Language complexity, narratives and theory of mind of Romani speaking children

Размещено на http://www.allbest.ru/

LANGUAGE COMPLEXITY, NARRATIVES AND THEORY OF MIND
OF ROMANI SPEAKING CHILDREN

roma children grammatical knowledge

Hristo Kyuchukov

University of Silesia in Katowice, Poland

Jill de Villiers Smith College in Northampton, USA

Abstract. The paper presents research findings with 56 Roma children from Macedonia and Serbia between the ages of 3-6 years. The children's knowledge of Romani as their mother tongue was assessed with a specially designed test. The test measures the children's comprehension and production of different types of grammatical knowledge such as wh-questions, wh-complements, passive verbs, possessives, tense, aspect, the ability of the children to learn new nouns and new adjectives, and repetition of sentences. In addition, two pictured narratives about Theory of Mind were given to the children. The hypothesis of the authors was that knowledge of the complex grammatical categories by children will help them to understand better the Theory of Mind stories. The results show that Roma children by the age of 5 know most of the grammatical categories in their mother tongue and most of them understand Theory of Mind.

Keywords: Romani, language assessment, Theory of Mind, language complexity.

Кючуков Хрісто, де Віллєрз Джіл

Мовна складність, наративи й «теорія розуму» ромських дітей

Анотація. У статті представлено результати досліджень, проведених за участі 56 ром-ських дітей з Македонії та Сербії у віці від 3 до 6 років. За допомогою спеціально розробленого тесту було оцінено знання дітьми ромської мови як рідної. Завданням тесту було визначенння рівня розуміння мови у дітей, а також їхнє володіння різними видами граматичних знань, таких як: wh - питання, wh - доповнення, пасивні дієслова, присвійний відмінок, час, граматичний аспект, здатність опановувати нові іменники та прикметники, а також повторення речень. Крім цього, діти отримали два ілюстрованих наративи про «теорію розуму». Гіпотеза авторів полягала в тому, що знання дітьми складних граматичних категорій допоможе їм краще зрозуміти розповіді про «теорію розуму». Результати довели, що ромські діти у віці до 5 років знають більшість граматичних категорій рідної мови, і більшість з них розуміють сутність «теорії розуму».

Ключові слова: ромська мова, мовна оцінка, «теорія розуму», мовна складність.

Introduction

Triple Jeopardy

The Roma children in Europe all too often grow up in poverty, marginalization and stigma. From the point of view of the language they speak - Romani - they grow up in a kind of triple jeopardy compared to other children in the world Their families are often poor and in most of the cases grow up in families where the parents are uneducated.

a) From a very young age they become multilingual, learning other dialects of Romani and/or other minority languages, but in most of the cases not learning the state language from birth.

b) They are part of a stigmatized group, a group that is often subject to discrimination in housing, employment and education (Vassilev, 2004; ERRC, 1999, 2015; Kyuchukov, 2005, 2010). Since the status of a language is perfectly predicted by the status of the people who speak it, Romani is often regarded as a deficient or degraded language.

Consider first the impact of socioeconomic status (SES) on language in the US and in West Europe, SES is blamed for the word gap (Hart & Risley 1995), in research that has demonstrated that lower class children are exposed to 30 million fewer words spoken to them than are spoken to middle class children between 2 and 5 years. In the US this lack of exposure has been linked to poorer performance in the early grades of schooling, and a large gap in achievement in education. However, many psychologists and linguists are not convinced that the count of words is the best index of rich language exposure, it is merely a proxy for conversational attention. More recent research findings show that it is not the count of words that matter, but conversational turns, and rich conversations that acknowledge the child as a participant (Hirsh-Pasek et al., 2015).

Kyuchukov (2014) has studied the language in the home of Roma families in Eastern Europe, and draws the conclusion that their style of interaction is in fact conducive to rich language learning despite the lack of material wealth and parental education. Although the Roma families are poor and the parents are uneducated the children grow up in an emotional and loving environment, which supports their language and cognitive development. The Roma children learn Romani in a way that is not traditional for educated families in the West. The Roma culture being an oral one is rich with anecdotes, songs, lullabies, fairytales, and folkloristic language games. From a very early age the children get exposed to complex grammatical structures and different registers, because living in extended families the care of the growing child is a responsibility of the whole family and very often of the whole community. Everyone in the family (and community) is free to talk and to play with the child. In Roma culture if a neighbor speaks to your child, makes jokes, sing songs to him or her, this is considered to be a sign of respect to the family. This is not the modern West European culture and style of communication, where the children get knowledge about the world through books and conversations in a nuclear family. The young children learn from everyone: from older brothers and sisters, grandparents, cousins, neighbors.

The expectation for the young children is that by the age of 2 years they should have not only enough lexical knowledge but also they should understand more complex sentences in instructions such as “Bring me X, which is in Y” The child might be in a situation such that s/he never before saw, heard or used that word or knew about the object (Kyuchukov, 2005). The parents expect that the child will have enough knowledge of the language and enough experiences to find the object they are asked for, by making inferences for example about what is new, or what the parent might need to complete a task. If the child cannot find the correct object and s/he brings another one, then the parent does not continue by showing the correct object and naming it, but rather the parent will repeat the request until the moment the child by himself finds the object the parent needs. The child is given autonomy to work things out, and so learning takes place in a quite different style. All these actions of the adults towards the children in Roma communities show the expectations of the parents that their children can master the language within communicative situations from an early age. They have to get the knowledge about the surrounding world through actions, and they have to repeat the activity until they are successful, until that moment they learn the new words and start to use them. (Reger, 1999; Reger & Berko-Gleason, 1991; Kyuchukov, 2015)

Nevertheless, differences in a style of language development do not necessarily result in the same readiness for Western schooling. Heath (1982) and others have shown that schools tend to represent the learning styles and conditions of Western middle class families, and coming to that by a different route can leave a child ill equipped to fit in the culture of school (Hoff 2013). In the famous study by Heath, lower class White and African American children were taught within their families in different ways than the middle class White families, despite all living in the same state (North Carolina) in the US. When they arrived at school, the middle class children's preparation with connecting the content of books to events in their lives, ostensive naming of objects, and “test” questions from their parents all placed them in an advantageous position with regard to school practices. That is not to say that the other groups experienced a less rich language environment, but the skills they acquired were at variance with the school's demands.

The Roma children thus face the same problems from their different style of language development when it comes to readiness for schooling as do other socioeconomically disadvantaged children from different cultures and children at risk (Merz et al., 2015; Landry et al., 2013). Even growing up in rich language circumstances, the disparities between school and parenting style reduce the probability of academic achievements of the children with low socio-economic status (Heath, 1981).

Turning to the second jeopardy, the multilingualism among the Roma children is considered to be a problem. It is widely acknowledged that learning several languages at once slows children down in mastering each language relative to a monolingual for about the first 7 years of learning (though of course there are individual differences) (Hoff & Elledge,2005). However, when one counts the knowledge that is distributed across the languages, then bilingual children are just as linguistically competent as monolinguals (Bedore, Pena, Garcia & Cortez, 2012). And the advantages of bilingualism are everywhere promoted. Multilingualism can help a child cognitively even before the languages reach mastery, in areas such as executive function and metalinguistic awareness in which a bilingual advantage has been found (Bialystok, 2001; Bialystok & Craik, 2010; Carlson & Meltzoff, 2008). Millions of dollars are spent on higher education to teach second languages to monolingual high school students, college students and adults so they can be functional in the modern world. Recent research has documented the protective effects of multilingualism on the aging brain, with reports of a two-year lag in the onset of senile dementia in people who speak more than one language (Bialystok, Craik & Freedman, 2007). It is important therefore to put the Roma child's bi- or multilingualism in perspective. The child may not be as competent in each of the languages as a monolingual peer at the younger ages, but it is important that this must not be judged as a problem in the early grades. By age 8 or 9 years, bilingual children typically catch up to their monolingual peers in their languages, unless there is an additional factor that comes into play, namely, the status of their home language. When there is stigma attached, as with African American English in the US, or even Spanish in some regions of the US, then a third source of jeopardy comes into play.

Usually the Roma children speak Romani, which commonly does not have prestige in the societies in which they live. Given their socio-economic circumstances, they often speak language(s) and dialect(s) of other minority groups who also are without respect in their societies. Some European researchers even consider that the greatest problem of the Roma children is their knowledge of Romani as a mother tongue, and that is seen as an obstacle for better acquisition of the official language of the country (Bakalar, 2004). In several European countries, if the Roma children do not have mastery of the official language of the country by the age of 6 when they have to enter primary school, they are considered to be cognitively impaired and are sent to special schools (ERRC, 1999; 2015; Fremlova & Ureche, 2011). For example, in Slovakia up to 65 % of the Roma children are in special schools because of lack of knowledge of Slovak language (Lajcakova, 2013). Some of the Roma children who speak Slovak to some degree, or who know a dialect variety of Slovak, are still treated as needing special schools designed for children with serious mental handicaps (Friedman, Gallova Kriglerova, Kubanova & Slosiarik, 2009; Kyuchukov, 2013, 2014)

The stigma against the Roma people (and therefore against their language) can be a major risk factor, just as with children who speak the stigmatized dialect of African American English in the US. That's why it is vital to make sure people do not consider differences as deficits, that educators and policy makers must acknowledge their prejudice and clear it away for fair testing. In the US, the second author has been involved in creating fair language tests for children who speak African American English, to demonstrate their competence in their home language (e.g. the DELV test: Seymour, Roeper & de Villiers 2005). However, there are no linguistic or psychological tests in Romani. In order to reduce the prejudice and discrimination against the Roma children, a first important step would be a test of the Romani language for children. If it can be demonstrated that the children have adequate mastery of their mother tongue there is less likelihood that they will be seen as needing special education, and appropriate provision can be made for introducing them to the language of the state without denigrating their home language. This is why we have developed such a test and we are trying to answer some important questions with it. Before we describe the test, it is necessary to situate the research we report within the special circumstances in Sweden.

Research Questions

How much do the children know of the Romani language by the age of 3-6 and which grammatical categories do they know? These are some of the research questions which we try to answer. However, the test we have developed is far from standardized yet. In some recent publications we have looked at how much the Roma children in Bulgaria and Slovakia know of the grammar of their mother tongue using this specially designed and developed test on language assessment. The first results on small samples (30 at a time, ages 3 to 6) show that most of the Roma children by the age of 5 perform at about the 80 % level on most of the grammatical categories of their mother tongue. Not only that but they can transfer the knowledge, e.g. applying the new rules to brand new words given to them in the test (Kyuchukov, 2013; Kyuchukov & de Villiers, 2014a, b).

Though we are encouraged by these initial results, we need ways to validate the test against other indices if the child's language competence. It is usually the case that a new language test is compared to linguistically sophisticated analyses of the spontaneous speech of the child, to ensure that it reflects the real life language the children can produce (Pearson, Jackson & Wu, 2014). But there are hazards here too, in that some young children may be shy or timid and not speak much, and a long transcript might not contain much of anything for the linguist to analyze. More fruitful might be to use more controlled circumstances to obtain some elicited speech, and then the use of a wordless picture book is helpful to have the child create a short narrative (Berman & Slobin, 2009). Therefore, another research question to answer is: can the same Roma children create a narrative looking at a series of pictures? We recognize that most of them do not have experiences creating a coherent and cohesive story looking at pictures. Children of this age often produce narratives that are barely describable as such, often being incoherent, and lacking any plot or connecting devices. Before age 5 or so, children are prone to tell stories using what Bruner calls “the landscape of action”, in which they simply recount or name the actions that happened. But true stories requite integrating this language of events with the landscape of consciousness: who were the characters, what were their intentions, goals and beliefs? Only older children can do this with some competence. It has been argued that narratives that include such mentalistic themes reflect the children's growing understanding that others have minds, with contents that may be different than their own, i.e. Theory of Mind. In fact, in the case of African American children studied by Curenton (2004), the children showed competence in their narratives even when they were not performing as well as their white peers on standardized tests of Theory of Mind reasoning.

The inclusion of a narrative task with specific questions relating to Theory of Mind was therefore justified on two grounds: one, to offer some convergent validity to the language test, and second, to investigate the relationship between language competence and Theory of Mind, both in the linguistic test and in the narrative itself. Until now in the scientific literature there is no information about the connection between language complexity, narrativity of Roma children and their Theory of Mind. This is a first attempt to give an answer to such a research question.

Design of the Test

The language test is composed of ten subtests (Table 1). Taking note of recent recommendations in the field of assessment for language, we measured not only what a child knows (language products), but also their ability to learn (language processes) (Hirsh-Pasek, Kochanoff, Newcombe, & de Villiers, 2005; de Villiers et al., 2014).

Table 1

Components and modules of Romani test