Chabacano/Spanish and The Philippine Linguistic Identity

by John M. Lipski, The Pennsylvania State University ( presented with permission by author )

click here for Penn State's home page

Introduction

In the linguistic profile of the Philippines, one of the most elusive elements to categorize and acknowledge the Spanish contribution. The amount of Spanish lexical incursions into the major Philippine language families is beyond dispute; less well understood is the extent to which the Spanish language is actively used and understood throughout the nation. Finally, and most germane to the topic of this symposium, is the inclusion, classification, and appreciation of a group of languages known to linguists as Philippine Creole Spanish, and to the speech communties themselves as Chabacano. Like many other creole languages, the name Chabacano itself stems from a derisive Spanish term meaning clumsy, ill-formed, and vulgar. Although within the PCS/Chabacano-speaking communities this word has lost the negative connotation and refers only to the language itself, many enlightened community members avoid the term Chabacano and prefer instead the regional designations of Caviteño, Ternateño, and especially Zamboangueño. Throughout the history of the Philippines Chabacano in its various manifestations has remained at the margins of the country's linguistic repertoire, ignored by many, repudiated by those who aware of its existence (this even includes many speech community members), ambiguously classified by Philippine language typologists, and truly appreciated only by the elite subset of professional linguists who specialize in creole languages.  Even among this group the creole status of PCS is not undisputed, all of which combines to make PCS/Chabacano a linguistic orphan surrounded by patrimonial languages and the unchallenged carryovers of colonial times.1

 

The reasons for the low profile of Chabacano in the Philippine linguistic consciousness are many, the most salient of which include:

 

(1) General unawareness of a language which nowadays is spoken extensively only in a geographically remote portion of the country, Zamboanga City and surrounding parts of Zamboanga del Sur province, as well as in other pockets in Mindanao and the islands of Basilan and Jolo. Philippine Creole Spanish dialects were once spoken more extensively in the Ermita district of Manila as well as in other enclaves on Manila Bay (Cavite and Ternate), but most non-speakers of Manila Bay Chabacano who were even aware that some Spanish-derived language was spoken in their midst assumed that this was some form of Spanish, possibly a broken second-language variety or pidgin.

 

(2) Among those Filipinos with second-hand knowledge of Chabacano, the notion that this is in reality a dialect of Spanish, a colonial language increasingly marginalized in Philippine society. In the contemporary Philippines, fluency in Spanish is generally restricted to a small and aging elite of mixed Philippine-Spanish heritage; the typical fluent Spanish speaker has at least one parent or grandparent born in Spain, and belongs to wealthy landowning or empresarial classes far-removed from the grass-roots level at which Chabacano is spoken. In Zamboanga, recent reintroductions of Spanish items (Lipski 1986a, 1987f) contribute to the mistaken notion that Zamboangueño Chabacano has been Spanish `all along,’ with only occasional deviations from standard usage.

 

(3) The mistaken notion among creolists (beginning with Whinnom 1956) that the largest Chabacano-speaking population, that of Zamboanga City, is small and moribund, when in fact it is a thriving first- and second-language speech community of perhaps half a million speakers.  Frake (1971) was the first to provide more accurate information on Zamboangueño, but to this day many scholars in the Philippines and abroad are unaware of the true strength of the Zamboanga Chabacano community.

 

(4) The fact that in the majority of surveys of Philippine languages (especially those produced within the Philippines, e.g. Llamzon 1978), none of the Chabacano varieties appears (at times `Spanish’ is listed), thus implicitly suggesting that Chabacano is not a `Philippine' language.

 

(5) The historical confusion, found in literature, travelers’ accounts, and official documents, between Philippine varieties of Spanish, rudimentary Spanish-based pidgins and trade languages, and Spanish-derived creoles spoken natively by Filipinos.

 

(6) The small number of native Spanish speakers in the Philippines has contributed to the lack of studies of contemporary Philippine Spanish; the majority of works which lay claim to such a description in reality deal with Hispanic lexical items in native Philippine languages, or with some aspect of the PCS dialects. At times, the latter dialects are mistakenly referred to as "Philippine Spanish," as though there were no legitimate non-creolized variant of metropolitan Spanish currently available in the Philippines. One example of this confusion is the statement (Diez, Morales, Sabin 1977:85) that `En la actualidad la situación del español es bastante precaria ... el dialecto español que se habla en aquellas islas recibe el nombre de chabacano’ [currently, Spanish is in a precarious situation in the Philippines ... the Spanish dialect spoken in that country is known as Chabacano]. Quilis (1975:34) speaks of ...el español como dialecto conservado en Cavite y Zamboanga ... este dialecto es el que se conoce con el nombre de chabacano. Su estructura es bastante peculiar: es un español con los recursos gramaticales del tagalo y del cebuano ... [Spanish as a dialect conserved in Cavite and Zamboanga ... this is the dialect known as Chabacano. It has a peculiar structure: it is Spanish with the grammatical resources of Tagalog and Cebuano]. In later writings Quilis clearly recognized Chabacano as a creole language and not simply a `mixed dialect' of Spanish. A more serious assessment (Whinnom 1956:2) states that `the modern Spanish of Manila has none of the characteristics of the South American or Andalusian Spanish ... the Philippine Spanish of today is the result of the second stage of the Spanish contact with the Philippines.’ National census data representing both the American administration and the Philippine national government provide confusing information, since PCS or Chabacano dialects are lumped together with modern Spanish, while there is a tendency to overlook potential Spanish speakers who have no formal training in that language. As a timely note, the web page provided by the Shangri-La hotels in the Philippines gives a description of Zamboanga which states `The dialect spoken there is a corruption of the Spanish language.’

 

1. The Chabacano presence in the Philippines

 

The Philippines is the only former Spanish (colony) where the Spanish language was never acquired by the majority of the native population, and which replaced no native language.  Among the other former colonies, only Equatorial Guinea shows a similar profile, but this colony was not effectively occupied by Spain until the 1860's (Fernando Poo) or the first decades of the 20th century (Río Muni), with the total period under Spanish colonial rule being no more than 60-100 years. Even so, the majority of the population speaks Spanish, albeit as a second language with varying levels of ability (Lipski 1985). In the Philippines, a very small and rapidly dwindling population—mostly recent descendents of Spaniards—speaks Spanish fluently; this group has little or no linguistic impact on the rest of the national population. In addition, several Spanish-derived creoles continue to be spoken natively; the largest community is found in Zamboanga City, with expatriate Zamboangueño enclaves in Cotabato, Jolo, and Davao. Along Manila Bay, vestigial speakers of Cavite and Ternate Chabacano are still to be found, while the once thriving Ermita variety of Manila has now disappeared.  Collectively, all varieties of Philippine Creole Spanish (PCS) are known as Chabacano, and three distinct dialects still exist, in Ternate, Cavite and Zamboanga. Whinnom postulated that Ternateño (T) was the first to be formed, suggesting that this creole has extraterritorial roots, descending from a largely Portuguese-based creole formed in the 17th century on the Indonesian island of the same name. Comparative work by Molony (1973, 1977a, 1977b) supports the hypothesis that T is the oldest of the Spanish-based creoles in the Philippines while observations by Batalha (1960) suggest more than casual parallels with some Asian Portuguese-based creoles.  Caviteño (C) was a later offshoot of T, one of many stable or fleeting Spanish-based contact vernaculars that arose in the fortified areas around Manila Bay, and it too can be traced back at least as far as the 18th century. T and C exhibit several grammatical features that make them contenders for being descendents or relexifications of earlier Portuguese-based creoles and pidgins (but cf. Lipski 1988). Both creoles have been influenced by Philippine languages, particularly Tagalog; this is evident not only in the preferred VSO word order (at least with pronominal subjects), but also in the incorporation of Tagalog particles and other syntactic structures. Varieties of PCS are also spoken on the island of Mindanao. The largest group of speakers is found in and around Zamboanga City. A small group, now largely dispersed and speaking central Philippine languages, previously existed in Davao, derived from immigrants from Zamboanga who arrived at the turn of the 20th century. Another small group is found in Cotabato City (Riego de Dios 1976a, 1978). Cotabatateño (Ct) is virtually identical to Zamboangueño (Z), with the few differences being mostly lexical. Riego de Dios suggests that the two dialects may have partially different roots, although admitting immigration from Zamboanga to Cotabato as the likely source of most of Ct.

 

Although unknown to or ignored by most Filipinos, the Chabacano dialects are of importance to creolists. Philippine Creole Spanish is the only Spanish-based creole in Asia, fitting in with the Portuguese-based creoles in India, Sri Lanka, Malacca and Macau; indeed, some (beginning with Whinnom 1956; cf. Lipski 1988 for differing views) assert that Philippine Creole Spanish is but a relexification of a pan-Asian Portuguese-based creole. All other Spanish-based creoles (Papiamentu, Afro-Colombian Palenquero, and vestigial enclaves found in Latin America) result from Afro-Hispanic language contacts, although some Hispano-Amerindian creoles may have existed in small numbers, and the possibility for a Hispano-Arabic creole as part of Mozarabic language cannot be totally discounted. Philippine creole Spanish is also important for theories of creole language typology. It breaks from the usual SVO patterns, in exhibiting a prototypically Austronesian VSO, albeit with many alternative possibilities. The verbal syntax is both tantalizingly similar and vastly different from other Iberian-based creoles, and many other unique features accrue to PCS. In order to assess the situation of Chabacano in the pantheon of Philippine languages, it is necessary to survey the nature and distribution of non-creole Spanish in the Philippines, since the most frequent misidentification responsible for the marginality of Chabacano awareness is the equation Chabacano = Spanish in the Philippines.

 

2. The Spanish language in the Philippines

 

The failure of the Spanish language to establish itself in the Philippines has been the subject of much prior commentary; suffice it to say that this linguistic situation stems from a combination of factors, among which are: the Spanish government's official and non-official policy of using the vernacular languages, particularly in religious functions; the relatively small number of Spanish natives in comparison with the indigeneous Philippine population; the lack of significant demographic shifts among native groups in the Philippines which would have precipitated the necessary use of Spanish as a lingua franca. With the exception of the Chabacano dialects, which arose around Spanish military garrisons and spread in multilingual commercial centers, Spanish never became the native language of any large sector of the native-born Filipino population, nor even became a widely used lingua franca outside of those (mestizo) groups most closely aligned with the colonial administration. With the coming of the American administration and the rapid and effective implementation of educational programs in English, Spanish was pushed ever further into the background, and its status as an obligatory part of the school curriculum is currently being called into question, as an apparent anachronism.

 

Currently, the majority of Spanish-speaking Filipinos belong to mestizo (Eurasian) families, directly descended from Spanish settlers. Moreover, this Spanish parentage is usually quite recent, in that nearly all Spanish speakers have at least one grandparent who was born in Spain; few Spanish speakers are found who cannot claim a Spanish-born relative at least two generations in the past. This Spanish-speaking nucleus is strengthened by intermarriage, since most Spanish speakers have married other Spanish speakers or have otherwise reinforced their Spanish language environment, including membership in clubs or in the Casino Español (in Manila and Cebu), trips to Spain, and choice of residential area.

 

It is also possible to find non-mestizo Filipinos who for one reason or another learned Spanish through contact with previous generations of Spanish speakers, but the number of such individuals is small in comparison with the totality of Philippine Spanish speakers. Spanish is still a subject in the university curriculum (despite current pressure to remove the requirement), and formerly Spanish was widely taught in the public schools. Although the majority of Filipinos who have studied Spanish under such circumstances have very little useful language ability, many individuals have a degree of passive competence which allows them to grasp the general meaning of Spanish phrases and expressions. Naturally, the high proportion of Hispanisms in the native Philippine languages aids in the recognition of current Spanish forms, and older Filipinos may recall the presence of Spanish priests, nuns and lay teachers, particularly in private schools, all of whom helped spread an awareness of the Spanish language. Lawyers in the Philippines have often studied Spanish more carefully, since much of the legal code was written in Spanish, and until relatively recently it was possible to use the Spanish language in the courtroom. Many Filipina nuns studied in convents directed by Spanish priests and nuns, where Spanish was the language of daily communication, and even today they may recall some aspects of that language. It thus becomes clear that, while the majority of true Spanish speakers come from mestizo families, there is an undetermined but not inconsiderable number of Filipinos with some knowledge of Spanish, below the level of native speakers but superior to that of foreign students.

 

A concomitant feature of most Philippine Spanish speakers is their socioeconomic level, which is usually toward the top of the scale. Spanish speakers are frequently members of Spanish landowning and commercial families, which have managed to retain and even expand their fortunes throughout the various post-colonial administrations in the Philippines. Naturally, not all such families have retained their wealth and social position, and there are other Spanish-speaking families which clearly belong to the middle classes, but among the wealthier Spanish speakers, use of the language is regarded as a source of pride and an unmistakable mark of aristocratic authenticity. These Spanish speakers continue to use the language at home, although it is difficult to use Spanish in public, due to general lack of interlocutors and a certain resentment among other Filipinos. Despite efforts of Spanish speakers to teach the language to their children, few true Spanish speakers under the age of about 40 are to be found, and it is unlikely that the language will survive another generation. Most Philippine Spanish speakers are also proficient in English, but few hold native Philippine languages in high esteem, often regarding with resentment and scorn the establishment of Tagalog (Pilipino) as a national language. As a result of these attitudes and behavior patterns, Philippine Spanish is characteristically refined, aristocratic, precise, and linguistically conservative, with none of the popular, regional and rural forms which are essential ingredients of the PCS dialects, and which are widespread in the Spanish dialects of Latin America. Also of note are the distinctly Castilian (i.e. central and northern Spain) traits of contemporary Philippine Spanish, where virtually no hint of Andalusian, Galician, Canary Island, Catalan, Valencian, or other regional features of vocabulary or pronunciation are found, despite the fact that many of the last wave of Spanish immigrants to the Philippines came from those regions. The highly precise and Castilianized Philippine Spanish reflects the influence of Spanish teachers, administrative personnel and religious figures, as well as literary and journalistic standards which were in wide usage until well after World War II, among the numerous newspapers, magazines and other documents published in Spanish.

 

Currently, the largest number of Spanish-speaking Filipinos is found in metropolitan Manila, although significant smaller groups are located in many provincial capitals, particularly in those regions characterized by large plantations and estates which have existed since the Spanish period. Among the latter zones are the sugar-producing regions of Negros (particularly in Bacolod but also around Dumaguete) and the fruit-producing regions of Mindanao, especially around Cagayan de Oro and Davao. Other nuclei of Spanish speakers are found in the Bikol area (Legaspi City and Naga), Iloilo, Tacloban, Cotabato, Vigan, Cebu and Zamboanga, being in the latter case bilectal Spanish-Chabacano speakers. Although the totality of the regions mentioned above represents a wide selection of regional languages, including Tagalog, Ilocano, Hiligaynon, Cebuano/Visayan, Waray, etc., there has been virtually no regionalized influence of these languages on Philippine Spanish, in that it is in general impossible to distinguish the geographical origin of a Spanish-speaking Filipino through features of spoken Spanish (unlike the case with spoken English).

 

The specific linguistic features of Philippine Spanish are of interest to the present enterprise only to the extent that they intersect with Chabacano and reflect the integration of both languages into a pan-Philippine linguistic matrix (Lipski 1987c, 1987d, 1987e). Phonetically, Philippine Spanish differs from other natively-spoken varieties in maintaining a uniformly occlusive pronunciation of intervocalic /b/, /d/, and /g/, which are fricatives in other Spanish dialects. This is especially noticeable in the case of intervocalic /d/, which may even overlap with the [r] articulation of /r/ in the Philippines, particularly when /d/ is given an alveolar articulation instead of the more universal dental pronunciation. Few current Philippine Spanish speakers utilize the fricative pronunciation of intervocalic and word-final /d/; those that do usually have at least one parent born in Spain, or have spent considerable time in that country.  This same trait has been carried over to the PCS dialects and to Hispanisms borrowed into native Philippine languages; an identical pronunciation is found among Spanish speakers in Equatorial Guinea, the only Spanish-speaking region of sub-Saharan Africa, and among bilingual indigeneous-Spanish speakers in many regions of Latin America. As an alternative to an occlusive articulatioin, intervocalic /d/ is frequently elided, particularly in the verbal affix -ado, used to form past participles, following the patterns current in peninsular Spanish dialects.  Unlike the dialects of Spain, and like those of Africa and bilingual areas of Latin America, Philippine Spanish intervocalic /d/ never passes through the fricative stage en route to deletion; the loss of /d/ is an imitation of an originally phonetically-motivated process, but there is no active reduction of /d/ among new words introduced into Philippine Spanish, or in other intervocalic contexts which in other Spanish dialects are also being gradually affected by the reduction process.

 

In other respects, Philippine Spanish retains a strongly resistant syllable-final /s/, unlike the Andalusian dialects that apparently provided the input to Manila Bay Chabacano dialects and most probably Zamboangueño as well (Lipski 1986b, 1987a). Word-final /n/ receives a uniformly alveolar articulation [n] in Philippine Spanish, despite the high frequency of word-final velar [˜ ] among the native Philippine languages, and the strong tendency to velarize word-final /n/ in the Spanish dialects of Andalusia, Galicia and other areas of Spain whence came many immigrants. Many Philippine Spanish speakers use the palatal lateral phoneme /ë/, characteristic of an ever smaller group of Peninsular Spanish dialects, but once prominent among varieties of Spanish taken to the Philippines. At times, /ë/ is realized as [ly] among the last generation of Philippine Spanish speakers, but merger with /y/, as has occurred in most dialects of Spain and Latin America, is extremely rare, and occurs only as an idiosyncratic trait. Unlike any dialect of Latin American Spanish or of PCS, contemporary Philippine Spanish exhibits the voiceless dental fricative phoneme /è/ (written z or, before e and i, as c), used in accordance with Spanish etymology and the norms of contemporary Castilian speech, although occasional discrepancies are observable.

 

Another outstanding characteristic of Philippine Spanish is the frequency of the glottal stop [q] at the beginning of words which nominally begin with a vowel: el hombre [el-qom-bre] `the man.' This is contrary to the normal Spanish phonotactic linking of word-final consonants to syllable-initial position if the following word begins with a vowel; in other Spanish dialects, the corresponding pronunciation would be [e-lom-bre]. The glottal occlusion [q] is also heard in some hiatus combinations, such as maíz [ma-qis] `corn,' the same pronunciation as is used among the native Philippine languages. Due to the extraordinary use of [q], the normal consonantal linking typical of Spanish phonetics does not as frequently occur in Philippine Spanish, with the result that phonetic boundaries between words are clearly perceivable in the spoken chain. This feature, evidently the result of influence from native Philippine languages, is found in the speech of nearly all contemporary Philippine Spanish speakers, regardless of their claimed or actual proficiency in Philippine languages; it is possible, however, that earlier generations of Spanish speakers, many of whom were nearly monolingual, may not have exhibited this trait, which is not found in any peninsular Spanish dialect.  In the morphosyntactic domain, fluent Philippine Spanish exhibits no deviations from other natively-spoken dialects. The second person familiar pronoun vosotros is used as in much of Spain, and le/les are used as direct object clitics, again following Peninsular usage. Lexically, there are few peculiarities. The number of indigeneous borrowings into Philippine Spanish is quite low, and is limited almost entirely to the flora and fauna which have no equivalent expression in Spanish, and to the formation of nicknames via the suffix -ng: Pedring (Pedro), Doming (Dominador), Carling (Carlos), Puring (Purificación), etc. In a few cases, a Tagalog root is combined with a Spanish morphological suffix, as in babaero/babayero `woman-chaser,' from Tagalog babae `woman.' Among the last generation of Philippine Spanish speakers, the use of the Philippine oo [oqo] instead of or in addition to the Spanish affirmative particle sí is relatively frequent, particularly in unguarded moments of reflection; presumably this did not occur among earlier generations of essentially monolingual Spanish speakers. Among the strictly Spanish elements in Philippine Spanish, there are a number of Americanisms, most of which are clearly Mexican in origin. These evidently date from earlier days of Hispano-Philippine contacts, when the Manila Galleon departed from the port of Acapulco, since recent linguistic contacts with Mexico have been almost nonexistent in the Philippines. Among the most prominent Mexicanisms still in use among current Philippine Spanish speakers are: zacate `grass,' petate `sleeping mat,' changue [tiangue] `market,' chili `pepper,' camote `sweet potato,' chongo [chango] `monkey,' palenque `market,' sayote [chayote] `a type of vegetable.' In order to ask for something not quite heard to be repeated, use of ¿mande? is the rule in Philippine Spanish, as in Mexico, and the three daily meals are el almuerzo `breakfast,' la comida `lunch' and `dinner,' following rural Mexican usage, and contrasting with more general Spanish el desayuno, el almuerzo, and la cena, respectively. Chabacano speakers also use Mexican curses and obscenities, and such words are not unknown in contemporary Philippine Spanish, although peninsular Spanish curses are more frequent. Curiously, despite the decidedly aristocratic character of modern Philippine Spanish, there is comparatively less reluctance to use these forms in mixed company or among women, perhaps reflecting lack of contact with contemporary sociolinguistic norms of Spanish-speaking nations. Other common Spanish words have undergone semantic shifts in Philippine Spanish. The word lenguaje has shifted from `style of speech' to `national language'; también no is used instead of tampoco in the sense of `not either,' possibly reflecting old Spanish usage; the expression hay que ver `it must be seen' is the most frequently used to express surprise or admiration. Also extremely frequent is the idiomatic expression la mar de `a lot of,' now outmoded in Spain, and the use of gracia for `given name' (as in ¿cuál es su gracia? `what is your name?') now typical only of some marginal areas of the Spanish-speaking world. Seguro, meaning `certain, sure' in standard Spanish, means `probably, maybe' in Chabacano and in contemporary Philippine Spanish; `sure' is rendered by (a)segurao `assured.' The most striking lexical innovation in Philippine Spanish is the conjugation of the word cuidado (pronounced cuidao) `caution, concern' in combination with subject pronouns; the derived meaning is roughly `whatever ... want(s)' or `... will take charge of it.' Combinations include tú cuidao, usted cuidao, ustedes cuidao, `it's up to you'; yo cuidao `I'll take care of it,' etc. This expression reflects the syntax of Tagalog bahala in combinations like ako ang bahala/bahala ko, corresponding to yo cuidao, ikaw ang bahala/bahala ka, equivalent to tú cuidao, etc., and represents the only widespread case of syntactic transfer from Philippine languages to non-creole Philippine Spanish

 

3. Early attestations of Chabacano and Spanish-based pidgins in the Philippines

 

Frake (1971: 223-224) is unequivocal in his classification of Chabacano, including Zamboangueño (the main focus of his article): Philippine Creole Spanish is not simply a Philippine language with unusually heavy Spanish lexical influence, nor is it Spanish with a large number of Philippine loan words. It is a distinct language, easily distinguishable from both its Romance and its Austronesian progenitors ... Philippine Creole Spanish shares enough in common with the classic creoles of the Caribbean that no one ... would, I think, challenge its assignment to the category `creole language.'

 

Despite Frake's impressive range of data and analysis, some of these conclusions are subject to reinterpretation. In particular, the notion that Chabacano is, or has always been, `easily distinguishable' from its Philippine language neighbors is not supported by the full range of available evidence. Moreover, although Z is definitively a creole, as are the remaining Chabacano dialects, its inevitable and undiluted origin in the Manila Bay Chabacano dialects is not a foregone conclusion.

 

One of the difficulties in tracing the presence and development of Chabacano in the Philippines is the common confusion of a coherent creole language with `broken Spanish' or even fluent Spanish. This is true not only for the greater Manila area, but also for the developing Spanish-derived creole of Zamboanga. Early visits to Zamboanga, in 1772 (Sonnerat 1776: 127) and in 1774-6 (Forrest 1780: 374-5) speak only of `Spaniards' (in reference to the garrison troops) and of escaped slaves from Jolo (largely of Visayan origin), without noting any special contact language that might have been in use. Even the language spoken by the `Spanish' troops must be suspect; Sonnerat (1776: 128-9) observed that the fort was guarded by `des gens bannis des Etats Espagnols, aussi prêts sans doute à le livrer qu'a le défendre' [men expelled from Spanish colonies, doubtlessly as willing to surrender the fort as to defend it].  Martínez de Zúñiga (1973: 236-7), describing the situation in the Philippines at the turn of the 19th century, noted that few Filipinos spoke Spanish. The exception was in the San Roque barrio of Cavite, where `they speak a kind of Spanish which has been corrupted and whose phraseology is entirely taken from the dialect of the country' (p. 250). The reference is clearly to Caviteño Chabacano, so that the author should have recognized Chabacano had he found it elsewhere. Upon describing Zamboanga, however, Martínez de Zúñiga only mentioned the Spanish garrison, `5,162 souls composed of natives, Spaniards, soldiers and prisoners,' with no indication that anything other than (non-creole) Spanish or Philippine languages were spoken there.

 

Some Spanish was apparently spoken in the Sulu Sea early in the 19th century. Moor (1837: 37) mentions Moslem Datus on Jolo who spoke Spanish, a fact also noticed by Yvan (1855: 230), and Saleeby (1980: 164-5). However, visitors to Zamboanga during the same period still note only Spanish (spoken by Spanish troops) or else `Moro' as spoken by Muslims from Jolo. This includes Keppel (1853: 70f.), St. John (1853: 131-2), Marryat (1848), Mallat (1846) and many others. In fact the general lack of knowledge of Spanish among Filipinos was frequently commented on by visitors to the islands. Bowring (1859: 28) speaking of the Manila working class, estimated that not one in a hundred spoke or understood Spanish. Lannoy (1849: 33) observed that indigenous political leaders were required to speak Spanish, but that this requirement was not enforced. Of Zamboanga, Lannoy noted that the garrison had roughly 380 men, of which were 11 were officers, 6 were sub-officers, and 24 were corporals. Of the linguistic and cultural problems, he noted (pp. 71-2) that `près de la moitié des officiers subalternes dans les régiments sont des indigènes, parlant la langue du soldat et jalousant les officiers espagnols, que parviennent seuls aux grades supérieurs. C'est là une cause constante de mésintelligence et d'irritation ...' [more than half the sub-officers are natives, speaking the soldiers' language and resenting the Spanish officers, who exclusively hold the higher ranks.  This is a constant source of misunderstanding and irritation ...]. MacMicking (1967: 92), writing in the 1850's, commented that most Filipinos could not speak Spanish, although `most of those in the neighborhood of Manilla can speak it after a fashion.' Jagor (1875: 156) stated that most soldiers spoke no Spanish.

 

Also instructive of the existence of Chabacano dialects in Zamboanga and elsewhere, and of the awareness of such varieties by outsiders, are observers' lists of languages spoken in each area of the Philippines. Jagor (1875: 55-6) assigned Spanish and Tagalog in that order to Cavite; Tagalog, Spanish, and Chinese to Manila; Spanish and Manobo to Cotabato; and `Mandaya' and Spanish to Zamboanga. Escosura and Cañamaque (1882: xxiii), writing in the 1860's, assign Spanish and Tagalog to Cavite; Tagalog, Spanish, and Chinese to Manila; Spanish and `Moro' to Basilan; and only Spanish to Zamboanga. This would indicate that Zamboanga was the most Spanish-speaking area of the Philippines in the mid 19th century. However, the same authors (p. 5) lament that native Filipinos speak only español de cocina, so that the designation `Spanish' assigned to Cavite, Manila and Zamboanga could well represent a Spanish pidgin, if not PCS. The information sifted and analyzed by Schuchardt (1883) would suggest that `Malayo-Spanish' was more typical of Manila and Cavite, and that Zamboanga might actually be Spanish-speaking.

 

The paucity of documentation on the language(s) spoken in colonial Zamboanga and the ambiguity of the existent attestations is surprising in view of the strategic importance of this port, the southernmost city in Spanish-controlled Philippines. Zamboanga was a way-station for travellers from every direction, and was constantly visited by Spaniards and foreigners alike. For the Spanish government, Zamboanga continued to be an important military defense against constant raiding by pirates and slavers from Jolo and other Moslem territories, and although the commercial importance of the city declined, overshadowed by growing urban areas such as Cotabato, Spain continued to maintain contact with Zamboanga until the end of the colonial period. It seems logical to surmise that had a Spanish-based creole significantly different from received Spanish been spoken in this important garrison town, it would not have escaped the scrutiny of Spanish authorities in Zamboanga. Alternative explanations must be sought, which take into account the constant Spanish and foreign observation of language and culture in Zamboanga, from the latter decades of the 18th century until the present time.  It is conceivable that Spaniards and other visitors regarded an evolving Z dialect of PCS as a broken language, the español de cocina or `kitchen Spanish' that was used to designate Spanish-based pidgins in Manila. This hypothesis seems unlikely, since extrapolation backward from the earliest attestations of Z (last decade of the 19th century) suggests that Z had attained its contemporary grammatical structures at least by the middle of the 19th century. By this time, explicit descriptions of what was known as español de cocina were widely available in 19th century Philippine Spanish literature, and this language is still recalled by some of the oldest residents of Manila. The latter language, however, was a rough pidgin, usually spoken by Chinese or other `foreigners,' and contained few if any of the consistent grammatical structures which characterize PCS: detailed TMA particle system, fixed syntax, hybrid Spanish-Philippine pronominal system, etc. `Kitchen Spanish' as spoken by Chinese sanglays as they were known in the Philippines is typified by the following example (López 1893: 58): `sigulo, señolía ... como no tiene ahola talabajo; como no tiene capé, y ha de ganalo la vida, sigulo tiene que hace tabaco' [of course, sir; since {I} do not have a job now, and since {I} don't have any coffee, and {I} have to earn a living, of course {I} have to make cigars]. Another example (Montero y Vidal 1876: 241) is `Mia quiele platicalo' [I want to speak with you]. Feced (1888: 77) gives examples like `guerra, señolía, malo negocio ... mía aquí vendelo, ganalo' [war is bad business, sir; I am here selling and earning {money}]. Moya y Jiménez (1883: 334) gives `mueno dia señolía ... ¿cosa quiele? mia tiene nuevo patila ...' [good day, Sir, what do you want? I have new merchandise]. Mallat (1846: 352) gives examples like `si que le compela cosa, cosa siñolita' [yes, buy many things, miss], and Saenz de Urraca (1889:142) gives todo balato, balato [everything {is} cheap]. Rizal (1891:121-2) has `siño Simoun, mia pelilo, mia luinalo’ [Mr. Simon, I lost it, I ruined it]; `Cosa? No tiene biligüensa, mas que mia chino mia siempele genti.  Ah, sigulo no siñola bilalelo …’ [what? Have you no shame; although I’m Chinese, I’m still a person. Surely {she} is not a true lady]; Mía cobalalo? Ah, sigulo suyo no sabe. Cuando pelilo ne juego nunca pagalo. Mueno suya tiene consu, puele obligá, mia no tiene’ [Me collect {the debt}? Oh, of course you don’t know. When {someone} loses in gambling, they never pay.  You have a consulate, you can oblige {them to pay}; I don’t have any]. In addition to the lack of archetypal Chabacano syntactic patterns, these examples illustrate at least three features which were never documented for Chabacano, but which do occur in other Spanish pidgins (as well as in Chinese pidgin English): use of mi/mia as subject pronoun, lateralization of intervocalic /r/, and pleonastic clitics, as in platicalo.

 

Authentic `kitchen Spanish' was used only between native Filipinos and Chinese merchants, or between these groups and Spaniards, much as the `bamboo Spanish' of Mindanao came to be used among and with Japanese arrivals in the early 20th century. It was never used natively, and was never used mutually by Philippine residents who spoke a common native language. In particular, the term `kitchen Spanish' was never applied to true PCS varieties such as C or T, except in error. For example, Montero y Vidal (1876: 97) offers the following excerpt from a conversation between a Spaniard recently arrived in the Philippines and a compatriot with long residence in the islands: `---¿Y eso de que los criados entienden todas las cosas al revés? ---Aprenda a hablarles en el idioma sui generis, que llamamos aquí español de cocina, repieiéndoles tres veces la misma cosa. Verá V. cómo lo entienden' [And what's that about how the servants get everything backwards? ---Learn to talk to them in that lingo that we call `kitchen Spanish' here, repeating everything three times. You will see how well they understand]. Escosura (1882: 5) lamented that `los indios mismos que se tienen por instruídos en castellano, lo están tan poco, que es preciso para que comprendan hablarles una especie de algarabía que vulgarmente se llama español de cocina; y para entenderlos a ellos, estar habituados al mismo bárbaro lenguaje' [even the Indians who supposedly have learned Spanish know so little that for them to understand one must speak to them in a sort of jargon known as kitchen Spanish, and to understand them, one must get used to the same horrendous language]. Given the Spaniards' strong negative feelings to `kitchen Spanish,' if they had observed that such a language was the predominant tongue of an entire population, whose speakers used it amongst themselves rather than only to foreigners, this would surely have been mentioned.

 

There are also many examples of Philippine pidgin Spanish as used by native Filipinos, with some creoloid characteristics but still representing an imperfectly acquired second language:

 

No puede, ama; aquel matandá Juancho, casado también `[it] isn't possible, ma'am; that no-good Juancho is also married' ¿Cosa va a hacer ya si nació viva? Siguro yo pegué plojo aquel día `what can [I] do if [the baby] was born alive? I must have been wrong that day.' (Rincón 1897: 22-3)

Pues suya cuidado, pero esa tiene novio castila y seguro no ha de querer con suya `That's your business, but that woman has a Spanish boyfriend and she surely won't have anything to do with you' (Montero y Vidal 1876: 240) 

Mira, jablá tú con aquel tu tata que no suelte el cualtas `Hey, tell your father not to give out the money' (López 1893: 35)

Camino, señor bueno `The road [is] good, sir'

Usted señor, bajar, y yo apartar animales `You sir, will get down [from the carriage]; I will disperse the animals'

Señor, malo este puente `Sir, this bridge [is] no good' (Feced 1888: 20-1)

Bueno, señor, aquí comer `Well, sir, here [you can] eat' (Feced 1888: 24)

Ese palo largo con cordeles atados a su punta y a las puntas de los cordeles anzuelos, cosa buena, señor. Cuando se escapa un preso, corro yo tras de él, se lo echo encima y queda cogido. `Sir, that long stick with ropes tied to the end and hooks on the ends of the ropes is a good thing. When a prisoner escapes, I run after him and I throw the thing over him, and he's caught' (Feced 1888: 34)

No hay ya, siñol; pudo quedá sin el plasa, porque sisante hace tiempo, cuando aquel cosa del flata ... pero no necesitá `He [doesn't work there] any more, sir; he lost the job, he's been out of work for some time, since the time of the money affair, but [he] doesn't need [it] (Rincón 1896: 16-17)

Siguro ha roto aquel rienda, pero en un poco arreglarlo `Those reins have probably broken, but [I] can fix them in a short time' (Rincón 1896: 27)

Metapísico pa, premature no más! Con que no se concibe, ja? `A metaphysician, eh?  You’re premature. So you don’t know, eh?’ (Rizal 1891:98) {mockingly said by a Spanish professor to a Philippine student}

Usté ya no más cuidado con mi viuda y mis huérfanos `You won’t take care of my widow and my orphan children’ (Rizal 1891:222)

 

None of these examples was presented as an instance of Chabacano; most were proffered as illustrations of imperfect acquisition of Spanish by natives of the Philippines, while other examples were presented without comment as `Philippine Spanish.' Nonetheless, the last set of examples represents neither creolized Spanish nor Chinese Spanish pidgin, but rather a wide gamut of L2 approximations to European Spanish by Filipinos who had only occasional opportunities to learn and speak Spanish. A comparison of Chabacano and Philippine `bamboo Spanish' shows that the latter shares some of the creoloid features of the former: word order, gravitation towards the 3 s. verb form, some Philippine and Spanish-derived particles, use of cosa as interrogative word, and some aorist constructions (derived from the Spanish infinitive) without TMA particles. However, `bamboo Spanish' lacks the full range of grammatical structures found in Philippine Creole Spanish, and to the extent that it was based on foreigner-talk proffered by expatriate Spaniards, makes greater use of the bare infinitive than occurs in actual Philippine L2 Spanish.

 

Finally, a few attestations of legitimate Chabacano crop up in late 19 th century literary texts, invariably from Cavite or Manila, and never identified explicitly as anything other than `broken Spanish’:

 

si vos quiere, yo ta emprestá con V. cuatro pesos para el fiestajan del bautizo `if you wish, I can lend you four pesos for the baptism celebration' (Rincón 1897:22-3) Siguro ese aquel que ta mandá prendé cunisós `He’s probably the one that had us arrested’ (López 1893:35)

 

¿Ya cogí ba con Tadeo? `Did they catch Tadeo yet?’ (Rizal 1891:220)

No jablá vos puelte, ñora, baká pa di quedá vos cómplice. Ya quemá yo ñga el libro que ya dale prestau conmigo. Baká pa di riquisá y di encontrá. Anda vos listo, ñora.

`Don’t speak so loud, ma’am, or you’ll be taken for an accomplice. I burned the book that [he] loaned me. Otherwise they could search and find [it]. Be careful, ma;am’ (Rizal 1891:220)

Conmigo no ta debí nada. Y cosa di jasé Paulita? `He doesn’t owe me anything. And what will Paulita do?’ (Rizal 1891:220)

 

These examples show the preverbal particle ta, the future/irrealis particle di, the accusative/dative marker con, and the first-person plural pronoun nisós, found in Cavite and formerly in Ermita (the Ternate form is mihotro, while Zamboanga has kamé [exclusive] and kitá [inclusive].

 

Despite the initial improbability, it is conceivable that Philippine residents of Zamboanga were in fact speaking Spanish at the time the earlier travel accounts were written, at least a close enough approximation to Spanish which Spaniards and other foreign observers would regard as a legitimate approximation to international standards, and not simply `kitchen Spanish.' The current grammatical structure of Z differs significantly from any variety of Spanish, and the two languages are to a large extent mutually non-intelligible between non-initiated speakers of each language. Thus the notion that any non-creolized form of Spanish was spoken in Zamboanga by native Filipinos initially seems unlikely; however, additional evidence deriving from reconstruction and extrapolation from current configurations suggests that this notion may be largely accurate. First, a large proportion of outsiders' descriptions of the linguistic profile of the Philippines in the 19th and early 20th centuries are very ambiguous, reflecting a combination of ingenuousness, ignorance, and undisguised xenophobic, and many of the descriptions could well encompass anything from a rudimentary pidgin to a close approximation to European Spanish, perhaps with only a segmental or suprasegmental accent. Consider the following typical traveller's view of the `Spanish' of the Philippines (Dauncey 1910: 212-3): I daresay you are surprised at my accounts of these and other conversations in Spanish, but the fact is, though I have not tried to learn the patois that obtains in the Philippines, I find it impossible not to pick up a good deal ... They speak badly, though, and the accent does not sound a bit like what one heard in Spain, besides which, there are so many native and Chinese words in current use.  Instead of saying andado, they say andao; pasao for pasado; and so on, with all the past participles, besides other variations on the pure Castilian tongue. I found that the Spanish grammars and books I had brought with me were of so little use for every-day life that I gave up trying to learn out of them ...

 

It is apparent that Mrs. Dauncey was ignorant of legitimately `Castilian' usage, where among other features the realization of -ado as -ao is frequent and socially accepted. Given her silence on more substantive grammatical matters (despite her claim that grammatical textbooks were of no use), we are left with no useful description of Philippine Spanish. Indeed, non-creole Spanish of the Philippines is quite close to Peninsular `Castilian' models, being spoken largely by families with recent ancestors from Spain, and differing from the dialects of the latter country mainly in pronunciation and the occasional slight grammatical or lexical difference. Dauncey's evident inability to understand the `Spanish' of the Philippines constitutes evidence that PCS is what she encountered, despite the lack of corroborative evidence that any Spanish-based creole was ever spoken outside of the Manila Bay enclaves and the previously-mentioned cities of Mindanao. She may also have encountered the `kitchen Spanish,' still alive and well only a few years after the official Spanish departure from the Philippines, and which would be offered to a foreign visitor who apparently did not speak `proper' Spanish herself.  Other observers of the linguistic situation in Zamboanga were less ambiguous. Thus, Worcester (1898: 130) noted that `On account of the multiplicity of native dialects, Spanish became the medium of communication, but they have long since converted it into a Zamboangueño patois which is quite unintelligible to one familiar only with pure "Castellano."'  This can only have been a very recent form of Z. Russell (1907: 172), who visited Zamboanga in 1900, and who had considerable knowledge of Spanish, referred to `Zamboanganese' as `a mixture of Castilian, Visayan and Malay.' She also refers to Zamboangueños speaking `unintelligible Spanish' in moments of excitement. The latter two descriptions are the exception rather than the rule, however, and the majority of 19th century descriptions of Zamboanga mention only `Spanish,' without any suggestion that Peninsular Spanish coexisted with a Spanish-based creole.

 

Z continues to be a vigorous living language, whose oldest living speakers were born towards the end of the 19th century, and who often recall even earlier speech patterns. Thus it is possible, through a combination of fieldwork and oral history, to at least partially reconstruct the linguistic situation of Zamboanga as far back as the middle of the 19th century, with some measure of certainty. This should allow the ambiguous and confusing travellers' accounts to be confronted with hard data, enabling a more accurate picture to emerge. In practice, despite the ready availability of field informants, matters are not always so simple. In my own fieldwork, residents of Zamboanga who had been born in the late 1800's were interviewed, as well as younger residents who accurately recalled the speech of parents and grandparents born even earlier, thus pushing back the date as far as the middle of the 19th century for reasonably trustworthy accounts of Z. In speaking with the oldest residents and in hearing accounts of earlier stages of Z, the observer is immediately struck by the much higher similarity to Spanish.  Most of the interview subjects were aware of later accretions to their own usage, typical of contemporary Z, particularly as regards more Philippine lexical items and syntactic particles.  Many older residents are fluent in Spanish, and their at times precarious awareness of the difference between `Spanish' and `Chabacano' suggests that the latter term was once applied to Spanish derivatives that were significantly less creolized than modern Z. On numerous occasions, although by that time I had mastered Z to the point that no patently Spanish items were slipping in, I was confronted by older `Zamboangueño' speakers who mixed unadulterated Spanish forms, including conjugated verbs, gender and number agreement and more Spanish-like word order, into spontaneous conversations in which the environment gave no indication of upwardly striving language. Such Spanish forms are never used by younger residents of Zamboanga (except occasionally by a handful of radio announcers), and most are not even accurately identifiable by younger community members. When asked to describe the Chabacano speech of older rural residents (known as Chabacano ondo `deep Chabacano'), younger informants could often come up with lexical items, but none explicitly commented on the use of Spanish morphological inflection. When I spoke in Spanish, upon request, younger Zamboanga residents were often unable to comprehend more than the bare minimum, while some commented (with dubious accuracy) that their grandparents or great-grandparents spoke similarly. Another indication of the indeterminacy concerning the use of `Spanish' came when the present writer inquired for names of Zamboanga natives who could speak Spanish, in addition to Z. Virtually every person consulted could come up with names, most representing either middle-aged residents who had studied in private schools run by Spanish religious figures, or old rural informants. Subsequent interviews with individuals so identified produced a wide range of results. A few were able to clearly differentiate `Spanish' and Z, but most individuals in this category are highly educated with considerable formal training in Spanish. Individuals identified by others as speaking `Spanish' but who lacked formal training in the latter language were more likely to speak a fluid mixture of indisputably creole forms common to all speakers of Z and inflected Spanish forms not found in Z.

 

Probing the existing population of Z speakers, and attempting to push back the time base for reconstructing the immediate precursor of Z leads back to the same indeterminacy and apparent muddle concerning the relationship between `Spanish' and Z as a legitimately different form of Chabacano. At the crux of the dilemma is the underlying assumption that the Spanish-based contact languages known collectively as Chabacano are the result of total creolization, i.e. representing an abrupt break from the patrimonial Spanish which was brought to the Philippines.  According to such a belief, the only possible scenario for the inability to assign an element unambiguously to `Spanish' or `Chabacano' is some type of `post-creole continuum,' in which decreolization or reintroduction of Spanish results in a more `Hispanized' Chabacano. The facts regarding Z point in the opposite direction, however. Spanish, at any level of fluency, has all but disappeared from Zamboanga City and its environs (except for some unadulterated Spanish forms reintroduced by radio broadcasters--cf. Lipski 1986a, 1987f--which, however show no signs of spreading to general usage). A century ago, however, Spanish was more widely known, and the further back in time the probe is pushed, the blurrier becomes the `Spanish'/`Chabacano' distinction. In other words, contemporary Z is much less like Spanish, and comes closer to fitting the diagnostics for abrupt creolization, than its predecessors. Clearly, such a configuration is not indicative of decreolization or a post-creole continuum, but rather of a significantly different model of formation, in which creolization in the sense of a geneological discontinuity played a minimal role. This in turn calls for a reevaluation of the theories regarding the formation of Z and other PCS dialects, and the proposing of alternative sources of creoloid structures in the formative stages of Z.

 

4. Towards a theory of the formation of Zamboangueño

 

Most descriptions of PCS have not distinguished between Z and the Manila Bay varieties, assuming implicitly or explicitly that Z is simply the offspring of an earlier transplant of Manila Bay PCS.  Whinnom (1956:3) hypothesized that the formation of the PCS dialects, including Z, was the result of linguistic and cultural mestizaje between Spanish-speaking garrison troops (soldiers from the lowest social classes) and Malay speakers: `only the convivence, and indeed intermarriage, of Spaniard and Malay can account for the fact that a creolized language emerged in the brief space of two generations.' More recently, McWhorter (2000:14) dismisses Chabacano as `having emerged via marriages between Iberian men and Philippine women,’ completely overlooking the origins of Zamboangueño and even Caviteño in nearly all-male military garrisons. The garrison troops, whose presence in Zamboanga was the strongest Spanish influence during the formative period of Z, were drawn from Mexico and from elsewhere in the Philippines, especially from Luzon and some central islands. The Philippine soldiers presumably learned Spanish from the Mexican troops, and from other Spanish speakers already in the Philippines.

 

Frake (1971) implicitly accepts Whinnom's hypothesis of the garrison-troop origin of Z, but makes the intriguing observation that many of the contemporary Philippine items in Z do not come from the geographically contiguous Visayan languages, but from Hiligaynon (Ilongo), spoken in the Central Philippines. Most of the words in question are lexical items with no particular semantic restrictions, but a number of core syntactic items are included. Frake gives no explanation for the presence of Ilongo items in Z, except to suggest that many garrison troops probably came from the Ilongo-speaking area. There may be additional or alternative routes of penetration; for example, Iloilo (the principle city in the Ilongo region) was one of the main stopover ports for ships travelling from Manila to Zamboanga (Warren 1981), and it is likely that Ilongo speakers were picked up along the way. Another potential missing link in the evolution and spread of the various PCS dialects comes from the indirect evidence that when Zamboanga was rebuilt in 1719, many PCS-speaking families from Cavite emigrated to Zamboanga, with some remaining in Iloilo (Germán 1984). Although PCS never became implanted in Iloilo, if family ties existed between Iloilo and Zamboanga, including the possibility for subsequent migration of settlers originally stopping in Iloilo, Ilongo words could have arrived in Zamboanga by this means. Maria Isabelita Riego de Dios (personal communication) has also discovered that many laborers were recruited from Panay (the main island where Ilongo was spoken) during the time period when Zamboanga and Cotabato were building up their military defenses, and she suggests that the Ilongo elements in the PCS dialects of both cities is a direct result of this immigration.

 

Much of the failure to separate the formation of Z from the Manila Bay PCS varieties comes from the status of the latter creoles in theories of Iberian-based creole formation.  Whinnom (1956) was the first to hint at a possibility which was later to become a full-fledged theory, namely that a large number of Asian-Iberian creoles, from India to Indonesia and including PCS in the Philippines, result from a single precursor, a Portuguese maritime pidgin which mixed with local languages as well as with other colonial superstrata to yield the variety of creoles now found in Asia and Oceania. Whinnom (1956), in a remarkable feat of historical reconstruction, postulated that the seeds of PCS were first sown on the Indonesian island of Ternate in the 17th century, where a Portuguese-based creole apparently arose on this important member of the Spice Islands. When the speakers of this proto-Portuguese creole were expelled sometime later, these Mardikas or Merdikas as they were known ended up in the Spanish colony of the Philippines, settling in several small villages along Manila Bay. The town of Ternate, where the PCS variety T is spoken, appears to have been named after the Indonesian island, and oral tradition among the Ternateños refers to the Mardikas. Since T seems to be the oldest surviving variety of PCS, the chronology is at least correct. Currently all traces of any Portuguese-based creole have disappeared from Ternate, Indonesia. A few years ago, a sultan from the latter island visited Ternate, Manila and found the T variety of Chabacano totally incomprehensible, not even recognizing individual words.

 

Despite Whinnom's pioneering contributions on the origins of the Manila Bay PCS dialects, his description of Z is sketchy, inaccurate, and based on second-hand sources. He assumed, naturally enough, that some form of Manila Bay PCS had been carried to the Spanish garrison at Zamboanga, where it continued to flourish and evolve; any differences between Z and the Manila Bay PCS dialects were presumably the result of local accretions rather than from a separate formative process. Subsequent in-depth investigations of Z, such as Frake (1971, 1980) and Forman (1972) implicitly assume some version of Whinnom's hypothesis. In view of the significant structural similarities between Z on the one hand and C and T on the other, it is not feasible to claim totally independent creolization in Zamboanga. However, additional evidence suggests that Z did not simply grow from a nucleus of transplanted Manila Bay PCS.  The only challenges to the notion that Z is somehow a transplanted variety of Manila Bay PCS with local overlays come from non-linguistic accounts, which often err in the direction of assuming no connection at all between Z and its Manila Bay homologues. The most elaborate alternative account of the formation of Z is suggested by Warren (1981). From the 17th century to well into the 19th century, Moslem pirates and slave raiders from Jolo and other islands in the Sulu Sea attacked many parts of the Philippines and carried off captives, who were pressed into slavery. Many of the slaves held on Jolo managed to escape to Zamboanga. Spanish and English military vessels also rescued slaves, usually depositing them at the nearest port under Spanish control; Zamboanga was a frequent dropoff point. Some freed slaves delivered to Zamboanga were once more forced to labor by the Spanish military authorities, and thus spent more time in the vicinity of Fort Pilar than they had originally intended, long enough, perhaps, for a Spanish-based contact vernacular to form or be extended by speakers of different Philippine languages. Warren (1981: 235-6) speculates thus: The fugitives established themselves with impoverished Chinese and vagrants in a community situated some distance from the presidio. Originating from different parts of the Philippine archipelago and lacking a common language, these degradados developed their own Spanish-Creole dialect -- Chavacano – to communicate. A large percentage of the surrounding rural population labelled Zamboangueno at the end of the nineteenth century were descendants of fugitive slaves who had lived on the margins of the presidio as social outcasts.  This idea is not new, for Worcester (1898: 129-30), in describing the population of Zamboanga, stated that: ... certainly a very considerable portion [of the Zamboangueños] are the offspring of slaves who have contrived to escape from the Moros ... the result has been that representatives of most of the Philippine coast-tribes have found their way to Zamboanga, where their intermarriage has given rise to a people of decidedly mixed ancestry. On account of the multiplicity of native dialects, Spanish became the medium of communication, but they have long since converted it into a Zamboangueño patois ...

 

In a later account, Worcester (1930: 512) noted that `Zamboanga was at the outset populated by escaped Moro slaves who had sought the protection of the Spanish garrison there. Coming originally from widely separated parts of the archipelago, these unfortunates had no common native dialect, hence there arose among them a Spanish patois known as Zamboangueño.'  Other descriptions of Zamboanga also speak of the mixed origins of its residents. Thus Vendrell y Eduard (1887:62), in speaking of Zamboanga, observed that `estos indgenas, la inmensa mayoría mestizos españoles, proceden en su origen de otras provincias del Archipélago, y muchos de Méjico, de donde llegaron á principios de este siglo, cuando perdimos aquel imperio’ [these indigenous people, the great majority of whom are Spanish mestizos, originally come from other provinces and from México, whence they arrived at the beginning of this century when we lost that empire]. These accounts suggest that Z arose in situ as a contact vernacular among transients and freed slaves. While it is likely that the linguistic heterogeneity of the Zamboanga garrison and its environs was conducive to the evolution of whatever Spanish-based lingua franca was adopted there, it is not possible to accept that Z arose ab ovo in Zamboanga or anywhere else in Mindanao. A number of factors militate against such a position.  First, the grammatical similarities between Z and Manila Bay PCS dialects are too striking to overlook, and point in the direction of importation of at least some grammatical elements from Manila Bay. It is likely that freed slaves in Zamboanga would adopt an already existent Spanish-based pidgin/creole, assuming such was already in existence in the military camp, and particularly if large numbers of former slaves were forced to labor in the fort. If these ex-slaves remained outside the pale of the fort, however (as observed by Sonnerat 1776: 127), it is more likely that a contact vernacular based primarily on Philippine languages would have arisen. My own research suggests that such a Philippine-based contact vernacular did indeed serve as the primary input to Z, but assuming the latter to have derived only from the speech of former slaves would not account for the similarities with Manila Bay PCS. At the same time, a much more heterogeneous mix of lexical items from scattered Philippine languages would be expected for such an extramural developing creole. Z lexical items of Philippine origin come, in descending order, from (i) regional Visayan, which has become a frequently spoken language in western Mindanao at least in the last century; (ii) Ilongo; (iii) very occasionally, Tagalog. Finally, known demographic and historical facts about Zamboanga fail to confirm the notion that a `large percentage' of rural Zamboangueños descend from former slaves, although the fact that a Spanish-based contact vernacular rather than a local Philippine language is spoken so far from Zamboanga City must be accounted for. Since the non-Moslem population of southwestern Mindanao was quite small prior to the establishment of Zamboanga, the answer may lie in the simple fact that there was no appropriate local language available to the developing rural Christian population. The views of Worcester, Warren and others who postulate that Z was essentially created in Zamboanga cannot be sustained as the principal hypothesis, although there is no doubt that returned slaves and other transients who made their way into Zamboanga were instrumental in increasing the number of speakers of any Spanish-based contact language.  A key factor in tracing the development of Z is the determination of the features already present in non-creole varieties of Spanish used by Philippine natives. Although little direct documentation is available on earlier stages of Philippine Spanish, data from the last decades of the 19th century suggest that many of the features of Z than cannot be directly traced either to European Spanish or to a cross-section of Philippine languages were to be found in the `Philippinized' Spanish used widely between Spaniards and Filipinos for several centuries. The ultimate source of such pre-creole items remains to be determined. Chinese merchants may have introduced some key items into `kitchen Spanish'; if some of the Chinese had previously learned a Portuguese-based pidgin or creole in Macau or Hong Kong, they might use the same words when attempting to speak Spanish, a language which they would identify with Portuguese. For example, the use of what appears to be a first-person subject pronoun derived from mi, which is never found in any variety of PCS, may have been transferred either from Portuguese pidgin or from Chinese pidgin English (e.g. Montero y Vidal 1876: 241, `Mia quiele platicalo' and Feced 1888: 77, mía aquí vendelo, ganalo'). Other items may have been introduced by Spaniards, based on stereotypes of pidgin speech which had already been solidified for Afro-Hispanic speech, including literary stereotypes dating from the beginning of the 16th century (cf. Lipski 1991). Most of the creoloid forms appear to be simply represent the natural amalgam of Spanish words and Philippine morphosyntactic patterns, calques into Spanish that would be immediately recognized by any Philippine speaker, and which could be understood at least partially by native Spanish speakers. Since identical or similar expressions are found in nearly all Philippine languages, Philippine listeners immediately understand the expression the first time they hear it, and Spanish-speaking arrivals in the Philippines picked up this expression as one of the first accretions to their `colonial' lexicon.

 

Many other pieces of the Z puzzle can also be found in non-creolized Philippine Spanish.  For example, the use of cosa as a generic interrogative is attested in Philippine Spanish, ranging from reasonably fluent to `kitchen Spanish' varieties:

(Feced 1888: 68-69): `¿También redactarás las actas de las sesiones? ---¿Cosa eso, señor?' [Will you also take minutes of the meetings? What is that, sir?] (Feced 1888: 91): `Quiero decir que tendrás muchos galanes. ---¿Cosa galanes?' [I mean that you must have many beaus. What are beaus?] Montero y Vidal (1876: 239): `¿Cosa, señolía?' [what is it, sir?]. (López (1893: 34): `¿Cosa? preguntó el maestro' [What is it? asked the teacher].  (Entrala 1882: 12): `¿Cosa dice?' [what is he saying?].  (Entrala 1882: 22): `Cosa Goyo? ... cosa tiene?' [What is it, Goyo? ... what is there?]  The latter quote also exemplifies the use of affirmative tiene to indicate `there is/are,' also found in Entrala (1882: 22): `Tiene canin, tiene nata, tiene coco ... ' [there is {cooked} rice, there is cream, there are coconuts ...]. Other interrogative words are used similarly, for example cual (Entrala 1882: 32): `Cual aquel?' [which one is that?].  PCS varieties, including Z, are characterized by an invariable verbal stem, usually derived from the infinitive minus final /r/. In the case of `modal' verbs and some other verbs, the third person singular form has been taken over: puede, tiene, sabe, etc. Vestigial and semifluent Spanish of many countries is noted for the gravitation of verbal paradigms to the third person singular, an attestation of partial agreement. Vernacular Brazilian Portuguese employs this strategy, as do vestigial dialects of Spanish. In contemporary Philippine Spanish, vestigial speakers occasionally reduce verbs to the third person singular. This is by no means an indication of a recent development found only among generations of Filipinos for whom Spanish is only a partially-learned ancestral language. There is ample evidence of this tendency in earlier stages of Philippine Spanish, when the language varieties in question represented the base variety of `Spanish'; for example: (Feced 1888: 90): `¿Capaz serás todavía de enamorar a algún capitán y casarte otra vez?  ---No sabe, señor.' [Could you still get some captain to fall in love with you and marry again? {I} don't know, sir.]  (Feced 1888: 103): `Oy, piloto, ven acá: ¿tú conoces esto? ---Conoce, señor ...' [Hey, pilot, come here. Do you know about this? {I} know, sir.]  (Feced 1888: 108): `¡Cochero! ¿Qué entiendes tú de eso? ---¡Sí, entiende, señor!'  [Driver! What do you make of this? Yes, {I} understand, sir!].  In Z, nuay < Spanish no hay `there does not exist' behaves exactly as its Visayan and Tagalog homologues, wala and dili, not only to refer to lack of existence, but to negate certain verbs. The same usage recurs consistently in non-creolized Philippine Spanish: speakers of Philippine Spanish at times used no hay to indicate that a person or thing inquired after was not present: Feced (1888: 15): `Que venga el médico,---dije al muchacho indio que me servía de ayuda de cámara. ---No hay más, señor. ---¿Cómo no hay más? ---Se ha marchado á recoger un muerto á dos días de caballo.' [Have the doctor come, I told the Indian {= native} cabin boy. He's gone sir. What do you mean he's gone? He went to pick up a dead person; it's two days on horseback].  Similarly, use of no hay to replace no tener `not to have,' is attested in Philippine Spanish: (Entrala 1882: 12): `que no hay cualtas' [{he says} that {he} has no money].  The use of con as objective case marker in Z and occasionally in the Manila Bay PCS dialects has already been commented on, together with the fortuitous similarity with some Visayan pronominals. In non-creole Philippine Spanish, use of con to signal accusative case is also attested: Señor, haga pabor de emprestar conmigo cuatro pesos' `Sir, please lend me four pesos' (Feced 1888: 42) 19  Pues suya cuidado, pero esa tiene novio castila y seguro no ha de querer con suya `Well, that;s his problem, but she has a Spanish fiancée and she probably doesn't love him' (Montero y Vidal 1876: 240) señor, V. sin duda no recordar conmigo `Sir, of course you don't remember me' (Moya y Jiménez 1883: 293) señor, más mejor que de usted conmigo seis pesos de sueldo ... `sir, it would be better for you to give me six pesos' salary ...' (Moro y Jiménez 1883: 285)  An example of Chinese `kitchen Spanish' is:

(López 1893: 58): `¡Ah! señolia, mucho disgustalo ele con suya, polque señolia manda plendé con ele' [Oh sir, he is very angry with you, because you had him arrested].  A characteristic of Z, calquing a wide variety of Philippine languages, is the lack of copula with Adj + Noun combinations. The same combinations are attested for non-creole Philippine Spanish:

Señor, malo este bache ... malo este puente grande `Sir, this pot hole {is} bad ... this big bridge {is} bad' (Feced 1888: 21)

Seguro tú grande el robo `You {can be} sure {it was a} big robbery' Entrala (1882: 22) V. magandang lalaque; fino el talle, bueno el cara `You are a fine lad; slim-waisted, good-looking' Entrala (1882: 23) pero malo ese ... `but that guy is bad' López (1893: 34)

 

The preceding examples show that Philippine Spanish, developing slowly throughout the major population centers of the islands and incorporating calques of regional Philippine languages, already contained the seeds of many creoloid structures, which when added to the mix of Spanish and cognate Philippine elements in the formative period of Z would enhance emerging creole structures. Philippine Spanish, at lower levels of fluency, also embodies considerable grammatical simplification which does not specifically reflect Philippine syntax, but which is common to reduced and vestigial forms of Spanish of other nations. Even in the 20 th century, travelers continued to describe rudimentary Philippine Spanish with the same terms used in previous centuries. Russell (1907) refers to the use of `broken Spanish' in several parts of the country, and at one point a man spoke to her in `what he was pleased to consider Spanish' (p. 81). Correa de Malvehy (1908:109), visiting the Philippines towards the end of the 19th century, makes similar reference to the fact that `también se habla generalmente en español más ó menos incorrecto, siendo la lengua oficial de la colonia y general de Manila' [more or less incorrect Spanish is also spoken, being the official language of the colony and generalized in Manila]. She also referred (e.g. p. 17) to the `broken Spanish' used by many Filipinos, and even gave an example (p. 135): `¡Calla castila, que corta aquel cabeza tuyo!' [shut up white woman, or I'll cut off your head]. Studies of contemporary Philippine Spanish by Lipski (1986b, 1987a) reveal the continued existence of similar structures, which when taken in their totality closely resemble the PCS dialects in many respects.

 

The examples just given show that many of the important building blocks which would coalesce to form Z were present in non-creole varieties of Philippine Spanish, as spoken by Filipinos and evidently also by Spaniards at times, in order to enhance communication. None of the examples comes from PCS-speaking communities, but rather provide a cross-section of usage which, extrapolating backwards only a few decades, could have been found in the rudimentary knowledge of Spanish shared by Spaniards, Philippine garrison troops, and former slaves in Zamboanga. These features alone do not suffice to explain all the creoloid traits in Z.  However, in combination with the Philippine common denominators surveyed in the preceding section, nearly all the major structures of Z can be accounted for without postulating a transplantation of a functioning PCS speech community to Zamboanga.

 

In partial summary, it has been proposed that Z did not arise as a radical creole from purely Spanish roots. Z came into being as residents of Zamboanga, both those of long standing and new arrivals, enhanced inter-ethnic communicability by drawing ever more heavily on the one extraterritorial language which had already begun to bridge the gap, namely Spanish. In Zamboanga, Spanish in its native or quasi-native form was principally the vehicle of some of the garrison soldiers (especially those from Mexico) and their commanding officers. This would account for both the phonologically more modern form of many Spanish items in Z and the noticeably vulgar, barracks-like nature of many of the borrowings. Items derived from earlier periods of Spanish are probably actually derived from Philippine languages, which began absorbing Hispanisms as early as the 16th century. During most of the 18th century, the non-Moslem population of the Zamboanga area was small, and the potential Philippine common denominators available to early generations of Zamboangueños would be smaller than if only the intersection of Tagalog and the major Visayan languages is considered. Many of the peripheral Philippine languages do not share the morphological and lexical similarities which have been proposed as having contributed to the formation of Z, so that recourse to common Spanish items, and incorporation of further Hispanisms (presumably including Manila Bay PCS forms carried to Zamboanga) would be the preferred means of expanding the Z lexicon. During the 19th century, the ethnic demographics of the Zamboanga area stabilized as piracy in the Sulu Sea was reduced.  During the same period, the presence of native or near-native Spanish speakers in Zamboanga City reached an all-time high, and Z absorbed most of its newest Spanish lexical items. Ilongo elements were introduced during this period, which may be explained by the the Cavite-Iloilo-Zamboanga connection suggested earlier.

 

The first stage of Z is assumed to have consisted mostly of Spanish items and of only the broadest Philippine common denominators such as the plural particle mga, interrogative particle ba, plus a few Ilongo words. Cebuano/Visayan accretions came later. Immigration to southwestern Mindanao of speakers of central Visayan languages, particularly Cebuano, became significant towards the end of the 19th century, a population shift rivaled in attested Philippine history only by the immigration of Tagalog speakers to central Mindanao (including Cotabato and Davao) as part of a homesteading movement in the 20th century. For the first time since its inception, Z was placed into contact with a cluster of mutually intelligible Philippine languages, whose speakers came to exercise a significant economic and social influence in Zamboanga City.  This fact, coupled with the rapid decline in the official Spanish presence, in the number of Spanish speakers, and in the social desirability of learning Spanish, caused the balance to tip in favor of Visayan as the source of new lexical items. That the mere presence of a common Philippine language was not sufficient to induce such lexical incorporation is indicated by the fact that Zamboanga City has always contained a large number of Tausug speakers, and yet Z has remained unaffected by Tausug and other languages of the Moslem population.  Although today Z has firmly integrated Visayan elements such as pronouns to such a degree that these elements appear to have formed part of the original language, they are in fact more recent accretions. Although use of Philippine pronouns in Z had already begun by the second half of the 19th century, the consolidation of the new paradigm to the nearly total exclusion of Spanish competitors (except marginally for bosotros/ustedes) occurred around the turn of the 20th century, i.e. within living memory of the area's oldest inhabitants. Even within the last two generations, the further introduction of Visayan elements, and a shift to more noticeably Visayan patterns of pronunciation, is observable.

 

To conclude the proposed reconstruction of Z, this language came into existence as an independent language towards the middle of the 18th century. This language has an especially rich history of partial relexifications, in a region characterized by multilingual contacts and a very fluid series of demographic movements. Z began not as a true creole, but as a natural common intersection of grammatically cognate Philippine languages which had already incorporated a lexical core of Spanish borrowings. The pool of speakers who provided the original input for Z did not constitute a single group, but included garrison troops, transients and later, former slaves recaptured from Moslem territories to the south. Over the period of a century and a half, Z partially relexified in a number of directions, with each stage of relexification responding to particular demographic or social events. At least the following stages may be tentatively proposed, based on direct documentation, indirect reconstruction, and chronological extrapolation:

 

STAGE I: (mid 1700's) Z arises in the Zamboanga garrison, as the common intersection of Spanish-laden Philippine languages.

 

STAGE II: (mid-late 1700's). Z absorbs grammatical and lexical structures from Manila Bay PCS, as the Spanish military presence in Zamboanga is consolidated.  Additional migrations of civilians from Cavite have a trickle-down effect on Z.

 

STAGE III: (1800's ?). Ilongo lexical elements are introduced into Z, possibly as the result of the use of Iloilo as a stopover for ships bound from Manila to Zamboanga. Ilongo grammatical forms could have been introduced at this time.

 

STAGE IV: (most of 1800's). Increasing presence of (civilian) native Spanish speakers in Zamboanga City results in incorporation of additional Spanish items, with structural differences between Z and (Philippine) Spanish reaching their alltime low point.

 

STAGE V: (Turn of 20th century onward). Large-scale immigration from the central Visayan region to southwestern Mindanao makes Cebuano Visayan the de facto number two language in Zamboanga City. Spanish lexical items are increasingly replaced by Visayan items. Word order begins to shift towards Visayan.

 

STAGE VI: (1930's onward). Increasing use of English in Zamboanga, not only in schools but even in casual conversations, results in growing incorporation of Anglicisms into Z. In the last two generations, this is leading the way to an eventual relexification of Zamboangueño away from its Hispanic lexical basis.

 

The reconstructed stages proposed above paint a picture considerably different from `typical' creole genesis occurring in other parts of the world, and explain the typological differences between Z and other Spanish-based creoles. This includes word order, use of particles, structure of the VP and NP, and many other features. Even among the PCS dialects, Z differs substantially from the Manila Bay PCS dialects, despite the fact that the adstratum Philippine languages in contact with Z and the Manila Bay PCS dialects are cognate and similar in basic structure. The difference, as outlined above, lies in the fact that the Manila Bay PCS dialects, especially T, apparently began life as true creoles, based on a nearly exclusively Spanish input (possibly with some creole Portuguese contributions), with Philippine elements added only later, without altering the basic patterns already developed. Z, on the other hand, began life as a hybrid pan-Philippine contact language whose Spanish items had already been filtered through Philippine languages, and which was therefore a Philippine language in the structural sense at every point during its existence.

 

5. The current status of Chabacano in the Philippines

 

Chabacano is currently spoken by a small number of elderly individuals in Cavite (especially the San Roque neighborhood), who may intersperse Chabacano expressions with the more frequently used Tagalog when speaking amongst themselves. A Circulo Chabacano informal group has from time to time attempted to revive the language, and church and town festival bulletins sometimes contains poems or sayings in Chabacano. In Ternate an equally small number of Ternateño speakers remains, but it appears that the average age of the speakers is somewhat younger. There is no indication that children in either community are being taught the language or even have opportunities to hear it. Ermita Chabacano has disappeared, although a few vestigial speakers undoubtedly remain somewhere in the Metro Manila area. It is in Zamboanga City and its environs that the Zamboangueño variety of Chabacano makes this the largest Spanish-derived creole speaking community in the world. In Zamboanga, despite the lack of official status, Chabacano is the de facto language of choice in nearly all circumstances.  In Zamboanga City, all the radio stations and local television productions broadcast in Chabacano (with constant code switching and introduction of English words and phrases), with a few newscasts in English and "public service" announcements in regional languages and Tagalog, as appropriate. The use of Chabacano in broadcasting is a relatively recent trend, since as late as ten years ago, most programming was in English (given the lack of acceptance of Tagalog materials), and it was only after the first experimental Chabacano programs were received with unequalled enthusiasm that the broadcast media gradually shifted over to this language. In the schools, English is the official medium of instruction from the outset, with Pilipino (based on Tagalog) taught as an obligatory subject. Despite official disapproval of halo halo (i.e. mixed or code switched) speech, in the classroom, most teachers use at least some Chabacano in presenting lessons, particularly in the lower grades, and also because the English abilities of many teachers are severely limited; for a general discussion of this type of situation.  During the 1960's and 70's, programs in "vernacular language instruction" were instituted for the first few grades, and while these have largely been suspended, some rural schools still implement the Chabacano textual materials, and de facto bilingual instruction is the rule in all public and many private schools.

 

Zamboanga City is one of the most thoroughly Hispanized cities remaining in the Philippines, and at one time, culminating in the first decades of the present century, some form of the Spanish language circulated rather widely in that city. Several Spanish-language newspapers were published until World War II, Spanish was used by (Spanish national) priests, and signs and announcements in Spanish were frequently seen about town. Although the general feeling among Zamboangueños is that Zamboanga was at least partially Spanish-speaking towards the end of the Spanish period and in the decades immediately following, it is more reasonable to suppose that