Some remarks on Spanish in the bilingual world

1 Introduction

The collection of articles of which the one article forms a part highlights and analyzes the many facets of Spanish in the modern world, with a special emphasis on variation and change. An approach that celebrates variation in Spanish is important for a number of reasons. The Platonic idea of language as a shared medium is beautiful as it is, but it is hard to uphold for a language spoken by 500 million people – clearly, they do not all share the same vernacular. A speaker of Mexican Spanish operates in a very different linguistic and cultural world than an average speaker living in Barcelona, even though they can understand each other. A recognition of variation and mechanisms that underlie this variation is extremely important. Equally valuable is the understanding that the factors behind variation in Spanish change over time, and the status and role of the language in the 2020s may be quite different from its status and role a generation or two ago. Hence, a continuing look into varieties of Spanish is very much needed, as it informs the rich knowledge base of Spanish linguistics as well as general dialectology and sociolinguistics. And finally, since Spanish is so widely spoken, that creates a sense of stability in its maintenance; after all, what can happen to a language spoken by so many millions of people? In other words, since Spanish is everywhere, it is not under any threat of change or extinction. The situation on the ground is of course more nuanced, as dialects of Spanish vary, some varieties emerge, some are on the wane, and social conditions for the existence or even recognition of some varieties vary vastly. The sense of stability is also false, since quite a few speakers of Spanish are bilingual, some of them subtractive, that is, heritage speakers. Understanding their language structure and use is of utmost importance to our understanding of different aspects of bilingualism in the modern world.^[1] After all, if a language as apparently indestructible as Spanish can become vulnerable, what does that tell us about smaller languages, ones with fewer speakers, ones with few or no homeland varieties? The vulnerability of Spanish in bilingual situations is thus very instructive for bilingual studies more generally. It is this bilingual use that I will briefly discuss below. My core goal in the following sections is to raise new questions that can be asked, based on the rich knowledge base accumulated by research on Spanish throughout the world, and to bring together subdisciplines that do not talk to each other enough: sociolinguistics, dialectology, applied linguistics, and analytical linguistics. The theories and methodologies may vary, but all these subfields share custody of Spanish in the world.

2 Bilingualism as the norm

In many textbooks on language sciences and language policy, monolingualism is assumed to be the norm, often without much discussion. Yet, the world is filled with bilinguals and multilinguals, and their social and linguistic biographies deserve to be acknowledged, celebrated, and studied. In studying the rich landscape of bilingualism, we can all come closer to understanding the concept of native speakerhood, linguistic development over a person’s lifespan, inter-generational transmission, and the complex interactions between language, identity, and culture. Bringing together these facets of human existence is important for our conceptions of the individual and the ambient community.

Quite a few speakers of Spanish are dominant in Spanish (and have another language in their bilingual inventory), and for many years, researchers have compared such speakers with L2 (second language) learners of the language. In doing so, such researchers have mainly assumed, if tacitly, a monolingual Spanish speaker as their standard of comparison. But what about a child in Mexico who speaks a Zapotec language at home but is schooled in Spanish and becomes dominant in Spanish? How much do we gain from comparing such people with L2 speakers of Spanish?

In general, since Spanish is so widespread and so widely taught in classrooms all over the world, the interest in Spanish as a second (foreign) language is quite understandable (consider Geeslin [2014, 2018] for a comprehensive overview). More recently, another large group of Spanish speakers has emerged as an important cohort in research on Spanish: subtractive bilinguals, also known as heritage speakers, that is, speakers who are more comfortable and fluent in the dominant language of their society. One of the more general, and widely used, definitions of heritage language is presented in the following quote:

A language qualifies as a heritage language if it is a language spoken at home or otherwise readily available to young children, and crucially this language is not a dominant language of the larger (national) society […] Differently [from monolingual acquisition], there is the possibility that quantitative and qualitative differences in heritage language input, influence of the societal majority language and differences in literacy and formal education, can result in what on the surface seems to be arrested development of the heritage language or attrition in adult bilingual knowledge. (Rothman 2009: 156)

For heritage varieties of Spanish that are not “mainstream” and that are spoken by heritage speakers, one of the vexing issues is whether or not speakers of such varieties should be recognized as native speakers. This question, one that researchers face, is reinforced by the feeling of insecurity or even shame experienced by heritage speakers who are more dominant in the other language of the bilingual dyad. These feelings of deficiency or inadequacy, combined with societal pressures to use the dominant language, can actually contribute to speakers abandoning their home language.

Only recently have researchers of bilingualism come to the recognition that differences between heritage and other speakers of the same language should be assessed in positive terms and viewed as the basis for further development. This development can take many forms, among them the acquisition of literacy skills in the heritage language, the incorporation of the structure and vocabulary of formal registers of the language, or the recognition of the heritage language as a sign of richer identity. In such contexts, the notion that one’s home language is their native language removes the stigma associated with not knowing the language “very well”. Yet another step in this direction has to do with the recognition of different types of bilingualism. It is common (and encoded in societal expectations) that a bilingual is viewed as two monolinguals in one; the more we all agree that mastery and use of languages for a bilingual is different from that of monolinguals, the better the social conditions on accepting and promoting bilingualism will become. The unique ways bilingual speakers innovate with their home language (including what is sometimes called code switching or mixing) are indicative of their sophisticated knowledge of both language systems – not a sign of lack of knowledge or lack of native speaker competence.

The recognition of heritage speakers’ competence has taken some time, and it has also gone through terminological pains. Original studies of such bilinguals characterized their language as “incompletely acquired” (e.g. Montrul 2008, 2009, 2016; Polinsky 2006): an awkward terminological choice designed to emphasize the differences between heritage language and the monolingual baseline or the language that these speakers receive as their input, which is typically the language of first-generation speakers of Spanish who emigrated from a Spanish-speaking country. The characterization of heritage languages in terms of “incomplete acquisition” has received serious pushback from researchers who focus both on sociolinguistic and grammatical properties of heritage languages (see especially Kupisch and Rothman 2018; Otheguy 2016; Putnam and Sánchez 2013).

As one of the users of the original term, Montrul noted that the connotation of the widely used term “incomplete acquisition” may in fact discomfit researchers and practitioners, as incompleteness may imply deficiency. She went on to clarify that:

[W]e can speak about heritage language grammars in general as having certain structural characteristics that differ from those of fully fluent bilinguals and monolinguals of the same variety. But these characteristics are most likely due to an interruption in the normal transmission of the language in childhood, rather than to exposure to a different language variety spoken by parents and siblings or their immediate network of heritage speakers. (Montrul 2011: iii–iv)

Similar to Montrul, but with rather different terminology, Lynch (2013) suggested that the term “U.S. Spanish”, in the sense of a “native” variety beyond first-generation adult immigrants, is highly questionable for two main reasons: first, a lack of community-based norms of usage owed fundamentally to the lack of generational continuity of Spanish in the U.S. (cf. Otheguy and Zentella 2012); and second, readily apparent patterns of sociolinguistic discontinuity according to the variationist model, an issue I will return to in Section 4.

As already mentioned, as the field of Spanish in bilingual contexts develops, there is a growing consensus that many (but not all) heritage speakers should be counted under the rubric of native speakers and are in fact more similar to L1 (first language) speakers than to L2 (second language) learners. At the same time, it is important to determine where the boundary between native speakerhood and less-than-native competence and use. As we consider this question, an important and largely overlooked group deserves more attention: the overhearers (Au et al. 2002, 2008): individuals who grow up hearing a language in their environment, but who do not speak it and may have not even been addressed in that language. The recognition of such speakers highlights naturally occurring subtypes of childhood exposure to language. In their studies of adult English-speaking overhearers of Spanish and Korean, Au and colleagues (Au et al. 2002, 2008) showed convincingly that these speakers had a modest advantage over L2 learners in sound perception and sound production, but beyond that, were closer to L2 learners. These observations raise the general question of the amount of exposure: what is the lower limit needed of time spent around native speakers to actually learn a heritage language? Understanding the amount and regularity of exposure could enrich our understanding of the impact of childhood language experience. At this point, we do not yet have a way of quantifying and operationalizing exposure, so it behooves us to pay attention to this question. Sociolinguistic and demographic data are particularly valuable in answering it, and Spanish, as is commonly the case, is going to be a major testbed of the new approaches.

As an interim summary, our understanding of native speakerhood has gradually grown more nuanced, in large part due to research on Spanish (Kupisch and Rothman 2018; Lynch 2022; Otheguy 2016; Potowski 2018; Potowski and Bugel 2015, among others). Until now, I have assumed the static approach to native speakerhood, where the linguistic repertoire of an individual is viewed as stable across their lifespan. This is clearly a simplification. In the next section, I will turn to different scenarios where the balance of languages controlled by a given speaker shifts throughout their life.

3 Reversals of dominance

If we now leave the static notion of bilingualism, we can immediately appreciate another facet of Spanish in multilingual and bilingual contexts: the relative status of the languages controlled by a bilingual and changes in this status across a person’s lifespan. Here, I will discuss two such scenarios, although some others may also be considered.

In the first scenario, a person’s heritage language undergoes change to become their dominant language, which typically happens if that individual moves to a country where their home language is dominant. This scenario is known as the situation of returnees. Returnees are bilingual speakers who, after spending their childhood, adolescence, or early adulthood in an immigrant setting, return to their parents’ country of origin (Flores 2010; Flores and Snape 2021, among others), or in the case of Spanish, to any country where Spanish is dominant. Thus, a speaker of Northern Mexican Spanish who grew up in Arizona may move to Peru and still find Spanish – but not their home variety of Spanish – the dominant language of their new country of residence. What characterizes returnees is the reversal of the dominant environmental language after their return (Flores and Snape 2021), albeit this dominant language may be of a variety different from what they were exposed to in the home. Montrul (2016) has proposed the term heritage language reversal to encompass the linguistic changes that occur in both directions in situations of return of heritage speakers to their homeland. She writes:

[…] return [to the homeland – MP] marks a reversal of the sociopolitical and functional status of the languages: what used to be the heritage language now becomes the societal majority language, and the L2 learned in the host country, even when it is not the primary language of the parents, becomes a heritage language of the child in the new environment. It would be considered a heritage language because in addition to having high proficiency in the language, the child still has a connection with the culture. The language and the culture were part of the child’s upbringing and socialization. (Montrul 2016: 15)

The reversal raises interesting questions that can offer profound insights not only into the processes of language acquisition and change, but also into social and educational implications of remigration, identity shifts, and pedagogical challenges that accompany the reversal (see Flores and Snape [2021] for the structural aspects of reversal, and multiple chapters in Potowski [2018], for the social and educational aspect, with a special emphasis on Spanish). In the case of Spanish, the move to a Spanish-speaking country different from the homeland of a given family is particularly intriguing: does such a move count as a heritage language reversal, or is it more of a case of adding a new dialect, an L3 (third language) of sorts (see Polinsky [2015] for more discussion)?

The second scenario is in some ways the opposite of what I just described, and it has to do with the downgrading of Spanish from the dominant language of the society that a family leaves behind. Spanish is the dominant language in many countries where other languages are spoken – consider Quechua in the Andes or Mayan and Zapotecan languages in Mesoamerica. The majority of Quechuan, Mayan, or Zapotecan speakers are also speakers of Spanish, and for most of them, Spanish is the language of social mobility. Thus, in the bilingual dyad Spanish/indigenous language, it is the latter that is likely to become heritage and subject to effects of subtractive bilingualism. However, if such bilingual speakers move to North America, English takes the prize as the dominant, societally valuable language, and Spanish may become less important in the immigrant family, hence moving to the role of another heritage language. For a second generation of such immigrants, two languages may become heritage (Spanish and the indigenous language of the homeland), which in turn creates intriguing patterns of identity and language use, patterns that little is known about.

4 Speakers are not all the same: addressing variation

Spanish holds great appeal to sociolinguists and dialectologists who can study similarities and differences in its linguistic properties, correlating those with geographical distribution, demographic changes, societal characteristics, and the status of Spanish in a given country. Advances in corpus linguistics and data mining from social media apps allow us to track changes in Spanish in real time; at the same time, they present strong evidence of the range of variation in modern Spanish – across geographical areas and also across classes of speakers. The range of linguistic variation manifest in bilinguals is vast, and that has compelled researchers to capture this variation not as a collection of discrete points but as the “bilingual continuum” (Mackey 1968; Polinsky and Kagan 2007; Silva-Corvalán 1994). Bilinguals in general, and heritage speakers in particular, project a general impression of native-like sound production and show great linguistic creativity that can actually mask some of the gaps in their knowledge (see Fairclough [2011], Lynch and Polinsky [2018], and Montrul and Foote [2014] for more discussion).

Typically, however, bilinguals show decreased confidence in interactions with educated native speakers or those who have more experience with formal registers. In this dimension of language use, some bilingual speakers describe their own abilities as wanting, inadequate, or inappropriate (Callahan 2010; Lynch and Potowski 2014; Potowski 2002; Urciuoli 2008). A lack of confidence in one’s linguistic ability may lead this type of bilingual to shy away from opportunities to speak the language altogether, which, in turn, creates a downward spiral (Abdi 2011; Goble 2016).^[2] This also leads to greater variation among speakers who we contend to qualify as native (see Section 2 above).

How can such variance be handled? In sociolinguistics, Labov’s framework for the study of language variation has been widely used. However, it is a crucial assumption of Labov’s framework that “the speech community [is] an aggregate of speakers who share a set of norms for the interpretation of language, as reflected in their treatment of linguistic variables: patterns of social stratification, style shifting, and subjective evaluations. This orderly heterogeneity normally rests on a uniform structural base […]” (Labov 1989: 2). The principle of uniformity of variable linguistic usage implies that within a given speech community, speakers share the same linguistic rules and constraints. Crucially, variationism as originally proposed by Labov (e.g. Labov 1972), largely relied upon the situation of monolingual speakers, however different geographically, educationally, or socially. Meanwhile, this principle of uniformity does not fare well in language contact situations. A crucial difference has to do with the presence of a number of non-uniform patterns in communities with extensive variance and a large number of bilingual speakers.

As an example, outside the Spanish-speaking world, Mougeon and Nadasdi (1998) document non-uniformity (in their terminology, discontinuity) in linguistic behaviors among French–English bilinguals in Ontario, as compared to monolingual French speakers. In particular, the authors observed that “cases of discontinuity manifest themselves either through the outright loss of vernacular features or loss of the social or stylistic constraints that are attached to them” (Mougeon and Nadasdi 1998: 50). Importantly, these discontinuities “constitute a solid body of empirical evidence that […] in minority speech communities, the principle of uniformity of rules and constraints is often falsified” (Mougeon and Nadasdi 1998: 50). Instantiations of discontinuity in the minority language speech communities, according to these authors, hold important insights into the sociolinguistic status of the minority language in contact.

In sum, variation among simultaneous or sequential bilinguals presents fascinating possibilities for investigation. In particular, bilingual language structure and use allow us to gain insight into the potential range of native speakerhood and the interface of bilingualism and L2 acquisition. The home language, heritage Spanish in particular, becomes a secondary language beyond childhood, that is, it is neither the language of schooling, nor the politically and culturally dominant language of broader society, nor the preferred language of interaction with same-age peers. A thorough consideration of bilingual speakers also allows researchers a unique opportunity to recreate some patterns of diachronic change (see also Kupisch and Polinsky 2021). Their potential for being a window on language change arises from the situation of very intense language contacts and from the reliance on colloquial input, rarely subject to standardization. For Spanish, this rapid change has been well documented by Silva-Corvalán (1994), who shows how intrasystemic changes are accelerated under transitional bilingualism, reliant upon parallel structures between the dominant language of society and the less dominant Spanish.

The limited exposure to and limited output in Spanish in lived, social practice, as well as lack of knowledge associated with formal education and exposure to a standard variety (e.g. Pires and Rothman 2009; Valdés and Geoffrion-Vinci 1998), are principal factors responsible for the variance in bilingual populations. In addition, focus should be placed on contexts in which the socio-linguistic behavior of bilinguals can be predictably different from that of NSs. This also necessitates further research on speaker identities (see Lynch [2014] and Potowski [2012, 2018] for overviews of this topic), the role of ideologies in bilingual settings (Leeman 2012), and community engagement and service learning (Leeman et al. 2011; Lowther Pereira 2015; Martínez 2010).

So far, I have discussed the situations of bilingual Spanish use with an implicit assumption that the knowledge of the language decreases with each new generation outside the homeland. But the discussion of variance in Spanish-speaking bilinguals would not be complete without a consideration of cyclical bilingualism, where the knowledge of Spanish does not follow Fishman’s famous Three-Generation Rule where it takes three generations of immigrants to assimilate to the dominant society and language (Fishman 1972; Fishman et al. 1966; see Alba et al. [2002] for a critical discussion). With respect to Spanish in the U.S., Villa and Rivera-Mills (2009), and Lynch (2012, 2022 point out that the concept of “generation” in sociolinguistic studies can be problematic, in at least two respects. First, in some places such as the U.S.-Mexico border region and South Florida, high levels of heritage-language proficiency can be observed in some third- and fourth-generation speakers, pointing to the cyclical nature of bilingualism. Second, and relatedly, in some cases, second-generation speakers manifest lower levels of proficiency than third- or even fourth-generation speakers. According to Villa and Rivera-Mills, “[i]t is the need to connect with a particular speech community that requires a change in the traditional unidirectional, linear pattern of change to a more circular one, in which at any given point and with any generation, there is the opportunity to recapture the H[eritage] L[anguage], thereby promoting a more stable bilingualism” (Villa and Rivera-Mills 2009: 30).

5 Conclusion

At this point, the rapidly growing field of study of Spanish-based bilingualism requires a dialogue among our various subdisciplines, recognizing shared challenges and goals. Perhaps the most immediate of those challenges and goals is the immense variance evident among bilingual speakers, something that I tried to outline in this contribution. On sociolinguistic grounds, much remains to be said about the acquisition and awareness of indexical variants in the HL; speaker agency and identities; “passing” as native in terms of contextually situated language performativity; the role of ideologies in bilingualism; and community engagement and service learning.