Diese ausgezeichnete Studie untersucht die Entwicklung schriftsprachlicher Kompetenzen mehrsprachiger Schülerinnen im syntaktischen Bereich im Türkischen und Deutschen. Hierzu werden bilinguale und monolinguale Beschulung anhand zweier Probandengruppen kontrastiert. Die Studie gehört zu den wenigen Arbeiten, die den Schriftspracherwerb mehrsprachiger Schülerinnen unter Berücksichtigung beider Sprachen sowie der erstsprachlichen Instruktion untersuchen. Die hier untersuchten Gymnasiasten setzen unabhängig von der Sprache und dem erstsprachlichen Input mit zunehmendem Alter syntaktische Subordinationen statt Koordinationen ein, dabei werden die Subordinationen zunehmend komplexer. Die Förderung der Erstsprache im bilingualen Schul-Kontext beeinflußt die Wahl und Verwendung der Erstsprache positiv, ohne dass die Zweitsprache dadurch in den Hintergrund rückt. Sprachliche Normabweichungen bedeuten hier keinen auffälligen Spracherwerbsprozess sondern sind Bestandteil dieses Lernprozesses.

Jens Walter, lehrerbibliothek.de

Summary
The PISA (Programme for International Student Assessment) study ranks coun-
tries according to the academic performance of the students in those countries.
In 2001, the PISA study had already identified a close connection between social
background and deficits in equal opportunities in Germany: Children from so-
cially disadvantaged families were significantly below the results of their better-off
classmates (Stanat et al. 2002). The German Federal Ministry of Education and
Research (BMBF) reports that after the so-called ‘PISA shock’ in 2001, Germany
was one of the few countries that had continuously improved its ranking (BMBF
2019). However, in 2019 the BMBF also states that since PISA 2012 the results have
either stagnated or declined. Although the influence of social background on edu-
cational achievement has decreased in Germany, it remains high in comparison to
other countries (BMBF 2019).
Given the importance of written language skills and multilingualism in the
school and social context, the present study examines two overarching questions
(a) at what level of complexity do Turkish-German bilingual students at secondary
levels I and II write and which type of compound and complex sentences do they
use when writing argumentative texts in their two languages; and (b) how much
does the first-language input in the school context influence the written language
development of students in their two languages (for research questions, see 2.1). To
answer these questions, written texts from Turkish-German bilingual students in
the 7th, 10th, and 12th grades in their first language Turkish and in their second
language German are studied, using texts from bilingual students from two differ-
ent schools. The first group of pupils received Turkish as their first language and
as a language of instruction (pupils with TU1), the second group of pupils received
Turkish as their second language and as a foreign language (pupils with TU7). The
data is taken from the MULTILIT project of the University of Potsdam.
The study is divided into a theoretical and an empirical part. In chapter 1.1, it
is shown that multilingualism is not a rigid term that describes a single phenome-
non. Rather, one can speak of the different types of multilingualism, e.g. individual
and societal multilingualism, which can in turn be divided into subtypes depend-
ing on the focus of the descriptive approach.
Chapter 1.2 addresses multilingualism in the migration context in order to
better understand and locate the multilingual everyday life and language use of
the pupils examined here with regards to the (educational) political framework.
First, a general report is given on “Europe and its multilingualism” (1.2.1). In this
sub-chapter it is shown that although the European Union values its linguistic
diversity in written documents, in practice, with regard to multilingualism and
the equal treatment of the languages of Europe, it mainly takes into account the
10 Summary
officially recognized languages of the European Union (see the Commission’s core
line on multilingualism, European Commission 2006: 1; European Commission
2012: 138). Subsequently, the offer of Turkish as a family language in the German
school system (1.2.2) is outlined, including the recruitment procedures and crite-
ria for teaching personnel in Germany. According to this, Turkish is rarely taught
as a regular foreign language in Germany, but more often as a heritage language.
Chapter 1.2.3 on the Turkish language in a migration context demonstrates that the
acquisition of Turkish in a multilingual migration context differs from monolin-
gual Turkish acquisition in Turkey.
Chapter 1.3 presents the historical development of the research area German as
a Second Language. It is shown that the main focus of most studies in the 1970s was
the language learning problems of “guest workers”, their reduced use of German,
the pidgin German (“Gastarbeiterdeutsch”). In the 1980s, the focus of research
shifted to the language behavior of Germans directed at migrant workers (For-
eigner Talk). Partly as a result of international comparative studies such as PISA,
IGLU and TIMSS, multilingualism and second language acquisition in the context
of school success as well as the educational opportunities of the “migrant children”
(in the study “multilingual children”) of the second or third generation moved
more and more into focus. According to the results of the international compar-
ative studies, the socioeconomic status of the students was and is a determining
factor with regard to equal education opportunities. Following the publication of
these results, further studies have been and are being carried out to investigate,
among other things, the second language acquisition of multilingual pupils (1.3.1).
There are more and more studies that also consider the development of the first
language and its influence on the second language acquisition of children and ad-
olescents growing up in Germany. A selection of these studies on the first and sec-
ond language acquisition of multilingual pupils was discussed in Chapter 1.3.2. In
addition, a number of studies were conducted to investigate the changes to which
first languages or family languages are subject in the migration context as a result
of language contact situations (1.3.3).
Chapter 1.4 discusses the different approaches to explain language acquisition.
After an overview of the most influential explanatory models (1.4), the language
acquisition theories relevant for this study, namely the interdependence hypothesis
(Cummins 1976) (1.4.1.1), the Cummins’ theoretical framework, BICS-CALP dif-
ferentiation (1.4.1.2) (1980; Cummins/Swain 1986), and the Interlanguage Hypoth-
esis (1.4.2) are described in more detail.
Chapter 1.5 presents the theoretical framework, which is divided into three
sections: First, the framework conditions are outlined by providing insight into
the schools’ (Turkish) instruction and curricula (1.5.1); then, chapter 1.5.2 discusses
the theoretical foundation, beginning with the parallels between the orate-literate
distinction of Maas (2010) and the BICS-CALP differentiation of Cummins (1980;
11Summary
Cummins/Swain 1986). It is then argued why the concept of language enhance-
ment (“Sprachausbau”) of Maas (2010) was chosen as the theoretical framework
for the present empirical study. Furthermore, the notion of syntactic complexity is
explained, with a description of earlier work on syntactic complexity that supports
Maas’ idea of language enhancement. The second part of the theoretical frame-
work focuses on Maas’ concept of language enhancement (1.5.2.1), which was the
basis for the linguistic analysis of the present study. In the presentation of the con-
cept of language enhancement, differences between Maas’ model on the degree of
syntactic complexity and its adaptation in the present analysis are mentioned. The
third part of the theoretical framework is the overview of complex sentences in
Turkish and German (1.5.3), to facilitate the understanding of the data analysis in
the empirical part (2).
In the part 2, the aim is not only to quantify the linguistic structures used by the
subjects (for subject profile, see 2.3), but also to create an overview of the frequency
and type of sentences used, in order to then be able to make a qualitative analysis
(2.5.2). In Chapter 2.4 – after the data elicitation (2.4.2) and the presentation of the
data corpus (2.4.3) – a complexity scale for both languages was designed for data
analysis based on Maas’ model of language enhancement while taking into account
the typological differences between Turkish and German (see 2.4.4.1 “Evaluation
procedure”). Using this tool, in the quantitative analysis first the text length results
(2.5.1.1) and then the results regarding compound and complex sentences in Turk-
ish texts (2.5.1.2) and in German texts (2.5.1.3) are described with regards to their
occurrence and frequency. In addition to frequently and rarely used compound
and complex sentences, the analysis also deals with those complex sentences that
appeared difficult for the students at the time of the survey (2.5.1.2 and 2.5.1.3). In
2.5.1.4 the morphological and syntactic errors as well as norm deviations in Turkish
and German texts are identified. For this purpose of qualitative analysis (2.5.2), the
metadata are presented in the form of language biographies (2.5.2.1). In chapter
2.5.2.4 possible correlations between the students’ language biographies and their
language use are discussed and reference is made to the working hypotheses (2.2)
at appropriate points.
The results of this study (2.5.2.2.4 and 2.5.2.3.4) show that both the Turk-
ish-German bilingually and German monolingually schooled groups of bilingual
pupils use complex sentences more frequently with increasing age in both languag-
es, similar to monolingual language acquisition. However, the two groups differ
from each other in that the bilingually schooled group uses simple and compound
sentences less frequently over time than the monolingually schooled group, and
instead uses complex sentences more often. Furthermore, the bilingually schooled
group uses participle constructions earlier than the monolingually schooled group.
Additionally, the bilingually schooled group also shows fewer errors and norm de-
viations in morphological and syntactic areas, especially in Turkish, but also in
12 Summary
German. The investigation of possible correlations between the language biog-
raphies of the students and their language use (2.5.2.4) shows that in contrast to
the monolingually schooled group that prefers German, the bilingually schooled
group (a) chooses both German and Turkish for reading newspapers and maga-
zines; (b) prefers both German and Turkish for writing; and (c) feels almost equally
familiar with both German and Turkish in spoken and written contexts. These
results, in turn, indicate that family language instruction provides expanded ac-
cess to literary structures through media in both languages, without neglecting the
second language in terms of language choice for both media and writing contexts.
The comparison of the two groups also demonstrates that written language acqui-
sition requires school instruction of that language to be integrated into the official
curriculum (see also Herkenrath 2012).
In the concluding section 3 (“Summary of the study and outlook”) – further
research perspectives (3.3) – reference is made to the didactic benefits of linguistic
analyses, such as those conducted in this study, in supplementing or modifying
curricula (3.4). Furthermore, the present study argues for a different approach re-
garding norm deviations and multilingualism in the classroom and – following
Diehl et al. (2000) and Sieber/Sitta (1994) – proposes a change of perspective from
the deficit-oriented perspective to the development perspective (3.4). This propos-
al as well as this study is addressed to all stakeholders in educational policy and the
education sector, such as schools, teachers, parents, but also universities, research
institutes and researchers, because social inequality and educational barriers are
also reproduced by and in higher education and research. As mentioned in the
opening paragraph, despite the improvement since the ‘PISA shock’ in 2001, the
results for Germany have stagnated since PISA 2012 and the influence of social
background on educational achievement is still high compared to other countries.
In view of this fact, it can be said that the starting point for the present study (2014)
is similar in 2020.
Keywords
Linguistics, multilingualism, multilingual education, written language acquisition
in L1 and L2, academic register, corpus linguistics, language typology, educational
inequality, teacher training, language policy.
Sprachwissenschaft, Mehrsprachigkeit, mehrsprachige Erziehung, Schriftspracher-
werb in L1 und L2, formelles Register, Korpuslinguistik, Sprachtypologie, Bildungs-
benachteiligung, Lehrerbildung, Sprachenpolitik.

Inhaltsverzeichnis