Overcoming language barriers : lessons learnt from migrant children

In this paper, we identify reasons for the high school achievement of some migrant children, in spite of the language barriers faced by themselves and their parents. We explore the literature to determine the factors used to overcome such barriers, particularly those beyond social economic status (SES) and other largely inherited factors that remain a common focus of migration and school effectiveness literature. We identify the need to pay greater attention to non-conventional factors, such as aspirations, expectations and creativities. We also examine school effectiveness literature in South Africa, arguably a typical case of a developing country, and note that much of the literature centres on analysis and lamentation of physical and human resource constraints, instead of experimenting on non-conventional factors.


Introduction
This paper asks why some migrant children around the globe perform well at school despite the predictable difficulties of navigating a new culture and language, often with constrained social and economic resources.We acknowledge that scholarly success is not the universal experience of migrant children.In fact, a large amount of research suggests that the majority of migrant children is disadvantaged for the aforementioned reasons and tend to underperform in comparison with their native population counterparts 1 .The United States, the country with the largest inflow of immigrants, generates the largest number of studies on this topic.Many U.S. studies reach a similar conclusion: "for more than 30 years, English language learners [those still not fluent in English, mostly migrant learners] have consistently lagged academically behind their English-proficient peers" (Golden and Fortuny, 2010: 5; also see Hagelskamp et al., 2010).
However, some scholars have found that "a small percentage of children of immigrants do manage to succeed" (Alba andSilberman, 2009: 1444) and even outperform natives.A longitudinal study conducted since 1991 (The Children of Immigrants Longitudinal Study in U.S., CILS

Ke Yu (Corresponding author)
hereafter) follows the progress of a cohort of over 5000 migrant children and finds that many achieve better grades and lower dropout rates than their fellow native students (Portes and Rumbaut, 2001;Rumbaut, 1999; also see Kasinitz et al., 2008).Kao and Tienda (1995) compare first generation migrants (born in the foreign land and migrated with their parents) with natives of the destination country and conclude that children born to migrant parents are relatively more successful in their educational outcomes than native children born to native parents are.Migrants out-performing natives are also reported in countries such as Sweden (Clark, 2003).
In this paper, we identify factors that explain the success or failure of migrant children at school, with a specific focus on their potential for use in overcoming language barriers and other inherited factors.Our analysis supplements existing migration and school effectiveness literature by expanding the focal discussion on those inherited factors.Our main finding points to a need for schools and parents to pay greater attention to non-conventional factors, such as aspirations, expectations and creativities, in order to improve learning outcomes.
We establish this through an extensive review of international migration literature.We also contrast our findings to a review of school effectiveness literature in South Africa that we expect to represent the context of a typical developing country.This paper is organised as follows: it starts with a brief review of the migration literature on the role of language in children's scholarly success.This is followed by an overview of why migrant children achieve (or not), and then the factors we choose to investigate in greater detail, namely various language interventions; parental involvement in terms of parental expectations, other ways parents facilitate the home learning environment; parental-school interactions, and school outreach programmes to maintain engagement.The South African literature is then examined and compared with the international literature.The article concludes by summarizing the findings and formulating the relevant implications.

Is language ability essential to scholarly success?
Linguistic skills are generally regarded as critical in a migrant's integration to their host country (Allen, 2006).(Early) literacy acquisition is often regarded as a prerequisite for academic success and accounts for persisting achievement gaps throughout the grades (Tienda and Haskins, 2011).In South Africa, inability to use one's home language at school is said to impede a learner's capacity to perform to the best of his or her ability and therefore to result in underperformance (Owen-Smith, 2010).The multilingual policies adopted after the democratic change of 1994 were particularly conceived to promote the usage of indigenous languages in schools and ultimately to improve learning outcomes.
But is language ability essential to scholarly success?One strand of migration studies compares different generations of migrant children 2 .Many first and second generation comparisons, mostly undertaken in the U.S., find that school outcomes typically improve between these two generations (Bonikowska and Feng, 2010), suggesting that children do perform better when they acquire better language skills and are better adapted to their new homes.Similar results emerge in other studies of college achievement and later life success (Forbus et al., 2011) and in non-English speaking countries (Di Liberto, 2015, looking at migrants to Italy; Driessen and Merry, 2011, migrants to the Netherlands).Allen (2006: 9) stresses the importance of "a 10-or 20-year adjustment period" for any migrants to achieve success in their new homes.However, a paradox appears when the third and subsequent generations are included in the comparisons: scholastic outcomes for the 3 rd generation reach a plateau or decline (Kao & Tiena, 1995; also see Kao, 1999;Hagelskamp et al., 2010;Telles & Ortiz, 2008).This suggests a limit to which linguistic competence contributes to academic success, at which point factors such as willingness to do homework or to work hard in school (Hagelskamp et al., 2010) might supersede fluency the language of the destination country.Alternatively, it is also possible that language competence is not essential to academic success at all (Rumbaut, 1999;Zehr, 2009).
In the U.S., there are extensive discussions on this topic under the term "immigrant paradox", a phenomenon observable even after controlling for SES and children's language skills (Tienda & Haskins, 2011).The paradox is "more pronounced among the children of Asian and African migrants than other groups, is stronger for boys than for girls, and is far more consistent in secondary schools than in elementary school" (Tienda & Haskins 2011: 8, also see Hagelskamp et al., 2010).Scholars claim that this phenomenon is not necessarily applicable to all countries (Zehr, 2009).They also suggest reasons such as unsatisfactory performance of the native population; the migrant children's higher school attendance rates and the migrants' self-selectivity (that they tend to come from more advantaged and ambitious families than those left behind in their home countries (Tienda & Haskins, 2011;Zehr, 2009) as possible explanations.
We are intrigued by the paradox as well as the pockets of success achieved by some migrant children.We seek to understand in this paper how they manage such achievement and what contributes to their successes.We aim to examine whether any factors might potentially be applicable to other developing countries, South Africa included, owing to contextual similarities.Contextual similarities here refer to the language barrier both for the children and their parents; family background where many are low-income parents with demanding work schedules and therefore little time to spend with their children; parents with low school attainment themselves; and limited family resources and support (e.g.financial resources, active parental involvement, ability to locate close to good schools).

Why migrant children achieve (or not)?
Many studies have attempted to explain school success or failure and the academic gap between different groups of students.Since the 1960s, the academic gap between white and black students in the U.S has prompted sustained investigations.Findings of these studies mainly point to the enduring and substantive impact of SES, including income, education level of the parents, mother's age first at birth (Lenkeit et al., 2015).SES has also been examined extensively in immigrant studies, in addition to other largely inherited factors -such as race, ethnicity and national origins.
Other common barriers to migrant children's academic performance include language skills, cultural differences and the time period since arrival in their new country (Crosnoe, 2010;Ji & Koblinsky, 2009;Poza et al., 2014;Schwartz & Stiefel, 2006).Summarising research on Mexican-Americans, Allen (2006: 23) suggests that "…[in addition to SES], having intact families and high educational expectations have been shown to be important in predicting school success, [although] these factors do not explain fully the low levels of Mexican-American education".For Tienda and Haskins (2011: 4), "the combination of poor parental schooling and not using English at home that is associated with poor scholastic outcomes for immigrant minority youth".Further trying to explain the immigrant paradox, Tienda and Haskins (2011) point to a few other factors that could impact on school achievement, including family, school and neighbourhood qualities (e.g.family structure, relationships within the family, diet and health conditions, also see Allen, 2006).Allen (2006) adds that racial stereotyping (both within the society and by the teachers) could also influence children's motivation to do well in school.Lastly, older children may be expected to fulfil various family responsibilities thereby reducing their studying time.
In this article we choose to focus on those relatively flexible factors, those less persistent across generations (e.g.SES) and therefore more malleable.Because our interest is in how children navigate the language barrier, interventions to improve their language proficiency are also included.Our review of the literature found that the following factors assist migrant children to achieve better results in schools: language intervention, parental involvement, parental-school interactions, and school outreaches (to maintain engagement).Each is discussed in detail below.

Language interventions
Different approaches to targeting migrant children's language inefficiency have ranged "from immersion or English-only schooling-where academic lessons are taught only in English -to a bilingual instruction to various strategies in between" (Golden & Fortuny, 2010: 6).Research starts to demonstrate the benefit of a bilingual approach compared to forcing all academic content in English (Garcia 2010) or mother tongue learning (Golden & Fortuny, 2010).Code switching (within the classroom), after been downplayed by education researchers for decades, has also regained popularity in recent years (Ferguson, 2003).These trends are also consistent with another trend towards greater acceptance of ethnic values, cultures and identities and utilising these as resources instead of eliminating them as burdens as in the assimilationist attitude that prevailed in the early 20 th century (Allen, 2006;Conger et al., 2011).
Despite consensus among the academics, however, this practice is still not always adopted owing to frequent strong objections from parents, who often believe English education enhances their children's future economic and social benefits (Khosa, 2012;Lafon, 2009).Interestingly, a number of migration studies do confirm that early exposure to English (e.g.speaking frequently at home or watching television) has a positive impact on both literacy and mathematics results (Taylor et al., 2013, Schneider et al., 2006).However, what also seems to be the case is that in the absence of this early exposure, children who undergo bilingual education are better off than those forced into English-only schooling.

Parental involvement and expectation
Research has repeatedly demonstrated the role of parental involvement in children's academic success and wellbeing (Turney & Kao, 2009;Mncube, 2010;Ndebele, 2015;O'Hehir & Savelsberg, 2014;Okeke, 2014), often under the broader concept of 'social capital'.Research suggests that parental involvement enhances student self-esteem; improves the child-parent relationship; helps parents to get to know the teachers, thereby facilitating joint monitoring of the child's performance; and helps children to have positive attitudes towards schools (Crosnoe, 2010;Marschall, 2006;Turney & Kao, 2009).Migrant children (and children with high-risk parents or parents with limited English proficiency in general) derive extra benefit from these involvements (Kao, 2004;Kim, 2002;Lahaie, 2008).In a modelling analysis of survey data (for over 1000 children from migrant families), Eunjoo and Yue (2016) conclude that parental involvement -parents' engagement with the child in home and school settings (together with the English proficiency of parents -a factor which is beyond the focus of this paper) are related to academic achievement, sometimes indirectly through children's own educational aspirations.
Literature indicates that parental involvement may have different dimensions and take different forms.A common typology is to distinguish between school-based and home-based parental involvement (Giallo et al., 2010;Jeynes, 2003;Lahaie, 2008;Turney & Kao, 2009).Language has been identified as a barrier to the involvement of parents.Not only does limited English proficiency affect parental ability to participate in school events but also it also affects the opportunity to contribute to home learning (Ji & Koblinsky, 2009;Wang, 2008), or even simple tasks such as comprehending their children's school reports or school expectations.This, together with cultural mismatches 3 , often leads to low involvement with the formal school system, especially parents with low SESs (Ji & Koblinsky, 2009).
Parental educational values and expectations stand out as the most important factors in parental involvement case studies (Crul & Mollenkopf, 2012), including those that do not focus exclusively on migrants 4 , and particularly for low SES parents.Regression analysis that includes parental expectation as a factor, such as that conducted by Feliciano and Lanuza (2016: 758), confirms that the immigrant paradox could "largely be explained by higher parental expectations, greater interest in school…[in addition to the use of English in early childhood]".As Hagelskamp et al. state, "Parental educational expectations and aspirations [about their children's ability to achieve academically] are among the most commonly studied explanations for both the academic success and the failures of immigrant-origin children" (2010: 720).
"Parents with scarcer resources may be less active in school activities, [however] they can still be entirely aware and supportive of their children's academic progress" (Centre on Education Policy, 2012: 2).For example, it has been noted that many such parents often engage actively in activities that facilitate home learning (Schneider & Lee, 1990), including allocating time and space for homework, ensuring homework is complete, setting limits on watching TV and arranging private tutoring (sometimes even with borrowed money or money from holding multiple jobs or work long hours) (Liu & Li, 2006).Interestingly, homework help is not commonly practiced, but instead often takes the form of supervising homework or establishing "rules about maintaining a specific grade point average" (Kao & Tienda, 1995: 13).Some parents have also been reported to have used their own life experience to teach their children both the importance of hard work and endurance, and to instil in them the value of higher academic aspiration (Fuligni & Fuligni, 2007;Lopez, 2001).
Other examples of parents translating high expectations into actions include: reading to children; asking children to read to them regardless of whether the parents can themselves read or understand; taking children to the library (Glick & Hohmann-Marriott, 2007) and talking about college (instead of school work; Kao, 2004).Some parents have implemented other practices aimed at enhancing their children's academic performance, for example, reducing all the non-academic related activities of their children such as household chores (Ji & Koblinsky, 2009;Kao & Tienda, 1995), watching TV, and extracurricular activities unrelated to academic performance (Pearce, 2006).
High expectations are common among migrant parents (Dyson, 2001;Ji & Koblinsky, 2009;Pearce & Lin, 2007).Many value education and believe that it is an important tool for their child to improve their status in life.For many, an improved life is the reason why they migrate in the first place.These values are often communicated to the children as a means of encouragement in overcoming difficulties faced at school.This sometimes results in migrant children feeling "a high family obligation and family cohesion" (Hagelskamp et al., 2010: 720) and in turn results in greater willingness to work hard and spend more time on their school work (Mau, 1997).Studies with Spanish parents have discussed this under the phrase consejosadvice composed of "spontaneous homilies designed to influence behaviours and attitudes" -regarding success in school, to provide an ongoing conversation about the importance of school (Plata-Potter & de Guzman, 2012).Similarly, appealing to family honour was also an approach identified in a study of Punjabi Sikh children in California's Central Valley (Gibson, 1988).Kasinitz et al. (2008) found that even poor, uneducated migrants have often "shown that they have the drive, ambition, courage and strength to move from one nation to another," and to transmit their determination to their children.

Creativities in facilitating parent-school interactions
The parents, the school or both parties might initiate parental-school interactions.As mentioned earlier, parents might attend school events, and enrol children in extra classes outside of school.Likewise, teachers and schools might reach out to parents, provide dual language in their written communication with parents, make use of interpreters when they engage with parents face-to-face (OECD, 2015), establish language classes for parents, and so on.These interactions ensure greater coherence in the messages sent and actions implemented.It also acts as a signal to both sides about the extent of their care about the children's education (Turney & Kao, 2009).
The many challenges to migrant parents' involvement in schools can be mitigated by creative alternatives that both parents and schools.These include sending children to additional schooling programmes, particularly those offered by their particular ethnic/ community schools (Zehr, 2009); schools organising presentations on school systems; providing college information programmes or adult education programmes; connecting parents to family service providers or other resources; providing advice on how parents could assist their children (Golden & Fortuny, 2010).To ensure attendance at parent-school meetings, the topics addressed at strategic meetings are canvassed with the parents beforehand; free childcare and/or token bonuses for students are offered (Zimmerman-Orozco, 2011).Inviting parent leaders from other schools as mentors, and being proactive in making home visits have also been reported and found to be effective (Zimmerman-Orozco 2011).Owing to the limited resources of many migrant parents, outreaches by schools are generally the most effective approach.

Creativities in school outreaches for better school attendance
School disengagement (e.g.absenteeism) has been pointed out as a common feature of non-achieving migrant children (Motti-Stefanidi et al., 2015) who are stuck in the vicious circle of low achievement and low engagement.In searching for school initiatives that focus on improving school attendance, we find a recurring theme of schools/teachers creating a sense of belonging for these children (Adelman & Taylor, 2015).Means to achieve this include the provision of financial support by schools; counselling and academic planning tailored to the needs of the students; additional (and free) classes (Tienda & Haskins, 2011); afternoon classes or summer school (Garcia, 2010); pairing children with college students as their mentors; and other tailored support.

The South African case
This section triangulates the South African education literature on the factors and interventions to boost learner achievement with the factors identified above.We choose South Africa as our case study, not only because both of the authors of this paper are South Africa-based and therefore is particularly interested in this case, but also because of the commonalities that we observed in our initial searching and reading of the South African education literature, with that of other developing countries (Yu et al., 2015).We also expect that the perceived value of children being educated in English -a second language and sometimes third language to many of the schoolchildren in South Africa -in addition to a great language diversity, is not unique to South Africa but applicable to other countries with populations whose first language is not English.
As mentioned above, language remains an undigested challenge in post-Apartheid South Africa.Concerns over the inability of children to use their home languages in schools have been frequently raised (Lafon, 2009;Owen-Smith, 2010;Reynders, 2014).This challenge has been used in explaining the continued discrepancy between learners with English as a first, second, or third language (Lafon, 2009).Simkins's (2010) multivariate regression on school mathematics scores shows that home language ranks second only to race as a primary factor influencing learning outcomes.Studies also show a strong correlation between literacy test scores and numeracy test scores (Besharati, 2016;Cummins, 2009), where language ability creates a domino effect for the learning results on other subjects.The current Language in Education Policy (LiEP), based on a multilingual principle, as well as the National Curriculum Statement, encourages the use of mother-tongue in the first three years of schooling (Grade 1-3).The first three years are also expected to strengthen learners' proficiency at a second additional language (typically English or Afrikaans), to which the learners then switch in Grade 4 (Van Staden et al., 2016).Incremental Introduction of African Languages (IIAL) is implemented in grades 1-12, aiming at promoting, strengthening and preserving African languages through teaching them.
Much like that of other developing countries where most discussions also centre on conventional factors such as physical and human resources, with rare mention of the nonconventional factors identified above, South African literature on the factors contributing to learners' underperformance also points to SES, family and community dynamics, and poverty and inequality (including lack of appropriate nutrition) (Besharati, 2016)."Other factors linked to learner achievement include household income, household size, presence of both parents, home language, race, access to television, toys and other learning materials" (Besharati, 2016: 124).Poor school infrastructure is also often in the spotlight as is the violence experienced by some learners at home and in the community (Bloch, 2009).Related to the factors discussed above from the immigrant literature, we find that the link between parental involvement and learning outcomes has been reported (Fleisch, 2008;Luxomo & Motala, 2012;Singh, Mbokodi & Msila, 2004).However, the South African literature mainly describes the difficulties associated therewith (Singh, Mbokodi & Msila, 2004;Stephinah, 2014).We could not find reference to any actual experiments or strategies.
Creative parent-home interface programmes have also largely been missed by the various education interventions in South Africa, except in some school development programmes implemented by Joint Education Trust (JET) 5 , while relevant government interventions have been the National School Nutrition Programme (Graham et al., 2015) and the No-Fee School Policy (Sayed & Motala, 2012).

Conclusion
This paper asks whether language ability is key to scholarly success and examines this factor in relation to the pockets of academic success that occur among migrant children.Because our interest is to understand what South African children (and their parents) can learn from migrant children about how to navigate the language barrier, we identify and focus our discussion on factors that are relatively flexible, less persistent across generations and therefore more malleable.
The main findings from our review of the relevant literature pertained to the importance of being conscious of and harnessing non-conventional factors, such as aspiration, expectation and creativities.This is reflected not only in the education policy shifts towards greater acceptance of diverse ethnic values, cultures and identities, thereby instilling a sense of belonging; and acknowledgement of the importance of parental expectations; but also facilitating the creativity that emerges in many schools and parental involvement programmes.These non-conventional factors have received increasing attention in school effectiveness and learner achievement in the international literature (Fouts, 2003;Lee & Shute, 2010;Marzano, 2003), yet recognition or acknowledgement in South Africa and other developing country contexts remains limited.Instead, literature on school effectiveness in developing countries usually still focuses on conventional tangible, observable, and measureable factors, such as school resources, teachers' qualifications and experience, class size and language of instruction (Yu et al., 2015).Much of the South African literature also discusses the different language policies in education and the factors that hinder their successful implementation, while failing to conceive, experiment with, or report on different interventions.
We recognise that migrants, despite difficulties in adapting to a new environment, tend to see their "adjustment problems as temporary…[and are therefore] more creative in inventing pragmatic solutions to their current predicaments," (Kao & Tienda, 1995: 5).This might be different in poor or local low SES communities that have become disillusioned with the educational outcome and system and have given up the dream of exiting poverty (Kao & Tienda, 1995).However, we reinforce the importance of having a different mind-set, a mindset that believes things can change and in turn induce creative and practical solutions to address the difficulties.
In terms of prevailing parental objection to bilingual education, we suggest that the message could be modified -that early exposure to English is good, but in its absence, bilingual schooling is better.We believe that this message might find wider acceptance amongst parents and could potentially rally their greater support for bilingual education.