Russian language proficiency of monolingual and Russian – English bi / multilingual children *

This paper reports the results of a study investigating the Russian-language proficiency of bi/multilingual (Russian–English [+additional language]) children in Saskatchewan, Canada, as compared to monolingual children in Russia. Very few studies of Russo-English bilingual children’s language performance are available in the Canadian context, and no studies have ever been conducted in Saskatchewan, where input is severely restricted compared to other contexts due to demographic reasons. The major impetus for the study was therefore to determine if in these settings, bi/multilingual children can develop minority language proficiency comparable to that of their monolingual peers in Russia. The methodology employed in the study focuses on the linguistic analysis of audio recordings of a picture description task performed by participants. Oral language proficiency parameters (including vocabulary, fluency, and syntactical complexity) in the speech of the 5–6-year-old bi/multilingual children were compared with the ones produced by a control group (monolingual children) from Russia. The results demonstrate that the oral language proficiency in the bilingual group is on a par with that of the monolingual group. However, reading and writing skills of the bi/multilingual group are less developed than in the control group.


Introduction
Canada is becoming increasingly more multilingual, and the proportion of the population speaking immigrant languages (i.e., languages other than Aboriginal, English and French) as mother tongues at home continues to grow, increasing from 10% to 20% of the Canadian population in just five years (Winzer & Mazurek, 2000) with more than 200 immigrant languages in use (Statistics Canada, 2012).In some areas, such as Toronto, 50% of schoolchildren have a first language other than one of the official languages (Byers-Heinlein & Lew-Williams, 2013).The issues of maintaining "linguistic equilibrium" or "linguistic plurality" in Canadian society in general, in education, and in individual families are, therefore, increasingly the objects of discussions (e.g., Veltman, 1998;Prevost & Beaud, 2002;Gabszewicz, Ginsburgh, & Weber, 2011;Armstrong, 2015).
Many immigrant parents in North America and in Canada are interested in raising their children to speak the immigrant (or heritage) language along with the majority language, and yet they do not always receive adequate support from specialists regarding their questions and concerns (Byers-Heinlein & Lew-Williams, 2013).All parents who attempt to raise their children as bilingual or multilingual experience difficulties in maintaining the balance between the languages (Caldas & Caron-Caldas, 2008).In an ideal world, with equal exposure to both languages, children could become balanced bilinguals (Davidson, 2009).However, the same amount of exposure to two languages is difficult to achieve in daily life; dominant bilinguals who speak one language better than the other are, therefore, more common (Paradis, Crago, Genesee, & Rice, 2003;Zaretsky & Bar-Shalom, 2010).Balance is even harder to achieve between a minority and a majority language because the heritage language has more limited resources and social support (De Houwer, 2007).
Because of widespread bilingualism in North America and in Canada particularly, a significant body of literature on heritage languages has accumulated.Heritage language speakers are defined as bilinguals who grew up in homes where a language other than the majority language was spoken, and have simultaneously, or with some delay, acquired the majority language as well (Scontras, Fuchs, & Polinsky, 2015).These are typically children of immigrant families (Scontras et al, 2015).
Heritage language acquisition research in North America has established some fundamental features of language development of bi/multilingual children.Their individual language acquisition paths and achieved proficiency levels are unique (Scontras et al, 2015).Their proficiency in the heritage language depends on the quantity and quality of language input (Hoff & Core, 2015;Jia & Paradis, 2015;Unsworth, 2016).The native language of both parents affects heritage language acquisition: the latter is more likely to be successful when both parents are native speakers of the heritage language (Hoff, Rumiche, Burridge, Ribot, & Welsh, 2014).Other factors in heritage language proficiency include age of arrival, and maternal education levels (Jia & Paradis, 2015).Some studies also suggest that language use is an important predictor of children's language fluency (Bedore, Peña, Gillam, & Tsung-Han, 2010, Hammer et al., 2012).By contrast, a few studies seem to indicate that the frequency of language use does not predict fluency (Schmidt, 2007;Zaretsky & Bar-Shalom, 2010).Therefore, we were motivated to see whether adequate language skill development by bi/multilingual children is possible in situations where social support for the immigrant language is minimal, and the language input and use are limited, as is the case of Russian in Saskatchewan.
A number of linguistic and applied linguistic studies have attempted to evaluate the language proficiency of multilingual children vis-à-vis their monolingual peers (e.g., Paradis et al., 2003;Bedore et al., 2010;Jia & Paradis, 2015).Some studies point to existing fundamental similarities in language acquisition and performance by young monolingual and bilingual children and infants (e.g., Fennell & Byers-Heinlein, 2014;Jia & Paradis 2015).And yet, these studies also show that from early infancy, bilinguals may experience more difficulties in some aspects of language acquisition.For example, bilinguals may have some deficits in their heritage language skills (particularly grammar) that result from incomplete acquisition or attrition, or the influence of the majority language (Montrul, 2008;Jia & Paradis, 2015).Some bilingual heritage language speakers may also have difficulties mastering phonology, have smaller vocabularies, have problems with some specific vocabulary items and morphemes or their bilingual acquisition may take a little more time (e.g., Paradis et al., 2003;Fennell, Byers-Heinlein, & Werker, 2007;Lew-Williams, 2013;Jia & Paradis, 2015).Overall, the speech of heritage bilinguals can be described as exhibiting signs of transfer from the majority language, divergence (incomplete acquisition), and attrition (for older bilinguals) (Scontras et al, 2015).Some researchers, however, find that the current state of research on heritage languages is insufficient (Byers-Heinlein & Lew-Williams, 2013, p. 96), and that too little is known about language development in bilingual environments (Hoff et al., 2014).It has been also suggested that presence or absence of differences in language performance by bi/multilinguals and monolinguals may be explained by the methodology and the approach to measurement (Peets & Bialystok, 2015).
One limitation of the above research is that, due to the status of various languages, the demographics of their speakers, and language policies, most research on bilingualism in North America is done with French-English and Spanish-English bilinguals (e.g., Paradis et al., 2003;Weikum et al., 2007;Sebastián-Gallés, Albareda-Castellot, Weikum, & Werker, 2012;Bedore et al., 2010;Hoff et al., 2014).However, the mechanics of bilingualism can be better understood if the scope of languages involved in research is broadened, since bilingual input is extremely diverse in terms of the languages involved, the linguistic level of the speakers, their age, gender, and other factors (Havy, Bouchon, & Nazzi, 2016, p. 420).
Russian is one of the immigrant languages of Canada that has so far been underrepresented in bilingualism research.The number of speakers of Russian as a mother tongue in Canada in 2011 was 169,950 (or 0.5% of the population).While the number of Russian-English bilingual children in North America has been growing, very little is known about Russian-English bilingualism not only in Canada, but in North America in general (Gildersleeve-Neumann & Wright, 2010).
The available studies of Russian as an immigrant language in Canada involved investigations of reading habits of Russian immigrants (Dali, 2005(Dali, , 2012) and a few selected aspects of Russian language acquisition (e.g., Kazanina & Phillips, 2007).So far, to the best of our knowledge, no studies in Canada have addressed the question of Russian language proficiency of Russian-English bi/multilingual children.The aim of the present study was to add to the palette of bi/multilingual studies in Canada by investigating Russian language proficiency of bi/multilingual children in Saskatchewan, as compared with that of their monolingual peers in Russia.
Most speakers of Russian in Canada reside in Ontario (93,080) (Statistics Canada, 2012).However, some other locations in Canada have significant di-asporas of Russian speakers.Noticeable among them is Kootenay in British Columbia, where due to Doukhobor heritage (e.g., Makarova, 2012), Russian-English bilingual children have a chance to attend Russian language classes or immersion centres.In some other locations, children's centres, such as the "Mechta" or "Dream" Children's Centre in Montreal, or Russian language and culture schools such as "Svetlyachok" or "Firefly" in Manitoba, are available for Russian-speaking children.In Saskatchewan, Russian speakers are relatively few in number: only 2,355 individuals in the province claim Russian as their mother tongue (Statistics Canada, 2012).Russian is not taught anywhere in the province at any level of education.A Russian Saturday school in Saskatoon is run by enthusiast parents, but it only meets once a week during the school year.Regular school subject classes are not offered in it, and few cultural activities are organized.The conditions for Russian language maintenance in Saskatchewan are therefore rather unfavourable.
The goal of the study reported here was to investigate whether, despite this unfavourable social environment, Russian-speaking children in Saskatchewan can develop Russian language proficiencies similar to those achieved by their monolingual peers in Russia.More specifically, the study compared some selected Russian speech fluency parameters of bi/multilingual children in Saskatchewan with those of monolingual peers in Russia.
The major research question of the study was: are there differences in basic oral communication skills between bi/multilingual (Russian+ English [+another language]) children in Saskatchewan and their monolingual peers in Russia?A secondary research question of the study was: do bi/multilingual children in Saskatchewan develop reading skills in Russian?
The age selected for the study was 5-6, since monolingual children of this age are known to have already formed basic Russian grammar skills (Zaretsky & Bar-Shalom, 2010), and since at this age, bi/multilingual children already started public schools, they are exposed to English, but are not yet fully assimilated into the majority language environment.

Participants
The participants were recruited via purposeful sampling.They were all between 5 and 6 years old, and included two groups: bi/multilinguals and monolinguals.
Bi/multilingual children were selected according to the following criteria: if they spoke Russian and English; if they were either born in Canada or brought to Canada before they were three years old; if they had been immersed in an English-speaking environment in Saskatoon on a daily basis for at least six months (in daycare, preschool or school), and if they had at least one Russian-speaking parent who satisfied the conditions of being born in a Russian-speaking country and having Russian as the mother tongue.
On the basis of the above, a total of 30 bi/multilingual children were recruited (12 boys and 18 girls).Twenty-three (out of 30) participants were born abroad, and seven were born in Canada.Seventeen children were bilingual (Russian and English) and 13 were multilingual (spoke Russian, English and at least one additional language).Of the multilingual subset, seven children spoke French, three Ukrainian, one French and Arabic, one French and Ukrainian, and one Ukrainian and Hebrew.The countries of participant children's parents were Russia (7), Ukraine (17), Kazakhstan (4), Kyrgyzstan (1), and Uzbekistan (1).All the children had Russian-speaking mothers.Two children had fathers who did not speak Russian, all other children had Russian-speaking fathers.As per the selection criteria above, all the children were brought up at home in a Russian-speaking environment, but attended an English-speaking pre-school, kindergarten, or elementary school for at least 6 months prior to the participation in the study.
Monolingual participants (N = 13 children: 6 girls and 7 boys) were all recruited in Kemerovo, Russia.None of them spoke any language other than Russian.The choice of the location was determined by two reasons.First, one of the authors secured recruitment access via personal contacts in the area.Second, Kemerovo is located in the geographic centre of Russia, similarly to Saskatoon's location in the middle of Canadian prairies.Further, both Saskatoon and Kemerovo are regional centres.

Materials
The study reported here is based on the linguistic analysis of audio recordings of a picture description task performed in Russian by child participants.A set of six pictures (from a children's online picture book, "Dobraya skazka v kartinkax" [A good fairy-tale in pictures]) was given to all participating children and they were asked to tell the story represented in the pictures.Before the picture description task, child participants were also asked interview questions in order to retrieve a self-assessment of their proficiencies.The questions addressed the languages that the children speak, the language they speak better, whether they can read in Russian and whether they had comprehension problems if they visited a Russian-speaking country.The same questions were asked of the parents, separately from the children's interview, so that they could not hear.The interviews and the picture-prompted narratives were recorded with a Zoom H2n Handy Recorder in Wave sound format.They were manually transcribed and analyzed to extract proficiency parameters.

Language proficiency assessment
A wordless picture book description task was selected as the basis of language fluency assessment in this study.Picture book narratives are commonly used for assessing bilingual children's speech, since they are easy to set up, pleasing and appropriate for children, and serve as an efficient way to solicit a complex language output that can be evaluated for multiple linguistic features (e.g., Bedore et al., 2010;Squires et al., 2014).
As with earlier studies of children's narratives (Bedore et al., 2010;Squires et al., 2014), the parameters employed for the description of children's speech included: speech rate (in number of words per minute); total vocabulary produced (the total number of words in the narrative); number of distinct lexical items; number of verbs and nouns; number of utterances; the longest utterance (in number of words); mean length of utterance (in word count); number of complex and compound sentences, and number of phonological, lexical and grammatical errors.
In order to assess their reading ability, we first asked the children if they could read, and if they answered "yes," they were given a passage to read from a children's book in Russian (Djuimovochka [Thumbelina]).Their reading proficiency was evaluated in words-per-minute and number of errors.
The proficiency parameters were entered on MS Excel 2013 sheets for analysis, t-tests (2-tail, assuming unequal distribution) were conducted for each parameter variance across the two participant groups (monolingual and bi/multilingual).It should be noted that the sample sizes were relatively small and unequal as often happens in bilingual research due to difficulties in obtaining samples (e.g., Fukuda, 2017), and the t-tests results therefore have to be treated with some caution.

Languages spoken by bi/multilingual children
When asked which languages they can speak, 16 children (53%) reported they could speak both Russian and English.Due to different backgrounds and schooling, almost half the children spoke another language or two in addition to English and Russian: seven (23.3%) spoke Russian, English and French; one girl spoke Russian, English, French and Arabic; one spoke Russian, English, French and Ukrainian; two participants knew Russian, English and Ukrainian; one spoke Russian, English, Ukrainian and Hebrew; and one spoke Russian and Ukrainian.Children's responses were fully confirmed by parents.

Language spoken better by bi/multilingual children
In response to a question about which language they spoke better: Russian, English, or both equally, children offered the total of 29 entries.Of these, Russian was selected by 11 (37.9%)children, English by nine (31%) and both were chosen by nine (31%) respondents.Results show that most of the children were comfortable with Russian.By contrast, only four parents indicated that their child spoke Russian better than English.There appear to be differences of opinion between parents and children with regard to the children's command of Russian vs. English.

Russian reading skills by bi/multilingual children
When asked whether they can read in Russian, 14 children (of 29), or 48% of respondents, answered "yes."Parents of these fourteen children also confirmed that their children could read in Russian at different levels.

Comprehension skills by bi/multilingual children when brought to monolingual environment
Since the Russian input that children are receiving in Saskatchewan comes mostly from within their families and from a few friends, we were interested in asking the children whether they had any problems comprehending Russian spoken in Russia.When asked whether they had been to Russia or another Russian-speaking country before, a total of 17 children answered positively.These children were asked whether they understood people during their trips to a Russian-speaking country.With one exception, all the children answered "yes."Parents' responses fully conformed to the children's answers.In other words, Russian language comprehension skills by bi/multilingual children appear to be sufficiently developed.

Reading skills of bi/multilingual children as compared to monolingual children
As reported above, when asked if they could read in Russian, 14 (out of 29) bi/multilingual children (48%) replied "yes;" however, only 10 of them (33% of participants) could demonstrate their actual reading skills by reading a passage from the children's book, Djuimovochka [Thumbelina].By comparison, nine out of 14 (64%) of monolingual Russian children could demonstrate their reading skills.Due to small values of expected probabilities (less than five), the chi-square test -that could help to demonstrate whether the reading skill abilities across the groups are statistically different -is unreliable, and could not be conducted.However, note that twice as many monolingual participants were able to read as compared to bi/multilinguals.
The reading rates of bi/multilinguals who were able to read ranged from three to 87 words per minute, with an average of 23 words per minute.The average reading rate of monolinguals was 34 words per minute.While the average reading rate of monolinguals is about 1.5 times higher, this difference is not significant for the given sample, due to the small number of children who could read.However, it should be noted that three bi/multilingual children displayed faster reading rates than those of the average Russian monolingual child.

Oral language proficiency (speaking skills) of bi/multilingual children as compared to monolingual children
The t-test results (see Table 1) did not show significant differences in most parameter values across the groups.While this could, of course, be partly due to the small samples, the average parametric values across the groups appear overall very similar.The only significant differences across the groups were found in the numbers of grammar and lexical errors as well as in the total number of errors, whereby bi/multilinguals make more errors.While there were no significant differences across the groups for most parameters in the given sample, it is interesting to note that some proficiency parameters in the speech of bi/multilingual children were marginally higher than those of their monolingual peers, specifically: total number of words produced, longest utterance, number of complex sentences, number of clauses per utterance, and speech rates.

Discussion
Some immigrant parents still express concern that bi/multilingualism may prevent their children from successful language acquisition, or that code-switching and code-mixing between languages may somehow "confuse" them (Byers-Heinlein & Lew-Williams, 2013).On the other hand, many linguistic and applied linguistic studies have been promoting the benefits of bilingualism (e.g., Kovács, 2007;King & Mackey, 2007;Bialystok, Craik, & Luk, 2012).It has been highlighted that bi/multilingual children's knowledge of more than one language is important for travel, employment, communication with family members abroad, maintaining connections to ancestral culture and history, and making more friends with diverse linguistic and cultural backgrounds (Byers-Heinlein & Lew-Williams, 2013, p. 98).In addition, bi/multilinguals have been shown to have advantages in communication overall as well as in some cognitive functions and activities (Bialystok et al., 2012;Brito & Barr, 2012;Byers-Heinlein & Lew-Williams, 2013).
Moreover, some studies have been issuing "warnings" that bilingual children should not be compared in their language performance with monolin- guals, since such comparisons are biased in favour of monolinguals and often demonstrate evidence of a bilingual delay in language development (Byers-Heinlein & Lew-Williams, 2013, p. 107).While it is obvious that different scales should be developed for evaluating bi/multilinguals' language performance, generally acceptable formats do not yet exist.The "monolingual" frame of reference is justifiable in our study, as we wanted to compare language fluency parameters of a population with limited input and context for acquisition with a population who has the maximally rich input (majority language situation for monolingual children).
In contrast to previous research studies showing that bilinguals/multilinguals tend to be behind monolinguals when compared on a single language performance variable (Hoff & Core, 2015;Jia & Paradis, 2015), our comparison of the speech of 5-6-year-old bi/multilingual and monolingual children across many variables align with other studies that showed that selected parameters in the speech of heritage bilinguals and monolinguals can be indistinguishable from each other (Fukuda, 2017).Our study can hopefully help to at least partially dispel parents' concerns that their children's Russian language proficiency is below what is expected for a monolingual child.For the given sample, we found few significant differences in the bi/multilingual and monolingual children's language performance, only the number of lexical and grammatical errors as well as the total number of errors differed between the groups.Furthermore, in our study, some functional parameters of bi/multilingual children's speech (including cumulative vocabulary, some measures of grammar complexity and speech rates) were above the average parameters for monolinguals (although insignificantly so).
It has been suggested in earlier studies that cumulative vocabulary is a reliable measure of linguistic capacity of young bilingual children (Hoff and Core, 2015, p. 89).Our comparison shows that cumulative vocabulary (as in the number of different lexemes employed) as well as the total number of words per narration did not differ significantly between the bi/multilingual and monolingual groups in the given sample.This finding may be indicative of the overall success in Russian language learning of bi/multilingual children.
In terms of grammar acquisition, earlier studies suggest that the grammatical development of bi/multilingual children may be behind that of monolinguals when their grammatical performance is measured separately for each language (Hoff & Core, 2015).Grammatical structures have been shown in earlier studies to be the "weak point" of heritage Russian-speaking children and adults (Polinsky, 2006(Polinsky, , 2008;;Scontras et al., 2015) as well as of heritage speakers of other languages (e.g., Montrul, 2005Montrul, , 2008;;Jia & Paradis, 2015).However, in our sample, bi/multilingual children show no less complexity of sentence structure than their monolingual peers.The number of grammatical errors in the speech of bi/multilinguals is, however, significantly higher than in the speech of monolinguals.The complex morpho-syntactic system of the Russian language is known to pose challenges for Russian-English bilinguals (e.g., Akhutina, Kurgansky, Polinsky, & Bates, 1999;Pereltsvaig, 2008;Smyslova, 2012;Scontras et al., 2015).The types of grammar errors in the Russian speech of bi/multilinguals are very similar to earlier studies and include errors in nominal and adjectival cases and verbal forms (e.g., Polinsky 2006Polinsky , 2008;;Pereltsvaig, 2008;Smyslova, 2012).
It is only possible to speculate what helped the bi/multilingual children to achieve this success in Russian language acquisition, despite the majority language environment that does not facilitate this particular language learning.In another sociolinguistic aspect of our investigation, one that is beyond the scope of this paper, both parents and their children expressed unanimous support for bilingualism.Most parents report solely or predominantly speaking Russian to their children at home.All the children -according to self-reports and the parental testimonies -have at least one friend to speak Russian with, and all the parents report having mostly Russian-speaking friends.Both parents of most of the children speak Russian in the home.Earlier research studies have shown that bilingualism is possible even if exposure to one of the languages amounts to only 10-25% of the total language input (Byers-Heinlein & Lew Williams, 2013).Even if a minority language such as Russian is only spoken by children at home with parents, siblings, and with a few friends, it may still develop along a predominantly typical path, a fact confirmed by this study.
The opportunity to interact with multiple speakers, as well as the quality of input, are known to be decisive in vocabulary acquisition by children (Byers-Heinlein & Lew-Williams, 2013).It is reasonable to suggest that for what they may miss in terms of the limited number of speakers of Russian around them, children are compensated by the duration and quality of language input, as well as by a strong pro-heritage-language policy in the home.Thus, all these factors may have served to facilitate a home language environment that stimulates Russian language learning by children.
It should also be noted that the influence of the majority language environment provided by English-language school education is still limited at the age of 5 or 6, after children have just started school.In the interview, one child expressed resistance to the school policies and majority language dominance.That same child had another Russian-speaking friend at school, and the two spoke Russian with each other during breaks and in class when children were supposed to communicate.The school teacher repeatedly tried to stop them from speaking Russian, rebuking and even punishing them a few times, yet the child concluded very proudly: "we kept talking in Russian no matter what she did, and she eventually gave up".
Given the fact that the majority of children and parents interviewed did not even come from Russia, but from other Russian-speaking countries where the language is currently in the minority, it is most surprising that both parents and children take so much pride in their Russian language heritage and invest such efforts in maintaining the language.Unfortunately, we did not ask the parents for their explanations of the significance of their Russian language and Russian heritage in this study, a subject that remains to be investigated in future work.
However, some areas of parental concern about the ability to maintain heritage language adequately may be justified, particularly as regards the chil-dren's ability to read and write.It may therefore be advisable for parents to spend more time reading to the children and encouraging them to read, since literacy is well known to be a major supporter of language acquisition and language proficiency (Dali, 2012).A study of reading patterns among Russianspeaking adult immigrants in the Greater Toronto area revealed a crucial connection between acculturation patterns of the ethnic group and reading behaviours and preferences (Dali, 2012).Specifically for English-Russian bilingual children in the USA, a negative correlation was found between reading skills and the degree of morphological attrition in oral skills (Zaretsky & Bar-Shalom, 2010).
It is an open question whether children can maintain and fully develop their Russian language abilities when the school environment takes precedence.It is almost inevitable that their Russian language skills will eventually remain underdeveloped as compared to their monolingual Russian peers, since they would not have the opportunities to advance their vocabulary, complex grammar, reading and writing skills; abilities that are typically acquired via the learning of various school subjects in upper elementary, high school, and university.The bulk of heritage language acquisition research suggests that heritage speakers eventually show structural transfers from the majority language, incomplete acquisition and attrition of the heritage language (e.g., Scontras et al., 2015).
This study is limited, and the results should be treated with caution: the sample in the study was relatively small, particularly for the monolingual group, due to difficulties in accessibility and constraints of time.We compared the general proficiency parameters by bi/multilinguals and monolinguals, but we did not address the issue of their English language ability and of language interactions that are known to arise in Russian-English language acquisition (e.g., Nicoladis, Da Costa, & Foursha-Stevenson, 2016).More detailed analysis of proficiencies in both languages as well as of the types of errors in children's Russian and English and of interactions between the languages are planned for future studies.

Conclusion
We examined Russian-language proficiency parameters (predominantly oral language proficiency and limited analysis of reading rates) of Russian-English 5-6-year-old bi/multilingual children as compared with the proficiency parameters of their monolingual peers.The results demonstrate the success of Russian language acquisition in the bi/multilingual group, despite the limitations of the social environment in Saskatchewan.Overall, bi/multilingual and monolingual groups were similar across speech fluency parameters.In a few parameters, however, bi/multilinguals perform insignificantly better than mono-linguals.On the other hand, bi/multilinguals are insignificantly behind their monolingual peers in the development of reading skills and make significantly more grammatical and lexical errors than monolinguals.

TABLE 1
Speech parameters of bi/multilinguals and monolinguals