21.1 Introduction

Within the field of contact linguistics and bilingualism studies, the Spanish language holds a prominent place. It is spoken in numerous regions across several continents and serves as an official language in 21 countries. It is also used by a number of minority communities, either as a native language or as a heritage language, and is rapidly growing worldwide as an instructed second language. As the field of contact linguistics continues to investigate the role of contact in language change, language maintenance, and language shift, among other key issues, the wide array of situations that involve Spanish have proven to be a valuable laboratory for advancing research into the processes and outcomes of language contact. The different varieties of Spanish existing in the world today offer a wealth of data and opportunities for research in bilingualism and contact linguistics in general. In this chapter, rather than offering a geographical mapping of bilingual communities and contact zones in the Spanish-speaking world, I will follow a thematic division that relies on findings regarding the major issues in contact linguistics that are carried out using Spanish data and the main contact phenomena that obtain across the Spanish-speaking world.

In Section 21.2, I describe the major sources of lexical borrowing into Spanish. The focus is on borrowing from Arabic, the indigenous languages of the Americas, and English. Section 21.3 describes cases of structural convergence in situations of contact between Spanish and other languages at the phonological and morphosyntactic levels. In Section 21.4, I examine the existence of Spanish-based contact varieties, including both creoles and non-creolized varieties. Section 21.5 analyzes the phenomenon of code-switching, with a special focus on English/Spanish code-switching. Section 21.6 presents two major macro-sociolinguistic issues as they pertain to Spanish: language maintenance and language shift. Finally, Section 21.7 summarizes the status of Spanish as a world language and offers an outlook into the future of bilingualism in the Spanish-speaking world.

21.2 Lexical Borrowing

Lexical borrowing is undoubtedly the most common outcome of language contact. It is a salient phenomenon – of which speakers are themselves often aware – and consists in the incorporation of words from other languages with, although not always, some degree of adaptation into the recipient language.Footnote ¹ Lexical borrowing is also a reliable tool for assessing the role of language contact in the historical development of a given linguistic variety. In the case of Spanish, contact with other languages is a significant factor in the building of its lexical stock from early on. As the diglossic situation of Classical and Vulgar Latin came to an end in the Middle Ages, Castilian was in contact with other Romance varieties, including Mozarabic, and most notably with Arabic. The latter had been introduced into the Iberian peninsula in 711 CE and was maintained in active use there for over eight centuries. Even with the completion of the Reconquista in 1492, some contact with Arabic continued, as bilingual Arabic speakers did not totally disappear from mainland Spain until the Great Expulsion of the Moriscos that was ordered by Philip III of Spain in 1609. It is known that during the Inquisition some Moriscos, especially women, needed Arabic translators (Galmés de Fuentes Reference Galmés de Fuentes1983:28) and that Christian missionaries attempting to evangelize Moriscos in the mountains around Granada had to rely on Arabic well into the 16th century (Martínez Ruiz Reference Martínez Ruiz, Aguadé, Corriente and Marugán1994:142).

Before the systematic purging of a large part of Arabic loanwords, Spanish in the 15th and 16th centuries counted Arabic as its second lexifying language (Lapesa Reference Lapesa1981:133), proof of the high degree of bilingualism that existed among Mozarabs and, later on, among Moriscos. Indeed, Menéndez Pidal (Reference Menéndez Pidal1986:502) states that the most important outcome of the contact of Castilian with Mozarabic was lexical borrowing. The Diccionario de la lengua española (DRAE; Real Academia Española 2001) indicates that the number of Arabic loanwords in Spanish is around 1,200 words, not counting derivations and toponyms, although Lapesa (Reference Lapesa1981:133) estimates the total number of all types of Arabisms in Spanish to be around 4,000 items. What is significant about Arabic loanwords in Spanish is not only the number of loans that persist in use today but also that some of them are extremely common words with relatively few lexical competitive variants, such as aceite ‘oil,’ almohada ‘pillow,’ and alcalde ‘mayor.’ Also significant is the high number of loans that refer to agriculture, as many of the Moriscos who stayed on after the Reconquista continued to work in this field (e.g. acequia ‘waterway,’ alberca ‘pond used for irrigation,’ and almazara ‘olive mill’). Finally, an interesting detail about Arabic loans in Spanish is the number of super-loans that spread via Spanish to other European languages including English (e.g. algebra ‘algebra,’ algoritmo ‘algorithm,’ alcohol ‘alcohol’). There are a few function words, such as ojalá ‘hopefully’ and hasta ‘until,’ that originate from Arabic, as does the derivational suffix -í, as in alfonsí ‘Alfonsine’ and marroquí ‘Moroccan,’ but Arabic influence on Spanish outside of lexical borrowing has not been solidly confirmed.

Similarly, Spanish borrowed extensively from indigenous Latin American languages and is responsible for the introduction of a considerable number of Amerindian lexical items into other languages. Indigenous loanwords vary from one variety of Latin American Spanish to another, but usually range from a few hundred to a few thousand words depending on the dialect zone (Lope Blanch Reference Lope, Blanch, Sánchez Romeralo and Poulussen1967:369). Spanish adopted several cultural borrowings that include words such as canoa and cacique from the Taíno language, chocolate and tomate from Nahuatl, choclo ‘corn,’ and carpa ‘tent’ from Quechua, among many other examples. Indeed, a significant part of the variation at the lexical level between different Latin American Spanish dialect groups can be attributed to contact with a specific indigenous language as in the case of words used for “corn” and the different products derived from it. Of course, in the other direction, the lexical influence of Spanish on Amerindian languages of Latin America still in use is much more significant and often serves as a reliable indicator of the degree of endangerment of these languages. For example, in the case of the Copala Triqui language of Mexico, Scipione (Reference Scipione2011) has shown that there are more than 1,000 Spanish loanwords that cover a wide range of semantic fields and include examples such as soldado ‘soldier,’ multa ‘ticket,’ and cultura ‘culture.’ While earlier loans are fully adapted to Triqui, bilingual speakers increasingly tend to maintain the phonological integrity of the Spanish words and even reanalyze established loans to restore their Spanish form. In the case of Nahuatl, studies have shown that lexical borrowing from Spanish has led to a process of relexification where up to 40 percent of its vocabulary is sourced from Spanish (Hill and Hill Reference Hill and Hill1977:62).

An additional source of lexical influence during the formation of Latin American Spanish has been the African languages, especially visible in the case of Caribbean Spanish varieties, with words such as guineo ‘banana’ and ñame ‘yam’ (Lipski Reference Lipski and Studerus1987:33). Finally, lexical influence of other European languages is also apparent, mainly in Southern Cone Spanish. For instance, the fact that Lunfardo is recognized as a distinct variety, partly for its Italian influence, is an indicator of the degree of contact between Spanish and other immigrant languages (Lorenzino Reference Lorenzino2014).

A third major wave of lexical items entered Spanish through its direct and indirect contact with English. In addition to frequent words such as fútbol, jersey, and detective, modern age advances in science and technology have contributed English loanwords to Spanish. The DRAE now contains items such as parking, USB, and blog. While the dictionary officially recognizes some 600 words as originating from English, more direct contact has resulted in an even higher number of loans in bilingual communities, especially in US Spanish (Pfaff Reference Pfaff1979; Smead Reference Smead1998; Lipski Reference Lipski2008). Adapted forms of English words such as high school, truck, and roof are examples that are often mentioned as among the most frequent loanwords used by different Hispanic communities in the United States.

In sum, contact with other languages has led to the incorporation of hundreds of loanwords from other languages into Spanish, reflecting the different situations of bilingualism that the Spanish language has had throughout its history and into the present. Both the need to fill a lexical gap, as in the case of fauna and flora words borrowed from Amerindian languages, and prestige, as in the case of contact with English now or Arabic in Muslim Spain, are factors for lexical incorporation (Weinreich Reference Weinreich1963). At the same time, equally important is the fact that Spanish is also the source language for hundreds of loanwords in indigenous languages. The wide-ranging diversity of situations has led to different types of lexical influence and associated phenomena that include lexical borrowing, lexical purging, and relexification.

21.3 Structural Innovation and Convergence

In cases of intense language contact, manifested through high levels of bilingualism and substantial lexical borrowing, it is not uncommon for structural convergence and transfer to occur between the different languages in contact (Thomason and Kaufman Reference Thomason and Kaufman1988). Convergence refers to cases where a certain feature acquires additional functions or increases its frequency as a result of contact (Heath Reference Heath1984).Footnote ² This process has been described by Silva-Corvalán (Reference Silva-Corvalán1994, Reference Silva-Corvalán2008) as “indirect transfer” and is different from the importation of features that were previously absent from the recipient language. The latter would be a case of structural borrowing – still a highly debated issue within the field of contact linguistics (Winford Reference Winford2003). Spanish as used in different bilingual communities offers many examples of indirect transfer, depending on the degree of bilingualism and the status of the languages involved. For example, in northern Morocco Spanish is acquired as a foreign language, often in a naturalistic setting, and shows in many cases the influence of native Arabic and Berber languages. This is especially evident in phonetic features such as vowel height – mid vowels tend to be raised since in Arabic there is no phonemic differentiation between mid and high vowels (e.g. tangerino > tang[i]rino) – and the realization of the palatal nasal as an apico-alveolar nasal plus palatal approximant consonant (España > Espa[nj]a), as well as in supra-segmental features including stress shift in cases of vowel reduction (Sayahi Reference Sayahi2006, Reference Sayahi and Díaz-Campos2011).

In Peninsular Spanish, among the most salient situations that present cases of convergence is that of Spanish in contact with Catalan, both in terms of the number of speakers involved and the historical antecedents that have led to a robust normalization of the use of Catalan in Catalonia, Valencia, and the Balearic Islands. Studies have shown that the influence of Catalan can be observed with regard to several features, for example intervocalic /s/ voicing and the devoicing of final /d/, higher rates of the use of the definite article before proper nouns, and variation in the use of deictic markers including the use of aquí for ahí and venir for ir to refer to the interlocutor’s location as opposed to the speaker’s location (Galindo Solé Reference Galindo Solé and Sayahi2003; Davidson Reference Davidson, Potowski and Bugel2015; Stokes Reference Stokes2015). Additionally, Blas Arroyo (Reference Blas and Arroyo2008, Reference Blas, Arroyo and Díaz-Campos2011) has shown that contact with Catalan is serving another purpose: that of slowing down some changes that are in progress in non-contact varieties. He argues that the maintenance of intervocalic /d/ (e.g. for forms ending with -ado: 74 percent for Valencian-dominant bilingual speakers vs. 37 percent for monolingual Spanish speakers) and higher rates of retention of the morphological future could be attributed to contact with Valencian Catalan. In the case of the latter, a rate of use (46 percent) that is much higher than that found elsewhere is attributed by Blas Arroyo (Reference Blas and Arroyo2007) to the use of an analytic form in Catalan to refer to actions in the preterit.Footnote ³

In Latin America, a major contact zone is the Andes region where Spanish is in contact with Quechua, in addition to other languages, and, as a result, shows instances of convergence (Austin et al. Reference Austin, Blume and Sánchez2015). Specifically, in the case of clitics, Klee and Caravedo (Reference Klee, Caravedo and Eddington2005) have discussed the use of leísmo and the archmorpheme lo for direct object regardless of gender in the speech of first- and second-generation Quechua migrants to Lima. Their results confirm the presence of leísmo in the speech of first-generation migrants and its transmission to members of the second generation, given that it is not an overtly stigmatized feature. They also find significantly high rates for the use of the archmorpheme lo to refer to plural masculine direct objects. Rates for the neutralization of the plural by first-generation migrants reach 64 percent while, for second-generation migrants, they reach 57 percent, which is significant especially if compared to speakers who are born in Lima to Limeño parents and who neutralize plural direct objects at the rate of 22 percent. More significant is the neutralization of feminine direct object pronouns: Klee and Caravedo (Reference Klee, Caravedo and Eddington2005) show that it reaches 76 percent in the case of first-generation migrants and 21 percent in the case of second-generation migrants, while the rate is only 2 percent in the case of native Limeño speakers. In another study, Klee et al. (Reference Klee, Tight and Caravedo2011) show that indigenous learners of Spanish in Lima use an OV word order more frequently than do L1 Spanish speakers. Both VO and OV are possible in Spanish, but influence from Quechua appears to increase the more marked use of an OV word order by bilingual speakers. Other instances of Quechua influence on Andean Spanish include the use of an evidential function for the present perfect, in contrast with the pluperfect that is used for a reportative function (Escobar Reference Escobar1997, Reference Escobar and Díaz-Campos2011), and the appearance of a double possessive marking in several Andean varieties (Clements Reference Clements2009).Footnote ⁴

A well-studied contact situation has been that of Spanish in contact with English in the United States. Although several scholars recognize that there are instances of convergence, the consensus is that US Spanish is not a restructured variety, and the claim that it represents a third stable new variety, popularly denoted Spanglish, is not accurate (Lipski Reference Lipski2008). Silva-Corvalán (Reference Silva-Corvalán1994, Reference Silva-Corvalán2001) describes several features that represent cases of convergence between Spanish and English including the elimination of the complementizer que ‘that’ in complement clauses, increased prenominal placement of adjectives, and the use of possessive adjectives instead of the definite article with inalienable nouns. Other pervasive features in US Spanish include the erosion of the subjunctive in favor of a more extended use of the indicative mood, and the extended use of the verb estar. Silva-Corvalán (Reference Silva-Corvalán2008) nevertheless argues that these features are already present in native vernacular varieties, as opposed to in the Academy-sanctioned standard Spanish, even though they become considerably more prominent in second- and third-generation bilinguals.

The case of the use of subject personal pronouns (SPPs) has in particular attracted a good deal of analysis. The fact that this is a conflict site between Spanish and English, as English generally requires overt pronouns while Spanish is a pro-drop language, has led to suggestions of a possible English influence on the rate of overt SPP use by speakers of Spanish in contact with English. While some scholars have argued that higher rates of overt pronoun use are not the result of influence from English (Cameron Reference Cameron1992; Flores-Ferrán Reference Flores-Ferrán2004), earlier studies had claimed the opposite (de Granda Reference de Granda1978; Navarro Tomás Reference Navarro Tomás1948). More recently, in a larger study, Otheguy et al. (Reference Otheguy, Zentella and Livert2007) and Otheguy and Zentella (Reference Otheguy and Zentella2012) have proved that competence in English in fact plays a role in a higher rate of subject pronoun use, as it sets apart New York-born speakers and those with higher levels of competence in English from recent arrivals. Otheguy et al. (Reference Otheguy, Zentella and Livert2007:779) conclude that: “there are, as predicted, positive correlations between rates of overt pronouns and years spent in NYC as well as English skills.” But here again, as with other cases discussed above, we have a feature already present in Spanish, and what we see are variable rates of its use as opposed to the incorporation of a totally new feature. It is important to keep in mind, then, that variations in the use of a feature in situations of language contact do not necessarily have as their source the other language. Additionally, Poplack and Levey (Reference Poplack, Levey, Auer and Schmidt2010:394) argue for a distinction between innovation, a possibly transient phenomenon, and change, which usually shows considerable diffusion and acceptability in the community.Footnote ⁵ An example of an item that has achieved diffusion in US Spanish is a form such as llamar para atrás ‘to call back’ (cf. standard Spanish llamar de vuelta/devolver la llamada (Otheguy Reference Otheguy, Roca and Lipski1993; Lipski Reference Lipski2008)).

21.4 Spanish-Based Contact Varieties

Contact between European languages and indigenous languages in different geographical regions across the globe, principally as a result of the European waves of colonization and slave trade, led to the appearance of several types of contact varieties, ranging from radical vernaculars to pidgins and creole languages. The status of Spanish-based creoles as prototypical creoles has been disputed in the literature with some scholars claiming their non-existence (McWhorter Reference McWhorter2000), while others provide significant evidence that Spanish-based creoles show similar grammatical features and processes of formation to other creoles (Clements Reference Clements2009 and references therein). Overall, studies have shown that there is a wide range of Spanish-based contact varieties, including creoles such as Palenquero and Chabacano, restructured vernaculars including Bozal Spanish and other Afro-Spanish varieties, and even intertwined systems such as Media Lengua.

In the case of creole languages, Palenquero has enjoyed a good deal of attention over the last few decades. Spoken in San Basilio de Palenque in Colombia, it has its origin in the speech of African slaves who escaped from Cartagena in the 17th century and remained isolated for a long period of time. Common creole structural features of Palenquero include a lack of gender marking and definiteness of noun phrases, and, at the verb phrase level, invariant verbs, as Palenquero relies on preverbal markers to express time, mood, and aspect (TMA), a feature common in creole languages. Increased access to Spanish and its use in Palenque is leading to a less natural inter-generational transmission of Palenquero, especially given the low number of its speakers (estimated by Schwegler (Reference Schwegler and Díaz-Campos2011) to range between 4,000 and 5,000). More recently, an increased awareness of the importance of Palenquero as an ethnic identity marker and its recognition as a valuable cultural product of community, national, and world heritage, has led to its introduction in the school system. Recent studies have shown that younger bilingual speakers present new features that are a result of contact with Spanish, such as the use of verb morphology and definite articles (Lipski Reference Lipski and Ortiz-López2011).

Another documented Spanish-based creole is Chabacano, spoken principally in and around Zamboanga City, and, less commonly, in other smaller communities in the Philippines. Chabacano emerged as a result of contact between Spanish, introduced in the Philippines as early as the 16th century, and indigenous Filipino languages such as Tagalog. These have had significant influence on the structure of Chabacano, including in its preference for a VSO word order (Lipski Reference Lipski and Jiménez2012). A different case is that of Papiamento, which combines grammatical features that are the result of Afro-Hispanic contact with other lexifying languages including Dutch, English, and Portuguese. Papiamento is spoken in the Caribbean islands of Curaçao, Aruba, and Bonaire; many of the grammatical features of its different varieties are shared with other Iberian-based creoles such as the use of pre-verbal TMA markers, including the marker ta (present in Palenquero and Chabacano (Andersen Reference Andersen and Singler1990:67)).

There also exist contact varieties of Spanish that represent the process of L2 acquisition rather than prototypical cases of creole genesis. These include Afro-Hispanic varieties such as the historical pidgin known as Bozal Spanish used by African slaves and possible post-Bozal varieties described by Lipski (Reference Lipski, Potowski and Cameron2007), Barlovento Spanish in Venezuela (Díaz and Clements Reference Díaz-Campos and Clancy Clements2008), Afro-Boliviano (Lipski Reference Lipski and Ortiz-López2011), and Chota Valley Spanish (Sessarego Reference Sessarego2014). The case of Media Lengua, spoken in the Ecuadorian Andes, is a particularly interesting case from a theoretical standpoint. Media Lengua is a variety formed by the superposition of Spanish lexicon onto Quechua grammar, used by rural migrants and their descendants as an intra-group code. The fact that about 90 percent of the lexicon comes from Spanish while the grammar remains that of Quechua (Muysken Reference Muysken, Highfield and Valdman1981; Gómez Rendón Reference Gómez and Rendón2008) has attracted considerable interest from scholars of mixed languages. In addition, the analysis of Media Lengua has contributed to our understanding of language contact resulting from rapid and massive population movements and the concomitant emergence of new mixed varieties, such as we find with several urban vernaculars in Africa (McLaughlin Reference McLaughlin2009).

Another example of a contact situation involving Spanish is that of Spanish in contact with Portuguese, whose outcome is generally referred to as Portuñol (Elizaincín Reference Elizaincín1992). Two Luso-Hispanic contact varieties have attracted interest: Fronterizo and Barranqueño. In the former, different degrees of mixing of Spanish and Portuguese in communities along the border between Brazil and Uruguay have resulted in Fronterizo (or Uruguayan Portuguese) as its major manifestation. Given the structural overlap between Spanish and Portuguese and the variable nature of the different contact systems grouped under Portuñol, no definitive lines have been drawn to account for clear structural distinctive features that could separate the different bilingual communities (Lipski Reference Lipski, Face and Klee2006:13). In the second case, the contact is between Spanish and Portuguese on the border between Spain and Portugal in the Iberian peninsula. Barranqueño, which emerged as a variety of Portuguese acquired by Spanish dominant speakers, is spoken by some 2,000 people in Barrancos (Clements et al. Reference Clements, Amaral and Luís2008). Among the most salient features accounted for as a direct influence of Spanish in Barranqueño, Clements et al. (Reference Clements, Amaral, Luís and Díaz-Campos2011) list the aspiration and deletion of final /-s/ and the deletion of final /-r/, both of which are common in the neighboring regions of Spain, and several phenomena related to the use and placement of pronouns including the doubling of indirect objects with full noun phrases and the preference for proclisis in contexts where Portuguese requires the use of enclisis with third-person indirect object clitics.

Finally, another case of language contact in the Spanish-speaking world is that involving Spanish/Limonese creole speakers in the province of Limón in Costa Rica, where we find an English-based creole in contact with Spanish. Limonese is a variety of the Jamaican creole, and its contact with Spanish is leading to substantial code-switching and influence from Spanish without the appearance of a distinct third variety. Spanish is the source language for frequent lone and multiword insertions in Limonese and for some morphosyntactic claques, including the predominant use of the “have + years” construction to express age as opposed to the “be + years” form (LaBoda Reference LaBoda2015).

In sum, the history of Spanish and its large-scale nativization by indigenous and creole populations has led to the creation of varieties that offer a window into different outcomes of intense language contact.

21.5 Code-Switching

Studies of code-switching between Spanish and other languages, especially between English and Spanish, have been among the most influential in the field. Early studies provided an initial description of bilingual speaker use of Spanish and English in the same communicative event (Gingràs Reference Gingràs and Bills1974; Timm Reference Timm1975; Pfaff Reference Pfaff1979, among others). These studies helped dissipate some of the misunderstandings about code-switching which even today continue to be perceived by non-specialists and users themselves as deviant behavior. They also established that, in bilingual communities, speakers frequently code-switch not randomly but according to a set of social, pragmatic, and grammatical constraints, and that the presence of code-switching may in fact serve as an indicator of the vitality of the languages in contact rather than of their mutual erosion. The occurrence of English/Spanish code-switching has been examined not only in oral conversation but also in other types of discourse, including in written and electronic communication (Callahan Reference Callahan2004; Toribio Reference Toribio and Díaz-Campos2011; Montes Alcalá Reference Montes Alcalá2015).

The variationist approach to code-switching introduced by Poplack (Reference Poplack1980) has provided proof not only of the quantitatively high frequency of code-switched occurrences in Hispanic communities, but also of the engagement by members of bilingual communities in different types of code-switching which often reflect the richness of their linguistic repertoire. A distinction is made between inter-sentential and intra-sentential code-switching. In the first case, switch points occur across sentence boundaries without major implications for the structure of either of the two languages (21.1).Footnote ⁶ In the second case, code-switching occurs within the same sentence but without violating the grammatical rules of either of the two languages involved (21.2).

1. (21.1) Está desconectado en este momento. I have a – I have a local number. Yo tengo teléfono en mi habitación si eso ayuda.

   ‘It is disconnected at the moment. … I have a phone in my room if it helps.’

1. (21.2) En la República Dominicana, oh it has to be in Spanish.

   ‘In the Dominican Republic, …’

Intra-sentential code-switching has been shown to be a good index of the degree of bilingualism: more balanced bilinguals tend to code-switch at the intra-sentential level. Lipski (Reference Lipski2014:41) has argued that code-switching also occurs in the speech of low-competency speakers, but that “in the case of low-fluency or semi-fluent bilinguals … not all language mixing may follow the ‘canonical’ code-switching trajectory” described above.

In the case of Spanish in contact with English, as with many other contact situations, lone noun insertion remains a very frequent phenomenon (21.3). Debates on how to interpret these – whether as instances of code-switching, of borrowing, or of nonce-borrowing – have been heated. While Myers-Scotton (Reference Myers-Scotton2002), for example, argues that non-established noun insertions are cases of code-switching, Poplack and Meechan (Reference Poplack and Meechan1998) have argued that these often behave like instances of borrowing.

1. (21.3) Yo trabajé en una fábrica donde hacían clothes de leather y jackets de leather. Yo peleaba mucho con el dueño.

   ‘I worked in a factory where they made leather clothes and leather jackets. I argued a lot with the owner.’

A particular focus within code-switching studies has been on the use of English and Spanish discourse markers, which include such frequent particles as “so,” “you know,” “entonces,” and “tú sabes,” among others (Aaron Reference Aaron2004; Lipski Reference Lipski, Sayahi and Westmoreland2005). Torres (Reference Torres2002) found that, within the Puerto Rican community in Brentwood (New York), speakers with varying levels of competence in both languages used a high number of English discourse markers when speaking Spanish, albeit to varying degrees and for different discourse functions. The English discourse markers “you know” and “so” are the ones most commonly used by speakers with different levels of bilingual competence, to the degree that Torres considers them to be borrowings integrated into Spanish discourse. At the same time, this does not cause the displacement of Spanish discourse markers, which continue to be used alongside English ones in an overlapping fashion.

In Spain, code-switching between Spanish and Catalan, Basque, and Galician has also been examined, with a special importance given to the emblematic use of minority languages and to the more common discourse functions carried out (Alvarez-Cáccamo Reference Álvarez-Cáccamo1990; Blas Arroyo Reference Blas and Arroyo1993). Code-switching is also common in other bilingual communities across the Spanish-speaking world including, to give an example of lesser-known cases, in northern Morocco, Ceuta, and Melilla, where code-switching between Spanish, Arabic, and Berber is frequent, and the British territory of Gibraltar, where, as Moyer (Reference Moyer1992) has shown, competence in both English and Spanish again results in frequent code-switching.

21.6 Language Maintenance and Language Shift

The particular ecology of a language contact situation (Mufwene Reference Mufwene2001) determines the domains of use of each of the languages involved and leads to the emergence of different phenomena. Contact can lead to language maintenance in cases of balanced bilingualism and a strong ethnolinguistic vitality of both languages, or to language shift in cases of an interruption in the natural transmission of the socially-subordinate language. Spanish as a world language has enjoyed a long history of expansion that in many cases has led to the restricted use of indigenous languages or even to their death.Footnote ⁷ In fact, Spanish is the dominant language in the majority of the cases where it is in contact with other languages. In Latin America, it has advanced quickly at the expense of Amerindian languages, with very few cases of balanced bilingualism such as that which we see in Paraguay (Gynan Reference Gynan, Sayahi and Westmoreland2005).

The other side of the coin is when Spanish is a minority language. There are cases where Spanish was not nativized and did not displace an indigenous language, but where it has the status of a foreign language coexisting with a local native language. In North Africa, Equatorial Guinea, and the Philippines, Spanish as a minority language enjoys considerable prestige, leading to its maintenance as a second language. In northern Morocco, about 80 percent of students in the 11 official Spanish education centers which still exist are Moroccan, and the Instituto Cervantes in many Moroccan cities often has long waiting lists for its classes. Spanish is perceived as a language that can provide educational and economic opportunities in addition to access to cultural and audio-visual products. In the disputed territories of the Western Sahara, where Spain was an occupying power until 1975, Spanish has been adopted as one of the official languages of the Sahrawis and is one used for education in the refugee camps. Equatorial Guinea, the only African country where Spanish is an official language, is a typical case of postcolonial language policies, as the colonial language is the language of education and administration, and even the language used for literary production.

In the United States, the status of Spanish as a minority language varies by region and by historical period (Amastae and Olivares Reference Amastae and Elías-Olivares1982; Lipski Reference Lipski2008). In addition to Puerto Rico and the regions with a traditionally heavy Hispanic presence, other areas have seen an increase in the number of Spanish speakers in recent years, including the Midwest (Escobar and Potowski Reference Escobar and Potowski2015). Moreover, Spanish is among the heritage languages in the US that is best transmitted to the second generation. According to the Pew Research Center (2013), eight out of every ten second-generation Hispanics claim to speak Spanish, which is much higher than is the case with Asian Americans (only four out of ten second-generation Asian Americans claim to speak their heritage language). However, as has been frequently argued, this is not enough to secure Spanish as a fully accepted language in the United States. Misconceptions about the threat of Spanish are abundant, and language shift continues to happen across generations (Potowski Reference Potowski, Wiley, Peyton, Christian, Moore and Liu2014; Escobar and Potowski Reference Escobar and Potowski2015). While new waves of immigration continue to sustain the presence of Spanish in the United States, in traditional communities where Spanish had been maintained in the past – such as in the US Southwest and enclaves like St Bernard Parish in Louisiana – we see Spanish disappearing quickly or being assimilated to other varieties (Bills and Vigil Reference Bills and Vigil1999; Torres Cacoullos and Aaron Reference Torres Cacoullos and Aaron2003).

21.7 Summary

The Spanish language today is used in very diverse settings that range from its status as a native language in Spain and Spanish-speaking Latin America, an official language in Equatorial Guinea, an “unofficial” second language in northern Morocco and the Philippines, and a language with variable status as a heritage language in the United States and other destination countries for Latin American and Spanish immigration. This diversity of situations has promoted the appearance of different language contact phenomena. At the same time, the continued unity of the language, in terms of mutual intelligibility among Spanish speakers regardless of native dialect, is remarkable. Despite its notable lexical and structural variability, Spanish continues to be considered, and rightly so, as a single language. In some 500 years Spanish has become a language used by many types of speakers in very different places: This supports its status as a global language offering a wide range of research opportunities for a better understanding of the nature of bilingualism and the processes and outcomes of language contact.

Moving forward, it appears that – in Spain – a greater recognition and revitalization of minority languages, with a concomitant increase in their normalization and use in a wider set of domains, will lead to increases in language contact. In North Africa, the two autonomous cities of Ceuta and Melilla continue to sustain the use of Spanish in northern Morocco and offer interesting cases of Spanish in contact with Arabic and Berber. In Latin America, as the spread of education and urbanization continues to promote the imposition of Spanish, there are increasing efforts to maintain and teach the indigenous languages. Although these efforts face an uphill battle, given the overwhelming symbolic and socioeconomic power of Spanish, they promise to at least increase awareness at the level of both affected communities and policy makers. Finally, in the United States, demographic changes and the long history of Spanish in the country support a continuing presence of this language well into the future.

In conclusion, a priority of research programs into bilingualism in the Spanish-speaking world and Spanish in contact with other languages should be to provide much-needed data on the many contact situations involving the hitherto far less researched languages with rapidly dwindling numbers of speakers. This can help achieve two things at once: first, to increase awareness about the situation of minority languages in general and to further document the processes of language maintenance and language shift; and, second, to cast additional light on the mechanisms of language variation and language change in situations of bilingualism, which is of interest to the field of linguistics in general given that the majority of people in the world today speak more than one language. Additionally, massive population movement and migratory trends continue to represent an excellent opportunity for specialists in contact linguistics, including those working on Spanish. Finally, in the age of “big data” and increasingly dominant electronic modes of communication, research on lexical borrowing and contact-induced structural change from and into Spanish should follow suit by becoming even more quantitative and usage-based.