Challenging deficit frameworks in research on heritage language bilingualism

Abstract

Recent years have seen an increased interest in the study of heritage language bilinguals. However, much of the research on heritage bilingualism is fraught with deficit framing. In this article, we demonstrate how many of the assumptions that underlie this growing field of research and the way that heritage speakers are positioned as research subjects reveal ideologies that center and value monolingualism and whiteness. We problematize a number of ways in which these ideologies commonly show up in the frameworks and methodologies used in psycholinguistics to study this population. We advocate for frameworks such as usage-based linguistics and multicompetence that center the multidimensional experiences of bilinguals and embrace nuance and complexity. We call on the research community to examine their research designs and theories to dismantle the systems that maintain heritage bilingualism at the margins of bilingualism research.

The study of bilingualism has come a long way from the predominantly negative views about bilingualism, particularly for Spanish-speaking children, that were prevalent in the 1940s and 1950s (see López et al., 2023). Research by Peal and Lambert (1962) challenged the view that bilingualism was harmful and ushered in a new wave of research that demonstrated the cognitive benefits of bilingualism and multilingualism. However, this change in perception of bilingualism did not apply to all types of bilinguals. Indeed, the ideologies underlying the bilingual deficit approach can still be found in research on bilingualism today. Despite decades of scholarly interest in bilingual language and cognitive processing, bilinguals continue to be compared to a monolingual standard. The monolingual standard becomes problematic when studying bilinguals in the USA because it fails to capture their unique linguistic experiences.

Critical scholars in linguistics have urged language researchers from all branches to “undo the racism and colonialism that were a founding motive of our discipline” with the goal of promoting racial equity within the field (Charity Hudley & Flores, 2022; Charity Hudley et al., 2020). In response to this call to action, we focus on how the study of Spanish as a heritage language reveals deep-rooted frameworks that continue to marginalize the language use patterns of racialized ¹ heritage bilinguals, and we suggest a critical self-reflection of the frameworks and methodologies used by psycholinguists to study heritage Spanish speakers, in order to seek linguistic and racial justice for speakers of this community. In our current sociopolitical context in the USA, where Spanish is one of the most spoken languages and Latinxs represent the biggest ethnically minoritized group, it is crucial that we critically examine how heritage Spanish speakers are perceived in institutional settings and scholarly works. Although a number of psycholinguists have made important contributions to challenging problematic research practices (e.g., Bayram et al., 2021; Kupisch & Rothman, 2018; López et al., 2023; Titone & Tiv, 2023; Vaid & Meuter, 2016), the field of psycholinguistics as a whole has largely failed to explore critical issues related to social categories such as gender, class, and race. In this article, we describe how raciolinguistics, the socially constructed ideologies about what constitutes “correct language” use that are circumscribed by our racial biases about the speakers, and standard language ideologies in research practices, along with a legacy that privileges monolingualism, results in the perpetuation of the marginalization of language practices by heritage Spanish speakers.

Research has long demonstrated that scholars and educators in higher education are not free from racial biases (Goar, 2008; Urciuoli, 2008). Experimental methods, in particular, have been upheld as models of neutrality and objectivity. However, as Goar (2008) points out, “research occurs within a wider field of social relations that is characterized by the agendas, interests, and values of the dominant group” (p. 162). Thus, it is important to recognize that the values and ideologies held by researchers as well as those that are valued by the broader research community affect various aspects of the research process, including the questions that get asked, the hypotheses that are generated as plausible explanations, the manner in which data are gathered, and how that data are analyzed and ultimately interpreted (Goar, 2008). In the last main section of this article, we argue that a usage-based approach and a multicompetence framework can assist us in accounting for and legitimizing the linguistic practices of heritage speakers.

Positionality statements

The authors all identify as highly educated women who are proficient in Spanish and English, but each have had a different language learning trajectory and racialized identity. The first author is a white woman who grew up in a low-income monolingual household and community in the USA and began learning Spanish in college. Her academic training has been primarily in psycholinguistics and neurolinguistics and has focused on crosslinguistic influence in the language systems of bi-/multilinguals. The second author is a bilingual heritage speaker of Spanish who grew up in California within a Mexican and Salvadoran immigrant family. Therefore, she understands the marginalization that heritage speakers endure as a group although her experience remains unique to her ethnic and white racial identity. Her academic training has been primarily in Spanish sociolinguistics, and her graduate work focuses on the critical role that race and ethnic identity play in the construction, maintenance, and dissemination of language ideologies in order to better understand language and identity negotiation outside of academia. The third author is a Mexican border woman and first-generation academic who can relate to bilingual Latinx people as someone with experience as a racialized subject in the USA, though without any experience with discrimination in the K-12 school system. Her academic training has been in Spanish linguistics and has focused on understanding Latinx bilingual speakers’ experiences as students of Spanish, as well as the racial relations they often navigate within and outside Latinx communities.

The authors’ combined experiences provide both insider and outsider perspectives on the experiences of heritage speakers, but all of them have witnessed the unjust way that heritage speakers are written and talked about in academic spaces and texts. Their different academic training has allowed them to draw on scholarship from different fields – sociolinguistics, psycholinguistics, cognitive science, race/ethnic studies, second language acquisition – to make sense of the deficit framing of heritage bilinguals and offer different approaches. The authors all mentor and teach heritage speakers and recognize their roles and responsibilities in challenging negative ideologies that marginalize and subjugate Latinx heritage Spanish speakers both inside and outside of academia.

Monolingual and native speaker ideologies

There is great variation in the language patterns seen among heritage Spanish speakers, much of it due to the specific contexts heritage speakers acquire their languages within, the language varieties they are exposed to, and predominant attitudes toward Spanish in different parts of the USA (e.g., Callesano & Carter, 2019; Kutlu & Kircher, 2021). Nevertheless, even heritage speakers whose language patterns most closely resemble those of Spanish speakers who grew up in predominantly Spanish-speaking contexts often have their language practices framed as deficient, with their competence and even their native-like practices called into question. In part, this framing is due to ideologies that position monolingual patterns and practices as the idealized target for bilinguals (Cook, 2016), a perspective labeled the “monofocal lens” of bilingualism (Vaid & Meuter, 2016). Two decades ago, Cook (2003) problematized the practice in second language acquisition studies of pitting “non-native” or L2 users against an idealized native speaker (i.e., monolingual) target. Cook argued that L2 users should not be subordinated to monolinguals and that holding L2 users to a monolingual standard would perpetually situate them as failures. Cook frames this in terms of linguistic rights: “One group of human beings should not judge other people as failures for not belonging to their group, whether in terms of race, class, sex, or language… The L2 user is a person in his or her own right… not an imitation of someone else” (Cook, 2003, pp. 3–4). These same problematic comparisons continue in many psycholinguistic studies of heritage bilingualism.

Another framework that is deeply entrenched in the study of heritage bilinguals is the idealized native speaker. The notion of the idealized speaker is associated with other social privileges and typically characterized as a monolingual (“uncontaminated” by bilingualism and language contact) of a standardized language variety, formally educated, and white (Gerald, 2020; Kubota, 2009). The truth is that the idealized native speaker exists only in linguistic mythology. The ideologies that privilege the idealized native speaker result in a monolingual bias in language research (Cheng et al., 2021), whereby bilingual speakers are, by definition, constantly found deficient. For example, heritage Spanish speakers tend to produce overt subject pronouns at higher rates than some other Spanish-speaking groups (Polinsky & Scontras, 2020). There is significant variation in overt subject pronoun production across Spanish varieties (Otheguy et al., 2007), and yet, heritage speakers’ use of overt subject pronouns is often framed as an error whereas it is considered dialectal variation for other groups of Spanish speakers.

These two ideologies, that of monolingual superiority and the idealized native speaker, form the foundation of the “incomplete acquisition” framework, which frames the practices of heritage speakers as simplified or arrested. When heritage speakers fail to exhibit the language features of monolinguals (typically living in a different country), their language is described as “incomplete” (Montrul, 2008), “defective” (Polinsky & Scontras, 2020), lacking “full competence” (Montrul, 2011), and demonstrating “interrupted or incomplete knowledge” (Alarcón, 2011). As Putnam and Sánchez (2013) point out in their critique of the term “incomplete acquisition,” it implies that there is some kind of ultimate attainment that heritage speakers never quite reach, suggesting developmental arrest and inferiority. This concept of ultimate attainment is an idealized abstraction that positions monolingualism as the norm and assumes homogeneity of linguistic representation and some kind of stable end state.

In many cases, heritage bilinguals in the USA are contrasted with “native speakers” who grew up abroad. This exclusionary language reveals that heritage Spanish speakers are often not considered to be native speakers of Spanish despite the fact that most of them acquired this language naturalistically, at home, prior to the onset of learning English, and that most continue using Spanish into adulthood, criteria typically agreed upon to be considered a “native speaker” (Cheng et al., 2021; Kupisch & Rothman, 2018; Putnam & Sánchez, 2013; Rothman & Treffers-Daller, 2014). To be deserving of the label “native,” heritage speakers are expected to speak like those who grew up in completely different countries, were educated in the language, and enjoy privileged language status in those countries. Yet, these groups of speakers grew up in very different circumstances from heritage speakers. Despite the fact that they may live in a different country, interact with a unique speech community, and are subject to all of the same evolutionary processes as other languages, the language patterns of heritage speakers are often expected to match a set of target forms that belong to a community where Spanish is the dominant language, or to “imagined communities” (Leeman, 2015).

Moreover, there is an insistence in psycholinguistic studies that bilinguals be compared to a monolingual “control” group. While this practice has been previously criticized on methodological grounds (Luk & Bialystok, 2013), López et al. (2023) further argue that comparing bilinguals to “white monolingual controls perpetuate[s] hegemonic whiteness” by positioning white monolingual speakers as the “gold standard” and treating bilingualism as an “anomaly.” Moreover, this type of group comparison often assumes that variability only exists for the bilingual group and not for the monolingual group. This assumption reflects the “idealized native speaker” in that monolinguals are assumed to have had similar experiences learning and using language and to have all reached a certain level of language competence (Cheng et al., 2021). Thus, many studies report extensive information for the bilinguals but very little for the monolinguals despite monolinguals also demonstrating meaningful within-group variability (Bice & Kroll, 2019; Pakulak & Neville, 2010).

Raciolinguistic and standard language ideologies

Monolingual and native speaker ideologies have their roots in social hierarchies of power and prestige, including systems of racism, nationalism, and elitism. Language practices associated with certain privileged groups are perceived as more valuable or inherently “better” than those of racialized groups. For example, in the USA, Anglo-American learners of Spanish are often praised, with their bilingualism framed as an asset, while the language abilities of Latinx Spanish speakers are stigmatized (what Clemons, 2022 refers to as the benefits versus burden of bilingualism for these different groups of Spanish speakers; see also Ortega, 2020; Rosa & Flores, 2017).

An important related construct to linguistic racialization is that of standard language ideology, “a bias toward an abstracted, idealized, homogenous spoken language which is imposed and maintained by dominant bloc institutions and which names as its model the written language, but which is drawn primarily from the spoken language of the upper middle class” (Lippi-Green, 2012, p. 67). In the case of Spanish, standardized forms maintain racial and class hegemony as well as institutional authority (Urciuoli, 2008). Spanish varieties associated with whiteness and economic prosperity are perceived more favorably (including more competent) than varieties used in places not associated with whiteness and economic success (Callesano & Carter, 2019). In the context of education, Spanish instructors in the USA tend to adhere to nationalistic ideologies that include the belief that there should be one national language in every country (Valdés et al., 2003). Monolingualism, viewed as the “correct” variety, is tied to nationalism, which is favored by language instructors in higher education invested in teaching Spanish as a foreign language spoken in other countries, but never as a variety in contact and/or from the USA. Thus, heritage Spanish speakers’ language is compared to normative varieties from Latin America and Europe and therefore within a system that privileges whiteness (Alfaraz, 2002; Zentella, 2017). García et al. (2021) describe this opposition as “abyssal thinking,” which, as a legacy of colonialism, positions certain language practices that are associated with “civil society” on one side of an imaginary line while the other side is delegitimized and stigmatized.

Moreover, mainstream discourses in the USA cast heritage languages as conflicting with the “American” identity. Within this ideological paradigm, language research tends to favor an idealized pursuit of understanding “the other,” which does not conform to these normalized frameworks. Through critical frameworks that respond to and disrupt the many ways in which language research and education continue to perpetuate a colonial and imperial system, researchers can critique the rigid and arbitrary nationalistic ideologies as well as racial dichotomy that only registers whiteness and its “other,” non-whiteness. While the ideology of whiteness centralizes around skin color, it is a social construction that privileges those whose identities reflect normalized cultural practices historically associated with people racialized as white (Matias et al., 2014). These attitudes result in double stigmatization: heritage speakers’ Spanish is devalued because it is a non-standardized (i.e., a marginalized) variety and their English is also perceived as “not good enough.” For instance, Chicano English varieties exhibit features that deviate from standardized American English, which are often perceived as an “accent” (Fought, 2003). A raciolinguistic analysis of this type of deficit framing reveals that heritage Spanish speakers’ language practices are perceived as problematic not because of any objective characteristics but based on the speakers’ racialized positions in society.

By analyzing racial relations and stratification, researchers can better understand how to disrupt systematic discrimination and racialization of heritage speakers. For instance, through Critical Race Theory and/or LatCrit frameworks, researchers can better understand the historical homogenization of Latinx people and how the predominant white/non-white binary in US racial politics impinges on the language and sociocultural identity construction of Latinx communities (Von Robertson et al., 2016) and other heritage speakers who do not conform to normalized sociolinguistic identity formations. White Latinx individuals, for instance, do not have the same experiences as Latinx bilingual speakers racialized as non-white. Racialization of Latinx heritage speakers, as well as other heritage speaker groups, should be considered within an intersectional and relational frameworks in order to counter social structures beyond fixed notions of identity formation or social categories (Holguín Mendoza et al., 2021). In order for heritage speakers’ varieties to be valorized and legitimized, we must question why and how the frameworks and methodologies that we currently apply when studying this community are a product of white colonialism that continue to pressure scholars and educators into legitimizing only the normative varieties. Instead, heritage language patterns should be studied as naturally evolving varieties in language contact environments (Silva-Corvalán, 1994).

Implementing usage-based, multicompetence frameworks in bilingualism studies

Theoretical frameworks used to study heritage bilingualism must incorporate the critical role of language usage in shaping language representation and processing as well as the dynamic and adaptive nature of the language system. While some versions of formal or generative approaches have attempted to incorporate the role of language usage and/or exposure into theories of language acquisition and representation (e.g., by proposing that these occur primarily at the interfaces, Sorace & Serratrice, 2009; or by proposing an “open UG” framework, Lightfoot, 2020), we find usage-based approaches and a multicompetence framework to be best suited for explaining the dynamic nature of heritage languages that is driven by individual experiences. Usage-based, multicompetence linguistic frameworks can account for various individual experiential factors and can predict language patterns found in heritage bilinguals without positioning them in a deficit relationship with other language practices.

Within the framework of usage-based linguistics, language is seen as a dynamic system that emerges as a product of language use and exposure (Bybee, 2010; Diessel, 2017). One of the basic principles of usage-based linguistics is that the language user’s linguistic knowledge is “constantly restructured and reorganized,” with structures remaining fluid as they continue to be shaped by communicative interactions as well as domain-general cognitive processes (Diessel, 2017). Within this framework, the language patterns observed for heritage bilinguals are a natural product of their lifetime of experience with language. Thus, framing the language produced by heritage bilinguals as incomplete is nonsensical since there is no defined “end state” of completion or idealized “target” within a usage-based framework. Moreover, positioning monolinguals or speakers of privileged/standardized varieties as the standard by which heritage bilinguals are assessed is also inconsistent with a usage-based framework since each of these groups’ language patterns is a product of a unique set of experiences, and thus none of them is inherently superior. In what follows, we examine how certain common research methods reflect negative language ideologies and offer alternatives that are consistent with a usage-based framework and that challenge the traditional deficit framework found in the study of heritage bilingualism.

Several researchers have pushed back on the dichotomous distinction between bilinguals and monolinguals, arguing that bi-/multilingualism is best understood not as a categorical variable but as a multidimensional construct (Bialystok, 2021; López et al., 2023; Luk & Bialystok, 2013; Pennycook, 2006). Indeed, language experiences that have been shown to influence language and/or cognitive processes include age of acquisition (Luk et al., 2011), amount of exposure and use of each language (Gullifer et al., 2021), language dominance (Puig-Mayenco et al., 2018), language immersion (Deluca et al., 2019), language brokering (López, 2020), and code-switching (Beatty-Martínez & Dussias, 2017). One can see that nearly all of these experiences exist along a continuum and are interdependent. Because of this complexity, researchers often employ different criteria for defining bilingualism and monolingualism, resulting in bilingual groups across studies that may be quite dissimilar from each other in important ways (Leivada et al., 2021; Ortega, 2020; Surrain & Luk, 2019).

Another common problem in research on bilingualism is that bilinguals are treated as a homogeneous group. The assumption of uniformity among bilinguals reduces the complexity of their experiences to a single dimension and ignores potentially meaningful individual differences (Adamou, 2021; Bialystok, 2021; López et al., 2023; Putnam et al., 2018). Group distinctions may be less informative for understanding language processing than exploring how individuals differ in their language processes. For example, both monolinguals and bilinguals are similarly impacted by factors such as quantity and quality of linguistic input and contexts of language use. Indeed, differences between these language users may be quantitative, not qualitative (Luk, 2015). Thus, researchers should consider what aspects of bilingual experience are most relevant to what is being studied and identify participants that vary along those particular dimensions. This approach is more nuanced and methodologically rigorous as it better addresses why and when bilinguals may show different outcomes (from monolinguals or from other bilinguals), not just that they show different outcomes (Salig et al., 2021; Takahesu Tabori et al., 2018). For example, the Adaptive Control Hypothesis proposes that the long-term consequence of engaging frequently in certain types of interactional contexts leads to a fine-tuning of the cognitive skills used to optimize language processing in those contexts as well as the neural networks that underlie them (Green & Abutalebi, 2013). The Adaptive Control Hypothesis exemplifies the way in which the study of bilingualism can probe how experiential factors result in various patterns of language and cognitive processing without employing a deficit framework or contextualizing bilingualism as a contrast to monolingualism.

Let us take an example from the field of cognitive neuroscience. Neuroimaging studies have shown that bilinguals often exhibit more extensive and/or stronger neural activity than monolinguals and greater gray and white matter density in certain brain regions (see Higby et al., 2013 for a review). A monolingual bias positions the brain structure and brain activation patterns of monolinguals as the standard by which the bilinguals should be compared or as the default brain state for individuals who are “uncontaminated” by bilingualism. This bias can show up explicitly, in deficit-framed wording, or less explicitly, by framing the monolingual data as expected and the bilingual data as unusual. An approach that does not privilege data from monolinguals explores how various types of experiences that differ between the groups contribute to the observed patterns. In this way, the diversity of experiences is used as a lens to understand language and cognitive processes and the neurophysiological processes that underlie them (Takahesu Tabori et al., 2018). Thus, when comparing the brain activity or structure of bilinguals and monolinguals, differences can be interpreted as a natural consequence of the different language experiences of these groups rather than as a failing of bilinguals to demonstrate the same patterns as monolinguals.

In recent years, a number of researchers have proposed new analytical methods for accounting for the diversity of bilingual experiences (e.g., Anderson et al., 2020; Gullifer & Titone, 2020; Kremin & Byers-Heinlein, 2021; Luk & Bialystok, 2013). For example, factor analyses have been used to identify the primary profiles characterizing the bilinguals in their sample (Anderson et al., 2020; Luk & Bialystok, 2013) and factor mixture and grade-of-membership models more flexibly incorporate both categorical and continuous variables (Kremin & Byers-Heinlein, 2021). A measure of language entropy has been derived to incorporate language use data for multiple languages across various communicative contexts and language domains (Gullifer & Titone, 2020). Some researchers have also advocated for more mixed-methods designs in bilingualism research, which provide a more thorough and nuanced picture of bilinguals’ language experiences by incorporating both quantitative and qualitative methods (López et al., 2023).

Bilingual experiences are highly dynamic, not only changing over time but also intersecting in complex ways to affect language and cognitive processes (Leivada et al., 2021; López et al., 2023). However, the dynamic nature of language and cognition is often ignored in psycholinguistic studies of bilingualism as single measurements are meant to represent static features of language and cognition. Salig et al. (2021) refer to the typical group comparisons as a “traits approach” because it assumes not only uniformity within each of the groups but also stability of the characteristics under investigation. They advocate for psycholinguistics researchers to instead take a “states approach,” which identifies dynamically fluctuating cognitive and neural states that emerge when bilinguals are placed in different linguistic or cognitive contexts. Indeed, within usage-based linguistics, language is considered to be a complex adaptive system (Beckner et al., 2009) in which the behaviors are the result of an interaction of various factors, which change based on past experiences or the interaction between current and past experiences (Beckner et al., 2009; Ellis & Larsen-Freeman, 2009; Filipović & Hawkins, 2019).

The dynamic processes inherent in bilingualism can be seen in patterns in crosslinguistic influence. Although crosslinguistic influence on a second-learned language is relatively well known, stronger evidence for the adaptability and flexibility of the language system comes from changes to native-language processing as a result of bilingualism and/or second language acquisition. For example, Spanish–English bilinguals who are English dominant or have had extensive exposure to English are more likely to demonstrate relative-clause attachment preferences that align more with English monolinguals than Spanish monolinguals (Dussias & Sagarra, 2007; Fernández, 2003). Additionally, highly proficient bilinguals can process ungrammatical constructions that are grammatical in the other language better than monolinguals and less proficient bilinguals (Fernández et al., 2017). By gaining knowledge of this syntactic structure and its semantic interpretations in English, these bilinguals presumably draw on the rich linguistic knowledge available to them from both languages to help them make sense of similar constructions.

The evidence for crosslinguistic influence provides strong support for the notion of multicompetence. Multicompetence refers to “the overall system of a mind or a community that uses more than one language” (Cook, 2016, p. 2). The principle is one of a linguistic supersystem housing all of the user’s linguistic knowledge, “with complex and shifting relationships between them” (Cook, 2016). Multicompetence is not limited to syntactic structures, and neither is crosslinguistic influence, but rather all aspects of language are part of a shared bilingual/multilingual mental linguistic system. A benefit of taking a multicompetence perspective in research is that, like usage-based linguistic approaches, it centers the experiences of bilinguals, positioning their language practices as “normal” (Cook, 2016; García et al., 2021). This perspective asserts the rights of bilinguals to exist on their own accord, not simply in “the shadows of native speakers” (Cook, 2016).

One of the primary tenets of usage-based linguistics is that language use is inherently social, and therefore language is shaped by social interactions. This implies that to understand language processes, one must consider current and historical social contexts. As Titone & Tiv (2023) illustrate, “people are embedded in a dynamic, multilevel system of sociolinguistic context whereby direct personal interactions and ambient language exposure constrain their everyday language behavior” (p. 10). However, most psycholinguistic studies of bilingualism examine language and cognition as isolated from the social contexts of the speakers. Surrain and Luk (2019) found that only 30% of studies on bilingualism described the participants’ sociolinguistic context. For example, in some US contexts, such as parts of California, Texas, and South Florida, heritage Spanish speakers may interact with a rich community of other Spanish speakers, while in other contexts their interaction with other Spanish speakers may be very limited. López et al. (2023) encourage researchers to incorporate not only individual factors such as language proficiency and age of acquisition but also environmental factors such as immigration, community language practices, and language policies. Titone & Tiv, 2023) outline a model called the Systems Framework of Bilingualism, which contextualizes cognition, language, and neuroplasticity within interdependent sociolinguistic layers that include interpersonal, ecological, and societal spheres of influence that are also temporally situated.

Incorporating bilinguals’ social context in our frameworks can also challenge the tendency to compare heritage bilinguals to speakers who grew up in very different circumstances and the deficit framing that is often inherent in that comparison. An important step for legitimizing the language varieties spoken by heritage bilinguals is situating them within their own sociolinguistic community without expecting them to have outcomes found among other speech communities. Additionally, we need to recognize the wide-ranging variability in experiences and language patterns that characterize the heritage-language-speaking community. By moving beyond binary classifications and deficit thinking that erase this community’s language traditions and abilities, we can utilize this rich variability to address many of the questions that are of interest to linguists, such as the effects of quantity and quality of language input, diversity of interlocutors, and literacy. We can still examine developmental trajectories, cross-language influence, language contact, and language change from a perspective other than the “white gaze” that positions racialized speakers as deviant. This approach would permit research on heritage speaker bilingualism to be more inclusive and also address the widespread belief in applied linguistics that heritage speakers’ language needs to be “fixed.” What if instead we consider whether heritage speakers have their own pragmatic constraints that do not match those of the standard variety, or any other variety for that matter? Spanish speakers in the USA, as any other social group, possess their own community’s indispensable rules and sociolinguistic nuances to signal a variety of social meanings for a broad array of social contexts (Holguín Mendoza, 2022; see also Parodi, 2011).

Conclusions and recommendations

Psycholinguistic research on the language practices of heritage bilinguals, in particular heritage Spanish speakers, continues to reflect deficit narratives that contribute to the marginalization of this population and the perpetuation of biased and misleading research results. While traditional group comparisons (e.g., bilinguals vs. monolinguals or heritage bilinguals vs. other types of bilinguals) may be useful for certain types of questions, they tend to flatten the multidimensional nature of bilingualism and ignore crucial differences among the bilinguals that are grouped together or ignore similarities that may exist across the groups. We have shown how these deficit approaches are linked to ideologies that favor white, monolingual speakers and always position others in relation to the “idealized native speaker” that emerges from that concept. We urge researchers to interrogate their frameworks and methods and identify any assumptions in their research designs that may reveal adherence to these problematic ideologies. Researchers have a responsibility to challenge the biases and harmful ideologies that permeate the field as our research, and the way it is framed, potentially impacts other fields.

The implementation of a usage-based, multicompetence approach challenges the ideologies that position heritage bilinguals at the margins by recognizing that their language practices “will depart from those of their monolingual counterparts in both L1 and L2, rendering the native/non-native dichotomy no longer relevant” (Fernández et al., 2017, p. 263). Thus, we urge researchers to recognize the multidimensionality of bi-/multilingualism, collect experiential data from all participants, not just those identified as “bilinguals,” that describe how participants vary across these dimensions, and to not automatically reduce bilingualism to a unidimensional, categorical variable. We also encourage more cross-disciplinary collaborations, which provide different perspectives, frameworks, and methods, as the relevant factors for the study of bilingualism go far beyond the narrow confines of linguistics and psychology. We hope that the field will move away from deficit frameworks of heritage bilinguals and toward a linguistic empowerment approach, one that recognizes the rich linguistic history of heritage speakers’ communities.