The Interacting Influences of Spanish and English on the Creole Papiamentu

Tara Sanchez
University of Pennsylvania

1  Introduction
    The Spanish gerundive and progressive morpheme -ndo is used in modern Papiamentu (PP) (Iberian creole) as a gerundive and progressive morpheme, though it is not found in the early creole. The morpheme appears to be borrowed. I argue that the modern use of -ndo in PP is the result of a two-step borrowing process. First, contact with Spanish led to the borrowing of the affix with its gerundive function (used in absolutive clauses). Next, contact with English led to the extension of the morpheme to a progressive function (used in periphrastic progressives) via surface string matching, a process through which bilinguals have been shown to import the pragmatic function of a syntactic construction from one language to another (Prince 1988, Silva-Corvalán 1993). In both steps, -ndo is an L1 interference feature of L2 PP speakers, which L1 PP speakers eventually incorporate into their grammars.

    Below, I present relevant sociohistorical information, describe Spanish -ndo and relevant PP structure, then present results of a real time, quantitative analysis of the affix in PP and Venezuelan Spanish texts supporting my hypothesis.

    2  Relevant Sociohistorical Facts

    PP is spoken on the islands of Aruba, Bonaire, and Curaçao, which lie near Venezuela in the Caribbean Ocean. Arawak Indians lived there when the Spanish arrived in 1527. The Dutch took possession in 1634, forcing the Spaniards and most Indians to leave. They took some Indians as slaves on Curaçao, and sent others to Bonaire and Aruba. This change in power did not necessarily lead to a change in language spoken, however, because the Dutch often preferred to use Spanish or Portuguese or creole Portuguese with conquered peoples, and Dutch amongst themselves (Holm 2000).

    The Dutch brought the first West Africans in 1648; at least some probably spoke pidgin Portuguese. Sephardic Jews relocated from Brazil beginning in 1659, probably speaking a regional variety of Portuguese or Spanish. After 1660, Jews played a major role in the administration of slave camps, often trading with Spanish Americans. By the 1680s, the African population equaled the white population. Most slaves did domestic work.

    Dutch and Jews learned the emerging creole for contact. Holm (2000) estimates that the creole stabilized on Curaçao around 1700, then spread to Bonaire and Aruba. PP words are attested in Jewish ship names in 1706, and Dutch documents in the 17th and 18th centuries. By the time the creole was fully established in the late 18th century, Dutch missionaries preached in PP.

    After slavery was abolished, the islands remained under Dutch control, but later achieved some autonomy. Curaçao and Bonaire now belong to the group of islands called the Netherlands Antilles. Aruba was part of the N.A. until 1986, when it became independent of this group. (It remains under the jurisdiction of the Netherlands.) Today, about 80% of island residents speak PP as a first language. Dutch remains the language of government and education. Spanish is culturally important. Portuguese was used during the early slave trade, but fell out of use by 1800. English has only recently (1915 on Curaçao, 1928 on Aruba) entered the picture on the islands with the introduction of the petroleum industry, and is economically important in oil and tourism. Today, residents speak PP, Dutch, Spanish, and English.

    3  Spanish -ndo

    In Spanish, verb + -ndo forms the progressive participle (1). The periphrastic progressive (2) and locative constructions (3) consist of estar, ?to be? + verb + -ndo. A verb marked by -ndo can be used after other auxiliaries ((4) such as ir ?to go? or andar ?to walk?). Stative verbs are prohibited in these constructions (5) though they are permitted in absolute clauses (6).

    (1)        a. hablar + -ndo  -->  hablando

      b. decir   + -ndo   -->  diciendo
      to speak +  GER   -->  speaking
    (2)         Ella está estudiando.
      she COP study-GERShe is studying.
(3)        Estaba Juan en el desierto bautizando?           (Mark 1:4) COP-IMP John in the desert baptize-GER
John was in the desert baptizing?
(4)        Todos te andan buscando.                                  (Mark 1:37) all-PL you walk-3pl look for-GER
All of them are [walking around] looking for you.
(5)        *Ella está queriendo estudiar. she COP want-GER to study.
She is wanting to study.
(6)        Queriendo estudiar, ella se fué a la biblioteca. want-GER to study she go-PAST to the library. Wanting to study, she went to the library.
4  Relevant Structure of Papiamentu

From the earliest evidence of the creole, we find preverbal markers of the imperfective-ta (from Sp. esta) and tabata (from Sp. estaba). Andersen (1990) analyzes tabata as specifically past marked and ta as unmarked for time. Both indicate general imperfective aspect, which may include a more specific progressive or habitual interpretation (7-9) (Maurer 1998:162). Ta in PP functions beyond Spanish esta, however. In the creole, Andersen shows that time of action can be marked at some point in the discourse, after which ta indicates agreement with the previously marked time of action. (10) (Andersen 1990). Ta and tabata also serve as forms of the copula in PP.

(7)         Awor mi ta lesa.

now I IMP read.
Now, I am reading/ I read.
(8)         Mi ta lesa tur dia. I IMP read all day.
I read every day.
(9)        Mi ta traha na postkantor. I IMP work in postoffice
I (regularly) work in the postoffice.
(10)       ...tempunan di nos ta'ata tempu ku dios ta'ata aki bou huntu ku nos. Tempunan ei bo ta tin ku risibi ku djesdos aña, no promé. ...time-PL of we IMP-P time that God IMP-P here below together with us. time-PL there you IMP have COMP receive with twelve year not before.
Our time was the time God was down here with us. Those days you would go?you had to receive (First Communion) at 12, not before.
5  Papiamentu -ndo in the Literature
    Gerundive -ndo is first mentioned in a grammar of PP in 1898 (Evertsz). Formation paradigms are given: -ando and -iendo surface, depending on the final vowel of the PP verb. Evertsz notes these irregular forms: ta ?to be? àestando, siendo; tin ?to have? àteniendo, which are identical to the Spanish gerunds of estar, ser, and tener. Hoyer (1918) and Lenz (1928) describe -ndo as a "recent borrowing" marking only Iberian verbs and occurring in absolutive phrases (11). They give additional, Spanish-like, irregular forms, including por ?to be able to? àpudiendo, kier ?to want? à keriendo, and kere ?to believe?à kreyendo. The existence of these irregular -ndo forms identical to their Spanish counterparts suggests that they were borrowed as unanalyzed wholes and the affix only later used productively.

    In 20th century grammars, authors have repeated that -ndo was "recently borrowed", but add the progressive function to the repertoire of the affix. None has noted that progressive -ndo used with ta or tabata (12) is not mentioned in early descriptions.

    (11)      ...i kantando na bos altu el a bolbe kas.

      ...and sing-GER in voice high he PER return home
      ...and, singing in a high voice, returned to his house.
    (12)       Nos ta yegando fin di aña, kual ta un temporada konosi kaminda ta toka hopi musika di tambu
      we IMP arrive-GER end of year, which COP a season known which IMP play much music of drum. (92:1)
      We are approaching the end of the year, which is a season known for (people) playing lots of drums.
    That -ndo is borrowed is not in dispute. In dispute is the fact that -ndo was borrowed with both its Spanish functions at the same time (gerundive and progressive), a "fact" only implied by recent literature.

    6  Methods

    Data are taken from 99 PP texts, of varied genres, dating 1775-1999. I coded every instance of ta or tabata marking a verb, and noted whether or not the verb was marked by -ndo. I also coded verbs marked with -ndo in other places (i.e. absolutive clauses, main verbs with auxiliary other than ta or tabata, etc.). Up to 100 tokens were taken from each text, though some contained more (e.g. Books of Mark, over 400 tokens each). Texts were grouped roughly according to decade for regression analysis. There were 2911 tokens. Of 2911 total clauses, 145 were cases of -ndo. Of 1419 main verb imperfective clauses, 52 were -ndo-marked progressives.

    I compare the PP data with analogous data from 45 Venezuelan Spanish texts of similar genres, dating 1831-1999. This variety is used because of its geographic proximity to the islands. There were 2255 tokens. Of 2255 total clauses, 126 contained -ndo. Of 1389 main verb imperfective clauses, 23 were -ndo-marked progressives.

    Finally, religious materials allow direct comparison of texts. There are two PP translations of the Book of Mark (1865 and 1916). I compare these to a version of the Spanish Bible which was published in 1850.

    7  Presentation and Analysis of Data

    7.1  Observations from the Early Texts

    The earliest texts show no -ndo (1775 letter,1776 court testimony, Curaçao). The earliest attestation is found in an 1803 letter from Aruba (13).

    (13)       ...y por Ser berdad noos ta firma Ees die noos mismo mano ofresiendo nos hoeramento delantie die triebunal die nos mayor gobierno Aruba 22de junio 1803
      ...and be able to be true we IMP sign this of our own hand offer-GER our oath in front of tribunal of our greatest governer (Maduro 1991)
      ...and for this to be true, we sign this with our own hands, offering our oath before the tribunal of our greatest governor.
    The next available texts date from the 19th century: 1865 Mark; 1871 Curaçao newspaper; 1882 Catechism (Curaçao); 1892 Catholic prayers; 1916 New Testament Bible. -Ndo is found in all but the catechism.

    None of these early religious works were originally written in PP. The New Testament Bibles could have been translated from Spanish or Dutch. Most Dutch were Protestants, the known Portuguese speakers were Jews, and there were few, if any, English speakers on the islands in the 19th century, so Spanish is the most likely source language for the Catholic texts. The -ndo affix was more than likely first used in PP as an interference feature of Spanish-dominant bilinguals.

    Unlike the religious texts, the 1871 Curaçao newspaper articles were originally written in PP (not translated) and were more than likely written by PP dominant speakers, or possibly even Dutch dominant speakers (about half of the newspaper articles are written in Dutch and half in PP; none are in Spanish). The -ndo examples in the Curaçao newspaper (e.g. (14)) provide evidence that -ndo was a productive morpheme used by L1 PP speakers at this time, not simply a Spanish interference feature.

    (14)      Toer miembronan tabata presente kitando doos, koe ta miembronan Royer i Heldewier

      all member-PL COP-P present subtract-GER two, who COP member-PL Royer and Heldewier               "Koloniale Raad," Civilisado, Curaçao newspaper, 7-8-1871, No.2
      All members were present except two: Royer and Heldewier.
    All of the 19th century -ndo cases have one important thing in common-none are periphrastic progressives. Other than in absolutive clauses, V-ndo appears a handful of times in serial constructions with an auxiliary verb other than the copula (e.g. (15)). Thus, while this affix was used productively in PP in the 19th century, periphrastic progressive -ndo constructions like those found in Spanish were not in use.

    (15)      ...eel a saka soe manoe afoor i a tokele bisando: mi kieer pa bo bira limpi

      ...he PERF take his hand out and PERF take-it say-GER I want COMP you become clean. Mark 1:41
      [Jesus] extended his hand: and taking it, he said: You are clean.
    Nonce borrowings provide evidence of the continuing influence of Spanish. The locative construction in (16) was the only one found in the older texts, and is characteristic of Spanish gerunds. The earliest example of what looks like a periphrastic progressive is found in the 1916 Bible (17).

    (16)      Spanish             Estaba Juan en el desierto bautizando...

      PP 1916             Juan tabata batizándo den desièrto?
      English              John was in the desert baptizing?                                      Mark 1:4
    (17)      ...y e Señor tawata trahando huntu cu nan?
      ...and the man [Jesus] COP-P work-GER together with them?               Mark 16:20
    7.2  Real Time, Quantitative Analysis

    7.2.1 The Rise of Progressive -ndo

    Figure 1 shows the increase in periphrastic progressives. There are no cases before 1920; by the 1990s, they comprise 11% of main verb imperfectives.

    Figure 1. Regression of main verb -ndo

    7.2.2   Comparison of Spanish and Papiamentu Progressive -ndo

    I examined Venezuelan Spanish texts to find the frequency of -ndo in main verb constructions. While the rate of estar + verb + -ndo (Figure 2) shows a gradual increase in Spanish, there is a sharp increase in the analogous ta/tabata + verb + -ndo in PP after 1920. This increase is important for two reasons. First, the rate of -ndo progressives in PP becomes dramatically greater than in Spanish, suggesting that something besides Spanish interference is influencing this rate. Second, the increase comes only after contact with English began in 1915. English progressives take the form copula + verb + -ing, which is analogous to the PP ta/tabata + verb + -ndo. English progressives are used much more frequently than Spanish ones and PP ones at the time contact was introduced, and so could cause the frequency of -ndo progressives in PP to increase if speakers relate the English and PP constructions (Prince 1988, Silva-Corvalán 1991).

    In contrast, Figure 3 shows that the decline in rate of -ndo in absolutive clauses is about the same in Spanish and PP. This use of -ndo does not appear to be influenced by English.


    Figure 2. Comparison of ta/tabata, estar + -ndo in PP, Venezuelan Spanish


    Figure 3. Absolutive -ndo in PP, Venezuelan, and Tejano Spanish

    7.3  Observations from the 20th Century Texts

    7.3.1  The Integration of -ndo into the Grammar of Papiamentu

    The PP periphrastic progressive construction is composed of a preverbal marker (ta or tabata) and a verb marked by -ndo. Certain verbs cannot be marked by ta, some are optionally marked by ta, and others must be marked by ta (or another preverbal marker). Only verbs that can be marked by ta can appear in a periphrastic progressive in PP. Thus, a PP stative verb that takes ta could be found with -ndo even though such a case would be ungrammatical in Spanish. Finding such cases should be possible since ta and esta are not exactly equal. Though they are rare, we do find them (18, 19). In these examples, PP allows stative live and keep to take -ndo, while Spanish would express these with the present indicative.

    (18)      Konsiderando ku nos ta bibando den un mundo kaminda notisia propaganda material i pelikula i novela konstantamente ta influensha nos mente...

      consider-GER that we IMP live-GER in a world where news, propaganda and movie and novel constantly IMP influence our thought... (96:14)
      Considering that we live in a world where news, propaganda, movies, and books constantly influence our thoughts?
    (19)      A medida ku sentro di Otrobanda ta kedando kla, Selikor,...mester kuminsá krea e sirkunstansha pa nos sentro di siudat por haña un bisto mas higieniko i keda limpi permanentemente
      while center of Otrobanda IMP keep-GER clean, Selikor, ...should begin create the circumstance for our center of city be able become obvious more hygenic and keep clean permanently
      While the center of Otrobanda is keeping clean, Selikor ...should begin creating a way for our downtown to become noticeably cleaner and keep clean permanently.  (31:154)
    7.3.2 The Influence of English

    Further evidence of borrowing from English is found in the form of nonce-borrowings of surface word strings that could only have come from English (20). Spanish uses an infinitive here.

    (20)      Pero aseptando e chens pa bai pariba ta grandi.

      but accept-GER the risk of go there COP great.
      *Pero aceptando la oportunidad para ir ariba está grande.
      But accepting the risk of going there [to jail] is a big deal.
    7.4   Analysis and Discussion

    The -ndo affix was probably introduced by Spanish-dominant bilinguals, who used the form when speaking and writing PP. They only added the morpheme to Iberian verbs. They would have used the affix as in the most common Spanish usage, absolutive clauses, and in the speech styles and genres in which it was typically used in Spanish.

    Both the irregular gerund forms reported in the literature and the texts in which they appear support this hypothesis. First, recall that stative verbs in Spanish can take -ndo in absolutive clauses, but not in progressive cases. The irregular forms found are all statives-?to be?, ?to have?, ?to be able to?, ?to want?, ?to believe?. Since none of these can take progressive -ndo in Spanish, the Spanish speakers who first used the affix in PP would not have used them in the context of a periphrastic progressive. Thus, -ndo was clearly used in absolutive clauses at the time the irregular forms were first borrowed into PP. Second, Álvarez (1995) notes that -ndo absolutive clauses are used frequently in legal and administrative Spanish. The first attested -ndo in PP is found in a document with some legal weight-a letter signed by individuals swearing to the veracity of its contents-which is just where we would expect to find a token.

    The interference hypothesis is also supported by the work of Prince and Pintzuk (2000), who showed that bilingual speakers do not seem to have full control of items in closed lexical classes (i.e. functional items) in their L2, as evidenced by the fact that closed class items from the L1 surface when they speak their L2, but not vice versa. The -ndo affix discussed here is a functional morpheme in Spanish, so should be considered a closed class item. As such, we can expect it to have surfaced in the PP of L1 Spanish speakers.

    Spanish-dominant bilinguals would have held a high social status. Their introduction of the affix may have been copied by members of lower social classes for prestige purposes, in change from above fashion. It would also have been copied by the children of Spanish-dominant bilinguals, simply because that is the variety of PP that they heard. Once L1 PP speakers began to use -ndo productively, we can say that it changed from being an interference feature of the Spanish speakers to a borrowing for PP speakers.

    It is difficult to say for certain when -ndo first began to be used by PP speakers, but we can use existing information to pinpoint the time frame. Integration had already occurred by the time of the publication of the Curaçao newspaper (1871). The first example is found in 1803, so we can suppose that integration occurred some time in the first half of the 19th century.The affix has been in use in PP since 1803 and PP speakers have been bilingual in Spanish and Dutch since before that time. Dutch has no progressive morphology. In Spanish, -ndo has both gerundive and progressive functions, but periphrastic progressive constructions do not appear in PP until the 20th century.

    In 1915, English entered the contact picture with the introduction of the petroleum industry in Curaçao. In cases of Spanish-English contact, it has been reported that English dominant bilinguals overuse the Spanish periphrastic progressive (e.g. Floyd 1978, Klein 1980). At the time when English speakers arrived on the islands, auxiliary + verb + -ndo was an infrequently used, though possible construction (15). I argue that PP-English bilinguals began to use the periphrastic progressive construction in PP after identifying it with the analogous English construction (copula + verb + -ndo) via the kind of surface string matching reported in Prince (1988) . This construction is more frequent in English than Spanish, and thus more salient. If English had not entered the contact picture, PP speakers may not have picked up the elusive Spanish periphrastic progressive.

    English-dominant bilinguals first used ta/tabata + verb + -ndo as a periphrastic progressive, but PP-dominant speakers (children of the English speakers and others) have now adopted the construction. Once speakers identified it with the English progressive construction, they began to use it more and more frequently, eventually surpassing the rate at which progressive -ndo is used in Spanish.

    8  Summary, Conclusions, and Future Work

    To summarize, when an inflectional morpheme is borrowed, its grammatical function may also be borrowed. Where a morpheme has more than one function in a language, it is not the case that all functions are borrowed with the morpheme. Here, English and Spanish have similar surface strings for the periphrastic progressive, but English uses this construction more frequently than Spanish, making it more salient. Contact with Spanish provided the morpheme, but surface string matching with the English progressive lead to the increase in use of periphrastic progressives, as indicated by the fact that this increase occurred only after English was introduced into the contact situation. Thus, the present state of -ndo in PP can only be explained by referencing the successive influences of Spanish and then English on the creole. The saliency of the borrowed item in the source language, time of contact, and nonce borrowings together explain how this borrowing took place.

Further work with -ndo is needed. Current data show that -ndo is used more frequently on Aruba and in newspapers (as opposed to other islands and genres), suggesting regional and possibly social and/or stylistic variation, but spoken data and ethnographic observations are needed to understand the variation. In addition, there is some debate in the literature about the degree to which the use of Spanish morphemes like -ndo is indicative of the overall hispanization of the creole. It is wrong to say that PP is undergoing heavy decreolization as there is no evidence of a continuum of lects from PP to Spanish. However, it is possible that there are regional, social, and/or stylistic varieties of the creole which are more or less ?hispanized?, in other words, which show more or less influence from Spanish. These can also be better understood through spoken data and ethnography.
Álvarez, Miriam (1995) Tipos de escrito III: Epistolar, administrativo y jurídico. Arco Libros, S.L.: Madrid.

Andersen, Roger W. (1990) Papiamentu Tense-Aspect, With Special Attention to Discourse. In Pidgin and Creole Tense-Mood-Aspect Systems, John Victor Singler, ed. Philadelphia:Benjamins.

Birmingham, John Calhoun (1971) The Papiamentu Language of Curaçao. University of Virginia, Dissertation.

Evertsz, N.J. (1898) Compendio de la Gramatica de Papiamento, ó sea método para aprender á hablarlo y á escribirlo en corto tiempo. Curazao: Tipografia Bethencourt:

Floyd, Mary Beth (1978) Verb Usage in Southwest Spanish: A Review. The Bilingual Review 5:1-2, 76-90.

Herrero, Gemma (1994) El Gerundio Independente. Estudios de Linguistica 10:165-178.

Howe, Kate (1994) Papiamentu Reader. Kensington, MD: Dunwoody Press.

King, Larry D. and Margarita Suñer (1980) The Meaning of the Progressive in Spanish and Portuguese. Bilingual Review 7:3 Sept-Dec, 222-238.

Klein, Flora (1980) A quantitative study of syntactic and pragmatic indications of change in the Spanish of bilinguals in the U.S. In Locating Language in Time and Space. New York: Academic Press. 69-82.

Kouwenberg,Silvia and Eric Murray (1994) Papiamentu. München: Lincom Europa.

Kouwenberg, Silvia and Peter Muysken (1994) Papiamentu. In Pidgin and Creoles: An Introduction, Jacques Arends, Pieter Muysken, and Norval Smith, eds. Philadelphia: Benjamins.

Lenz, Rodolfo (1928) El Papiamento: Lengua Criolla de Curazao. Santiago, Chile: Balcells & Co.

Maduro, A.J. (1953) Ensayo pa yega na un ortografia uniforme pa nos papiamentu. Curaçao: Prijs.

Martinez, Glenn (2000) A sociohistorical basis of grammatical simplification: The absolute construction in nineteenth-century Tejano narrative discourse. Language Variation and Change 12:3.

Maurer, Philipe (1998) El papiamentu de Curazao. In Américan negra: panorámica actual de los estudios lingüisticos sobre variedades hispanas, portuguesas y criollas, Matthias Perl and Armin Schwegler, eds, with Gerardo Lorenzino. Frankfurt am Maim: Verveut Verlag.

Maurer, Philipe (1986) Les modifications temporelles et modales du verbe dans le papiamento de Curaçao. Hamburg: Helmut Buske Verlag.

Munteanu, Dan (1996) El Papiamento, Lengua Criolla Hispánica. Madrid: Editorial Gredos, S.A.

Ozete, Oscar (1983) On the So-Called Spanish Gerund/Participle. Hispania 66:1 Mar, 75-83.

Prince, Ellen (1988) On Pragmatic Change: The Borrowing of Discourse Functions. In Journal of Pragmatics, 12, 5-6, Dec, 505-518.

Prince, Ellen and Susan Pintzuk (2000) Bilingual code-switching and the open/ closed class distinction. In Penn Working Papers in Linguistics, 6.3, 237-258.

Römer, René (1998) Ethnicity and Social Change in Curaçao. In The White Minority in the Caribbean, Howard Johnson and Karl Watson, eds. Kingston: Ian Randle Publishers.

Sanchez, Tara (2000) The Papiamentu Progressive: Syntactic Conditioning and Dialect Variation. University of Pennsylvania, MA Thesis.

Sankoff, Gillian (2001) Linguistic Outcomes of Language Contact. In Handbook of Sociolinguistics, Chambers, Trudgill, Schilling-Estes, eds. Basil Blackwell.

Silva-Corvalan, Carmen (1991) On the Permeability of Grammars: Evidence from Spanish and English Contact. In Linguistic Perspectives on the Romance Languages; Ashby, Mithun, Perissinotto, Raposo, eds. Amsterdam: Benjamins.

Teza, Emilo (1863) Il dialetto curassese. Pisa. pp.542-552.

Torres Cacoullos, Rena (2000) Grammaticization, Synchronic Variation, and Language Contact: A study of Spanish progressive -ndo constructions. Philadelphia:Benjamins.

Torres Cacoullos, Rena (1999) Variation and Grammaticization in Progressives: Spanish -NDO Constructions. In Studies in Language 23:1, 25-59.

Wattman, Francine Harriet (1953) Papiamentu Morphology and Syntax. MA Thesis: Cornell University.

Zdenek, Joseph (1972) Another Look at the Progressive. Hispania 55:3 498-9.

Department of Linguistics
University of Pennsylvania
619 Williams Hall
Philadelphia, PA 19104