“My Child Is a Perfect Bilingual”: Cognition, Emotions, and Affectivity in Heritage Language Transmission

Abstract

One of the key questions of studies on heritage language (HL) transmission is which factors most likely foster the intergenerational transmission of HL and more saliently favor its acquisition in second-generation speakers. The present study explores the effect of the cognitive and affective disposition of first-generation speakers on the subjective HL proficiency level in the second generation of Russian-speaking immigrants in the town of Salamanca, central Spain. Based on a scalar questionnaire which enquires into the language practices, language attitudes and language motivations of the first-generation speakers, the study analyzes the effect of self-categorization, attitudes towards HL utility, and strategies of HL intergenerational transmission in ten mixed families. The main results of the study show that positive HL affectivity is key to assuring proficient HL acquisition in second-generation speakers, while negative HL affectivity systematically drives unbalanced Spanish–Russian bilingualism in children. The final results are consistent with those of other recent studies on affectivity in HL and suggest the importance of positive attitudes towards HL in its transmission.

1. Introduction

Heritage language (HL) speakers are assumed to be bilinguals in a non-restrictive sense of the term (Rothman 2007). However, it was already well described in studies on heritage language transmission that achieving full bilingual competence is the exception rather than the rule for second-generation migrants. Bilingualism in heritage speakers is frequently based on misbalanced proficiency of the majority language and heritage language (Montrul 2016) and is usually highly varying among heritage speakers in what concerns their language skills (Chevalier 2004). Despite the level of proficiency in inherited language, heritage speakers are frequently characterised as speakers lacking functional literacy (Rothman 2007), which means they have a restricted stylistic ability to use HL in all contexts. Indeed, it is common for heritage speakers themselves to be aware that their proficiency in the majority language (the language of the host country) is stronger than their skills in the HL (Chevalier 2004).

Unlike other naturalistic bilinguals, heritage language (HL) speakers’ bilingualism depends to a higher degree on social circumstances surrounding HL acquisition (Rothman 2007). Considering that heritage speakers are those “whose home language is much less present in their linguistic repertoire than the dominant language of their society” (Scontras et al. 2015), unbalanced HL development is closely linked to environmental factors of language acquisition during childhood.

The HL learner’s competence is conditioned by what Polinsky (2017) calls “the baseline”, or the language input transmitted by the first-generation immigrant to the second-generation speaker. But, apart from the language system itself, sociolinguistic factors like language ideologies, language loyalty, and attitudes to languages play an important role in what is actually transmitted from parents to children. At its core, HL is the result of a language shift from first-generation speakers (Polinsky and Kagan 2007), which makes it an essentially sociolinguistic phenomenon. Further, HL transmission directly correlates with language attitudes, one of the key factors in understanding linguistic functionality. Recent studies have put forward that the attitudes from both the first-generation speakers (Au and Oh 2005) and the second-generation speakers themselves (Boon and Polinsky 2015) predict the ultimate success of the latter in HL.

One of the key aspects underlying attitudes to HL is the specific combination of the HL and the majority language (ML), as well as their social values. The higher the prestige and the social value of HL, the more probable its intergenerational maintenance. To mention two illustrative examples, one can focus on two extensive HL communities in the United States: Chinese and Spanish-speaking. While Mandarin speakers share positive attitudes towards both their HL and ML, and consequently support Mandarin in HL promotion and transmission between generations (Wiley et al. 2008), some Spanish speakers in the same country exhibit negative attitudes towards their local varieties of HL (Achugar and Pessoa 2009), which gives rise to the replacement of Spanish by English (Montrul 2018).

The social value of a HL relies on different factors: the economic status and independence of the HL community, usually estimated in terms of professions and the dominating type of work; the social importance and relevance of the HL community, referring to the degree of its self-esteem as a historical entity; and the linguistic status of HL community, which is based on the general perception of HL (Balestra et al. 2008). From these three factors, linguistic status seems to be the most unstable in what concerns its possible impact on HL transmission. This is due to the direct dependence of linguistic status on the cognitive and affective disposition of first-generation speakers.

Starting out from this premise, the present study proposed to explore how the cognitive and affective disposition of first-generation speakers determine the strategies of HL transmission to second-generation speakers and, accordingly, the level of proficiency the latter achieve in HL. For this purpose, the present study focused on the Russian-speaking community living in the town of Salamanca, central Spain, which is characterized by some specific sociolinguistic features. Russian is one of the most numerous minority languages in Spain, with more than 50,000 immigrants living in Spain as of 2019 according to the Spanish National Statistics Institute. The share of the Russian-speaking population, however, is even higher in Spain if one takes into account that many immigrants arriving from post-Soviet countries other than Russia (Ukraine, Belarus, Moldavia, Georgia, etc.) may also be Russian-speaking. Some sources (Riazantsev and Grebeniuk 2014) sustain that immigrants from post-Soviet countries amount to almost a million speakers, and one can only estimate how many of them use Russian as their habitual language.

There have been several waves of immigration from post-Soviet countries to Spain since the dissolution of the USSR. While immediate post-dissolution immigration was characterized by high social and ethnic diversity and followed the purposes of immigration started during perestroika (Kopnina 2005), succeeding migratory waves obeyed purely economic aims first (Tishkov et al. 2005) and both economic and affective aims later. As a result, the Russian-speaking population follows different HL transmission strategies with second-generation learners: from explicitly reinforced maintenance of Russian as a HL (RHL) to total renunciation of this language for being difficult, unpopular, and unnecessary in future (this is a paraphrase of some opinions captured from Facebook pages of Russian-speaking communities in Spain).

In the small town of Salamanca, located two hours to the northwest from Madrid, the members of one of the Russian-speaking communities share many sociolinguistic features. This is a small community of young women who are married to Spanish-speaking men and have been residing in Spain for an average of 13 years. In addition, all of them belong to the same medium–high sociocultural level, assessed from their professions and educational level (for methodological background, see (Ghezzi and Sampedro Mella 2015)). However, their children vary significantly in what concerns proficiency in RHL: while some of them may be easily described as full bilinguals (Russian as heritage language and Spanish as majority language), others have notably limited proficiency in RHL.

Considering this background, in the present study it was hypothesized that HL transmission strategies might depend on the cognitive and affective aspects of first-generation speakers’ attitudes. There could be underlying psychological reasons which do not allow first-generation speakers to strongly support HL development in children, and, eventually, these reasons might be identified through a set of sociolinguistic characteristics related to cognition (what kind of speaker am I?), emotions (does my language really serve me for anything?), and affectivity (will my language really serve my children for anything?). The perspective adopted in this work is in line with previous research which showed a strong relationship between positive attitudes towards HL in the family and HL proficiency (Guardado 2018b) and between positive HL attitudes of the second-generation speakers and their language development (Kagan 2012). It goes, however, further in order to explore the link between HL attitudes in the first generation and HL proficiency in the second generation as deeply related factors.

Indeed, the present research could prove this assumption by an attitudes study in the mentioned Russian-speaking community. The members of this community adopt different HL transmission strategies because of the different cognitive, emotional, and affective dispositions they have towards RHL. Accordingly, these psycholinguistic factors seem crucial for understanding both the success and the failure of HL transmission between generations.

2. Materials and Methods

Language loyalty, also known as intergenerational transmission and considered the background for HL maintenance, may be measured as the set of variables of mother tongue, use of language, language proficiency, and attitudes to languages (Rivera-Mills 2012). Following this claim, in this research it was considered that understanding all these aspects is crucial to finding out the strategies of HL transmission in the Russian-speaking community in Salamanca.

In order to collect necessary data, a sociolinguistic questionnaire was designed. The design of the questionnaire did not follow any previously used model. This decision was mainly driven by the needs of the study, which sought to measure three key sociolinguistic parameters: cognitive aspects, emotional aspects, and affective aspects of RHL transmission strategies altogether. Even so, the choice of the items for measuring these parameters relied on both sociolinguistic theory and the results from other studies, as will be described below.

Overall, the questionnaire included 34 scalar items organized into eight sets concerning:

• Personal data (4 items): The collection of these data is conventional in sociolinguistic research and plays a particular role in studies on social networks, where demographic variables refer to attributes of network members (Vetter 2011).
• Data for calculating the sociocultural level (4 items): Here, sociocultural level was considered in terms of a set of static traits which might drive language behavior in HL transmission. In many sociolinguistic studies, as He (2010) reviewed, sociocultural level was taken as a predictor of HL development and the varying degree of proficiency in the second generation. In the present research, sociocultural level was calculated as a variable measured by the profession of the participants, the profession of their parents, and the educational levels of both. This measurement is standard for sociolinguistic research (Ghezzi and Sampedro Mella 2015).
• Data for calculating the language self-assessment of first-generation speakers (4 items): Self-reporting of language abilities is considered to be a reliable measurement and is consistently used as a standard method of language assessment elsewhere (Kim and Chao 2010). Including language self-assessment is common in research on HL, where several key studies (see, for example, Riehl 2017) have analyzed first-generation language self-assessment and included it as a variable which determines HL transmission strategies.
• Data for calculating characteristics of the social network of first-generation speakers (5 items): Sociolinguistic studies put forward that the structure of the social network may deeply condition language uses. Dense social networks usually determine more uniform language practices, while loose (or diffuse) social networks imply non-homogeneous language practices (Blanco Canales 2000). For HL transmission in particular, dense social networks were argued to be responsible for better HL acquisition (see Shively 2018 for an overview).
• Data for analyzing language use in first-generation speakers (6 items): Self-reported language use is a central part of any sociolinguistic questionnaire, since it allows for contextualizing language development. Although many deeper assessment tests may be used for analyzing this aspect (see Polinsky 2017 for an overview), in the present study, self-reported language use in first-generation speakers, in particular, with their children and in the family context was considered sufficient for contextualizing RHL development in the second generation.
• Data for calculating the instrumental and integrative motivation of language use from first-generation speakers to second-generation speakers (4 items): As a key attitudinal factor, motivation is described as predictor of language attainment (Lu and Li 2008), both from the bottom-up perspective of the second generation and from the top-down perspective of the first generation. Frequently, HL parents show more integrative than instrumental motivation for language development in their children (Lindholm-Leary 2001), which might directly influence the way in which HL is transmitted and encouraged.
• Data for calculating the language competence of second-generation speakers (3 items): Proficiency in HL may be assessed in very different ways (see Polinsky 2017 for an overview). Self-reported language proficiency is one of the conventional methods in HL studies, since many studies have found this technique to be highly consistent with real language achievement (Mu 2016). In this research, language proficiency is assessed from reports from first-generation speakers.
• Data for analyzing explicitly stated HL transmission strategies (4 items): Finally, the questionnaire included items concerning HL transmission strategies from the first-generation speakers. Language management strategies proved to be key in the development of language proficiency in the second generation (Brehmer and Kurbangulova 2017) and thus play an important role in understanding language attainment.

The purpose of all included items consisted, on the one hand, of obtaining data for correct sociolinguistics classification of the participants and, on the other hand, of gathering data for cross-analysis of (a.) the cognitive, affective, and emotional characteristics of first-generation speakers and (b.) the RHL proficiency of second-generation speakers. In this way, verification was sought of the correlation between the language self-esteem and motivation of first-generation speakers and HL competence and transmission towards second-generation speakers.

The questionnaire was completed by 10 speakers with similar sociolinguistic profiles: all participants are young women, average age 39, with high education and enrolled in intellectual activity, with a long-term residence in Spain (

Table 1. Research sample.

Participant	Age	Education Level	Years in Spain
1	34	Tertiary	11
2	35	Tertiary	8
3	40	Tertiary	17
4	49	Tertiary	3
5	39	Tertiary	15
6	46	Tertiary	15
7	38	Tertiary	20
8	42	Tertiary	12
9	44	Tertiary	14
10	29	Tertiary	16
Average	39.6 (±5.98)		13.1 (±4.86)

). Participants were contacted via a “snowball” technique: new potential participants were found and recruited through social networks (a participant recommends his or her acquaintances to take part in the study), which was previously described as an effective method in sociolinguistic research of small communities (Milroy and Gordon 2003). The selection criteria for participating in the study were (a.) to have Russian as one’s mother tongue; (b.) to have a Spanish-speaking partner; (c.) to have been living in Spain for several years; and (d.) to have a child/children exposed to both languages at home. All subjects participated voluntarily and were aware of the purposes of the research and of that they could withdraw from the study at any time. Participation was anonymous and followed the principles of research ethics.

Apart from this, participants share other important sociolinguistic characteristics. All of them are married to Spanish-speaking men, of whom most (a.) either do not speak/do not understand Russian (40%) or (b.) understand some but do not speak any Russian (40%). Consequently, the home language policy of the participants is usually based on the simultaneous use of Russian and Spanish, and, in most cases, Spanish is used more frequently than Russian, although all participants started to use RHL with their children from birth.

Before the data are analyzed and discussed, it is necessary to comment on the small sample size used in this research. There are several reasons why only 10 participants took part in the study. First of all, the Russian-speaking community of Salamanca is not the biggest one in Spain; most Russian-speaking migrants live by the Mediterranean coast, while Central Spain has a smaller migrant population originating from post-Soviet countries. Secondly, the sociolinguistic profile this research was looking for—young Russian-speaking women married to Spanish-speaking men—is not the dominant one within the Russian-speaking community in Salamanca. In this respect, the present research resembles other studies on attitudes in HL transmission which were based on small samples (for example, Park and Sarkar 2008).

3. Results

On the whole, second-generation speakers are mainly characterized by their mothers as assymetric bilinguals, who share native-like comprehension (listening and reading) but restricted verbal production (speaking and writing) of RHL. In addition, some first-generation speakers underline the presence of a certain foreign accent in their children when speaking Russian. In general, second-generation RHL speakers are dominant bilinguals, some of whom tend to be receptive speakers of Russian. These general results were predictable, since limited exposure to RHL and the lack of formal instruction are hindering factors in HL intergenerational transmission.

Nonetheless, the overall perspective on RHL competence in second-generation speakers can be discussed in terms of internal variation. During their participation in the survey, some first-generation mothers mentioned that their children know Russian as well as their monolingual Russian-speaking peers; others complained that the RHL proficiency of their offspring either started to worsen or did not improve over the years. In order to check the influence of cognitive, affective, and emotional factors on this internal variation between second-generation speakers, their RHL proficiency was analyzed from all three perspectives separately.

3.1. Cognitive Characteristics of First-Generation Speakers and Language Competence in the Second Generation

Applying a cognitive filter to the general perspective on RHL competence in second-generation speakers makes an important difference. Cognition here implies the speaker’s understanding of oneself as a speaker and is primarily embodied in self-categorization parameters: an open category which defines the speaker’s identity and awareness in interaction (Ziegler 2011).

Four variables from the questionnaire were used for calculating the self-categorization of the first-generation speakers of RHL: (a.) language self-assessment in Spanish (What is your proficiency in Spanish?); (b.) language self-assessment in Russian (Did your competence in Russian change since you moved to Spain?); (c.) language self-assessment as a Russian–Spanish bilingual (What is your language proficiency as a Spanish–Russian bilingual?); and (d.) attitudes towards one’s language proficiency as a Russian–Spanish bilingual (Would you change anything in your proficiency as a Spanish-Russian bilingual?). Together, the four parameters make up self-categorization profiles which are based on the participants’ assessment of their proficiency in both Russian and Spanish in terms of fossilization and attrition of the mother tongue (Russian), attainment of a second language (Spanish), bilingual competence, and bilingual performance.

According to the obtained scores (maximum 20 points, 5 points maximum for each item), speakers were divided into two groups: a positive self-categorization group and a negative self-categorization group. The positive self-categorization group included speakers who very positively assessed their own proficiency in Spanish and Russian, their bilingual competence, and linguistic self-awareness. These participants achieved scores of 16–20 from 20, or 4 points minimum for each measured variable. The four variables (competence in Russian, competence in Spanish, bilingual Russian–Spanish competence, and attitudes towards one’s own language competence) point in the same direction: all of them express positive language ideologies and, thus, positive self-categorization.

The negative or critical self-categorization group included speakers who felt critical about their proficiency in Spanish and Russian, their bilingual competence, and their linguistic self-awareness. These participants achieved scores of 10–15 from 20; that is, from 2 to 3 points for each measured variable. Scores under 10 were not obtained, but if they were, they would point to almost total absence of Russian language maintenance and use in the participant. In this respect, the baseline at 10 points served as a control criterion for assuring a participant’s adequacy for research objectives. Their joint characteristics also pointed in the same direction: all of them expressed non-positive language ideologies and, thus, negative self-categorization (

Table 2. Sample grouping according to self-categorization.

Group	Score Range	Participants	Features
Positive self-categorization	16–20	4	Self-assessed native-like proficiency of Spanish (C1–C2) Self-assessed absence of native language fossilization Self-assessed symmetric bilingualism Positive attitudes towards one’s language competence (self-assessed full bilingualism)
Critical self-categorization	10–15	6	Self-assessed varying proficiency of Spanish (from B1 to C2) Recognition of certain fossilization of native language Self-assessed asymmetric bilingualism Critical attitudes towards one’s language competence (self-assessed dominant bilingualism)

).

Overall, the results showed that belonging to the critical self-categorization group increases the potential of transmitting RHL to a lower level of competence in the second generation. Within the positive self-categorization group, this study found a relatively uniform position in the language competence of the second generation; however, the critical self-categorization group is notably heterogeneous when assessing the same parameters.

The following illustrative materials summarize the most salient aspects of RHL proficiency dependence on the self-categorization of the first generation (

Table 3. Language competence of the second generation in Russian as a heritage language (RHL) according to first-generation speakers.

Aspect of Language Competence	Positive Self-Categorization Parents	Critical Self-Categorization Parents
Bilingual competence	Unbalanced Spanish–Russian bilingualism	General tendency to unbalanced Spanish–Russian bilingualism at different degrees
Comprehension in RHL	Fully proficient comprehension	Good comprehension
Speaking in RHL	Good speaking skills	Moderate speaking skills
Accent in RHL	Little accent	From little to noticeable accent
Reading skills in RHL	Good reading skills	Good reading skills
Writing skills in RHL	Developed writing in most children	Developed writing in few children

, Figure 1 and Figure 2). Overall, second-generation speakers have more balanced bilingual competence and more advanced proficiency in RHL if their mothers self-assess positively. Children from such families have better comprehension skills and speaking abilities; they also develop less of a foreign accent in RHL and usually can both read and write in RHL. Contrarily, children of mothers who critically assess themselves may develop highly unbalanced bilingualism in which RHL is the subordinate language. They tend to have less speaking fluency in RHL, in which they also develop a moderate accent; and although all of them read, only few can write.

3.2. Emotions and Affectivity of First-Generation Speakers and Language Competence in the Second Generation

The notion of “language affect” is usually applied to second language learning to describe the set of feelings one has about the target language (Pavlenko 2013). The emotional charge of heritage languages also implies such feelings and refers to diffferent associations the speaking community shares about HL, like its esthetic value (Leedom Shaul 2014) or its value as a bridge to the community’s past (Guardado 2018a). As one of the most important affective factors (Mishan 2005), motivation clearly stands out for its direct influence on language acquisition and transmission: positively-oriented affectivity drives better outcomes in language attainment. In this sense, the emotional component is embedded into the motivational angle, and several studies on emotions in L2 acquisition considered their analysis through motivation research (see Oxford 2011 for an overview).

In order to determine to which extent first-generation affectivity and emotional position influence the level of acquisition of RHL in the second generation, the following four variables were analyzed from the questionnare: (a.) the assessed usefulness of RHL for the second generation in the future (Do you believe Russian language will be useful for your children in future?); (b.) the assessed importance of RHL for the second generation (Do you believe Russian language is important for your children?); (c.) the assessed usefulness of RHL in comparison with other foreign languages (Do you believe Russian language is more important for your children than English?); (d.) the assessed expectations for RHL in second-generation speakers (Do you believe your children should know Russian language as well as possible?). These variables refer to both instrumental and integrational motivation: that is, both utilitarian and social and interpersonal motives for language learning (Baker and Prys Jones 1998).

Surprisingly, all participants coincided in considering RHL to be extremely useful and important for second-generation speakers, and they strongly believed that second-generation speakers should know Russian as well as possible. The point of difference between speakers lay in perceptual instrumental motivation concerning the usefulness of Russian against foreign languages, which in the questionnaire explicitly referred to English.

According to the obtained scores, speakers were divided into three groups: a pro-Russian motivated group; a pro-English motivated group; and an equality motivated group. The pro-Russian motivated group considered Russian to be more important for the second generation than other languages; the pro-English motivated group considered English to be more important for the second generation than Russian; and the equality motivated group considered both Russian and English to be equally important for second-generation speakers.

The overall results show that first-generation speakers’ positive affectivity towards RHL determines higher levels of assessed language proficiency in the second generation (

Table 4. Language competence of the second generation of RHL according to first-generation speakers.

Aspect of Language Competence	Pro-Russian Group	Equality Group	Pro-English Group
Bilingual competence	Unbalanced Spanish–Russian bilingualism	Balanced and unbalanced Spanish–Russian bilingualism	Unbalanced bilingualism with prominent dominance of Spanish
Comprehension in RHL	Proficient comprehension	Proficient comprehension	Good comprehension
Speaking in RHL	Good speaking skills	Good speaking skills	Moderate speaking skills
Accent in RHL	Little accent	From little to noticeable accent	Noticeable accent
Reading skills in RHL	Good reading skills	Good reading skills	Good reading skills
Writing skills in RHL	Developed writing in few children	Developed writing in most children	Developed writing in few children

). Children from families who believe Russian to be more or as necessary and important as the English language for the second generation are usually assessed by their parents as either balanced or slightly unbalanced bilinguals, with native-like comprehension and reading skills, good speaking abilities with almost no accent, and moderate writing development. Conversely, children proceeding from families who believe Russian not to be so important as English are usually assessed by their parents as true unbalanced bilinguals with none of the skills being at a native-like level: although they have good comprehension and reading skills, their speaking is limited and marked by a noticeable accent, while writing skills are developed in very few of them.

Notably, the pro-Russian group and equality group mostly coincide with the group of positive self-categorization parents. This is a predictable observation, since one may expect positive attitudes to language transmission and learning in every respect.

Figure 2 summarizes the most salient aspects of RHL proficiency dependence on the affective position of the first generation.

3.3. Who Transmits RHL to Whom and Why

The participants of this study pointed out the strategies of RHL transmission they usually apply with the second-generation speakers. They were asked about their own practices in Russian language at home (I watch TV/movies in Russian; I read Russian books; I read Russian press; I listen to music in Russian language); about the practices of their children in the Russian language at home (They watch cartoons/movies in Russian; They read/I read them Russian books; They listen to music in Russian language; They play games in Russian language); and about their involvement in the explicit transmission of Russian (Do you teach your children Russian language at home?) and the strategies of such explicit transmission (I read them Russian books; I teach them Russian alphabet for reading; I teach them Russian alphabet for writing; I teach them new words and sentence composition; We describe pictures; We do not have time for learning Russian).

The answers given by the participants allowed us to identify two groups of speakers: (a.) those who make efforts to explicitly transmit RHL to the second generation (overall score from 11 to 20) through different everyday activities and target tasks for contextual maintenance of RHL; and (b.) those whose efforts are weak in transmitting RHL to the second generation (overall score from 0 to 10) since everyday activities and target tasks for contextual maintenance of RHL are random and non-systematic.

Group (a.) almost completely coincides with the positive self-categorization parents and, thus, with speakers who possess positive affectivity towards RHL and its intergenerational transmission. First-generation mothers who believe themselves to be proficient speakers of Russian as their mother tongue and Spanish as their second language also believe Russian is of high importance for their children and act accordingly by making all efforts in order to assure native-like proficiency in RHL in the second generation. As a result, they assess their children’s competence in RHL as totally or almost native-like—some of them referring to it as “My child is a perfect bilingual”—and even equivalent to that of their Russian monolingual peers.

Conversely, Group (b.) coincides with the critical self-categorization parents and, thus, with speakers who do not share positive affectivity towards RHL, but rather consider Russian not as useful as other languages for their children. First-generation mothers who believe themselves to be unbalanced bilinguals and bilinguals with fossilized competence are not always sure whether Russian is of any importance for their descendants, and they do not actually make significant efforts to maintain RHL in the second-generation speakers. As a result, they assess their children’s competence in RHL as unbalanced and subordinate to Spanish, with most children being unable to fluently speak and write in their heritage language.

4. Discussion

Recent studies on attitudes to heritage languages evidenced the importance of positive cognition and affectivity in understanding why some communities succeed in intergenerational transmission of HL and others do not. Considering HL as a prestigious language and displaying positive attitudes towards HL in the first generation is crucial for avoiding an intergenerational shift to the majority language and for enhancing HL second-generation proficiency (Li 2006). Such proficiency directly correlates with the strength of the parents’ positive attitudes towards HL and their involvement in HL transmission (Simpson 2019). Although “one parent, one language” was demonstrated not to be the best context for successful HL intergenerational transmission (Brehmer and Kurbangulova 2017), a positive speaker profile is key for assuring successful HL development in second-generation speakers, even if only one parent is responsible for it.

In this sense, the present study is consistent with other previous research (Park and Sarkar 2008; Pérez-Leroux et al. 2011) which demonstrated the salient importance of language positiveness in the first generation. A small community of Russian-speaking immigrants living in a town in central Spain proves this claim. First-generation speakers with positive self-assessment as Spanish–Russian bilinguals, positive attitudes towards Russian as an ideological and functional tool, and positive motivation for intergenerational transmission of Russian consider their children to be more proficient RHL speakers. Although assessed second-generation RHL proficiency is not homogeneous, positively oriented first-generation speakers are more likely to believe they are raising perfect Russian–Spanish bilinguals with almost no language limitations. Some speakers even believe their children to be fully balanced functional bilinguals, able to natively communicate in RHL as their monolingual peers would do. The family language policy in these cases is based on default language transmission, which is reinforced by conscious and explicit strategies for RHL acquisition in second-generation speakers.

Conversely, first-generation speakers with critical self-assessment as Spanish–Russian bilinguals (usually perceiving themselves as semilinguals: non-native Spanish speakers with fossilized proficiency in Russian) usually share less positive attitudes towards Russian as a functional tool and, as a result, foster RHL intergenerational transmission to a smaller extent. Believing Russian to be less important than English, for example, is more characteristic of critically self-assessed first-generation speakers; consequently, family transmission strategies are more likely to be based on receptive, rather than default, transmission.

In the explanation of these results, the present study does not consider first-generation attitudes to be a misleading variable which would influence the positive assessment of second-generation speakers linearly. Quite the opposite, during surveying, the researcher was surprised to discover that positively self-assessed first-generation speakers were more prone to be demanding about their own and their children’s HL proficiency. Overall, they tended to comment more on their concerns about RHL maintenance and competence, particularly in the second-generation speakers.

Despite several limitations of this study, it also offers several interesting implications for the scientific comprehension of heritage languages. Certainly, the small size of the sample and its specificity—all participants belong to a reduced community living on a non-typical territory for Russian-speaking immigrants in Spain, most of whom are concentrated on the Mediterranean coast—may influence the final results. Nonetheless, these results highlight an important conclusion about which sociolinguistic context benefits HL intergenerational transmission. First of all, the study puts forward that HL transmission relies on the versatile psycholinguistic profiles of the first-generation speakers, including their cognitive, emotional, and affective predispositions. Although this research shows positivity in cognitive, emotional, and affective factors to coincide in Russian-speaking first-generation immigrants in Salamanca, it may theoretically imply that the three factors should be positively valenced in first-generation speakers to assure successful HL intergenerational transmission.

Either way, the present study highlights positive affectivity towards HL in the first generation to be key to promoting native-like or almost native-like proficiency in HL in the second generation. This idea was also reached by other studies (see Hsieh 2012) which showed the direct dependence of children’s proficiency in HL on parents’ attitudes to HL. In this respect, and with a focus on future research, it is possible to suggest that first-generation speakers driven to migrate by affective or non-compulsory motives are more prone to transmit their HL successfully. This is a pending task, to which one should add the necessity to compare the results obtained in this study with similar data from other Russian-speaking communities in Spain.