Language biographies and multilingual language use: A sociolinguistic study of young refugees from Syria, Iran, and Afghanistan living in Germany

: This paper describes the individual experiences with migration - related multilingualism of young refugees from Syria, Iran, and Afghanistan living in Germany. Young refugees are a social group of particular linguistic interest because their late L2 learning takes place with limited family background and limited maintenance of heritage language and culture. These interview excerpts are connected to the current dissertation project titled “ Sprachbiographien: Das Beispiel junger Ge ﬂ üchteter aus Syrien, Iran und Afghanistan ” ( “ Language Biographies: The example of young refugees from Syria, Iran and Afghanistan ” ) . The nonrepresentative corpus of the study comprises 21 language biographical interviews with female and male refugees between 17 and 27 years. Speci ﬁ c guidelines were adhered to obtain insights into partici pants ’ languages and cultural identities with a special focus on strategies of multilingual performances. The aim of this article is to show how young refugees ’ language biographies are in ﬂ uenced by multilingual language use, based on selected samples from this survey.


Introduction
For the European Member States, 2015 was characterized above all by the immigration of asylum seekers. The rise in the number of asylum seekers led to administrative and infrastructure crises in many countries, which became the dominant topic of political debates under the term "refugee crisis"both at European and national levels. Approximately 890,000 asylum seekers entered Germany in 2015 (c.f. BAMF 2015): 476,649 asylum applications (first and subsequent applications) were registered compared to just 202,834 in 2014; this corresponds to an increase of 135% in just one year.
The increase in immigration between 2015 and 2016 was mainly due to refugees from Syria, Iraq, and Afghanistan. The substantial group of unaccompanied minor refugees who entered the country, mostly from Afghanistan, Syria, Eritrea, and Somalia, presented a particular challenge. The Bundesamt für Migration und Flüchtlinge (BAMF, "Federal Office for Immigration and Refugees") registered approx. 67,000 unaccompanied minor refugees with and without asylum applications in Germany (c.f. BAMF 2017). This quantitative representation of the current migration movements (around 60 million people are displaced worldwide) and the heterogeneity of refugee groups regarding their origin, language, and the cause of flight are not included.
The divergent concepts of German language and inclusion policies and the efforts to develop didactic concepts to improve the linguistic inclusion of young refugees in school have not included the subjective experience of those affected by educational policy measures so far. Therefore, the focus of the present work is migration-related multilingualism using the example of young refugee's language biographies. As a generation of migrants with traumatic experiences fleeing their home countries, their needs should receive special attention. These unaccompanied minor refugees work to fulfill the requirements of Germany's integration policy, while they are at the same time dealing with their own sense of identity and language development in an unknown culture.
This article will discuss the exploratory qualitative analysis of young refugees' insights regarding language learning, the connection between language and positioning, and the impact of biographical experiences of language learning on multilingual language use. For this purpose, the theoretical background of language biographies will first be examined. To conclude, it will be shown how language biography research can be reconciled with the empirical study of migration and multilingualism.

Language biographies in multilingual research
The narrative-qualitative approach of German language biography research offers the opportunity to "complement in particular studies related to migration and multilingualism or to lead to a better understanding of the results" (Holzhauser and Ritter 2019:37, my translation). The different forms of migration share a resulting necessity: a new beginning in mostly all areas, often combined with the acquisition of a new language. The consequences can shape subsequent generations and are defined as a "biographical break" in research (ibid.: 34, my translation). The latter describes the consequence of a "language break when they are cut off from their previous sphere of communication, or a language transition, when contact communication is possible in transitions as well as in language attempts" (cf. ibid, my translation). Thus, the migration-related multilingualism is the object of investigation in language biographical studies.
With the narrative or discursive turn of the 1970s, narratives were manifested as a scientific subject of study in the humanities and social sciences (cf. Busch 2011;Pavlenko 2007). This speaker-centered approach was and is characterized by a wide variety of disciplines: psychological, sociological, and literary moments play an important role in language biography research. Autobiographical narration in the context of interviews finally found its way into linguistics and foreign language research at the end of the 1990s (see Busch 2017). The aim of studying autobiographical narratives should be to capture individual experiences of learning a foreign language in a social context. This development led away from the idea of a linear language acquisition towards a speaker-focused situation of acquisition (see Pavlenko 2007), which considers social, ethnic, gender-specific, and emotive contexts.
Thus, in second and foreign language research, the focus was no longer only on the degree of successful language acquisition with failure analyses, but also on the personal-social level of multilinguals, such as language attitudes, language use, and the context of (multilingual) practice. Language biographical data can be collected using a wide variety of written or oral methods which are often combined: language biographical interviews, language learning diaries, or literary autobiographies or memoirs (see Busch 2011Busch , 2017. The visualization of language biographies by language portraits (see Krumm and Jenkins 2001) or the written collection of language biographical data (cf. Franceschini 2002) represents further possibilities for language biographical data collection.
Until today, language biographical surveys have been conducted from a wide range of perspectives (see Keim and Knöbl 2007;König 2014;Veronesi 2010). In addition to the preoccupation with individual multilingualism (see Franceschini 2010), social multilingualism (Riehl 2014:12) is also in the foreground of linguistic studies. Aneta Pavlenko (2005) engaged the interrelationships between multilingualism and emotions. Bärbel Treichel (2004) takes a collective-social perspective on multilingualism using the example of Wales. In addition, the inter-influential correlation between language and identity plays a decisive role in the field of recent language biography research and migration-related multilingualism (König 2014;Thiez 2018).
In the context of the present study, language biographies can be defined as sociolinguistic research subject, which is constructed by a dialogical interaction/stimulus with the focus on language acquisition, (multi-)language use, context of language use, and language attitudes.

Analytical method
The survey design is based on Rita Franceschini's Center-Periphery-Model, CPM (c.f. 2001). As mentioned earlier, Franceschini emphasizes that a person's linguistic autobiography is not static, but dynamic. Initial competence of an individual's first language is enhanced by competence of other languages and varieties. However, linguistic competences are not all connected with the same functions and identifications (ibid.). The development of each language and code has to do with the individual and their experiences over the course of a lifetime. One reason for the resulting linguistic repertoire can also be migration. The resulting feature of the migration-related language biography is the highly uncontrolled language acquisition, which is particularly linked to situations and contexts. The dominance of different languages changes over the course of life and enters the center or periphery of language use.
With reference to the language biographical interviews of Rita Franceschini (cf. 2002;Franceschini and Miecznikowski 2004), a guide was developed that focuses on biographical data of language acquisition as well as on language experience and language use. In summary, the guide includes the following topics: the context of language acquisition, the juxtaposition of the different language-cultural communication practices, and the corresponding language attitudes as well as linguistic positioning in the context of linguistic identity. The theoretical discussion of the terms multilingual language use, language acquisition, language learning, or context of language use, which are necessary for this study, cannot be taken into account exhaustively in the context of this article, but is sketched by the presented CPM. As already stated in the introduction, five linguistic study features can be mentioned for migration-related language biographies. The analysis presented in this paper refers to the areas of language acquisition, language use, and context of language use. At this point, the focus of the analysis can be broken down into the following three research questions on young refugee's use of multilingualism: 1. What are the language learning situations in the countries of origin and the countries of migration? 2. What is the pattern of language use in daily life before and after forced migration? 3. What is the context of multilingual practice in the countries of origin and the country of migration?
Based on these criteria and questions, the study focuses on young refugee's multilingual use in hosting society, influenced by migration-related language biographies. In the context of the CPM, the participant's view on controlled or uncontrolled learning situations, language use in private and public settings, and functional language use, limited to social situations/acting, and communication agents are categorized (see Figure 1).
All mentioned categories and definitions must be considered in a dichotomous positioning process between the society of origin and the host society, or between the L1/other varieties and the migrationrelated foreign language.
The selection of the sample was carried out in cooperation with a housing project initiated by the Munich City Youth Office at the beginning of the 1990s and the school "Flüchtlinge in Beruf und Schule (FLÜB&S)" "refugees in education and school." The survey is conducted with the help of lay interpreters, who translated in moments when language comprehension was difficult (written in italics in the excerpts). Based on the dissertation project, data were collected through semi-structured interviews. The qualitative and exploratory approach of data collection is also reflected in the evaluation: content-structuring analysis according to Mayring (2015) serves to categorize the interview data, based on MAXQDA. The procedure in the qualitative data analysis is pointed by integrating key examples.
The participants in the nonrepresentative study are female and male unaccompanied minors (UM), accompanied minors (AM), and unaccompanied adolescents (UA) between 17 and 27 years of age, who had been in Germany on average between two and three years at the time of the survey. The group of respondents consists of a total of 21 participants, seven persons per country of origin. Two thirds of the participants are male (see Table 1). On average, the refugees surveyed are 20 years old. Nine of the participants were one year away from receiving their Qualifizierender Hauptschulabschluss "Qualifying lower secondary school certificate," which qualifies them for vocational training. Eight of the respondents had already passed the "Qualifying lower secondary school certificate" at the time of the survey. Respondents mainly had a language level between B1 and B2 (according to the Common European Framework of Reference). Three of the interviewed participants fled to Germany with their families and belong to the group of accompanied minors (AM). Five of the interviewees belong to the group of unaccompanied adolescents (UA). Thirteen young refugees can be added to the group of unaccompanied minors (UM). All the mentioned classifications of social status were based on the time of migration to Germany.  As the figure shows, most participants speak German as a foreign language at language level between B1 and B2. Most B1-speakers belong to the UA-group and have been living in Germany for two to three years (see Figure 2). Seven to eight young refugees have a certified B2 language level, seven of whom are also members of the UA-group. Three Syrian participants who have language levels between A2 and B1 and have been living on average 3.2 years in Germany fled to Germany with their families as accompanied minors (AM) and were unable to attend school because of the ongoing war situation in their country of origin.

Findings and discussion
In the following, the biographical and linguistic characteristics are summarized. The data are defined in more detail by transcription¹ excerpts from the interviews. It is important to underline that all respondents belong to the group of late L2-or L3-learners who learn German as a first foreign language or further foreign language. The majority of Afghan adolescents belong to the ethnic group of Hazara. Some of them grew up in Iran, in a region around Tehran. However, they consciously belong to the Afghan group of Hazara which is the ethnic-national identity of the parents' generation. Before fleeing to Germany, the two official Afghan languages, Dari (Persian Dialect) and Pashto, were in the focus of attention. Urdu and Arabic can also be counted to the languages that have been used regularly in everyday life. The first language was mentioned mainly in the context of films, the latter was the language that was used in religious education. However, both languages can be placed in the context of prescriptive language use (i am able to read but i do not know what it means): Participant P11 (cf.  I: aber du hast ja dann arabisch gelernt P12: ja das arabisch aber die muslim mussen das lernen aber ich kann lesen aber ich weiß nicht was bedeutet dann Thanks to the migration to Germany and the German school system, Afghan refugees had the first experience of attending a foreign language course (or language lessons with appropriate literacy). This was partly due to the fact that all male participants described that they had to support their families through additional work, for example at the market, and therefore only attended the first three school classes at the maximum. Moreover, the situation in Kunduz, Zindan, or Parwan (three of the places mentioned in the interview, before the adolescents fled from the Taliban to Kabul) was too dangerous for the young refugees.
The Afghan informant P13 (see Figure 1) was also unable to attend school for security reasons and was taught by a private teacher until the time of her marriage. According to this, P11 defines Arabic as a peripheral language, also because, according to its own statements, he has no opportunity to pray (not even in the mosque) because of everyday school life in Germany. As the following example shows, the uncontrolled learning situation is linked to educational policies and self-motivation. In the following excerpt of an interview with a young Afghan woman who was used to learn with a private teacher in Afghanistan developed her own learning strategies (then with a pen always i sat in the hand in subway) to teach herself German, as she had no opportunity to attend a German school. The motivation seems to be particularly high, as she explicitly sought help from contact persons (here: carers) of the hosting society. The associated learning investment is based on the motivation of the respondents to counteract the lack of language opportunities themselves.

Participant P1 and Holzer, February 2018
Excerpt Time: 00:58:12 to 00:59:46 English when i was so until i had not gone to the German school i have watched movies and words what i have always heard in the subway i had extra a small booklet and then with a pen always i sat in the hand in subway and then the people what they said i did not understand i did not care how you write so i did not pay attention to spelling so i just wrote like that and then i came home and then my carers and so i asked what does mean so i did not show that just say that as i wrote this well what does mean water then they told me German als ich so bis ich nicht in der deutsche schule gegangen war hab ich selber so filme angeschaut und hab immer so wörter was ich immer gehört habe in der ubahn oder so ich hatte extra eine kleine so heft gehabt und dann mit ein stift immer in der hand in ubahn gesessen und dann die leute was die gesagt haben nicht verstanden ich hab egal wie das man schreibt also ich hab nicht auf rechtschreibung so geachtet ich hab nur so geschrieben und dann bin ich zuhause gekommen und dann meine betreuerinnen und so hab ich gefragt was bedeutet also ich hab nicht das gezeigt einfach so gesagt wie ich das geschrieben habe naja was bedeutet wasser halt dann die haben mir gesagt All the Iranian interviewees grew up in Iran. Since their parents came from Afghanistan, they feel related to the Afghan minority of Hazara. Afghan Iranian respondents grew up with the local language Farsi and spoke Dari or Hazaragi, the dialect of the ethnic group of Hazara, with their parents at home. Urdu and Arabic were also listed as peripheral languages. The latter were also related to the attendance of Islamic classes (we had to learn until we are able to read it and referring to meaning too but the focus is more on reading):

Participant P1 and Holzer, February 2018
Excerpt Time: 00:15:14 to 00:16:02 English it is an islamic country where we learn yes I had have learned for two years but we had to learn until we are able to read it and referring to meaning too but the focus is more on reading German ist ein islamische land da lernen wir halt ja das hab ich zwei jahre gelernt aber nicht also da nur müssen wir so viel arabisch können dass nur wir lesen können und vom bedeutung auch ein bisschen aber mehr konzentration ist darauf dass wir lesen können halt koran P1 also describes Arabic as a functional-receptive language. Arabic is not included in the language biography constructed by P1 (which consists of Dari, Farsi, Urdu, English, and German). Arabic also seems to be perceived as a peripheral language.
Seven Syrian refugees reported that before fleeing to Germany, they mainly spoke Syrian dialects of Arabic (e.g., Ash Shami, spoken around Mu'addamiyyat al-Sham) and Kurdish in everyday life. They come mainly from the areas around Aleppo, Idlib, Deir-ez-zor, and Al Hasaka. Standard Arabic was taught at the school, also in the context of Islamic classes. Almost all Syrian refugees reported that they also speak Turkish (or have receptive Turkish skills) because they stayed in Turkey for a long time before migrating to Germany. Syrian respondents also stated that they regularly attended English lessons at school. Except for one participant, all Syrian respondents are accompanied minors or accompanied adolescents (at the time of the survey, but not at the time of migration). All accompanied refugees mentioned here speak their L1 (Arabic, Persian) at home for obvious reasons (we are in seven together my parents my sisters and brothers i forgot german a bit).

Participant P6 und Holzer, April 2018
Excerpt Time: 00:22:03 to 00:22:47 English i had more words in english but at the moment i forgot because we are in one up to two rooms and we are in seven together my parents my sisters and brothers i forgot a bit German ich hatte also mehr worte in englisch aber zur zeit habe ich vergessen weil wir sind jetzt in ein zwei zimmern und zusammen sieben also wie es waren meine eltern meine geschwister kann ich nicht habe ich ein bisschen vergessen Young refugees have opportunities to speak German mainly in the context of educational institutions or in contact situations with the hosting society.
Depending on the context, they must move to the center or the periphery of attention via their L1 and L2/L3 according to educational policy guidelines (in the example above, German is L3). Therefore, the acquisition of German as L2 or L3 also depends on the opportunities in the private environment. In the context of the language biographical interviews, young refugees reported that their everyday multilingual practice was predominantly influenced by their respective L1 and German. This applies not only to refugees with the social status of AM/AA (accompanied adolescents, as fourth category of young refugee's social status), but also to UM or UA. German is mainly used for public contact and contexts, while L1 is mainly used in private settings, in peer groups, and also at school, which is demonstrated by the following excerpt:

Participant P11 and Holzer, September 2018
Excerpt Time: 00:13:47 to 00:14:33 English i usually speak german when i go to school yes we talk in german but in the break there are nine afghans in our class yes then i speak more persian German normalerweise wenn ich in die schule geh ja wir sprechen deutsch aber in der pause es gibt so in unsere klasse es gibt neun afghaner ja dann ich spreche ich mehr persisch German influences on L1 are noticeable in two aspects: On the one hand, through code mixing phenomena, especially in the field of Insertion (cf. Muysken 2000), as the following example shows (words which are often used by people for example just): Participant P11 and Holzer (here I.), September 2018 Excerpt Time: 01:18:34 to 01:21:12 English P12: if i talk to my friends for example he can understand German he has c1 now he knows what i mean when i talk in persian the i use one or two words in german I: ok and when do you mix the languages P12: words which are often used by people for example just German P12: wenn ich mit meinem freund beispiel spreche er kann auch er hat jetzt ce eins (c1) er weiß was ich meine wenn ich persisch spreche dann kommt ein oder zwei wörter auf deutsch I: ok und wann mischst du die Sprachen P12: die worte das die leute viel benutzen beispiel einfach On the other hand, influences can also be detected in the context of lexical language, especially among young refugees who were between fourteen and fifteen years old at the time of migration (see P1/P2):

Participant P2 and Holzer, with lay interpreter May 2018
Excerpt Time: 00:39:01 to 00:39:28 English i am shy when i talk in arabic in comparison to german because there are words that are normal in german but if you use them in arabic then they are so gigantic and therefore you cannot use them German im arabischen bin ich etwas schüchterner als beim deutschen sprechen, weil es gibt wörter die im deutschen voll normal sind aber wenn man sie im arabischen benutzt dann sie sie so gigantisch und deswegen kann man sie nicht verwenden At the time of this survey, young refugee's daily life is still predominantly influenced by their L1. The German language is mainly used in relation to the public environment, the educational system, and in contact with the hosting society. In addition, about half of the respondents from Syria, Iran, and Afghanistan speak up to four languages in everyday life: English (language of education), Turkish (as another migration language), Arabic (as a language of religion and contact, depending on the refugees' housing situation), or the dialects of the respective L1. Languages are strongly linked to ethnicity, especially for Persian speakers. The following example of a young Afghan shows the connection between language use and positioning in relation to language and ethnic affiliation: Participant P4 and Holzer, with lay interpreter, May 2018 Excerpt Time: 00:18:40 to 00:19:44 English in general i feel discriminated in almost every language i am able to speak there is always something said or misunderstood even within my own circle when I speak pashto dari speakers do not accept it because of the ethnic conflicts in afghanistan and when i speak pashto the pashto guys do not like it either because they say I am alienating myself and even when i speak German in the afghan context it is perceived as alienation and arrogance.
German generell erfahr ich fast in jeder sprache in der ich spreche diskriminierung sei es deutsch da erfahr ich jeden tag hier diskriminierung weil es wird immer irgendwie was dazu gesagt oder das wird falsch aufgefasst und auch innerhalb des eigenen kreises wenn ich pashto spreche gefällt es den dari sprechern nicht weil eben auf grund dieser ethnischen konflikte in afghansitan und gleichzeitig wenn er dari spricht gefällts den pashto jungs nicht weil sie sagen der entfremdet sich und auch wenn er deutsch spricht in diesem afghanischen kontext wird das als eine art entfremdung oder überheblichkeit wahrgenommen The young man describes his difficulties in communicating with the hosting society, but the language contact situation with the hosting society leads to conflicting positions: ethnic conflicts which were originally prevalent in Afghanistan also lead to a linguistically constructed external positioning, which also concerns German as language of migration.

Conclusion
The analysis of the presented CP-Model and the selected excerpts of interviews demonstrated that refugee's L1 is predominantly in the focus of their language biography, living in Germany with an average stay of three years.
Young refugees used German as a public language, independent of the mentioned social categories UM, AM, UA, and AA. The participants including those living in ethnically heterogeneous housing initiatives are mainly ethnically equal to their peer groups. Languages from the center of attention that led to multilingual everyday life in the country of origin have moved to the periphery of language attention, unless they are used as a contact requirement (e.g., in housing initiatives or schools). For all refugees, there is usually no contact with the hosting society in the private setting, especially not with age-appropriate groups. In exceptional cases, young refugees were brought into contact with the hosting society mainly through voluntary initiatives over specific periods of time, with most respondents believing that only contact with persons who speak German as L1 would enable language skills. German teaching in the hosting society was mainly associated with learning grammatical structures. It was criticized that writing skills and text comprehension would not be sufficiently facilitated. In addition, the language biographical narratives referring to the country of origin showed that the teaching concept was highly authoritarian, gendersegregated, and above all in frontal interaction. This learning concept led to a contrast situation in German classrooms.
The presented analysis focuses on both the language biographical experiences related to multilingual use/language learning and the connected formation of the multiethnic group of young refugees, living in Germany. The exploratory analysis of the connection between language biographies and multilingualism has shown that language biographies created in narrative interviews allow conclusions about the use of multilingualism: Center languages pass over into peripheral languages and multilingual practice is transformed into a functional, context-sensitive bilingualism, which is separated in public and private language use. In addition, it was shown that multilingual use is not only tied to personal and situational contexts, but is also linked to biographical experiences. Depending on the language positioning in the CP-Model mentioned here, multilingual use is influenced by language biographical experiences. Not only the multilingual use, but also the language biography as an overall production are influenced. The observed characteristics of multilingual use, which are influenced by language biographies, are based on exemplary case analyses. Whether these are characteristics that occur among young refugees at the beginning of their migration must be checked on a broader empirical basis within the framework of this dissertation. engagement. Thank you to all the employees of Auerhaus München and FLÜB&S for their trust and support. Last but not least, I am grateful to my colleagues Christina Mutter and Melanie Keller for their patience and help.
Funding information: Authors state no funding involved.
Data availability statement: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.