“I hardly ever practice the real Standard German.” Self-reported language use and language proficiency in South Tyrol (Italy)

This paper deals with qualitative data on language use and self-assessments on language proficiency in multilingual speakers in South Tyrol (Italy). The first aim is to investigate empirically whether there has been a change in young people’s language use within certain domains (family, friends, and school). The second aim describes whether, after years of segregation between the language groups, a change in pupil’s language proficiency could be observed. Moreover, we also examine which extra-linguistic factors may influence their language competences in German, Italian and English. One way to investigate these aspects is through an empirical study on self-reported language use and language proficiency by the speakers themselves. The current survey was conducted in South Tyrol, an official trilingual province in northern Italy. The qualitative analyses are based on biographical interviews provided by 65 high school graduates (mean age: 19.3 years) attending German-language high schools in South Tyrol.


Introduction
South Tyrol is a trilingual province in northern Italy, where three languages are officially recognized: Italian, German, and Ladin. Within Italy, the German-language group is a minority, while in South Tyrol itself it forms the overwhelming majority (section 2). Besides a Bavarian variety (Wiesinger 2000), which is their first language, members of the German-speaking community have also a regional German standard variety (Abfalterer 2007; Ammon/Bickel/Lenz 2016; Ciccolone 2010) as well as Italian in their linguistic repertoire.
In this research, we focus on qualitative data on self-reported language use and self-assessments on language proficiency in multilingual speakers attending German-language 1 high schools. In order to see whether there happens to be a shift in language in the South Tyrolean context, language use in the following three domains -family, friends, and school -is investigated. Contact with the other language group as well as the use of the second language (L2) are indispensable for individual bilingualism of the South Tyrolean population. Although direct contact with the Italian language and its speakers may indeed positively influence subjects' motivation to learn the language, the heterogeneous geographical distribution of the language groups as well as the parallel existence of linguistically separated institutions, such as schools, cultural associations or sport clubs, often constitute an obstacle to effective/advanced L2 acquisition (section 2).
Data comes from 65 biographical interviews conducted within the KOMMA study 2 (section 3). By choosing to focus on self-reports of usage and proficiency, a researcher decides to work with subjective data, participants' interpretations and personal perceptions of language competences. Although the data does not (necessarily) represent facts as such or objective measures, it should instead be understood and analysed as experience reports (Franceschini/Miecznikowski 2004: XIII;Pavlenko 2008: 324). The subjects' descriptions of their own linguistic behaviors and competences, therefore, do not necessarily correspond to reality. Additionally, biographical interviews are always filtered by numerous factors such as desired self-image or memory, personal attitudes toward one's languages, the setting in which the interview takes place, the interviewer herself/himself, and many more, all of which must be considered while analyzing and interpreting the subjects' responses (for a critical overview regarding autobiographical narratives, cf. Pavlenko 2007).
This paper is organized as follows. Section 2 provides an overview of the (socio)linguistic situation and the territorial distributions of the language groups in South Tyrol, while section 3 discusses the methodology employed in the study. This section also presents the research questions in section 3.4. Section 4 represents the main body of the study: the qualitative findings obtained from the biographical interviews. A discussion and final conclusion are presented in section 5.

Sociolinguistic Situation in South Tyrol
According to the 2011 Census, which also provides information on the ethnic-linguistic affiliation of the residents of South Tyrol, 69.6% of the population declared themselves as belonging to the German-speaking language group, 25.8% to the Italian-speaking group, and 4.6% to the Ladin-speaking group 3 (ASTAT 2019: 15). The two main linguistic groups, Italian-and German-speaking inhabitants, are not distributed equally throughout the five cities where data collection took place (section 3). In Bozen/Bolzano, one of the five cities where data collection took place, the largest linguistic proportion belongs to the Italian-speaking group ( (ASTAT 2019: 16-18). This inhomogeneity of the territorial distribution of the language groups in cities, towns, and valleys might suggest that South Tyrolean inhabitants have widely differing language skills due to, for example, language contact and/or varying amounts of exposure to the other language.
The majority of the Italian-speaking group is characterized by external multilingualism, whereas the German-speaking group is characterized by both external as well as internal multilingualism (Lanthaler 2012a: 140-143, 149-151;Wandruszka 1979). External multilingualism refers to the knowledge of unrelated varieties -in the case of South Tyrol this refers to a speaker who has knowledge of German and Italian. Internal multilingualism, on the other hand, refers to the control and/or usage of related varieties -in the present linguistic context this refers to a speaker with knowledge of a regional variety of German (i. e. Tyrolean dialect 4 ), which belongs to the Southern Bavarian dialect group, henceforth Bavarian variety, and a German standard variety (Abfalterer 2007;Ammon/Bickel/Lenz 2016;Riehl 1999). The German-speaking community normally uses the Bavarian variety, a non-standard variety, in intimate domains (e. g., with family, friends, in the neighbourhood), in everyday communication, and at one's place of employment. The standard variety (or written dominant norm) is used in more formal contexts/official situations, for written communication (e. g., education, politics, media, religion) and with tourists (ASTAT 2015: 142-147, 151f.;Lanthaler 1990;Riehl 2000Riehl , 2007Saxalber Tetter 1982).

Methodology and Research Questions
Before introducing the research questions of the current qualitative study (section 3.4), an overall description of the KOMMA-corpus will be provided.

Corpus Design
The KOMMA-corpus consists of written and spoken data collected in solely German-language high schools in the cities of Bozen/Bolzano, Brixen/Bressanone, Bruneck/Brunico, Meran/Merano, and Schlanders/Silandro (a more detailed corpus description 5 can be found in Glück/Leonardi 2019: 448-451). Subjects were pupils who were in their final year of high school and had German as (one of) their first language(s) (see section 3.3).
In the present analyses, only the spoken corpus will be considered, more specifically, just the biographical interviews (25h and 15min; 260,822 token), which make up approximately 60% of the spoken corpus. All interviews were performed according to a semi-structured guideline and lasted between 13 and 50 minutes (mean duration 24 minutes). Using ELAN 6 (Sloetjes/Wittenburg 2008), the spoken data was transcribed according to German orthography (however, in lowercase letters) and not phonetically. The transcription 7 included pauses, paralinguistic markers (laughter or coughing), fillers but not features such as intonation, word-stress or volume.

Procedure of Data Collection
Data collection took place between 2012 and 2015 during regular school hours in the subjects' high schools. At the onset of each biographical interview, which was conducted in Standard German, the pupil completed a brief questionnaire providing socio-demographic as well as sociolinguistic information, including: (i) year of birth, (ii) first language (L1), (iii) mother's L1, (iv) father's L1, (v) languages spoken by the subject, and (vi) the percentage of time one uses German and/or Italian when interacting with friends. Results will be presented in the following section. The topics addressed in the biographical interviews 8 ranged from family background information, to language learning experiences, self-evaluation of language skills, language use, and attitudes towards different languages/varieties. In the present contribution we will focus on pupil's language use and their self-reported language skills.

Subjects
In total, 65 pupils -41 female (63%) and 24 male (37%) -with an age range of 19 to 22 years (Mage=19.3 years) were interviewed. At the time of data collection, 65% attended a Gymnasium (an upper secondary school that provides general education), 29% a Fachoberschule (an upper secondary school that provides technical education), and 6% a Waldorf School, experiencing a Waldorf oriented education. Out of 65 subjects, 20% were interviewed in Bozen/Bolzano as well as in Brixen/Bressanone, 9% in Bruneck/Brunico, 26% in Meran/Merano, and 25% in Schlanders/Silandro. Data emerging from the questionnaire revealed that 61 subjects (94%) named German 9 as their first language. Two (3%) subjects declared to have Italian as their L1, and two more subjects (3%) claimed to have two languages as their L1: German/Italian, and German/Spanish. Table 1 provides a summary. The parents' L1 10 was collected exclusively from the pupils themselves. According to them, the majority of their mothers had German as L1 (n=59; 90.8%), followed by Italian (n=4; 6.2%), Ladin (n=1; 1.5%), and Spanish (n=1; 1.5%). Similarly, the majority of subjects' fathers had German as L1 (n=59; 90.8%), followed by Italian (n=4; 6.2%). Two subjects (3%) stated that their fathers had two first languages, namely both German and Italian (see Table 1).
8 For a more detailed description of the questions asked during the interviews, cf. the KOMMA website (Free University of Bozen-Bolzano). 9 In the questionnaire pupils did not differentiate between the different German varieties (i. e., Tyrolean, Austrian Standard German, Swiss German) when using the term "German". During the interviews, then, they used the following expressions: Südtiroler Dialekt, Bayerischer Dialekt, Österreichischer Dialekt, Saarländischer Dialekt, Standarddeutsch or Hochdeutsch. 10 The questionnaire contained the singular form "language" ("Your first language", "Mother's first language", "Father's first language"). Within the biographical interviews, the interviewer was able to ask more detailed questions concerning subjects' home language(s) revealing that 13 pupils grew up with parents who had either different first languages or with a parent who had grown up herself/himself bilingually (cf. Leonardi 2020). In addition to the Bavarian variety, the standard variety of German and Italian, all subjects claimed to know (it was not further specified what is meant by "knowing") English. Depending on the particular school and what languages were taught, but also due to the pupils' private circumstances or interests 11 , languages such as French, Spanish, Latin, Greek, Ladin, Polish and Swiss German were listed on the questionnaire and/or named during the biographical interviews.

First Language(s)
According to the questionnaire, 56 (86.2%) out of 65 pupils claimed to use German 70% or more of the time while interacting with their friends. As shown in Graph 1, only nine pupils (13.8%) stated that they use both Italian and German in this domain on a regular basis. "On a regular basis" here means that the pupil claimed to use both German and Italian almost equally: German is used between 45% and 60% of the time, while Italian is used between 40% and 55% of the time. This, however, is not that surprising if we consider that: (i) data collection took place in German-language high schools only; and (ii) the language groups are not distributed equally throughout South Tyrol, meaning that the opportunity to use and hear Italian is very limited in certain valleys, towns, and villages due to the absence of Italian native speakers. The geographic distribution of the language groups -with a concentration of the Italian-speaking population in cities and the majority of the German-speaking population living in rural communities -has a 11 For instance, some pupils explained that they learned Spanish through private lessons outside of school.

Research Questions
The present paper addresses the following research questions, which will be analysed and discussed separately: 1. How do young people in South Tyrol describe their use of language(s) in different domains, focusing on family members, friends and the school context? 2. How do they rate their language skills in Standard German, Italian, and English? Which factors influence subjects' linguistic abilities while living in an officially bilingual/trilingual area?
The first research question focuses language use in certain domains and will be examined in section 4.1. The second concerns the relationship between contact and language competences, which will be analysed in section 4.2. To answer these research questions, we conducted interviews with fifth-grade high school pupils. As argued by Pavlenko, biographical interviews or linguistic biographies, also known as Sprachbiographien 12 , provide important contributions to the research on bi-and multilingualism as well as on second language acquisition, since "they offer insights into people's private worlds, inaccessible to experimental methodologies" (Pavlenko 2007: 164). Such autobiographical narratives "are a subtype of personal experience narratives" (Pavlenko 2008: 319) and are "a unique means of gaining first person insights into the processes of language learning, attrition, and shift" (Pavlenko 2008: 321), since biographical interviews deal with the acquisition of certain varieties or languages in relation to subjects' life histories. Biographical interviews provide several advantages by offering the possibility to analyze and investigate experiences from the subject's perspective. Such interviews can be very detailed and can capture, for instance, a speaker's complex linguistic construction within the home language use, the functionally adequate use of various varieties in everyday life, or decisions for or against using a particular variety. Moreover, via biographical interviews, one can get more insights into the language use of speakers who use multiple varieties but who are not perfectly balanced bilinguals.

Results
I shall now turn to the main body of the study: the qualitative findings gained from the biographical interviews. Language use in different domains will be answered in section 4.1, revealing whether the present findings are consistent with results obtained in previous studies (e. g., ASTAT 2015; Riehl 2007; see section 2). Section 4.2 goes on to present qualitative results gained from the pupils themselves regarding their language competences 13 in Standard German, Italian and English as well as providing possible explanations for their answers.

Language Use
As described in section 2, there is a clear-cut distinction between private/informal domains and official/formal domains in the South Tyrolean context, i. e. informal domains are reserved for the Bavarian variety, while more formal domains favour the use of the standard variety of German. As shown in (1), both the standard variety and the non-standard variety are often strictly correlated to certain places, conditions, or people. For speaker B09, Standard German is a language used in Germany, at school, or in interaction with tourists, while the Bavarian variety is used with family members, friends, and in her spare time. In (2), speaker B39 claims to use Standard German at school and she switches back to the Bavarian variety when communicating with his friends.
( Moreover, it is not just the physical environment of the school that is associated with the use of Standard German but also communication with teachers. As explained by speaker B55 (example (3)), she never knows how to behave when meeting a teacher outside school hours, even if they do not teach her anymore.
(3) B55: das ist immer ganz komisch wenn man einen lehrer trifft weil ich weiß dann nie ((laughs)) wie ich wie ich mich verhalten soll (-) weil auch wenn ich (-) lehrer treffe die ich schon gar nicht mehr habe (-) aber ich tendiere dazu auf hochdeutsch zu sprechen (B55: 'It is always quite funny when you meet a teacher. I never know how I should behave because even if I meet teachers I do not have any more, I tend to speak Standard German.') [Schlanders/Silandro, female] The fact that Standard German is the preferred language spoken at school is rarely questioned among our subjects; it is taken for granted, therefore it is necessary -according to pupils' opinions -to practice and learn Standard German at school since it is not actively spoken at home or in everyday life. Moreover, it is often argued by our subjects that Standard German is an official language, which enjoys status and prestige both within and outside South Tyrol and that it is therefore quite important to have an adequate knowledge of this language. In (4) and (5), speaker B33 and speaker B29 explain it the following ways: (4) B33: mein dialekt ist jetzt keine offizielle sprache und wir sind hier an der schule um (-) die form einer sprache zu lernen und das hochdeutsche und äh (-) wenn wir schon sonst nicht so viel möglichkeiten dazu haben (-) dann ist es schon gut find ich schon gut wenn wir in der schule deu/ äh hochdeutsch sprechen (B33: 'My dialect is not an official language and we are here at school to learn the form of a language, Standard German, because otherwise we do not have opportunities to use it. Then, in my opinion, it is good if we speak Standard German at school.') [Meran/Merano, female] (5) B29: jeder hat einen anderen dialekt und den spricht man ja in der familie und mit freunden und so weiter (-) die hochsprache ist halt trotzdem eine offizielle sprache und wenn man nach deutschland geht dann (.) kann man sich halt verständigen und ich glaube unseren dialekt versteht nicht jeder (B29: 'Everyone has a different dialect 14 and you speak it within the family and with friends and so on. The standard language is still an official language and when you go to Germany you are able to communicate, and I think our dialect is not understood by everyone.') [Bruneck/Brunico, female] Several aspects can be derived from the two statements provided in (4) and (5): language status, one's relationship with the standard language, and intelligibility. First of all, the Bavarian variety and Standard German have different status among their speakers. Standard German is "an official language" while the Bavarian variety, as a non-standard variety, has no official status. Secondly, in (4) speaker B33 explains that South Tyroleans do not have "many opportunities" to use Standard German (see also section 4.2), therefore -according to her -it is "good if we speak Standard German at school". Thirdly, according to speaker B29 in (5), not everyone understands the Bavarian variety, especially not those who live outside of South Tyrol. Therefore, when using Standard German, you are not only using an official language, but you make yourself understandable to those who are not from the area, such as people from Germany.
These results confirm what has already been shown in previous studies (see section 2): there is still a clear-cut distinction between informal domains and formal domains, even among young South Tyroleans.

Language Proficiency
During the biographical interviews subjects were asked to rate their different language skills. The data reported below is exclusively qualitative. The following statements do not only provide information on subjects' perceived language skills in Standard German, Italian, and English, but also potential reasons for their "good" or "bad" language proficiencies. This is an advantage which we gain from using qualitative data.

(Lack of) Practice
Most participants claim that they do not have any issues or problems with the standard variety of German. Nonetheless, as we have already seen in section 2 and section 4.1, the standard variety is rarely spoken in everyday life (Abel 2007). In (6), speaker B28 considers it important to transmit the standard variety of German within the school context, since she claims that "one has to practice" in order to learn the language, and one cannot learn "just through television, movies, and books". In (7), speaker B38 explains that she "hardly ever practice[s] the real Standard German" and that she does not "speak Standard German so often". Moreover, she further explains that in interaction with teachers her "dialect always comes out again", meaning that her non-standard variety interferes when speaking the standard variety of German.

Place of Residence
Similarly, a factor influencing the pupils' language skills in Italian is the rare opportunity to use the language due to the asymmetric distribution of the language groups within South Tyrol, as explained in (8). Speaker B50 points out that people living in the valleys often "speak bad Italian" because they do not have the "opportunity to use the language" actively. In (9), speaker B61, who lives in the Western part of the province, the Vinschgau valley, which is predominantly Bavarian speaking, highlights this fact by explaining that "we speak far too little Italian" and therefore she has "problems with the Italian language". Moreover, she also struggles with the fact that she has to justify herself to strangers: she is from Italy but does not speak Italian properly.
(8) B50: in den tälern also in der peripherie (-) sprechen sehr viele leute ein schlechtes italienisch (.) weil eben auch nicht die möglichkeit da ist weil man wirklich nur unter deutschen lebt und nur s/ sehr vereinzelt italiener auch (.) hier leben (.) und dann ist es natürlich schwierig ein gutes verhältnis zu dieser sprache zu pflegen (B50: 'In the valleys many people speak bad Italian because there is no opportunity to use the language. You live among German-speaking inhabitants and only very rarely there are Italian-speaking inhabitants too. Then, of course, it is difficult to maintain a good relationship to this language.') [Schlanders/Silandro, female] (9) B61: tja mit dem italienischen hab ich so meine probleme ((laughs)) (-) äh ja ähm wir sprechen zu wenig also hier im vinschgau wir sprechen viel zu wenig italienisch (-) und wenn man dann irgendwo hinkommt und jemand fragt ja von wo bist du ja von italien ja und du kannst fast kein italienisch (B61: 'I have problems with Italian. We speak too little here in the Vinschgau valley, we speak far too little Italian. If you go somewhere and someone asks you: Where are you from? Yes, from Italy. And you speak almost no Italian.') [Schlanders/Silandro, female]

Usefulness/Uselessness of a Language
As argued by speaker B41 in (10), it is important to be able to speak Standard German; otherwise, it is going to be difficult, "even if we live in a country where you can speak dialect".
(10) B41: als (-) oberschulabgänger wenn man die hochdeutsche sprache nicht sprechen kann wird man sich schwer tun (-) auch wenn wir in einem ( ) land sin/ äh leben in dem man dialekt sprechen kann (B41: 'As a high school graduate you will have difficulties if you are not able to speak Standard German, even if we live in a country where you can speak dialect.') [Meran/Merano, male] Although Italian is considered an important language for social and professional life in South Tyrol (Abel 2007: 7;Baur 2009: 22), as also shown in (11) and (13), English is sometimes perceived as even more important than Italian, as demonstrated in (12). For speaker B65 (example (12)), English is "much more useful than Italian", which he claims not needing on a daily basis.
(11) B21: ich finde es ist sehr wichtig mehrere sprachen zu können (-) englisch ähm generell also um sich ähm überall auf der welt ähm mitteilen zu können (-) und italienisch besonders hier (B21: 'I think it is very important to be able to speak several languages. English in general, so you can communicate anywhere in the world and Italian especially here.') [Brixen/Bressanone, male] (12) B65: englisch hingegen hatte ich meistens eher gern also i/ ich wollte die sprache englisch gerne lernen (-) ich fand sie sehr viel nützlicher als italienisch ((laughs)) weil wie gesagt italienisch brauchte ich im alltag nie (-) und it/ und englisch kam ich immer mehr fand ich immer öfter auch im internet und computerspielen (B65: 'English, on the other hand, I wanted to learn English. I found it much more useful than Italian because, as I said, I do not need Italian on a daily basis and I encounter English more often on the internet and in computer games.') [Schlanders/Silandro, male] The South Tyrolean bilingual exam 15 (German: Zweisprachigkeitsprüfung, Italian: Patentino), which examines language proficiency in the standard variety of German and Italian, is the official exam one must pass in order to be able to work in public services. Therefore, an adequate knowledge of both languages is necessary, as explained in (13) by speaker B53.
(13) B53: es gibt prüfungen (.) wo man dann diese zwei-und dreisprachigkeit ja zertifiziert (-) diese werden an fast allen arbeitsstellen (-) eben in südtirol (.) verlangt und ohne die hat man eigentlich (.) weniger chancen eine arbeit zu haben also es werden immer die bevorzugt die dieses zertifikat besitzen (B53: 'There are [language] exams where one certifies this bi-or trilingualism [Italian/German or Italian/German/Ladin]. This certificate is required for almost every job in South Tyrol and without it, one has actually fewer chances to get a job. Those who possess it will always be the preferred ones.') [Schlanders/Silandro, male] In reality, it is not just that "those who possess it will always be the preferred ones", but if a job requires such certificate, a person without one is not able to get the job.

Second Language Teaching
The physical environment of the school itself is important for teaching a language as well as creating positive attitudes and opinions towards that language/community since "the role of the language teacher is not just to teach the language and impart knowledge, but also to foster positive attitudes towards different languages and varieties through the right atmosphere" (Lasagabaster 2005: 403). Example (14) demonstrates how important the school environment is to the language learning process, which in this case can be attributed to those who teach the language. Speaker B29 explains that she had several different Italian teachers throughout her primary school years and therefore she claims that she was not able to "learn the language properly".
(14) B29: in der grundschule wir hatten (-) ich glaub ich hatte (-) zwölf dreizehn italienischlehrer die haben (-) gewechselt ich in einem jahr hatte ich sechs (-) und dann logisch lerne ich die sprache nicht richtig (B29: 'In primary school we had, I think I had twelve or thirteen Italian [language] teachers. I had six in one year and then, of course, I do not learn the language properly.') [Bruneck/Brunico, female] It can be summarized that the following factors have an impact on speakers' language proficiency: place of residence, the ability of practicing the other language, subjects' L2 learning experiences as well as their opinion towards the general usefulness of a language.

Discussion and Conclusion
The present empirical study had two interrelated aims, which are outlined below.
15 Some Italian certificates (e. g., PLIDA, CIC) and German certificates (e. g., Goethe, TELC) are also recognized as a valid alternative to the South Tyrolean bilingual exam.
The first aim was to examine whether -over the past decades -there has been a change in pupils' language use within certain domains (family, friends, and school). Domains of language usage are defined as abstract constructs, which determine the choice of a language/variety in a multilingual language community (Werlen 2004: 335). According to Fishman's (1964Fishman's ( , 2000Fishman's ( [1965) domain analysis, in certain situations one language might be more appropriate than another language. Therefore, Fishman presumes that language choice is not influenced by language competence, but by the situational appropriateness (Werlen 2004: 336). The object of this investigation was the German-speaking community in South Tyrol, a region characterized by internal and external multilingualism, whose multilingual speakers have competences in a Bavarian variety, in a German and an Italian standard variety. As we have seen in section 4.1, the uses of German varieties in South Tyrol complement each other: the Bavarian variety is reserved for the domains family, friends, and leisure time, while the German standard variety is predominantly used at school, with teachers at school, and tourists from Germany. The present qualitative results demonstrate that the Bavarian variety remains a very vivid variety among its speakers.
The second aim, examined in section 4.2, investigated the pupils' self-reported language skills in Standard German, Italian, and English. Interestingly, the social life between the German and Italian language group still appears to be fairly segregated and individual bilingualism does not yet seem to be the reality in South Tyrol (e. g., Forer et al. 2008;Forer/Paladino/Wright 2012). Moreover, we also investigated which extra-linguistic factors influence the pupils' language skills. Both, L2 learning motivation and positive attitudes towards the L2 are key factors for successful language learning, since positive attitudes foster motivation to learn the L2 and consequently influence speaker's L2 learning achievement (e. g., Gardner/Lambert 1972). As demonstrated by speaker B53 in example (13), South Tyrolean pupils are aware about the importance and benefit of the Zweisprachigkeitsprüfung. In such a case, the L2 acquisition is defined as instrumental bilingualism (Mackey 2005(Mackey : 1486, since the other language is learned for purely utilitarian motives (better career advancement or higher income). However, according to Vettori/Wisniewski/Abel (2012: 452), who also investigated South Tyrolean pupils' motivation and orientation towards the L2, "instrumental orientation does not correlate with L2 skills", and consequently "does not contribute to improving the pupils' competences at all". This is in line with previous research, demonstrating that integrative motivation is linguistically more successful than instrumental motivation (e. g., Gardner/Lambert 1972).
Moreover, the asymmetric distribution of the language groups within the territory also influences speakers' L2 language proficiency, since some areas are nearly monolingual. A survey conducted by ASTAT revealed that in an (almost) exclusively German-speaking environment, 50.2% of the respondents claimed to "understand everything" in Italian, while this number increases drastically (84.2%) when people live in an area with more than 15% Italian-speaking inhabitants (ASTAT 2015: 125-136). Besides these results, which are based on self-evaluation methods, a more recent study, which examined South Tyrolean pupils' L2 competences (German or Italian) according to the CEFR (Common European Framework of Reference) levels, confirmed these results too ; Forer/Paladino/Wright 2012; Vettori/Wisniewski/Abel 2012). Therefore, in line with previous investigations, the current qualitative results demonstrate that the place of residence has an impact on peoples' passive and/or active secondlanguage knowledge.
As mentioned above, favourable attitudes towards the L2 community, which accounts for increasing motivation to learn and practice the L2, are not only influenced by contact experiences (direct and indirect contact patterns, cf. Forer/Paladino/Wright 2012) but also by pupils' learning experiences. Mioni (1990: 31) reported that the school context is important not only for teaching a language, but also for creating attitudes and opinions. In other words, positive learning experiences in the L2/foreign language and its culture can have a positive impact on pupils' attitudes and their language learning success. This has recently been proven by ASTAT's 2015 report, demonstrating that positive learning experiences in subjects' L2 have had a positive impact on their self-reported language proficiency in German or Italian respectively (ASTAT 2015: 59). Nonetheless, in the same report it was also shown that -based on subjective evaluations -only 19.6% of the German-speaking subjects were satisfied with the ability to express themselves adequately in the L2 provided by L2 teaching at school, 50.3% were partly satisfied, and 25.4% were not satisfied at all (4.6% claimed to have had no experience) (ASTAT 2015: 57-58). These numbers show that an important foundation for advanced L2 knowledge is already laid during school years, therefore also underlining the important role of a (language) teacher.
The current investigation is not based on objective criteria, but on respondents' self-assessments of their proficiency or lack of proficiency in a given language. Italian and German are both official -and therefore also important -languages in South Tyrol. By extracting individual statements from subjects, we have tried to demonstrate that pupils show different language abilities due to different speaker-internal factors as well as extra-linguistic factors. These factors are motivation to learn and attitudes towards a certain language or community, place of residence, learning environment, and learning/teaching strategies. The domain school seems to be important for subjects' competences in the standard variety of German, since it provides the rare opportunity to practice the language. In addition to the learning experiences at school, the opportunity to meet and interact with people outside of school is very important for learning Italian, which ultimately contributes to active bilingualism. However, due to the asymmetric distribution of the language groups and the separate education systems, there is often far too little contact between individuals belonging to the German-speaking and the Italian-speaking community, which, consequently, impacts speakers' language proficiency.
The present qualitative analysis of pupils' biographical interviews has produced a complex picture, which is based on statements made by individual young adults living in South Tyrol. The advantage of using narratives when collecting data on language use and language skills is the acquisition of more detailed information provided by the subjects themselves. This is a major aspect that can hardly be achieved through quantitative data collection only.