Identity, language socialization, and family language policies in dialect contact: the case of Argentinean immigrants in Malaga, Spain

5.1 Language socialization at home: what children do with language and what their mothers think about it

The ideologies or language attitudes that parents have about the two varieties in contact shape language policies at home which are, in a way, one of the biggest pressures our younger immigrants can experience: family language policies are important as they shape children’s developmental trajectories, connect in significant ways with children’s academic achievement at school, and collectively determine the maintenance of the D1-BAS variety.

Following Bürki (2019: 197), transmission of a monoglossic and purist ideology is usually given by mothers, who are also transmitters of the linguistic norm.^[3] After analyzing our data, we found two sets of mothers with very different language ideologies or policies: (a) those who corrected their children’s linguistic performance when they diverged from the home vernacular variety (D1-BAS) and (b) those who understood that accommodation or acquisition of a new variety is “normal” or even desirable when moving to another country.

Regarding the group of mothers, Purist Caregivers, who corrected their children’s linguistic performance at home, we must first consider that acceptance, discrimination, and stigmatization issues of less socially accepted varieties of language are important for this group (language attitudes). The Andalusian variety of Malaga (D2-MAL) is perceived by these Purist Caregivers as incorrect, inappropriate, and less desirable for their children to acquire than their home dialect (D1-BAS). Many of them commented that they did not want their children to adopt Andalusian features of speaking (above all, phonological features) due to various reasons. On the one hand, they argued that adoption of phonological and some lexical features would impair mutual intelligibility in normal day-to-day interactions within the household; on the other hand, it would make the children sound uneducated or rough – which clearly reflects negative attitudes towards the Andalusian variety. Therefore, Purist Caregivers decided that D2-MAL was not socially or linguistically prestigious enough to be inserted within the household, and they forbid its use, as we see in the following examples:

Inf. 64. ARG.

[00:21:59] Es que es muy raro, es muy raro […] Y yo en mi casa las corrijo (habla de sus hijas), en mi casa no se habla […] No se habla malagueño […].

[00:22:08] Ajá.

[00:22:09] En mi casa se habla argentino.

[00:21:59] It’s all very weird […] I correct them at home (speaking about their daughters), at home we don’t speak Malaguenean.

[00:22:08] Aha.

[00:22:09] At home we speak Argentinean.

Inf. 101. ARG.

[00:33:24] A veces le digo a F (su hija): “Vos no hablarás así, ¿no?” (se refiere a si habla como malagueña). F que si algún día te escucho […] ay, yo te juro que me pongo re nerviosa con eso […] es que los malagueños hablan mal y cuando yo la escucho a F que habla como malagueña no, no puedo con eso […].

[00:33:24] Sometimes I tell F (her daughter): “You won’t talk like that, will you?” (She is referring here to her daughter speaking like someone from Malaga). “F. if I hear you talking like this” […] I can’t, I swear to you that this makes me so nervous […] because people of Malaga speak badly and incorrectly and when I listen to F. speaking like someone from Malaga, I can’t, I can’t stand that.

Other mothers or caregivers, however, accepted and considered the acquisition of D2-MAL as normal, and they did not forbid its use at home. These Tolerant Caregivers did not consider this new variety as incorrect or less desirable; quite the contrary, they perceived this new variety as new linguistic capital ^[4] (Bourdieu 1991; Sankoff and Laberge 1978), which would allow their children to adapt better to the new environment and have fuller or better experiences in Malaga.

Inf. 50. AMAL.

[00:14:30] Hay madres que no les importa como hablan sus hijos, pero había madres que sí les importa y les decían: “No, tú en casa hablas como un argentino”.

Entrevistadora: ¿Y qué opinas de eso? ¿Quieres que en tu casa hablen argentino?

[00:14:41] No […] A mí me encanta que hablen como malagueños […] Me encantaría hablar yo […] Como española, pero no me sale […].

[00:14:30] There are mothers who do not care how their children speak, but there were mothers that did care about that and told their children: “No, you have to speak as an Argentinean at home”.

Interviewer: And how do you feel about that? Do you prefer that your children speak Argentinean at home?

[00:14:41] No […] I love that my children speak as Malagueneans […] I would really love to speak like that […] like a Spaniard, but I’m not able to do that […].

Inf. 59. ARG.

[00:15:40] Yo creo que se sienten cómodas hablando como argentinas con nosotros (habla de sus hijas) […] A la mayor le dije: “Adáptate, no te diferencies. Mejor que seas igual” (en la forma de hablar). Pero no, no la forzábamos para nada a hablar como argentina en casa. Al revés, al revés.

[00:15:40] I think that they feel comfortable speaking to us in Argentinean (she is speaking about their daughters). I told the eldest: “Adapt yourself, don’t be different. It’s better if you speak in the same way”. But no, we never forced them to speak Argentinean at home, quite the contrary.

These ideologies or different linguistic attitudes towards D2-MAL transcend the family circle and have a big impact on the degree of acquisition of the new variety both by children and their mothers. It is very important to point out that we were also able to interview some of the fathers of these children, but none of them seemed to be interested or commented about language ideologies or policies at home.

Therefore, on the one side, we extracted information about Purist Caregivers. They usually were the children’s mothers who pressured them to speak D1-BAS at home, considered themselves exclusively as Argentineans. They did not accept for themselves a mixed identity and took part in close-knit social networks with Argentineans, strenuously rejecting as objectionable those Argentinean immigrants that accommodate to D2-MAL, did not attend any kind of formal education in Malaga since they arrived there when they were more than 18 years old, and hence, had the firm idea of returning to Argentina in the near future. These biographic and mesosocial characteristics of the mothers could be responsible for the bi-dialectal behavior of their children (von Essen 2020a, 2021). It is also relevant to point out that these mothers belong to a group we labeled as the Argentos (von Essen 2020a: 34, von Essen 2021: 304), who seem to be far from accommodating to the southern Spanish variety and are uni-dialectal, divergent, and conservative of their original variety (D1-BAS).

On the other side, Tolerant Caregivers were usually the children’s mothers who did not force their children to speak Argentinean (D1-BAS) at home, considered themselves integrated in Malaga, had social networks that included Andalusian as well as Argentinean speakers, did not reject immigrants that accommodate to D2-MAL, attended perhaps some kind of formal education in Malaga, as they arrived in Spain when they were more than 18 years old, and did not envisage returning to Argentina. These biographic and mesosocial characteristics could be underlying the relatively high degree of their children’s acquisition of D2-MAL linguistic features. Children of Tolerant Caregivers showcased monodialectal patterns of linguistic use; they “sounded” very close to how a speaker born and raised in Malaga would speak. Most of these caregivers belonged to a group we labeled as the Amalgamados (von Essen 2021: 304), who seemed to be halfway in their accommodation process to the southern Spanish variety of Malaga (D2-MAL).

Table 1 shows how each of these groups are connected and how the speakers are related (mothers and children): (a) Purist Caregivers are conservative mothers who forbid D2-MAL use at home or in any other social interactions (n = 2); (b) Bidialectal Children (BD Children) are the most sensitive group to the interviewerer effect: they show evident patterns of style-shifting and are, stricto sensu, bidialectal. In other words, they accept and abide by the norms of D2-MAL whenever only the Spanish interviewer is present, but switch to the D1-BAS norms as soon as the Argentinean interviewer takes over the interaction (n = 2); (c) Tolerant Caregivers are a group of mothers who permit D2-MAL use at home (n = 5); (d) Nuevos Malagueños Children (NM Children, n = 6) have clearly accepted the Malaga norms of use (D2-MAL) and are, in fact, uni-dialectal.

As for how these characteristics affected immigrants’ accommodation to Malaga linguistic features, Figure 1 and Table 2 show a comparison between frequency of use of several Malaga variables by four groups of children and caregivers as defined above.

Figure 1:

Percentage of D2-MAL acquisition among different immigrant children and their mothers.

First, percentage of acquisition of coronal fricative /θ^s/ phonemic split, which is one of the most salient features of the Andalusian middle-class variety (Villena-Ponsoda et al. 2022). When comparing both phonological inventories (D1-BAS vs. D2-MAL), the latter “[…] includes the phonemic split and reallocation of the previous merged phoneme /θ^s/ resulting in sibilant /s/ versus non-sibilant /θ/ contrast (Villena-Ponsoda 1996, 2018; Villena-Ponsoda and Vida-Castro 2017: 131). […] This split or phonemic contrast of the Andalusian coronal fricative /θ^s/ known as distinción is the result of convergence towards the Castilian Spanish northern varieties, which preserved contrast between word sets with (inter) dental /θ/ caza [‘kaθa] ‘hunting’ and alveolo-palatal /s/ [‘kasa] casa ‘house’. Distinción, on one hand, is very common among highly educated middle-class speakers of D2-MAL and, on the other hand, it is unthinkable in D1-BAS, where the pronunciation of /θ^s/ as [s̪] (seseo) is prestigious, and the split is unlikely to occur. So, word sets caza and casa are pronounced the same way, i.e. [‘kas̪a] (von Essen 2020a: 12–13)”. Hence, acquiring this split or phonemic contrast is related to accommodation or acquisition of D2-MAL.

Second, percentage of acquisition of the less noisy allophones of /ʝ/ or /ʒ/. When comparing both phonological inventories (D1-BAS vs. D2-MAL), intervocalic variants of /ʝ/ in D2-MAL “[…] tend to be approximant, open fricative allophones with a small index of noise like the voiced palatal approximant [ʝ̞] or the voiced palatal fricative [ʝ]. […] The variants of D1-BAS are the voiced postalveolar fricative [ʒ] and the voiceless postalveolar fricative [ʃ] […] which exhibit long duration, are tense, have high indexes of noise, and are often voiceless (von Essen 2020a: 11)”.^[5]

Third, under the category “Tú forms”, we contrast the percentage of acquisition of morpho-syntactic features, more specifically, of the pronoun paradigm and verbal forms used in D2-MAL: whereas in D2-MAL, the use of the second person singular pronoun you is tú (second person singular in Castilian Spanish and many other varieties), in D1-BAS. The same category is produced as vos. Also, some corresponding verbal forms change, so whereas for D1-BAS you eat regularly occurs as vos comés (voseo), the D2-MAL unchangeable solution is tú comes (tuteo).^[6]

Finally, labelled as “Malaga words”, we show frequency of acquisition of the lexical features of D2-MAL that are salient and/or could affect intelligibility (von Essen forthcoming). Our results are in line with Chambers’ (1992: 677) who considers that “[…] lexical variants are acquired faster than pronunciation and phonological variants”. This is particularly true for all the groups included here, where we establish what percentage of D2-MAL lexicon is produced by the immigrants during the semi-guided interviews.

As we infer from Table 2 and Figure 1, ideologies, pressures, and family linguistic policies promoted by caregivers at home, on the one hand, and acquisition of D2-MAL variants by both groups, on the other hand, could play a relevant role in the children’s dialect acquisition. All these caregivers, mostly, the children’s mothers, are users of a linguistic variety that is their country’s standard (D1-BAS). Therefore, they are supposed to have a strongly hierarchical national ideology in which their variety of Spanish is at the top of the linguistic scale. The Andalusian variety is not the national standard variety of Spain, and these mothers are aware of this situation.

It is also important to point out that whereas the differences in accommodation of the children are related to their mothers’ sociolinguistic stances, these children (NM and BD) have extremely similar biographic and mesosocial characteristics: they arrived when they were less than 14 years old (most of them between the ages of 7 and 14), they attended elementary school in Malaga, they have social networks composed mostly of speakers from Malaga, they have positive language attitudes towards D2-MAL, and they do not plan to return to Argentina. One of the explanations for these differences between NM-children and BD-children could lie in the fact that NM-children are not prohibited to adopt D2-MAL, whereas BD-children are strongly advised not to adopt nor use this new variety at home (or in any other social interactions). These differences are also related to the background of the interviewer (see Section 4). When comparing results of the Spanish interviewer (cf. Table 2, only Spanish interviewer), NM-Children and BD-Children show very similar linguistic behaviors, especially if we look at the morpho-syntactic and lexical level, which show no significant differences.

In sum, if we look at results from Table 2 and Figure 1, above, we can conclude that:

1. Although these results account for a small number of speakers (15 speakers out of 72), dependence between linguistic variation and the speaker’s inclusion in one of the groups (NM Children vs. BD Children vs. Tolerant Caregivers vs. Purist Caregivers) is very high (Cramer’s V): 0.943 for [ʝ] allophones, 0.829 for /θ^s/split, 0.738 for “Tú forms”, and 0.517 for lexical acquisition of D2-MAL features.

2. We also establish, as in previous results (von Essen 2020a: 35, 2021: 213), an accommodation hierarchy (cf. Table 2): lexis is the feature most permeable to accommodation (68.8% of “Malaga words”). Morpho-syntactical features seem also to be easier to master (58.8% of “Tú forms”). Phonological features are the hardest to acquire, and accommodation depends on their specific complexity: splitting of /θ/corresponds to a complex rule phenomenon, while open [ʝ] realization is a simple rule phenomenon.^[7]

3. All groups overall show some degree of accommodation to D2-MAL features. However, the exception being, very likely, the splitting of /θ^s/.^[8]

4. Differences between age groups are apparent. Both groups of caregivers tend to acquire fewer phonological features than any group of children, no matter their ideologies or sociolinguistic stances. Of course, this should be related to late contact with D2-MAL.

5. Morpho-syntactic and, above all, lexical accommodation, is quite common among caregivers (both, purist and tolerant), no matter what they “think they do”. This is particularly interesting for the Purist Caregivers, since what they think about language, i.e. that D2-MAL must be forbidden at home since it is not prestigious enough and appropriate only for uneducated speakers, on one hand, and what they do with language they actually acquire and produce D2-MAL features, on the other hand, shows that their language interventions at home, i.e. what the mothers try to do about language, do not always match their own linguistic performance. In fact, Purist Caregivers’ linguistic behavior involves a considerable degree of accommodation (perhaps unconscious) of morpho-syntactical (18.8%) and lexical (40.1%) D2-MAL features.

6. Purist Caregivers’ policies could lead, or even, unconsciously force, bidialectal children (BD Children), on the one hand, to adopt less convergent or even divergent linguistic behaviors towards D2-MAL. Tolerant Caregivers’ more adaptative attitudes, on the other hand, could be responsible, among other small-scale variables (von Essen 2020a, 2021), for the high degree of acquisition of D2-MAL variants by the New Malaguenian children (Nuevos Malagueños, NM Children). This is particularly clear if we compare the rate of acquisition of D2-MAL features by BD Children and NM Children. In spite of the differences (see Table 2), it should be pointed out that BD Children tend to react to the interviewer’s background (i.e. presence of an Argentinean speaker during the interview). When speaking to the Malaga interviewer, they produced similar results to those of the NM Children (von Essen 2020a, 2021). One would suggest, then, that the second generation of immigrants (BD Children and NM Children) understand varieties as communication tools, so that every element is equally valid, depending on the context, to help fulfill communicative interactions (Bürki 2019). This is particularly clear for the NM children group, who have acquired more D2-MAL than D1-BAS linguistic features (97.3% lexis, 98.5% morpho-syntax, 99.4% of [ʝ] and 84% of [θ], for an overall frequency of 91.7% of phonological accommodation).

5.2 Language socialization at school: what children experience during their formal education in Malaga

According to García and Torres-Guevara (2009: 182), in educational activities that take place at school and the classroom, language becomes paramount for three reasons: (a) schooling often takes place in the language of power in a nation state (here, D2-MAL); (b) language is the medium through which instruction takes place; and (c) it is an important school subject.

Here, we examined personal experiences at school of the immigrants of our sample and the impact of formal education in their degree of accommodation towards D2-MAL. To do so, we observed and analyzed data from our whole sample (n = 72). Based on the information that the immigrants shared about their experiences, we were able to portray how relationships between young immigrants and their classmates worked out and how teachers dealt with this issue. In order to analyze the impact of school on the immigrant children’s accommodation and acquisition, we focused on the same linguistic levels considered so far (phonology, morpho-syntax, lexis), dividing the speakers into three groups:

(a) those who attended elementary school in Spain (age of arrival between 3 and 14) (n = 16); (b) those who attended Spanish high school or university (age of arrival between 14 and 18) (n = 16); and (c) those who did not attend any kind of formal education in Spain and arrived in Malaga when they were more than 18 years old (n = 40).

In previous results from this sample (von Essen 2020c: 323–333; 2021: 260), we found that the immigrants’ age of arrival and their access to formal education in Spain had a moderate effect in explaining linguistic accommodation. More precisely, age of arrival explained 68,3% of /θ^s/variation, whereas access to formal education explained 60,8% of /ʝ/variation, 34,3% of “Tú forms” and 31,2% of lexical variation (cf. Table 3).

However, despite these, to a certain extent, remarkable results, we see as necessary to broaden the information about the personal experiences of the speakers, who are culturally and, somewhat, linguistically very close to the recipient community. None of them commented that they are perceived as a threat (García Sánchez 2009), nor are their cultural practices extremely different from those of the D2-MAL speakers. Some of them exclusively speak, at least with their parents and family, a different variety of Spanish (D1-BAS), which was sometimes a problem for themselves, for some of their classmates, and, even, for some teachers, who ended up correcting their linguistic productions when they diverged from the local variety (D2-MAL).

Although we found different experiences among the immigrants that arrived between the ages of 3 and 14 years old, we found some groups with common experiences as to how they were treated by their classmates during elementary school. It seems that accommodation pressure at this stage was greater than that felt further on in formal education. Some of the children from this group (n = 8) mentioned that they all felt lack of attention from their parents to language planning at home, i.e. parents never openly pressured them to speak in a certain way, which could lead to language-free accommodation to the dominant variety in the speech community. All of them mentioned, however, two stages in their acquisition of D2-MAL, strongly dependent on school attendance. First, they started changing some words (for example, they picked fresa for strawberry and left out frutilla), because they needed to be understood by their teachers and classmates. Second, this lexical accommodation derived or even pushed unconscious accent acquisition:

Inf. 92. NM.

[00:17:40] Los primeros meses sí en el colegio tenía un acento muy cerrado y se metían un poco conmigo algunas personas y tal. Pero no sé, a los cuatro meses o así, empecé a cambiar la manera de hablar porque si no, no me entendían.

[00:18:01] ¿Tú crees que el colegio hizo que tú empezaras a cambiar tu acento? (Entrevistador)

[00:18:02] Sí. A lo mejor usaba en vez de fibra, rotulador; o cosas así. Porque estás en clase y necesitas saberlo. Y ya luego, de hablar así y de escuchar en el colegio se te pega y se te cambia solo el acento.

The first few months at school, I did have a very strong accent (D1-BAS), and some classmates made fun of me a little and such. But I don’t know, after four months or so, I started to change the way I spoke, because otherwise they wouldn’t understand me.

[00:18:01] Do you think that school made you start to change your accent? (Interviewer)

[00:18:02] Yes. Maybe during class, I used rotulador (“marker” in Spain) instead of fibra (“marker” in Argentina); or things like this. Because you’re in class and you need to know that. And then, from talking like that and listening at school, the new accent sticks to you, and it changes.

A second group of speakers (n = 4) commented that during the first period at elementary school, they were assigned an “interpreter–translator” classmate, that is, another immigrant child who had been living in Malaga for a longer period of time. This “interpreter-translator” child oversaw the teacher’s explanation of situations, words, expressions, or lexical items that these younger immigrants did not understand during the first year of school. In a way, they were a link between the newest immigrant and the teacher.

This “interpreter–translator” child was never negatively perceived by our immigrants. When asked about this, they explained that either teachers maybe decided to use this link because they were not able to stop the class to attend the new immigrant students’ doubts, or perhaps they tried to ease new students with immigrant background into integration. Although the experience with the help of this “interpreter–translator” child was perceived positively by all the immigrant of this group, the decision of these teachers could alienate and even separate children in the classroom into two groups: native versus immigrants. At the end of the day, this decision could have been the consequence of defective preparation of teachers on how to deal with different languages or linguistic varieties in the classroom and would have caused alienation rather than integration. In the following excerpt, we can get a better insight into how this teacher’s strategy worked out and how both immigrants and the rest of the classroom perceived this link:

Inf. 84. NM.

[00:06:05] Y son palabras diferentes, y tú cuando eres tan chica ¡no te enteras de nada! Yo no me enteraba ni de la mitad, te lo juro, lo que pasa, que había un niño de mi clase que se llamaba Oki, que me cae muy bien, que llevaba más tiempo allí, y como que me arropó, y me apoyó y me explicaba y tal y cual […] la verdad que es que la gente de mi clase, bueno yo no sé, pero a mí trataron muy bien, vamos hice amistades muy rápido […] Me costó eso, no es el idioma es como la lengua, ¿sabes? Que no es la misma. Sí, pero que al año o dos yo ya estaba hablando andaluza. O sea que es que los niños aprenden muy rápido, hmm.

[00:06:05] And they are different words, and when you are so young, you don’t understand anything! I didn’t even understand the half of it, I swear (she is speaking about the first months at school). What happened is that there was a boy in my class named Oki, who I like very much, who had been living there longer, and he kind of tucked me in, and supported me and he explained to me everything […] the truth is that the people in my class, well I don’t know, but they treated me very well, I made friends very quickly […] What was difficult for me, is not the language, it is kind of the way of speaking. Which is not the same. Yes, but after a year or two I was already speaking Andalusian. Children learn very quickly, hmm.

One of the children (a boy) commented that he used sports, particularly football, to better integrate among the children of the new classroom:

Inf. 87. BD.

[00:06:09] Me integré un poquito así. Con el fútbol […] Sí. Porque no sé porque la sociedad esta, al que se le da bien el fútbol, está como un poco mejor valorado que uno que no haga deporte en el colegio […] Entonces al principio, yo me muero por el fútbol y los primeros días como era nuevo no me dejaban jugar, porque es que yo me acuerdo. Y claro, cuando uno de mis mejores amigos me dijo que me probarían a jugar. Claro, los partí a todos, entonces ellos dijeron “uhh, uhh”. Que fue una manera de relacionarme, no te voy a mentir. No es que tuviera problemas para relacionarme, pero […] No sé, un chaval que viene de […] Que llega sin amigos, con otro acento […] Me ayudó mucho a integrarme.

[00:06:09] I integrated a little like that. With football […] Yes. Because I don’t know why this society, whoever is good at football is somewhat better valued than someone who doesn’t play sports at school […] So at first, I was dying to play football, and the first few days as I was the new student, they didn’t let me play. And, of course, when one of my best friends told me that they would try me out to play and I played great and broke all of them, they said “uhh, uhh”. Football was a way of relating to others. It’s not that I had problems relating to others, but […] I don’t know, a kid who comes from […] Who comes with no friends, with a different accent […] Football helped me integrate a lot.

However, not all experiences were positive. Among the children that arrived at the age of three, we found a female speaker that was forced and actively corrected by her teachers to speak the D2-MAL variety, not only in the classroom but also at home. Otherwise, she commented, she would have been excluded from her new school, as we see in the following example:

Inf. 82. BD.

[00:06:35]

Porque yo vine con tres años, y yo hablaba argentino, yo llegué aquí y empecé a hablar, o sea yo en el colegio hablaba argentino. Lo que pasa es que la profesora me decía que o hablaba español o me echaban del colegio. Me obligaba a hablar español […] y mi madre fue a hablar y le dijo con la directora del colegio, que le dijo que yo tenía que empezar a hablar así porque los chicos tan chicos de mi clase no me entendían […] Porque el argentino y el español tienen por ejemplo palabras diferentes, o las mismas palabras que allí significan diferente. Pero ya ahora no, o hablo todo argentino o todo español. Hablo argentino con los argentinos y malagueño con los malagueños.

Because I came when I was three years old, and I spoke Argentinean (D1-BAS), I came here and began to speak, that is, I spoke Argentinean at school. What happened is that the teacher told me that either I spoke Castilian Spanish (D2-MAL) or they would kick me out of school. She forced me to speak Castilian Spanish […] and my mother went to talk to the school principal, who told her that I had to start speaking D2-MAL because my classmates did not understand me […] Because Argentinean Spanish (D1-BAS) and Castilian Spanish have, for example, different words, or the same words that mean something different. But not anymore, either I speak Argentinean or Castilian Spanish. I speak Argentinean with Argentineans and Malaguenean (D2-MAL) with the people of Malaga.

We were also able to interview her mother (Inf. 101 ARG), who commented on this very traumatic experience with the school principal and her daughter’s teachers, and what it meant for her daughter and for her:

Inf. 101. ARG.

Me decía la maestra a mí, que yo en casa hablara, que tratara de hablar andaluz. Por qué, ¿qué pasaba con F.? Es muy extrovertida. Entonces ella, hablaba todo el tiempo, y la maestra, iban todos los chiquitos ‘¡ay F. dijo canilla! ¿Qué es la canilla? Y bueno, canilla es grifo, almohada es un cojín y la maestra ya estaba harta, al mes, harta de esa situación. Y entonces me llamó y me dijo que ella la iba a corregir a F. en el colegio, y que yo la corrigiera en casa […] Y yo imaginate […] Le dije que yo en casa no lo iba a hacer, que era un trabajo de ella si quería […] Pero ella tiene eso tiene que cambia su forma de hablar y habla malagueño con andaluces y argentino con argentinos. Pero ¿por qué? Porque en el colegio la obligaban a hablar andaluz y en casa hablaba argentino.

The teacher told me to try to speak Andalusian at home. Why, what happened to F.? She is very outgoing. So, she talked all the time, and the classmates asked the teacher ‘oh F. said canilla (“tap” in Argentina instead of grifo in Spain) What is the canilla?’ And well, canilla is a tap, and the teacher was fed up, after a month, fed up with that situation. And then she called me and told me that she was going to correct F. at school, and that I should correct her at home […] I told her that I was not going to do it at home, that it was her job if she wanted to do that at school […] But she has that, she changed her way of speaking and now she speaks D2-MAL with Andalusians and D1-BAS with Argentineans. But why? Because she was forced at school to speak Andalusian, and at home, she spoke Argentinean.

Although this was an isolated case and not the common scenario, the teachers’ ideologies on correctness, their negative language attitudes towards different linguistic varieties, and their influence on students has been largely confirmed in literature: corrective feedback in the classroom is consequential for students’ learning and affects performance (García Sánchez 2009: 179).

If we consider immigrants who attended high school or the university (n = 16), we find a common denominator in this group: attendance at high school or the university had a big impact in the speakers’ degree of accommodation. As we commented previously, only in late adolescence (between 17 and 18 years), do speakers come to be aware of the social significance of prestigious variants both of their own variety (D1-BAS) and of other varieties of Spanish (D2-MAL). Therefore, almost everyone commented on how specific experiences and memories regarding the different varieties they had contact with, as well as their underlying ideologies, shaped the way they speak today, as we observe in the following example:

Inf. 49. Imp.^[9]

Y poco a poco, al empezar sobre todo en la Universidad yo creo que me influyó mucho. Porque me acuerdo que exponíamos trabajos y al final me daba cuenta que nadie se centraba en lo que estaba diciendo, sino que se fijaban en mi acento. Y eso me empezó a molestar, mucho, muchísimo […] Y a partir de ahí bueno, forzado empecé a cambiar la forma de hablar, en un principio fue forzado, ahora es natural. A ver si cambiando no me interrumpen cuando hablo o se centran más en lo que digo en los trabajos.

And little by little, starting especially at university, I think it influenced me a lot. Because I remember that we had a presentation and, in the end, I realized that nobody focused on what I was saying, but rather they noticed my accent. And that started to bother me, a lot. And from then on, well, in the beginning like forced, I began to change the way I speak, at first it was forced, now it’s natural. Let’s see if by changing my accent, they don’t interrupt me, and they focus more on what I say, not how I say it.

As to the effect of language socialization at school on linguistic accommodation or acquisition (cf. Table 4 and Figure 2), we found that immigrants who attended elementary school in Malaga and were incorporated into the school system in early stages of language development, had bigger opportunities and, in fact, acquired a higher percentage of D2-MAL than D1-BAS features (see Section 5.1 above). However, attending high school or the university also had a moderate effect in accommodation, although for this group, acquisition of phonology seemed to advance at a slower pace than for other language components. Nevertheless, attendance at any kind of formal education in Spain was not a sine qua non condition to accommodate to the D2-MAL variety: speakers that were not exposed at all to formal education in Spain did also accommodate to D2-MAL.

Figure 2:

Impact of formal education in Malaga, Spain (n = 72) (Adapted from von Essen 2020c: 323–330).

This effect was particularly strong on phonological features; for example, in the case of /θ/ and /ʝ/, dependence of D2-MAL variants used by immigrants and formal education rates is relevant and attains Cramer’s V values of 0.602 and 0.608, respectively. This comes as no surprise given the fact that phonological features are faster acquired in younger stages of exposure to D2, and that elementary school takes place during the optimal age of the acquisition period. Chambers (1988, 1992, Kerswill (1994, 1996, Payne (1976, 1980, and Trudgill (1986), among many others, referred to these differences in acquisition by considering a critical period in acquisition: until the age of 14, children can certainly acquire phonological features, although the optimal age of acquisition is before the age of 7, especially for the so-called “complex-rule” phonological features, like the acquisition of /θ^s/ (von Essen 2021: 244). Therefore, the older the immigrants are exposed to D2 features, the smaller or lesser their phonological acquisition will be. As Chambers (1988: 689) suggested “[…] a person seven or under will almost certainly acquire a new dialect perfectly, while a person 14 or over almost certainly will not. In between those ages, people will vary”.