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

represent respectively the weak and optionally voiced consonants, and the unvoiced and optionally long consonants) and an optional (but common) deletion of the vocalic endings of the words are relevant, along with an optional deletion of the last syllable of the grammatical words, like postpositions. Displaying these elements of word reduction and with lexical complexities, the language shows great variability, including ritual and cryptic uses (see Sherzer 1975), and beyond more obvious differences, linked to the age of the speaker, and between men and women (see Pacini 2001, 2002).

2. Phonetic and syntactic interferences A possible influence of Spanish on Kuna can be seen in two cases of phonetic variation, in the sense that two phonological rules of the regional Spanish apply also to Kuna. This coincidence can, of course, be casual. The first variable rule is the strengthening, by affrication, of postnasal /s/, as we can see in binsae ‘think’, whose pronunciation can be both [pin'sae] and [bin'sae] or [pin'tsae] ~ [bin'tsae].4 The word is likely a Spanish loanword (cf. pensar ‘think’) and we have to add that the /ns/ cluster is very rare in Kuna, avoided at the morpheme border (gunn(e) + -sa (eat + PAST)  ['guta]), but possible in external samdhi (mimmigan sate ‘without children’). On the other hand, a clearly native sound, initial [j], can be strengthened optionally like in the (widely) regional Spanish. For example, the words yer ‘very’, yaug ‘tortoise’, yala ‘side, country’ can be pronounced [jer, 'jauk, 'jala] as well as [er, 'auk, 'ala]. This (possible) Spanish pronunciation

80 Luciano Giannelli and Raoul Zamponi can be easily accepted as it is integrated in a general process of strengthening of the initial consonants, that traditionally applies to obstruents and, in particular, to stops (['kuna] or ['guna], while ['nega] ‘house, space, land’ is practically the only permissible form). In any case, the acceptance of the Spanish variation is selective. We do not know of any example of a Kuna word beginning with /j/ pronounced as [d], which is common in Spanish pronunciation in Latin America, even if it is not very frequent in Panama or perhaps less frequent than in other countries. These facts lead us to reject the possibility that the strengthened pronunciation of initial /j/ is due to Spanish pressure. But, in general, we cannot forget that every variation and every change of today’s Kuna occurs within the context of Kuna’s relationship to Spanish through a (strong) minority of bilingual speakers, even if these changes are mediated to a great extent by the islanders, or at least those not urbanized. So, if we cannot perceive exactly the motivations and the “meaning” of the quick loss of some variable pronunciations which correspond to the loss of sounds that are unknown in Spanish, we have to at least consider the pressure of Spanish. This is the case of [  :] respectively for /s e r/, in words like wisi ‘to know’, usu ‘rabbit’, degi ‘so’ or bane ‘tomorrow’ (where /e/ is respectively stressed and unstressed), ari ‘iguana’, and so on. In the case of the palatalized realization of /s/, we believe that [] had been the most common realization, which easily justifies [t] as the strong counterpart of /s/ (see wis suli [wit:uli] ‘I do not know’). The loss of [  :] from Kuna eliminates the sounds that are not present in Spanish, so that the only difference in sounds between Kuna and Spanish remains the indifference of Kuna regarding voicing, the source of many orthographical errors in Spanish (consider the written words amico for amigo ‘friend’, castar for gastar ‘expend’, and so on), and the (possible) long pronunciation of strong consonants. In other words, Kuna speakers, and we can say, the bilingual Kuna speakers, introduced and are developing a normalization of Kuna pronunciation with respect to the Spanish one, in a process of convergence. This can be very important in consonantal clusters and their constraints. The Kuna language does not present initial consonant clusters, apart from the vexata quaestio of the possible presence of the (also initial) cluster /kw/ or, alternatively, of a pair of labiovelar phonemes, weak and strong (see Giannelli 2002). Our opinion is that Kuna has lost this phoneme, so that it represents the “exceptional” situation of the presence of a lone initial cluster gw- (realized as [gw]- ~ [kw]-) in words like gwenad ‘relative, kin’, gwakwa ‘parrot’, and gwensak(a) ‘one’). On the other hand, Kuna does have the phoneme /w/ (see the minimal pairs wala ‘large place; trunk’ contrasting

Hispanisms in Kuna 81

with yala ‘land, zone, mountain’, and ua ‘fish’ vs. wa ‘smoke’) just as Spanish does (cf. huevo ‘egg’ and hueso ‘bone’). As happens in many languages, the CV structure of the first syllable of the word can be enriched internally in the present-day Kuna, while heavy syllables at the end of the word are widely present. In spite of this, there are Kuna scholars who are working to demonstrate that the syllabic structure of Kuna, and its word structure, was, or – more relevantly – is subjacently CVCV(CV). As regards the current conditions, this is possible because of the “living etymology” current in Kuna (see Wagua and Green 2004), thanks to the current use of long and short forms of the same word, for simple terms, but also complex and “hard” terms, as: (i) The Kuna language has reduced forms of words and a good number of the reduced forms have a complete form well known by every speaker, e.g. nega ~ neg ‘house’. (ii) In many cases, the reduced forms – and the derived compounds – have a complete counterpart which is known only by a few speakers, e.g. ablis ‘blood’ from abe ‘blood’ + nisa ‘liquid, juice’. (iii) Of some reduced forms, only a handful of people (generally the saila, political and spiritual leaders) know the full version – see argana ‘hand’, commonly argan, and ogoba ‘coconut’, commonly ogob – while some forms can only be reconstructed with difficulty. One of these is neggwebur, interpreted as a word that was originally the sentence nega gue bula ‘the houses stay together’, with the typically polysemic word bula used as an adverb (‘all within’) and also as a noun (‘gathering’) to produce the meaning of ‘weed’ (maleza in Spanish). From this, we can observe in the Kuna culture a continuous and tremendous interest in discovering etymologies, not in any way blocked by real difficulties in providing them or by weak evidence. In any case, we can say that clusters Cr are not original and are (were) to be avoided, as we can see in loanwords like nibir ‘pound, scales’ (from Spanish libra ‘pound’), badir ‘priest’ (from Spanish padre), bate ‘dish’ (from English plate), or reduced compounds like ablis ‘blood’ from abe-nisa, and the problematic bergwable ‘all’. The idea that Cr clusters are not indigenous can be accepted, while this seems impossible for rC clusters, which are present in too many common roots like garba ‘basket’, burba ‘soul’ and birga ‘year’. rC clusters are also found in lexemes like urba ‘younger brother’ and ‘under’, very probably ulu (~ ur) + ba(li) ‘under the canoe’, as well as loanwords like garda ‘paper, letter, book’ from Spanish carta ‘paper’, and garbinder ‘carpenter’ from

82 Luciano Giannelli and Raoul Zamponi English. Cr clusters remain rare in Kuna, but current Kuna presents, in its roots and in many bimorphemic words, the same constraints on consonant clusters that occur in Spanish, if we also consider the avoiding of some clusters as in anmar ‘we’ pronounced as ['am:ar], abnisa reduced to [a'blisa], gun-le ‘be eaten’ with a taxative pronunciation ['gul:e], agsu ‘brother-in-law of a woman’ commonly pronounced ['ajsu]. (Different realizations have to be considered residual.) In other words, the allowed syllabic structure with consonant clusters are both CVr
CV and CV
CrV (with the important exception of some clusters at some morphological border, like dadgan and babgan, that is dada + -gana (grandfather + COLL5) ‘grandfathers’, and baba + -gana (father + COLL) ‘fathers’, whose respective pronunciations are ['datkan] ~ ['tatkan] and ['babgan] ~ ['pabgan]), or gobsa ‘I wept’ (gob(e) ‘weep’ + -sa ‘past’). We can add that a reconstruction of the original conditions shows that heavy syllables existed only if it were possible for the syllable to have a coda, while a heavy onset was impossible. Today these conditions have changed, and the elimination of [gw] as a phoneme can be seen as obeying this shift. This “silent pervasion” or – in the other direction – this “underground” hauling process and drift, through the convergence, can be found – or seen – also in idioms and pragmatic uses as well as in syntax. We find it difficult, for example, not to consider bane malo ‘good bye, see you tomorrow’ as a reproduction of the Spanish hasta mañana; another example is nued(i) ‘good’, that, with the meaning ‘all right’, is too similar to Spanish ¡bueno! to not consider it as a calque. The use of the future tense as a condescending or educated form of imperative (so that one can say dago(e), literally ‘you will come’ instead of dage to say ‘come!’) descends from Spanish, in the opinion of many (educated) speakers. In syntax, we find (in an educational book, whose author is a linguistically aware native speaker) the sentence in (1) translated as mi sobrina está parada ‘my niece is standing’. (1)

an sia gue gwichi 1PRO niece be standing ‘My niece is standing’

We cannot forget that, in everyday language, this sentence is often uttered – and seems more “correct” – without gue: an sia gwichi, as gwichi means ‘standing’. Nor can the language be defined as a copular one, in spite of some possible and sometimes necessary copular uses of gue. The an sia gue gwichi version seems a perfect calque of the organization of the Spanish sentence, as a heavy marker of the Spanish linguistic hegemony also on bilingual speakers engaged in the defence and development of the native language.

Hispanisms in Kuna 83

The last case we discuss is highly problematic, as the interpretation of the situation is unclear in basic present-day Kuna. (2)

segar an nika / sate match 1PRO have / have.not ‘I have/have not got matches’

In this example, the element indicating possession (nika), or its contrary (sate), is posed after the pronoun an ‘I’ just like a postpositional element. In spite of this all, we are not sure whether or not nika and sate are verbs. In common use, segar an nika/sate can alternate freely with an segar nika/sate, that seems to reproduce a simple SOV structure, without any particular significance. Of course, one of the two constructions depend on an extraposition or a raising process (very frequent in Kuna in a stylistic-pragmatic function), but here, understanding which is the basic form and the derived one is not easy. An auxiliary in the past is required, as in mani an nik gu-sa (money 1PRO have be-PAST) ‘I had money’, and these past forms seem to indicate only a non-verbal nature of nika and sate. Discussing the whole, very complex matter is, of course, impossible here (see Arias Díaz, Giannelli, and Zamponi forthcoming), and in any case we can say that a class of elements (that we can define as verbs) occur in past tense clauses in a similar way, instead of by the simple addition of the suffix -sa, as in an obe + -sa  an obsa ‘I washed myself’. What we can say is that the use of the probably most recent form, an mani nik gu-sa and of an mani nika ‘I have money’ is present in a scenario where the other language, Spanish, uses a transitive construction (tengo dinero ‘I have money’) for the possessive sentence.

3. Lexical borrowing At the lexical level, we may observe in Kuna a significant (but not particularly pervasive) presence of elements borrowed from Spanish within the classes of nouns, numerals, and verbs. As we may expect, the majority of these elements are cultural borrowings, i.e. terms for items or concepts originally alien to Kuna culture that came into the language along with their denotata. Among the apparently few “unnecessary” borrowings, i.e. loanwords for already known items or concepts, we may recall benao ‘deer’, gwebu ‘egg’, gwendo ‘story, tale’, and neche ‘milk’ (Sp. Venado, huevo, cuento, leche) – which are in competition, respectively, with the properly indigenous equivalents goe, allakwa, soged (from soge ‘say, tell’), and nunis (nu ‘breast’ + nis(a) ‘juice’). Olo ‘gold’ (Sp. Oro) has now completely

84 Luciano Giannelli and Raoul Zamponi replaced the traditional term gungi (Holmer 1952: 60), in all probability also a loanword (cf. Imperial/Cuzco Quechua qulqi ‘silver’6). While many loanwords have been adapted to Kuna phonology, others have not and retain their foreign shapes, albeit with minor segmental and suprasegmental modifications. The adaptation of Spanish loans fundamentally consists of (a) the replacement of the sounds foreign to Kuna (see Table 1) with the closest available sounds (in terms of features and acoustic properties), the vast majority according to the series of correspondences that appears in Table 2, and (b) the replacement of consonant clusters not permitted by the phonotactic rules of Kuna with permitted sequences or simple consonants (see Table 3). Table 1. Kuna and Spanish phoneme inventories In Spanish and Kuna Vowels

Only in Spanish

Only in Kuna

p, t, k

b, d, g

p:, t:, k:

s

f, , x

s: ([tS])

i, e, a, o, u

Stops Fricatives Affricates

t

Nasals

m, n

Trills

r

r:

Laterals

l



Glides

m:, n: l:

j, w

Table 2. Adaptation of foreign phonemes from Spanish loanwords Spanish

Kuna

Examples

b

p ()

abena ‘oats’, bunidu ‘bonito’ basu ‘drinking glass’, obisbo ‘bishop’ (Sp. Avena, bonito, vaso, obispo)

[d]

t ()

[]

t (), r, Ø

Dios ‘God’, dotor ‘doctor’, dordo ‘curtain, net’, dosin ‘dozen’ (Sp. Dios, doctor, toldo, docena) udio ‘Jew’, argar ‘mayor’, benao ‘deer’ (Sp. Judío, alcade, venado)

d

g

k ()

sigali ‘cigar(ette)’, ganso ‘hook’, iglesia ‘church’, gobierno ‘government’ (Sp. Sigarillo, gancho, iglesia, gobierno)



s

serbesa ‘beer’, mansan ‘apple’, oros ‘rice’, sabad ‘shoe’ (Sp. Cerveza, manzana, arroz, zapato)

Hispanisms in Kuna 85 Spanish

Kuna

Examples

x



wes ‘judge’, efe ‘boss’, eneral ‘general’, unio ‘June’ (Sp. Juez, jefe, general, junio)

r:

r

radio ‘radio’, rei ‘king’, bari ‘barrel’, oron ‘rum’ (Sp. Radio, rey, barril, ron)a

y, l, b

esdambiya ‘stamp’, yabi ‘key’, badila ‘watermelon’, basdia ‘sweet, pastille, tablet’, mardo ‘hammer’, sigali ‘cigar(ette)’, bote ‘bottle’ (Sp. Estampilla, llave, patilla, pastilla ‘pastille, tablet’, martillo, sigarillo ‘cigarette’, botella)c



a

b

c

Since the language does not admit /r/ in word-initial position in its indigenous vocabulary, Spanish words with an initial /r/ are realized in Kuna, in some cases, with a prothetic vowel that harmonizes with the following vowel (cf. also the personal names Araur, Aramon, and Oroberdo; Sp. Raúl, Ramón, and Roberto).  if // occurs in final open syllable. The vowel that precedes or follows // may also be deleted. Note the Spanish Americanism patilla ‘watermelon’.

Table 3. Adaptation of some aberrant consonant clusters from Spanish loanwords Spanish

Kuna

Example(s)

kt

t: ()

dotor ‘doctor’(Sp. Doctor)

mp

p ()

gaban ‘bell’ (Sp. Campana)

nx

s

naras ‘orange’ (Sp. Naranja)

ngl

l

iles ‘Englishman’ (Sp. Inglés)

lC

rC

dordo ‘curtain, net’, garson ‘trousers’, grub ‘club’, sorda ‘soldier, companion, partner’ (Sp. Toldo, calzón, club, soldado ‘soldier’)

Cr

CVra

dorombo ‘top (toy)’, nibir ‘pound, scales’, badir ‘priest’ (Sp. Trompo, libra ‘pound’, padre)

a

The V tends to harmonize with either the preceding or following V.

The replacement of sounds sometimes concerns vowels and consonants that Spanish and Kuna share. Spanish words with /e/ can sometimes appear with /i/; see dosin ‘dozen’ from docena, yabi ‘key’ from llave, simendo ‘cement’ from cemento, and badir ‘priest’ from padre. A loanword with /o/, especially in word-final position, can surface with a final /u/; see bunidu ‘bonito’ from bonito, badu ‘duck’ from pato, basu ‘drinking glass’ from vaso, gesu ‘cheese’ from queso, and gwebu ‘egg’ from huevo. A loanword with the affricate [t], which is the only manifestation of /s:/ in Kuna, as indicated

86 Luciano Giannelli and Raoul Zamponi above, can be realized with a short /s/; see musila ‘knapsack’ from mochila and also – not unexpectedly since /s:/, as well as all long consonants, is only intervocalic in Kuna – sino ‘Chinese man’ from chino and sogolade ‘chocolate’ from chocolate. Finally, some Spanish loanwords with /r/ and /l/ have realizations in Kuna with /l/ and /n/, respectively; see olo ‘gold’ from oro, bandela ‘flag, banner’ from bandera, nibir ‘pound, scales’ from libra ‘pound’, neche ‘milk’ from leche. As a rule, however, Spanish loans with (sequences of) segments that have Kuna equivalents, exhibit no significant phonetic adjustments; see arbon ‘harpoon’, biba ‘pipe’, de ‘tea’, dinda ‘ink’, mayo ‘May’, and semana ‘week’ for example (Sp. Arpón, pipa, té, tinta, mayo, semana). Words borrowed from Spanish that contain /f/, in actual occurrences, maintain this consonant, even though foreign to Kuna, e.g.: febrero ‘February’, fiesta ‘holiday, festivity, feast’, fudbol ‘football’ (Sp. Febrero, fiesta, fútbol), and Fransia dule, which has now replaced the more assimilated barses ‘French man’ (Sp. Francés), as originally recorded by Puig (1944), with /f/ converted into /p/ (cf. also gabi ‘coffee’ from English coffee). At the semantic level, Spanish loanwords can be divided into those that have (largely) maintained their original meaning and those that have not. The latter, typically, have an extended meaning being used, for example, to refer to the particular object designated by the word in the source language and also the whole class of which the original referent is a part or member (e.g. sigali ‘cigarette, cigar’ from sigarillo ‘cigarette’, sisa ‘alcoholic beverage’ from chicha ‘maize liquor’), and/or what the object is made of (e.g. garda ‘letter, book, paper’ from carta ‘letter’ and isbe ‘mirror, glass, crystal’ from espejo ‘mirror’). Two interesting cases of semantic widening can be observed in nibir ‘pound’ also meaning ‘scales’, from Spanish libra ‘pound’, and sorda ‘soldier’ also meaning ‘companion, comrade, partner, consort’, from Spanish soldado ‘soldier’. The class of nouns is that which, more than the other word classes, accepted lexical elements of Spanish. The borrowed nouns belong to specific semantic fields, the most significant of which seem to concern (A) objects, (B) materials, (C) foods and drinks, (D) flora, (E) fauna, (F) institutions, (G) Christianity, (H) professions/jobs, and (I) time. A short list of these is given in Table 4.

Hispanisms in Kuna 87 Table 4. Some Spanish loanwords in Kuna Meaning

Spanish

Kuna

Meaning

campana gaban

‘bell’

pato

badu

‘duck’

llave

yabi

‘key’

venado

benao

‘deer’

motor

modor

‘motor’

zapato

sabad

‘shoe’

Spanish A

B

espejo

D

E

b c d e

f

‘glass’

garda

‘paper’

oro

olo

‘gold’

b

G

blasdigo ‘plastic’

gobierno gobierno ‘government’ ley

lei

‘law’

policía

bolisia

‘police’

rey

rei

‘king’

ángel

anger

‘angel’ f

cruz

nagrus

‘(holy) cross’

chocolate sogolade ‘chocolate’

hostia

osdia

‘host’

pastilla

basdia

‘sweet (n.)’

misa

misa

‘mass’

té

de

‘tea’

doctor

dotor

‘doctor’

H

vino

bino

‘wine’

enfermero enfermero ‘nurse’

arroz

orosc

‘rice’

juez

wes

‘judge’

mango

mango

‘mango’

maestro

maesdro

‘schoolteacher’

martes

mardes

‘Tuesday’

mayo

mayo

‘May’

manzana mansan

‘apple’

naranja

‘orange’

gallina micho

a

isbe

F

a

carta plastica C

Kuna

naras gannir mis(i)

‘hen’ (met.) ‘cat’

e

I d

miércoles miergoles ‘Wesnesday’ semana

semana

‘week’

isbe ‘glass, crystal, mirror’. carta ‘letter’; garda ‘paper, letter, book’ aro in the Río Caimán dialect. galina in the Río Caimán dialect. The Spanish word micho (or michino), in all probability, reached Kuna via one or more indigenous South American languages (among these, apparently, Imperial/Cuzco Quechua, where michi means ‘cat’). A form mitSi or misi for ‘cat’ is also widespread in numerous native languages of the western sector of South America. The initial syllable na of this name, in all probability, is the Spanish (feminine singular) article la ‘the’ or una ‘a’.

The class of objects is the sector of Kuna nominal lexicon most enriched by borrowings, and not solely from Spanish. It is precisely this lexical category that shows the majority of English loanwords which can be found in Kuna, whose number is however rather limited; see, for example, bensil ‘pencil’ and Grismas ‘Christmas’, coexisting with the Hispanisms labis and Nabidad

88 Luciano Giannelli and Raoul Zamponi (Sp. lápiz, Navidad), sugar ‘sugar’ which forced itself onto the Hispanism achugwa (Sp. azúcar; Prince 1913: 489), siman or simen ‘cement’ on which the Hispanism simendo (Sp. cemento) forced itself, gidar ‘guitar’, mani ‘money’ (and also ‘silver’), nusbebar ‘newspaper’, wachi ‘watch’ (and also ‘hour’ and ‘time’), and mergi ‘American person’.7 Some Spanish loanwords denoting units of measurement are commonly used as nouns, which are also “classifiers for themselves”, lexicalizing, in an NP with a numeral (or with the interrogative quantifier bigwa ‘how many’), both the numeral classifier and the head noun (see Simpson 2005: 831), analogous to the native Kuna nouns with similar denotata (see Arias Díaz, Giannelli, and Zamponi forthcoming: sec. 5). This is an important testimony for the complete lexical assimilation of the loanwords in question. (3)

san nibir bo meat pound/CL two ‘two pounds of meat’

(4)

mor yarda bo cloth yard/CL two ‘two yards of cloth’

(5)

gwebo dosin bo egg dozen/CL two ‘two dozens of eggs’

(6)

mani bo five.cents/CL two ‘ten cents (of balboa/dollar)’

Taking into consideration that Spanish numerals are those which are used today by the Kuna in order to count, as indicated above, we may signal the presence in the present-day (rarely used) Kuna numeral system of the Hispanisms sero ‘zero’(Sp. cero), siendo ‘hundred’ (Sp. ciento), mili (~ mir) ‘thousand’ (Sp. mil), and miyon ‘million’ (Sp. millón). The use of siendo and mili imposes, we have to say, a violence on the Kuna numeral system, which is fundamentally vigesimal and, from 400 onwards, uses a base of 400 (dula classified by the numeral classifier wala- ~ war- ‘(ob)long’), acting as a basis for, respectively, the hundreds and the thousands (see Table 5).

Hispanisms in Kuna 89 Table 5. A few traditional and non-traditional numerals compared

100

400

1000

2000

Traditional numeral

Non-traditional numeral

dul-atar twenty-five [20  5] dula war-gwen twenty CL:OBL-one [400  1] dula war-bo gaka dul-ambe four.hundred CL:OBL-two CONN twenty-ten [(400  2) + (20  10)] dula wal-atar four.hundred CL:OBL-five [400  5]

siendo gwen hundred one [100  1] siendo bake hundred four [100  4] mili gwen thousand one [1000  1] mili bo thousand two [1000  2]

The Hispanisms in the domain of verbs are extremely rare, in particular, in the speech of old people. The existing “loanverbs” are uninflected forms accompanied by the indigenous Kuna verb sa(e) ‘do, make’, which follows the original Spanish verb (in the infinitive) and is inflected for tense-aspectmood: an basear sae ‘I go for a walk’ (Sp. pasear), an grabar saoe ‘I will record’ (Sp. grabar), an delefonar sasa ‘I telephoned’ (Sp. telefonear).8 A strong resistance opposes Spanish lexical borrowing in Kuna. This has meant that relatively few new (European) words were totally accepted, including terms common in many other native American languages as a consequence of the penetration of objects and Western cultural models in the respective societies. A notably heightened productiveness of the Kuna lexicon in supplying labels to new meanings, which manifests itself in a great capacity for neologisms (see Table 6) and also in a more moderate capacity of assuming indigenous terms into new meanings (see Table 7), inevitably relates to this resistance of Kuna to borrowings. Table 6. Neologisms for non-indigenous things Meaning(s)

Neologism

Literal meaning

‘film’ ‘steamer’ ‘lemon’ ‘potato’ ‘photo’ ‘patrol, kerosene, Diesel’

garda binnied so ur naras sole mergi gwalu wagar wilup gwallu di

‘paper which changes’ ‘fire ship’ ‘sour orange’ ‘American sweet potato’ ‘face print’ ‘oil/fat water’

90 Luciano Giannelli and Raoul Zamponi Table 7. Indigenous words for non-indigenous things Word

Traditional meaning(s)

Present-day additional meaning(s)

duba gala gingi gwallu

‘thread, rope, string’ ‘bone, pipe, whistle, flute, music’a ‘bow’ ‘fat, oil, lamp, light’

moli

‘tapir’

‘(electric) cable’ ‘radio, record player’ ‘firearm’ ‘(electric) lamp, bulb, chandelier, electric light’ ‘horse, donkey, bovine’

a

Pipes, whistle, and flutes were originally made of bone

Interestingly, the still flourishing Kuna language shows a strong resistance to the Spanish pressure at all its levels, we have to say, in spite of ever greater numbers of bilingual speakers. This does not mean that there is a total impermeability to the influx of the language that has, in this case, a hegemonic role. As in other situations, also those far from the Amerindian context, avoiding a manifest absorption of foreign elements does not preclude the hegemonic language from influencing the language at a deeper and less evident level.

Notes 1.

2.

3.

4.

Even though this article is a joint production by the two authors, in the present edition, sections 1 and 2 are to be attributed to Luciano Giannelli and section 3 to Raoul Zamponi. Within this family, Kuna does not seem to be closely related to any other particular language. A recent study on Chibchan comparative grammar (Quesada 1999) highlights that Chibchan languages spoken in Central America are differentiated from those of Colombia by a series of areal features. In light of this work, Kuna figures as a “Central American isthmian language of the Chibchan family in the Intermediate Area” (see also Quesada 2000: 8–10). Other works dedicated to Kuna phonology include Branca (1973b), Sherzer (1973, 1992, 1997), and Baptista and Wallin (1974), this last describing briefly the Panamanian continental varieties (Bayano valley). Kuna words are transcribed according to the principles of the official Kuna orthography (see Wagua and Green 2004). In this orthography, as indicated above, render /p t k/,

represent /p: t: k:/, denotes /s:/ (= [t]), is assigned to [j], while the other graphemes have their traditional IPA value. (Two adjacent identical consonant graphemes indicate a long consonant.)

Hispanisms in Kuna 91 5.

6.

7.

8.

1 = first person, C = consonant, CL = classifier, COLL = collective, CONN = (numeral) connective, F = feminine, met = metathesis, NEG = negative, NOM = nominalizer, OBL = (ob)long, PRO = pronoun, SG = singular, SOV = subjectobject-verb, Sp = Spanish, V = vowel. The Panamanian isthmus, since the first half of the 16th century, has acted as a crossing point for the precious metal route from Peru to Spain (Squillacciotti 1998: 19). Within the category of names for new objects certain Gallicisms (perhaps terms of Lesser Antillean Creole French) could have penetrated, one of which appears to be muswe ‘necktie, (hand)kerchief’ from mouchoir. Cf. the analogous treatment of verbs borrowed from Russian in Udmurt, a Finno-Ugric (Uralic) language of the region of the Urals (Winkler 2001: 56).

References Arias Díaz, Gervasio, Luciano Giannelli, and Raoul Zamponi forthc. Morfologia della lingua cuna (dule gaya). Munich: Lincom Europa. Baptista, Priscilla M. and Ruth Wallin 1974 La jerarquía fonológica del Bayano Cuna. In Lenguas de Panamá. Tomo I: Sistemas fonológico, Patricia Baptista (ed.), 7–16. Panama City: Instituto Nacional de Cultura and Instituto Lingüístico de Verano. Berengueras, José María 1934 Rudimentos de la gramática karibe-kuna. Panama City: Tipografía Católica. Boyd, Hedyx et al. 2003 Anmar gaya narmaked absoged durdakmala. Panama City: Ministerio de Educación. Branca, Gabriel B. 1973a An introduction to a linguistic description of the San Blas Cuna dialect. These presented to the Faculty of the English Dept. of the University of Panama. Panamá. 1973b Reglas morfonológicas para la asimilación de las consonantes en la lengua cuna. In Actas del IV Simposium Nacional de Arqueología, Antropología y Etnohistoria de Panamá. Panama City: Instituto Nacional de Cultura, 177–183. Erice, Jesús 1980 Gramática de la lengua kuna. Panama City: Impresora de La Nación. 1982 Diccionario de la lengua kuna. Panamá: Editorial Mariana Arosemena. Gassó, Leonardo 1908 Gramática karibe-kuna. Barcelona:Tipografia Católica.

92 Luciano Giannelli and Raoul Zamponi Giannelli, Luciano 1999 Abia Yala Inmargan – Americana: studi intorno alle lingue native di un antico/nuovo mondo. Siena: Protagon Editori Toscani. 2000 Breve profilo morfologico della lingua cuna (dule gaya); con un’ appendice di Réuter Orán Bodin. http://www.unisi.it/ricerca/centri/ cisai/index.htm, Linguistico, Lingua Cuna. 2002 Los problemas de la fonología cuna In Fonologia della lingua cuna (dule gaya), Luciano Giannelli, Giovanna Marotta, and Beatrice Pacini (eds.), 3–33. Munich: Lincom Europa. 2003 Per un sistema ortografico cuna (dule) oltre l’inventario alfabetico. In Scrittura in Centroamerica, Luciano Giannelli, and Maurizio Gnerre (eds.), 3–17. Siena: Università degli Studi di Siena, CISAI. Giannelli, Luciano, Giovanna Marotta, and Beatrice Pacini (eds.) 2002 Fonologia della lingua cuna (dule gaya). Munich: Lincom Europa. Holmer, Nils M. 1947 Critical and comparative grammar of the Cuna language. Göteborg: Etnografiska Museet. 1951 Cuna chrestomanthy. Gothenburg: Etnografiska Museet. 1952 Ethno-linguistic Cuna dictionary. Gothenburg: Etnografiska Museet. Ibelele 1981 Kuna-alfabeto. E burba oduloedigar. [Usdup/Ustupu]. Llerena Villalobos, Rito 1986 Lingüística amerindia: fonología de la lengua kuna. Lingüística y Literatura 10: 31–53. 1987 Relación y determinación en el predicado de la lengua kuna. Bogotá: Universidad de los Andes. Montalván, Ana 2001 [1973] Interpretación fonológica de las consonantes oclusivas geminadas en la lengua kuna, y la definición del alfabeto kuna. In Studi americani, Luciano Giannelli (ed.), 31–53. Siena: Università degli Studi di Siena, CISAI. Montalván, Ana and Doris Rojas 2000 Panamá y sus lenguas indígenas. Siena: Università degli Studi di Siena, CISAI. Orán Bodin, Réuter 1992 Ue an ai: libro de lectura para niños kunas. Primer grado. Panama City: ADS Publisher. 1994 Ue an ai: libro de lectura para niños kunas. Nba 2°. Panama City: ADS Publisher. Pacini, Beatrice 2001 La variabilità sociolinguistica in lingua cuna. Rivista Italiana di Dialettologia 25: 327–360

Hispanisms in Kuna 93 2002

Un primo approccio quantitativo alla fonologia cuna. In Fonologia della lingua cuna (dule gaya), Luciano Giannelli, Giovanna Marotta, and Beatrice Pacini (eds.), 49–80. Munich: Lincom Europa. Prince, J. Dyneley 1912 Prolegomena to the study of the San Blas language of Panamá. American Anthropologist 14, n.s.: 109–126. 1913 Grammar and glossary of the Tule language of Panamá. American Anthropologist 15, n.s.: 480–528. Puig, Manuel María 1944 Diccionario de la lengua caribe cuna. Panamá: La Estrella de Panamá. Quesada, J. Diego 1999 Chibchan: with social reference to participant-highlighting. Linguistic Typology 3: 59–108. 2000 A grammar of Teribe. Munich: Lincom Europa. Sherzer, Joel 1973 Interpretación sociolingüística de la fonología cuna. In Actas del IV Simposium Nacional de Arqueología, Antropología y Etnohistoria de Panamá, 187–195. Panama City: Instituto Nacional de Cultura. 1975 A problem in Cuna phonology. The Journal of the Linguistic Association of the Southwest 1: 45–53. 1978 Kuna numeral classifiers. In Linguistic and literary studies in honor of Archibal A. Hill, Mohammad Ali Jazayery, Edgar C. Polomé, and Werner Winter (eds.), 331–337. The Hague: Mouton. 1992 Formas del habla kuna. Quito: Abya Yala. 1997 Guide to Kuna pronunciation and grammar. In The art of being Kuna, Mari Lyn Salvador (ed.), 131–134. Los Angeles: UCLA Fowler Museum of Cultural History. Simpson, Andrew 2005 Classifiers and DP structure in Southeast Asia. In The Oxford handbook of comparative syntax, Guglielmo Cinque and Richard Kayne, (eds.), 331–337. Oxford: Oxford University Press. Smith, Lino et al. 1975 Informe de la reunión de la Subcomisión Técnica de la Comisión Especial del Ministerio de Educación sobre un alfabeto para la lengua cuna. Hombre y Cultura 3: 185–191. Squillacciotti, Massimo 1998 I Cuna di Panamá: identità di popolo tra storia e antropologia. Turin: L’Harmattan Italia. Wagua, Aiban 1984 Corso elementare di lingua cuna. Siena: Università di Siena. 2001 Corso elementare di lingua cuna =Curso elemental de lengua cuna. 2nd online edition, edited and commented by Luciano Giannelli. www.unisi.it/ricerca/centri/cisai/index.htm, Linguistico, Lingua Cuna.

94 Luciano Giannelli and Raoul Zamponi Wagua, Aiban and Abadio Green (eds.) 2004 Bases de lecto-escritura kuna: aspectos unificados bajo la dirección y decisión de los Congresos Generales Kuna, Onmaked Dummad Namaked – Onmaked Dummad Sunmakaled. Kuna Yala. Winkler, Eberhard 2001 Udmurt. Munich: Lincom Europa. Zamponi, Raoul 2000 Postilla. In Panamá y sus lenguas indígenas, Ana Montalván, and Doris Rojas (eds.), 83–94. Siena: Università degli Studi di Siena, CISAI.

Spanish lexical borrowing in Imbabura Quichua: In search of constraints on language contact Jorge Gómez Rendón

1.

Introduction

Imbabura Quichua1 (henceforth IQ) is one of several Ecuadorian dialects of the Quechua language family spoken in the Andean province of Imbabura. The permanent contact of IQ and local Spanish during the last 450 years has resulted in a significant influence from Spanish on the lexicon and the grammar of the native language. The present article deals with Spanish lexical borrowing in IQ. It explores the statistical results from the investigation of a corpus of spontaneous speech collected during fieldwork in Ecuador (2003–2004). The purpose is to identify possible linguistic constraints by analyzing statistical frequencies in the borrowing of major word classes. The present research is framed in the parts-of-speech theory developed by Hengeveld (1992) and Hengeveld et al. (2004). A more extensive analysis of lexical and grammatical borrowing in several Ecuadorian dialects from a cross-linguistic perspective is presented in Bakker et al. (2008). Section 2 makes the reader familiar with the sociolinguistic aspects of SpanishQuichua contact in Imbabura. Section 3 outlines the theory of parts of speech and develops individual hypotheses according to the typological classification of the languages involved in the borrowing process. Section 4 deals with the phonological and morphological changes to which Spanish lexemes are subject in the process of borrowing. Section 5 discusses the statistics concerning the borrowing of lexical classes in the light of Hengeveld’s theory and the implications for lexical borrowing. Section 6 offers some conclusions and suggestions for further research. 2.

Sociolinguistic aspects of language contact in Imbabura

2.1. Background Although the Northern Andes were subject to the rule of the Incas for only sixty years before the Spanish conquest, by the second half of the seventeenth-century Quechua had replaced the pre-Inca language of the Cara

96 Jorge Gómez Rendón people who inhabited the territory of the present province of Imbabura (Jijón y Caamaño 1941; Salomon and Grosboll 1986). Scholars have amply discussed the reasons for such a rapid expansion, identifying mainly two major factors. On the one hand, as it becomes evident from early historical records, Quechua was spoken in the (northern) Andes well before the Inca invasion, being introduced by long-distance traders from Chincha in central Peru (Torero 2003: 93–105). On the other hand, the use of Quechua as a lingua franca for the evangelization of native peoples during the first two centuries of Spanish colonization encouraged the language shift in non-Quechua speech communities. A more detailed history of Quechua in the Andes from the Spanish conquest to the present has been presented elsewhere (Bakker et al. 2008) and will not be addressed here. I turn now to describing the sociolinguistic profile of the communities where the data for this research were collected.

2.2. The communities of this study As far as the number of speakers is concerned, the province of Imbabura ranks second among the eleven Quichua-speaking provinces of Ecuador (Haboud 1998: 91–92). At the same time, Imbabura shows the largest number of bilingual Quichua-Spanish speakers in the country (Büttner 1993: 48–49). Imbabura Quichua speakers are deeply concerned with language maintenance and have been struggling for an egalitarian use of Quichua vis-à-vis Spanish in the last decades. The data for this research were collected in three Indian communities in Imbabura: Angla, Casco Valenzuela and El Topo. The communities are located on the skirts of the Imbabura Mountain in the Northern Andes of Ecuador, at scarcely forty minutes by bus from the cities of Otavalo and Ibarra and two times the same distance from the capital. The population of the three communities altogether amounts to some 1000 speakers. Historically, Angla was an Indian town separated from the surrounding estates, while the other two communities were located within Hacienda El Topo. Regardless of their adscription, the three communities delivered their workforce to the neighboring estates under different labor systems. Angla is the least traditional of the three communities as it maintains stronger links with the Mestizo society and shows higher levels of bilingualism. It is also known for speaking chapu-shimi, a mixed language of Spanish and Quichua2 that is being disseminated to neighboring communities through contact. On the contrary, Casco Valenzuela and El Topo have become ac-

Spanish lexical borrowing in Imbabura Quichua

97

tively involved in the Spanish-speaking society only in recent times. Casco Valenzuela and El Topo show higher levels of Quichua dominance over Spanish (Büttner 1993: 80–117)3, with gender and age playing an important role. There are more women (92.6%)4 than men (57.4%) who speak Quichua while the percentage of Quichua-speaking elders doubles that of youngsters. There appears to be a strong process of Hispanisation in the three communities as well as a steady increase of bilingualism across generations. The social and economic conditions underpinning this scenario are well known: on the one hand, the existence of impoverished Indian peasants living of subsistence agriculture on small, usually unproductive pieces of land, who are forced to hire their workforce in the big cities in order to supplement their family income5; on the other hand, the aggressive assault of the Spanish-speaking society on traditional Quichua communities through media and schooling6.

3.

Methodology

3.1. Data collection The data were collected during a four-month stay in the province of Imbabura, Ecuador. The author benefited from previous fieldwork in the communities, where he conducted research on a local mixed variety (Media Lengua). Although the data were collected in normal conversational settings, they were produced by different speakers and belong to different speech genres. Seven informants produced individual samples in seven recording sessions. All informants were in their late fifties or sixties at the time of recording and represented the most conservative speech in their communities. Only one woman (69) is represented in the samples. Three informants were born in Casco Valenzuela, two in Angla, and two in El Topo. For all the speakers, the birth community is the same as the community of residence7. In five cases the recording setting involved other people who were present but did not participate. The samples cover different topics ranging from family and community issues to more specific topics such as local festivals and life in the former haciendas and the parish town of San Pablo. Interethnic relations between Mestizos and Indians are a recurrent topic. The informants set their own narrative agenda and showed a strong involvement with their topics, as it is typical of testimonial narratives8.

98 Jorge Gómez Rendón 3.2. Data processing As for the processing of data, recordings were transcribed phonemically, with exclusive consideration of distinctive features. In the second step Spanish borrowings were identified between slashes, without including Quichua morphology. In marking the Spanish lexicon on Quichua texts, strains of Spanish speech that were considered code switches were not included9. The next step was to run a specially designed computer program on the parsed transcripts in order to produce statistical information with respect to the number of tokens and types for both languages (Quichua and Spanish) and the morphological context in which they appear (i.e. whether a Spanish borrowing occurs with other borrowings or with Quechua words). In the next stage, types and tokens10 were classified according to their parts of speech in Spanish and their syntactic position in Quichua. As no univocal correspondence between parts of speech and syntactic position exists across languages, a random sample of 1000 tokens including borrowed and autochthonous items were analyzed for their syntactic position in order to correlate the parts-of-speech distribution of borrowed items with the general distribution of parts of speech in the target language. A detailed visualization of the processing tasks is offered in Bakker et al. (2008).

3.3. The theory of parts of speech and the analysis of data The theoretical framework of this research is the theory of parts of speech as elaborated by Hengeveld (1992) and Hengeveld et al. (2004). Hengeveld’s theory is concerned only with content words, including nouns, verbs, adjectives and manner adverbs11. Hengeveld defines parts of speech primarily on syntactic grounds, with the phrase as the basic syntactic unit. Phrases can be referential (noun phrase) or predicational (verb phrase). Each phrase is composed of two slots, one for heads and one for modifiers. The four possible combinations for English are shown in Table 1. Table 1. Lexemes and syntactic slots

Referential Phrase Predicational Phrase

Head

Modifier

Noun Verb

Adjective (Manner) Adverb

Spanish lexical borrowing in Imbabura Quichua

99

In the frame of this theory, parts of speech are defined according to their non-predicative uses. Thus, “verbs are characterized by the fact that they have no non-predicative uses, i.e. they can be used predicatively only. Nonverbal lexemes, on the other hand, may have additional predicative uses, but their defining use is a non-predicative one” (Hengeveld et al. 2004: 6). A survey of typologically different languages spoken in several areas of the world showed that languages have specific distributions of lexical classes in what Hengeveld calls ‘parts-of-speech systems’ (cf. Table 2). While some languages have one lexical class for each syntactic slot (differentiated systems), others have more than one (flexible systems), and still others lack lexical items to fill syntactic slots (rigid systems). English is a typical example of a differentiated system in which nouns, verbs, adjectives and manner adverbs can be clearly identified in specific slots, as opposed to, for instance, Samoan in which virtually any lexeme can be used in any syntactic position. Table 2. Parts of speech systems Part of speech system Flexible

Head Pred. Phrase

Head Ref. Phrase

Modifier Modifier Ref. Phrase Pred. Phrase

1 2 3

Verb Verb

Lexeme Non-verb Modifier Noun

Differentiated

4

Verb

Noun

Adjective

Manner Adverb

Rigid

5 6 7

Verb Verb Verb

Noun Noun –

Adjective – –

– – –

There also exist languages that fit into two adjacent types and constitute intermediate systems. In the case of flexible languages, this situation occurs when derived lexemes cannot be used in all the syntactic slots in which their base lexemes are used. In the case of rigid languages, an intermediate system is attested when “the last class of lexemes on the hierarchy that is relevant for that language [is] a small closed class of items” (Hengeveld et al. 2004: 25). Including intermediate systems, the full inventory of language types amounts to thirteen. As it becomes evident from Table 2, parts-of-speech systems can be ordered from the least flexible system on one end (type 1) to the most rigid on

100 Jorge Gómez Rendón the other (type 7) through several less flexible or more rigid systems, including those with full differentiation. This pattern shows an implicational order in that, for instance, a language with a specific lexical class for predicate phrase modifiers always has individual lexical classes for each of the previous syntactic slots (system 4). Accordingly, no languages are attested which have a specific lexical class for predicate phrase modifiers but none for referential phrase modifiers, etc. The combination of syntactic positions and lexical classes reflects the cross-linguistic variation of languages around the world but also the developments in the history of individual languages whether they are product of internal or external changes such as contact with other languages. The result of synchronic and diachronic variation in parts of speech can be captured in an implicational hierarchy as shown below. Table 3. The parts-of-speech hierarchy Head of > 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)’

<

toca-

‘play’

(14) akaba-chi-shpa-mi < acaba-CAUS-GER-VAL ‘making [someone] finish [something]’

acaba- ‘finish’

(15) chay tanda dali-kpi-ka < DEM bread give-when-TOP ‘when they give the bread [to someone]’

dali-

‘give (to)’

108 Jorge Gómez Rendón How productive morphological changes are is demonstrated in (14), where a string of verbal and pragmatic particles are suffixed to the base lexeme. Besides, the two original meanings of the Spanish verb (acabar means ‘finish something’ and ‘be finished’) collapse in one by the addition of the causative marker chi. Example (15) seems to be a case of frozen borrowing in which two morphemes are assimilated as a single unit. A number of morphological processes may be operative in nouns and adjectives too. The most noticeable of them has to do with frozen borrowing and results in the double marking of plurals. (16) kosaskunaka kosas-kuna-ka thing:PL-PL-TOP ‘the things’

< cosa + s

< cosas

Spanish plural markers on nouns preceded by numerals are rather infrequent in the corpus. This tendency conforms to IQ, in which plural marking is not obligatory. Spanish gender markers occur often on borrowed adjectives (17). Spanish diminutives (18) and superlatives (19) are rather scarce. (17) kuriusu yana curious be:INF ‘to be curious’

< curioso (masculine)

(18) mamita mother:DIM ‘dear mother (19) papaso father:SUPL ‘grandfather’

4.3. Frozen borrowing of Spanish lexemes Frozen borrowing is the second most common of all accommodation processes of Spanish lexemes into IQ structure, with important consequences for the morphosyntax of the target language. In the present analysis, frozen borrowing not only includes such obvious instances of borrowing as (20a) and (20c) but also more complex expressions like those already reported for massively relexified varieties in other areas of Ecuador (Muysken 1985). A

Spanish lexical borrowing in Imbabura Quichua

109

great number of frozen borrowings include adverbial periphrases and conjunctions frequently used in Spanish. Consider the following examples: (20) a. b. c. d. e. f. g. h. i. j. l. m. n. o.

dirripinti diunabes diltodo masuminos namaski unsolu diuslupagui kalsunbaju diberas kaduno ilkimas loki inki purraya

< de repente < de una vez < del todo < más o menos < nada más que < uno solo < dios le pague < calzon bajo < de veras < cada uno < el que más < lo que < en que < por raya

(suddenly) (at once) (completely) (more or less) (only) (only) (thank you) (long-trousers) (really, actually) (each) (at most) (which) (what) (following the line)

The criteria used to distinguish frozen borrowings from code switches are both phonological and morphosyntactic. On the one hand, frozen expressions have a distinctive stress pattern – main stress on the penultimate syllable on the model of IQ – and differ from Spanish patterns:

(21) ilkimas (22) masuminos (23) diuslupagui

IQ /_-_/ /__-_/ /__-_/

Sp /__-/ /-_-_/ /-__-_/

On the other hand, their individual components cannot be subject either to derivational or inflectional processes. Such is the case of adverbial phrases and noun phrases as shown in (24–25). (24) kalsunbaju long-trousers

kalsunbaju-ta long-trousers-ACC kalsun-ta baju*

kalsunbaju-kuna long-trousers-PL kalsun-kuna baju*

(25) diunabes at.once

diunabes-mari diunabes-EMP diuna-mari bes*

diunabes-tak diunabes-MIT diuna-tak bes*

The syntactic position of frozen borrowings is determined by IQ syntax, in which the components cannot be dislocated (26–27).

110 Jorge Gómez Rendón (26) kunanka ilkimas rundin (ilkimas)* tukarian kunan-ka ilkimas rundin tuka-ria-n pan.flute play-DUR-3PRES today-TOP only ‘Today they play the pan flute only’ (27) kaduno sirbintikunaka (kaduno)* purijurka kada-uno sirbinti-kuna-ka puri-ju-rka go-DUR-3PST each servant-PL-TOP ‘Each employee used to go’ Another process counted as frozen borrowing is the reduplication of Spanish lexemes. Reduplication is common in IQ and serves emphasis. As for parts of speech, the most common of reduplicated items are nouns and manner adverbs (35). (28) bulla13 bulla shamurkanchika bulla bulla shamu-rka-nchi-ka noise noise come-PST-1PL-TOP ‘We came making much noise’ A final issue to be addressed here is that of phrases borrowed seemingly as entire units but interpretable as instances of code switching. Consider the following excerpt: chayka /primero/ watataka /nivela/shpa /cuinta/ /kida/parka, chayra kay watami /primero/guman yaykushka kapan, [primer año de colegio]man, chayka ninanda adilantashka kashka chaymanga.

In this sample text, an entire noun phrase in Spanish (in square brackets) contrasts with other borrowings (in slashes). While I consider these cases arguably ambiguous, I have counted them as cases of code switching and thus not included in this study.

5. Spanish borrowing of lexical classes in IQ: frequencies and tendencies In this section I present the statistics of Spanish borrowings in IQ according to (1) types and tokens; (2) the parts of speech in the source language; and (3) the use of Spanish borrowings according to their syntactic function in IQ. The statistics are analyzed in relation to the predictions made by the theory of parts of speech discussed above. The following table contains the

Spanish lexical borrowing in Imbabura Quichua

111

number of Spanish borrowings in IQ. The figures correspond to three different types of information: lexemes that may be assigned to one part of speech (tokens); lexemes with one or more bound morphemes affixed to them, i.e. words with identifiable phonological and semantic boundaries (types); and the numerical relation between tokens and types which shows the number of occurrences of single words in the corpus (token-type ratio). Numbers in parenthesis show average percentages for all the samples. Table 7. Spanish borrowings in IQ: general percentages Types, Tokens, TTR

Target Language (IQ) Source Language (Sp)

Tokens (mean) Types (mean) TTR

29296 (78.72%) 7660 (78.12%) 3.824

7921 (21.28%) 2145 (21.88%) 3.692

Total 37217 9805 3.795

While the percentage of Spanish borrowings (roughly one fifth of the corpus) evidences a strong lexical influence, these figures are by no means uncommon to other Ecuadorian Quechua dialects. The comprehensive analysis of several Ecuadorian dialects in Bakker et al. (2008) confirms this tendency. Given the amount of lexical borrowing in the language, one may expect several changes in IQ morphosyntax along the scale of borrowing proposed by Thomason and Kaufman (1989: 74–76). And in fact a few Spanish morphemes have been incorporated in IQ together with other deviations from IQ morphosyntax such as the drop of accusative markers, inverted possessive constructions and verb-medial position (cf. Gómez Rendón, forthcoming). The second step of the analysis was to identify to which part of speech borrowings belong in the source language. Table 8 shows the classification of Spanish borrowings according to parts of speech, including tokens and percentages. Table 8. Spanish borrowings in IQ: parts of speech

Tokens

Verbs

Nouns

Adjectives

Adverbs

Others

1336 (16.87%)

4401 (55.56%)

629 (7.94%)

145 (1.83%)

1410 (17.80%)

Nouns (55.56%) represent the largest number of Spanish borrowings, followed by verbs (16.87%), adjectives (7.94%) and adverbs (1.83%). The category “others” include conjunctions, numerals, discourse markers, and

112 Jorge Gómez Rendón other types of adverbs. The results presented here are identical to those found across dialects of Ecuadorian Quechua (Bakker et al. 2008). In order to evaluate the results on the use of Spanish borrowings in IQ, it is necessary to know first the distribution of lexemes across syntactic positions in the recipient language. For this purpose both native and borrowed lexemes were counted in a sample text of 1000 words. The results give a picture of the lexical blueprint of IQ as mapped onto syntactic slots. Table 9. Lexical distribution of syntactic functions in IQ Syntactic Function

Head Pred. Phrase 280

Head Ref. Phrase 273

Modifier Ref. Phrase 50

Modifier Pred. Phrase 125

Table 9 shows no significant difference in the distribution of heads. At the same time modification of predicate phrases is notoriously prolific. The number of heads is larger than the number of modifiers, but the headmodifier ratio is not proportional. Roughly speaking, of two verbs, at least one is modified, while of five nouns, only one is so. This tendency is reflected on the ratio between modifiers in the sample, with more than two predicate phrase modifiers for one referential phrase modifier. In all, the number of lexemes in the position of modifier of predicate phrase shows a strong preference for verbal over nominal modification in the language. Let us consider now the use of Spanish borrowings in the corpus. Loanwords were analyzed for their function in IQ in order to know whether they occupy the same syntactic slots (and thus fulfill the same syntactic function) as in the source language. The following are the results for tokens and types of the main lexical classes.

Spanish lexical borrowing in Imbabura Quichua

113

Table 10 a. Spanish borrowings in IQ: parts of speech versus function (tokens) Syntactic Function

Verbs (Tokens)

Nouns (Tokens)

Adjectives (Tokens)

HP HR MP MR MM

1332 3

49 3971 118 256 7

12 161 89 367

1

Adverbs (Tokens)

125 5 15

Totals 1393 4135 332 629 16

Table 10 b. Spanish borrowings in IQ: parts of speech versus function (types) Syntactic Function

Verbs (Types)

Nouns (Types)

Adjectives (Types)

HP HR MP MR MM

537 3

19 1659 28 181 1

6 117 55 234

1

Adverbs (Types)

39 4 8

Totals 562 1779 122 420 9

HP: Head of Predicate Phrase; HR: Head of Referential Phrase; MP: Modifier of Predicate Phrase; MR: Modifier of Referential Phrase; MM: Modifier of Modifier (of Predicate/Referential Phrase); NP: Noun used as postposition.

Two tendencies may be inferred from the tables above. On the one hand, the prototypical functions of Spanish borrowings according to parts of speech are the most frequent in the recipient language. Accordingly, verbs are used most of the times as heads of predicate phrases while nouns and adjectives are used more often than not as heads and modifiers of referential phrases respectively. This would confirm the hypothesis of functional specialization, but other interpretations are more plausible (cf. infra). On the other hand, while the theory predicts that Spanish nouns, adjectives and adverbs will be used in more than one syntactic position, flexibility seems to be disproved for adverbs but finds strong support for nouns and adjectives. Spanish nouns in IQ seem to be extremely flexible in use, occupying verbal, adjectival and adverbial positions. Likewise, Spanish adjectives are flexible to the extent that they are used not only as heads of referential and predicate phrases but also as modifiers. Not comparable with nouns and adjectives, adverbs show a much lesser degree of flexibility; interestingly Spanish adverbs may be used as modifiers of modifiers just like in Spanish. I explain this tendency

114 Jorge Gómez Rendón of adverbs by resorting to the constraint of directionality as explained in Section 2.3. Examples of the flexible use of Spanish borrowings are discussed in Bakker et al. (2008). I focus now on the interpretation of the data. Different interpretations may be given to the issue of flexibility depending on the approach of the analyst. One approach interprets the use of borrowings across different syntactic functions on the basis of their distribution without any other considerations. From this viewpoint, a widespread distribution such as the one found for nouns and adjectives in Table 10 is sufficient proof of flexibility. Another approach analyses the distribution by taking the prototypical function as the measuring stick. From this viewpoint, the flexible use of borrowings is significant only for nouns and adjectives. Taking a more radical position, one would be tempted to say that only adjectives support the claim of flexibility, as they are the only lexical class that truly shows a widespread distribution across syntactic slots. Whatever the stance we take in the end, we should consider that the data come from a corpus of spontaneous speech and not from elicited texts. Accordingly, we are not to expect a symmetrically distributed set (i.e. 25% of cases for each of the four possible syntactic slots) but instead skewed distributions. Additionally, it is expected that the prototypical functions of borrowed lexemes are preferred more often than not, simply because the forms and meanings are well known to speakers considering their high degree of bilingualism and exposure to Spanish. With these caveats in mind, I make a case for flexibility in the use of Spanish borrowings in IQ. Having analyzed data from other Ecuadorian Quechua dialects from the ‘prototype’ perspective, Bakker et al. (2008) find less strong support for the claim of flexibility, which appears to be even weaker for the other languages of their sample. To the extent that flexibility implies the adaptation of borrowed lexemes to the system of parts of speech of the target language, we are also in a position to state that there is no conclusive evidence pointing to the lexical specialization of non-verbs into nouns and modifiers (type 3). The corollary is that IQ is not undergoing a typological shift and its parts-ofspeech system remains the same in spite of the large number of lexical borrowings in the lexicon. The results for grammatical borrowing for Ecuadorian Quechua (Bakker et al. 2008) are consistent with this view.

Spanish lexical borrowing in Imbabura Quichua

115

6. Final remarks From the social and historical background and the linguistic data collected in fieldwork, it is safe to identify IQ on the third stage of Thomason’s borrowing scale, with high degrees of bilingualism (Büttner 1993: 48–49) and borrowing of non-basic and basic vocabulary. Contact-induced changes in the lexicon include the borrowing of a significant percentage (20%) of vocabulary, including conjunctions and Spanish numerals. Contact-induced structural changes are also present and include Spanish-derived SVO word order, subordinate constructions headed by Spanish conjunctions and SVO clauses, and a few Spanish inflectional affixes of gender and number (cf. Gómez Rendón, forthcoming). The sample analyzed here represents a point of departure from which hypotheses and predictions may be tested on an individual language basis. The next step is to conduct the same analysis on other language pairs and finally compare results cross-linguistically to arrive at possible generalizations about constraints operating on lexical borrowing.

Appendix 1: Informants, age and community of origin and residence Name

Gender

Age

Community

Bernardo Juan María Baltazar Rafael Amalia Ascencio Lizardo

Male Male Male Male Female Male Male

66 67 60 69 56 59 66

El Topo El Topo Casco Valenzuela Casco Valenzuela Angla Angla Angla

Notes 1.

2.

I follow here the convention of calling Quichua the Ecuadorian dialects that form a branch of the Quechua language family. Ecuadorian Quichua is classified as part of Quechua II (Torero 1964) or Quechua A (Parker 1963). All dialects of this subfamily share a number of traits with other varieties from Southern Peru, Bolivia and Argentina (Adelaar 2004: 185 ff). For a linguistic and sociolinguistic description of this language, see GomezRendón 2005.

116 Jorge Gómez Rendón 3.

The survey makes a distinction between Quichua-Spanish bilingualism (i.e. where Quichua is dominant) and Spanish-Quichua bilingualism (where Spanish is dominant). Altogether, the percentages of bilingualism in Casco Valenzuela and El Topo are higher than those reported in other communities of Imbabura, but lower than those for Angla. 4. The few Quichua monolingual speakers in Casco Valenzuela and El Topo are all (old) women. No Quichua monolingual speakers have been found in Angla. 5. The three communities show large numbers of labor migrants to the cities of Ibarra and Quito, with almost one half of the population under thirty years working outside. In the past, virtually all labor migrants were men. Nowadays, an important number of young women travel for work on a regular basis. Labor migration contributes more than fifty percent to the family income. 6. Although there is a bilingual elementary school in El Topo, a considerable percentage (30%) of children attend nearby Spanish monolingual schools. Furthermore, many bilingual teachers are not bilingual per se and teach their subjects either in Spanish or in a mixture of Spanish and Quichua. As it seems, the scenario for a rapid shift to Spanish is set in the classrooms themselves. 7. A list of informants with data about age, community and gender is included in the appendix. 8. For the form and content of testimonial narrative in the context of cultural studies in Latin America, see Beverly (1987). 9. While lexical borrowing includes the incorporation of lexical units, be they free or bound, code switching implies the change of languages in alternating strings of speech or phrases. Particularly difficult for the analysis are cases in which words from two different languages alternate within the same phrase while the structural framework belongs to only one of the languages. The resulting bilingual phrase shares features of lexical borrowing and code switching and may be a good example of intrasentential code switching. For the purposes of this study, cases of intrasentential code switching are not included in the analysis. For a discussion of lexical borrowing, code switching, and their intertwining, see Muysken 2000, Backus 2005, and Field 2005. 10. The statistical results show a difference between tokens and types (expressed in the Token-Type ratio or TTR), which can be explained by the agglutinative nature of the language, with one and the same suffix occurring several times in one sentence. The mean token-type ratio for all the samples is 2.200, meaning that as many as 2,2 tokens for one type occur in the corpus. Note also that neither tokens nor types do strictly correspond to words in Quichua, again because of its agglutinative nature. 11. The reason for restricting adverbs to manner adverbs is that only the latter subclass of adverbs modifies the head of predicate phrases, while other types modify the sentence as a whole (Hengeveld et al. 2004: 6). 12. Abbreviations: 1 = first person, 2 = second person, 3 = third person, ACC = accusative, ADIT = additive, ALL = allative, ART = article, CAUS = causative, COM =

Spanish lexical borrowing in Imbabura Quichua

117

comitative, DAT = dative, DEM = demonstrative, DIM = diminutive, DIR.EV = direct evidence, DUR = durative, EMP = emphasizer, GEN = genitive, GER = gerund, INF = infinitive, INFER = inference, INST = instrumental, INTERP = interpelative, LIM = limitative, LOC = locative, NEG = negative, OBJ = (in)direct object, PERF = perfect, PL = plural, PRES = present, PROL = prolative, PST = past, REP = reportative, SG = singular, SUB = subordinator, SUPL = superlative, TOP = topicalizer 13. Originally, Spanish bulla means both ‘crowd’ and ‘screaming’. The meaning of the borrowing points to the number of people who came and the noise they made.

References Adelaar, Willem, in collaboration with Pieter Muysken 2004 Languages of the Andes. Cambridge: Cambridge University Press. Backus, Ad 2005 Code-switching and language change: one thing leads to another? International Journal of bilingualism 9 (3/4): 307–340. Bakker, Dik, Jorge Gómez Rendón, and Ewald Hekking 2008 Spanish meets Guaraní, Otomí and Quichua: A multilingual confrontation. 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.), 165–238. Berlin/NewYork: Mouton de Gruyter. Beck, David 2002 The typology of parts of speech systems: the markedness of adjectives. New York: Routledge. Beverly, John 1987 Anatomía del testimonio, Cap. en Del lazarillo al sandinismo: Estudios sobre la función ideológica de la literatura española e hispanoamericana. Minneapolis: Prisma Institute. Büttner, Thomas 1993 Uso del quichua y el castellano en la Sierra ecuatoriana. Quito: Ediciones Abya Yala. CIEI 1982a Caimi Ñucanchic Shimiyuc-Panca. Quito: Ministerio de Educación y Cultura, PUCE. 1982b Ñucanchic llactapac shimi. Quito: Ministerio de Educación y Cultura, PUCE. Cole, Peter 1982 Imbabura Quechua. Amsterdam: North-Holland Publishing Company.

118 Jorge Gómez Rendón Dikker, Suzanne 2003 Adpositions in linguistic theory and language contact. Master Thesis. Amsterdam: University of Amsterdam. Fauchois, Anne 1988 El quichua serrano frente a la comunicación moderna. Quito: Proyecto EBI, Ediciones Abya Yala. Field. Frederic 2005 Long-term effects of CS: clues to structural borrowing. International Journal of bilingualism 9 (3/4): 341–360. Gómez Rendón, Jorge 2005 La media lengua de Imbabura. In Encuentros y conflictos. Bilingüismo y contacto de lenguas en el mundo andino, Pieter Muysken and Hella Olbertz (eds.), 39–57. Madrid: Vervuert Iberoamericana. forthc. Spanish grammatical borrowing in Imbabura Quechua: an overview. Manchester Project on Grammatical Borrowing. Haboud, Marleen 1998 Quichua y Castellano en los Andes Ecuatorianos: los efectos de un contacto prolongado. Quito: Ediciones Abya Yala. Hengeveld, Kees 1992 Non-verbal predication: theory typology and diachrony. (Functional Grammar Series 15.) Berlin: Mouton de Gruyter. Hengeveld, Kees, Jan Rijkhoff, and Anna Siewierska 2004 Parts-of-speech systems and word order. Journal of Linguistics 40(3): 527–570. Jijón y Caamaño, Jacinto 1941 El Ecuador Interandino y Occidental. Quito: Editorial Ecuatoriana. Muysken, Pieter 1985 Contactos entre Quichua y Castellano en el Ecuador. In Memorias del Primer Simposio Europeo sobre Antropología del Ecuador, Segundo Moreno (ed.), 377–472. Quito: Abya Yala. 2000 Semantic transparency in Lowland Ecuadorian Quechua morphosyntax: Linguistics 39 (5): 873–988. Myers-Scotton, Carol 2002 Contact linguistics. Oxford: Oxford University Press. Parker, Gary 1963 La clasificación genética de los dialectos quechuas. Revista del Museo Nacional 32: 241–252. Schachter, Paul 1985 Parts of speech systems. In Language typology and syntactic description, Timothy Shopen (ed.), 3–61. Cambridge: Cambridge University Press.

Spanish lexical borrowing in Imbabura Quichua

119

Salomon, Frank and Sue Grosboll 1986 Names and peoples in Incaic Quito: retrieving undocumented historic processes through anthroponymy and statistics. American Anthropologist 88: 387–399. Thomason Sarah and Terrence Kaufman 1989 Language contact, creolization and genetic linguistics. Berkeley: University of California Press. Torero, Alfredo 1964 Los dialectos quechuas. Anales Científicos de la Universidad Agraria 2 (4): 446–478. 2003 Idiomas de los Andes. Lima: Centro Cultural Bartolome de las Casas. Van Hout, Roeland and Pieter Muysken 1994 Modelling lexical borrowability. Language variation and change, 6 (1): 39–62.

Spanish prepositions in Media Lengua: Redefining relexification Suzanne Dikker

1. Introduction Media Lengua, which means as much as ‘half language’ or ‘halfway language’, is a contact language between Spanish and Imbabura Quechua1. The variety described here is spoken in central Ecuador, currently by about 1000 native speakers. Media Lengua is an in-between language in many respects. First of all, it forms a linguistic mixture of Quechua and Spanish, which leads to its classification as a Bilingual Mixed Language (Thomason 2001). Geographically speaking, the Media Lengua-speaking community is located between the Spanish blanco world in the valleys and the purely Quechua-speaking villages higher up the mountain slopes. In sociolinguistic terms, Media Lengua was created by men who had Quechua as their native language, but left to work in Spanish-speaking areas (the obreros). When they returned to the community, they had been using Quechua on an infrequent basis, while having acquired a relatively fluent urban Spanish. Media Lengua is spoken by the bilingual middle generation, forming a link between the older, Quechua-speaking generation, and the younger, Spanishspeaking generation. Finally, while Quechua is used in traditional contexts and in situations of contact with the Quechua-speaking farmers, and Spanish is used in school and with the non-Indian world, Media Lengua is the primary language for communication in the every-day life of the community. This paper aims at establishing whether or not Media Lengua should be analyzed as a unique case of language mixture, by studying the incorporation of Spanish adpositions. Section 2 discusses Muysken’s (1981) relexification model, which he designed to account for the structure of Media Lengua. Section 3 explores whether Media Lengua can be accounted for in terms of general principles and constraints on contact-induced language change. Section 4 treats the linguistic status of adpositions, followed by a description and a discussion of the behavior of adpositions in Media Lengua.

122 Suzanne Dikker 2. Media Lengua as a relexified language Muysken (1981; 1997) defines Media Lengua as a relexified language: while the Quechua phonology, morphology, and syntax are preserved, virtually all of the lexical elements are provided by Spanish. Muysken (1997) argues that the different components of a lexical entry (the phonological form, the syntactic properties, the subcategorization features, the semantic properties and the selection criteria) function independently of each other. He formulates this point in the following way: “New relexified forms do not enter as full lexical entries, but as phonological shapes which are grafted onto the original lexical entries. Thus the original entry is not replaced, but merely altered in its own shape.” (Muysken 2000: 266). He furthermore emphasizes that the only requirement for relexification to occur is that the lexical entries of the recipient language (RL, here: Quechua) and the source language (SL, here: Spanish) share some semantic features. As a result, Media Lengua has the same lexical categories as the RL. Finally, there are considerable differences between the relexification of content words, facilitating a straightforward relexification, and function words, which imply a drastic restructuring of the system, because function words do not carry any meaning outside the linguistic system that they are part of. This explains why virtually only content words are relexified. The results of relexification in Media Lengua are illustrated by the following example: 2 (1)

a. Media Lengua unu fabur-ta pidi-nga-bu bini-xu-ni one favor-ACC3 ask-NOM-BEN come-PROG-1 ‘I come to ask a favor’ b. Quechua shuk fabur-da

maña-nga-bu

shamu-xu-ni

c. Spanish vengo para pedir un favor This example clearly shows how the replacement of phonological shapes of Quechua stems with Spanish forms comes about, in for example the replacement of maña by pidi. It also shows that the original structure of Quechua is maintained. From the Spanish forms in the Media Lengua sentence, it can be seen that in the process of incorporation, they undergo regularization (e.g., Sp: vengo4  ML: bini) and adaptation to the Quechua morpho-phonology (e.g., Sp: favor  ML: fabur-ta). From the fact that

Spanish prepositions in Media Lengua: Redefining relexification

123

Media Lengua is not random and creative, in addition to the observation that lexical items come predominantly from one language (90% Spanish), it can be discarded that Media Lengua is an instance of ‘deep’ intrasentential codeswitching (Muysken 1997). 3.

Relexification and ‘normal’ contact-induced language change

Muysken’s account for the structure of Media Lengua suggests that the process of relexification may be qualitatively distinct from other, ‘normal’, instances of contact-induced language change (CiCh). This section discusses general principles and restrictions on CiCh, exploring whether Media Lengua follows these, or, alternatively, a separate account like Muysken’s is indeed warranted. 3.1. Social factors Various viewpoints have been defended with regard to the identification of the main triggers of CiCh. It is largely uncontroversial to assume that the social situation of actual contact between two or more languages (or dialects, for that matter), is a necessary first requirement for any form of this type of change to occur. Thomason (2001) stresses this point specifically, since she feels that a number of scholars seem to have lost touch with this common-sense point. The intensity and duration of contact are of considerable influence on the extent to which the linguistic systems of the languages in contact are affected (Thomason 2001; Campbell 1998). Another quantitative measure, group size, is subordinate to the relative social positions of the communities in contact, since the minority status of a group, usually consisting of speakers of the language(s) most affected in terms of CiCh, is defined in terms of social status (subordination and superordination) rather than size. The relative social positioning of the languages in contact determines whether the type of CiCh that is observed either classifiable as interference or as borrowing. Interference occurs when a superordinate language is imposed upon a minority group. This situation of massive L2 acquisition often results in an imperfect and fossilized non-target final stage, generally simplified and containing many L1 elements. While interference occurs from the first to the second language (L1  L2), in the case of borrowing non-native items are taken over from the second language into the native language (L2  L1). Within this dichotomy, Media Lengua could be classified as a case of borrowing rather than interference,

124 Suzanne Dikker since what is observed is the incorporation of non-native elements (Spanish) into the first language (Quechua). Milroy (1993) emphasizes – for both diachronic and contact-induced change – the role of specific members of a language community in language change, namely those with the highest number of ‘weak-ties’ within and outside of the community. In the Media Lengua-speaking community, the obreros can be identified as forming such a group. Milroy concludes that linguistic change is a social phenomenon, which comes as a result of the desire to mark social identity. He goes as far as saying that if no social meaning is attached to it, change will not occur. Thomason (2001) also states that the speakers’ attitude is ultimately responsible for the failure of other substantive predictors. This entails that whatever change is triggered, it can be halted or accelerated by the speakers’ attitude. Thomason’s (2001) viewpoint on CiCh is reflected in the following hierarchy: (2)

Speakers’ attitude > 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

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)

Lexicon

Syntax 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 <

botella ‘bottle’ cuchillo ‘knife’ casillas ‘bathroom, toilet’ cebolla ‘onion’ caballo ‘horse’

The old pronunciation of Spanish // accounts for the apparently strange realizations of Spanish words like sabón < jabón ‘soap’ or sáro < jarro ‘jug, pitcher’. This explains exactly why there are in many cases two forms of one and the same lexeme of Spanish origin in Philippine languages, namely, one old and popular form and one modern form orientated towards the standard (written) language, e.g. labása vs. nabáha, both from Spanish lavaja and navaja ‘small knife’. There is sugaról and hugadór, both from Spanish jugador ‘player’ (the popular forms are taken from López 1965: 474). Since most Philippine languages have the word-initial phoneme /h-/ like in Tagalog háwak ‘hold, grasp’, the conservation of initial /h-/ in Spanish

Hispanisation processes in the Philippines

209

words is frequent, but not always present, e.g. Tagalog húrno < Spanish horno. This conservative feature is also found in the Chabacano-variety of Ternate. All together these “inconsistencies” demonstrate the introduction of Hispanic words by spoken, spontaneous speech in earlier times and on the other hand a lot of cases with “spelling” (or educated) pronunciation introduced during the later period of Spanish rule in the Philippines. Generally, this same dualism is also noticeable when comparing the Chabacano-varieties in Ternate with those in Zamboanga and in Cavite respectively.

2.3.

Morphology

2.3.1. Reanalysis of Spanish words and phrases There a quite a few cases where Spanish words or phrases were reanalyzed by native speakers of Philippine languages. For example, following López (1965: 482) Spanish ¿como está? ‘how are you?’ was reanalyzed as containing the infix -um-, used in Tagalog to form active transitive forms, and with the elision of unstressed e of está the Tagalog reproduction is kumustá (k-um-ustá), leaving a meaningless base kustá. (9)

TAGALOG & CEBUANO Spanish como está? ‘how are you?’ > 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’ fañocho Maila’ ya

(= fan- + chocho)

ta

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

famboka (= fan- + boka) IMP:eat

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

li’e’ i

I.ERG see the ‘I saw the teacher.’

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

<1 <1 3

5,445 5,445 7,381

250

1

18,271

We already see that it is a relatively close run between Austronesian and Spanish because both oscillate around the 50%-mark. However, the Austronesian figures are generally a tick higher than the ones reported for Spanish. In this way, our initial assumption according to which speakers value their native language higher than the foreign one is corroborated at least in the guise of a moderate tendency. If we break down the above general statistics over the two genders, the following picture emerges, cf. (8)–(9). There were 56 male participants accounting for 46% of the informants as opposed to 65 female participants making up for 54% of the interviewees.

Pro or contra Hispanisms

(8)

(9)

249

Male respondents Austronesian

Spanish

other

part

answers

%

answers

%

answers

1 2 3

1,316 1,311 1,695

52 52 49

1,201 1,206 1,630

47 47 47

sum

4,322

51

4,037

47

%

total

3 3 91

<1 <1 2

2,520 2,520 3,416

97

1

8,456

%

total

Female respondents Austronesian

Spanish

other

part

answers

%

answers

%

answers

1 2 3

1,478 1,468 1,951

50 50 49

1,437 1,447 1,881

49 49 47

10 10 133

<1 <1 3

2,925 2,925 3,965

sum

4,897

49

4,765

48

153

1

9,815

Both male and female informants made ample use of Austronesian and Spanish in their solutions. However, there is a small difference between the two genders as male respondents give preference to Austronesian terms more often than their female counterparts opt for Austronesian. For the female group, the gap separating Austronesian from Spanish is minimal for parts 1–2 and only slightly more pronounced for part 3. With the men, especially for parts 1–2, the Austronesian words score somewhat higher than the Spanish ones. Moreover, the women are also responsible for the bulk of the third category (= 62%) i.e. for those unexpected terms which do not belong to the word pairs. Some of these items provided by the female informants were inappropriate in the sense that they did not convey the intended meaning. For the subsequent statistics of other sociolinguistic parameters, we have reduced the number of categories listed in our catalogue above in order to work on the basis of categories of comparable sizes. For the parameter age, for instance, the inclusion of the single respondent with an age below 30 would have distorted the entire picture as this person behaved in singularly deviant fashion.12 We thus decided to drop his/her data from the statistics and reorganise the remaining age groups into three categories: recently born (age 31–50), post-war born (age 51–60) and wartime and pre-war born (age 61–90), cf. (10)–(12).

250 Rosa Salas Palomo and Thomas Stolz (10) Recently born respondents (35 = 29%) Austronesian

Spanish

other

part

answers

%

answers

%

answers

1 2 3

807 804 1,024

51 51 47

766 758 1,022

48 48 47

sum

2,635

49

2,546

48

%

total

2 3 89

<1 <1 4

1,575 1,575 2,135

94

1

5,275

%

total

(11) Post-war born respondents (43 = 35%) Austronesian

Spanish

other

part

answers

%

answers

%

answers

1 2 3

987 990 1,236

51 51 47

937 935 1,304

48 48 49

11 10 25

<1 <1 <1

1,935 1,935 2,623

sum

3,213

49

3,176

48

46

<1

6,493

(12) Wartime and pre-war born respondents (42 = 35%) Austronesian

Spanish

other

part

answers

%

answers

%

answers

%

total

1 2 3

965 956 1,345

51 50 52

925 934 1,167

48 49 45

0 0 50

0 0 1

1,890 1,890 2,562

sum

3,266

51

3,026

47

50

<1

6,342

From these tables, we learn that the two “younger” groups display a relatively similar overall behaviour as to their preferences. Austronesian and Spanish elements are statistically close to each other although in parts 1–2, both prefer Austronesian terms over Spanish ones. The “youngest” group, however, has a comparatively high share of third-category items in part 3. This high share affects the ratio of Austronesian solutions which drops to a low of 47%. Incidentally, this is exactly the same percentage we get for Austronesian terms in (11) (part 3). However, with the post-war born respondents, this relatively low percentage is not caused by third-category items because the Spanish solutions are more numerous than the Austronesian ones. It looks like this stronger position of Spanish with the age group 51–60 has been giving way to elements intruding from the side of third-

Pro or contra Hispanisms

251

category items used by younger speakers. Interestingly, the oldest group in (12) does not favour Spanish-derived words in any way. For them, the Austronesian elements are the first choice for all three parts of the questionnaire. For part 3, Spanish words get the lowest ratio of all age groups. Moreover, third-category items are marginal. Thus, the very old have a pronounced preference for Austronesian words and disfavour the use of Spanish-derived ones. They also make seldom use of third-category items which are absent completely from their responses in parts 1–2. In (13)–(14), we give the results for the parameter ethnicity. Since the Chamoru group outnumbers all other ethnic groups by a ratio of more than 9-to-1 and no other ethnic groups are represented by more than two informants, we subsume informants of Tagalog, mixed Chamoru (and Tagalog, Japanese, American) and other origin under the perhaps not absolutely appropriate label non-Chamoru. (13) Chamoru respondents (114 = 94%) Austronesian

Spanish

other

part

answers %

answers

%

answers

1 2 3

2,630 2,617 3,439

51 51 49

2,488 2,501 3,307

48 48 47

sum

8,686

50

8,296

48

%

total

12 12 208

<1 <1 2

5,130 5,130 6,954

232

1

17,214

%

total 315 315 427

(14) Non-Chamoru respondents (7 = 6%) Austronesian

Spanish

other

part

answers

%

answers

%

answers

1 2 3

164 162 207

52 51 48

150 152 204

47 48 47

1 1 16

<1 <1 3

sum

533

50

506

47

18

1

1,057

The results for both groups are very similar. The little differences there are may be explicable in terms of chance factors such as the relatively small size of the non-Chamoru group. In any case, Austronesian words score usually higher than Spanish ones. However, in actual usage the gap between the two options is relatively narrow. Third-category items are well represented in part 3 for both groups. It seems that the rise in frequency on the part of

252 Rosa Salas Palomo and Thomas Stolz third-category items again affects mostly the Austronesian options whereas the Spanish share remains relatively stable. For the parameter of level of education, we also provide a bipartition of the interviewees’ population into one that combines elementary and secondary education and another one which comprises all kinds of tertiary education from the college level upwards, cf. (15)–(16). (15) Elementary and secondary education (72 = 60%) Austronesian

Spanish

other

part

answers %

answers %

answers %

total

1 2 3

1,633 1,627 2,188

50 50 49

1,604 1,610 2,067

49 49 47

3 3 137

<1 <1 3

3,240 3,240 4,392

sum

5,448

50

5,281

48

143

1

10,872

%

total

(16) Tertiary education (49 = 40%) Austronesian

Spanish

other

part

answers %

answers %

answers

1 2 3

1,161 1,152 1,458

52 52 48

1,040 1,049 1,444

47 47 48

10 10 87

<1 <1 2

2,211 2,211 2,989

sum

3,771

50

3,533

47

107

1

7,411

Again, the two groups yield similar overall results which show Austronesian items on the first rank. However, the more educated speakers favour Austronesian words slightly more in parts 1–2 whereas their actual usage is such that Austronesian and Spanish are treated more or less on a par. For both groups, third-category items coming in with part 3 tend to affect the share of Austronesian items more than they affect the ones of Spanish. Speakers with an education below the tertiary level are more prone to using third-category vocabulary. Spanish is slightly more popular with this group, too. Next in line is the geographical site of the place of residence during the informants’ formative years. Since almost half of our informants originally come from Central Guam, we have distributed the informants over two groups: one representing natives of Central Guam and the other covering all other regions of Guam and elsewhere, cf. (17)–(18).13

Pro or contra Hispanisms

253

(17) Respondents originating from Central Guam (59 = 49%) Austronesian

Spanish

other

part

answers %

answers %

answers

1 2 3

1,319 1,324 1,771

49 49 49

1,327 1,322 1,704

49 49 47

sum

4,414

49

4,354

48

%

total

9 9 124

<1 <1 3

2,655 2,655 3,599

142

1

8,909

%

total

(18) Respondents originating from elsewhere (62 = 51%) Austronesian

Spanish

other

part

answers

%

answers

%

answers

1 2 3

1,475 1,455 1,875

52 52 49

1,311 1,331 1,807

46 47 47

4 4 100

<1 <1 2

2,790 2,790 3,782

sum

4,805

51

4,449

47

108

1

9,362

For speakers with a formative background in Central Guam, Austronesian and Spanish items yield similar results for parts 1–2. Their frequencies are almost identical. Only for part 3, the Spanish share drops lightly whereas the Austronesian share remains unaffected by the introduction of third-category items. More generally, informants who grew up in Central Guam make a difference between Austronesian and Spanish only when it comes to actually employing the terms. For the respondents from other regions, however, there is a relatively clear preference for words with an Austronesian etymology in all three parts of the questionnaire. The Spanish share is comparatively low. However, third-category items only affect the percentage of Austronesian elements. As to the present residence of our informants, we follow the same pattern as with the previous criterion. We divide the population into two groups – one of which includes all those who are currently living in Central Guam whereas the other group comprises residents of all other regions of Guam, cf. (19)–(20).

254 Rosa Salas Palomo and Thomas Stolz (19) Residents of Central Guam (56 = 46%) Austronesian

Spanish

other

part

answers %

answers

%

answers

1 2 3

1,277 1,284 1,704

50 50 49

1,234 1,226 1,607

48 48 47

sum

4,265

50

4,067

48

%

total

9 10 105

<1 <1 3

2,520 2,520 3,416

124

1

8,456

%

total

(20) Residents of other regions (65 = 54%) Austronesian

Spanish

other

part

answers

%

answers

%

answers

1 2 3

1,517 1,495 1,942

51 51 48

1,404 1,427 1,904

48 48 48

4 3 119

<1 <1 3

2,925 2,925 3,965

sum

4,954

50

4,735

48

126

1

9,815

Like in the case before, the overall results are strikingly similar for both groups. Some differences occur however when we look at the various parts individually. For residents of Central Guam, Austronesian and Spanish elements in parts 1–2 are not separated by as huge a gap as is the case in (20). Austronesian solutions are always more frequent for both groups in all parts of the questionnaire. However, for residents of regions outside of Central Guam, the actual usage test reveals a loss on the Austronesian part which results in a kind of draw between Austronesian and Spanish. For this difference, the third-category items can be held responsible. Note however, that the third-category items do not have the same effect with speakers who live in Central Guam as both Austronesian and Spanish shares are affected to a similar extent. For the size of families, we operate on the basis of three categories. The first category covers all informants who have six or less siblings, the second and largest group contains a number of siblings from 7 to 12, the third category embraces all groups of siblings of 13 and beyond, cf. (21)–(23).

Pro or contra Hispanisms

255

(21) Six siblings or less (42 = 35%) Austronesian

Spanish

other

part

answers %

answers

%

answers %

total

1 2 3

1,003 987 1,253

53 52 49

885 902 1,238

46 47 48

2 1 61

<1 <1 2

1,890 1,890 2,552

sum

3,243

51

3,025

47

64

1

6,332

(22) Seven to twelve siblings (66 = 54%) Austronesian

Spanish

other

part

answers %

answers

%

answers %

total

1 2 3

1,505 1,495 2,006

50 50 68

1,454 1,463 770

48 49 26

11 11 134

<1 <1 4

2,970 2,970 2,909

sum

5,006

56

3,687

41

156

1

8,849

(23) Thirteen and more siblings (13 = 11%) Austronesian

Spanish

other

part

answers %

answers %

answers %

total

1 2 3

286 297 378

48 50 47

299 288 386

51 49 48

0 0 29

0 0 3

585 585 793

sum

961

48

973

49

29

1

1,963

On this parameter, the statistical values are more varied than on the previous ones. For the first time, we also get results which go far beyond or fail to meet the range of the variation between 55–45% we had so far. For “small” families (21), the results still meet our expectations. Austronesian items are generally preferred over Spanish ones. In part 3, the Austronesian items loose some territory because third-category items get more numerous in this part of the questionnaire. Surprisingly, this effect is inverted to the detriment of Spanish items in (22): informants with seven to up to twelve siblings perform in such a way that the actually used forms show an unprecedented high of Austronesian forms and a record low of Spanish items which account only for slightly more than a quarter of all answers. Clearly, in middle-

256 Rosa Salas Palomo and Thomas Stolz size families, the Austronesian share does not suffer from the growth on the part of third-category items. This trend however, is not continued in still more sizeable families as (23) shows. In this class, the Austronesian segment looses ground and thus Spanish has a much stronger positions with informants from very large families. Last but no least, we look at the position our informants have among their siblings. In accordance with our previous practice, we render the data handy for our purposes by reducing the number of categories to three. We distinguish the group of those informants who are the eldest among the siblings from those who have both older and younger siblings and a third group for which informants qualify who are the youngest in their family, (24)–(26). (24) Oldest among siblings (24 = 20%) part 1 2 3 sum

Austronesian

Spanish

other

answers %

answers

%

answers %

total

556 542 721

51 50 49

522 536 713

48 49 48

2 2 30

<1 <1 2

1,080 1,080 1,464

1,819

50

1,771

48

34

<1

3,624

%

total 3,555 3,555 4,819

(25) With older and younger siblings (79 = 66%) Austronesian

Spanish

other

part

answers %

answers

%

answers

1 2 3

1,800 1,801 2,368

50 50 49

1,746 1,744 2,281

49 49 47

9 10 170

<1 <1 3

sum

5,969

50

5,771

48

189

1

11,929

%

total

(26) Youngest among siblings (18 = 15%) part 1 2 3 sum

Austronesian

Spanish

other

answers

%

answers

%

answers

438 436 557

54 53 50

370 373 517

45 46 47

2 1 24

<1 <1 2

810 810 1,098

1,431

52

1,260

46

27

<1

2,718

Pro or contra Hispanisms

257

Tables (25)–(26) correspond largely to the general picture with their moderate but nevertheless clear preference for Austronesian terms and the slight losses for the Austronesian terms in part 3. Somewhat different is the behaviour of the youngest siblings. Their preference for Austronesian terms is more pronounced than with older siblings. Third-category items are again responsible for some losses on the part of Austronesian but the values remain high enough to account for half of all answers given in part 3.

4.2. Evaluation How do our hypotheses correspond to the above results? For the initial hypothesis (1H), the answer is that there is a tendency throughout the entire population of interviewees to give preference to Austronesian over Spanish. However, table (7) suggests that this tendency is moderate as the Spanishderived terms have still a rather strong position independent of the fact hat they are most often only second best. Thus, our hypothesis is only partially borne out by the facts. This also applies to (2H) where it was hypothesised that the informants would direct their loyalty always to Austronesian no matter what the task was. However, there are clear differences between the three tasks. Parts 1–2 invited subjective statements based on individual preconceptions of correctness and frequency of use. In these two categories, the Austronesian items fared best by far. In this sense, our hypotheses (2Ha) and (2Hb) are largely corroborated by the evaluative behaviour of the informants. As to (2Hc), the situation is different because the actual use of items could be shown not to correspond to the speaker judgments about correctness and frequency. Not only do Austronesian items drop below the 50%-mark but third-category items gain in importance (without being a threat to either of the other two categories). Their gains seem to be counterbalanced mainly by losses on the side of Austronesian terms whereas the Spanish items are less strongly affected by the rise of the third category. For the individual sociolinguistic parameters, our hypotheses most often turn out to be wrong. Contrary to our expectation, Tables (8)–(9) are indicative of a higher loyalty of men to Austronesian who demonstrated that they are more knowledgeable of Austronesian-derived Chamoru forms than women. Women’s performance is also characterised by a slightly higher share of third-category items. The better knowledge of Austronesian Chamoru reported for men especially applies to the solutions given in parts 1–2 of the questionnaire. Remarkably, men and women display almost identical distribution patterns for Austronesian and Spanish items when it comes to

258 Rosa Salas Palomo and Thomas Stolz actually using words in part 3. For the age parameter, we predicted in (3Hb) that the older generations will tend towards a higher proportion of Spanish. Again, this assumption is disproved by the facts as none of the age groups in (10)–(12) displays an overall preference for Spanish. However, those who were born immediately after the 2nd World War have a higher percentage of Spanish in their actual use whereas for the younger generation, the figures of Austronesian and Spanish are almost identical. This is the more remarkable because the oldest informants whose birth dates back to the wartime and pre-war period make far less use of Spanish than those born later. This contradicts our hypothesis. What needs to be mentioned too is the fact that third-category items are statistically negligible with the oldest and the middle generation. However, with the youngest group, third-category items become stronger in actual use and occur at least four times as often as in the performance of the older informants. Ethnicity proved to be a rather weak criterion. Contrary to what we said in (3Hc), there is no difference in the behaviour of the Chamoru group and the non-Chamoru group. Both prefer Austronesian over Spanish with percentages which are perfectly in line with the general picture. Speakers from a non-Chamoru background display a slightly stronger tendency to introduce third-category items in part 3 but this is counter-balanced by the fact that they otherwise have slightly lower percentages for Spanish than the informants from homogeneous Chamoru families. According to our original ideas expressed in (3Hd), a higher level of education should go hand in hand with a preference for the putatively prestigious Spanish. As Tables (15)–(16) suggest, this idea is somewhat mistaken as, at least for parts 1–2, higher educated informants opt preferably for Austronesian terms – with a percentage which goes beyond the one reported for less educated speakers. As so often before, part 3 reveals that the actual use of terms does not necessarily follow the subjective image of correctness and frequency. Better educated speakers get even scores for Austronesian and Spanish items in their performance whereas Austronesian elements are still slightly better off in the performance of less educated speakers. Be that as it may, no group displays any pro-Spanish preference. Third-category items are more frequent with less educated speakers without occurring significantly more often than in other groups. In contradistinction to several of the previous parameters, the place of residence during the interviewee’s formative years has a stronger effect on the choice the informants had to make. In line with our hypothesis (3He), speakers originating from Central Guam are more tolerant towards Spanish items. In part 1, these even outnumber Austronesian ones though only by a

Pro or contra Hispanisms

259

narrow margin. In part 2, the figures for both etymological sources are almost the same. Only in the translation task do Spanish items fail to reach the values given for Austronesian ones. This is the effect of the relatively high number of third-category items which seem to intrude mainly into the territory originally occupied by the Hispanisms. Respondents from outside Central Guam behave strikingly different. In their answers to parts 1–2, Austronesian is indisputably solution number one and the percentages for Spanish are never as high as with speakers originating from Central Guam. Again, part 3 is crucial in the sense hat the results demonstrate that the preference given to Austronesian is dependent upon the laboratory situation of parts 1–2. Nevertheless, even in actual use, the gap between Austronesian and Spanish items is considerable. Third-category items seem to be competitors mostly of Austronesian items, though. Less clear are the results for the parameter of actual residence. In (3Hf), we assumed parallel to the previous criterion of place of origin that living in Central Guam would have an effect such that Spanish elements are more common with residents of that area. However, the results in tables (19)–(20) do not reflect any major differences between the two groups of informants. The best one can make of the figures is noticing that for parts 1–2, informants living elsewhere on Guam have slightly higher percentages for Austronesian terms than those speaker who reside in Central Guam. However, this is owed to the fact that for these tasks the informants used fewer thirdcategory items. In part 3, third-category items are used with identical shares in both groups. In point of fact, speakers living outside of Central Guam used Austronesian and Spanish terms with similar frequencies in the translation tasks whereas interviewees from Central Guam retained higher rates for Austronesian even in part 3. In sum, the data are inconclusive if added up. Only for the individual tasks do we find evidence that in the less urbanised areas of the island, Austronesian Chamoru is thought of as more correct and more frequent. Family size is a difficult criterion to evaluate. According to (3Hg), we expect a levelling effect in the larger families such that their choices fall square with the ones of the average of our informant population. However, the data are confusingly different from what we predicted. In what we term small families, Austronesian is better off with Spanish reaching values only which are lower than the average in (7). This is especially valid for parts 1– 2, whereas the figures given for the actual use of items are back to normal, in a manner of speaking, as they no longer display a huge discrepancy between the two etymological sources. The share of third-category items is rather moderate. The figures in (22), however, are astonishing. In middle-

260 Rosa Salas Palomo and Thomas Stolz size families, parts 1–2 do not yield figures worthwhile commenting upon as they show the usual slight majority of Austronesian terms. However, part 3 provides a remarkable increase of Austronesian solutions with the parallel downsizing of the Spanish share. Similarly high figures do not occur elsewhere in the statistics, nor do we find equally low figures as the ones for Spanish in (22). We could take this as evidence for a levelling effect – however it is an effect different from the one envisaged by us. The scores in part 3 are also responsible for the relatively high overall figures for Austronesian and the corresponding low figures for Spanish in this group. In addition, third-category items reach 4% which is the reported maximum for this category. Families can be even bigger – as (24) shows. However, for informants with thirteen and more siblings the striking dominance of Austronesian terms is turned into a slight preference for Spanish items. In part 1 and part 3, Spanish solutions outnumber Austronesian ones. This is against the general trend towards more Austronesian solutions and thus proves our prediction wrong. For the time being, we can only state the facts without having an explanation at hand. The last criterion is the relative position in the order of siblings. We assumed in (3Hh) that older/oldest siblings will prefer Spanish. There is no compelling evidence for a preference of this sort in the tables. Oldest siblings and those who have both older and younger siblings produced relatively similar figures all of which attest a preference for Austronesian along the line of most of the other tables presented above. Oldest siblings even behaved more pronouncedly pro-Austronesian in their answers to part 1. It hardly needs to be stated that in actual use, the percentages of Austronesian and Spanish are closer to each other than in parts 1–2. Third-category items are relatively frequent in the performance of “middle” siblings. Somewhat unexpectedly, the results for youngest siblings are special as they include unusually high percentages for Austronesian and low values on the Spanish side. The gap between Austronesian and Spanish is remarkably huge for parts 1–2 and it is still sizeable for the performance data of part 3 (although the values given for third-category items are by no means exceedingly high). Thus, informants who have younger siblings behave more or less like the average whereas informants who have only older siblings display unexpectedly high scores for Austronesian. The vast majority of the hypotheses we formulated in section 2 could not pass the test. The role of Spanish as a kind of prestige language implicit in many of the hypotheses is obviously an illusion. On the other hand, there is also no witch-hunt underway which prevents speakers from making freely use of Spanish-derived items. It is important to repeat that Austronesian and

Pro or contra Hispanisms

261

Spanish are no longer alone in their competition. There is now also a third category of items – many with an English etymology – which participate in this competition although their numbers are still relatively small. Since we have seen above that for many speakers, it is next to impossible to tell Austronesian etyma and Spanish etyma apart, even our initial hypothesis is standing on shaky grounds: The average speaker of modern Chamoru is aware of the fact that historically, his/her language underwent massive Hispanisation. However, this superficial awareness does not imply that the individual speaker can easily identify each and every Hispanism. For many of these, identification requires philological expertise. In the absence of this skill among laymen, Spanish words and Austronesian words form an amorphous mass, in a manner of speaking, and thus, native speakers of Chamoru do not necessarily have different attitudes towards the items in question. That there is a moderate pro-Austronesian tendency cannot be denied. However, this tendency is at its strongest when people are asked to reflect about language. When they are made to use language, their behaviour often contradicts their previously expressed attitudes.

5. Conclusions Our concluding remarks are meant as an outlook on research that has to be done in the future. In the above, we have demonstrated that native speakers of Chamoru have attitudes which reflect a non-egalitarian status of Austronesian and Spanish elements. However, the subjective statements about correctness and frequency of use are often at odds with the same speaker’s actual performance in translation tasks. This observation is something one would expect as self-assessment and actual performance are linked together closely only with very alert speakers of any language. In this way, Chamoru speakers most probably do not behave differently from native speakers of other languages when asked to make statements about their etymologically mixed lexicon and are then observed in their actual language production. Since Spanish as a prestige language has disappeared from the island a century ago already and Chamoru-Spanish bilingualism has given way to Chamoru-English bilingualism in the meantime, native speakers of Chamoru normally do not know any Spanish at all. This lack of familiarity with Spanish is the major cause for the uncertainty of some Chamoru informants to identify Spanish-derived items. For them at least, Spanish-derived items and their Austronesian counterparts are integrated parts of present-day Chamoru and can no longer be separated from each

262 Rosa Salas Palomo and Thomas Stolz other. On the other hand, the absence of Spanish as a competitor of Chamoru on Guam also reduces the necessity to censure the use of Hispanisms as Spanish is no threat to the autochthonous language (any more). One lesson our study teaches is that excessive purism aiming at eliminating all Hispanisms from everyday language is bound to fail because speakers will not be inclined to give up using words which are fully integrated. This is in line with what has been observed in other indigenous speech-communities worldwide whose attempts at ridding their language of foreign elements had disastrous effects (Flores Farfán 2003: 308–309). There remain many tasks for the future. Our study is only the first step towards a more extended and more thorough investigation of the etymologically mixed lexicon of Chamoru. On the sociolinguistic side, there is an urgent need for us to check our findings with a much larger group of informants among whom younger speakers (below the age of 30 including teenagers and children), less educated people and residents of regions outside Central Guam should be given more space. Some of the parameters – especially the number of siblings – are less enlightening than others and thus may be cancelled from future studies. However, there are other sociolinguistic parameters whose inclusion in our research program may make more sense such as e.g. ethnicity and linguistic background of husband/wife, use of languages in the family (with partner, with children), writing/reading competence in Chamoru, knowledge of Spanish as a foreign language, etc. Moreover, research on the mixed lexicon also opens up many possibilities for structure and content oriented linguistic investigations. As a byproduct, our study has revealed that major semantic changes have been taking place such that differences in the meaning of Austronesian-Spanish word pairs which were still clear to the older speakers have been blurred and thus the younger generations treat the words as full synonyms. This is a promising field of investigation – a micro-study of the historical semantics of (Spanish elements in) Chamoru. At the same time, working with a larger set of word pairs may lead eventually to the creation of an etymological dictionary of the language. Another important aspect is the possibility to see which of the words get which scores, i.e. we can determine whether or not, for a given word pair, the Austronesian option is chosen more often than the Spanish one and vice versa. From this procedure, a ranking will result which allows us to single out rare birds among the alternatives. As we have seen above, identifying Hispanisms is a difficult task for native speakers, at least sometimes. This effect could also be researched in the sense that one makes an inventory of those Hispanisms the average speakers fails to recognise as such. This list could be complemented by a catalogue

Pro or contra Hispanisms

263

of putative Hispanisms, etc. Taken together, all these micro-studies contribute to something one could term the “realistic” lexicon of modern Chamoru which comprises all those items speakers normally consider proper Chamoru. To achieve this, more attention has to be paid also to third-category items. We need to know which words belong to this category and with which items from the word pairs they compete. We are confident that all this is not only feasible but also to the benefit of general linguistics, language contact studies and, last but not least, the Chamoru people themselves who learn more about the properties of their language. Furthermore, we are convinced that it is worthwhile carrying out similar studies for other indigenous languages.14 Given the many cases of Hispanisation constellations worldwide, for instance, chances are that these studies of individual languages can be integrated into one huge comparative study which looks at the similarities and dissimilarities of attitudes and structural aspects associated with mixed lexica.

Notes 1.

2. 3.

4.

The research on which this study is based was originally conducted by Rosa Salas Palomo in Spring 2005. She likes to express her gratitude to her students at the University of Guam, Omaira Brunal-Perry, Velma A. Sablan, the University of Guam, the staff at the Micronesian Language Institute (Mye, Doreen and Zander), Marie (an important volunteer), and her family (Joe, Joey, Angel, John and Corey). Thomas Stolz joined her as co-author during his sojourn as visiting scholar at the Micronesian Language Center (Guam) in May 2007. He is grateful to the University of Bremen and the Sparkasse in Bremen for their financial support. Words of thanks also go to the people at the Micronesian Language Institute whose kindness made working there a real pleasure. Christel and Alma deserve thanks, too, for their patience. Jessica Messerschmidt kindly clarified some bibliographical issues. Both authors are also indebted to Dik Bakker for his valuable comments on a draft version of this article. Dångkolo na si Yu’os ma’åse! For a detailed account of Hispanisation processes in Chamoru going beyond the lexicon, cf. Pagel (2003). The exact amount of Hispanisms in contemporary Chamoru is still a matter of debate. Albalá and Rodríguez-Ponga (1986: 25) count 54% which is also the estimate in Rodríguez-Ponga (1995: 91) who quotes a number of older sources which give much higher (but unrealistic) percentages for the Spanish share. Stolz (1998) has much lower values for Hispanisms in Chamorro. Cf. also Solenberger (1962) who discusses prestige issues at a time when native speakers of Chamoru still formed the vast majority of the inhabitants of Guam.

264 Rosa Salas Palomo and Thomas Stolz 5.

A pertinent example of this cultural practice is the popular group Guma’ Pålu Li’e’ who de-Hispanise their lyrics by way of reconstructing Chamoru/Austronesian forms or find equivalents in other Austronesian languages. 6. For general questions related to language loyalty of the inhabitants of Guam, cf. the two articles by Riley (1975, 1980). 7. Our assumptions are based on the established “traditional” categories of sociolinguistic research (gender, age, education, residence, etc.) as reiterated in Stockwell (2006), Ungeheuer et al. (2006) and Llamas (2007). For obvious reasons, this study looks at a selection of parameters. Other relevant parameters, such as profession and, religion have not been included here. Note that the same categories are also crucial for studies on language endangerment (Tsunoda 2006). 8. Our ideas about the role of siblings are based on Cicirelli (1995) who provided a longitudinal psychological study of sibling relations. 9. AUS = Austronesian, ERG = ergative, H = hypothesis, IMP = imperative, INCL = inclusive, Q = question, SP = Spanish 10. Only to give an impressionistic idea of the uncertainty of speakers of this age group as to the etymology of the vocabulary they use, we mention that in the course of this investigation, native speakers suggested to us to also look at additional word pairs like katsunes/pantalones ‘trousers’, busuelo/tason ‘kitchen utensil’, boti/batko ‘boat’, mahot/atkade ‘mayor’, espia/aligao ‘look for’, etc. It was believed that one of these items was Austronesian and the other Spanish. However, in these cases, both items have a Spanish etymology, namely calzones and pantalones ‘trousers’, pozuelo and tazón ‘cup’, bote and barco ‘boat’, mayor ‘head of X’ and alcalde ‘lord mayor’, espiar ‘to spy on’ and alegado ‘put forward’ (Rodríguez-Ponga 1995). In some of these and similar cases, sound changes and semantic developments have blurred the etymological ties of the Chamoru word and its Spanish etymological source. Interestingly, speakers considered peskadot ‘hunter’ (< pescador ‘fisherman’) a Hispanism but often failed to recognise the related verb peska/piska ‘to hunt’ (< pescar ‘to fish’) as originally Spanish-derived. In addition, for the ordinal numbers, tetseru ‘third’ (< tercero ‘third’) was usually identified as Spanish whereas the synonym minatres ‘third’ passed as Austronesian although it contains the Spanish-derived numeral tres ‘three’ (combined with the Austronesian ordinal prefix mina-) which is also the usual cardinal numeral for three in Chamoru (Rodríguez-Ponga 2001). 11. Some of these semantic distinctions are still reflected in the dictionary by Topping et al. (1975) as e.g. AUS addeng ‘human leg/foot’ ~ SP patås ‘animal leg/foot’; AUS ha’of ‘interior of chest’ ~ SP pecho ‘fleshy part of chest’. 12. This informant has unusually high percentages for Austronesian in all three parts of the questionnaire, namely 78% for part 1, 64% for part 2 and 67% for part 3 yielding an average of 70% pro Austronesian. This is a ratio that is matched by none of the other 120 respondents. Thus, we felt entitled to ignore this singularity in the statistics.

Pro or contra Hispanisms

265

13. This bipartition is also justified by the fact that internal migration on Guam has been such that for practical reasons, population exchanges between the northern and southern regions of the island are common whereas residents of Central Guam are less prone to moved house. 14. Stolz and Stolz (1996, 1997 and 2001) have shown that parallel processes of Hispanisation occur in scores of languages spoken on both shores of the Pacific and thus provide the ideal testing ground for theories of language contact.

References Albalá, Carmen-Paloma 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, Carmen-Paloma and Rafael Rodríguez-Ponga 1986 Relaciones de España con las Isalas Marianas. La lengua Chamorra. Madrid: Fundación Juan March. Cicirelli, Victor G. 1995 Sibling relationships across the life span. New York: Plenum Press. Flores Farfán, José Antonio 2003 Nahuatl purism: between language innovation, maintenance and shift. In Purism in minor languages, endangered languages, regional languages, mixed languages, Joseph Brincat et al. (eds.), 281–314. Bochum: Universitätsverlag Dr. N. Brockmeyer. Fought, Carmen 2006 Language and ethnicity. Cambridge: Cambridge University Press. Hernández-Legorreta, Cutberto 2001 En búsqueda de un nacionalismo en Guam. In Imperios y naciones en el Pacífico. Vol. II: Colonialismo e identidad nacional en Flilipinas y Micronesia, María Dolores Elizalde et al. (eds.), 407–418. Madrid: Consejo Superior de Investigaciones Científicas. Kumisi’on i Fino’ Chamorru 1983 Kuenta – Report. Agaña: Government of Guam. Llamas, Carmen (ed.) 2007 The Routledge companion to sociolinguistics. London: Routledge. Odo, Carol 1971 Variation in Hawaiian English: underlying R. Working Papers in Linguistics 3 (2): 1–30. Pagel, Steve 2003 Zum Sprachkontakt Spanisch-Chamoru auf Guam und den Nördlichen Marianen. Unpublished M.A.-thesis. Halle: Martin-Luther-Universität Halle-Wittenberg. Institut für Romanistik.

266 Rosa Salas Palomo and Thomas Stolz Riley, George A. 1975 Language loyalty and ethnocentrism in the Guamanian speech community. Anthropological Linguistics 17: 286–292. 1980 Language loyality and ethnocentrism in the Guamanian speech community: seven years later.’ Anthropological Linguistics 22: 329–333. Rodríguez-Ponga, Rafael 1995 El elemento español en la lengua chamorra (Islas Marianas). Unpublished PhD-thesis. Madrid: Universidad Complutense de Madrid, Facultad de Filologia. 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 a.M.: Vervuert. Solenberger, Robert R. 1962 The social meaning of language choice in the Marianas.’ Anthropological Linguistics 4: 59–64. Stockwell, Peter 2006 Sociolinguistics: a resource book for students. London: Routledge. Stolz, Christel and Thomas Stolz 1996 Funktionswortentlehnung in Mesoamerika: Spanisch-amerindischer Sprachkontakt. Sprachtypologie und Universalienforschung 49(1): 86–123. 1997 Universelle Hispanismen? Von Manila über Lima bis Mexiko und zurück: Muster bei der Entlehnung spanischer Funktionswörter in die indigenen Sprachen Amerikas und Austronesiens. Orbis 39: 1–77. 2001 Hispanicised comparative constructions in indigenous languages of Austronesia and the Americas. In Lo propio y lo ajeno en las lenguas austronésicas y amerindias, Klaus Zimmermann and Thomas Stolz (eds.), 35–56. Frankfurt a.M.: Vervuert. Stolz, Thomas 1998 Die Hispanität des Chamorro als sprachwissenschaftliches Problem. Iberoamericana 70 (2): 5–38. 2003 Not quite the right mixture: Chamorro and Malti as candidates for the status of mixed language. In The Mixed Languages debate. Theoretical and empirical advance, Yaron Matras, and Peter Bakker (eds.), 271–315. Berlin /New York: Mouton de Gruyter. Topping, Donald, Pedro Ogo, and Bernadita C. Dungca 1975 Chamorro-English Dictionary. Honolulu: University of Hawaii Press. Tsunoda, Tasaku 2006 Language endangerment and language revitalization. An introduction. Berlin /New York: Mouton de Gruyter.

Pro or contra Hispanisms

267

Underwood, Robert 1999 The practice of identity for the Chamorro people. The challenge of Hispanicization. In 1898: España y el Pacífico. Interpretación del pasado, realidad del presente, Miguel Luque Talaván et al. (eds.), 527–531. Madrid: Asociación Española de Etudios del Pacífico. Ungeheuer, Gerold et al. 2006 Sociolinguistics: an international handbook of the science of language and society. Berlin /New York: Mouton de Gruyter. Zimmermann, Klaus and Thomas Stolz (eds.) 2001 Lo propio y lo ajeno en las lenguas austronésicas y amerindias. Frankfurt a.M.: Vervuert.

Index of subjects

absolutive, 22, 32–35, 44, 193–194 accent, 68, 207 accommodation, 106–108 accusative, 111, 116, 143, 193 acoustic, 84 acquisition, 27, 52, 123, 156, 189 active, 53, 150, 160–161, 183, 209 adaptation, 65, 84–85, 100, 102, 104, 106, 114, 122, 214 adjacent, 90, 99 adjective, 5, 10, 18, 22, 50, 55, 59, 72, 98–100, 102–108, 111, 113–114, 131, 174–175, 183–184, 191, 217 adoption, 154 adposition, 4, 7, 15, 22, 121, 127, 133, 135, 139, 141–142 adverb, 50, 55–56, 72, 81, 98–100, 102– 104, 106–107, 109–114, 116, 134– 136, 138, 212, 220, 223 adversative, 44 affiliation, v, 190, 193 affix, 4, 50, 115, 128–131, 184, 210, 216–217 affricate, 23, 84–85, 189, 206 agent, 5, 134, 187 agglutinating, 3, 132, 136, 187 agreement, 180, 183, 210, 231 allative, 116, 134–135 allophone, 30, 182, 205 alphabet, 149, 205, 245 alveolar, 68, 189 analogy, 135, 141 analytical, 39 archaic/archaisms, 157, 224, 232 areal, 13, 90 assimilation, 31, 33, 61, 88, 106, 175

autochthonous, v–vi, 98, 155, 167–168, 170, 176, 180, 183, 185, 190, 192, 237, 262 auxiliary, 83, 186 beneficiary, 143 bilabial, 29–30, 44 bilingual, vii, 32, 39, 52, 55–56, 59, 73, 78, 80, 82, 90, 96, 116, 121, 124–125, 130, 143, 170–172, 174, 177, 187, 192–194, 212 bimorphemic, 82 borrow, 101–102, 141 borrowability, 127–128 borrowed, 20, 36, 55, 59–62, 69, 83, 86, 91, 98, 100–103, 106–108, 110, 112, 114, 124, 126–133, 137, 140–142, 160, 175–177, 183, 185, 187, 206 borrowers, 55, 129 borrowing, 20, 38, 51–53, 55, 62–63, 69– 70, 73, 83, 89, 95, 98, 100–101, 103, 106, 108, 110–111, 114–117, 123–124, 126–132, 136, 138–139, 141–142, 151, 160, 163, 171, 174–177, 182–184, 187, 190, 193–194, 227 borrowings, 53, 56, 59–62, 70, 83, 87, 89, 98, 101, 103–107, 109–114, 124, 143, 157, 160, 174–175, 182, 194, 206, 213 borrows, 101–102 calque, 40–41, 82, 231 catholicism, 86 causative, 5, 12, 18, 22, 36, 44, 50, 108, 116, 176, 217 christian, 18 church, 84, 152, 155, 215 circumpacific, 176, 185

270 Index of subjects classification, 95, 103–104, 111, 121, 134, 138, 142 classifier, 13–14, 19, 22, 79, 88, 91 clause, 4, 8, 16, 187, 194 clauses, 4, 8, 16, 22, 83, 115 clitic, 36 coda, 82 cognitive, 154, 187, 193 collective, 91, 180, 189, 237 colloquial, 136, 141, 206, 214 colonial, 3, 6, 14, 16, 18, 20–23, 28, 31, 38, 40, 43, 49, 153, 169–170, 211, 229, 231 colonialization, 203, 208 colony, 203 comitative, 117, 221, 232 comparative, v, 90, 167, 177, 184, 191, 193, 250, 253, 263 competence, 52, 65, 70, 155, 157, 190, 262 complement, 31, 127, 177, 184–185 complementizer, 78 compound, 81, 105, 136, 138 conflict, 33, 137, 168 conjugated, 69 conjunction, 50, 59, 69–70, 72–73, 109, 111, 115, 126, 131, 174–177, 182, 185, 191, 211–212, 227 conquest, 95–96, 213–214 conservation, 65, 78, 208 consonant, 79, 80, 82, 84–86, 90–91, 175, 182, 185, 206 constituent, 4, 6, 10, 14, 16, 22, 45, 187, 245 constraint, 80, 82, 95, 100, 114–115, 121, 126, 132–133, 141 construction, 4–5, 8–9, 11, 15–18, 21–23, 40, 45, 65, 69, 72–73, 83, 111, 115, 125, 128, 138–140, 142, 154, 186, 191, 210–212, 216, 221, 231, 232 contact, v–vii, 6, 14, 17–21, 27–28, 39, 43, 49, 51–53, 55–56, 59–64, 70–73, 95–96, 100, 121, 123, 125–126, 128–

133, 136, 141, 151, 161, 163, 167– 170, 174–177, 179–181, 183, 185, 187–188, 191–192, 194, 203–204, 213, 215, 221, 223, 240–241, 263, 265 contiguous, 101, 105 continuum, 49, 51–53, 61, 63, 71–72, 160, 171–172, 174–175, 177, 190, 192 control, 152, 175 convergence, 31, 34, 37, 42–44, 80, 82 conversation, 53, 55–56, 59, 63, 71, 173– 174 conversion, 156 coordinator, 125 copula, 4, 22, 216, 232 corpus, 49, 95, 108, 111–112, 114, 116, 182, 193, 230 courtesy, 40, 42 creation, 36, 45, 71, 130, 159, 262 creative, 123 creolization, 203, 214 culture, 55, 65, 68, 81, 83, 89, 116, 149– 150, 156–158, 162–163, 167–170, 183, 188, 190, 213, 224, 237, 240, 245, 264 dative, 117, 193 deaffricativization decimal, 21–22, 248 decline, 170, 188, 230 decrease, 73 deletion, 35–38, 41, 45, 79, 209, 230 demonstrative, 22, 117, 176, 193 dental, 23, 189 depuration, 64 derivation, 9, 17, 22, 41–42, 50, 109, 128–129, 135, 210, 216–217 designation, 55, 160, 172 determinant, 44, 192 diachronic, 100, 124, 126, 134, 171, 172 dialect, 15–19, 27, 29, 73, 87, 95, 111– 112, 114–115, 123, 141, 149, 152, 206, 215, 239 dialectal, 6, 33, 50, 72, 78, 88

Index of subjects diaphasic, 72, 171 diastratic, 72, 171 diglosia, 52 diminutive, 72, 108, 117, 131 discontinuous, 8, 49, 63–64 discourse, 33, 53, 55–56, 59–61, 71, 111, 125–126, 174–175, 185, 193, 213, 216, 223–224 displacement, 29, 35, 51 distinctive, 98, 109 distribution, 71, 98–99, 102, 112, 114, 212, 220–221, 243, 257 ditransitive, 11, 216, 221 diversification, 50 donor, v, vi, 150, 244 drift, 82, 176 dual, 185, 212 ecology, 159 education, 50, 65, 78, 82, 152, 156, 158, 161, 204, 230, 239–243, 244, 252, 258, 264 elicitation, 37, 45 elite, 169, 230 elliptical, 45 emphasis, 50, 101, 110, 180, 190 encode, 5, 11–12, 14–16, 18, 137, 175, 216 endangered, 29–30, 34–35, 159, 161 environment, 106, 141, 179–180, 192 epenthetic, 45 equative, 216 equatorial, vi equilibrium, 149–150, 158, 160–163 ergative, 185, 193–194, 264 etymology, 45, 78, 81, 218, 237–238, 245, 253, 259, 261–262, 264 etymon, 241, 244 evangelization, 96 evolution, 31, 158, 191 exception, v, 80, 82, 105, 143, 160, 187, 190–191, 214, 219 exclamation, 190, 214 exiles, 149, 155

271

existential, 4, 9, 178, 185, 232 expedition, 169 exploitation, 153, 156 expression, 5, 9, 36, 39–41, 108–109, 176, 189, 191, 214 extinction, 27, 29, 34, 43, 159, 162 female, 115, 174, 241–242, 248–249 feminine, 44, 87, 91, 183, 193, 210, 216 fieldwork, 29, 50, 78, 95, 97, 115, 224 finite, 37, 107 fluency, 55, 121, 128–129, 152–156, 160– 161 focus, v–vi, 39, 114, 150, 159, 169–170, 184, 187, 193, 204, 214, 244 foreign, 63, 84, 86, 90, 151, 153, 155– 158, 161, 168, 173–175, 182, 190, 231, 240, 248, 262 foreigner, 43, 45, 63, 152, 188 formal, 31, 78, 132, 134, 136, 137, 141– 142, 151–152, 160–161, 183 fossilized, 28, 40–42, 45, 123 frequency, 59, 95, 101, 106, 110, 161, 174–175, 204, 246, 251, 253, 257– 259, 261 fricative, 23, 29–30, 44, 84, 189 gallicisms, 91 gap, 125–126, 138–139, 141–142, 249, 251, 254, 259, 260 gender, 50, 97, 108, 115–116, 174, 178, 183–184, 187, 210, 231, 239–240, 242, 246, 264 generation, 29, 32, 39, 97, 121, 132, 150– 151, 153, 155, 161, 169, 171, 242, 244, 247, 258, 262 generative, 133 genesis, vi, 45, 130, 143, 217 genetic, v, 127, 159, 193 genitive, 117, 143 gerund, 117, 131, 143 glide, 30, 84, 206 glottal, 23, 31–32, 44, 216 government, 84, 87, 128, 152–153, 188 grammaticalization, 100, 187, 193

272 Index of subjects grammaticization, 64 graphemes, 90 heads, v, 8, 12, 66, 69, 88, 98–105, 112– 113, 116, 131, 264 heritage, 156, 158 hesitative, 73 hierarchy, 99–101, 124, 132, 153 hispanicisation, viii hispanism, vii, 77, 87–89, 173–177, 179, 182–186, 193, 203, 208, 214, 237–238, 240–241, 259, 261–264 hispanisms, vii, 77, 87–89, 173–175, 177, 179, 182–183, 185–186, 193, 203, 208, 214, 237–238, 240–241, 259, 262–263 hispanized, vii, 28, 30, 33, 35, 37–39, 42–43, 61, 69, 71, 153, 184, 187–188, 212, 231 hispanophone, 55, 61, 230 honorific, 23, 27, 44–45 hybrid, 133, 155, 157 hypercorrection, 31 ideology, 38, 174, 189, 194 idiolect, 171 idiom, 44, 82, 138 idiosyncratic, 107 imitation, 176 immigrant, 152, 155, 190 imperative, 23, 32, 37, 44, 82, 264 imperfect, 72, 123 imperfective, 23, 218, 232 inanimate, 4, 7, 13, 19, 22–23 inchoative, 134 incorporation, 5, 11, 18, 22, 38, 116, 121– 122, 124–128, 132–133, 136–137, 141– 142, 175 indefinite, 184, 193 indian, 96–97 indicative, 37, 43, 85, 88, 90, 137, 176, 193, 229, 232, 239, 248, 257 indigenous, v, vi, 43, 63, 72, 81, 83, 85, 87, 89–90, 143, 149–153, 155–156, 158, 161–163, 262–263 indios, 168

infinitive, 20, 34, 37, 40, 89, 107, 117, 186, 193, 229 inflection, 23, 79, 107, 109, 115, 128–129, 210, 229 informant, 97, 115–116, 171, 239–244, 246–249, 251–262, 264 innovation, 28, 31–32, 35, 37, 40–42, 78, 128, 149, 157–158, 183, 186–187 input, 154, 215–216, 226 insertion, 177 instrumental, 5, 11, 45, 117 integration, vii, 128, 153, 174, 178, 209 interaction, 45, 132, 191, 193, 239 interethnic, 97 interference, 52, 72, 79, 123, 161–162, 171, 179, 189 interjections, 50, 78 interlanguages, 52, 71 interlocutors, 45 intermarriage, 169 international, v, vi internet, 155, 181, 190–191 interpelative, 117 interpersonal, 156 interrogative, 44, 88, 185, 213 intertwined, 151, 214, 226 intervocalic, 44, 86, 172, 208, 230 intransitive, 36, 38, 41, 50, 194 intrasentential, 116, 123 invasion, 3, 6, 96 irrealis, 23, 218, 232 jargon, 229 jesuit, 168 langue, 193 lateral, 23, 84 layers, 28, 203, 219 lexeme, 61–62, 81, 95, 98–99, 101–103, 106, 108, 110–112, 114, 136, 151, 157, 175, 182–183, 187, 206, 208–209, 214, 217 lexical, 32, 34, 38, 40, 42, 44, 49, 51, 53, 59, 64, 79, 83, 86–89, 95, 99–107, 110– 112, 114–116, 122–124, 127–128, 130,

Index of subjects 132–138, 140–142, 175, 177, 179–180, 182, 187, 189–191, 203–204, 220, 226, 231, 237–238 lexicalization, 100, 103 lexicon, vii, 41, 49, 55, 63–64, 68, 70– 71, 78, 87, 89, 95, 98, 114–115, 129– 131, 133, 135, 140, 143, 154, 160, 174, 182, 204, 213, 215, 231–232, 237, 240, 244–245, 261–263 limitative, 117, 143 literacy, 157–158 literature, 43, 63, 68, 89, 100, 126, 133, 203, 205, 220 loan, 14, 22, 32–34, 36, 39, 44, 53, 55, 84, 86, 129, 174, 182, 184, 204–207, 214, 230–231 loanword, 20, 33, 59, 73, 78–79, 81, 83– 88, 106, 112, 206, 238 locative, 117, 134–135, 216, 232 mainland, 157 maintenance, 51–52, 96, 106, 124, 171, 192, 193 marked, 13, 19, 40, 68, 174, 183–185, 215–216, 221, 224 marker, 31–35, 37, 40, 45, 82, 108, 111, 176, 184–187, 192, 210, 212, 216–218, 220–221, 224, 227, 229, 232, 240–241 marking, 5, 98, 108, 138, 218, 220 meaning, 17, 20, 23, 34, 44, 59, 71, 80– 82, 86–87, 89–90, 107–108, 114, 116– 117, 122, 124, 137, 167, 191, 194, 214, 224, 232, 244, 247, 249, 262 mechanism, vii, 130, 188 mestizo, 96, 169 metaphor, 70, 167, 172 metathesis, 91, 107 metatypy, 188 migrants, 116 military, 152–153 minority, 73, 80, 123, 159, 203–204, 237 missionaries, 77, 205, 208 missionary, 172, 211 missions, 168

273

mixed, vi, vii, 49, 52, 62, 69, 71–72, 78, 96–97, 121, 125, 130, 143, 155–156, 167, 170, 180–181, 187, 192, 194, 203, 214, 228, 239, 251, 261–263 mixing, 49, 51–52, 61, 78, 155, 191, 203, 226 modal, 50, 177, 187, 211–212 modification, 84, 103, 112, 167, 175, 183, 185, 187 modifier, 10, 98–106, 112–114, 131 monolingual, 28, 31, 39, 53, 61, 78, 116, 124, 143, 155, 170, 174, 207 monotransitive, 12, 18 monovalent, 135–136 mood, 178, 185, 216–217 morpheme, 14, 22, 30, 79, 106, 108, 111, 128, 208, 216–217 morphology, 3, 21, 31–32, 35–37, 41– 42, 45, 77, 79, 82, 95, 98, 106–108, 122, 129–130, 175, 183, 187, 209– 211, 215, 217, 229 morphophonemic, 28, 174 morphosyntactic, 49, 69, 109, 170, 176, 178, 184, 188–189, 212, 230, 231 mother, 34–35, 39, 43, 50, 108, 179, 222, 228–229 narrative, 69, 97, 116, 150–151, 224 nasal, 31, 84 nasality, 31 native, vi, viii, 29, 35, 65, 77, 79, 82, 87, 88–89, 95–96, 102, 112, 121, 123, 125, 126, 128–129, 151, 153, 167–168, 170, 175, 206, 209–210, 214, 225, 230–231, 237–238, 240–242, 244, 248, 252, 261–264 navigator, 149 negation, 23 negative, 44, 91, 117, 185 negator, 193, 220, 232 neighbour, 225, 226 neologism, 20, 33, 38, 70, 89, 155 neuter, 50 newspaper, 88

274 Index of subjects nominalization, 184 nominalizer, 91, 143, 176, 184, 217, 232 nonbasic, 55 nonce, 174 normalization, 77, 80 normative, 211, 231 notion, 43, 62, 126 notional, 105, 134 noun, 4–5, 7–22, 33, 55, 59, 62–63, 73, 81, 83, 86, 88, 98–114, 128, 140–141, 174–175, 183–184, 216, 246–247 number, v, vii, 5, 7, 13, 18–19, 23, 27, 33, 39, 45, 50, 55, 59, 70, 72, 77, 79, 81, 87, 90, 96, 98, 100–101, 108–109, 111–112, 114–117, 123–126, 141, 149– 150, 153–154, 157, 159–161, 169– 170, 178, 180, 182, 185, 194, 204, 212, 214, 217, 231, 237–239, 241– 244, 248–249, 254, 256, 259, 261–264 numerals, 5, 13–14, 19, 21–22, 38–39, 50, 55, 73, 79, 83, 88–89, 91, 108, 111, 115, 182, 264 object, 11–13, 23, 36–38, 40, 44–45, 86– 89, 91, 117, 177, 194, 216, 220–221, 232 obligation, 59, 176–177, 181, 187, 193 oblique, 227, 232 onset, 82 opaque, 44–45 oppression, 151, 188 optative, 32, 37, 213 oral, 31, 49, 50, 64, 70, 158, 161, 215 orthographic, 43, 65, 80 palatalized, 80, 206 paradigm, vii, 37, 185, 212 paratactic, 69, 73, 125, 176 parents, 153, 181, 194 participant, v, 49, 193, 238, 241–242, 247–248 participle, 23, 131 particle, 23, 50, 55–56, 59, 73, 108, 124, 126, 174–175, 184, 191, 212–213, 216, 223–224, 232

passive, 50, 53, 55, 134, 153, 156–157, 190, 193, 216 penultimate, 109 perception, 52, 151, 160–162 perfect, 36, 44, 82, 117, 143, 218, 229, 232 perfective, 229 periphrases, 109 person, 7, 9, 14, 16–17, 23, 30, 35, 39, 44, 50, 73, 88, 91, 116, 143, 169, 186, 189, 191, 193, 216, 217, 224, 229, 232, 242–243, 249 phase, 79 phenomenon, v–vi, 27–28, 31–32, 35, 43, 51, 53, 71, 73, 78–79, 107, 124, 128, 151, 153, 159, 172, 182, 191, 220, 223, 231 philological, 261 phoneme, 29, 34, 44, 68, 80, 82, 84, 129, 205–206, 208 phonemic, 28, 43, 174, 206 phonetic, 31, 43, 51, 53, 61, 79, 86, 106, 175, 189 phonological, 30–32, 34, 43–44, 79, 95, 106–107, 109, 111, 122, 128–129, 133, 137, 139, 141, 175, 178, 207, 213–214, 230 phonology, vii, 28, 34, 77, 79, 84, 90, 107, 122, 129–130, 175, 182, 204–205, 232 phonotactic, 84 phrase, 8, 69, 98–106, 109–110, 112–113, 116, 160, 185, 191, 194, 206, 209, 245 pidgin, v, 43, 45, 209, 221, 228–229, 232 pidginization, 43, 203, 228 pidginized, 228 plural, 5, 13, 16, 19, 22–23, 28, 31–32, 44, 50, 73, 108, 117, 143, 185–186, 193, 212, 216, 219, 232 plurality, 31 politeness, 213, 219 politics, 63, 81, 152, 155–157, 159, 161– 162, 169, 170, 175, 193, 214, 232, 238 polyglot, 52

Index of subjects possessive, 4, 8, 9, 15–17, 21–22, 33, 35, 39, 44, 50, 83, 111, 136, 143, 175, 193, 229, 232 possessor, 4, 8–9, 12, 14, 16, 23, 175 postnominal, 137 postposition, 4, 7, 15, 22, 39, 83, 113, 127, 137, 139, 142–143 postpositions, vii, 22, 79, 127, 133, 137, 141 pragmatic, 44, 82, 108, 125, 184 pragmatics, 27, 40, 45, 191, 232 prayer, 152 predicate, 4, 9, 100–105, 112–113, 116, 133–136, 138–141, 227 predication predicative, 99, 134–142 prediction, 102–103, 105, 110, 115, 155, 260 prefix, 8–9, 14, 16–17, 23, 36–40, 50, 73, 176, 211, 216, 264 prehead, 4, 8, 16 prenominal, 212 preposed, 136 preposition, vii, 16, 21–22, 44, 50, 69– 70, 121, 125, 127–128, 132–142, 175, 182, 185–186, 212, 216, 232 prepositional, 39, 62, 69, 105, 134, 137– 138, 142–143 present, 21–22, 27–28, 31–34, 36–37, 39, 44, 49, 69, 73, 78–83, 90, 95–97, 100, 108, 110, 115, 117, 126, 133, 137, 139, 141–143, 150, 158, 160–162, 167, 169, 172, 174, 176, 179, 182, 185, 188, 193, 203, 209, 214, 231–232, 238, 243, 253 prestige, 52, 64, 124, 156, 158, 231, 240, 260–261, 263 prestigious, 64, 172, 237, 258 preterit, 10, 23 preverbal, 216–217 priest, 81, 85, 152, 206 productive, 18, 72, 108, 131, 138, 183, 216–217 productivity, 30, 38

275

proficiency, 180 progressive, 44, 49, 63, 71, 143 pronominal, 4, 9, 172 pronoun, 23, 37, 73, 83, 91, 176–177, 212, 219 pronouns, 194, 212, 216, 219, 229 pronunciation, 79–80, 82, 207, 208–209 prosodic, 129, 207 prototype, 114, 171 prototypical, 113–114 psycholinguistic, 151, 154 psychological, 264 public, 152, 156–157, 160–161, 170, 188, 215, 231, 241, 245 punctuation, 68–69 purism, 49, 63–65, 70–72, 262 purist, vi, 31, 38–39, 161, 163, 174, 194 qualitative, 126, 130 quantifier, 88 rank, 96, 104, 241, 252 ranking, 232, 262 ranks, 96, 104, 241, 252 rapanuiisms, 179, 193 reanalysis, 16, 209 recipient, v–vii, 88, 112–113, 122, 216, 221 reciprocal, 217 reconstruction, 82, 150 reduplication, 23, 39, 44, 110, 176, 185, 209 referential, 98, 100–105, 112–113 reflexive, 23, 44, 50, 143 regional, 50–51, 65, 72, 79, 215 register, 31, 43, 70, 237, 239 regularization, 122 reinforcement, 107, 170 reinvention, 159 relexification, vii, 121–123, 127–128, 130, 133, 137, 139, 141–142, 232 relexified, 108, 122, 131, 136, 141, 203, 214, 221 religion, 67, 170, 264 renaissance, 160, 170

276 Index of subjects repair, 245 repertoire, 30, 32, 40, 44 reportative, 117 residence, 115, 239, 243, 252–253, 258– 259, 264 restriction, 123, 127, 129, 132–133, 137, 142, 191, 193 restructuring, 122, 172 resyllabification, 206–207 retroactive, 63, 70 reverse, vii, 49, 155, 160 –162 revitalization, 65, 70, 167–169 romancisation, v–viii romanicisation, viii romanisation, 72 root, 7, 9, 11–12, 15, 17, 20, 38, 50, 81– 82, 132, 136–137, 141, 183, 210, 217 rule, 40, 42, 68–69, 79, 84, 86, 95, 133, 174–175, 178–179, 186, 203, 205, 209, 229 sailors, 91 scenarios, 83, 97, 116, 124, 141, 167, 171– 172 script, 43, 45, 150 segmental, 84 semantics, 40, 213, 262 sense, v, 27, 34, 79, 132, 151, 167–168, 175–176, 193, 244, 249, 257, 259, 262 sentence, 14, 68–69, 73, 81–83, 116, 122, 139, 176, 204, 220, 240, 245–246 settlement, 149, 151, 162, 170, 231, 243 settlers, 149 sibilant, 107 significant, 69, 71, 83, 86, 95, 112, 114– 115, 177, 187, 231, 248 similarities, 44, 203, 208, 211, 229, 232, 263 simplification, 28, 30, 33, 35, 39, 43, 45 simplified, 34–36, 43, 45, 123, 143 simplify, 34–36, 43, 45, 123, 143 singular, 5, 9, 13, 17, 23, 32–33, 44, 73, 87, 91, 117, 143, 176, 185–186, 193, 216, 232

situation, v–vii, 27, 35, 43, 45, 49, 51– 53, 55, 61, 64, 72, 78–80, 83, 90, 99, 121, 123, 129–130, 132, 136, 150– 156, 171, 174–176, 183, 185, 192, 203, 217, 227, 231, 240, 257, 259 situational, 154 slave, 9, 169, 172 slavicisation society, 28–89, 96–97, 149, 158, 162, 214 sociolectal, 6 sociolinguistic, 27, 29, 43, 71–72, 77, 95– 96, 115, 121, 125, 150, 153, 155, 160, 163, 187, 192, 203, 238, 242, 249, 257, 262, 264 sociopsychological, 190 soldier, 85–86, 208, 228–229 spatial, 14 –15 speaker, 34, 37, 49, 60–61, 65, 77, 79, 81–82, 106–107, 153–154, 175–176, 178, 180, 190–191, 238, 240, 242, 244, 257, 259, 261 speakers, vi–viii, 6, 14, 20–21, 29, 34, 41–43, 45, 50, 52, 55–56, 59, 61, 63, 65, 73, 78, 80–82, 90, 96–97, 100, 114, 116, 121, 123–124, 128, 141, 143, 151–155, 159–160, 170–172, 174–175, 177–180, 182–183, 188, 190–194, 204, 207, 209–210, 214– 215, 219, 230–232, 237–241, 244, 246–248, 251–254, 258–264 speech, 5–6, 14, 37, 45, 51, 55, 60, 65, 69–70, 78, 89, 95–101, 103–107, 110– 111, 113–114, 116, 156, 160, 171– 172, 174–175, 178, 180, 182–183, 185, 189, 191–192, 194, 209, 213– 215, 229–230, 237, 240–241, 245 spelling, 68, 192, 208–209 stable, 252 standard, 43, 52, 65, 70, 72–73, 162, 172, 178, 189, 208, 230 standardization, 49, 63–64, 65, 71, 194 statistical, 95, 98, 116, 247–248, 250, 255, 258

Index of subjects statistics, 95, 110, 143, 155, 240, 244, 247–249, 260, 264 stem, 7, 11, 13, 18, 20, 31, 33, 36, 107, 122, 151, 230 stigmatization, 192 stigmatize, 193 stigmatized, 43 stops, 23, 31–32, 44, 78, 80, 84, 216 stratification, 242 stress, 28, 109, 207 structural, vii, 115–116, 126, 132, 142, 177–179, 181, 187, 203–204, 215– 216, 229, 231, 263 structurally, 125, 141, 215, 226, 229 style, 70, 172, 191, 194, 239 stylistic, 31, 154 subcategorization, 122, 140 subclass, 116 subfamily, 115 subgroup, 149 subject, 9, 14, 17, 23, 31, 37–38, 95, 109, 116, 149, 153, 162, 168, 177, 184, 194, 209, 216, 219, 220 subordinate, 115, 123 subordinators, 78, 117, 125 subsistence, 97 substantive, 21, 50, 124, 127 substrate, 149, 171, 193, 223 subvariety, 178 –179, 215 suffix, 5, 9, 12, 17, 20, 23, 31–33, 36, 50, 73, 83, 116, 130–131, 134, 137–139, 142, 176, 210 superlative, 108, 117, 184, 193 superstrate, 170 –171, 193 supplement, 97 suprasegmental, 68, 84 switch, 98, 109, 143, 174 syllabic, 81–82, 107, 130, 205 syllable, 68, 79, 81–82, 85, 87, 107, 109, 175, 205–206, 216 symbol, 205–206, 240 synchronic, 100, 171–172, 192–193, 221 synchronically, 5, 194, 214, 231

277

syncretic, 155, 157, 160, 162, 172, 193– 194 syncretism, 160, 193 synonym, 70, 139, 244–246, 262, 264 synonymous, 142, 189, 244, 247 syntactic, 4, 6, 13, 18, 40, 69, 79, 98–106, 109–110, 112–114, 122, 125, 127–129, 175–177, 245–246 syntagma, 62 syntax, 38, 77–79, 82, 109, 122, 125, 129, 131–132, 141, 211, 220 topicalizer, 117 topics, v, 21, 27, 53, 97, 143, 157, 167, 184, 189, 238 toponym, 7, 15–16 towns, 17, 56, 59–62, 96–97, 215, 225, 229 tradition, v, 44, 65, 154, 205 transcription, 53, 56, 60, 72–73, 138 transcripts, 69, 98 transfer, 60, 62, 70, 179, 190–191 transformation, 130 transitive, 6, 36, 38, 40–42, 45, 50, 83, 194, 209, 221, 232 transitivity, 36 translation, 14, 39, 68, 72, 78, 215, 226– 227, 244, 246–247, 259, 261 travel, 116 tribes, 213 trigger, 123 typological, vi–vii, 3, 6, 14, 21–22, 95, 101–104, 114, 127, 129, 130–131, 133, 136–137, 141, 180, 211 typology, viii, 41, 127, 133, 143, 192 unanalyzable, 106 ungrammatical, 35, 221 unidirectional, 44 unification, 169 uninflected, 89 universal, 125–128, 139, 141–143 universals, 125, 141, 143 unmarked, 6, 14, 41, 55, 175, 177, 183, 185, 193, 214, 220

278 Index of subjects unspecified, 5, 168, 242 unstressed, 80, 209, 216 urban, 78, 121, 204 usage, 27, 33, 37–39, 52–53, 55–56, 59, 60, 70, 246, 251–252, 254 use, 20, 27, 29, 34, 50–52, 65, 68–70, 73, 78–79, 81–83, 88, 96, 99–103, 105, 110, 112–114, 121, 131, 135– 136, 141, 150–151, 153–154, 156– 159, 161, 163, 171, 173–176, 178– 179, 183–184, 186–188, 190–191, 194, 210–211, 215, 218, 229, 238, 240, 244, 246, 248–249, 251–252, 257–262, 264 uses, 69, 79, 82–83, 88, 99, 101, 134, 160 utterance, 33 valency, 21, 135, 136 variability, 27, 31, 79 variable, vi, 79–80 variant, 33, 187 variation, 6, 51, 65, 72, 79–80, 100, 153, 174, 208, 255 variety, vi–vii, 27–30, 33–35, 37–39, 41– 43, 45, 49–53, 55, 56, 59–65, 68, 70– 72, 78, 88, 90, 97, 108, 115, 121, 127, 134, 143, 154, 160, 167, 170–172, 177–180, 189–190, 192–193, 203, 206, 208, 215, 217–218, 224, 230 velar, 68 verb, 4–6, 9, 10–14, 16, 18, 20, 32, 36, 40–41, 45, 50, 55, 62, 69–70, 73, 83,

89, 91, 98–99, 101–103, 106–108, 111–113, 128, 143, 174, 176, 178, 185, 189, 209, 212, 217–218, 221, 229, 231, 246–247, 264 verbal, 31, 40, 56, 59, 61–62, 107–108, 112–113, 131, 138, 172, 175, 185, 212, 241 verbalizer, 39, 217, 232 verbs, 10–11, 18, 20, 50, 55, 69, 83, 89, 91, 98–99, 101, 106–107, 111–113, 128, 174, 178, 185, 189, 209, 218, 221, 246–247 vernacular, 131 vigesimal, 13–14, 19, 22, 88 village, 56–57, 60, 63, 78, 121, 170, 173, 214–215, 225 vitality, 159, 215 vocabulary, 55, 59, 70, 85, 115, 129–130, 132, 175, 178–179, 194, 213, 215– 216, 229, 252, 264 vocalic, 79, 206 voice, 134, 156, 168–169, 189 voiced, 79, 151 voicing, 28–29, 79–80, 107 vowel, 23, 29–30, 33, 36, 45, 84–85, 91, 106–107, 130, 175–176, 182, 185, 205–206 women, 9, 17, 53, 79, 82, 97, 116, 151, 158, 183, 210, 216, 222–223, 225, 230, 240, 249, 257

Index of languages

Achugwa, 88 Amerindian, v–vi, 50–51, 65, 71–73, 90, 124–125, 143, 194, 214, 244 Antillean, 91 Austronesian, v–vi, 149, 168, 180, 183– 184, 204, 212, 215, 232, 237, 238– 241, 244–262, 264 Breton, 73 Candonga, 228 Caribbean, 77 Castellano, 44 Cebuano, vii, 204–217, 228, 231 Chabacano, vii, 203, 204, 205, 207, 208, 209, 214, 215, 216, 217, 218, 219, 220, 221, 223, 224, 229, 231, 232 Chamorro/Chamoru, vii, 167–170, 180– 194, 237–248, 251, 257–259, 261– 264 Chibchan, 77, 79, 90 Creole, 51, 91, 181, 187, 192, 194, 203, 208, 214–216, 220, 221, 232 Cuzco, 84, 87 Dutch, 149, 169 German, 152, 188, 224 Guano, 169 Hatun, 104 Hispanic, 39, 167, 204–205, 207, 209, 211–215, 231 Hueyapan, 41 Imbabura, 95–97, 104, 116, 121, 134, 143 Inca, 96 Indoeuropean, 180, 183

Italian, v Japanese, 228, 251 Magandang, 228–229 Malay, 221, 232 Mangareva, 149, 169, 172 Maori, 151, 156, 193 Mayan, vi Nahuatl, vi, 3, 6, 10–14, 16–23, 27–45, 126, 203, 214, 216 Otavalo, 96 Papiamentu, 219, 232 Pipil, 125 Portuguese, v, 203, 215, 217, 219, 221 Quechua, vi, 84, 87, 95–96, 98, 104–105, 111–112, 114–115, 121–122, 124, 127, 130–143, 194 Quichua, vii, 95–98, 104–106, 115–116 Quito, 116, 139 Rapanui, vii, 149–162, 167, 169–180, 182–183, 185, 189–194 Romance, v, 167, 187, 204, 232 Russian, 91 Tagalog, vii, 204 –206, 208 –215, 228 – 231, 251 Tahitian, 150–151, 153, 173 Tonga, 157 Udmurt, 91 Uralic, 91 Usa, 171, 237 Warfare, 169 Watami, 110 Watataka, 110 Zapoteco, 51, 63, 65–67, 71

Index of toponyms

Africa, v–vi America, vi–vii, 1, 77, 80, 87, 90, 116 Argentina, 115 Atlantic, 150, 214, 216, 221 Batavia, 220, 221 Bolivia, 115 California, 190 Chile, 152–153, 156, 159, 170, 172, 179 China, 203, 229, 232 Colombia, 77, 90 Ecuador, 95–97, 108, 121, 138, 143, 194, 226 Europe, v, 149 Filipino, 194, 203, 213–214, 226, 230 Germany, viii, 169, 237 Hispanoamerica, 149 Jakarta, 220 Japan, 169 Jerusalem, 186 Madrid, 229 Malacca, 203, 219, 220–221, 232 Malaysia, 215 Marianas, vii, 167, 169–170, 180, 182, 189, 192–194, 213, 231, 237, 243

Mesoamerica, 6 Mexico, 32, 34, 44, 49–50, 72, 194, 203, 216, 229, 244 Micronesia, 167, 186 Mindanao, 203, 213, 215, 228–230 Morelos, 41–42 Netherlands, viii New Zealand, 151, 156, 158 Oaxaca, 49–51, 65, 72 Oceania, 150, 167 Pacific, 149, 162, 167–168, 170, 186, 237, 265 Panama, vii, 77–78, 80 Peru, 91, 96, 115 Philippines, vii, 169, 186, 194, 203–207, 209, 213–215, 219, 221, 226, 228– 232 Polynesia, 158, 167 Salvador, 125 Samoa, 157 Santiago, 152, 158 Spain, 91, 169, 194, 232, 237 Tahiti, 157–158, 169, 172

Index of authors

Adelaar, Willem, 115 Albalá Hernández, Carmen Paloma, 180, 182, 194 Albalá, Paloma, 180, 182, 194, 214, 244, 263 Anderson, Stephen R., 23, 213 Andrews, Richard, 27, 44–45 Arias Díaz, Gervasio, 77, 83, 88 Backus, Ad, 116 Bakker, Dik, v–viii, 43, 95–96, 98, 101, 104, 111–112, 114, 125–127, 132–133, 135, 137–138, 140, 143, 263 Bakker, Peter, 62, 71–72, 181, 187, 194 Baptista, Priscilla M., 90 Barlow, Roberto, 35 Barón Castro, Radolfo, 213 Barthel, Thomas S., 172 Baxter, Alan N., 219, 221, 232 Beck, David, 104 Berengueras, José María, 77 Beverly, John, 116 Bosque, Ignacio, 136 Bowen, J. Donald, 210, 212 Boyd, Hedyx, 77 Branca, Gabriel B., 77, 90 Brody, Jill, 124, 143 Bunye, Maria Victoria R., 205–207, 211– 212, 214 Burrus, E. J., 180 Büttner, Thomas, 96–97, 115 Campbell, Lyle, 123–128, 143 Canger, Una, 27 Cicirelli, Victor G., 264 Clements, Joseph Clany, 143, 221 Clyne, Michael, 172 Cole, Peter, 134, 143 Constantino, Ernesto, 205–206 Cooreman, Ann M., 194

Couto, Hildo Honorio do, 181 Cruz, Víctor de la, 64 Cruz-Jiménez, Roberto, 63 Dakin, Karen, 31 Dalgado, Sebastiao Rodolpho, 219 DeBruyne, Jacques, 138 DeGraff, Michel, 126 Demonte, Violeta, 136 Díaz, Norma, 192 Dik, Simon C., vi–viii, 43, 132–133, 143, 263 Dixon, R.M.W., 163, 194 Dorian, Nancy, 43 Du Feu, Veronica, 149–150, 176–177 Erice, Jesús, 77–78 Evans, Nicholas, 45 Ferguson, Charles A., 43, 45 Fernández López, María del Carmen, 134 Field, Frederic, 116 Fill, Alwin, 192 Fischer, Steven Roger, vii, 149–150, 152, 155, 160, 162–163, 169–173, 175–177, 192 Flores Farfán, José Antonio, vi, 27, 31, 35, 37–38, 43–45, 72, 262 Forman, Michael L., 222–223 Fought, Carmen, 240 Frake, Charles O., 215 Garza Cuarón, Beatriz, 72 Garzon, Susan, 145 Gassó, Leonardo, 77 Giannelli, Luciano, vii, 77, 80, 83, 88, 90 Godínez, Esteban, 29, 34, 45 Gómez Macker, Luis A., 193 Gómez Rendón, Jorge, vi–vii, 95, 111, 115, 138 Green, Abadio, 77–78, 81, 90

282 Index of authors Greenberg, Joseph H., 127 Grice, Paul, 191 Grosboll, Sue, 96 Haboud, Marleen, 96 Hartau, Claudine, 24 Haspelmath, Martin, 191 Hawkins, John A., 127 Heath, Jeffrey, 126–128, 140 Heimann-Koenen, Marie-Luise Heine, Bernd, 128, 139, 187 Hekking, Ewald, vi, 72, 125–127, 132, 133, 137–138, 143 Hengeveld, Kees, 95, 98–99, 116, 133, 139, 143 Hernández Legorreta, Cutberto, 237 Hill, Jane H., 27, 33, 35, 38, 44 Hill, Kenneth C., 27, 33, 35, 38, 44 Himmelmann, Nikolaus P., 204, 211 Hinz, Eike, 9, 12–13 Holm, John A., 181 Holmer, Nils M., 77–78, 84 Hopper, Paul, 187 Huijbregts, Riny, 134 Jensen, Anne, vi, 3, 28, 38 Jijón y Caamaño, Jacinto, 96 Johanson, Lars, 194 Karttunen, Frances, 19–20, 27, 31, 33, 43–45 Kaufman, Terrence, 55, 111, 126, 171, 175, 177–179, 187–188, 192 Kitano, Hiroaki, 213 Klaus, Susanne, 23, 149 Klee, Carol A., 131, 143 Kloss, Heinz, 64 Koontz-Garboden, Andrew, 221 Labov, William, 27 Langacker, Ronald W., 22 LaPolla, Randy J., 138 Launey, Michel, 27 Le Page, Robert B., 180 Lefebvre, Claire, 181 Lightfoot, David, 125 Lipski, John M., 204, 215, 228–229

Llamas, Carmen, 264 Llerena Villalobos, Rito, 77 Lockhart, James, 15, 17, 19–20, 27, 31, 33 Lope Blanch, Juan, 72 López, Cecilio, 207–211 López, Felipe, 50 Ludwig, Ralph, 172, 180 Mackenzie, J. Lachlan, 133–135, 137, 140, 142–143 Makihara, Miki, 153, 155–158, 163, 169– 180, 189–194 Marotta, Giovanna, 77 Matras, Yaron, 163, 194 Maude, Harry Evans, 169 McWhorter, John H., 181 Milroy, James, 124 Mondada, Lorenza, 193 Montalván, Ana, 77 Montrul, Silvina A., 178 Moravcsik, Edith A., 126–127 Mous, Maarten, 181, 187, 194 Mufwene, Salikoko S., 181, 192–193 Mühlhäusler, Peter, 169, 192 Munró, Pamela, 50 Munteanu, Dan, 181, 219 Muysken, Pieter, 71–72, 108, 116, 121– 125, 127–128, 130–131, 133, 136–137, 141, 143, 194 Myers-Scotton, Carol, 174 Nava Nava, Refugio, 33, 35 Ocampo, Alicia M., 131, 143 Odo, Carol, 188 Oliver, Douglas L., 167 Onedera, Peter, 168, 189 Orán Bodin, Réuter, 77 Oroz, Rodolfo, 172 Pacini, Beatrice, 77–79 Pagel, Steve, vii, 167, 183, 185, 263 Palemón Arcos, Francisco, 30 Pangilinan, Michael R.M., 213 Parker, Gary, 115 Pavón Lucero, María Victoria, 138 Peralta Ramírez, Valentín, 39, 45

Index of authors 283 Pérez Tejedor, Juana Pabla, 219 Pfänder, Stefan, 185, 191 Pigafetta, Antonio, 168 Prince, J. Dyneley, 77, 88 Puig, Manuel María, 78, 86 Quesada, J. Diego, 90 Quilis, Antonio, 204, 213, 230 Rabanales, Ambrosio, 172 Riego de Dios, Maria, 215 Riley, George A., 264 Rodríguez-Ponga, Rafael, 180–182, 184– 185, 237, 263–264 Rogers, Robert F., 168–169, 180, 188, 193 Rojas, Doris, 77 Rubino, Carl, 206, 232 Sabater Fuentes, Anna, 184 Sakel, Jeanette, 163 Salas Palomo, Rosa, vii–viii, 183, 194, 237, 241, 263 Salomon, Frank, 96 San Miguel, Rachel, 33, 35, 214 Sankoff, David, 174 Santos Ferreira, José dos, 219 Schachter, Paul, 104, 213 Schrader-Kniffki, Martina, vi, 49–50, 69, 71–73 Schuchardt, Hugo, 191, 226, 228–229, 232 Sherzer, Joel, 77, 79, 90–91 Siewierska, Anna, 135, 140, 143 Simpson, Andrew, 88 Smith, Geoff P., 186 Smith, Ian, 143 Smith, Lino, 77 Solenberger, Robert R., 263 Squillacciotti, Massimo, 77, 91 Steinkrüger, Patrick O., vii, 203, 216, 217, 231 Stockwell, Peter, 264 Stolz, Christel, 71, 73, 124–125, 128, 138–139, 176, 185, 212, 265

Stolz, Thomas, v–viii, 71, 73, 124–125, 128, 138–139, 176, 181, 183–186, 194, 212, 237, 263, 265 Suárez, Jorge A., 27 Tabouret-Keller, Andrée, 180 Thiesen de Weber, Nancy, 170, 193 Thomason, Sarah, 21, 51, 55, 72, 111, 115, 121, 123–137, 141–143, 171, 175, 177–179, 187–188, 192 Thompson, Roger M., 204, 214 Thornton, Agathe, 193 Topping, Donald M., 182, 185–186, 192, 264 Torero, Alfredo, 96, 115 Traugott, Elizabeth, 187 Tsunoda, Tasaku, 127, 133, 137, 139, 143, 264 Underwood, Robert, 189, 237 Ungeheuer, Gerold, 264 van Coetsem, Frans, 143 van der Sijs, Nicoline, 65, 73 van Riemsdijk, Henk, 134 Van Valin, Robert D., 138 Veyra, Jaime C. de, 231 Wackernagel, 213 Wagua, Aiban, 77–78, 81, 90 Wallin, Ruth, 90 Weber, Robert, 170, 193 Weinreich, Uriel, 126, 128 Weitlaner, Roberto, 35 Whinnom, Keith, 228–229 Whorf, Benjamin Lee, 38, 41 Wickberg, Edgar, 230 Winford, Donald, 174 Winkler, Eberhard, 91 Wolfenden, Elmer Paul, 206 Wolff, John U., 210, 212–213 Yap, Elsa Paula, 205–207, 211–212, 214 Zamponi, Raoul, vii, 77, 83, 88, 90 Zimmermann, Klaus, vi, 51–52, 71–73, 149–150, 154, 158–159, 161–163 Zorc, R. David, 214

Contributors

Dik Bakker Department of General Linguistics Universities of Amsterdam & Lancaster Spuistraat 210 NL-1012 VT Amsterdam The Netherlands

Luciano Giannelli Centro Interdipartimentale di Studi Cognitivi sul Linguaggio (CISCL) Complesso S. Niccolò Via Roma, 56 I-53100 Siena, Italy

[email protected]

[email protected]

Suzanne Dikker Department of Linguistics

Jorge Goméz Rendón Department of Theoretical Linguistics

New York University 719 Broadway, 4th Floor New York NY 10003, USA

University of Amsterdam Spuistraat 210 1012 VT Amsterdam The Netherlands

[email protected]

[email protected]

Steven Roger Fischer

Anne Jensen

Institute of Polynesian Language and Literature PO Box 6965, Wellesley St. Auckland, New Zealand

University of Copenhagen Nørregade 10 DK-1017 Copenhagen Denmark

[email protected]

[email protected]

José Antonio Flores Farfán CIESAS

Steve Pagel Institut für Romanistik

Juárez 87, Tlalpan 14000 México D.F., Mexico

Martin-Luther-Universität HalleWittenberg 06099 Halle/Saale, Germany

[email protected]

[email protected]

286 Contributors Rosa Salas Palomo Micronesian Language Institute

Thomas Stolz Fachbereich 10

University of Guam, House #4 UOG Station Mangilao, Guam, USA 96923

Universität Bremen Postfach 330440 28334 Bremen, Germany

[email protected]

[email protected]

Martina Schrader-Kniffki Fachbereich 10

Raoul Zamponi Facoltà di Lettere e Filosofia

Universität Bremen Postfach 330440 28334 Bremen, Germany

Via Roma 47 I-53100 Siena Italy

[email protected]

[email protected]

Patrick O. Steinkrüger Zentrum für Allgemeine Sprachwissenschaft, Typologie und Universalienforschung (ZAS) Schützenstr.18 10117 Berlin, Germany [email protected]