Literature review

Lavelle (2007, p. 219) defines writing as a cognitively complex task, involving multiple attentional demands, strategies, and processes, which is affective involving intentionality and self-expression as well. She claims, that it is both an art and a science, inspired yet routine, reflective yet directive.”

Learner language is the oral or written language produced by learners. It serves as the primary data for the study of L2 acquisition (Ellis & Barkhuizen, 2005, p. 4). One of the most striking differences in gathering samples to analyse learner language is that written samples are permanent, and for this reason, they are easier to collect (Ellis & Barkhuizen, 2005, p. 28).

Beyond oral and written distinction, we can also distinguish between naturally occurring, clinically elicited, and experimentally elicited performances. In the case of clinically elicited writings, the focus is on the message, whereas in experimentally elicited writings, the focus is on the form.

It is not always simple to put writings into one type, as they can appear on a continuum.

Weigle claims that while writing in the first language is challenging, it is more so in second language: in addition to limited linguistic resources, second-language writers may be disadvantaged by social and cultural factors. She adds that motivational and affective factors play a role as well. Furthermore, the issue of time constraints is salient for them, because they are unable to write as fluently and quickly as their native speaker peers (Weigle, 2002, pp. 36-38).

A corpus can be defined as a collection of texts or parts of texts upon which some general linguistic analysis can be conducted (Meyer, 2002, p. xi.)

Modern-day corpora are of various types ranging from general/reference corpora vs. specialized corpora (e.g., British National Corpus and Bank of English), through historical corpora vs. corpora of present-day language (e.g., Helsinki Corpus, ARCHER); regional corpora vs. corpora containing more than one variety (e.g., Wellington Corpus of Written New Zealand English); learner corpora vs. native speaker corpora (e.g., International Corpus of Learner English); multilingual corpora vs. one-language corpora); to spoken vs. written vs. mixed corpora (e.g., LLC = London-Lund Corpus of Spoken English) (Nesselhauf, 2005).

Meyer claims that corpora have numerous uses, ranging from the theoretical to the practical, making them valuable resources for descriptive, theoretical, and applied discussions of language. He also adds that as corpus linguistics is a methodology, all linguists could use corpora in their studies of language.

(Meyer, 2002, p.45).

The benefits of collecting linguistic data from language learners are manifold. By analysing them with appropriate computer software, new scientific results can be gained and used for compiling new curricula or monolingual dictionaries. Furthermore, they can create a basis of a new programme that is able to correct mistakes made by the learners (Durst, Szabó, Vincze & Zsibrita, 2014a). Conrad (2005) urges an increase in such second language learner corpora. She suggests that larger scale, cross-sectional, longitudinal studies should take place in SLA corpus linguistics, therefore, from larger databases greater influence could be exercised on L2 teaching in the future.

For analysing Hungarian as a foreign/second (HFL/HSL) (learner) language, two corpora have been compiled so far. One was published in the USA at Indiana University (Dickinson & Ledbetter, 2012, 2015), the other one, called

HunLearner, was published at Szeged University, Hungary (Durst, Szabó, Vincze & Zsibrita, 2014b). Currently, the Indiana University corpus contains data from 14 learners at three different proficiency levels. Nine are so-called beginners, one is intermediate and four are advanced level. They total in 9391sentences with ten annotated journals. The texts are composed of entries on various topics chosen by the student, and each one is ten to fifteen sentences in length. HunLearner is a new corpus that includes essays written by learners of Hungarian. In the project launched by Szeged University, written data were collected from 35 students majoring in Hungarian studies at the University of Zagreb, Croatia. Texts were morphologically and syntactically analysed by the magyarlanc tool (Zsibrita, Vincze & Farkas 2013). The topics were Egy szimpatikus ember [A nice person] and Magyarországról és a magyarokról [On Hungary and Hungarians]. Learners were at A2, B1 proficiency levels.

Currently, the corpus consists of 1427 sentences and 22000 tokens.

3. The study

3.1 The research questions

The aim of the study was to investigate the language proficiency and typical linguistic patterns of immigrant students throughout their written performances in their target language, Hungarian, as well as to study their attitude and feelings towards the topic of mother tongue presented in their course book.

The research questions were as follows:

•! Which language proficiency levels are students at?

•! What are the typical lexical and linguistic patterns in their written performances?

•! What types of mistakes do they make?

•! How do they express their identity through the essays?

3.2 Method

3.2.1 Research context and participants

The context of my study is the immigrant community in the Hungarian – Chinese Dual Language Primary School, Budapest. At present, this is the only Chinese – host country bilingual school in Europe, established in 2004.

Currently, 360 pupils are enrolled at the school. Some 110 of them speak Mandarin Chinese as their first language, and there are altogether 123 students whose mother tongue is not (or not only) Hungarian. The number of Hungarian pupils has outnumbered immigrant students’ year by year. There seems to be a growing interest in and popularity of the Chinese language among Hungarian parents. In this school, all students learn both Chinese and Hungarian languages

and civilization. A more detailed description of the academic context can be found in Lukácsi-Berkovics (2015, pp. 159-160).

The participants of the study were eighteen 8^th grader students, whose first language (L1) was Mandarin Chinese with one exception. There was also a Russian as a first language speaker pupil. These boys and girls were at about the age of 14 at the time of the study. Seven of them learned Hungarian as a second language (HSL). These students belonged to the so-called MID class.

MID is an abbreviation of “magyar mint idegen nyelv” meaning Hungarian as foreign language class (HFL). Although the school management is aware that Hungarian here is a second and not a foreign language, the reason for calling it

“idegen” [foreign] is simply practical: it is easier to pronounce it MID for Hungarians in everyday use (instead of MMINY).

Eleven pupils were considered proficient enough in Hungarian to take part in fully immersed education. It means that these students learned Hungarian literature and grammar, and all the other subjects along with their Hungarian L1 classmates. In the analysis section of my study, I will compare the performance of these two subgroups as well.

Besides the mother tongue being mostly Chinese, there was another common feature of all the students: as all of them had been taking part in the Hungarian education system for more than one academic year, by law they were all required to sit for the annual test called ‘The National Assessment of Basic Competencies’ (Oktatási Hivatal, 2014, p. 7), a test of mathematics and reading comprehension skills originally designed for HL1 students. Officially, L1 is not a differentiating factor when assessing learner competence.

3.2.2 Data collection instruments and procedure

To answer the research questions, my data collection included language proficiency tests, a questionnaire, classroom observations, and students’ written performances. The instruments used for measuring students’ language level were the following: (a) an A2 language exam validated by ELTE Origo Examination Centre Ltd. (ELTE Origó Nyelvi Centrum Kft., 2013); (b) a B1 language exam also validated by ELTE Origo Examination Centre Ltd. (ELTE Origó Nyelvi Centrum Kft., 2013); (c) a Student’s questionnaire (pilot version);

and (d) a Teacher’s questionnaire (pilot version). Some of the findings related to the level of proficiency were discussed earlier (Lukácsi-Berkovics, 2015, p.162). For further information about linguistic and lexical patterns used by these students, I collected essays from them on a complex topic. In the present study, I will focus on this area.

In the grammar course books for 8^th graders (Lerchné Egri, 2013), there is a section devoted to languages, especially focusing on the mother tongue among the other languages in the world. The book provides texts in the topics of mother tongue, bilingualism, the nature, history and typology of languages, and also further activities, typically research based ones and written tasks. One of the written tasks by the end of the section is to write an essay titled Édes anyanyelvem [My sweet mother tongue]. As L1 was not Hungarian for all the

students, they were required to do some preparatory research about their mother tongue and participate in class discussions before embarking on the task. The participants were encouraged to use their notes in the process of writing the essays. The course book and work book also suggest process writing for completing a final essay.

The study was carried out in two consecutive academic years, in 2013 and 2014. The data for language proficiency were collected in both years in two waves: October and April, the essays were collected between the two waves each year.