rapa Nui) Steve Pagel

1. Introduction This paper addresses two cases of language contact involving Spanish in the Pacific, both of which exhibit comparable traces of Hispanisation in and around the autochthonous languages Chamoru and Rapanui.1 While intensive linguistic contact, particularly with Spanish, has led Chamoru to enter a new language stage in the 18th century, the present interfering of Spanish in Rapanui is a smooth continuation of the thorough modifications this language has experienced since the end of the 19th century. Additionally, the linguistic scenarios in both areas can be said to ‘expand’ at present due to the emergence of a range of mixed varieties which accomplish an important symbolic function in the revitalization of the native identity. A rather unusual context for the research on Hispanisation, Oceania in fact hides some of the more intimate secrets of Romance studies: “The rape of Oceania began with Guam”, states Oliver (1951: 234), meaning that it commenced under none other than the Spanish flag. The rape of Oceania might end, one could say with respect to recent developments, on Easter Island, and is this time carried out under the Chilean flag. The two places referred to in this metaphor, Guam in the Mariana archipelago of Micronesia and Easter Island (or Rapa Nui) on the eastern edge of Polynesia, mark the beginning and the present end of a Hispanic era in Oceania, and they set the thematic framework of the present paper. The aim here is to contribute to the investigation of the Spanish legacy in the autochthonous cultures and languages of the Pacific region (which in a broader sense would include the Americas), a topic which so far has been neither studied methodically nor, in most cases, from a socio-historical critical perspective2. Although separated by time and space, the contact scenarios on the Marianas and Easter Island share not only a similar socio-historical background, but also particular features of linguistic Hispanisation. In both cases, the autochthonous language (Chamoru from the West-Malayo-Polynesian branch, and Rapanui

168 Steve Pagel from the East-Polynesian branch of the Austronesian family) has been subject to intensive language change triggered by contact with Spanish. The following paragraphs will deal with the major aspects of these two dimensions of Hispanisation (and additionally Americanization on Guam), the socio-historical (§2) and the linguistic (§3, 4, 5), before concluding in the last chapter (§6). 2. Socio-historical aspects of Hispanisation Lamentably a warrior is the title of a poem written by the Guamanian author Peter Onedera, probably one of the most prominent voices of current Chamoru revitalization on Guam. It tells a short story of the colonization of the island, emphasizing the – ever persistent – tension between what is considered autochthonous and what is considered foreign. The poem ends with the following, somehow belligerent lines: We may be uncivilized and ugly, pray tell / we proved him and so he set sail / He called us thieves in the night / but we did put up a fight / In the P.I., they sent him to hell! // (Onedera 1994: 12)

The unspecified ‘he’ and its respective forms probably refer to Magellan, who in March 1521 sighted Guam and neighbouring Rota, the first stretches of inhabited land discovered in a largely unknown ocean. When he approached the shores of Guam, so tells us Magellan’s chronicler Pigafetta, his ships were plundered by natives, who took away, among other things, a small boat of essential importance to the Europeans. In the attempt to get it back, and as an act of revenge, Magellan’s men killed seven of the ‘indios’ and destroyed numerous huts and boats on the shore (Pigafetta 1969: 60; Rogers 1995: 7–8). Based on his impression of the event, the captaingeneral named the newly discovered archipelago Islands of the Thieves. It is this first encounter between Europeans and a people of the Pacific which is taken up in the verse above, an unequal encounter in Onedera’s opinion, balanced only by Magellan’s death, which indeed occurred in the Philippine Islands (P.I.) soon after. However, there may be more to this encounter than the immediate impression it suggests, in a sense it “set a precedent that became a tragic pattern in later Spanish-Chamorro relations. The great cultural differences in the values of the two sides led repeatedly to hostility and armed conflicts” (Rogers 1995: 9). Thus, the stage was set when in 1668 Jesuit Father San Vitores arrived to install the first mission on Guam, thereby establishing more permanent

Comparative aspects of Hispanisation

169

contact with the Spanish crown which had already claimed the islands in 1565. Of the approximately 50,000 Chamoru populating the archipelago at this time less than 4,000 survived the next forty years of repeated epidemics of introduced diseases and atrocious warfare (Rogers 1995: 70–71, Mühlhäusler 1996: 105–108). Those who did survive were resettled to the southernmost islands Guam and Rota, where in the early 18th century a new Chamoru elite began to form, one that developed out of intermarriage with, above all, Spaniards, Mexicans and Filipinos. In fact, within a few generations, the remains of the Chamoru ethnic group were entirely absorbed by this mestizo group, appropriately called ‘Neo-Chamoru’ by Rogers (1995: 104). In 1898 Spain lost its Micronesian possessions as a consequence of the Spanish-American War, and subsequent colonizers were the U.S. on Guam (and in the Philippines) and Germany in the northern Marianas. Spanish on Guam was rapidly substituted by English in administration, business and tuition, and only in the private sector did Chamoru continue to be spoken (Rogers 1995: 134). In the course of the two World Wars, first the northern Marianas and later Guam fell to Japan, but both were (re)conquered by U.S. troops in 1944. Despite unification under U.S. sovereignty they have continued to form different political entities to the present day3. Due to pro-English language policies and other factors such as massive immigration and urbanization, in recent decades Chamoru has become a minor language, being spoken by every 5th person of the approximated 220,000 inhabitants of the Marianas today (Census 2000). There are, however, a number of voices promoting revitalization of Chamoru cultural and linguistic values, though it remains to be seen what their impact will be. Contact history on Rapa Nui began on Easter Sunday 1722 when Dutch captain Roggeveen reached this remote place and, in honour of the day, named it Paasch Eyland, ‘Easter Island’. A Spanish expedition under the command of Felipe Gonzalez set foot on the island in 1770, claiming it for the Spanish crown, although Spain never made an effort to redeem that claim. It was not until the mid 19th century that Easter Island moved closer into the focus of the colonial powers. Peruvian ships passed the island between 1862 and 1863 to recruit labour for guano plantations, kidnapping more than 1,400 of the estimated 4,100 Rapanui (Fischer 2005: 89, Maude 1981: 192), most of whom never returned. It is supposed that essential parts of the original Rapanui culture have vanished as a consequence of these slave raids (Fischer 2005: 63–64, 1997: 355–362, Makihara 1999: 169–172). The population crash further continued with the outbreak of smallpox imported by the few returnees, as well as the subsequent migration of islanders to Tahiti and Mangareva. At an all-time low, in 1877 only 110 Rapanui re-

170 Steve Pagel mained on the island. Although Chile’s annexation of Rapanui in 1888 put an end to this decline, it implied for those remaining the loss of their native land and confinement to the village of Hanga Roa. Permission was needed to leave the village since the entire island had been turned into a sheep farm by the Chilean-Scottish company Williamson & Balfour, to which the island was handed over. It was only in 1966 that the Rapanui were granted full citizenship rights, although they have faced different problems since then, such as massive immigration from the continent and an increasing hegemony of Spanish over Rapanui in the administrative, the tuitional, and since the 1990s in the public sector as well (Weber and Thiesen de Weber 1990, 1998). Today half of the approximately 3,700 inhabitants of the island consider themselves Rapanui, of which about two thirds are bilingual Spanish-Rapanui speakers, the remaining being monolingual speakers of Spanish (Makihara 2004: 4, 2005: 728). Recently, however, a reversion of Hispanisation as a result of the reinforcement of native identity has come into discussion (Makihara 1999: 160–163, 2005: 755, Fischer 2001: 316, this volume), paralleled by similar anti-(post)colonial movements elsewhere in the Pacific. To sum up, the contact with Spanish on the Marianas and Easter Island, although taking place neither simultaneously nor being initiated by the same political entity, has favoured a number of analogous processes on the sociohistorical sector of both areas. Among them are 1) a fatal demographic crash of the autochthonous population, convoyed by 2) the breakdown and partial loss of traditional structures (e.g. in politics, religion, settlement and language), 3) mesticization with the superstrate culture, but 4) continued sensation of alterity, 5) outnumbering by migration, and 6) a recent reassertion of native identity.

3. Linguistic aspects of Hispanisation on Easter Island A question arising from chapter 2 may be put as follows: If the traditional structures of the Chamoru and Rapanui have been dissolved as a consequence of demographic breakdown and continuous cultural and linguistic contact, on what basis can a renaissance of traditional values be promoted, and on what basis can a native identity be reasserted? In the following paragraphs a possible answer is to be approached from a linguistic perspective, the focus being on Hispanisation processes particularly with respect to the morphosyntactic level of the autochthonous languages and mixed varieties that have emerged through contact.

Comparative aspects of Hispanisation

171

A fascinating statement concerning the status quo of Rapanui language on Easter Island is given by an informant of Makihara: (1)

Rapanui (Makihara 1999: 169) e ai ro  te vananga Rapanui e noho mai ena, ‘ina e ttou he usa ‘the Rapanui language exists, we just don’t use it’

The statement can be interpreted in at least two ways, both of which seem equally plausible and true. First, there is a Rapanui language that is in danger of becoming lost because there is no generation of speakers to which it could be passed on. Second, there is an older form of the Rapanui language which is considered to be ‘original’, surviving between the lines of an otherwise altered or modernized code, probably identical with the code used in (1). The two approaches display what Thomason and Kaufman (1988) identified as the main routes of contact-induced language change: language shift (with interference from the substrate language), and language maintenance (with borrowing from the superstrate language). Both routes have played an important role in the formation of the synchronic linguistic scenario of Easter Island, which, in fact, does not seem easy to grasp: Fischer (2001: 237), e.g., states that “[a]t the beginning of the twenty-first century there is not “one” Rapanui language but a wide variety of idiolects that range between Rapanui and Spanish”. Correspondingly, Makihara finds that “[t]he bilingual situation now existing on Rapanui can be described as a continuum spanning Rapanui and Spanish speech varieties” (1999: 165). This continuum is structured by Makihara along the prototypes of four sets of varieties, whose main features shall be briefly reviewed in the following paragraphs: Modern Rapanui

Old Rapanui

Chilean Spanish

Rapanui Spanish

Figure A. The bilingual continuum in the contemporary Rapanui speech community (adopted from Makihara 1999: 167)

The term Old Rapanui comprises a diverse set of diachronic, diastratic and diaphasic varieties that were spoken on Easter Island prior to the 1860s, i.e., before the above mentioned demographic collapse occurred (Makihara

172 Steve Pagel 1999: 169, Fischer 1997: 359–360). It is assumed that “most speakers of prestigious speech style(s) did not survive the slave raids and the subsequent population crash” (Makihara 1999: 169). Accordingly, the transitory migration of many survivors to Tahiti and Mangareva resulted in a complex restructuring of the Rapanui language under the influence of other East Polynesian languages, with further influx coming from the missionary and mariner languages French and English. The outcome of this restructuring is Modern Rapanui, the name being “a blanket designation for the everchanging late nineteenth- and twentieth century stages of the language” (Fischer 1997: 360). Old Rapanui survives as a metaphor in Modern Rapanui, to recall the statement in (1), and has a physical presence only in early word lists and traditional verbal arts (e.g. in the recitation of traditional poems, called patu’atu’a; see Barthel 1960, Fischer 1994). Such being the case, it is technically not a part of the synchronic, but of the diachronic linguistic scenario of Easter Island, that’s why the bracket in figure A is a broken line. Consequently, Old Rapanui is not explicitly mentioned in Makihara (2005), instead the range of the continuum is set along these four benchmarks now: syncretic (or Modern) Rapanui4, R1S2 Rapanui Spanish, R2S1 Rapanui Spanish, and Chilean Spanish. See figure B, which is based on Makihara (2005: 733): Modern Rapanui  R1S2 Rapanui Spanish  R2S1 Rapanui Spanish  Chilean Spanish Figure B. Synchronic bilingual continuum in the Rapanui speech community

Placed at the right end of the continuum, Chilean Spanish comprehends “a set of varieties of Spanish spoken in continental Chile, particularly in the Santiago-Valparaiso-Viña del Mar area” (Makihara 1999: 172). Chilean Spanish can be understood as being standardized within a pluricentric conception of Spanish (see Clyne 1992), its features, which for the sake of brevity cannot be repeated here in detail (see, e.g., Oroz 1966, Rabanales 1992, Ludwig 2002), comprise the aspiration or loss of implosive s (loh doh instead of Standard SP los dos) and loss of intervocalic d (too instead of Standard SP todo) as well as typical phenomena of many Latin American varieties such as seseo (s, z and c e,i are realized as /s/), yeismo (both y and ll are realized as /j/) and voseo (in Chile mostly pronominal tuteo with verbal voseo, e.g. tu jugai instead of Standard SP tu juegas). At present virtually everyone on Easter Island manages to speak some Spanish, the majority of whom as an L1 or with close to L1-like perfection (Makihara 2005: 728).

Comparative aspects of Hispanisation

173

Based on Old Rapanui, Modern Rapanui has incorporated linguistic material from Tahitian, Mangarevan, English, French and only recently from Spanish (Makihara 1999: 178–181, Fischer 2001: 314). From the perspective of this paper, however, the latter is of essential importance. See the following excerpt from a conversation between Rapanui elders, two teachers and a school class, recorded in December 2003 in Easter Island’s only village Hanga Roa; Hispanisms are bold: (2)

Rapanui [RA-KONU-3]5 01 IO 02 03 04 05 06 V 07 IO 08 09 10 11 12 V 13 IO 14 15 16

ya… te probrema o mtou, o te puebro te probrema, well, the problem is ours, the problem of the people, ‘ina he angi-angi te me’e nei e necesita nei, who don’t know that it [Rapanui lang.] is necessary, te ng poki nei e noho mai nei the children living on this island necesita ki te rua arero mo ‘ite, need to know their language, ki te arero extranjero mo ‘ite mo te rua anga (and) they need to know the foreign language for their working hmhm hmhm ya, he va’ai te importante he k ki te:: profesora era, so it is necessary to make their teacher understand, o mu’a jefe era, or their future boss, te me’e importante mo te ng poki nei (that) it is important for the children he aumento haka’ou i te hora, to increase the time [of instruction in Rapanui] again,

fifty-fifty!

yes indeed! i a au e vive atu ena hai vananga rapanui, (like this) I am impoverishing with Rapanui language, a au e vananga nei i roto i te rario manu kena (but) I talk in Radio Manu Kena hai vananga rapanui using Rapanui language ‘e/‘e ‘ina kai veve ‘ a au and, and (then) I am not impoverishing

174 Steve Pagel A first observation to extract from (2) is that Hispanisms in Modern Rapanui are frequently used, but appear to be somewhat restricted to the level of lexicon. We find nouns, verbs and adjectives of Spanish origin (e.g. probrema ‘problem’, necesita ‘to need’, extranjero ‘foreign’), and except for one conjunction (o ‘or’), one discourse particle (ya ‘well’) and the use of a noun marked for gender in Spanish (profesora ‘female teacher’), example (2) lacks grammatical Hispanisms. Yet, before general conclusions about the quantity and quality of Hispanisms in Modern Rapanui can be drawn, an important theoretical issue must be addressed. It relates to the distinction between processes of code-switching and borrowing, about which no consensus has yet been reached in theoretical contact linguistics. Winford (2003: 107) resumes what appear to be the “chief criteria that have been used to distinguish them”, both of which are equally problematic when applied to the present situation on Easter Island: 1. the degree of use of a loan by monolingual speakers 2. the degree of its morphophonemic integration The first criterion is difficult to apply to Easter Island because there are few, if any, monolingual speakers of Rapanui left. The second criterion is equally weak, since the frequency of use of Hispanisms in Modern Rapanui, as well as at least their phonemic integration varies heavily among speakers and speech situations, this variation itself being an important symbolic act of the Rapanui speech community (Makihara 1999, 2005)6. From a sociolinguistic point of view, Makihara’s (1999, 2005) bilingual continuum takes account of that fact; from the perspective of descriptive linguistics, however, the case is somewhat more difficult. Note that excerpt (2) is one with a comparably large portion of Hispanisms (17 of 110 tokens), but there are others, even in the same conversation, showing no element of Spanish origin at all. Accordingly, the question as to the portion of Hispanisms in Modern Rapanui cannot be answered objectively, with our still popular conception of language as a monolithic complex of material and rules being, of course, somehow at the core of the matter. Nonetheless, this is not supposed to mean that no statements can be made at all. Both inter-sentential and intra-sentential code-switching between Rapanui and Spanish are common strategies among the Rapanui speech community, the latter category comprising a continuum that ranges from spontaneous single-word switches (or ‘nonce loans’, Sankoff et al. 1986) down to fully integrated borrowings (Myers-Scotton 1993). Makihara, however, reminds us that “[t]he awareness of the Spanish provenance of these borrowings […] is heightening as purist ideology spreads” (2001b: 200). Still,

Comparative aspects of Hispanisation

175

speaking ‘pure’ Rapanui (which is not identical with Old Rapanui) by completely avoiding Hispanisms is considered unnatural by most speakers, and seems to be restricted to the milieu of local politics and the institutional tuition of Rapanui language. Thus, the following sections will concentrate on the more unmarked areas of Modern Rapanui, focusing primarily on items closer to the nativized-borrowing end of the above mentioned continuum in intra-sentential code-switching. The term borrowing will be used in its classical sense as “the incorporation of foreign elements into the speaker’s native language” (Thomason and Kaufman 1988: 21), implying that there is a somewhat intuitive knowledge about the degree of that incorporation, but one that is based at least partly on the frequency of use among the speech community (this being a result of a process Croft 2000 calls ‘propagation’) and the awareness of the foreign provenance of the borrowed element. Borrowings from Spanish in Modern Rapanui comprise all parts of speech, including function words such as conjunctions, prepositions and discourse particles, with nouns and adjectives being, as in many other contact situations, the most frequently borrowed elements. Borrowing usually involves phonetic assimilation, although assimilation processes differ among speakers and speech situations (Makihara 2001b: 195). The phonological set of Rapanui comprises five vowels but only ten consonants (Fischer 2001: 314), hence phonetic assimilation of Spanish vocabulary is most salient in the consonantism, even more so since Rapanui traditionally has open syllable structure (C)V. Thus, the Spanish noun problema ‘problem’ could be found in Rapanui speech as problema, probrema (2), porobrema or poroborema, while the introduction of the respective consonant clusters (as in (2) puebro, importante, extranjero, profesora, aumento) can be considered a contact-induced alteration of Rapanui phonology (Fischer 2001: 318). Lexical items borrowed from Spanish are usually incorporated into Rapanui grammar, i.e., they meet the terms of Rapanui syntactic rules and receive morphological modification from Rapanui techniques. For example, the Rapanui distinction between possessive categories A and O applies also for borrowed lexemes used in the function of a noun. These categories encode the relation between the possessor and the possessed in terms of dominance and control in many Polynesian languages7. The unmarked O-category expresses lack of dominance or control over the possessed (used for one’s name or age, friends, land, feelings etc.) while A-category marks the opposite, i.e., dominance and control of the possessor over the possessed (used for one’s wife and children, tools, food etc.). Thus, in Modern Rapanui one will find, e.g., to’ona v a (SP vida) ‘his/her life’ but ta’ana pepe (SP bebé) ‘his/her baby’. Accordingly, the modification of Hispanisms in verbal func-

176 Steve Pagel tion is processed by means of partial or total reduplication, the causative prefix haka- or the nominalizer suffix –[VOWEL]nga, to name just a few techniques. Recent contact-induced alterations of Rapanui morphosyntax are discussed in detail by Makihara (1999, 2001b) and Fischer (2001). Among the most prominent is the increasing use of Spanish coordinating conjunctions o ‘or’, pero ‘but’, y ‘and’, and sino (que) ‘but, except (that)’ instead of the traditional paratactic sentence organization (Du Feu and Fischer 1993: 167; Makihara 2001b: 210–212). Noteworthy in this context is a study on the borrowing of function words in nine language contact situations with Spanish in the circumpacific area, carried out by Stolz and Stolz (1996, 1997), where pero and o turned out to be the most often borrowed conjunctions; a finding that is hence confirmed in Rapanui. Fischer (2001: 322–323) notes a general drift of Rapanui syntactic patterns towards Spanish patterns, another indication being the prepositive use of a demonstrative pronoun that traditionally was only postpositive. Instead of te me’e nei ‘this thing’ or te hare (e)na ‘that house’ one is likely to find nei me’e or n hare today, seemingly an imitation of SP esta cosa, esa casa etc. Further evidence leads to the assumption that Rapanui neutral word order is shifting from VSO towards Spanish SVO (Fischer 2001: 323). Fischer contrasts the following examples of which (3) represents the more traditional form and (4) how it would frequently be expressed today: (3)

Rapanui (Englert 1948, in Fischer 2001: 323)8 he ‘aroha, he tatangi ararua ML suffer ML weep both ‘they both suffer and weep’

(4)

Rapanui (Fischer 2001: 323) ararua he ‘aroha, he tatangi ‘ ML suffer ML weep DEM both ‘they both suffer and weep’

Another prominent morphosyntactic Hispanism is mentioned by Makihara (2001b: 206–207): the expression of obligation with SP tiene que. In contrast to the respective autochthonous form, which makes use of the TMA marker e (state, non-past), obligation with tiene que “allows the speaker to strengthen the sense of obligation and necessity”. Like most Spanish verb forms borrowed into Rapanui, SP tener is used invariably in its third-person singular present indicative form tiene, a fact that is of particular interest if

Comparative aspects of Hispanisation

177

one takes into consideration that neither subject nor object are expressed in Rapanui if evident from the context (Du Feu and Fischer 1993: 167, Du Feu 1996: 140). Obligation with tiene que can be further emphasized via the insertion of a subject pronoun, portraying then, however, a violation of Rapanui neutral word order VSO, which would require to postpose it after the modal complement (Makihara 2001b: 207), and by doing so providing further evidence for an ongoing change in word order under the pressure of Spanish: (5)

Rapanui (RA-KONU-3) he noho i roto o te tokerau tiene koe que hoki ki ML stay in inside of D wind OBL 2SG OBL return to te vnanga hiva era D language continent that ‘to keep your sails in the wind you have to fall back upon Spanish’

In sum, due to the relatively recent language contact, there is significant lexical but comparatively little structural borrowing from Spanish into Modern Rapanui. Borrowed Hispanisms are usually embedded in East-Polynesian morphosyntax, although minor alterations towards Spanish patterns are undeniable and the ‘hybridization’ of Rapanui is probably further on the rise. However, the conception of a linguistic continuum, as adopted here from Makihara (1999, 2005), makes it difficult to speak of the Rapanui language, even more since this is lacking a broadly accepted standardization9. Nevertheless, applied to the rather unmarked areas of the linguistic complex Modern Rapanui, we can translate the above reviewed observations of Makihara (1999, 2001b) and Fischer (2001) into the terms of Thomason’s and Kaufman’s borrowing scale (1988: 65–109). Then, the alteration of Rapanui caused by language contact with Spanish has reached the second (borrowed conjunctions instead of parataxe) or third (minor syntactic changes) of five possible levels, depending on one’s interpretation of the attribute ‘minor’. The term Rapanui Spanish on Makihara’s continuum comprises varieties of Spanish spoken on Easter Island that are not identical with Chilean Spanish. Rapanui Spanish is subdivided into two varieties, R1S2 Rapanui Spanish (L1 Rapanui, L2 Spanish) and R2S1 Rapanui Spanish (L1 Spanish, L2 Rapanui), that are distinguishable both in the way they interfere with Rapanui patterns and on the basis of their speakers’ age. Roughly spoken, R1S2 Rapanui Spanish covers those speakers of the bilingual community above thirty-five years of age, R2S1 Rapanui Spanish those under thirty-

178 Steve Pagel five. A reason for this ‘neat cut’ can be found in the community-wide language shift from Rapanui to Spanish that has occurred from the 1960s onwards, i.e., following the full integration of Easter Island into the Chilean state and an increasing migration to and from the island (Makihara 1999: 107–111)10. In accordance with Thomason’s and Kaufman’s (1988: 37–45) observations on language shift, subvariety R1S2 Rapanui-Spanish – i.e., that of speakers whose L1 is Rapanui – shows influence from Rapanui primarily at the phonological and morphosyntactic level, and not in terms of vocabulary. In other words, structural patterns from Rapanui overlay with those of Spanish; to what degree this takes place largely depends on the speaker and his or her knowledge of Spanish, though also on the context in which the speech is placed, as well as on the respective addressee(s) (Makihara 1999: 174–175, 2005: 739–740). Salient morphosyntactic features of R1S2 Rapanui Spanish comprise, e.g., missing concordance in number (examples 6 and 7) and gender (6), incorrect use of the existential verbs ser and estar (6), violation of tense, mood and aspect rules as well as the consecutio temporum (6 and 7), many of which are paralleled in other second language varieties of Spanish (exemplarily Montrul 2004): (6)

R1S2 Rapanui Spanish (RA-TEHU-1) cualquier cosa que yo hago every thing.SG.FEM that 1SG.NOM make.1SG.PRES.IND estaré bonito porque me be.1SG.FUT.IND beautiful.SG.MASC because 1SG.DAT gustan hacer-lo like.3PL.PRES.IND make.INF-3SG.ACC.MASC ‘everything I make is (going to be) good because I enjoy working on it’ (vs. Standard SP: cualquier cosa que yo haga será bonita porque me gusta hacerla)

(7)

R1S2 Rapanui Spanish (RA-TEHU-1) por.eso yo te dijo, si éste te therefore 1SG.NOM 2SG.DAT say.3SG.PRT if this 2SG.DAT ofreció veinticuatro tu dice ‘bueno’ offer.3SG.PRT twenty-four SG.NOM say.3SG.PRES.IND good ‘that’s why I told you, if I sold it to you for twenty-four, you’d say ‘ok’’ (vs. Standard SP: por eso yo te dije, si éste te ofreciera por veinticuatro tu dirías ‘bueno’)

Comparative aspects of Hispanisation

179

Makihara (2005: 740) argues that there is a functional aspect of R1S2 Rapanui Spanish beyond that of a mere variety of Spanish as a second language. According to her the “[c]ontinued R1S2 Rapa Nui Spanish use with outsiders and among themselves […] can highlight ethnic differences and underline ethnic solidarity, respectively”. Although this claim might be valid for the present, it is doubtful that this subvariety will stabilize and develop rules of its own right, in other words: that it will survive any longer than the last speakers of Spanish as a second language on the island. Instead of structural interference, the second subvariety R2S1 RapanuiSpanish – i.e., that of speakers whose L1 is (Chilean) Spanish but who hold at least a marginal knowledge of Rapanui – is characterized by interference primarily on the level of vocabulary, and hence corresponds to what Thomason and Kaufman (1988: 37–45) state for contact induced change in language maintenance11. Speakers of R2S1 Rapanui Spanish incorporate lexical items from Rapanui in an otherwise (Chilean) Spanish linguistic environment, but in contrast to Hispanisms in Modern Rapanui that can be regarded as more or less nativized, these ‘Rapanuiisms’ are essentially spontaneous lexical transfers, moving R2S1 Rapanui Spanish conceptually closer to the code-switching complex12: (8)

R2S1 Rapanui Spanish (RA-KONU-3) 01 AL mi nua trató de/ de darme too su me’e my grandmother tried to, to teach me her things pero lo que pasó fue que mi mamá… le da::/ 02 but what happened was that my mother was, 03 le daba miedo ense/ que me=enseñaran Rapanui, was afraid of teaching, that they taught me Rapanui, porque si/ sabía que si yo me iba 04 because if, she knew that if I would leave 05 allá=en/ en el continente, for the continent [=Chile], 06 se=iban=(a)=hacer kata de mí, they would make fun of me (over there) 07 de/ de:/ desde=o de como yo hablaba, of, of, from, or of how I speak 08 V hmhm hmhm por eso mi mamá no me quiso enseñar Rapanui 09 that’s why my mother didn’t want to teach me Rapanui

180 Steve Pagel As Makihara (1999: 173–175) points out, speakers of Rapanui Spanish will deliberately choose to ‘colour’ their speech by transferring items from Rapanui to Spanish, in order to emphasize their authenticity and solidarity among themselves, as well as their alterity towards outsiders (see Le Page and Tabouret-Keller 1985, Ludwig 1995). In other words, Rapanui items used in an otherwise Spanish linguistic environment establish an “[e]thnic identification […] ACROSS a language-based distinction that had previously served to separate Chilean from Rapa Nui” (Makihara 2005: 754, emphasis original), and thus become essentially important for younger Rapanui who do not have sufficient proficiency in their ancestors’ language13. Accordingly, the speaker in (8), who is being addressed in Rapanui during the whole dialogue but consistently responds in Spanish, adds more and more lexical material from Rapanui to her speech over several conversational turns. By doing so, she is stressing her Rapanui-ness towards her communication opposites – fellow Rapanui – without talking to them in their own language, which she is able to but of which she is also afraid of, as she clearly demonstrates elsewhere14. 4. Linguistic aspects of Hispanisation on the Marianas As a result of the demographic collapse, in the course of the 18th century the autochthonous language of the Marianas changed significantly. Few documents about the old Chamoru language are preserved (see Burrus 1954) but it is probably safe to say that the linguistic contact with Spanish has led Chamoru to a new language stage, considering also Rogers’ assumption of an absorption of the ethnic Chamoru by what he calls ‘Neo-Chamoru’ (Rogers 1995: 104) before the end of the Spanish era in 1898. Accordingly, in the following section I will distinguish between Old Chamoru as a collective term for the autochthonous varieties spoken on the Marianas prior to the 18th century, when the number of ethnic Chamoru hit rock bottom and mesticization progressed, and Modern Chamoru, comprising the varieties spoken from then onward, the latter being characterized, as we will see, by an amazing coexistence of Austronesian and Indoeuropean patterns. Before going into the details of this coexistence, it must be pointed out that due to this there seems to be no agreement among scholars about the typological status of Modern Chamoru. In their pioneer study, RodríguezPonga and Albalá Hernández considered it a ‘mixed language of Austronesian basis with strong penetration by Spanish’ (“lengua mixta de base austronésica, con fuerte penetración española”, Albalá Hernández and Rodrí-

Comparative aspects of Hispanisation

181

guez-Ponga 1986: 66), but the same term was later engaged by Bakker and Mous (1994) to refer to a specific type of contact language which is not suitable for Modern Chamoru, as Stolz (1998, 2003) has shown15. Since then, Chamoru has been labelled quite generously ‘a Creole language’ (“una lengua criolla”, Rodríguez-Ponga 2001: 277), ‘an atypical Creole’ (“un criollo […] atípico”, Munteanu 1997: 962) and even ‘the opposite of an Anti-Creole’ (“o oposto de anticrioulo”, Couto 1996: 89), all of which is, in my opinion, and with regard to the specific socio-linguistic and structural criteria usually assumed for Creole languages, inappropriate (see, e.g., Holm 2000; Mufwene 2001; Lefebvre 2004; and McWhorter 1997, 2005 for discussion).16 I am not aiming at a solution of that matter here, but propose for the moment the more neutral assumption of a new language stage Chamoru has entered due to language contact and change in the course of the 18 th century. See example (9) from an open letter in the internet, which gives a good impression of how Modern Chamoru is mixed: (9)

Chamoru (CH-CHB-2) Lao kao diebi na ta sokni i famaguon? But should we accuse the children? Pues hayi siempre ta achaka put este na baban prublema? Who else is to be held responsible for this bad problem? Bai atotga sumangan na hita-ha' buenti niman Saina. I will dare to say it is maybe us alone, the parents. Diebi na ta gof attendi i famaguonta. We must take good care of our children. Metton yu nu esti. I am certain about this. Taimanu siha ufan atituyi hit yanggin meggaiña hit man tinani How can they pay attention to us if we are more busy nu asunto siha fuera ki i familia-ta. with the things happening outside of our family? Yanggen ta puni na hita na gaige i chetnot, If we deny that the problem is with us, pues manmachochomguan i mata-ta. then our eyes are closed. Entonses na ti poddong i obligasion giya hita Do you mean that the obligation

…

182 Steve Pagel na diebi na ta fa'naguen maolek i famaguonta? to teach our children well does not fall on us? Yanggen ti hita, pues manu na poddong i obligasion? If not we, who then has abandoned this duty? Asiguru na ti poddong i obligasion gi iman maestro Surely it wasn’t the teachers yan maestra pat i pulisman siha. or the policemen who abandoned this duty. Fanmanhasso achataotaota, sa' hita siempre manmañotsot. Let’s think, people, because (if not) we will surely repent it. The problem of distinguishing between code-switching and borrowing, as has been addressed above with regard to Modern Rapanui, does not appear here; Spanish as a spoken language has vanished from the Marianas since at least the middle of the last century (see Albalá Hernández and RodríguezPonga 1986), hence Hispanisms in Modern Chamoru speech must be considered nativized borrowings for the most part, whose foreign provenance is by and large unknown to the speaker17. Quite the opposite is, by the way, true for loans from English, which frequently appear in Modern Guamese Chamoru (e.g. pulisman ‘policeman’ in (9)), but are fairly salient to the speakers and should therefore be considered as code-switching rather than as borrowing phenomena. Modern Chamoru phonology encompasses several sounds that have been introduced via borrowing from Spanish; among the most salient are consonant clusters, as in the words siempre, prublema or entonses in (9), and most likely the mid vowels /e/ and /o/, which in pre-contact Chamoru were probably allophones of /i/ and /u/ (Topping 1973: 23; Blust 2000: 92). Hispanisms are not limited to any parts of speech or semantic fields, they include numerals and the terms for weekdays, and provide also the vast majority of Chamoran function words (i.e. lexemes with a grammatical function, such as conjunctions, prepositions, etc.)18. From an analysis of the most detailed dictionary of Modern Chamoru (Topping et al. 1975), RodríguezPonga (1995) calculates an astonishing number of 54.9% Hispanisms in the lexical inventory of the language. Although his analysis is not based on real samples of the language, spoken or written, he is assuming that “las palabras de origen español que usan los hablantes actuales del Chamorro constituyen entre un 50 y un 60% de su léxico” (‘the words of Spanish origin used by present Chamoru speakers constitute some 50 to 60% of their lexicon, Rodríguez-Ponga 1995: 92). My own impression from an extensive corpus of spoken and written Guamese Chamoru is, however, that the pro-

Comparative aspects of Hispanisation

183

portion of Hispanisms in use varies a good deal among speakers and speech situations, and is all at once notably lower (it rarely exceeds 30%, see Pagel 2003: 36–38). Correspondingly, in example (9) only 27 of 113 tokens are of Spanish origin.19 One explanation for this disproportion can be found in the observation that most Hispanisms in Chamoru have an autochthonous substitute (as in (9) prublema vs. chetnot), which, in the light of an increasing awareness of their cultural roots, might be preferred by Chamoru speakers, given, of course, that the provenance of the according element is known to them (see Salas Palomo and Stolz, this volume). Unlike in Rapanui, the language contact with Spanish has prompted also considerable modifications of Modern Chamoru morphosyntax, with these modifications often exhibiting noteworthy innovation rather than simple borrowing20. For the sake of brevity only the most important features shall be presented here (see Stolz 1998, 2003 and Pagel 2003, 2005, forthcoming, for details). An unknown category to Austronesian languages, grammatical gender and agreement has been introduced to Modern Chamoru via borrowing from Spanish, but is neither obligatory nor otherwise an accurate reproduction of the category in Spanish. Gender in Modern Chamoru has three dimensions: first, nouns with the semantic feature [+ANIMATE] can be marked for gender by simply adding the lexemes lahi ‘man’ or palao’an ‘woman’, as in che’lu lahi ‘brother’ and che’lu palao’an ‘sister’. Second, a morphological gender distinction remains active on those nouns with the same semantic feature borrowed from Spanish, such as maestro/maestra ‘teacher’ in (9), this distinction being, however, not productive since it does not extend to autochthonous material. Third, an optional gender agreement has come into being between adjectives borrowed from Spanish in their masculine and feminine form, and nouns with the semantic feature [+HUMAN], now of both Indoeuropean and Austronesian origin. Therefore, in (10) the Austronesisms lahi ‘man’ and palao’an ‘woman’ activate the gender distinction on the borrowing bunitu/bunita ‘beautiful’, while in (11) the Hispanism isla ‘island’ (feminine in Spanish) does not, and the adjective remains unmarked (i.e. formally masculine) bunitu. (10) Chamoru (CH-SAMI-1) nu guahu i.mas bunitu na lahi si Hritik, yan NF 1SG.EMPH SUP beautiful.MASC L man D.F.PS Hritik and i.mas bunita na palao'an si Mahdri Dixit SUP beautiful.FEM L woman D.F.PS Mahdri Dixit ‘for me, Hritik is the most beautiful man and Mahdri Dixit is the most beautiful woman’

184 Steve Pagel (11) Chamoru (CH-SAMI-4) gofmahalang yu' nu.i bunitu na isla very miss 1SG.ABS D.NF beautiful.Ø L island ‘I miss the beautiful island very much’ A related topic is the issue of articles in Chamoru, since some are loans from Spanish where this category is always marked for gender. The indefinite article un can be considered a direct borrowing, but it is not subject to gender distinction in Chamoru (see Stolz and Sabater Fuentes 2002). More difficult to define is the status of the definite article for common nouns in focus i, which is also not marked for gender and whose origin is a little controversial. Although similar particles exist in more or less closely related Austronesian languages such as Ilokano ti and Malagasy ny, the function of i as a nominalizer, and the use of i mas or itmas as markers for the superlative in Chamoru (cf. SP el más) strongly indicate a Spanish origin. Rodríguez-Ponga (1995: 72) synthesizes the problem, suggesting that SP el might have merged with a pre-contact Chamoru particle *i. See (12) for the use of the indefinite article un and (13) for nominalization with the definite article i. (12) Chamoru (CH-ESCH-1) guaha un che´lu-ta exist ID sibling-POSS.1PL.INCL ‘we have a brother’ (13) Chamoru a. i dikike na taotao D.F little L man ‘the short guy’ (SP el hombre pequeño)

b. i

dikike little ‘the short one’ (SP el pequeño) D.F

Another morphosyntactic category that has been Hispanised to a considerable degree is comparison. In the comparative the Spanish pattern mas + adjective + que in the form mas ki or mas kinu complements the Austronesian affixes la’- and -ña, the three differing mainly in terms of pragmatic nuances. The superlative in Chamoru is formed by a combination of the definite article for common nouns in focus i and the Hispanism mas ‘more’ (sometimes itmas). Example (14) shows a combination of comparative markers of both Spanish and Austronesian origin, for the superlative see again (10).

Comparative aspects of Hispanisation

185

(14) Chamoru (Topping et al. 1975: 137) metgot-ña i kabayu mas ki i ga'lagu. COMP COMP D.F dog strong-COMP D.F horse ‘The horse is stronger than the dog.’ Borrowed function words such as prepositions, conjunctions, discourse markers, negative and interrogative markers indicate further substantial modification of Chamoran morphosyntax (Stolz 1998; Rodríguez-Ponga 1998; Pagel 2003). Unlike in Rapanui, and in a number of other contact situations with Spanish in the circumpacific area investigated by Stolz and Stolz (1996, 1997), no coordinating conjunction has been borrowed into Chamoru; the set comprises the Austronesisms ya(n) ‘and’, lao ‘but’ and pat ‘or’. Hispanisms like komo ‘if’, antes ki ‘before’ or mientras ki ‘while’ constitute, however, the greater part of the subordinating conjunctions, as Topping (1973: 151) and Stolz (1998: 24–25) show. Probably the deepest marks of the language contact with Spanish are exposed by the categories tense, mood and aspect (Pagel and Pfänder 2001; Pagel 2003: 53–64, 2005). Unmarked, a verbal action is interpreted as noncontinuative and completed in Chamoru (CH táitai ‘read’); continuative aspect is marked by reduplication of the stressed vowel and the preceding consonant (tátaitai ‘reading’). The Hispanisms era, esta and estaba were borrowed and accomplish similar functions in Chamoru as they do in Spanish (Rodríguez-Ponga 1996); e.g. estaba, among other functions, indicates habitual existence in the past, and by doing so complements the autochthonous set of existential verbs, consisting of gaige ‘be present’, taigue ‘be not present’, guaha ‘have, there is’ and taya ‘not have, there is not’. An issue of particular importance is future tense, which is grammatically marked by at least two different elements, although in everyday speech, if a phrase is marked for future by one element, the other becomes pragmatically redundant and can be omitted. In both marker classes Hispanisms are indispensable, as the following table from Pagel (2003: 59) shows: Table 1. Future paradigm in Modern Chamoru Marker class I

Marker class II

Ergative pronoun21

bai Ø u

hu un Ø

Singular 1. para/siempre/debi di 2. para/siempre/debi di 3. para/siempre/debi di

186 Steve Pagel Plural 1. para/siempre/debi di 2. para/siempre/debi di 3. para/siempre/debi di

EXCL bai INCL u

Ø u

or Ø

EXCL

in

INCL

ta

ha

PL ma

en DUAL

Before considering the first marker class, which is composed exclusively of Hispanisms, the marker bai in class II deserves our attention. The use of bai as a future marker particularly for the first person (singular and plural) indicates a provenance from SP voy (1SG.PRES.IND of ir ‘to go’) and hints automatically at the analytic Spanish future with auxiliary ir + preposition a + infinitive (Stolz 2003: 275). In addition to this correspondence in form and function, the absence of similar markers in typologically neighbouring languages of the Philippines and Micronesia strongly speaks in favour of a Spanish origin of bai (Topping 1973: 262). However, one cannot entirely rule out the possibility that this marker is a construction compliant with the future marker bai in several English-based Creoles in the Pacific, ultimately originating from English by and by (see, e.g., Smith 2002: 125–129). In contrast to marker class II, class I is not structured by grammatical person but reflects an innovative modalization of the category future in Chamoru, making use of highly grammaticalized Hispanisms. Compare the following examples: (15) Chamoru (CH-ESCH-1) para u ma protehi este na ressource FUT FUT.3 3PL.ERG protect this L resource ‘they (3+) will protect this resource’ (16) Chamoru (CH-KIN) klaru na yanggen ti ma sapotte i kinalamten-ña sure that if NEG PASS support D.F activity-POSS.3SG i lenguahi, kun tiempo siempre u måfnas giya Guåhan D.F language with time FUT FUT.3 die.out in.PLA Guam ‘it is clear, that if the activities of the (Chamoru) language are not supported, it will certainly die out on Guam’ (17) Chamoru (CH-OBI-2) debi.di bai hanao para Jerusalen para bai famadesi meggai FUT FUT.1 go to Jerusalem FUT FUT.1 suffer much ‘I will have to go to Jerusalem to suffer many (things)’

Comparative aspects of Hispanisation

187

While future with para is of a rather neutral or uncertain kind, siempre makes an additional reference to the possibility of the action happening. Strictly said, siempre implies a firm certainty or determination that the action will happen under the conditions stated or otherwise known to the addressee(s). Future with debi di, on the other hand, implies an obligation for the agent of the action, reflecting the modal origin of the marker in Spanish deber de ‘have to’. A highly Hispanised category without doubt, Chamoran future is nevertheless not a reproduction of the Spanish category, neither in form nor in function, but must be considered a remarkable innovation that has emerged through language contact. However, a hint to the paths of this emergence is given in (17), where the use of the para-bai-paradigm as a marker for final clause (best translated as ‘in order to’) indicates a wider functional scope of the same. The cognitive proximity between finality and future tense has been emphasized already in connection with grammaticalization processes, e.g. in the development of Romance and English analytic future (Hopper and Traugott 1993; Heine et al. 1991; Heine and Kuteva 2002); accordingly, the grammaticalization of CH para (bai) might have followed (or continued) the path from finality to future tense pursued before by the analytic Spanish future, where the pattern ir a // hacer algo ‘go to do something’ evolved out of ir // a (X para) hacer algo ‘go to (X in order to) do something’. In sum, the contact-induced language change in Chamoru can be considered extensive. There has been enormous lexical and also significant structural borrowing from Spanish, although in many cases Spanish elements were not just copied but used to create something entirely new. In an admittedly liberal interpretation of the criteria of Thomason’s and Kaufman’s borrowing scale, the contact-induced change in Modern Chamoru reaches the fourth (gender distinction, future markers) of five possible levels. However, with the outstanding exception of future tense and some borrowed lexemes of central grammatical function, the affected areas are situated on the more peripheral parts of Chamoru structure, or are, as in the case of grammatical gender and comparison, mere options or variants. Due to its lacking of required sociolinguistic and especially structural criteria, Modern Chamoru must hence not be considered neither a Spanish based Creole nor a (bilingual) mixed language, as defined by Bakker and Mous (1994) and Thomason (2001). Chamoru remains Chamoru, i.e., despite major modifications due to intensive contact with Spanish, the language is safely rooted in the West-Malayo-Polynesian group, featuring typical constituents such as agglutinating morphology, ergativity, and elaborated strategies of grammatical focus. In other words, Chamoru did not undergo

188 Steve Pagel what Ross calls metatypy, i.e. “[t]he change in morphosyntactic type which a language undergoes as a result of its speakers’ bilingualism in another language” (Ross 1999: 1). From these observations, it seems adequate and justifiable to distinguish – consciously in accordance with what has been done for European languages such as English, Spanish and German – between two language stages, a pre-contact (Old) and a contemporary (Modern) Chamoru, as has been done in the course of this paper. 5. Some linguistic aspects of Americanization on Guam A straightforward resume of the linguistic changes occurring on Guam during the last centuries gives Rogers: “[t]he precontact Chamorros said hacha, hugua, tulo; the partially Hispanised Chamorros said uno, dos, tres; the Americanized Guamanians say one, two, three” (Rogers 1995: 245, italics original). However close the outcome seems here, the mechanisms of language change set in motion by the arrival of the U.S. on Guam were entirely different to those characterizing the contact with Spanish. Until 1898 Chamoru had always had a firm stand in Guam’s public domain, and Spanish language policy never intended to remove it from this position, while the cultural pressure from the U.S. has been overwhelming from the beginning, and the use of Chamoru was legally restricted. According to Thomason and Kaufman (1988: 100), one possible outcome of overwhelming cultural pressure is language shift to the dominating language, similarly to what has been stated above with regard to Easter Island. To be precise, language shift occurred almost simultaneously in both places from the 1960s onwards, when Spanish on Easter Island and English on Guam began to conquer the households, after having, at least in the latter case, conquered most other domains beforehand. Rogers shows that by 1914 “English had replaced Spanish in schools, government offices, courts, and most businesses” on Guam, but Chamoru was “still the dominant language in local homes and gatherings” (Rogers 1995: 134). This remained the case until the late 1950s, despite continued linguistic oppression by U.S. administration. From the 1970s on, however, Chamoru has been repeatedly reported as being in a state of decline (e.g. by Odo 1972), and this downward trend has continued until the present day. The timing of this language shift, which in fact would have been expected to occur much earlier, can be explained by at least three factors: 1) the broadcasting of radio and television programs from the U.S. from the mid 1950s onwards, 2) the gradual opening of Guam for foreigners, and

Comparative aspects of Hispanisation

189

consequently 3) the advent of tourism in the 1960s. Thereby, an ideology of English that had been promoted since the turn of the century finally appeared to come true: in terms of this ideology, the acquisition of English was synonymous with economic success, with technological progress and overall improving conditions for Guam (Underwood 1987). Similar arguments can be used to explain the communal language shift to Spanish on Easter Island, occurring almost simultaneously from the 1960s onwards (Makihara 1999: 107–111). Given this parallelism of language loss and reassertion, it does not seem surprising that local varieties of English, similar to the Rapanui Spanish varieties described by Makihara, developed throughout the Marianas, but particularly on Guam and in the Chamoran community on the American continent. See the following excerpt of another poem by the aforementioned Peter Onedera, in which the English speech of Chamoru children on Guam around the 1960s is imitated. The topic of the poem is, appropriately enough, an English lesson: So everyone keep quiet and speak Chamoru secretly / an’ in da class we spik englis in high voice, very fancily / I read in da book, “da jibra in da joo, di’åpblo.” / she say, “no, no, no, not to be” / den I laff and she look at me, face red, very angrily // (Onedera 1994: 7)

Maybe not accurate in every detail, Onedera’s example gives a good impression of the phonetic interference of Chamoru in what I will call CH1E2 Chamoru English (L1 Chamoru, L2 English) here, corresponding to the terms coined by Makihara for Rapa Nui, i.e., a collective term for second language varieties of English spoken by Chamoru who were born, for the most part, well before the 1950s. The phonetic realization of dental fricative // as alveolar plosive /d/ in da and den, and that of alveolar fricative /z/ as affricate /d
/ in jibra and joo may serve as simple examples here; in general, of course, phonetic influence from Chamoru in CH1E2 Chamoru English varies according to the individual’s mastering of English phonetics. Analogous to what has been stated for R1S2 Rapanui Spanish, further influence from Chamoru is most salient at the morphosyntactic and the semantic level of CH1E2 Chamoru English. Therefore, in the above excerpt the verbs say and look have no 2nd person ending -s, a fact that is understandable from Chamoru being a non-flectional language. Aside from that, there is only one lexical element from Chamoru in the above excerpt: the exclamation di’åpblo ‘devil’ (ultimately from SP diablo). In contrast to what Makihara claims for R1S2 Rapanui Spanish, however, I am not aware of CH1E2 Chamoru English being particularly used for emphasizing ethnic difference

190 Steve Pagel or ethnic solidarity, with the exception, naturally, of an understanding of the above poem itself as such an emphasis. In contrast to Easter Island, Guam is a melting pot for immigrants from many different places, often speaking rudimentary varieties of English themselves. The effect of using CH1E2 Chamoru English might hence run counter to the intention: instead of highlighting the speakers affiliation to the autochthonous culture, it would mark him as foreign. Sociopsychological factors, however, play a key role in the varieties of English found among younger Chamoru, especially on Guam, who have grown up speaking English for the most part, and hold only passive or marginal competence in Chamoru. Corresponding to what Makihara states for R2S1 Rapanui Spanish, CH2E1 Chamoru English (L1 English, L2 Chamoru) can be said to highlight the speaker’s affiliation to the autochthonous culture without actually speaking in the according tongue. In other words, the use of CH2E1 Chamoru English marks in-group and out-group relations in terms of ethnic affiliation, and includes the deliberate choice to speak English in a distinct way by ‘disguising’ it with, above all, lexical items transferred from Chamoru. As in the case of R2S1 Rapanui Spanish, the question as to what extent the process behind these transfers is code-switching or borrowing cannot be exhaustively answered here; placed on a continuum, however, most items are arguably closer related to the first than to the latter. See examples (18) to (20) from prominent Chamoru internet communities: (18) CH2E1 Chamoru English (CH-CHACO-3) Magahet hao Chamoruborn, most of the posters in here are Chamorron Stateside and from all over, but mostly all over California I think. Lao Mr 1s yan 0s I will be in Guam for most of the summer, so when you're headed back post here, send me a message here. (19) CH2E1 Chamoru English (CH-CHACO-2) Pretty maolek chick, how'd you find this site? (20) CH2E1 Chamoru English (CH-CHACO-17) Buenas yan Hafa Adai to all!!! Yan’ Si’ Yu’us Ma’ase to all those involved in the making of this wonderful site. This is such a great way for all of us Islanders to keep in touch with the Familia, friends and our Culture [.] The quantity and type of Chamoruisms vary among the postings. While (18) starts with the expression magahet hao ‘you’re right’ and later employs the

Comparative aspects of Hispanisation

191

coordinating conjunctions lao ‘but’ and yan ‘and’, (19) makes use of the adjective maolek ‘good’ in a comparative construction with the English particle pretty. It is important to state that it is almost always basic concepts of Chamoru language that are used in CH2E1 Chamoru English, as becomes very clear in (20). Here the community is addressed with the common phrases buenas ‘greetings’, hafa adai ‘hello’ and si yu’us ma’ase ‘thank you’, and apart from these stereotyped expressions only one lexical transfer appears, familia ‘family’, whose meaning, however, is easy to grasp for every speaker of English. An explanation for this implicit restriction, also with regard to what Makihara states for Easter Island, can be approached from two ways: 1) younger Chamoru (especially on Guam) often lack a far-reaching knowledge of the Chamoru language, probably more so than most Rapanui do with regard to their ancestors’ language; and 2) the Chamoru community is bigger and further spread than that of the Rapanui. Unlike the latter, the Chamoru cannot presume to know every member of the community in person, which makes it more difficult to judge the language skills of the respective communication opposite as well as to negotiate the appropriate style of interaction (see Makihara 1999: 234–271), even more so if communication takes place via a medium like the internet. In other words, and in the terms of pragmatics: speakers of CH2E1 Chamoru English are constantly balancing between the maxim of extravagance (Haspelmath 1999) and that of clarity (Grice 1968, 1975). CH2E1 Chamoru English must be sufficiently different from American English in order to underline ethnic alterity, and it must be sufficiently similar to it, in order to be understood by all members of the speech community.

6. Conclusions Controversial in itself, Hugo Schuchardt’s (1884: 4) famous statement that “[e]s gibt keine völlig ungemischte Sprache” [‘there is no completely unmixed language’] implies an even farther reaching consequence: far from being exceptional, language contact, the unintentional and intentional mixing or syncreticizing of communication codes must be considered a phenomenon found universally among the world’s languages (Pfänder 2004). Nevertheless, there are periods of more and less intensive contact and contact-induced change in the evolution of every language, just as there are periods of more and less intensive socio-cultural pressure which the speakers of one language exert on those of another.

192 Steve Pagel This paper has illustrated two situations of substantial linguistic contact involving Spanish which, at first glance, seem only to a small extent related, but upon closer inspection show many tendencies of convergent development, due to similar socio-historical and linguistic conditions. Among the most important of these analogous affairs is the rapid change of the autochthonous languages Chamoru and Rapanui as a result of demographic collapse, ethnic mesticization and, at the same time, remarkable language maintenance. The outcome of this change has been interpreted in both cases as a new language stage (Old vs. Modern) instead of arguing pro or contra certain categories of contact typology (Creole, mixed language etc.). The main reasons for the rejection of these categories are their narrow and often controversial definition criteria as well as the a priori stigmatization of their speakers as, e.g., speakers of ‘non-genetic languages’ (Thomason and Kaufman 1988), illegitimate offsprings of other (European) languages, etc. (see Mufwene 2001). A second process that has been described for both situations is the emergence of vivid and rather spontaneous varieties ‘in-between’ the two languages involved in the contact, at least some of which accomplish an important function as a marker of identity in a linguistic environment that has been characterized in the last decades by an increasing language shift towards the hegemonic languages. From this perspective, the synchronic linguistic continuum modelled for the bilingual speech community on Easter Island (Makihara 1999, 2005) could be adapted easily to the bilingual Chamoru community on Guam, and it might be applicable to other situations of linguistic contact exhibiting comparable conditions as well. Notes 1.

2.

There is disagreement about the spelling of both terms. Without going into the details of the corresponding polemics, the form Chamoru will be used throughout this paper to refer to the autochthonous people and language of the Marianas, the form Rapanui for the autochthonous people and language of Easter Island, and the form Rapa Nui for Easter Island itself (see Topping 1973: 58–67; Fischer 1993a, 1993b, 2005: 91). The role of socio-historical parameters in contact-induced language change cannot be overestimated (Díaz et al. 2002: 389), or as Thomason and Kaufman (1988: 35) put it: “it is the sociolinguistic history of the speakers, and not the structure of their language, that is the primary determinant of the linguistic outcome of language contact. Purely linguistic considerations are relevant but strictly secondary overall”. From a fairly different but related perspective, the ecological approach to linguistics also emphasizes this indivisibility (see, e.g., Mühlhäusler 1996; Fill and Mühlhäusler 2001).

Comparative aspects of Hispanisation 3.

193

Although both Guam and the Northern Marianas are unincorporated territories, i.e., “[a]n area under U.S. sovereignty that is not intended by the U.S. congress to become a U.S. state as defined by the Supreme Court in the Insular Cases, 1901” (Rogers 1995: 347, italics original), their political status diverges. While Guam is an organized unincorporated territory, the Northern Marianas are a Commonwealth in political union with the U.S. For a discussion on the struggle of Guamanians for citizenship, stateship and reunification with the northern islands, see Rogers (1995). 4. Makihara (2005) prefers syncretic Rapanui to contrast present varieties of Rapanui with pre-contact Rapanui. From a synchronic perspective this seems to make sense, since syncretism is one of the most salient characteristic of Rapanui. Diachronically, however, this syncretism is relevant for linguists rather than for speakers. Moreover, a naming practice that is based on a singular genetic affiliation of languages as the unmarked case might stigmatize present and future speakers of Rapanui, as has been argued already by Mufwene (2001: 106–125) with regard to some varieties of English. 5. If not quoted otherwise, examples are taken from my own corpus of spoken and written language, collected on Guam and Easter Island from 2001 to 2004. Bold characters mark Hispanisms in Rapanui and Chamoru examples, or Rapanuiisms/ Chamoruisms in Spanish and English examples, respectively. 6. See also Mondada (2007: 168) who considers code-switching a “ressource mobilisée par les participants de manière contingente, localement située, sensible à l’organisation séquentielle de la interaction en cours”. 7. Thornton (1998) discusses the cognitive relation of these categories in Maori with the pan-Polynesian concept of tapu and hereby uncovers a possible path of grammaticalization. 8. ABS = absolutive, ACC = accusative, CH = Chamoru, COMP = comparative, D = definite article, DAT = dative, DEM = demonstrative, EMPH = emphasizer, ERG = ergative, EXCL = exclusive, F = focus, FEM = feminine, FUT = future tense, ID = indefinite article, INCL = inclusive, IND = indicative, INF = infinitive, L = linker, MASC = masculine, ML = main discourse line, NEG = negator, NF = nonfocus, NOM = nominative, OBL = obligation, PASS = passive, PL = plural, PLA = place, POSS = possessive, PRES = present tense, PS = person, PRT = preterite, SG = singular, SP = Spanish, SUP = superlative. 9. Makihara (2001b: 193) circumvents this problem by distinguishing with Saussure’s terms the levels of langue (Modern Rapanui as a language variety) and parole (bilingual and syncretic ways of speaking Rapanui). 10. A careful documentation of this shift is provided by Gómez Macker (1982) and Weber and Thiesen de Weber (1990, 1998). 11. With the (not insignifant) restriction that maintenance does not concern the dominated but the dominating language here, and borrowing is thus not from the superstrate but from the substrate language. 12. Similarly, Makihara cannot “predict if this new way of speaking Rapa Nui Spanish will stabilize” (2005: 755), while in Modern Rapanui “there are a

194 Steve Pagel

13. 14.

15.

16. 17.

18.

19. 20. 21.

number of highly nativized originally Spanish elements […] considered as Rapanui by most speakers” (2001b: 200). An interesting question arising from Makihara’s observation is, whether the syncretic speech styles subsumed under Rapanui Spanish, for the same reasons, are spoken in the Rapanui ‘enclaves’ on the continent. A frequent argument is that younger Rapanui are often corrected or teased by parents when speaking Rapanui. The outcome is said to be linguistic ‘frustration’ or the refusal to speak Rapanui. For a deeper analysis of unreciprocal language choice on Easter Island, see Makihara (1999: 249–253, 2005: 742–746). An often cited example for these (bilingual) mixed languages is Media Lengua in Ecuador, a language composed of Quechua morphosyntax and Spanish vocabulary (Muysken 1994, 1997). It is precisely this clear separation of two subsystems from two different languages that is the most defining characteristic of these contact languages. Accordingly, the process by which mixed languages evolve is called ‘language intertwining’ by Bakker and Mous (1994). See Matras and Bakker (2003) for the state of the art. One might add that the assumption of a synchronically definable contact type called Creole with specific language internal features is in itself problematic. However, what Makihara (2001b: 200) stated for Rapanui, that “[t]he awareness of the Spanish provenance of these borrowings […] is heightening as purist ideology spreads”, could in a weaker form become true for the Marianas too. See Salas Palomo and Stolz, this volume. Additionally, Modern Chamoru shows numerous borrowings from Amerindian languages since contact with Spain almost always took place via the Americas or the Philippines. Thus, many, if not most ‘Spaniards’ on the Marianas were actually of Mexican or Filipino origin (Albalá Hernández 2000, 2001). As for Rapanui, there is no standardization for Modern Chamoru finding an overall acceptance. Counting here is based on the orthography of the author of the text. Johanson (e.g. 2002) emphasizes that employing the term copying instead of borrowing implies that there is always an alteration in meaning and/or function when an item is transferred from one language to another. Ergativity is defined by Dixon (1994: 1) as “a grammatical pattern in which the subject of an intransitive clause is treated the same way as the object of a transitive clause, and differently from transitive subject”. In contrast to nominative-accusative languages such as English or Spanish, Chamoru makes use of two different classes of personal pronouns, one in the absolutive and one in the ergative case, the latter of which is used in the future phrase. Note that Chamoru is a language with split ergativity (Cooreman 1987).

Comparative aspects of Hispanisation

195

References Albalá Hernández, Carmen P. 1994/ Americanismos en las Indias del Poniente. Voces de origen indígena 2000 americano en las lenguas del Pacífico. Madrid /Frankfurt a.M.: Iberoamericana Vervuert. 2001 Voces indígenas americanas en las lenguas del Pacífico. In Lo propio y lo ajeno en las lenguas austronésicas y amerindias, Klaus Zimmermann and Thomas Stolz (eds.), 279–294. Frankfurt a.M: Vervuert. Albalá Hernández, Carmen P. and Rafael Rodríguez-Ponga y Salamanca 1986 Relaciones de España con las Islas Marianas. La lengua chamorra. Madrid: Fundación Juan March. Bakker, Peter and Maarten Mous (eds.) 1994 Mixed Languages: 15 case studies in language intertwining. Amsterdam: IFOTT. Barthel, Thomas S. 1960 Rezitationen von der Osterinsel. Anthropos 55: 841–859. Blust, Robert 2000 Chamorro Historical Phonology. Oceanic Linguistics 39 (1): 83–122. Burrus, E. J. 1954 Sanvitores’ Grammar and Catechism in the Mariana Language. Anthropos 49: 934–960. Clyne, Michael (ed.) 1992 Pluricentric languages. Differing norms in different nations. Berlin / New York: Mouton de Gruyter. Cooreman, Ann M. 1987 Transitivity and discourse continuity in Chamorro narratives. Berlin / New York: Mouton de Gruyter Couto, Hildo H. do 1996 Introdução ao estudo das línguas crioulas e pidgins. Brasília: Editora Universidade de Brasília. Croft, William 2000 Explaining language change: an evolutionary approach. Harlow: Longman. Díaz, Norma, Ralph Ludwig, and Stefan Pfänder 2002 Procesos lingüísticos en situaciones de contacto: Parámetros y perspectivas. In La romania americana. Procesos lingüísticos en situaciones de contacto, Norma Díaz, Ralph Ludwig, and Stefan Pfänder (eds.), 389–441. Madrid /Frankfurt a.M.: Iberoamericana Vervuert. Dixon, Robert M. W. 1994 Ergativity. Cambridge: Cambridge University Press. Du Feu, Veronica 1996 Rapanui. (Routledge Descriptive Grammars.) London /New York: Routledge.

196 Steve Pagel Du Feu, Veronica and Steven R. Fischer 1993 The Rapanui language. In Easter Island studies: contributions to the history of Rapanui in memory of William T. Mulloy, Steven Roger Fischer (ed.), 165–68. Oyford: Oxbow Books. Fill, Alwin and Peter Mühlhäusler (eds.) 2001 The ecolinguistics reader: language, ecology and environment. London /New York: Continuum. Fischer, Steven R. 1993a The naming of Rapanui. In Easter Island Studies: contributions to the History of Rapanui in memory of William T. Mulloy, Steven Roger Fischer (ed.), 63–66. Oxford: Oxbow Books. 1993b Rapanui or Rapa Nui? Rapa Nui Journal 7 (4): 73–75. 1994 Rapanui’s great old words: E timo te akoako. The Journal of the Polynesian Society 103 (4): 413–443. 1997 Rongorongo: The Easter Island Script. History, Traditions, Texts. Oxford: University Press. 2001 Hispanicization in the Rapanui Language of Easter Island. In Lo propio y lo ajeno en las lenguas austronésicas y amerindias, Klaus Zimmermann and Thomas Stolz (eds.), 313–332. Frankfurt: Vervuert. 2005 Island at the end of the world: the turbulent history of Easter Island. London: Reaktion. this vol. Reversing Hispanisation on Rapa Nui (Easter Island). Grice, Paul 1968 Utterer’s meaning, sentence meaning, and word meaning. Foundations of linguistics 4: 225–242. 1975 Logic and conversation. In Speech acts. Syntax and semantics 3, Peter Cole and J. Morgan (eds.), 41–58. New York: Academic Press. Gómez Macker, Luis A. 1982 El bilingualismo en Isla de Pascua. Signos 14 (19): 90–99. Haspelmath, Martin 1999 Why is grammaticalization irreversible? Linguistics 37 (6): 1043– 1068. Heine, Bernd, Ulrike Claudi, and Friederike Hünnemeyer 1991 Grammaticalization: a conceptual framework. Chicago: University of Chicago Press. Heine, Bernd and Tania Kuteva 2002 World lexicon of grammaticalization. Cambridge: Cambridge University Press. Holm, John A. 2000 An introduction to Pidgins and Creoles. Cambridge: Cambridge University Press Hopper, Paul and Elizabeth Traugott 1993 Grammaticalization. Cambridge: Cambridge University Press.

Comparative aspects of Hispanisation

197

Johanson, Lars 2002 Contact-induced change in a code-copying framework. In Language change: the interplay of internal, external and extra-linguistic factors, Mari C. Jones and Edith Esch (eds.), 285–313. Berlin /New York: Mouton de Gruyter. Lefebvre, Claire 2004 Issues in the study of Pidgin and Creole languages. Amsterdam: Benjamins. Le Page, Robert B. and Andrée Tabouret-Keller 1985 Acts of identity: Creole-based approaches to language and ethnicity. New York: Cambridge University Press. Ludwig, Ralph 1995 Sprache als Kultursymbol. In Kulturelle Perspektiven auf Schrift und Schreibprozesse: Elf Aufsätze zum Thema Mündlichkeit und Schriftlichkeit, Wolfgang Raible (ed.), 187–214. Tübingen: Narr. 2002 Urbanidad, migración e hibridación de la lengua: procesos de contacto en el español de Santiago de Chile. In La romania americana. Procesos lingüísticos en situaciones de contacto, Norma Díaz, Ralph Ludwig and Stefan Pfänder (eds.), 357–386. Madrid /Frankfurt a.M.: Iberoamericana Vervuert. Makihara, Miki 1998 Rapanui-Spanish bilingual choice and code switching. In Easter Island in Pacific context, South Seas Symposium: Proceedings of the 4 th International Conference on Easter Island and East Polynesia, University of New Mexico at Albuquerque, 5.–10. August 1997, C. M. Stevenson, Georgia Lee, and Frank Morin (eds.), 33–38. Los Osos: Easter Island Foundation. 1999 Bilingualism, social change, and the politics of ethnicity on Rapanui (Easter Island), Chile. Dissertation, Yale University. 2001a Changing Rapanui language and identity. In Pacific 2000. Proceedings of the Fifth International Congress on Easter Island and the Pacific, C. M. Stevenson, Georgia Lee, and Frank Morin (eds.), 425– 428. Los Osos: Easter Island Foundation. 2001b Modern Rapanui adaption of Spanish elements. Oceanic Linguistics 40 (2): 191–224. 2001c Rapanui-Spanish Bilingualism. Rongorongo Studies 11 (1): 25–42. 2004 Linguistic syncretism and language ideologies: Transforming sociolinguistic hierarchy on Rapa Nui (Easter Island). American Anthropologist 106 (3): 529–540 2005 Rapa Nui ways of speaking Spanish: language shift and socialization on Easter Island. Language in Society 34: 727–767. Matras, Yaron and Peter Bakker 2003 The mixed language debate. Theoretical and empirical advances. Berlin /New York: Mouton de Gruyter.

198 Steve Pagel Maude, Harry E. 1981 Slavers in paradise: The Peruvian slavetrade in Polynesia, 1862– 1864. Stanford: Stanford University Press. McWhorter, John H. 1997 Towards a new model of Creole genesis. New York: Lang. 2005 Defining Creole. Oxford: Oxford University Press. Mondada, Lorenza 2007 Le code-switching comme ressource pour l’organisation de la paroleen-interaction. Journal of language contact 1: 168–199. Montrul, Silvina A. 2004 The acquisition of Spanish. Morphosyntactic development in monolingual and bilingual L1 acquisition and adult L2 acquisition. Amsterdam /Philadelphia: Benjamins. Mühlhäusler, Peter 1996 Linguistic ecology. Language change and imperialism in the Pacific region. London: Routledge. Mufwene, Salikoko S. 2001 The ecology of language evolution. Cambridge: Cambridge University Press. Munteanu, Dan 1997 Notas sobre el léxico de orígen español en chamorro. Anuario de lingüística hispánica 12/13 (2): 959–974. Muysken, Pieter 1994 Media Lengua. In Mixed languages. 15 case studies of language intertwining, Peter Bakker and Maarten Mous (eds.), 201–205. Amsterdam: IFOTT. 1997 Media Lengua. In Contact languages: a wider perspective, Sarah G. Thomason (ed.), 365–426. Amsterdam: Benjamins. Myers-Scotton, Carol 1993 Duelling languages. Grammatical structure in codeswitching. Oxford: Clarendon. Odo, Carol 1972 A survey of language use and attitudes on Guam. Agaña: Guam Department of Education. Oliver, Douglas L. 1951 The Pacific Islands. Cambridge: Harvard University Press. Onedera, Peter 1994 Fafa´ña´gue yan hinengge siha. Agaña, Guam. Oroz, Rodolfo 1966 La lengua castellana en Chile. Santiago: Facultad de Filosofía y Educación, Universidad de Chile. Pagel, Steve 2003 Zum Sprachkontakt Spanisch-Chamoru auf Guam und den Nördlichen Marianen. Master thesis, University of Halle.

Comparative aspects of Hispanisation 2005

199

Rapa Nui and the Marianas: approaches to a comparative analysis. In The Reñaca Papers. VI International Conference on Easter Island and the Pacific, Christopher M. Stevenson, José Miguel Ramírez Aliaga, Francis J. Morin, and Norma Barbacci (eds.). Los Osos: Easter Island Foundation. Pagel, Steve and Stefan Pfänder 2001 Französisch und Spanisch als Input im Sprachkontakt. Grammatikalisierung, ‘Grundkonzepte’ und emergent grammar. Sprachkontakt in der Romania. Arbeitspapiere zur Linguistik 36: 71–109. Pfänder, Stefan 2004 Creating Latin American Spanish. Language contact and language change in Bolivia. Habilitation, University of Halle. Pigafetta, Antonio 1994 Magellan’s voyage. A narrative account of the first circumnavigation. New York: Dover Publications. Rabanales, Ambrosio 1992 El Español de Chile: situación actual. In Historia y presente del español de América, Cesar Hernández Alonso (ed.), 565–592. Valladolid: Junta de Castilla y León. Rodríguez-Ponga y Salamanca, Rafael 1995 El elemento español en la lengua chamorra (Islas Marianas). Tesis doctoral, Universidad Complutense de Madrid. 1996 Formas de ‘ser’ y ‘estar’ en Chamorro. In 1996 Actas del Cuarto Congreso de Hispanistas de Asia, 49–55. Seoul: Asosiacíon Asiática de Hispanistas. 1998 Las preposiciones hispano-chamorras. In La Torre. Revista de la Universidad de Puerto Rico, 511–522. Tercera Època. 2001 Los numerales hispano-chamorros. In Lo propio y lo ajeno en las lenguas austronésicas y amerindias, Klaus Zimmermann and Thomas Stolz (eds.), 253–278. Frankfurt: Vervuert. Rogers, Robert F. 1995 Destiny’s landfall. A history of Guam. Honolulu: University of Hawai’i Press. Ross, Malcolm D. 1999 Exploring metatypy: how does contact-induced typological change come about? (Keynote talk, Annual meeting of the Australian Linguistic Society, Perth). http://rspas.anu.edu.au/linguistics/projects/mdr/Metatypy.pdf Salas Palomo, Rosa and Thomas Stolz this vol. Pro or contra Hispanisms: attitudes of native speakers of modern Chamoru. Sankoff, David, Shana Poplack, and Swathi Vanniarajan 1986 The case of the nonce loan in Tamil. Language variation and change 2 (1): 71–101.

200 Steve Pagel Schuchardt, Hugo 1884 Dem Herrn Franz von Miklosich zum 20. November 1883: Slawodeutsches und Slawo-italienisches. Graz: Leuschner & Lubensky. Smith, Geoff P. 2002 Growing up with Tok Pisin. Contact, creolization and change in Papua New Guinea’s national language. London: Battlebridge. Stolz, Christel and Thomas Stolz 1996 Funktionswortentlehnung in Mesoamerika: Spanisch-amerindischer Sprachkontakt (Hispanoindiana II). STUF 49 (1): 86–123. 1997 Universelle Hispanismen? Von Manila über Lima bis Mexiko und zurück: Muster bei der Entlehnung spanischer Funtionswörter in die indigenen Sprachen Amerikas und Austronesiens. Orbis 39: 1–77. Stolz, Thomas 1998 Die Hispanitaet des Chamoru als sprachwissenschaftliches Problem. Iberoamericana 2 (70): 5–38. 2003 Not quite the right mixture: Chamorro and Malti as candidates for the status of mixed language. In The mixed language debate. Theoretical and empirical advances, Yaron Matras and Peter Bakker (eds.), 271–315. Berlin /New York: de Gruyter. Stolz, Thomas and Anna Sabater Fuentes 2002 El uso del artículo indeterminado en la primera traducción chamorra del nuevo testamento. In Traspasando fronteras: el reto de Asia y el Pacífico (II), Francisco Javier Antón Burgos, and Luis Oscar Ramos Alonso (eds.), 731–739. Asociación Española de Estudios del Pacífico, Centro de Estudios de Asia, Universidad de Valladolid. Thornton, Agathe 1998 Do A and O categories of ‘possession’ in Maori express degrees of Tapu? The Journal of the Polynesian Society 107 (4): 381–393. Thomason, Sarah G. 1997 A typology of contact languages. In The structure and status of Pidgins and Creoles, Arthur Spears and Donald Winford (eds.). Amsterdam: Benjamins. 2001 Contact-induced typological change. In Sprachtypologie und sprachliche Universalien. Handbuch Sprache und Kommunikation, vol. XX, Martin Haspelmath, Ekkehard König, Wulf Oesterreicher, and Wolfgang Raible (eds.), 1640–1648. Berlin /New York: Mouton de Gruyter. Thomason, Sarah G. and Terrence Kaufman 1988 Language contact, creolization, and genetic linguistics. Berkeley: University of California Press. Topping, Donald M., with the assistance of Bernadita C. Dungca 1973 Chamorro reference grammar. Honolulu: University of Hawai’i Press. Topping, Donald M., Pedro M. Ogo, and Bernadita C. Dungca 1975 Chamorro-English dictionary. Honolulu: University of Hawai’i Press.

Comparative aspects of Hispanisation

201

Underwood, Robert A. 1987 Language survival, the ideology of English and education on Guam. In Chamorro language issues and research on Guam, Mary L. Spencer (ed.). Mangilao: University of Guam. U.S. Census Bureau 2004 2000 Census of population and housing. Guam. http://www.census. gov/prod/cen2000/island/GUAMprofile.pdf Weber, Robert, and Nancy Thiesen de Weber 1990 ¿Podrá sobrevivir el idioma Rapa Nui? Revista Signos, Estudios de Lengua y Literatura 23 (8): 119–136. 1998 Diagnóstico lingüístico de los educandos de Kinder a séptimo año básico de la escuela ‘Lorenzo Baenza Vega’ de Isla de Pascua al final del año escolar 1997. Valparaíso: Programa Lengua Rapa Nui, Universidad Católica de Valparaíso y SIL International. Winford, Donald 2003 An introduction to contact linguistics. Malden: Blackwell.

Hispanisation processes in the Philippines Patrick O. Steinkrüger

1. Introduction Beginning in the first half of the 16th century and ending by the end of the 19th century, Spanish rule in the Philippines made an important impact on the linguistic situation of the archipelago which is still present today. Like other Austronesian-speaking areas and China, the Philippines were connected to Mexico for around 300 years – a fact which is lexically reflected in nearly all linguistic systems of the Philippines – and with the Portuguese colonies in Asia, especially Macau, Malacca and India, but also Moluccas. This last fact accounts for lexical, but also structural similarities, between the Portuguese-based Creoles in Asia and Chabacano, the Spanish-based creole of the Philippines relexified in the 16th or 17th century. During after these more than three centuries of Spanish colonialization, several contact situations and languages influenced by Spanish emerged in the Philippines: 1. Contact-induced influence: lexical & structural Hispanisms in most Philippine languages 2. Creolization: the varieties of Chabacano in the Bay of Manila (Caviteño and Ternateño) and the island of Mindanao (Zamboangueño and its layers) 3. Language mixing: at least Ilokano and Spanish mixed as HispanoIlokano (†) 4. Pidginization: “Bamboo Spanish” and “Español de Cocina” 5. Philippine-Spanish: Spanish as a mother-tongue of a small Filipino minority (in the past and today). In the literature, there is quite frequently confusion about these different contact-induced systems, especially referring to Spanish and Chabacano. Disregarding the structural differences between them, in all these systems pan-Philippine Hispanisms are found at the lexical level, like sigúro ‘probably’ or sakáte ‘grass’ (from Nahuatl via Mexican Spanish). Concerning the current circumstances and comparing it with the Spanish period we can see that the sociolinguistic situation has dramatically changed: Spanish-

204 Patrick O. Steinkrüger Philippine bilinguism, a marginal fact even in the past, does not exist any more, except for a very small Spanish speaking minority in some places (cf. 4.3). However multilinguism and frequent code switching has been increasing, especially with English and Tagalog in urban areas like Metro Manila (for Taglish see Thompson 2003). Code switching, as in example (1), is quite widespread among young speakers and is different from the high frequency of loans in an average sentence in one of the major languages of the Philippines (cf. example 2): (1)

CHABACANO ZAMBOANGUEÑO1 Dale kumígo ditúyo cellphone number para I can call you later. give I.OBJ POSS.2SG cellphone number so-that I can call you later ‘Give me your cellphone number so that I can call you later.’

(2)

CEBUANO Pwéde ka mag-translate para sa áko? can you VRBLZ-translate for LOC I ‘Can you translate (something) for me?’

This essay will focus on the structural Hispanisation of Philippine languages, but also gives information on the other contact results which have been mentioned in the short list above. This attempt is far from being exhaustive, but it gives at least an idea of the linguistic consequences of the Spanish colonization of the Philippines, which have not been in the centre of interest either in Romance or in Austronesian studies. The reader who is more interested in external factors like education and socio-historical background, can refer to Lipski et al. (1996: 271–280) who describe the basic facts. 2. The Hispanic influence on Philippine languages Unfortunately, there has been little work done on the structural influence of Spanish on the Philippine languages. Most work is concerned with lexical items (see e.g. Quilis 1973 and other publications by the same author) since “[t]his influence [i.e. of Spanish on Philippine languages; P.S.] is most visible in its lexicon and phonology.” (Himmelmann 2005a: 350). Romance studies generally shows little interest in this field. Studies from an Austronesian perspective also have little information on the Spanish impact in the grammars of Philippine languages (cf. the essays in Adelaard and Himmel-

Hispanisation processes in the Philippines

205

mann 2005). Most literature deals with the Hispanic influence on the two major languages Tagalog and Cebuano, so we know nearly nothing about its influence on the minor languages. In addition, there is still a great lack of studies treating its syntactical and discursive influence on Philippine languages. 2.1. Writing system First of all, the old Philippine writing tradition had disappeared by the 16th century (except Palawan) and was substituted by the Latin alphabet used by Spanish missionaries. The Baybayin-scripture was of Indic origin and, consequently, a syllabic system (see Constantino 1978: 115). Interestingly, the system only possessed 3 symbols representing vowels, which is additional evidence for the assumption that the vowels /e/ and /o/ were introduced to the Phillipines by the Spanish. 2.2. Phonetics and phonology 2.2.1. Inventory It is nearly uncontested that after the Spanish rule, two new vowels had been introduced into the inventory of most Philippine languages. Usually, they occur in Spanish loans and in non-accentuated syllables which frequently become neutralized: [e]~[i]; [o]~[u] (a process which also can be observed in Chabacano). In non-Spanish words, they are treated as allophones, or there is at least no evidence for minimal pairs. For one of the two major languages of the Philippines, Cebuano there is no evidence that they are even integrated: “The underlying vowel system in Cebuano has three phonemes, /i, a, u/. When Spanish loan words are considered, the phonemes /e, o/ occur as members of the system.” (Bunye and Yap 1971: 8). Examples from Cebuano (Bunye and Yap 1971: 9): (3)

CEBUANO STRESSED /E/ AND /O/ /e/ ékstra ‘extra’ pétsa ‘date’ /o/ pósporo ‘match’

létse ‘milk’

In the case of Tagalog, this fact was contested by several scholars (cf. Constantino 1978: 128 fn. 27 and 141–142). He writes as follows:

206 Patrick O. Steinkrüger It is now generally accepted that Tag. e and i and o and u are separate phonemes. The vowels e and o occur not only in borrowed words but also in native words. For example, e and i contrast in the following native words: hé:to ‘here it is’ and hí:toq ‘catfish’, é:wan ‘I don’t know’ and í:wan ‘leave’. On the other hand, o and u contrast in the following forms: á:patnapúq ‘forty’ and á:patnapóq ‘four already sir’, isú:si mo ‘look it’ and isó:li mo ‘return it’. (Constantino 1978: 128 fn. 27)

One argument which weakens his statement is the fact that Constantino contrasts lexemes with phrases and not with other lexemes. Furthermore, the old Tagalog writing system with only three vocalic symbols is also strong evidence for the former existence of only three vowels. Surprisingly, there are minimal pairs of Spanish loans in Philippine languages: Wolfenden (1971: 14) reports that Hiligaynon, a language of Central Philippines, contrasts pari ‘priest’ and pare ‘compadre (godfather)’, both from Spanish padre. Not only in Tagalog and Cebuano, but also in Ilokano, the third largest language of the Philippines, new vowels from Spanish had been introduced. Unlike Cebuano, Ilokano possessed four native vowels, namely /i, e, u/ and /a/. A new opposition between /e/ and / / arose which is neutralized in some varieties (Rubino 2005: 328): “The new phonemes /o/ and / / are found in only in loanwords. In the northern dialects, the phoneme /e/ is pronounced as / /.” Apart from vowels, another controversy is the phonemic status of several affricates which occur as palatalized /t, d and s/ after the glide /y/ (Rubino 2005: 328): “Because of many borrowings from English, Spanish, and colloquial Tagalog where these palatal sounds are not complex phoneme sequences, the phonemic status of [tS], [Z] and [S] is open to debate.” 2.2.2. Phonotactics and resyllabification Tagalog and other Philippine languages tend to have open syllables without consonant clusters. Initial consonantal clusters and closed syllables (#CCV(C)) only occur in Spanish loans. All the following examples in Cebuano are taken from Bunye and Yap (1971: 6–7) in Cebuano: (4)

CEBUANO: syllable structure /CCV/ kwélyo ‘collar’, kwárto ‘room’, pyáno ‘piano’ /CCVC/ pwérsa ‘force’, dwénde ‘dwarf’

Hispanisation processes in the Philippines

(5)

207

CEBUANO: initial clusters (examples) /pw/ pwéra ‘go, leave’ pwérto ‘opening’ pwéde ‘okay’ /pr/ príto ‘fry’ prísyo ‘price’ priméro ‘first’ /bw/ bwérta ‘go back’ (< Span. vuelta ‘turn’) bwéno ‘well’ /ts/ tséke ‘check’ tsa ‘tea’ tsinélas ‘slippers’

However, there are also many cases of resyllabification: brazo ‘arm’ > ba.ra.so in the popular (monolingual) pronunciation of Filipinos (López 1965: 475). 2.2.3. Stress patterns of Hispanic loans In Cebuano, as in many Philippine languages, stress is on the ultima or paenultima (Bunye and Yap 1971: 11). Since Spanish has introduced many loans with an original stress on the antepaenultima, Philippine speakers must have had to handle this prosodic deviation. López (1965: 476) claims that “[t]he stress of S[panish] forms reproduced in T[agalog] cannot be predicted.” It seems that López’ statement is correct since loans from Spanish with an original accent on the antepeanultima show alternation: (6)

CEBUANO & CHABACANO (stress in Spanish loans) teléfono > teleponó ‘telephone’ (CEBUANO) página > pahína ‘page’ (CEBUANO) páharo/paháro/páhro ‘bird’ (CHABACANO)

2.2.4. Phonological “inconsistencies” of Spanish words in the Philippines Historically, it is interesting that loans from Spanish show different behaviour. This is quite logical if we look at the different periods of Hispanisation represented by at least two different forms of Spanish introduced in the Philippines: An older form of Spanish in the 16th and 17th century intro-

208 Patrick O. Steinkrüger duced by early colonialization (spoken by soldiers and missionaries) and a modern one used in school and the print media during the 19th and 20th century. This is the reason why we have old forms which show similarities with Ternateño – the oldest Spanish-based creole spoken in the Bay of Manila – but also Spanish spoken by Filipinos and other forms which represent a kind of “spelling pronunciation” unknown to those older varieties. The variation or “inconsistencies” concern at the same time both the following Spanish sounds and morphemes: 1. -ado vs. -aw 2. -ly- vs. -y3. h- vs. sWhereas Philippine Spanish and all varieties of Chabacano realize Spanish -ado as [-'aw], Philippine Hispanisms have [-'ado]: (7)

CEBUANO kumplikádo ‘complicate’ sigurádo ‘sure’

Sometimes the intervocalic // is realized as /j/ (so called yeísmo) in Cebuano words, and sometimes not: (8)

CEBUANO botílya < kutsílyo < kasílyas < but: sibúyas < kabáyo
Tagalog/Cebuano kumusta (ka)?

Another morphological feature of Spanish verbs which occurs in Tagalog is the deletion of the final -r: apurá ‘hurry’, sará ‘close’, intendé ‘understand’ etc. (López 1965: 477). This is a feature also found in Chabacano verbs which have a Spanish antecedent. It could argued that an already restructured form of Spanish (Chabacano or a pidgin) was the origin of these items. Usually, the final /-r/ of Spanish origin is reflected as /-l/: Tag. kasál < Span. casar ‘to marry’ or asúkal < azúcar ‘sugar’. 2.3.2. Morphological integration of Spanish lexemes How deeply Hispanic words are integrated in Philippine languages is shown by the following cases where they are subject to different morphological processes:

210 Patrick O. Steinkrüger a. Partial reduplication (from López 1965: 493): (10) TAGALOG dusédusena ‘by the dozen’ < Span. dozena ‘dozen’ parépareho ‘all equally alike’ < Span. parejo ‘equal’ b. Spanish roots with Philippine affixes: (11) CEBUANO tienda ‘shop’+ -han (LOC) tindáhan ‘shop, grocery’

entráhan ‘entrance’ entra- ‘enter’+ -han (LOC) 2.3.3. Derivational suffixes There are some Spanish derivational affixes with Philippine roots; examples from López include (1965: 489–490): (12) TAGALOG dalága ‘maid’

dalagíta ‘teenager’ kabiláq ‘other side’

kabilóso,-a ‘fickle’ lamón ‘voracity, greediness’ lamonéro,-a ‘voracious, greedy’ (13) CEBUANO & TAGALOG babáe ‘woman’

baba-éro ‘womanizer, philanderer’

2.3.4. Inflectional affixes Following Bowen (1971: 946–947), morphological gender markers -o/-a have been introduced and even the consequent gender agreement within NPs is necessary. López (1965: 503) points out that, “[m]orphological constructions have introduced a distinction of gender, masculine and feminine, based on sex of animate beings, which was not recognized before in Tagalog”. The following examples are from Cebuano: (14) CEBUANO lóla, lólo ‘grandmother, grandfather’ gwápo, gwápa ‘nice’ amígo, amíga ‘friend’ However, following John Wolff (p.c.), many native speakers of Philippine languages do use these forms without any distinction.

Hispanisation processes in the Philippines

211

2.4. Morphosyntax Spanish has exerted less influence on the syntactical structure of Philippine languages, including Tagalog. López (1965), whose examples nearly all refer to morphological features, may overrate the typological similarities of both languages. He concludes as follows: “Because of the similarity in function of grammatical categories, S[panish] forms are integrated in the system of T[agalog] syntax and construction of composition” (Lopez 1965: 491). Himmelmann holds the early normative influence of Spanish responsible for the Hispanic grammatical features in Tagalog; he states: As for morphosyntax, it is not unlikely that there has been some normative influence by colonial institutions and practices because the Spanish started using Tagalog very early on as a missionary language, writing grammars and preparing catechisms in Tagalog (a Doctrina Christiana, dated 1593, is the oldest preserved Tagalog document). (Himmelmann 2005b: 1475)

Clear influences of Spanish can be observed in the morphosyntax of comparison, some modal adverbials and conjunctions. 2.4.1. The morphosyntax of comparison For the purpose of comparison, the Spanish elements más ‘more’ and paréjo ‘equal’ were introduced as shown in the examples (15) and (16): (15) CEBUANO (Bunye and Yap 1971: 23) Mas asúl ang lápis kay sa papél. more blue DET pencil as LOC paper ‘The pencil is bluer than the paper.’ (16) CEBUANO (Bunye and Yap 1971: 21) parího maítum equal black ‘as black as’ 2.4.2. Modals Cebuano, for example, has introduced the Spanish modals gústo ‘want, will’ (< gústo ‘I like’; cf. example 17) and pwéde ‘possibly’ (< puede ‘s/he can’). In addition to pwéde there is still the modal prefix n/maka- is still

212 Patrick O. Steinkrüger available. In Tagalog, there is also the adverb maaári ‘it is possible’. As Stolz and Stolz (1997: 37–38) point out, it is remarkable that the Hispanic modal puwéde in Tagalog has not substituted former categories but coexists with the Austronesian element. A detailed study of the distribution of both strategies must still be done. 2.4.3. Other morphosyntactical influences Bowen (1971: 947) claims that Spanish has influenced the application of categories for certain number distinctions in Tagalog. Following him, the optionality of the prenominal particle mangá (literally ‘some; approximately’) has been diminished and mangá has become a grammaticalized plural marker. Secondly, the paradigm of pronouns has been changed because of the reduced importance of the pronoun dual forms kata, nata and kanita which have nearly disappeared from spoken language. There are a lot of functional words of Spanish origin, especially connectives and prepositions, a fact which has been already shown by Stolz and Stolz (1997: 36–38) giving examples from the North-Philippine languages Tagalog and Bikolana. Examples for the conjunction pero ‘but’ and the preposition para ‘for’ which takes NPs and VPs as in Spanish are very frequent. The connective pero also exists in Cebuano (see example 17), but as in the case of puede (cf. 2.4.2), the Spanish element coexists with the Austronesian apan which has the same function (see Bunye and Yap 1971: 62). (17) CEBUANO (adapted from Bunye & Yap 1971: 62) Gústo ko nímo, pero póbre man ka. like I you:DO but poor PART you ‘I like you, but you’re poor.’ Wolff (2001: 248) found that there is a preference of analytic constructions in extreme Hispanised forms of Cebuano formerly used by SpanishCebuano bilinguals. Instead of the affixed verb, a Spanish verbal form is used: (18) CEBUANO (from Wolff 2001: 248; fn. 4) Sígi siya ang ínum. continued he LNK drink ‘He drank and drank.’

Hispanisation processes in the Philippines

213

Surprisingly, none of the numerous discourse particles in Tagalog are of Spanish origin. This is unlike the case of the Americas or the Marianas, where Spanish discourse particles are widespread: Table 1. Wackernagel particles in Tagalog (from Anderson [adapted from Schachter] to appear, p.3) ba kasi kaya daw din ho lamang man muna

(interrogative) ‘because’ (speculation) (reported speech) ‘too’ (politeness) ‘only’ ‘even’ ‘for a while’

na naman nga pa pala po sana tuloy yata

‘already’ ‘instead’ ‘really’ ‘still’ (surprise) (politeness) (optative) ‘as a result’ (uncertainty)

2.5. Lexicon We can say that around 20 % of the vocabulary, depending on the textsorte, in a Tagalog text, and probably also for any other Philippine language which had been in contact with Spanish, is of Spanish origin. In the case of Tagalog, this influence was very strong since the Spanish chose Manila as the capital of the Philippines. Manila is the centre of the Tagalog-speaking culture of the Archipelago. Interestingly, Hispanic elements are nearly unknown in languages without intensive contact, like those of the mountain tribes, the so called negritos. This is the case for Aita Mag-anchi, spoken in Central Luzon (see Kitano and Pangilinan 2003). However even in geographically more peripheral languages like Sama in Southern Mindanao, Spanish influence on its lexicon is “heavy” (statement by Jun 2005: 377). A series of studies have been done of cultural items from Spanish in the major Philippine languages Tagalog and Cebuano (see e.g. Barón Castro 1972, Quilis 1973, 1976). However linguistic analysis must go beyond the listing of all the new concepts introduced by the Spanish with their corresponding words. Wolff (2001: 248) rightly states: Clearly the cultural borrowings from Spanish confirm what we know about the areas of Filipino life which were affected by the Spanish conquest. More interesting are forms which deviate from the Spanish original in semantics and in grammatical or phonological features. These deviations reveal the extent to Spanish concepts were not taken over but reinterpreted into a Filipino understanding of the world.

214 Patrick O. Steinkrüger There are examples like e.g. Cebuano pwéra ‘go’ < Spanish fuera ‘out’ (from Bunye and Yap 1971: 10). We can also generalize from other cases that most semantic change took place by inference or metonomy. Other cases from Tagalog include: óras ‘time’ < ‘hours’ báryo ‘village’ < barrio ‘quarter’ or tirá ‘rest’ < tirar ‘throw away’ and from Hiligaynon mánu ‘right’ < mano ‘hand’. There are also expressions from colloquial speech which in Philippine languages are stylisticly unmarked but in Spanish, e.g. péra ‘money’ < pela ‘Peseta (coll.)’. Phonological adaptation of Spanish words is discussed in section 2.2. How important the relationship to the Americas was, is shown by Albalá (2003) who analyzes words from Amerindian origin like, e.g. tianguis ‘market’ from Nahuatl. It is also important to say that Hispanic lexemes are still present in modern Filipino society and are not restricted to the historical past. Alcántara’s study (1996) focuses on the Hispanisms in modern communication in the Philippines, and Thompson (2003) shows the Spanish presence in Taglish, the code-switched or already intertwined form of English and Tagalog used in and around Metro Manila. There are not only items for juridical or political terms in the modern language, but also for cursing and other exclamations. Spanish is one of the sources of expressions like: satanás ‘Satan’, púta ‘whore’, karámba ‘jar’, bastós! ‘What a cheek!’ (< basto ‘coarse’) and susmaryosep (< Jesús, María, Josef), an exclamation of surprise (cf. also Zorc and San Miguel 1991). Synchronically, it is not correct to call them all “Spanish loan words”, since the native speakers of Philippine languages accept them without identifying them as “Spanish”.

3. Creolisation The Spanish-based Creoles in the Philippines came into being after the Spanish conquest in the 16th century and are commonly known as Chabacano. Despite its despicable original meaning (‘tasteless’, ‘vulgar’), this name is used by the speech-community as a self-designation. A Portuguesebased creole was certainly the forerunner of Chabacano which was relexified during the 16th or 17th century. Chabacano is one of the three Spanishbased Creoles worldwide, alongside Papiamento, which is also a former Portuguese-based creole, and Palenquero; which is socio-historically an exceptional case. It shares interesting features with other European-based Creoles in Asia; a fact differentiating it from the Atlantic creoles. In addition, unlike the latter, Chabacano has been acquiring a mixed character; especially from the 20th century onwards. The total number of Chabacano

Hispanisation processes in the Philippines

215

speakers can be estimated at no more than 500,000. Geolinguistically, Chabacano can be divided into two main varieties: Manila Bay Creole (MBC) and Southern Mindanao Creole (SMC). Dialects of the former are found on the northern island of Luzon and are spoken by relatively small communities in the towns of Cavite and Ternate. The dialect of Ermita, a district of Manila, can now be regarded as extinct. SMC is spoken on the island Mindanao in the south of the archipelago. The highest degree of vitality of SMC, and of Chabacano generally, is observed in the city of Zamboanga and the surrounding area in the extreme western part of Mindanao (approximately 300,000 speakers). The creole is also used here as a lingua franca by diverse ethno-linguistic groups. It is codified for public and private purposes with a Hispanic orthography; there are Bible translations, literary writings, and until recently press publications. Public oral use has recently been increasing in broadcasting, in church, and in some public schools. A subvariety of Zamboangueño is also spoken in the town of Cotabato and there is a small speech community of SMC in Davao. In addition, there are also reports of a speech community of Zamboangueño in one village near Semporna (Malaysia, Sabah). With respect to the structural traits of Chabacano, many features are due to contact with Austronesian or, more exactly Malayo-Polynesian, in particular its lexicon and morphological inventory of Zamboangueño. For example, the basic word order in Chabacano is VSO (cf. also 3.3.1), except in some Indo-Portuguese Creoles where we find SOV basic word order, this is a marked case in comparison to other European-based Creoles. As for the lexicon, the major input from Philippine languages, apart from Spanish, in the case of MBC, is from Tagalog. Zamboangueño has also adopted much from Ilonggo (the proper name is Hiligaynon) which was formerly used as a lingua franca in the region (following Frake 1971; also Lipski 1992) and also Cebuano, the major Western Visayan Language in the Southern Philippines, and Tagalog (Pilipino) which has only recently started to influence the Philippine languages all over the Islands. Besides Spanish and Hiligaynon, national, regional and different local contact languages have influenced the lexicon of Zamboangueño. The local languages Tausug, Subanon and Yakan, despite their close geographical contact with Chabacano, have had very little influence on the Chabacano lexicon. In her analysis of a basic vocabulary of more than 6000 entries, Riego de Dios (1989: 188) states that 83% of the items are of Spanish, ca. 15% of Philippine and 2,5% of English origin. More than 30 years after her analysis, it can be assumed that English influence has increased and Spanish influence has diminished. Very few words are of Portuguese origin, but these are sometimes structurally important:

216 Patrick O. Steinkrüger e.g. kiláya ‘how’ < Port. que laia ‘what kind’, ele < Port ele ‘s/he’ and na, a locative preposition < Port. na, a merger of the preposition em and the feminine singular article a. The importance of Mexican input is shown by words like angge ‘market’ < tiangue (a Mexican Spanish word of Nahuatl origin). The current structure of the language, especially concerning the vocabulary and word structure, is heavily influenced by Philippine languages. The sound system shows Philippine features like the merging of [o]~[u] and [e]~[i] in unstressed syllables and the occurrence of the glottal stop. Some of the productive Zamboangueño word formation morphemes originally come from Spanish, and many are from Hiligaynon or Cebuano. Frequently occurring derivational affixes are the verbalizing prefix man-: kwénto ‘story’ man-kwénto ‘to tell’ and the adjectivising prefix ma-: pwérte ‘strength’ ma-pwérte ‘strong’. Their forms are often identical with those of the source languages, but in many cases they have different functions. The word formation system seems to be unusually complex for a Creole (see Steinkrüger 2003 for details). In Zamboangueño, first person plural pronouns have a Philippine form encoding inclusion or exclusion of the addressee. As in Philippine languages, human proper names in subject function are marked with si: ta-kantá si Maria ‘Maria is singing’. Plurals of nouns are frequently marked with the particle mga [maNa]: el mga péhro ‘the dogs’. Both MBC and SMC mark primary objects with kon, that is, direct objects and recipients of ditransitive constructions are treated in the same way; human direct objects are obligatorily marked with kon: konosé le kon ese muhér ‘he knows that women’ (see examples in section 3.2.2). There are many discourse particles of Philippine origin, e.g. daw for reported information (cf. Table 1 and section 3.3). However, Chabacano also has features typical of Atlantic creoles, e.g. no morpho-syntactic passive construction, lack of an equative copula (soltéro el hénte ‘the man is a bachelor’), identity of ‘have’ and ‘there is’ (both tyéne), preverbal markers for tense, mood and aspect (e.g. ya-andá yo na Samboángga ‘I went to Zamboanga’, ta-kusiná le ‘(s)he is cooking’ and ay-kantá silá ‘they will sing’). In the following, some of the above mentioned structural characteristics will be discussed in detail.

Hispanisation processes in the Philippines

3.1.

217

Morphology

3.1.1. Word-formation It is characteristic of Zamboangueño, but not of the MBC-varieties, to have a lot of productive affixes which are obviously from Philippine origin. This may have created the impression that language intertwining of Spanish and certain Philippine languages was the background of the genesis of Zamboangueño (for more details see Steinkrüger 2003 and 2004). Besides these, there are still a few affixes of Spanish origin, but whose application is restricted to Spanish roots, e.g. -da (a nominalizer as in eskribida ‘writing/s)’, -ante (indicating nomina agentis like in trabahánte ‘worker’) or -ón (like in olbidón ‘forgetful person’). The situation is sometimes complicated by the fact that Spanish morphemes and lexemes have been introduced into Chabacano via Philippine languages where these items are still productive and frequent. The following table lists the most frequent affixes found in Zamboangueño. Nearly all of them are of Hiligaynon or Cebuano origin: Table 2. Derivational affixes of Philippine origin in Zamboangueño ikamamakamanmanmagpapaka-han, -an -hán, -án -hin, -in -un

ordinal numbers adjectives (of abundance) causative verbalizer collectivity (relationship) collectivity (relationship) causative, direction nominalizer (abstraction) reciprocal verb location nominalizer for persons nomina agentis

ika-dós ‘second’ ma-pyédra ‘full of stones’ maka-tríste ‘make sad’ man-kwénto ‘tell (a story)’ man- ermána-y-ermáno ‘group of brothers’ mag-ermáno ‘group of brothers’ man-pa-balábak ‘bend, buckle’ paka-alísto ‘gift, talent’ man-kwentó-han ‘tell each other’ tubu-hán ‘sugarcane plantation’ myedú-hin ‘anxious person’ bigáqun ‘person who flirts a lot’

3.1.2. TMA-markers Like many other Iberoromance-based creoles around the world, especially Portuguese-based ones in Asia, Chabacano varieties possess three preverbal markers indicating tense, aspect and mood (see Table 3). They are all of Spanish and/or Portuguese origin and their morphonological status is that of affixes. The main difference between MBC and SMC lies in the different

218 Patrick O. Steinkrüger marking of the irrealis, that is di- for the first and ay- for the latter varieties. Ternateño even shows tendencies to fuse the markers with the verbs, frequently with andá ‘go’ and asé ‘make ’ (cf. example 23). Table 3. The inventory of TMA-markers in Chabacano Marker

Category

Etymology

(y)aay-/deta-

perfect(ive) irrealis imperfective

Span. ya ‘already’ Span. hay de ‘one must/has to’ and/or debe de ‘id.’ Span. está COP.3SG.PRES.IND

Examples of application of the TMA-markers in Chabacano: (19) ZAMBOANGUEÑO Ya-uyí yo ta-lyurá el dalagíta. PRF-hear I IMPFV-cry DET girl ‘I heard the girl crying.’ (20) CAVITEÑO Di-andá yo na plása. IRR-go I LOC market(-place) ‘I will go to the market.’ (21) ZAMBOANGUEÑO El péhro ta-buská konel palakáq. DET dog IMPFV-seak OBJ.DET frog ‘The dog is looking for the frog.’ (22) TERNATEÑO Ta-pegá éli kung kel péhro. IMPFV-hit s/he OBJ DET dog ‘S/he is hitting the dog.’ (23) TERNATEÑO indá ‘go, come’ (< Span. andar ‘walk’) a + indá andá: andá yo ‘I went’ di + indá dindá: dindá yo ‘I will/would go’ ta + andá tandá: tandá yo ‘I am going/use to go’

Hispanisation processes in the Philippines

219

3.2. Personal pronouns The internal and paradigmatic structure of Chabacano personal pronouns reflect the historical connection of the Philippines with other regions in the world, especially with Portuguese possessions in Asia. The pronouns also demonstrate that a clear-cut distinction between Portuguese and Spanish during the 16th and the 17th century2 was difficult to establish by non-native speakers, at least for their restructured layers. As an example, vosotros ‘you (PL)’ did not exist in Portuguese but it is obviously the protoform of this pronoun in the Portuguese-based creoles in Malacca, Macau and India. The subject pronouns of the following Iberoromance-based creoles will be compared in the following table (in exactly in this order): T = Ternateño, C = Caviteño, M = Macaísta (Santos Ferreira 1978: 23), PK = Papia Kristang (Baxter 1988: 52–53), IP = Indo-Portuguese (Norteiro) (Dalgado 1906: 154– 155), PP = Papiamentu (Munteanu 1996: 295), P = Palenquero (Pérez Tejedor 2004: 56) and Z = Zamboangueño: Table 4. Free subject pronouns in Iberoromance-based creoles SG

T yo bo, tédi éli

C yo vos éle

M iou vôs

PK yo bos eli

IP eu ós éll(a)

PP mi bo ele

P yo bo ele

Z (i)yo (e)bós, tu, ustéd Éle

T mihótro

C M nisós nôs

PK nus

IP nós

PP nos

P suto

Z kitá (incl.) kamé (excl.)

tédi(s) lohótro

vusós vosôtro bolotu3 usôt ilós ilôtro olotu illôt

PL

boso enú (polite), utére kamó nan ané silá

In the table only full forms are listed, not weak nor clitisized forms. With the exceptions of Palenquero and Zamboangueño, where the whole set of the plural is taken from Niger-Kongo and Malayo-Polynesian respectively, we can assume that the proto-paradigm of all Iberoromance-based creoles was the following: yo/mi ‘I’, vos ‘you’, ele ‘s/he’, nosotros/misotros ‘we’, vosotros ‘you (PL)’ and (e)lesotros ‘they’.

220 Patrick O. Steinkrüger 3.3.

Syntax

To give some examples, two typical properties of Chabacano syntax are discussed in this paragraph. The first phenomenon shows how “Philippine” the syntax of Chabacano is. The second phenomenon is not only typical for Chabacano, but also for Portuguese-based creoles in Asia, as for example Papia Kristang in Malacca, the creole in Macau or the former creole of Batavia/Jakarta. 3.3.1. VSO and P2-movement As in Philippine languages, VSO is the unmarked order within the sentence (see example 22). However if certain elements are inserted into the initial position of the sentence the subject moves into the second position. These elements are negators, temporal adverbs etc. Examples (25) and (26) demonstrate this behaviour in Chabacano: (24) ZAMBOANGUEÑO Ta-abahá yo na eskaléra. IMPFV-go.down I LOC stairs ‘I’m going down the stairs.’ (25) ZAMBOANGUEÑO Hendéq yo ta-asé ni náda. NEG I IMPFV-do not.even nothing ‘I don’t do anything at all.’ (26) TERNATEÑO Manyána lótru dindá na Maníla. tomorrow they will.go LOC Manila ‘Tomorrow they will go to Manila.’ 3.3.2. Differential object-marking One striking feature of Chabacano syntax is the marking of objects with the lexical item ‘with’ which is phonologically kon or kung. It is the element itself and also its distribution which attract a linguist’s interest. As for the origin of the marker there are several theories in the literature including my own view, which is point 1 and 5 in the below list. In the following we try

Hispanisation processes in the Philippines

221

to review some arguments and observations concerning the object-marker kon in Chabacano. There are the following arguments as to the origin of this marker, which are not always contradictory but mutually additional: 1. Spanish origin: In the 16th and 17th some comitative objects are marked with con, as e.g. with the verbs ver ‘see’ or matar ‘kill’. Chabacano, which arose in this time, could have grammaticalized this marker. 2. Luso-asiatic origin: In many luso-asiatic creoles (e.g. Malacca, Macau, Batavia etc.) com is an object-marker. If Chabacano is really a relexified form of an Portuguese-based creole or pidgin this would not be a surprise. In addition, at least until around 1800, the Philippines had constant contact with Portuguese possessions in Asia. The place of origin could be South-India from where it spread over to other creoles in Asia (cf. Clements and Koontz-Garboden 2002 for this argument). 3. Malay origin: Baxter (1995) argues that at least in the Portuguesebased creole of Malacca ku was functionally influenced by the Bazaar Malay sama ‘with’. 4. Philippine origin: Recently, Fernández (p.c.) argues that kon has its origin in phonologically similar markers for objects before proper names. 5. Hokkien-Chinese influence: There is the coverb kâng ‘with’ marking recepients. A further puzzling issue is the synchronic distribution of the marker. So far, we have made some general observations concerning the application of the marker: 1. Human (animate) objects are most frequently marked with kon (e.g. examples 30–36). 2. Recipients (nearly always humans) and (human) objects of transitive constructions are both marked with kon. As in Spanish, the language tends to differentiate primary and secondary objects (cf. examples 33 and 36). 3. Unlike in Atlantic Creoles, double-object-constructions are always ungrammatical. 4. Double-marking of objects (for example including the theme) in ditransitive constructions is possible but rare (= marked; cf. example 28).

222 Patrick O. Steinkrüger (27) ZAMBOANGUEÑO (from a folktale) Mánga paharyadór lang ta-pwéde bisitá koneste lugár. PL hunter only IMPFV.can visit OBJ.this place ‘The place was visited only by hunters (of birds).’ (28) ZAMBOANGUEÑO Ya-dále yo kuneste líbro kunel hénte. PRF-give I OBJ.this book OBJ.DET man ‘I gave this book to the man.’ (29) ZAMBOANGUEÑO Ya-mirá yo el hénte ya-embuná konel irúq. PRF-see I DET man PRF-hit OBJ.DET dog ‘I saw how the man hit the dog.’ (30) ZAMBOANGUEÑO Konosé le konese muhér. know s/he OBJ.that woman ‘S/he knows that women.’ (31) ZAMBOANGUEÑO Páula ta-amá kun Pédro. Paula IMPFV-love OBJ Pedro ‘Paula loves Peter.’ (32) ZAMBOANGUEÑO Si Páula ta-libá líbro kon Pédro. A Paula IMPFV-carry book OBJ Pedro ‘Paula carries Peter the/a book.’ (33) ZAMBOANGUEÑO Ta-libá yo un butélya de ágwa kun mi nána. IMPFV-carry I ART bottle of water OBJ my mother ‘I carry a bottle of water to my mother.’ (34) ZAMBOANGUEÑO (Forman 1972: 157) Ya-dále abíso el maga páharo konel maga peskadór. PRF-give warning DET PL bird OBJ.DET PL fisherman ‘The birds gave a warning to the fishermen.’

Hispanisation processes in the Philippines

223

(35) ZAMBOANGUEÑO Dále komigo ágwa! give OBJ.1SG water ‘Give me water!’ (36) CAVITEÑO María ya-regalá un relós kun su nóbyo. Mary PRF-give ART watch OBJ her boyfriend ‘Mary gave her boyfriend a watch.’ (37) ZAMBOANGUEÑO (Forman 1972: 199) Ya-mirá silá síne. PRF-see they movie ‘They saw a movie.’ (38) ZAMBOANGUEÑO (Forman 1972: 157) Ya-asé le kebrá konel báso. PRF-cause s/he break OBJ.DET glass ‘He broke the glass.’ (39) ZAMBOANGUEÑO (Forman 1972: 199) Ya-mirá le el páto. PRF-see s/he DET duck ‘He saw the dug.’ (40) ZAMBOANGUEÑO (Forman 1972: 167) Ya-saká le konel muhér konel dragón. PRF-seize he OBJ.DET woman OBJ.DET dragon ‘He seized the woman from the dragon’

3.4. Discourse particles As we see in Table 5, there is only one particle of Spanish origin in Zamboangueño. This is the temporal adverb ya ‘already’ in post-verbal position (see the last line in Table 5) which indicates completiveness. This is good evidence towards demonstrating how strong the influence of Philippine languages on Chabacano currently is. In the case of Zamboangueño, however, we do not know if this is an old substrate influence or a more or less recent contact-induced phenomenon introduced by intensive contact.

224 Patrick O. Steinkrüger Table 5. Discourse particles in Zamboangueño (a provisional compilation) Particle ba (1) ba (2) gaháq galéq gáneq gayót kwan man pa ya

Short forms

ga gan gat

Function/Meaning question marker reassuring = German bloß ‘maybe’ mirativity ‘indeed’, ‘even’ etc. ‘pretty, very’ dummy = German denn (Wo denn?) ‘still, yet; again’ ‘already’

The question particle is grammatically not obligatory, since questions in Chabacano are prosodically marked, as shown in the following example: (41) ZAMBOANGUEÑO a. Tyene tu anak? b. Tyene ba tu anak? EXIST you child EXIST Q you child ‘Do you have children?’

3.5. Text (Ternateño) The following text is a narrative of the MBC-variety Ternateño, obviously the most fascinating variety of Chabacano since it has preserved its quite archaic character. The story is about a the so-called asuwáng which here appears as a “wer-pig”. This monster is a pan-Philippine motif in traditional culture and not restricted to the Ternateños. (42) TERNATEÑO (fieldwork data from 2004) Tyéni pwérku na kamínu EXIST pig LOC way The pig on the way Kel

tyéni un grúpu di hénti ayá na Báhra a group of person there LOC Ternate There was a group of men in Ternate, PART EXIST

Hispanisation processes in the Philippines

225

t-asé

harána kun mánga muhér gwápa. serenade OBJ PL woman nice who used to make serenades for nice woman. IMPFV-make

ay

na

pwéblu tandá na ótru pwéblu si village IMPFV-go LOC other village if They used to go serenading not only in the native place but also in other towns if PART LOC

tyéni lótru ta-uyí ta-yegá tyéni EXIST they IMPFV-hear IMPFV-arrive EXIST ever they come to know that there were beautiful ladies around. gwápa muhér tandá lótru par’ asé harána. Un nóchi nice woman IMPFV-go they for make serenade a night bung ridóndu el lúna ta-kaminá lótru a-kuntrá very round the moon IMPFV-walk they PRF-meet One night, it was full moon, when they were walking, they used to encounter lótru un grándi pwérku. Kel pwérku kel, a-gahryá lótru they a big pig that pig that PRF-throw they a big pig. Then they used to throw stones to the pig, but the pig did not do anything to pyédra, nung kyéri salí. Mas tasé mas ta-gritá stone no want leave more IMPFV-make more IMPFV-cry them but only shouting louder and louder. kel pwérku pero no ta-salí na húntu di luhótru that pig but not IMPFV-leave LOC near of they It didn’t want to go away from them. i asé lótru a-kohré bwénu, a-kohré résyu para and make they PRF-run good PRF-run quick for So what they did was to run as fast as they could because no […] para purki a-myédu lótru kun kel pwérku. A-hablá PRF-say not for because PRF-fear they OBJ that pig they were scared of the pig.not […] góra lótru kun kel dilótru bisínu kel pwérku kel now they OBJ that of-they neighbour that pig that And then they told it to their neighbour.

226 Patrick O. Steinkrüger a-mirá lótru kel na kamínu. A-hablá kel dilótru PRF-see they PART LOC way PRF-say that of-they And the neighbour told them: bisínu Nakú! Kel el asuwáng ‘ki na báhra kayáq neighbour oh! that the Assuwang here LOC Ternate so.that “Oh! That is the Asuwang here in Ternate!” So from that time on, they did not go serenading impesá dikél nu kyer’ ma lótru asé harána. begin so not want more they make serenade anymore because they were very, very afraid of the big pig. 4.

Other contact-induced languages as products of Hispanisation

4.1. Language mixing We have no exact knowledge on the extent of language mixing in the Philippines. However, at least one intertwined language existed on the Philippines which probably died out by the end of the 19th century. This intertwined language is reported by Schuchardt in his article “Über das Malayo-Spanische der Philippinen” (Schuchardt 1883: 125–126, footnote 1) who mentions a letter written in this language. It seems that this language was structurally comparable to Media Lengua in Ecuador: a lexical input mainly from Spanish embedded in Ilokano morphosyntax. According to Schuchardt, this intertwined language was spoken by Filipino mestizos in the city of Vigan in Luzon, Northern Philippines. Below we cite both the letter in Schuchardt’s article and a glossed version adding a rough translation into English: (43) HISPANO-ILOCANO (19th century; from Schuchardt 1883: 125–126, fn. 1) Mi estimado amigo: iparticiparco qca á nanombraranac á cabo del barrio qt sentirec unay ti caasanmo ditoy porque convidarenca met comá á maquipagdespachar itoy bassit á napreparar ditoy balay. Unica á noticia á maipropocionarca qca: ni comadrem buntis manen qt idi inda cobraren ti buisna, timmacbu qt natnag idiay batalan qt nabiac diay quiliquilina. Na castigar ngarud, pues naarimuhanan la unay. Toy amigo qt servidormo Z.

Hispanisation processes in the Philippines

227

The following abbreviations are used in the text: qt = ket (predicate marker and conjunction) qca = kenca (2SG oblique) Glossed version of the text:4 Mi estimado amigo: i-participar-co qca á na-nombraran-ac TH-inform-I 2SG.OBL LNK PRF-nominate-1SG My dear friend á cabo del barrio qt sentir-e-c unay ti ca-asan-mo5 LNK chief of the barrio and feel-TR-1SG very the NOM-absence-your ditoy porque convidaren-ca met comá á here because invite-1SG>2SG also hopefully/should LNK maquipag-despachar itoy bassit á na-preparar ditoy balay. COM-serve this small/little LNK PRF-prepare here house Unica á noticia á mai-propocionar-ca qca: only LNK news LNK able-deliver-1SG>2SG to you ni comadre-m buntis manen qt idi in-da ART godmother-2SG pregnant again and before went-they cobrar-en ti buis-na, timmacbu qt natnag idiay charge-TR ART tax-3SG ran (Tag) and fell over there batalán qt na-biac diay quiliquili-na. bamboo back porch and PRF-break that armpit-3SG Na-castigar ngarud, pues naarimuhanan PRF-punish then than/so thrifty (Tag) la [=laeng] unay. Toy amigo qt servidor-mo Z. [just, only] very this friend and servant-2SG Z. Translation of the letter into English: ‘My friend: I inform you that they appoint me chief of the “barrio” and I regret a lot your absence because I also wanted to invite you for the preparations I have at home. The only news I have for you: Your godmother is again expecting and one day, when they wanted to collect her tax, she ran way and fell on the terrace: the consequence of that accident was the injury of her shoulder, a worthy punishment of her stinginess. Your friend and servant Z.’ Since we have only this one document, another explanation could be heavy borrowing and/or borrowing: the situation in the 19th century was probably similar to the modern one. However, instead of “Taglish”, a mixture of Ilokano and Spanish, “Ilocañol” was spoken.

228 Patrick O. Steinkrüger 4.2. Pidginization There are several reports on pidginized forms of Spanish in the Philippines (see e.g. Lipski et al. 1996: 280): These restructured forms are called ‘Bamboo Spanish’, ‘Español de Cocina’, or Chinese Tagalo-Spanish. The first label refers to a restructured from of Spanish mixed with Tagalog and Japanese which was used between Japanese soldiers and Filipinos during the Second World War around Davao (Mindanao, South-Philippines). In the following two examples the items ikáw ‘you’ and nánay ‘mother’ are Tagalog or Cebuano and oksan or okusan ‘wife’ is from Japanese: (44) BAMBOO SPANISH (from Whinnom 1956: 16) a. Cuando sale ikaw nanay? when leave you mother ‘When were you born?’ b. Nombre oksan tiene? name wife has ‘What is your wife’s name?’ The difference between Kitchen Spanish and Chinese Tagalo-Spanish is not really clear. It may be that the latter is still spoken by Chinese shop keepers in Davao but there is a lack of data. As we see in the following dialogue, Español de Cocina was also spoken by the Chinese and Chinese mestizos. The following dialogue was found by Schuchardt in a hispano-filipino short-story by Montero y Vidal from the 19th century: (45) SPANISH-BASED PIDGIN (adapted from Schuchardt 1883: 115) SHE: ¿ Cosa quiere suya conmigo? [‘What do you want from me?’] HE: Mia quiele platicalo. [‘I want to speak to you.’] SHE: ¿Y para cosa? [‘For what?/Why?’] HE: Porque vos magandang dalaga. [‘Because you are a nice girl.’] SHE: ¡Abá! Está enamorando conmigo este chino! [‘Oh! He is falling in love with me this Chinese!’] HE: ¡Oh, oh! icao mariquit. [‘Yes, you are beautiful!’] SHE: Kánsia. [Thanks.] HE: Mia quiele mucho con suya y tiene cualtas para puede compla saya y candonga. [‘I like you a lot and I have money for buying skirts and candongas (the traditional scarf).’] SHE: Bien, sigue suya conmigo, para habla bueno-bueno con aquel mi tia. [‘OK. Follow me and we will convince my aunt.’]

Hispanisation processes in the Philippines

229

This short paragraph shows some interesting features. Both persons still speak their mother tongue (Tagalog and Chinese respectively) and use a restructured from of Spanish with each other. Consequently, Whinnom’s criterion of tertiary hybridization is fulfilled. In addition they probably have a basic knowledge of each other’s language: The Filipina girl thanks in Hokkien-Chinese (from gn-xiè) and the Chinese boy pays compliments to the girl in Tagalog: magandáng dalága ‘nice girl’ an ikáw marikít ‘you are beautiful (lit. ‘shining’). Structurally, there are similarities in their speech but also differences in comparison to Chabacano, e.g. the use of con ‘with’ as an object-marker, the question-word cosa ‘what’ and the use of infinitive forms but without the final -r: platicá ‘speak’ (= Chabacano Ternateño), complá ‘buy’ (cf. also 2.3.1). On the other hand they use the Spanish possessive pronouns suya and mia6 instead of yo ‘I’ and eli ‘s/he’. There also are vestiges of inflectional morphology: está enamorando ‘is falling in love’. This is quiet unusual for fully restructured languages but underlines Schuchardt’s assumption that Kitchen Spanish was more of a jargon because “it exists on innumerable levels with greater or lesser approximation to Spanish grammar and structure and with greater or lesser use of Malay7 vocabulary’ (quoted from Lipski et al. 1996: 280). After the verb platicá ‘speak’ there is the object-clitic -lo which only looks like the post-verbal or clause-final perfective marker -lo used in China Coast Pidgin. KitchenSpanish shows structural similarities to Chabacano and other Portuguesebased creoles in Asia since Manila was constantly connected to the places where these languages were spoken during the Spanish colonial period – including demographic exchange. 4.3. Philippine Spanish After more than one hundred years since the end of Spanish rule, the census from the year 1990 indicated that less than 0.1% of the population spoke Spanish as a mother tongue, i.e. around 2700 people. The few soldiers, merchants, and civil bureaucrats who came to the Philippines tended to remain in the walled city of Manila or in other garrison towns. Other places where many Spanish or Mexican whites settled was Cebu and other places in Central Philippines, and finally Zamboanga in Mindanao. It was only in 1863 that Madrid ordered that primary schools use the Spanish language. However, this was not practiced in many cases and expanded only very late and superficially throughout the islands. Spanish became a language of the upper class but was exclusive of social strata particularly during the 19 th

230 Patrick O. Steinkrüger century. In this context, we should emphasize the importance of the Chinese mestizos in the late 19th and early 20th century who to a great extend belonged to the Hispanophone elite of the Philippines and mainly built the Filipino identity (see e.g. Wickberg 2001). At least two external factors caused the decline of Spanish in the Philippines: After the PhilippineAmerican War (1898–1900) the U.S. pushed English as a language of administration and higher education. Unlike in the late 19th century when the Filipinismo expressed itself in Spanish, the new Filipino elite in Manila, also supported by the U.S., choose Tagalog as the national language with the foundation of the new Republic in 1946. Geographically, there are currently still three main centres of Spanishspeaking populations: 1. Luzon (especially in and around Metro Manila) 2. Central Philippines (especially Cebu and Iloilo) 3. Mindanao (especially Zamboanga City) The following is based on a corpus which the author of this essay has compiled, including the speech from nearly 20 native speakers from different places in the Philippines. Consequently, this sample represents around 1% of the app. 2700 Filipinos who declare Spanish to be their mother-tongue. At the phonological level the following characteristics can be generalized: a. There is never yeísmo. It means that intervocalic // is realized as such and not as [j]. This is unlike Ternateño which represents the oldest form of Spanish in the Philippines and can also be observed in some loans in Philippine languages (cf. example 8). b. Merging of the historical sounds // and // as [h]: mujer [mu'her] ‘woman’ and bajo ['baho] ‘down’. The same occurs in most American and Andalusian varieties. c. There is always seseo. It means that this not like in Standard European Spanish with a phonematic distinction of / / and /s/ but both sounds are realized as [s]: hacer [a'ser] ‘make’ vs. caso ['kaso] ‘case’. d. Deletion of intervocalic /d/ in the group /-ado/ even within the stem: lado ['law] ‘side’, prado ['praw] ‘meadow’. On the morphosyntactical level, no particular feature has been found. It seems that Quilis (1992: 188) was right in saying “En el nivel gramatical, no se producen muchos fenómenos dignos de mención en el español de Filipinas como lengua materna”. Nevertheless, many speakers show un-

Hispanisation processes in the Philippines

231

steadiness in the agreement of number and gender, especially of objectpronouns. The last phenomenon could be due to English and Philippine influence as in these languages this category nearly does not exist. In addition, it shows that most speakers lack regular practice because communication in Spanish is not dominant. Some particularities are found in its lexicon. The lexicon of Philippine Spanish reflects all the centuries of Hispanic dominance in the Philippines and, since the mid-20th century, the impact of Tagalog and English in public and private life. Since English has become the language of social prestige, especially after WWII, there are quite a lot of lexical loans or calques in the language, comparable to the situation in Puerto Rica or in the United States. Concerning the lexical structure of Philippine Spanish, see for example Steinkrüger (2007: 259–260) and also the interesting if not scientific debate in de Veyra (1951). 5. Some conclusions As we have seen, the Spanish impact on the linguistic landscape of the Philippines is still significant today. Most visible is the influence on the lexicon (including onomastics) in most Philippine languages. Following my own observations it is synchronically incorrect to call Spanish lexical items “loans” since there are phonologically and morphologically deeply integrated into the respective Philippine languages and they are not considered as “foreign” by their native speakers. Alongside lexical material, there are also are quiet a few grammatical features, like functional words (e.g. pero ‘but’), but also other morphosyntactic constructions (e.g. comparison) and categories (e.g. gender). The most Hispanised languages of the Philippines are probably Tagalog and Cebuano, which represent 50% of the population in the Philippines. The latter is the language of the first Spanish settlement Cebu in Central Philippines, and Tagalog, which became the language of the colonial and now national capital, was from the 16th century onwards influenced by Spanish normative efforts. If we contrast the case of the Philippines with the Americas and Marianas, we can assume that it is probably because of the lack of an extended bilinguism in the past and in the present that the structural influence of Spanish was less significant. It is still not entirely clear if restructured forms of Spanish influenced the Philippine languages (e.g. forms of the verb, c.f. 2.3.1). The other important result of Hispanisation is Chabacano. With its high number of speakers in Zamboanga and surrounding places, Chabacano is

232 Patrick O. Steinkrüger the largest speech-community of an European-based creole in Asia and of a Spanish-based creole world-wide. Its structure is of interest for Romance studies, since it has preserved an archaic character of Spanish – especially Ternateño – and it shows similarities with other Romance-based creoles around the world but mainly with the Portuguese-based ones in Asia. It is therefore certain that Chabacano is a case of relexification of a former Portuguese-based creole or pidgin. On the other hand, Chabacano is also a challenge for Austronesian studies because of its high rate of MalayoPolynesian features ranking from phonology to pragmatics and the lexicon.

Notes 1.

2. 3.

4. 5. 6. 7.

1SG = first person singular, 2SG = second person singular, 3SG = third person singular, ART = article, COM = comitative, COP = copula, DET = determiner, DO = direct object, EXIST = existential, IMPFV = imperfective, IND = indicative, IRR = irrealis, LNK = linker, LOC = locative, NEG = negator, NOM = nominalizer, OBJ = object, OBL = oblique, PART = particle, PL = plural, POSS = possessive, PREP = preposition, PRES = present, PRF = perfect, Q = question marker, SG = singular, TH = theme, TR = transitive marker, VRBLZ = verbalizer. Between 1580 and 1640 Spain and Portugal formed a political unity. Baxter (1988: 53) also reports on “older” forms not used any more but known to the speakers: nos-turo ‘we’, bos-turo ‘you (PL)’ and elotru ‘they’. He assumes that bos-turu in Papia Kristang is derived from vos todo(s), a convergent construction influenced by Malacca Bazaar Malay lu semua ‘you all’. (Baxter 1988: 82, fn. 8) Many thanks go to Carl Rubino for helping me to gloss the text. Should be caawanmo instead of caasanmo. This fact is reminiscent of Papiamentu ‘mi’ and English-based China Coast Pidgin ‘my’, both meaning ‘I’. With the label “Malay”, Schuchardt refers to the Malayo-Polynesian languages of the Philippines.

References Adelaar, Alexander and Nikolaus P. Himmelmann (eds.) 2005 The Austronesian languages of Asia and Madagascar. London /New York: Routledge. Albalá, Paloma 2003 Hispanic words of Indoamerican origin in the Philippines. Philippine Studies 51 (1): 125–146.

Hispanisation processes in the Philippines

233

Alcántara, Teresita A. 1996 Los hispanismos en los medios de comunicación social filipinas. Actas del 40 Congreso de Hispanistas de Asia, 569–576. Seúl/ Corea: AAH. Anderson, Stephen R. to appear Second position clitics in Tagalog. To appear in The nature of the word, Sharon Inkelas and Kristin Hanson (eds.), MIT Press [pdfversion from the net, 1–18]. Barón Castro, Radolfo 1972 Estudio preliminar. In Hispanismos en el Tagalog, Adolfo Cuadrado Muñiz (ed.), XVII–LXIV. Madrid: Educación Iberoamericana. Baxter, Alan N. 1988 A grammar of Kristang (Malacca Creole Portuguese). Canberra: Australian National University. 1995 Um importante sincretismo no português crioulo de Malaca: a preposição multifuncional ku. In Miscelânea de estudos lingüísticos, filológicos e literários in memoriam Celso Cunha, Celine da Cunha Pereira and Paulo Roberto Pereira (eds.), 15–33. Rio de Janeiro: Nova Fonteira. Bowen, J. Donald 1971 Hispanic languages and influence on Oceania. Current Trends in Linguistics 8: 938–952. Bunye, Maria Victoria R. and Elsa Paula Yap 1971 Cebuano grammar notes. Honolulu: University of Hawai’i Press. Clements, J. Clany and Andrew J. Koontz-Garboden 2002 Some theoretcal implications of adpositions in Spanish and Portuguese-based creoles. Paper presented at the annual meeting of the SPCL, San Francisco. Constantino, Ernesto 1978 Tagalog and other major languages of the Philippines. Current Trends in Linguistics 8: 112–154. Dalgado, Sebastião Rodolpho 1906 Dialecto Indo-Português do Norte. Revista Lusitana 9 (3/4): 142–228. Fernández, Mauro (ed.) 2001 Shedding light on the Chabacano language [Special issue of Estudios de Sociolingüística. Linguas, sociedades e culturas 2:2], Vigo, Spain: University Press. Forman, Michael L. 1972 Zamboangueño texts with grammatical analysis: A study of Philippine Creole Spanish, Ph.D. diss., Cornell University. Frake, Charles O. 1971 Lexical origins and semantic structure in Philippine Creole Spanish. In Pidginization and creolization of languages, Dell Hymes (ed.), 223–242. Cambridge: Cambridge University Press.

234 Patrick O. Steinkrüger Himmelmann, Nikolaus P. 2005a Tagalog. In The Austronesian languages of Asia and Madagascar, Nikolaus P. Himmelmann and Alexander Adelaar (eds.), 350–376. London: Routledge. 2005b Tagalog. In Morphology. A handbook on inflection and word formation, vol. 2, Gert Booij, Christian Lehmann, and Joachim Mugdan (eds.), 1473–1490. Berlin /New York: Mouton de Gruyter. Jun, Akamine 2005 Sama (Bajau). In The Austronesian languages of Asia and Madagascar, Nikolaus P. Himmelmann and Alexander Adelaar (eds.), 377– 396. London: Routledge. Kitano, Hiroaki and Michael Raymon Manaloto Pangilinan 2003 Overview of Aita Mag-Anchi in Central Luzon, Philippines. In Descriptive and theoretical studies in minority languages of East and South East Asia, Hiroaki Kitano (ed.), 169–223. Kyoto: Aichi University of Education. Lipski, John M. 1986 On the reduction of /s/ in Philippine Creole Spanish. Diachronica III (1): 43–66. 1992 New thoughts on the origins of Zamboangueño (PCS). Language Sciences 14: 197–231. Lipski, John, Peter Mühlhäusler, and F. Duthin 1996 Spanish in the Pacific. In Atlas of languages of intercultural communication in the Pacific, Asia, and the Americas, Stephen A. Wurm, Peter Mühlhäusler, and Darrell T. Tryon (eds.), 271–298. Berlin / New York: Mouton de Gruyter. López, Cecilio 1965 The Spanish overlay in Tagalog. Lingua 14: 467–504. McKaughan, Howard 1971 Minor languages of the Philippines. Current Trends in Linguistics 8: 155–167. Munteanu, Dan 1996 El Papiamento, lengua criolla hispánica. Madrid: Gredos. Pérez Tejedor, Juana Pabla 2004 El Criollo de Palenque da San Basilio: Una visión estructural de su lengua. Bogotá: Universidad de los Andes / CESO / CCELA. Quilis, Antonio 1973 Hispanismos en tagalo. The Canadian Journal of Romance Linguistics 1: 68–92. 1976 Hispanismos en cebuano. Madrid: Alcalá. 1992 La lengua española en el extremo oriente ibérico. In La lengua española en cuatro mundos, Antonio Quilis, 109–187. Madrid: MAPFRE.

Hispanisation processes in the Philippines

235

Riego de Dios, María 1989 A composite dictionary of Philippine Creole Spanish (PCS). Studies in Philippine Linguistics 7 (2): 1–210. Rubino, Carl 2005 Iloko. In The Austronesian languages of Asia and Madagascar, Nikolaus P. Himmelmann and Alexander Adelaar (eds.), 326–349. London: Routledge. Santos Ferreira, José dos 1978 Papiá Cristám di Macau. Epítome de Gramática comparada e vocabulário. Macau: Fundação Macao. Schuchardt, Hugo 1883 Kreolische Studien IV: über das Malaiospanische der Philippinen [Creole Studies IV: on the Malayo-Spanish of the Philippines]. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften zu Wien 105: 111–150. Steinkrüger, Patrick O. 2003 Morphological processes of word formation in Chabacano (Philippine Spanish Creole). In Phonology and morphology of Creole languages, Ingo Plag (ed.), 253–268. Tübingen: Niemeyer. 2004 Philippine Spanish Creoles. In Encyclopedia of Linguistics, Philipp Strazny (ed.), New York: Routledge. 2007 La influencia léxica en Filipinas. In Lenguas en contacto y cambio lingüístico en el Caribe y más allá, Wiltrud Mihatsch and Monika Sokol (eds.), 249–263. Frankfurt a. M. etc.: Peter Lang. Stolz, Christel and Thomas Stolz 1996 Funktionswortentlehnung in Mesoamerika, Spanisch-amerindischer Sprachkontakt. STUF 49: 86–123. 1997 Universelle Hispanismen? Von Manila über Lima bis Mexico und zurück: Muster bei der Entlehnung spanischer Funktionswörter in den indigenen Sprachen Amerikas und Austronesiens. Orbis XXXIX: 1– 77. Thompson, Roger M. 2003 The Spanish overlay. In Filipino English and Taglish, Roger M. Thompson, 59–66. Amsterdam /Philadelphia: Benjamins. Veyra, Jaime C. de 1951 Filipinismos en Lengua española. Manila: Nueva Era Press. Whinnom, Keith 1956 Spanish contact vernaculars in the Philippine Islands. Hong Kong: Hong Kong University Press. 1954 Spanish in the Philippines. Journal of Oriental Studies I: 129–194. Wickberg, Edgar 2001 The Chinese Mestizo in Philippine History. Manila: Kaisa para sa Kaunlaran.

236 Patrick O. Steinkrüger Wolfenden, Elmer Paul 1971 Hiligaynon reference grammar. Honululu: University of Hawai'i. Wolff, John U. 1973/4 The character of borrowings from Spanish and English in the languages of the Philippines. Philippine Journal of Linguistics 4 (1): 72–82. 2001 The influence of Spanish on Tagalog. In Lo propio y lo ajeno en las lenguas austronésicas y amerindias, Klaus Zimmermann, and Thomas Stolz (eds.), 233–253. Frankfurt a.M./Madrid: Vervuert/Iberoamericana. Zorc, R. David and Rachel San Miguel [eds. Annabelle M. Sarra and Patricia O. Afable] 1991 Tagalog slang dictionary. Kensington, MD: Dunwoody Press.

Pro or contra Hispanisms:1 Attitudes of native speakers of modern Chamoru Rosa Salas Palomo and Thomas Stolz

1. Introduction Chamoru (aka Chamorro) is the autochthonous language of the Marianas which are politically divided into two separate entities, namely the unincorporated US territory Guam and the autonomous Northern Marianas (associated with the USA). Except otherwise stated, we refer exclusively to the Chamoru of Guam in this article. Genetically Chamoru belongs to the Austronesian phylum and therein more specifically to the Philippinian branch. The estimates for the size of the Chamoru speech community calculate a maximum of about 90,000 native speakers of whom two thirds reside on Guam. The survival of Chamoru is presently at stake as the language has become a minority language on Guam where it has to compete with dominant American English with a ratio of 40% Chamorus to 60% other ethnic groups (Hernández Legorreta 2001: 416). In the past, Chamoru had to endure pressure from another prestigious language. The Marianas were ruled by Spain for a period of 234 years (1665–1899) before the Spanish-American War of 1898 and the subsequent purchase of Spain’s remaining island possessions by imperial Germany put an end to the Spanish dominance in the Pacific. Nevertheless, the Spanish language has left visible traces in modern Chamoru2 as a considerable number of lexical items – both content words and function words – have a Spanish etymology (Rodríguez-Ponga 1995; Stolz 1998)3. Many of these words occur frequently in everyday communication of Chamoru native speakers. These Hispanisms are currently used in spoken as well as in written Chamoru of all genres (Stolz 2003). Spanish-derived words and those inherited from Austronesian form part of the contemporary lexicon of Chamoru and thus qualify equally as “Chamoru words”. Underwood (1999: 528–530) demonstrates how important the question of Hispanisation is in the Chamorros’ search for identity4. Since activists of Chamoru culture and language are making an effort to create a purified register of Chamoru for instance for the purpose of reviving the ancient practice of preserving the collective knowledge of the pre-Hispanic past in traditional chants and

238 Rosa Salas Palomo and Thomas Stolz dances5, the problem arises for them to identify Hispanisms and provide an Austronesian replacement for them. However, not all of the Hispanisms strike the eye as loanwords and thus they may escape easily the observer’s attention (just as some putative Hispanisms might turn out to be of a different origin). What this difficulty to recognise Hispanisms at first glance also suggests is that not all native speakers of Chamoru are aware to the same degree of the Spanish origin of some of the items they use on a daily basis. The awareness/unawareness of the Spanish background, in turn, may also affect the speaker’s attitude towards a particular lexical item. The attitude may be such that a speaker rather tries to avoid a certain word if s/he knows that it is derived from Spanish (for instance, to make a conscious though indirect political statement). This is the general background for the questions and hypotheses of this study which we consider a pilot study from whose results other like-minded investigations (not only of Chamoru) may develop. For the first time in the history of research on Chamoru, we want to find out where the language loyalty of the native speaker of Chamoru is directed.6 Are there preferences for Spanish to the detriment of Austronesian or the other way round? If there are any preferences at all, are they random or are they determined by sociolinguistic parameters? These questions and hypotheses are spelled out in more detail in section 2. The method employed and the population of participants are presented together with an extract of the questionnaire and additional observations in section 3. Section 4 is dedicated to the presentation of the results which are checked against our original hypotheses. In the conclusions (section 5), we evaluate our findings and suggest topics for follow-up studies.

2. Questions and hypothesis We start from the general assumption that it makes a difference to Chamoru speakers whether they use Spanish items or Austronesian items. For the purpose of this study, we also assume that speakers distinguish the two etymological components of present-day Chamoru. From this basis, a number of questions and hypotheses can be derived. For each of the questions to be posed, we provide a working hypothesis which serves as our guideline and reference point in the remainder of this contribution. In (1)–(3), we present the questions and hypotheses in the briefest possible form. Below we explain why we opt for these particular hypotheses as potential answers to the questions.

Pro or contra Hispanisms

239

(1)

Q: Do modern Chamoru speakers direct their language loyalty towards Austronesian or Spanish? H: Modern Chamoru speakers are loyal to Austronesian.

(2)

Q: Is the language loyalty identical to (1) when it comes to selecting (a) the more correct form, (b) the more frequently used form, (c) the form actually used? H: There is a preference for Spanish in all three cases (a)–(c).

(3)

Q: Is there a relationship between language loyalty and (a) gender, (b) age, (c) ethnicity, (d) education, (e) place of (birth and) formative years, (f) place of residence, (g) size of family (in terms of siblings), (h) relative order among siblings (oldest, youngest, in between) H: Yes, namely (a) men prefer Spanish. (b) the older the informant the more likely is a preference for Spanish. (c) speakers from Chamoru families will have a stronger loyalty to Austronesian than speakers from ethnically mixed families. (d) the more educated the informant the more likely is a preference for Spanish. (e) there is a clear correlation between the place where the informant lived in his/her formative years and his/her preference of Spanish or Austronesian. (f) the place of residence is indicative of the extent to which an informant is exposed daily to certain styles/registers/dialects and thus correlates with his/her preference of Spanish or Austronesian. (g) the larger the number of siblings the more communicative interaction with the effect of levelling, i.e., in larger families speakers’ preferences will resemble closely the average computed for the entire population. (h) first born siblings will function consciously as patterns for the linguistic behaviour of their younger brothers and sisters and

240 Rosa Salas Palomo and Thomas Stolz thus make an effort to use correct/educated language (= give preference to Spanish). To make things as simple as possible, we deliberately take a relatively naïve laymen-like stance as to the hypotheses. Of course, we know perfectly well that what we claim here and only for practical reasons, is not universally true. However, the hypotheses are such that they reflect what often comes to mind first.7 For (1H), we simply assume that native speakers of a language inherently identify with their language (in its “pure” form) and tend to value it higher than other languages. This attitude extends to elements which have a foreign origin. The function of language as a symbol of cultural identity (Fought 2006) is at stake here and will influence the speaker’s choice and bias him/her towards Austronesian. For (2H), we assume the opposite: Spanish wins out because (a) it is associated with a higher degree of education (“learned forms”) and thus appears to be ultimately more correct. This is where the function of language as a social marker comes into play. We pretend that Spanish still retains some of its former prestige. This assumption is responsible for many of the subsequent hypotheses. For (b), the informant identifies what s/he perceives to be the more frequently used form (although this subjective view might turn out to be in disagreement with the actual statistics). The reason why speakers also consider the Spanish form the more frequent one may or may not have to do with (a). It is also possible that the hypothesised option reflects their idea of the degree of Hispanisation of their native language. They know that Chamoru is influenced by Spanish and accordingly they assume that the Spanish item is the one that occurs more often. For (c), the situation is different as, this time, we look at what the respondents actually use themselves in the sample sentences. We hypothesise that they will choose preferably Spanish only because we assume that the speakers’ linguistic behaviour tends to be consistent with their previous choices. What are the motivations for our hypotheses in (3H)? As to gender differences, we start from the assumption that women are linguistically more conservative. Given that the Hispanisms are relative newcomers to the Chamoru lexicon (while Spanish disappeared from the linguistic scene of Guam a century ago), we do not expect to find them in women’s speech but rather with men who usually have a more outwardoriented attitude. The age parameter overlaps to some extent with the one of education as being older often implies that one has finished schooling whereas others have not yet. In addition, the members of the oldest age group may look back to a time in which Hispanisms were even more widespread than nowadays (and/or contacts with people who actually knew

Pro or contra Hispanisms

241

Spanish were still frequent). Ethnically homogeneous families mean also linguistically homogeneous surroundings and thus speakers from families with a purely Chamoru background are believed to show preferences to Austronesian. Education and assumed correctness are related to each other. We assume that informants with a higher degree of education opt for the Spanish words because they are available to them as social markers. The formative years of one’s lifetime are not labelled formative for nothing. The early childhood and adolescence are periods during which speech habits develop which later on are often difficult to change. Thus, whoever grew up in surroundings where Hispanisms were scarce or frowned upon, will opt for the Austronesian etymon. Who passed his/her formative years in an ambient where Hispanisms were employed liberally or even highly praised, will opt for the Spanish word. On the other hand, who currently lives in the metropolitan area of Central Guam is exposed to a cosmopolitan ambient where languages other than Chamoru dominate. We assume that this intensity of multi-ethnic and multi-lingual contacts brings about a more tolerant approach towards Spanish whereas in the less urbanised regions of Guam, Austronesian should rank at the top. In large families, siblings8 have a higher chance of interacting communicatively and the constant verbal exchanges may lead to a levelling effect. In contrast to this levelling effect, older siblings, especially first born ones assume the role of a social model for their younger siblings. Since they know about their prominent position and the influence they are exerting, they opt for what is considered more correct language and thus introduce more Spanish. In section 4, we will see whether or not these hypotheses are borne out by the facts. Before we look at the results, however, it is necessary that we familiarise ourselves with the project design and its implementation.

3.

Method and participants

3.1. Who worked with us In Spring 2005, 16 students of the Junior and Senior level of the Intermediate Chamoru I class taught by Rosa Salas Palomo at the University of Guam were instructed to carry out the questionnaire-based data collection. The interviewers themselves are all Chamoru native speakers and work as Chamoru language teachers in public schools. They were asked to collect data from ten informants per interviewer. The group of informants was intended to comprise an equal number of male and female participants. The

242 Rosa Salas Palomo and Thomas Stolz selection of the participants was strictly volunteer-based and thus, neither stratification nor random selection apply. The only criterion which had to be fulfilled was that the informant be a native speaker of Chamoru. Originally, a population size of 160 informants was aimed at. However, owing to various reasons, only 121 informants could be included in this study. For a start, we break down the group of participants according to sociolinguistic parameters. * Gender:

With 53.7% = 65 participants, the female group is larger than the male one which covers 46.3% = 56 participants.

* Age:

Divided into groups according to decades, the informant population is characterised by a strong bias in favour of the older and oldest generations: (a) 21–30 years: 0.80% = 1 person, (b) 31–40 years: 6.60% = 8 persons, (c) 41–50 years: 22.31% = 27 persons, (d) 51–60 years: 35.54% = 43 persons, (e) 61–70 years: 19.80% = 24 persons, (f) 71–80 years: 9.90% = 12 persons, (g) 81–90 years: 5.00% = 6 persons.

* Ethnicity: This parameter is clearly biased in favour of ethnic Chamorus whose group accounts for 94.2% of the informant population which is equivalent to 114 persons. The non-Chamorus represent 5.8% (= 7 persons), namely 2 Chamoru-Japanese, 1 Chamoru-Tagalog, 1 Chamoru-American, 1 ethnic Tagalog and 2 other with unspecified ethnicity. * Level of education: According to the number of years spent at school or the level of tertiary education, the informant population can be characterised as well educated. There are 40% = 49 persons who attended college and/or university. The minimum of 0–3 years of schooling applies to 4.10% = 5 informants. Only one informant (= 0.80%) looks back at 4–6 years of education. 10 informants (= 8.30%) stayed in school for 7–9 years whereas the largest group of informants (46.30% = 56 persons) have a schooling of 10–12 years i.e. secondary education. As to tertiary education, 5 informants (= 4.10%) reached the AA level, 17 informants (= 14%) have a B.A./B.S., 5 informants (= 4.10%) passed 1–2 years of graduate studies, 9 informants (= 7.40%) have a M.A./M.S. degree and one informant has a PhD degree.

Pro or contra Hispanisms

243

* Early residence: The place of residence during the formative years of our informants reflects the current distribution of the population over the island of Guam whose main centre of settlement is in Central Guam (= metropolitan area). 48.80% = 59 of our informants spent their formative years in this area whereas the other areas display much lower scores, namely 22.30% = 27 persons from South Central Guam, 13.20% = 16 persons from Northern Guam, 9.10% = 11 persons from Southern Guam. 8 informants (= 6.66%) originally lived outside Guam (4 in the Northern Marianas and 4 elsewhere). * Present residence: The values for early residence and present residence are similar to each other, though not absolutely identical. Expectedly, the largest group lives in Central Guam (46.30% = 56 persons) and South Central Guam is again second-best with 24.80% = 30 informants. Northern Guam boasts of the highest gain (8% = 10 persons more) in comparison to the figures given for early residence as nowadays 21.50% = 26 informants live in the north of the island. In contrast, Southern Guam has lost (5% = 3 persons) as it is represented by only 6.66% = 8 persons. One individual (= 0.80%) resides outside Guam. * Number of siblings: Chamorro families are sizeable. There are only 17 informants (= 14.10%) with 0-3 siblings. All other informants have more and even many more sisters and brothers: (a) 4–6 siblings: 20.70% = 25 persons, (b) 7–9 siblings: 36.40% = 44 persons, (c) 10–12 siblings: 18.20 = 22 persons, (d) 13–15 siblings: 6.6% = 8 persons, (e) 16–18 siblings: 2.5% = 3 persons, (f) 19 and more siblings: 1.7% = 2 persons. * Order among siblings: Of our informants, 20% = 24 persons are oldest among the siblings whereas 15% = 18 persons are youngest. The remaining 65% cover 41 persons (= 34%) who belong to the upper age group among the siblings, 31 persons (= 26%) belong to the lower age group. 7 informants (= 6%) qualify as representatives of the middle age group. Owing to the voluntary recruitment, the internal make-up of the group of informants is not strictly balanced. For age and education, the biases are especially strong in a way that might have repercussions on the results. On

244 Rosa Salas Palomo and Thomas Stolz the one hand, the youngest generation of speakers (age < 20 years) is completely absent from the survey. This can be explained by the fact that young speakers lack confidence as to their command of Chamoru and thus did not volunteer. Well educated speakers are clearly overrepresented in our group. This bias has a purely practical reason because the interviewers were all Chamoru language teachers who recruited informants from their immediate social surroundings where people with at least the same level of education are expected to be numerically strong. We are aware of the necessity that in future studies, the research focus should be on the youngest speaker generation and less well educated speakers (perhaps with special reference to those who live outside the metropolitan area of Central Guam). To avoid skewing the statistics, we will subsume several of the above numerically insignificant categories under major umbrella categories which allow us to generalise in lieu of paying too much attention to the particulars.

3.2. The questionnaire The questionnaire has been developed specifically for this study. It was administered individually in sessions of 30–40 minutes at a location of the interviewee’s choice. In the questionnaire, the interviewee had to make decisions about pairs of synonymous words/forms all of which are currently in use by speakers of Chamoru. Each pair of synonymous words consists of an Austronesian item and its Spanish-derived equivalent. All Spanish-derived items had been checked with an educated native speaker of Spanish in order to determine their origin and meaning (in the donor language Spanish). Some of the non-Austronesian words are ultimately of (mostly Mexican) Amerindian origin but have been subsumed under Spanish as they also form part of the lexicon of Mexican Spanish (Albalá 2001: 282). The interviews were conducted in Chamoru. The questionnaire is divided into three parts. In part 1, the interviewee was asked to identify the more correct form of the word pair. In part 2, the informant had to state which of the two forms is the more frequently used. Part 3 is different in the sense that no statement was elicited of the informant. In this part, s/he had to fulfil a translation task which enabled us to determine which of the two forms of a given word pair is put to actual use by the speakers. For parts 1–2, the same set of pairs of (partial) synonyms was used to elicit statements of the interviewees. The words used in the 45 pairs are enumerated in (4) with their English translation and the Spanish etymon also accompanied by an English translation. In a small number of cases, the

Pro or contra Hispanisms

245

word pairs also included phrases. The examples appear in alphabetical order of the first content word on the Austronesian side. As both Austronesian and Spanish elements form part of the Chamoru lexicon, none of the two constituents of the word-pairs has the exclusive right to being classified as Chamoru. Thus, the elements are identified according to etymology, namely Austronesian vs. Spanish. (4)

Pairs of synonyms 9 AUS addeng = SP patås (< patas ‘feet [of animal]) ‘foot’; AUS adingani = SP kuentusi (< cuento ‘tale’) ‘talk to’; AUS agagå’ = kueyu (< cuello ‘neck’) ‘neck’; AUS amko’ = SP bihu (< viejo ‘old’) ‘old’; AUS båba = SP målo (< malo ‘bad’) ‘bad’; AUS baba = SP atbieto (< abierto ‘open’) ‘open’; AUS chålan = SP kåyi (< calle ‘street’) ‘road’; AUS chocho = SP boka (< boca ‘mouth’) ‘eat’; AUS cho’gue = SP bida (< vida ‘life’) ‘to do, to work’; AUS fafa’nå’gue = SP ma’estro/ ma’estra (< maestro/maestra ‘teacher’) ‘teacher’; AUS fa’maolek = SP arekla (< arreglar ‘repair’) ‘repair’; AUS fanha’aniyan = SP kalendario (< calendario ‘calendar’) ‘calendar’; AUS fanhaluman = SP entråda (< entrada ‘entrance’) ‘entrance’; AUS fantinane’an = SP kommon (< común ‘[inter alia] public toilet’) ‘toilet’; AUS fine’nena = SP primet (< primer ‘first’) ‘first’; AUS fino’ = SP lengguåhi (< lengüaje ‘language, speech’) ‘language’; AUS ga’chong = SP amigo (< amigo ‘friend’) ‘friend’; AUS gefpa’go = SP bunitu (< bonito ‘nice’) ‘pleasant’; AUS gupot = SP fi’esta (< fiesta ‘feast’) ‘feast’; AUS ha’åni = SP dihas (< días ‘days’) ‘day’; AUS ha’of = SP pecho (< pecho ‘chest’) ‘chest’; AUS hokso’ = SP sabåna (< savana ‘grassland’) ‘mountain’; AUS malåte’ = SP intelihente (< inteligente ‘intelligent’) ‘intelligent’; AUS i mamaila’ = SP otro (< otro ‘other’) ‘next’; AUS maolek = SP bueno (< bueno ‘good’) ‘good’; AUS ogga’an = SP manana (< manaña ‘morning’) ‘morning’; AUS pa’a = SP kottura (< cultura ‘culture’) ‘culture’; AUS på’go na ha’åni = SP este na ha’ani (< este ‘this’) ‘today’; AUS puengi = SP noche(s) (< noche[s] ‘night[s]’) ‘evening, night’; AUS sakkan = SP años (< años ‘years’) ‘year’; AUS ta’chong = SP siya (< silla ‘chair’) ‘seat’; AUS iya tåtatte = SP i ettimo (< ultimo ‘last’) ‘last’; AUS umeguihan = SP pumeska guihan (< pescar ‘hunt, fish’) ‘to fish’; AUS yan = SP kon (< con ‘with’) ‘with’; AUS yini = SP este (< este ‘this’) ‘this’;

These pairs of synonyms were presented in pairs of otherwise identical sentences. The terms under scrutiny were made to occupy the same syntactic

246 Rosa Salas Palomo and Thomas Stolz slot, i.e., care was taken that the synonyms corresponded to each other not only semantically but also in terms of word-class membership/syntactic function. In this way, we assured that grammatical differences could not interfere with the informants’ solution. An excerpt of the first two tasks on the questionnaire is shown in (5). Part 1 required of the informant to identify the more correct word (= Chamoru dinanche-ña ‘its correctness’) whereas the same set of sentences was then used for the informants’ statements about the frequency of use (= Chamoru sinesso-ña ‘its frequency’). (5)

Pairs of sentences a. Verbs: chocho vs. boka ‘eat’ Maila’ ya

ta

come and we.INCL ‘Come and let us eat!’

fañocho

(= fan- + chocho)

famboka (= fan- + boka) IMP:eat

b. Nouns: fafa’nå’gue vs ma’estro/maestra ‘teacher’ Hu

li’e’ i

I.ERG see the ‘I saw the teacher.’

fafa’nå’gue ma’estro/ma’estra teacher

For (5b), the possibility to distinguish gender on the Spanish-derived noun and thus make the reference more distinct may have had an effect on the respondents’ choice of term. This and similar things need to be looked at more closely in subsequent studies. The sentences and word-pairs were offered in the same way and order to all respondents, namely Chamoru and Spanish simultaneously (on paper). All informants had to judge the same forms. For part 3, the actual usage of 61 items was tested on the basis of translations of English sentences into Chamoru by the informants who were asked to translate the entire English sentence. Elderly speakers had more difficulties with the English sometimes i.e. their degree of bilingualism tends to be somewhat lower than the one observed with younger speakers. The two word pairs of (5) were kept in play by sentences like the one in (6). The translations were coded as 1 if the Austronesian form was chosen, as 2 if preference was given to the Spanish form and as 3 if a form different from the two previous ones was used.

Pro or contra Hispanisms

(6)

247

Translation a. Verbs (cf.[5a]) The mayor of Barrigada came to eat.

1 2 3

chocho boka ____

b. Nouns (cf. [5b]) I have a Chamoru teacher.

1 2 3

fafa ’nå ’gue ma’estro/ma’estra ____

3.3. Observations In the course of the interview sessions, some rather interesting facts came to the fore. Older participants showed a much higher awareness of the differences between Spanish and Austronesian members of a word pair whereas younger speakers (i.e. those of an age below 50) most often could not tell Spanish and Austronesian elements apart, meaning: both word forms were fully integrated Chamoru words for them10. In contrast to these younger speakers, members of the older generation were often able to make out very subtle semantic distinctions and thus argued that, on reflection, the word pairs were not always made up of fully synonymous items.11 Chances are that these erstwhile semantic differences are no longer accessible to the younger speakers of Chamoru. In general, the older speakers needed more time to work with the questionnaire. That is why the interview often extended over two sittings. In addition, part 3 proved to be somewhat challenging for many older informants as they were not always familiar with the English terms. 4.

Results

4.1. The statistics The word pairs provided the informants with exactly two options from which to choose and thus, ideally, there is a 50% chance probability that either of two alternatives is selected. However, occasionally, the respondents diverged from the prescribed path by way of introducing an unexpected third form. These unpredictable deviations account only for less than 1% of the solutions offered for the tasks in parts 1–2 and thus are statistically marginal. However, with part 3, where options other than the two

248 Rosa Salas Palomo and Thomas Stolz members of the word pairs were in principal possible, the share of deviations rose considerably to 3% of all answers. This is indicative of the difference of subjectively perceived preferences and actual language use. We have to bear this in mind when we proceed to the evaluation of the data. What is also important to note is the fact that none of the statistics below yield values which are significant at the 5%-level according to the chisquare test. Thus, we refrain from applying this test to our data. The failure to yield statistically significant values translates into a largely egalitarian status of Austronesian and Spanish items in contemporary Chamoru. In what follows, we emphasise the differences which can be observed. In (7), we give the general survey of the choices for each part of the questionnaire and the respective totals and sums. The column headed total contains the number of answers per part (and their sum) calculated according to the formula 45 word pairs (or 61 choices for part 3) multiplied by 121 informants. The percentages refer always to the share a given cell has of the total for a given row (i.e. this also applies to the bottom row labelled sum). For all percentages given in this and other tables, decimal units were truncated in order to avoid further distortion of the statistics by non-significant factors. (7)

General statistics Austronesian

Spanish

part

other

answers %

answers

%

answers

1 2 3

2,794 2,779 3,646

51 51 49

2,638 2,653 3,511

48 48 47

sum

9,219

50

8,802

48

%

total

13 13 224