Introduction

The nation-state of Belize, with a population of only about 325, 000 people (Statistical Institute of Belize, SIB), is both Central American and Caribbean, yet seemingly belongs to neither. This is true not only politically and economically as reflected in its membership to both the Caribbean Community (CARICOM) and the Central American Integration System (SICA) but also linguistically and culturally. Like the Anglophone Caribbean, Belize is English- and Creole-speaking which represents strong historical, social and linguistic ties to the Anglophone Caribbean, particularly to Jamaica where the administrative centre of the colony was located (Young). At the same time, more than half the country speaks Spanish (SIB) linking Belize to other Central American nations as well as Mexico both culturally and linguistically. In addition, Spanish, English and English-lexified Creoles are also spoken along the Central American Coast in Honduras, Costa Rica, Nicaragua, Panama and Guatemala (Holm; Leung and Loschky). Likewise, Maya languages spoken in Belize are also spoken in Guatemala (Mopan and Ketchi) as well as Mexico (Yucatec), and Garifuna, spoken primarily in southern Belize, is also spoken in Honduras, Guatemala and Nicaragua. Thus, language reflects and positions this small nation as both Caribbean and Central American/Latin American.

Spanish, English and Belize Kriol (hereafter Kriol) are numerically the majority languages in Belize and represent one aspect of this in-between status that Belize holds.¹ An outcome of these three languages co-existing in the same geographic space i.e., being in contact, is that many speakers develop multilingualism and some engage in language mixing.² As shown below in (1),³ speakers use single English (endorse) or Kriol words (bway, don) in Spanish discourse or alternate between multi-word fragments, in this case between Spanish (yo sabía…) and English or Kriol (I work…).⁴

In this paper, the empirical linguistic practices of multilingual Spanish speakers (Fuller Medina forthcoming, 2016) are examined within the Belizean semiotic landscape. This landscape is elucidated by triangulating data from official and unofficial language policies; language attitudes from interview data; and observational material garnered while I was a professor at the national university in Belize. It is demonstrated that speakers must not only navigate linguistic systems but also ideological ones in order to achieve their social and communicative goals. Connecting language choice to the social in the Belizean context is not new of course; the foundational work of LePage and Tabouret-Keller analyzed linguistic behavior as acts of identity developing a model by the same name. However, this work was done at a time when Belize as a nation was in some ways still finding its footing during the period leading up to independence in 1981.⁵ At the time, alternations between languages were noted but mixed discourse, or the use of multiple languages in the same utterance was not the focus of analysis. Likewise, by 1978, LePage and Tabouret-Keller reported that, in contrast to earlier interviews, Kriol was being explicitly named as a Belizean language, as was the identity of “Belizean” (rather than British Honduran) with a strong association between the two (220). Thus, the current paper is part of a larger project examining the ways in which empirical linguistic practices connect to and interact with race, ethnicity and citizenship in Belize offering a snapshot, 40 years later, of language and social meaning in post-independence Belize.

Few studies have taken a global approach to Spanish as spoken in Belize and fewer have provided an examination of the Belizean sociolinguistic landscape since LePage and Tabouret-Keller. Not surprisingly, the studies which examine the varieties of Spanish spoken in Belize have been carried out by dialectologists (Quilis, Quesada Pacheco, Cardona Ramírez) with quantitative studies being in the minority (Hagerty 1979, Fuller Medina forthcoming, 2016).⁶ In fact, Belize is largely absent in the Hispanic linguistics literature, receiving scant mention in foundational works on Latin American Spanish (e.g., Hualde et al.; Lipksi, 1994). No doubt due, in part, to its status, and relative visibility, as an English-speaking nation which overshadows the fact that more than 50% of the population claim Spanish as a language they speak (SIB). Not surprisingly, the majority of the literature on Spanish-English bilingualism and contact has been based on varieties of Spanish in the U.S. particularly Chicanx Spanish (e.g. Aaron, Amastae and Elías-Olivares; Lipski, 2008). Yet sociohistorical circumstances are hypothesized to lead to various outcomes of language contact depending on the status of the language, size of the community, and intensity of contact (Thomason and Kaufman). Even where the same languages are used in two different communities, distinct patterns of language use can emerge (Poplack; Muysken). Thus, examining Spanish-English contact in non-U.S. communities is instructive in determining if the U.S. patterns are generalizable. This paper also contributes to filling the gap in Hispanic linguistics with respect to Belizean varieties of Spanish.

In what follows, section 2 provides a historical sketch of Spanish, English and Kriol followed by an overview in section 3 of the empirical linguistic practices of a cohort of Spanish speakers who engage in robust language mixing (Fuller Medina forthcoming, 2016). This section also examines the status of Spanish, Kriol and English within the ideological semiotic landscape where these practices take place. The final section concludes the paper.

Historical sketch of language in Belize

The territory now known as Belize was originally home to Maya groups whose resistance impeded Spanish settlements (Buhler). British buccaneers⁷ who would lay in wait off the coast of Belize to rob Spanish flotillas of their wood cargo took advantage of the lack of Spanish settlements to encroach on the territory (Shoman 19). The buccaneers later turned their attention to cutting their own logwood and mahogany⁸ eventually establishing settlements and later obtained formal permission from Spain in 1763 for woodcutting (Ibid: 173). At this point, enslaved Africans were brought to Belize to work in the wood cutting industry. This is the context which gave rise to the English-lexified Creole language spoken in Belize, Kriol. Creole languages are predominantly those speech varieties which were formed as a result of contact between European languages and West African languages in the context of slavery during colonization.⁹ Kriol shares much of its lexicon with English but also draws grammatical and lexical features from the West African Languages Akan, Bantu (Escure), Yoruba, Twi, and Igbo as well as the indigenous Miskito language of the region (Young). Because enslaved people were often brought through Jamaica, and Belize had an administrative relationship with Jamaica, it is also believed that there is a close relationship between the Creole spoken in Jamaica and Belize Kriol (Young 34-37).¹⁰

Spain and Britain eventually entered into a conflict over territorial rights, which culminated in the Battle of St. George’s Caye in 1798 where the Spanish were defeated.¹¹This battle serves as the decisive moment in the history of Belize leading to its current status as the English-speaking nation in a region dominated by the Spanish language. English remains the language of power of and prestige in Belize. This, despite the fact that only 63% of the population report English as one of the languages they speak; a number which may actually be lower since the census cautions that “some respondents… might in fact have been referring to Creole” (SIB 21).

Both contemporary migrations from other Central American nations as well as historic migrations in the mid-19th century from Mexico and Guatemala provide the input varieties for Spanish in Belize (Hagerty; Shoman; Young). The varieties established as a result of the early migrations are considered to be Belizean varieties of Spanish.¹² The 1850’s saw large-scale migration of the Mestizo and Maya people fleeing the Caste War in Yucatan who settled in the northern region of the country (Shoman 86). This population, included Maya and Spanish speakers, many of whom were bilingual. Similarly, in the 1860s, Spanish-speaking and Maya-speaking people migrated to western Belize (Shoman 87-88; Hagerty, 1979, 22).¹³ This migration stemmed in part from political unrest and in part from economic factors as the Chicle¹⁴ industry attracted workers from both Guatemala and Mexico (Shoman 87; Hagerty, 1979, 22). Both these groups of migrants were sizeable enough to define the regional linguistic landscape and therefore serves as the historical benchmark for 150 years of Spanish-English contact.¹⁵ As a result of isolation, each variety developed along separate trajectories maintaining their distinctions to become Western Belizean Spanish (WBS) and Northern Belizean Spanish (NBS).¹⁶In addition, these varieties developed in the absence of Spanish prescriptive norms because the language of instruction in Belize is English, (Hagerty, 1992).¹⁷ The details of the linguistic contact in the period between the first migrations to present day are unclear but many speakers were Maya speaking and/or Spanish-Maya bilinguals making Yucatec Maya likely the lingua franca in northern communities in the first few decades (Cal, p.c. April 2, 2020).¹⁸ As a majority language, Spanish has since displaced Yucatec Maya in various communities and therefore, to speak of Belizean varieties of Spanish is to also speak indirectly of language shift from Maya to Spanish. These communities are undergoing or have undergone a second shift towards English and, more recently, Kriol particularly among younger speakers (Cal and Fuller Medina, 2017).

The Spanish spoken in Belize reflects these shifts. In addition to the Kriol and English features noted above, the Maya influence can be seen in loanwords such as xix (crumbs, small amount), chichi (grandmother), and tuuch (navel) which are still used today (Fuller Medina, 2016; Hagerty, 1979; Quilis) as well as in the phonological system as per Hagerty’s phonological analysis (1979, 1992). In addition, a number of non-standard features characterize Belizean varieties of Spanish. These include duplicated possessives (su….mi) (2), third person auxiliary forms (ha) for first person forms (he) (3) and non-standard use of demonstrative “otro uno” constructions (4).

The duplicated possessives are well documented in Spanish (Company Company; RAE 349) and are likely a retention from Old Spanish. The third person auxiliary in (3) is also found in Southwest Spanish (Torres Cacoullos and Travis) suggesting an English source. However, this feature has been documented in “rustic” or rural Spanish dialects which presumably have little contact with English (Lipski, 2008, 96). The same is true of the usage shown in (4), which is attributed to calquing from English but may well be a retention from Old Spanish as well (Hagerty, 1996, 137).

Some features which distinguish the two varieties are the production of rhotics, word-final nasals and the use of verbal voseo. Northern Belizean Spanish has a retroflex approximant~tap contrast whereas Western Belizean Spanish maintains the canonical tap-trill contrast typical of mainstream varieties of Spanish (Hagerty, 1979, Fuller Medina, 2016). In the west, word final alveolar nasals tend to be produced as velars and there is widespread use of verbal voseo (Cardona Ramírez; Fuller Medina, 2016). As will be discussed further below, some of these features and, in particular, the use of mixed discourse shown in (1), above, lead to Spanish being stigmatized. In fact, Spanish is Belize is called “kitchen Spanish” or frequently described as “broken” or “bad” Spanish.¹⁹This perception of Spanish in Belize as mixed or “broken” is widespread; however, it is not the case that all speakers engage in mixed discourse. In fact, in a corpus of 51 interviewees, 40% of speakers use highly monolingual speech making very little use of codeswitching or borrowings (Fuller Medina, 2016, 86-88). This paper has as its focus a specific cohort of speakers who do engage in high rates of language mixing in order to understand how this discourse mode connects to the semiotic landscape. Speakers of NBS, WBS or other Central American varieties of Spanish who do not make use of language mixing as an everyday discourse mode may move through and interact with the landscape in distinct ways and are not analyzed at this time.

The main ethnic and linguistic groups in Belize are: Creole/Kriol, Mestizo/Spanish/Latino, Maya, Garinagu²⁰ with Mestizos currently forming the largest ethnic group as seen in Table 1. This demographic has been shaped by both continued migration of Creole people to the United States since the mid -1940’s coupled with migrations from other Central American nations to Belize as of the early 80’s (Vernon). The term Mestizo²¹ has its origin in the Spanish colonial caste system referring to a person of mixed indigenous and European ancestry. This definition still holds today in Belize but is not categorically accepted. Some reject the label Mestizo on the basis of its colonial roots and erasure of indigenous identity. On the other hand, in response to the more inaccurate term Spanish and the more offensive term Paña, many advocate for the use of Mestizo to name ethnic identity and move away from pejorative, inaccurate terms. Creole is also a term from colonial times, originally denoting a person born in the colonies, but in Belize it now generally refers to a person of (non-Garifuna) African ancestry or mixed European and African ancestry (Vernon, Bradley qtd in Judd). This is the usage adopted in the current paper. However, Creole is also sometimes used as a descriptor meaning mixed and can be conflated with language. LePage’s notes that for many participants the terms mixed and Creole were synonymously used in describing their ethnicity. For example, one participant stated that they speak “…broken Spanish and broken Cr…. English, so I would call myself Creole” (221). Another participant similarly uses the term Creole to describe Spanish, calling it “Creole Spanish” because it is “not grammatical Spanish” (167) in which case the notion of Creole being a “broken” version of language is applied to Spanish.

As the discussion of the semiotic landscape will show, empirical linguistic practices further complicate this since the language, Kriol, is adapted into the linguistic identities of non-Creole people. While a full treatment of race and ethnicity is beyond the scope of this paper, it is worth noting some distinctions in Mestizo and Creole relationships to Belizean identity. According to Macpherson, the Mestizo identity calls on indigeneity, aligning itself more closely with a connection to the land and indigenous ancestors while remaining in a hierarchical relationship to Maya people in Belize. The Belizeanness of Creoles on the other hand, tends to be traced back to the formation of Belize as a nation state, set in motion with the defeat of the Spanish at the Battle of St. George’s Caye. This battle is the mythological basis for the Creole claim, particularly elite Creoles, as the rightful inheritors of Belize in the post-colony and independence period (Premdas). At the same time, persisting racist colonial legacies define contemporary ideologies which privilege lighter skin and “straighter” hair [Bolland qtd in Premdas]. Premdas suggests that “A light-skinned Mestizo child who learns English and attends a prestigious St. John's College is more likely to rise into the Belizean elite than a dark-skinned Creole-speaking child of African descent." Thus, the colonial ideologies also create hierarchies of language where English is simultaneously positioned at the top and Spanish and Kriol at lower levels as stigmatized codes relative to English. As illustrated in Premdas’ assertion, race and language interact in this landscape.

The census data show that, numerically, Spanish,²³ English and Belize Kriol are the majority languages of Belize, though English is the only one that is socially a majority language. To a large extent the main ethnic groups in Belize have historically corresponded to the linguistic groups in Table 1. While this mapping is by no means categorical, it forms the basis of the racialization of language. Spanish, for example, is generally associated with those who present phenotypically as Mestizo and Kriol with Afro-Belizean Creole people. However, these associations are nuanced. As the census data indicates, the number of speakers who claim Kriol (44.6%) far outnumbers those who claim Kriol ethnicity (25.9%) suggesting that it is spoken widely by various ethnicities (See also LePage and Tabouret-Keller 220-221).

Semiotic landscape: empirical practices, ideology and linguistic agency

The semiotic landscape refers to the multidimensional structure where linguistic styles, features, and choices connect to the social (Eckert). At any given moment a speaker makes choices based on where they are located in this structure and in response to the range of possibilities available to them. As Lippi Green (72) notes, the social space between speakers is rarely neutral. Thus, as speakers make linguistic choices and position themselves in the world, linguistic ideologies, i.e., beliefs about language and language use, mediate these acts. These choices in language use and social positioning via language, which need not be conscious, can be considered to constitute linguistic agency. Accordingly, the empirical practices of Spanish speakers who engage in language mixing reflects this agency, both defining and being defined by the Belizean semiotic landscape.

Empirical Practices

In order to identify the major linguistic strategies employed by Spanish speakers who make use of Spanish, English and/or Kriol in the same utterance, naturalistic data from interviews with a cohort of high language mixers from northern and western Belize was analyzed. A comparative quantitative analysis of all instances where Spanish speakers used more than one language in the same utterance revealed that they employ three main strategies: (i) categorical integration of English-origin (EO)²⁴ nouns and verbs into Spanish grammar, (ii) alternation between languages in mid-sentence which do not violate the grammars of the respective languages and (iii) use of English-origin discourse markers for pragmatic functions in Spanish (Fuller Medina forthcoming, 2016, 2015).²⁵ These are described further below.

The comparative analysis of bilingual data (Poplack and Meechan, 1998), in this case multilingual data, takes advantage of typological similarities and dissimilarities to determine which grammars are operational in mixed discourse. Spanish, English and Kriol are typologically similar languages. This means that they share a number of grammatical features such as word order. The sentences below for example, illustrate their classification as Subject-Verb-Object (SVO) languages, indicating that the order of words in a sentence is subject (John/Juan) followed by verb (read/lee) followed by object (the book/el libro). Thus, areas of grammar where these systems coincide, such as the SVO order, can be considered sites of coincidence or coincidence sites.²⁶

An exception to word order similarities is related to the order of attributive adjectives and pronominal arguments of the verb in object position (it in 7, above). In Spanish, adjectives are generally postnominal (casa roja) as compared to English and Kriol where they precede the noun (red house).²⁷ Thus, the grammars are dissimilar in these respects representing a conflict site (Poplack and Meechan).²⁸ As discussed further below, Spanish differs from English and Kriol with respect to the placement of pronominal objects. Additional conflict sites include number agreement and verbal inflectional morphology. Spanish employs more inflectional verbal morphology than either Kriol or English and also requires number agreement between nouns and modifiers in contrast to the latter languages.

In Spanish when a noun is plural, the determiner must also be marked plural whereas this number agreement is only required in English and Kriol in limited cases (e.g., demonstratives); furthermore, Kriol nouns are not inflected for plural reference (Decker; Young). Where EO nouns were used in Spanish discourse this number agreement rule was applied near categorically as in (8) below where the determiner los is the plural in accordance with the noun it is modifying.

With respect to EO verbs in Spanish discourse, these appear overwhelmingly in a construction composed of a verb meaning ‘to do’ or ‘to make’, hacer, and an EO verb as shown in (9) (Fuller Medina forthcoming, 2016, 2015). In these constructions, hacer serves a functional role carrying verbal inflectional morphology as required by Spanish and the EO verbal item (mix) supplies the semantic information i.e., naming of the event as well as argument structure (Fuller Medina, 2020). In these constructions, Spanish inflectional morphology categorically appears on hacer and where the verb has a pronominal object (or reflexive pronoun), the placement is consistent with Spanish grammar (Fuller Medina forthcoming, 2016, 2005, 80). In (9), for example, the object pronoun lo, capitalized for clarity, may be attached to the embedded verb (hacer mix) because the verb is non-finite, while in (10) the object pronoun must be preverbal because hacer is finite in this construction.

In the case of both EO nouns and verbs, the rules of Spanish grammar are systematically applied evidencing speaker knowledge of the Spanish grammatical rules and how to apply them. Likewise, the use of intrasentential codeswitching as in (11, repeated from 1) points to speaker knowledge of the syntactic rules of the languages in question. These alternations, only occur at sites of coincidence, or where grammars are similar, and none occur at conflict sites which would entail violating the grammars of the respective languages.

This sensitivity to conflict and coincidence sites among the respective languages is further elucidated through the use of English-origin adjectives (Fuller Medina, 2016, 228-229). Recall that Spanish differs from English and Kriol regarding adjective-noun order. When using EO adjectives in Spanish, speakers show clear strategies to resolve the conflict of grammars. The first strategy is to use them only at sites of coincidence as in (12), below, where the EO adjective negative would occupy the same position in both English and Kriol (postposed to algo, ‘something’).

The second strategy is to integrate them into Spanish in which case the adjective (strapless) occupies the post-nominal position as required by Spanish as seen in (13). Finally, speakers circumvent the conflict site by switching languages before the [ADJ+N] constituent as shown in (14), rather than violating Spanish word order which would lead to ungrammatical structures such as *la red casa. The treatment of adjectives provides clear evidence of speaker sensitivity to sites of coincidence and conflict between multiple grammars.

Another area where speakers make use of English-origin items in Spanish discourse is at the discourse-pragmatic level. They frequently employ EO discourse markers such as Like and so, shown in (15), for this purpose (Fuller Medina, 2016).³⁰

Previous work on discourse markers in bilingual speech has attempted to analyze them within the codeswitching-borrowing binary in order to determine their status as one or the other (Lipksi, 2005; Torres; Flores Ferrán) or to determine if they have the special function of introducing codeswitches (Aaron). Since discourse markers are not syntactically bound, it has proved challenging to determine if they are borrowings or codeswitches. This is because they primarily serve pragmatic rather than syntactic functions such as ensuring discourse coherence, bracketing units of talk (D’Arcy; Schiffrin 57), marking speech as more colloquial in nature (Sankoff et al.) and signaling cooperative aspects of communication D’Arcy (394-395). In other words, discourse markers can easily be adapted into bilingual speech since they are not semantically heavy and speakers need not circumnavigate syntactic boundaries or apply inflectional rules; speakers appear to do just that.³¹ In the data discussed here, the EO discourse markers appear in various syntactic positions in Spanish sentences, consistent with pragmatic functions. In addition, they do not appear to serve any special bilingual function such as introducing codeswitches since the majority of EO discourse markers do not precede alternations to Kriol or English (Fuller Medina, 2016, 215-220). This use suggests multilingual competency at the level of discourse-pragmatics. Discourse markers are generally not a part of prescriptive language instruction and are instead features of local vernaculars (Sankoff et al. 193). Thus, speakers have knowledge of this area of grammar for all three languages which allows them to integrate discourse markers into their multilingual narratives.

Mixed discourse has historically been viewed as a compensatory strategy for lack of proficiency in Spanish (cf. Poplack; Valdés-Fallis). However, the preceding discussion of the use of discourse markers, EO nouns, EO verbs and codeswitches suggests that speakers are simply recruiting all the linguistic resources available to them, because it serves their communicative goals and because they can. None of these strategies appear to be compensatory. These practices show clear evidence of competency in Spanish. In addition, the codeswitches and bilingual verbs, in particular, also evidence competency in Kriol and English since knowledge of these grammars is also required. This language use is systematic and rule-governed, not in the sense of a set of bilingual or codeswitching rules, but according to the constraints of the languages themselves. Thus, the data further reveal the speakers’ skillfulness in navigating multiple grammars when used in the same utterance (Bullock and Toribio; Flores; Otheguy and Stern; Poplack).

Ideology and linguistic agency

While the above-described patterns demonstrate un-inhibited use of Spanish, English and Kriol, this discourse mode and the languages themselves are embedded within a semiotic landscape where racialized colonial linguistic ideologies figure prominently.³² As previously noted, English is the only language in Belize with official status and is required for access to education, employment, services (though some services are available in Spanish) and broader participation in citizenship. This privileging of English along with the distancing and exclusion of other languages is rooted in the rhetorical traditions of the colonial project (Marfield). Monoglossic ideologies and the racialization of language are additional components which work together to mediate linguistic choices.

Linguistic choices or style can be said to operate in three dimensions: the social, the situational and the transactional (Winford 32). The social may relate to language choice as an act of identity, be this individual identity or group membership. The situational, in this case, refers to the contexts where languages are generally used and with whom. Thus, for example, English is the prestige language in Belize and, as such, the mandated language of instruction. Accordingly, English would be used in the classroom. The transactional dimension, the most basic form of language use, may be where a speaker recruits and deploys linguistic resources to achieve a particular goal. This goal may be to satisfy a non-social need such as purchasing food. Thus, codeswitching or any other language choice may be transactional. These three dimensions are described here as separate and discrete alongside terms such as identity and ethnicity which are purposely left vague as a matter of methodological convenience. As more research is done on language in Belize, the ways in which communities themselves conceptualize these categories can be ascertained as well as the ways in which traditional macrosocial categories of class and gender both operate in and interact with language in this context.

Linguistic ideologies are closely tied to the social and situational dimensions since they become manifest in linguistic hierarchies. These hierarchies, in turn signal the social spaces for which a language or discourse mode is appropriate. They further signal which codes are available for doing social work such as marking identities. These ideologies can be formalized in written policies, codified in practices and expressed via attitudes or other metalinguistic commentary by speakers themselves. Thus, while Belize Kriol and Spanish are both numerically majority languages in Belize they are simultaneously socially minoritized, Kriol to a much larger extent than Spanish.

Kriol has no official status but is largely considered to be the lingua franca in Belize as it is widely spoken (Escure; Udz; Decker; Young; Le Page and Tabouret-Keller). It should be noted, however, that not all Belizeans speak and/or understand Kriol therefore it is more useful to speak of Kriol as the dominant national language which functions as a lingua franca in many, but not all, communities.³³ Like other Creole languages, it has historically been considered “bad English” or “broken English” and, as such, holds little or no overt prestige (Udz; Young; Le-Page and Tabouret-Keller). In other words, it does not have recognition in formal spaces such as education, government functions, or legal documents or proceedings where the only language sanctioned is English. Kriol may be rebuffed or even penalized in formal settings in favor of standard English. For example, the Ministry of Education (MoE) states that “in classrooms where Kriol is the dominant language teachers should model English to the maximum extent, using Kriol only for special purposes” emphasizing the importance of students learning the difference between English and Kriol (167).³⁴ Similarly, course outlines at the national university explicitly state that assignments must be written in standard English indirectly proscribing against the use of Kriol. The belief that Kriol is a barrier to acquiring the standard, as alluded to in policies, is also one that is held by speakers as seen in (16).

In addition, as a language that emerged from the sociohistorical circumstances of slavery, spoken by people who were viewed as primitive rather than full intelligent beings, the belief that the language is “broken” is one that has strong historical roots.³⁵ Kriol is not simply racialized as Black, i.e. the language of Creole Belizeans, but it holds a historical association with enslaved people. In the metalinguistic commentary below (17), an older parent rejects Kriol telling their young adult daughter that she should not be speaking Kriol because she is Mestiza. This is a clear expression of raciolinguistic ideology: Creole Afro-Belizeans speak Kriol, Mestizos/as speak Spanish, Kriol is ugly.

At the same time, attitudes have been shifting in the past 20 years, particularly among younger people who view Kriol more positively; no doubt as a result of the work of the National Kriol Council over the past 20 years (Udz). Furthermore, the domains in which Kriol is used outnumber the spaces where it is deemed unacceptable. Kriol can be used at home, among friends, in informal spaces at work, during breaks, and can also be used in written form in online news outlets,³⁶ with varying degrees of approval or rejection, as well as on social media as is easily observed on #Belizeantwitter or other social media sites. Not surprisingly then, as Escure notes, Kriol is gaining ground linguistically and despite rejections such as those expressed above, this “low prestige” language has continued to cross ethnic boundaries since LePage and Tabouret Keller first noted its association with Belizean identity. In fact, it appears to be moving towards becoming a marker of pan-Belizean identity rather than solely the language of those speakers who are ethnically Creole (Udz 201). This has led to shift in some Garifuna communities in favor of Kriol (Ravindranath; Bonner) and may be taking place in some Spanish-speaking communities as well. Further indications of this this spread is the incorporation of Kriol into the everyday language of the Mestizo Spanish-speakers described above. Thus, speakers navigate the intricacies of multiple grammars in tandem with the ideological practices, tied to these languages. These ideologies on the one hand militate against the use of Kriol, marking it as ugly and as a failure to acquire English, effectively shaming speakers, and on the other hand reinforces its use as a source of in-group membership and national pride. An outcome of this contradiction is found in the empirical practices which are in direct opposition to the belief that Kriol is a hindrance or notions that Kriol is a “broken language”. The vitality of Kriol itself also contradicts these ideologies but so does the fact that Spanish Speakers expand their repertoire of semiotic symbols by incorporating a lower prestige language (Kriol) along with the higher prestige language, English, into their discourse and identities. Both English and Kriol have been agents of shift becoming part of mixed discourse, however, it can be argued that this expansion is motivated by very different circumstances. English relies on hegemonic ideologies and potential for economic access and power, while Kriol relies on the social meanings it indexes, potentially piggy-backing linguistically on English due to typological similarity.

Like Kriol, Spanish is a majority language which is also stigmatized; however, it lies in between English and Kriol in the hierarchy of the Belizean linguistic landscape. Despite the fact that Spanish has long been described as “kitchen Spanish”, Creole Spanish or “jus laik yu di taak Kriol” (example 19 below), no one would argue that Spanish is not a language, even with reference to local varieties. There is no proscription against Spanish, in fact, quite the opposite. Spanish is a de facto second official language of Belize and this is codified in various ways such as in education policy which recognizes “Belize’s geo-political situation and the status of Spanish as a major language of business and trade, and will support school and community efforts to enable students to acquire functional skills in the Spanish language by the end of primary school” (MoE 4). Spanish is also required at the secondary and tertiary level. Its status is further codified in employment, where Spanish is now a common requirement, and access to tertiary level education where it is an important factor for accessing scholarships in the region. There is at least one Spanish language radio station in Belize and, most recently, government offices provide official written communication in both Spanish and English regarding Covid-19 and related employment relief.³⁷ Thus, Spanish has overt prestige potentially translating to economic, educational and information access. In contrast to Kriol, there are no direct or indirect interdictions against it. Thus, Spanish is valued in a variety of domains including formal spaces.

On the other hand, these ratifications implicitly refer to standard Spanish. Because of the retentions noted previously and mixing with English as well as Kriol,³⁸ Belizean varieties of Spanish are stigmatized as shown in the attitudes expressed by speakers in metalinguistic commentary below (Fuller Medina 2016).

These purist ideologies which extend to language mixing are also expressed in teaching practices. One in-service high school teacher reported “correcting” students’ use of voseo as this usage did not align with the course textbook. Similarly, a university instructor was heard describing a student’s use of espelear, a calque from the English verb to spell, as an aberration and unacceptable for someone graduating with a university degree. NBS speakers who study in Mexico also reported experiencing linguistic discrimination (Fuller Medina, 2016, fn77). The empirical practices, described above, directly contradict these attitudes. The criticism of espelear is a purely ideological one since all languages borrow lexical items and this calque, in particular, has been integrated in the major verb class for Spanish. This is a well-documented strategy and one which also points to competency in the verbal inflectional system. Nonetheless speakers express similar beliefs regarding language mixing (Fuller Medina, 2016, 2015):

Notwithstanding these attitudes, local varieties and mixed discourse are valued in the arena of informal spaces, both holding covert prestige because, like Kriol, these discourse modes align with local cultural practices and identity (LePage and Tabouret-Keller, 1985). For example, the same speaker in (18), above, also says that mixed discourse “becomes a part of you” in (22), reminiscent of Anzaldua’s conception of language: “Ethnic identity is twin skin to my identity. I am my language” (59). While language mixing may simply be transactional, depending on the situational dimension, it clearly does social work such as marking bilingual/multilingual identity.

Furthermore, when speakers engage in mixed discourse they are simultaneously negotiating a multidimensional landscape where hegemonic ideologies constrain language use. The data described above were collected within the specific context of the sociolinguistic interview which aims to create a space where informal, naturalistic speech can be captured thereby mitigating these interwoven pressures.⁴⁰ Nonetheless the interview itself is marked as a formal space as seen in (23) where the interviewee interrupts to ask, in English, which language they should use.

The interview structure marks the space as formal; however, previous to the interviewee’s question, both Spanish and mixed discourse were being used and, presumably, the latter would index informality.⁴¹ Yet, decisions regarding language use along the situational, social and transactional dimensions are always being made as speakers recruit then deploy the linguistic resources available to them. The exchange in (23) simply makes this explicit. In the same way that conflicting grammars are resolved in the patterns described above, the conflicting ideologies also find resolution despite what might be suggested by speakers’ use of mixed discourse to express negative attitudes to language mixing. Often, speaker reports of language use, for example, do not match their actual language use; likewise, mismatches are often found between empirical practices and overtly expressed ideological stances with respect to these practices (Labov, 1966; Romaine). However, rather than demonstrating the unreliability of self-reports, the contradictions illustrated in the data validate linguistic agency and the limits of hegemonic linguistic ideologies.

While policies and practices have concrete consequences for speakers ranging from linguistic discrimination to limited access to education and economic power, these ideologies must compete with the value placed on these discourse modes in multiple domains as well as community level rejection of these hegemonic ideologies. Standard forms of English and Spanish, for example would be rejected within informal spaces.⁴² Language among the speakers analyzed here is doing key social work in preserving culture (Charity Hudley), revealing or indexing identity (LePage and Tabouret-Keller) and co-defining the social landscape (Eckert). Speakers neither passively accept the two colonial languages nor do they seem to aim to master standard varieties, but rather, they demonstrate linguistic agency in reshaping them to meet their communicative goals along social and situational dimensions. Speakers even incorporate Kriol, a low prestige language not associated with their ethnic identity, into their linguistic repertoire and identity. This adoption of Kriol is not necessarily the adoption of a black identity, as this is rejected, but rather it is an expression of Belizean identity. Recall that in the late 70’s Kriol was becoming a national language marking Belizean identity. Despite racialized ideologies which constrain its use and the mythology of Creoles as rightful inheritors, Kriol became a good candidate for a unifying symbol of Belize on the road to independence. First, it is a language that was created out of the same sociohistorical circumstances from which the nation of Belize emerged. Second, it is the only language not shared by other nation-states. Garifuna, Yucatec, Mopan, Ketchi, and Spanish are all spoken in other countries but Kriol is not. Finally, it could also symbolize an identity independent of Britain but different from the more powerful Spanish-speaking country, Guatemala, laying claim to it.⁴³

Both the use of mixed discourse as well as its persistence as a community discourse mode reflects how speakers resolve conflicting ideologies and grammars. This is in spite of and in direct opposition to all that mitigates against their language use, i.e., educational policy, monoglossic ideologies and the privileging of English. Consequently, empirical practices of language mixing are not solely sites of grammatical encounter but sites of resistance. Perhaps this is best exemplified by the speaker who uses mixed discourse to express metalinguistic commentary proscribing its use: no tienes que hacerlo mix. It should be noted, however, that not all language mixing is resistance, as similar language mixing practices take place in other contexts but these do not necessarily reflect resistance. Spanish-English mixing, for example, can index elite status in Puerto Rico (Pérez Casas) and, similarly, a “fresa” identity in Mexico (Holguín Mendoza) because in these cases, mixed discourse marks the economic access associated with the acquisition of English. Thus, while all language mixing practices challenge purist notions, they are not always cases of linguistic resistance as this is dependent on the particularities of the linguistic semiotic landscape and how speakers position themselves and are positioned in it.

Discussion

Much of the literature on Spanish-English contact and mixed discourse has focussed on the grammatical constraints of language mixing and where ideologies are examined, consideration is often limited to language attitudes. The empirical linguistic practices of high language mixers discussed here have been shown to entail much more than resolving the combinatorial intricacies of the simultaneous use of more than one language. Each language, Spanish, English and Kriol, as well as mixed discourse, was shown to be tied to particular attitudes, hegemonic ideologies and shifting ethnoracial identities, all of which inform linguistic choices and practices.

It is essential to expand analyses of language ideologies to include state sanctioned rhetoric which codifies and imposes norms on language use. For example, the course policies requiring standard English and ministry policies which recognize the importance of Spanish discussed above, both embody monoglossic ideologies. The former insists on monolingual English and the latter on monolingual (standard) Spanish rather than local varieties. Likewise, the language of government and education is English which effectively minoritizes other national languages. The only mention of Kriol in educational policy is as a home language, as a vehicle to acquiring subject content (e.g. math) or standard English and as a variety to be used only sparingly. Additionally, the view that Kriol is an impediment to the acquisition of English is embedded in educational policies. Yet Kriol is recruited to serve a unifying role for the nation-state despite exclusion from economic and educational spheres. By including these aspects in the analysis of language use, it becomes possible to elucidate how a language like Kriol, racialized as Black, historically marginalized and potentially rejected by at least some Mestizo speakers, becomes part of the linguistic practices and identity for the Mestizo Spanish-speakers discussed above. In Belize, the rhetorical re-crafting of Kriol as part of a nationalist project allowed it to be reinterpreted in such a way that it ceases to be solely the language of Creole people. As a consequence, it becomes available to the Spanish speakers discussed here, all of whom were born in or grew up in post-independent Belize. This pattern of language use differs from other Spanish-Creole contexts along the Central American Coast where Spanish-English-Creole multilingualism is both geographically and ethnoracially constrained and where, at least some Creoles may be linguistically vulnerable (Leung and Loschky). In addition, the analysis of linguistic practices together with individual attitudes and state ideologies has shown that community vernaculars hold strong social significance and are used in multiple domains which together enable resistance to the hegemonic ideologies.

Thus, while it is well-established in the literature that intrasentential codeswitching reflects a high level of bilingual proficiency (Bullock and Toribio; Otheguy and Stern; Poplack), as has been demonstrated in the preceding discussion, the empirical practices of the Belizean Spanish-speakers within this landscape suggests additional competencies. To describe these speakers solely as bilingual or multilingual would simply reference knowledge of multiple languages, traditionally viewed as multiple, discrete monolingualisms (cf. Grosjean). This view is not consistent with their linguistic behaviour. Plurilingualism, on the other hand, refers to the speakers’ complex linguistic and cultural practices; dynamic and evolving repertoire; and ability to navigate various contexts which may require different modes of communication (Galante 2). This navigation is made explicit when the speaker above verbalizes her negotiation of language and situation by asking which language to use in the interview. Plurilingualism moves beyond communicative and grammatical competence and, additionally, places no emphasis on balanced or equal proficiency in the speakers’ languages. This approach accounts more accurately for how speakers may enact agency and meaning-making along multiple dimensions within the semiotic landscape. While first theorized within the European context, in fact, plurilingualism may be useful in describing the dynamics of linguistic practices in Creolophone societies and not just the Spanish-English-Kriol context described here.

As Belize moved through stages from colony to independent nation followed by stages of major migrations (to and from Belize) and developing infrastructure, the linguistic semiotic landscape also transformed. This paper has provided some insight into developments in Belize since LePage’s sociolinguistic surveys were carried out in the 70’s (LePage and Tabouret-Keller). Forty years later the observations regarding Belize Kriol appear to have played out as it is strongly associated with a pan-Belizean identity, to the point of being adapted into the mixed discourse of Spanish speakers. In addition, the shifts to Kriol first observed by LePage and Tabouret-Keller, along with the shift to English, may be quite advanced in some Spanish-speaking communities. Fuller Medina (2016, 259), for example, reported difficulty locating native speakers of NBS under the age of 60 in one northern community.⁴⁴ This, despite the strong vitality reflected in the census for Spanish, making clear that these global measures mask local realities. Further research is needed in this area to determine the vitality of Belizean varieties of Spanish as well as the extent of shift towards Kriol. Likewise, more research is needed on other languages in Belize, particularly indigenous languages, in order to more fully understand the semiotic landscape given that the nationalist project created one official language and promoted another as a national language leaving out various ethnolinguistic groups. In addition, in contrast to the speakers analyzed in this paper, other ethnolinguistic groups in Belize may or may not index Central American and/or Caribbean identities through their linguistic practices. As Belize moves forward in asserting its autonomy and identity, renewed migrations from the rest of Central America are taking place (Marshall and Corrigan), indigenous rights are being rightly asserted (Ramos), and Guatemala’s territorial claim may finally be addressed in the International Court of Justice (Caribbean Council). It remains to be seen how the semiotic landscape may once again transform.

Acknowledgments

I am grateful to Shelome Gooden, Nicole Ramsey and anonymous reviewers whose valuable comments helped to improve the current paper. Earlier versions of this work also benefited from helpful discussions with Vikram Tamboli, Karina Alma and the audience at a 2019 invited lecture I gave (co-sponsored by the UCLA Central American Studies Working Group and Latin American Institute). Any remaining errors are my own. This research was supported by the Social Sciences and Humanities Research Council of Canada (award # 767-2007-0686).

References

Publication Dates

History