Esther Berendse - Utrecht University Repository

advertisement
Acquiring L2 English Prepositions in an
L1 Dutch Environment:
The Effect of Immersion Through CLIL
Teaching
Master’s Thesis Education and Communication, Utrecht University
Esther Berendse
3467740
Supervisor: Dr. M.C.J. Keijzer
June 2013
2
Contents
Chapter 1 – Introduction
3
Chapter 2 – Theoretical Framework
2.1 Historical Overview of Language Teaching Methodologies
2.2 CLIL in Europe
2.3 Language Teaching in the Netherlands
2.3.1 CLIL in the Netherlands
2.3.2 CLIL Standards
2.3.3 Research on CLIL
2.4 Implicit and Explicit Knowledge
2.5 The Role of Language Input
2.6 Implications for Practice
6
7
9
10
11
12
13
16
17
18
Chapter 3 – Research Questions and Predictions
20
Chapter 4 – A Comparison Between English and Dutch Prepositions
21
Chapter 5 – Method
24
5.1 Participants
5.2 Materials
5.3 Procedure
5.4 Data Analysis
25
27
28
Chapter 6 – Results
6.1 Demographic Background Information of the Students
6.2 GJT Results
6.2.1 Instances of Insecurity
6.2.2 Disregarding Prepositions
6.2.3 Students Unable to Correct Incorrect Trials
6.3 Elicited Production Task
31
Chapter 7 – Discussion
41
Chapter 8 – Conclusion
52
References
56
Appendices
59
34
35
35
37
38
3
Chapter 1 – Introduction
The past two decades have seen the development of Content and Language Integrated
Learning (CLIL) in secondary schools in The Netherlands as well as elsewhere in Europe. In
CLIL, the second language is used as a medium through which a content-based subject is
offered. Pupils thus acquire the second language not only within the English language
classroom but also through studying content-based subjects. CLIL can be said to be a
functional approach to second language teaching, as meaningful and authentic communication
is believed to be essential to the second language acquisition process, according to Second
Language Acquisition theory (Richard and Rodgers 161). It is believed that “using the
language for content learning will enhance implicit learning of that language, leading to
higher levels of proficiency than can normally be achieved” (Admiraal, Westhoff, and de Bot
76). This implicit knowledge, Krashen argues, is more important than learned knowledge in
acquiring second language proficiency (15). According to him, “fluency in production is
based on what we have ‘picked up’ through active communication” rather than learned
knowledge (Introduction 2). As such, it could be argued that CLIL teaching facilitates
students’ second language acquisition process.
A large and growing body of studies has investigated the effects of CLIL teaching.
Most studies have examined the success of CLIL as well as the differences between
mainstream education and CLIL teaching. Researchers seem to agree upon the positive effects
of CLIL: CLIL teaching appears to have no negative effects (Admiraal, Westhoff, and de Bot
77; Huibregtse in De Graaff et al. 605) and students acquire a higher level of proficiency than
their mainstream education peers (Huibregtse in De Graaff et al. 605; Verspoor and
Edelenbos 11; Admiraal, Westhoff, and De Bot 86-89); in addition, their language use is more
authentic (Verspoor and Edelenbos 11; Verspoor et al. 47). The research to date, however, has
tended to focus on general language proficiency in writing, speaking, and reading. No data is
4
currently available regarding the immersion effect of CLIL teaching on the acquisition of
(specific) grammatical features by pupils.
The present study, therefore, seeks to investigate the immersion effect of CLIL on
acquiring a specific grammatical feature. Such an approach is interesting since the role of
grammar has gained a prominent position in the second language classroom nowadays, and
people agree that “without a good knowledge of grammar, learners’ language development
will be severely constrained” (Widodo 122). The grammatical feature under investigation in
this study is English prepositions. Since there are hardly any rules as to when to use which
preposition the only way to learn them is to learn collocations by heart or to simply look up
prepositions in a dictionary. The frequency of input may, therefore, play an important role in
acquiring prepositions. In accord with the so-called “critical mass” hypothesis, learners
exposed to more language input are expected to attain a higher language proficiency level
(Marchman and Bates qtd. in Kimbrough and Eilers 286). Since CLIL pupils are exposed to
more English language input from an early age, it could be expected that they acquire English
prepositions quicker and significantly better than their mainstream education peers, as they
become more entrenched. This thesis attempts to examine the immersion effect of CLIL on
acquiring English prepositions. It aims to first of all explore why CLIL teaching is believed to
be effective. It will also discuss what prior studies have found as to differences between the
quality and quantity of input of mainstream language teaching and CLIL teaching. In
addition, this present study will attempt to answer the following research question through a
data-driven approach: Is CLIL education more effective than mainstream education in
acquiring English prepositions in the context of Dutch secondary school students learning
English?
This thesis is organised in the following way. It first sketches the theoretical
framework for the study and will then go on to review the linguistic properties of Dutch and
5
English when it comes to prepositions. It will then discuss the methodology, results, and
discussion of this study. Finally, the conclusion will consolidate the research findings, outline
the limitations of the present study, and propose future directions.
6
Chapter 2 – Theoretical Framework
Since it is argued that there are more multilingual people than monolingual people in the
world today (Saville-Troike 8), it is fair to say that foreign language learning is an important
matter. Throughout history, numerous language teaching methods and approaches have been
developed, attempting to bring about improvements in language learning. Professional
organisations, academics, educators, as well as publishers are constantly looking for “the
‘best’ method of teaching a language” (Richards and Rodgers 15). Changes in language
teaching methodology reflect “recognition of changes in the kind of proficiency learners need,
such as a move toward oral proficiency rather than reading comprehension as the goal of
language study” as well as “changes in theories of the nature of language and of language
learning” (Richards and Rodgers 3). Discussions on teaching methodology today demonstrate
“contemporary responses to questions that have been asked often throughout the history of
language teaching” (Richards and Rodgers 3). For example, the question ‘should we teach
grammar?’ has dominated the language teaching field for many years; however, it is no longer
questioned whether we should teach grammar at all, but rather what the most effective way to
teach grammar is (Richards and Renandya 145).
It could be questioned how innovative and effective contemporary teaching
methodologies are. Lange argues that “the latest bandwagon ‘methodologies’ [came] into
prominence without much study or understanding” (qtd. in Richards and Rodgers 15). In
addition, De Bot and Maljers argue that the only true educational innovation in Dutch
secondary schools is CLIL that is integrated in so-called ‘tweetalig onderwijs’ (in De Graaff
and Tuin 132). De Bot and Maljers claim that the language teaching practice in the
Netherlands is still very much focused on outmoded ideologies such as the GrammarTranslation Method and the Audiolingual Method, which has “hindered a true revolution” (in
De Graaff and Tuin 134, my translation).
7
2.1 Historical Overview
This section will adopt a historical approach in briefly reviewing some of the most widely
known language teaching methods and approaches: the Grammar-Translation Method, the
Direct Method, the Audiolingual Method, the Communicative Approach, and, lastly, the
Natural Approach. In comparison with general approaches to language teaching, teaching
methods “tend to have a relatively short shelf life” as they are associated with specific
practices and therefore become outdated relatively quickly (Richards and Rodgers 244).
The Grammar-Translation Method mainly focused on reading and translating and paid
only little attention to speaking or listening skills. This method implied studying grammatical
aspects and translating complex texts. It dominated the European foreign language field from
the 1840s to the 1940s and was originally used to teach Latin and Greek. Opposition to the
Grammar-Translation Method arose in the mid- and late nineteenth century (Richards and
Rodgers 7), which gave rise to the development of new teaching methods such as the Direct
Method. Unlike the Grammar-Translation method, this new method integrated use of the
foreign language in the classroom. Students now had to speak and think in the foreign
language. The Direct Method attempted to imitate the conditions under which a first language
is learned: through implicit input and imitation of the target language. Like the Direct
Method, the Audiolingual Method also actively used the foreign language as a means to
teach; however, the latter focused on drills and practice as it viewed language learning as
habit formation. Since this method only focused on drills and disregarded interaction as a
component of the learning process, questions were also raised regarding its usefulness.
Such concerns related to the authenticity of language learning methods laid the basis
for the Communicative Approach. This approach builds on the notion of “language as
communication” (Richards and Rodgers 159) as it is believed that “linguistic theory need[s]
to be seen as part of a more general theory incorporating communication and culture”
8
(Richards and Rodgers 159). In the Communicative Approach, pupils learn a language by
using it as a means to communicate; they thus develop speaking-, writing-, listening-, as well
as speaking skills. Communicative competence is believed to be an important aspect of
communication; however, the learning process also involves practice and error. In order to
achieve communicative competence, authentic and meaningful communication should be the
goal of classroom activities in the Communicative Approach; group activities that involve real
communication are, therefore, stimulated. Instead of learning grammar by means of drills,
such as was common practice in the Audiolingual Method, the Communicative Approach
promotes activities that involve meaningful and authentic language use such as “completing
tasks that are mediated through language or involve negotiation of information and
information sharing” (Richards and Rodgers 165). It could be argued that the language
teaching field has showed some significant progress with the rise of the Communicative
Approach. In one form or another, it is still the predominant approach used in foreign
language classrooms today.
Debates in the second language teaching field nowadays also focus on aspects of the
Natural Approach, which builds on Krashen’s views of language acquisition such as the
Monitor Theory. The approach, developed by Krashen and Terrell, is based on the acquisition
process of first and second languages in informal contexts and attempts to promote implicit
language acquisition. Krashen and Tarrell see the Natural Approach as a model of the
Communicative Approach (Richards and Rodgers 178) and just like the Communicative
Approach, the Natural Approach also sees “communication as a primary function of
language” (Richards and Rodgers 179). The difference is that the Natural Approach spends
almost no time on the theoretical aspect of the language: it does not refer to “grammatical
analysis, grammatical drilling, or a particular theory of grammar” (Richards and Rodgers
178). In addition, it does not focus on explicit correction of errors. Instead of focusing on
9
practice and correctness, this approach focuses on language input and exposure.
Comprehensible and meaningful input is therefore believed to be essential: “acquisition can
take place only when people understand messages in the target language” (Krashen and
Terrell 19).
2.2 CLIL in Europe
Reactions to the Communicative Approach and the Natural Approach inspired the
development of a new foreign language teaching approach: CLIL. Inspired by the success of
Canadian immersion programmes and because of the need for higher level of foreign
language proficiency by pupils, CLIL began to develop in Europe in the 1990s (De Graaff et
al. 603; De Graaff et al. 604). The term CLIL was adopted by the European Network of
Administrators, Researchers and Practitioners (EUROCLIC) in order to define the nature of
the approach more clearly “midst a plethora of related approaches such as content-based
instruction, immersion, bilingual education and so on” (Coyle 2). Although CLIL shares some
characteristics with content-based instruction, immersion programmes, and bilingual
education, it “[places] both language and non-language content on a form of continuum,
without implying preference for one or the other” (Marsh 58); in other words, it sees language
and content learning and teaching “as integral parts of the whole” (Coyle 2). CLIL gives
“language and non-language subject matter a joint curricular role” (Marsh 58). In short, CLIL
uses the foreign language as a tool for learning a content-based subject but, at the same time,
views it as a goal of the learning process itself.
CLIL is in agreement with the Communicative Approach as it stresses the idea that
proficiency and fluency in the foreign language can best be achieved by using it as a
“functional medium of communication and information” rather than “making it the object of
analysis in class” (De Graaff et al. 606). According to De Graaff et al., the most important
10
difference between CLIL and “teaching the subject in the mother tongue” is the fact that the
former involves “additional language learning objectives” and “specific opportunities for
communication and language use” (606).
2.3 Language Teaching in The Netherlands
Traditionally, foreign language teaching in the Netherlands was very limited as it tended to
focus only on written skills. Listening and speaking skills were only considered relevant for
“waiters and sales representatives” (Kwakernaak 29, my translation). Motivation for this was
the desire to give foreign language education the same “intellectual status” that the education
of classical Latin received (Kwakernaak 29, my translation). With the rise of the
Communicative approach in the 1980s, this changed, and oral skills gained in importance in
core goals and exam programmes (Kwakernaak 29). This does not, however, mean that oral
skills have acquired a prominent position within classroom practice.
In informal contexts, a second language is acquired mostly through oral
communication rather than written communication; however, reading and writing remain the
skills most practised in Dutch secondary schools (Kwarkernaak 29; Kwakernaak 30).
Especially in the earlier years, oral skills are very important; foreign language education in
the later years, on the other hand, can zoom in on reading skills as this is important for
students’ higher educational career (Kwakernaak 29). Although the use of the target language
as the language of instruction is a great step towards promoting oral skills, practical
experiences show that the belief that pupils only learn grammar and vocabulary in the
classroom in order to communicate outside the classroom is still very common (De Bot and
Maljers in De Graaff and Tuin 135). In addition, De Bot and Maljers point out that Dutch
second language education still incorporates only little interaction and group work in
11
classroom practices (in De Graaff and Tuin 135). In the past decade, the rise of CLIL,
however, has generated significant educational innovation in the Netherlands.
2.3.1 CLIL in the Netherlands
In the Netherlands, CLIL has become a popular teaching practice and is incorporated into socalled ‘tweetalig onderwijs’ (TTO, bilingual education). TTO originates from 1989 when the
Alberdingk Thijm college in Hilversum decided to offer a more “international form of
education” as a reaction to the need of Dutch parents (De Bot and Maljers in De Graaff and
Tuin 140, my translation). As it was found that CLIL education did not interfere with the
development of Dutch and the level of other subjects taught in English, CLIL began to appeal
to more students in the latter half of the 90s (De Bot and Maljers in De Graaff and Tuin 140141). Now, in early 2013, there are 120 Dutch secondary schools that offer pupils DutchEnglish bilingual education (EP “Tto scholen”).
Bilingual education is not imposed upon Dutch schools for secondary education by the
Dutch ministry of education; instead, schools can choose freely whether or not to offer CLIL
programmes and they are relatively free in their design of CLIL curricula, which are
developed through collaboration between language teachers and teachers of non-language
subjects. However, in order to safeguard the quality of bilingual education, CLIL programmes
have to meet several standards and requirements as to the desired level of English of both
students and teachers, the use of authentic English lesson materials, and the central position of
internationalisation within the school policy.
The goals for pupils attending CLIL teaching programmes in the Netherlands are
defined as follows by the European Platform (“CLIL”):

Development of intercultural knowledge and understanding;

Development of intercultural communication skills;
12

Improvement of both active and passive language skills;

Potential to study materials from different (international) perspectives;

More contact with the target language without having more contact hours;

Diversity within the lesson methods applied in the classroom;

Improved motivation of pupils through extra confirmation of language skills.
However, “[f]rom a Dutch perspective, the main goal of CLIL is functional,” De Graaff et al.
point out; that is to say, “to develop proficiency in a foreign language alongside knowledge of
non-language subject area” (607). Regarding the proficiency level, students attending CLIL
programmes are expected to have a higher proficiency level in English than their mainstream
education peers. More concretely, mainstream VWO1 education students are required to have
a B2/C1 level in English at the end of their sixth year, whereas CLIL students are expected to
have a B2 level at the end of their third year already. At the same time, the overall scores of
CLIL students on their final exams are, furthermore, not allowed to divert negatively from the
national average (EP “Standaard” 2).
2.3.2 CLIL Standards
The goals defined above are realised by the implementation of several standards. Regarding
the quantity of English input CLIL pupils receive, at least fifty per cent of the subjects in the
first three classes of secondary education are taught in English. In the higher years, the target
language of most subjects is Dutch since the national final exams are in Dutch as well.
Guidelines stipulate that in the higher years, out of a total of 4,800 (VWO) and 3,200
(HAVO) hours of study load, 1,150 (VWO) and 850 (HAVO) should be in English (EP
1
Dutch secondary education is offered at different levels: VWO, HAVO, and VMBO, with
VWO being the highest level and VMBO-T being the lowest. At most secondary schools,
CLIL is offered only to VWO, that is, pre-university education students.
13
“Standaard” 3). The position of the Dutch language is, furthermore, equal to that of English.
In Dutch bilingual education, CLIL is integrated in subjects like Geography, History, Physical
education, and Arts. Although schools can select which subjects to teach in English, this has
to be at least one creative or Physical education subject, one scientific subject, and one social
sciences subject.
As to the quality of input, at least two subjects should be taught by English native
speakers. These native speakers play an important role in the acquisition of the form of the
foreign language by Dutch pupils (EP “Standaard” 3). The ERK level of non-native speakers
of English should be at least B2 for all language skills. Pupils are, furthermore, exposed to
authentic English materials and have sufficient access to materials in English at all times (EP
“Standaard” 3; EP “Standaard” 5).
Bilingual education, furthermore, values internationalisation. The international
orientation of CLIL is reflected in the curriculum by special projects, programmes, and
activities such as exchange programmes, international collaboration projects, Englishspeaking excursions, and participation in the European Youth Parliament.
2.3.3 Research on CLIL
A large and growing body of literature has investigated the effects of CLIL. Several
international studies have reported positive results of CLIL regarding foreign language
proficiency while no negative consequences for other subjects have been reported (Admiraal,
Westhoff, and De Bot 77). These results, however, do not necessarily apply to Dutch settings,
as CLIL in the cases analysed in the most international studies “is implemented under
conditions that are different from the Dutch educational context” (Admiraal, Westhoff, and
De Bot 77).
14
Admiraal, Westhoff, and De Bot investigate the effect of English as the language of
instruction on the proficiency levels of Dutch students in English. They investigate aspects
such as receptive word knowledge, reading comprehension, and oral proficiency. Although
students who attend the CLIL programme do not have a significantly higher level in receptive
word knowledge, they do show a significantly higher level in reading comprehension (84-86).
In addition, the CLIL group has a higher oral proficiency level as well as a better
pronunciation (86-89).
Several other studies have reported positive effects of CLIL in the Netherlands on the
proficiency level in English. De Graaff et al. bring in Huibregtse who concludes that Dutch
CLIL students attain a significantly higher proficiency level in English than their mainstream
education peers (605); in addition, she points out that this higher proficiency level is not
accompanied by “any negative effects on … academic proficiency in L1 or on other school
subjects” (605). In their study investigating writing skills, Verspoor and Edelenbos also claim
that students who attend a CLIL programme are faster to reach higher levels of proficiency in
the foreign language than mainstream education students, even when general intelligence and
contact with the foreign language outside school hours were taken into account (11). What is
more, Verspoor and Edelenbos note that the language use of CLIL pupils is more authentic
and fluid as they tend to make more use of so-called chunks (11). Verspoor and Edelenbos
thus conclude that CLIL students must learn a foreign language and its grammar implicitly
(11). The positive learning outcomes of CLIL must, however, not be ascribed to the quantity
and quality of input alone; opportunities to produce the foreign language are essential as well,
Verspoor and Edelenbos point out (12). This is in agreement with Marsh who draws attention
to the fact that CLIL does not imply that the more input one receives the better he or she will
become (“Adding Language,” par. 6). According to Marsh, the “essence” of CLIL is
integration of core variables that are “interwoven into the curriculum, and realised through
15
classroom practice” (par.4). In their large-scale study, Verspoor et al. also compare the
language proficiency levels of CLIL pupils and their mainstream education pupils in their
earlier years; in addition, they include a control group in that they investigate the proficiency
levels of pupils who attend mainstream education at a secondary school that does not offer
CLIL education. Their main conclusions largely corroborate previous findings. First of all,
Verspoor et al. conclude that students of both educational types make many mistakes in the
initial stages of their language acquisition process, a common aspect of the acquisition
process (47). They also report that grammatical errors by CLIL students are no longer visible
at the end of their third year; the only errors they make are of a lexical nature (62). In
agreement with Admiraal, Westhoff, and De Bot, Verspoor et al. reveal that CLIL students
retain a significant lead on their mainstream education peers, especially with regards to their
writing skills (61). However, whereas the former study did not find a significant difference
with regard to vocabulary knowledge, Verspoor et al. conclude that CLIL students have a
significantly larger vocabulary size (61). However, as the proficiency level increases towards
the end of the lower years, Verspoor et al. show that development begins to level out for both
educational types (61). On the other hand, CLIL students are found to retain their lead that is
impossible to overtake by mainstream education pupils (Verspoor et al. 61). In agreement
with the study of Verspoor and Edelenbos, Verspoor et al. conclude that the foreign language
acquisition process by CLIL students mainly occurs implicitly as they learn the foreign
language with only little explicit attention to grammar (62). Because CLIL students receive
more authentic input, they use relatively more grammatical as well as lexical chunks and
collocations (Verspoor et al. 47). What is interesting, Verspoor et al. point out, is that CLIL
students make more errors in collocations and chunks because they take more risks and their
output is more substantial than their mainstream peers (47). These errors will, however, cease
to exist over time (Verspoor et al. 47).
16
2.4 Implicit and Explicit Knowledge
One of the most controversial issues in the second language teaching field concerns the value
of teaching a language and its grammar explicitly in class. Explicit knowledge “consists of the
facts that speakers of a language have learned” states Ellis (95). It is “held consciously, is
learnable and verbalizable, and is typically accessed through controlled processing when
learners experience some kind of linguistic difficulty in using the L2” (Ellis 95). Implicit
knowledge, on the other hand, “is held unconsciously, and can only be verbalized if it is made
explicit. It is accessed rapidly and easily and thus is available for use in rapid, fluent
communication” (Ellis 95). As Ellis points out, there is a general consensus that competence
in a second language is “primarily a matter of implicit knowledge” (95).
The question, thus, is whether aspects such as grammar and vocabulary should be
taught explicitly at all. Although young children are not taught grammatical rules of their
mother tongue explicitly, they are able to form grammatical sentences. On the other hand,
although learners of English in secondary schools are often taught grammar rules explicitly,
they still do not acquire the foreign language on a native level. Krashen suggests that this can
be explained by the fact that explicit grammar teaching does not bring about acquired
knowledge which is “needed to participate in authentic communication” (qtd. in R. Ellis 167).
The question is to what extent learners use their consciously acquired knowledge in
performance. According to Krashen’s Monitor Theory, “learning never turns into acquisition”
(83) and conscious knowledge only plays a “limited role” in second language performance
(16). He claims that learned knowledge only has one function: that as a monitor or editor (15).
Foreign language utterances are initiated by acquired knowledge rather than conscious or
learned knowledge, which only alters the output (Krashen 15). CLIL thus heavily ties in with
Krashen’s theory as it promotes implicit language acquisition: Verspoor and Edelenbos and
Verspoor et al. also found that the language acquisition process in CLIL occurs primarily
17
implicitly (11; 62). Because CLIL uses language as a medium to learn, grammar is implicitly
used when speaking or writing. Difficult grammatical structures are not left ‘until later,’
which is often the case in the mainstream education curriculum. The unconscious knowledge
pupils acquire is important as this generates higher levels of language proficiency: “our
fluency in production is based on what we have ‘picked up’ through active communication”
(Krashen Introduction 2). In CLIL programmes, this language fluency is believed to be more
important than accuracy. Verspoor et al. point out that making mistakes is natural and
necessary to the language acquisition process; errors become less frequent over time and only
have to be corrected when they become entrenched (47).
2.5 The Role of Language Input
It is widely believed that frequency is an essential part of language acquisition: language
learners analyse the input they receive and eventually, “practice makes perfect” (Ellis
“Frequency Effects” 178). Since CLIL students consistently receive more input than their
mainstream peers, it could be suggested that they acquire the English language better. This is
supported by Kimbrough and Eilers who concluded that learners of English who are “more
consistently” exposed to language input perform better in the language (286). These effects
are largest among younger children, Kimbrough and Eilers reported, and diminish after a
certain critical mass of input has been reached (286). In addition, the effects are largest for
language features learned “item by item,” such as vocabulary acquisition (Kimbrough and
Eilers 286).
As to the role of language input, Krashen hypothesises that language acquisition takes
place when a language learner receives natural input that contains “i+1” and focuses on
meaning, not form (21). In other words, acquisition only takes place when the learner
18
understands input that is “a little beyond” his or her current level of competence (Krashen 21).
Krashen points out that this input will be automatically arrived at when communication is
“successful” and when there is sufficient input that is being understood by the learner (22).
Krashen’s Input Hypothesis thus relates to acquisition rather than learning: a learner must first
understand meaning and acquires the structure of the utterance as a result (21). Although early
speech may not be grammatically correct, Krashen argues, “[a]ccuracy develops over time as
the acquirer hears and understands more input” (22). In other words, language fluency and
correctness cannot be taught, but emerge over time.
2.6 Implications for Practice
As European Framework guidelines for CLIL teaching and prior studies (Verspoor et al.;
Verspoor and Edelenbos; Admiraal, Westhoff, and De Bot) have shown, CLIL programmes at
Dutch secondary schools stimulate implicit learning as they offer more and more qualitative
input in the target language than mainstream curricula and focus on communication rather
than correctness. Moreover, pupils are encouraged to produce English as often as they can,
which implicitly boosts their proficiency in the foreign language. Because pupils learn
implicitly, they are more fluent in the foreign language than their mainstream education peers.
Research on the effects of CLIL indicates that this type of education is effective in acquiring a
foreign language in a formal setting. For example, CLIL students, when tested both at the end
of the earlier years as at the end of the later years, have a higher level of proficiency in
English than their mainstream VWO peers.
Based on the positive outcomes of CLIL, several suggestions can be offered to foreign
language teachers. Naturally, mainstream students should be exposed to as many authentic
input as possible. Language teachers should therefore use the target language as the language
of communication during class as often as possible. In addition, instead of teaching single
19
words, teachers could draw attention to the larger chunks in which words occur as these are
very important in acquiring fluency (Verspoor et al. 47). In addition, grammar should be
taught more implicitly. Teachers should not only draw specific, in other words explicit,
attention to rules, but give students the opportunity to use the language actively, allowing
them to discover grammar rules for themselves and correct their own mistakes. Moreover,
more attention should be drawn to testing communicative skills rather than testing
grammatical competence as a goal in itself. Rather than tests that deal explicitly with
grammatical structures, grammar-integrated speaking-, writing-, reading-, or listening tests
could be used to assess students on their language skills. In other words, teachers should focus
more on the communicative effect of utterances and less on correctness. After all, Verspoor et
al. found that if students are exposed to sufficient language input and are provided with
enough opportunities to use the foreign language, any errors will disappear over time (47).
Last but not least, as to all the implications described above, teachers should implement these
suggestions from the very first year onwards.
20
Chapter 3 – Research Questions and Predictions
This thesis attempts to examine the immersion effect of CLIL on acquiring English
prepositions. Chapter 2 already explored why CLIL teaching is believed to be effective and
what prior studies have found as to differences between the quality and quantity of input of
mainstream language teaching and CLIL teaching. In the pages that follow, this thesis will
attempt to answer the following research question through a data-driven approach: Is CLIL
education more effective than mainstream education in acquiring English prepositions in the
context of Dutch secondary school students learning English?
Since several studies have reported positive effects of CLIL on the proficiency level in
English (Huibregtse in De Graaff et al. 605; Verspoor and Edelenbos 11), it could be expected
that CLIL teaching will have a positive influence on the acquisition of prepositions. The
quality and quantity of input CLIL students receive could facilitate the acquisition process of
English prepositions significantly. In addition, immersion in the CLIL classroom could
support CLIL students to acquire the language implicitly, which could help them gain a lead
on their mainstream peers with regards to the knowledge of English prepositions.
21
Chapter 4 – A Comparison Between English and Dutch Prepositions
In acquiring a second language, “cross-linguistic influence,” that is to say, transfer of prior L1
knowledge to the L2 system, has been found to occur (Saville-Troike 19). Two main types of
transfer may occur: positive transfer and negative transfer. The first occurs “when an L1
structure or rule is used in an L2 utterance and that use is appropriate or ‘correct’ in the L2”
(Saville-Troike 19) and the latter “when an L1 structure or rule is used in an L2 utterance and
that use is inappropriate and considered an ‘error’” (Saville-Troike 19). Since English and
Dutch are both West-Germanic and thus typologically related languages, they have several
features in common, which stimulates positive transfer from L1 to L2 by Dutch learners of
English. Both languages have, for example, a rather similar sound system which assists Dutch
learners of English to into producing and perceiving English sounds. Both languages,
however, differ greatly with regard to grammatical features such as word order and tense.
Dutch learners of English therefore tend to negatively transfer L1 knowledge into English
quite regularly, which results in ungrammatical English utterances (Saville-Troike 19).
Besides word order and tense, the feature under investigation in this thesis, prepositions, also
poses problems for Dutch learners of English. Whereas native speakers use prepositions
correctly and unconsciously, learners of English often have problems with English
prepositions due to various reasons.
First of all, English and Dutch often have completely different prepositions; for
example, “over tien minuten” in Dutch is “in ten minutes” in English. As a result, negative
transfer from Dutch to English often occurs by Dutch pupils. In addition, an English
preposition often has multiple translations in Dutch: “over the years,” “over the table,” and
“over 400 euros” translate into three prepositions in Dutch: “over,” “op,” and “meer dan.” In
short, a one-to-one translation is often not possible since a specific Dutch preposition cannot
consistently be translated into one and the same preposition in English.
22
There are, furthermore, also problems that simply have to do with the English
language itself, as Lindstromberg points out. The prepositions “at” and “by” are often rather
similar in meaning; in addition, expressions such as “at random” and “by chance” have a
similar meaning, which could result in the “cross-association of words” (Lindstromberg 5).
Many English prepositions are in fact similar in meaning in some circumstances whereas they
are not interchangeable in others: “by/at the seaside” but “*by random” or “*at chance”
(Lindstromberg 5). In addition, prepositions are, in general, pronounced rather softly and are
not stressed, which makes it difficult for learners of English to notice and differentiate
prepositions. For example, as Lindstromberg points out, learners might have difficulties
discriminating the prepositions “on” and “in” since they are “phonologically [as well as]
orthographically small and similar” (5).
In addition, problems may also have to do with the spatial representation of
prepositions. Bowerman argues: “[a]lthough all languages make categorical distinctions
among spatial configurations … they do not do so in exactly the same way” (144); in other
words, the classification of spatial relationships is language-specific. Whereas English uses
the same preposition in “on the shoulder,” “on the wall,” and “on the napkin,” Dutch
describes these spatial relationships with three different prepositions: “op de schouder,” “aan
de muur,” and “om de servet” (Bowerman 146). As illustrated, Dutch differs from English
with regards to the “classification of ‘on’ relationships” (Bowerman 146). In contrast to
English, Dutch is “sensitive” to the manner in which objects are attached, Bowerman points
out (146). For example, the choice of prepositions in Dutch depends on whether the contact
between two objects involves a horizontal surface or vertical surface and whether it is
“attached by one or more points” (Bowerman 146). Dutch pupils who are learning English
will thus have to learn the English categorisation of spatial relationships through experience
just like L1 learners do.
23
Although pupils are generally able to learn numerous prepositions that express time or
place by heart, they experience more serious problems with English collocations: words that
are often combined together. Prepositions in collocations often receive a special meaning and
collocations in general are, therefore, “hard to guess” (McCarthy and O’Dell 6). As a result,
Dutch pupils often use collocations incorrectly as they apply incorrect prepositions to a verb
or noun. For example, Dutch pupils often say “*He is engaged with Sarah” instead of “He is
engaged to Sarah” or “*I voted on Obama” instead of “I voted for Obama.” Since collocations
come naturally to native speakers and since they make speech sound more fluent and natural
they are good predicators of (near-)native fluency.
24
Chapter 5 – Method
The purpose of this study was to investigate the immersion effect of CLIL teaching on
acquiring English prepositions in an ESL classroom by Dutch secondary school pupils. In
general, Dutch pupils find English grammar rather hard, which is why it seemed interesting to
investigate the immersion effect of CLIL on acquiring grammatical structures. Since there are
hardly any rules as to the use of prepositions, it was decided to use this grammatical feature in
particular for this study.
5.1 Participants
The participants in this study were 80 secondary school pupils within an age range of 12-19.
Students were either in their first year or in their sixth year of O.R.S. Lek en Linge, which is a
regional, public comprehensive school in the rural area of Culemborg. This school attracts
students from Culemborg as well as students from other local areas such as Geldermalsen,
Tricht, Buren, and Beest. Lek en Linge has offered CLIL teaching since 2002. The team of
CLIL teachers mainly consists of Dutch teachers who have taken intensive courses in English
in order to maintain a high level of English and they are certified as proficient through
Cambridge. In addition, there are three English native speakers and, regularly, there are native
speaker trainees as well. The CLIL groups under investigation in this study both receive
English lessons from a native speaker.
First graders had an age range of 12-13 (mean age 12,39) and sixth formers of 17-19
(mean age 17,57). As illustrated by Table 1, the students were divided over four groups: a
6TVWO (CLIL) group of 13 pupils, a 1TVWO group of 29 pupils, a 6VWO group of 16
pupils, and a 1VWO group of 22 pupils. There was a roughly equal distribution of gender
within the first-year TVWO group and the sixth-year mainstream group: 14 females and 15
males in the first group and 7 females and 9 males in the latter group. Within the first-year
25
mainstream education group, however, there were visibly more females than males: 16
females in contrast to 6 males. In addition, within the sixth-year TVWO group there were
more males than females: 8 males in contrast to 5 females. It should be noted that the
mainstream group consisted of 5 additional students; however, their tests were disregarded as
they were obviously not filled in seriously and might therefore have influenced the outcomes
of the study negatively.
Students were not given the choice to participate as the study was carried out during
their regular English classes, which were mandatory for them. By means of a brief
questionnaire preceding the grammatical judgement task (GJT), it was found that TVWO
students enjoyed their English classes significantly more than their VWO peers. This might
have influenced the results, as the former group could have been more motivated in
participating than the latter one.
Table 1: Distribution of gender across all groups under investigated
Group
Females
Males
Total
1VWO
16
6
22
1TVWO
14
15
29
6VWO
7
9
16
6TVWO
5
8
13
5.2 Materials
In each group, one grammaticality judgement task in which students had to judge sentences
on a binary scale: (‘correct’ and ‘incorrect’ or ‘I do not know’) was administered. As part of
the GJT, students had to correct the sentences they believed were incorrect. By asking
26
students to underline the incorrect part in a given sentence, students could not simply guess
but had to think about their answers. In addition, such a design can be informative in showing
when students disregard a sentence based on grounds other than the preposition. In short, the
production component was deemed to make the GJT more reliable. In addition, an Elicited
Production Task was administered to assess the use of prepositions by students. Students were
asked to produce a written piece relating something about their last shopping experience in
order to evaluate the quality and quantity of their use of prepositions. Lastly, a brief
questionnaire was administered before the GJT to collect more background information on all
the students. These questions focused on aspects such as motivation, interest, and proficiency
in English. The questionnaire can be found in Appendix A; the GJT and the production task
can be found in Appendix B; the Elicited Production Task can be found in Appendix C.
The GJT consisted of 40 sentences: 30 sentences focusing on the use of prepositions
and 10 fillers to distract students from the actual feature under investigation. All sentences
were equally split into correct and incorrect sentences and all sentences were randomly
shuffled. All sentences containing prepositions could, in turn, be divided into three main
categories: 10 sentences focusing on time prepositions, 10 sentences focusing on place
prepositions, and 10 sentences that tapped into English collocations. The first category
consisted of prepositions that dealt with the days of the week, time of the day, etcetera. See
(1) below for an example.
1. George is leaving Monday at noon.
The second category consisted of prepositions that dealt with position as well as direction.
See (2) below for an example of the first and (3) for an example of the latter.
27
2. *In our house, there is a large painting at the wall.
3. They strolled along the beach watching the sunset.
The third category tapped into common English collocations containing prepositions. See (4)
below for an example.
4. *She apologised about the mistake.
Lastly, the fillers dealt with matters such as English tenses, word order, and English plural
formation in order to distract students from the actual feature under investigation.
Prepositions were, thus, assessed in every conceivable form.
In the Elicited Production Task, students had to relate something about their last
shopping experience since this was expected to elicit quite some prepositions. In order to
promote the use of prepositions, several prompts were presented, such as ‘when and where
did you go shopping?’ and ‘how did you get there?’ In order to give students the opportunity
to write down as much as possible, they did not necessarily have to answer in full sentences
but were allowed to write down short phrases instead.
5.3 Procedure
Each group took part in one session of 30 minutes. All sessions took place during regular
English classes. The tests in both the first-year groups and the sixth-year TVWO group were
administered in two weeks’ time; however, due to miscommunication, the test was
administered 2,5 weeks later in the sixth-year mainstream education group, that is, on their
penultimate day of school. The wrong group of students, that is, another CLIL group instead
of a mainstream education group, was tested initially. The language of instruction was
28
English in the TVWO classes and Dutch in the mainstream education classes. Groups first
received instruction as to why they were asked to participate and what they were going to do.
Students were told they were to participate in a MA study that would investigate the
difference in language proficiency between TVWO students and mainstream education
students. Students knew they would not get a mark for the test but that the point was to test
their intuitions about the sentences, so as to reduce anxiety levels. It was stressed that students
had to correct sentences they believed were incorrect since this was an important aspect of the
study. Because the concept ‘grammatical’ was not clear to many first-year students, this was
explained as well. Students were also told that an initial hunch usually is the best one, and that
they had to judge the sentences immediately after reading. Lastly, they were told they had to
do the assignment individually and finish it within twenty minutes. Since some sentences
consisted of rather difficult words, first-year students were allowed to ask the meaning of
words unknown to them during the task; this was done in English in the TVWO group since
this was their standard language of instruction. Whereas about half of the students needed the
full 20 minutes, the other half did not need that much time.
When the twenty minutes were over, students received instruction about the Elicited
Production Task. Students were told that they would have 5 minutes to complete the task,
which proved ample time as all students finished the task fairly quickly. It was stressed that
they did not necessarily have to write down full sentences but were allowed to write down
phrases; this was illustrated by means of a Dutch example (see Appendix C). Students were
also told that they had to answer at least the eight prompts in the assignment (see Appendix
C), but that they were free to write down anything they wanted.
5.4 Data Analysis
29
Data of the GJT was analysed by the administrator on the base of 13 different categories (see
Table 2 below). The first column of Table 2 indicates the category; the second column
indicates whether the trial was correct or incorrect; the third column indicates whether the
student believed the sentence to be correct or incorrect; the fourth column indicates whether
the student corrected the trial he or she believed to be incorrect correctly or incorrectly; the
fifth column signifies what the category implies in terms of preposition knowledge; the last
column indicates whether the student’s answer was believed to be incorrect or correct by the
administrator. In order to inspect students’ performance with respect to prepositions, data was
entered into SPSS. Categories 1, 4, 11, 12 were recoded as variable 1 (correct), whereas
categories 2, 3, 5, 6, 7, 8, 9, 10, 13 were recoded as variable 0 (incorrect). In order to inspect
instances of insecurity by students, only category 13 was taken into consideration and recoded
as a variable. In order to investigate how often students rejected a sentence in the GJT on
grounds other than the preposition, categories 9 – 12 were recoded as a variable as well.
Lastly, categories 6 and 7 were recoded as a variable in order to assess the mean number of
times students knew something was wrong with the preposition but did not know how to
correct that or corrected it wrongly.
Table 2: Categories on the base of which data in the GJT was analysed.
1
Trial +
Pupil +
Student understands that preposition is
correct.
+
2
Trial +
Pupil -
Correction
-
Student does not understand that preposition is correct; he/she corrects it wrongly.
3
Trial +
Pupil -
Correction
+
Student does not see that preposition is
correct; he/she comes up with a correct
alternative.
-
4
Trial -
Pupil -
Correction
+
Student understands that preposition is
wrong; he/she corrects it correctly.
+
5
Trial -
Pupil +
Student does not understand that preposition -
30
is wrong.
6
Trial -
Pupil -
Correction
-
Student understands that preposition is
wrong, but he or she corrects is wrongly.
-
7
Trial -
Pupil -
No
correction
Student does understand that preposition is
wrong, but he or she does not know how to
correct it.
-
8
Trial +
Pupil -
No
correction
Student does not see that the preposition is
correct, he or she also does not give an
alternative.
-
9
Trial -
Pupil -
Correction
+
Student does not recognise incorrect
preposition. Instead, he or she focuses on
another aspect in the sentence and corrects
this correctly.
-
10
Trial -
Pupil -
Correction
-
-
11
Trial +
Pupil -
Correction
-
12
Trial +
Pupil -
Correction
+
Student does not recognise incorrect
preposition. Instead, he or she focuses on
another aspect in the sentence and corrects
this correctly.
Student understands that preposition is
correct. He or she focuses on another aspect
in the sentence and corrects this wrongly.
Student understands that preposition is
correct. He or she focuses on another aspect
in the sentence and corrects this correctly.
‘I don’t know’
13
+
+
-
Data of the Elicited Production Task was analysed by first of all counting the number of
prepositions that were used. The number of errors made in preposition use was subtracted
from this total number in order to arrive at a raw index of preposition use. As such, there
would be no maximum score, but a higher figure would indicate a better command of English
prepositions.
31
Chapter 6 – Results
This chapter presents the results of several (independent and paired sample) t-tests that were
carried out in order to determine whether any significant difference could be found between
Dutch CLIL students and mainstream education students in their performance with respect to
English prepositions on the basis of a Grammaticality Judgement Task including a correction
format and an Elicited Production Task in which students were asked to relate something
about their last shopping experience, as this was expected to elicit a considerable number of
prepositions (see section 5.2).
6.1 Demographic Background Information of the Students
Before the outcomes of the GJT and the free use of prepositions are reported, Table 3 below
first of all lists some background information on all the participants that took part in this
study. In particular, it indicates whether students enjoyed their English classes; how difficult
they considered English to be; how they rated their own L2 proficiency; whether they were
also interested in Anglophone culture; and how much contact they had with the language
outside of their school hours. In addition, it indicates students’ average grade for English and
whether they though that TVWO students had a better command of English than mainstream
VWO students.
Students were asked to provide a rating from 1 to 5, with 1 being the lowest and 5 the
highest for the first five questions. For the sixth question, that is, whether students thought
that TVWO students had a better command of English than mainstream VWO students,
students were presented with a multiple choice question consisting of three options: there is
no difference at all, TVWO is better than mainstream education, or TVWO is much better
32
than mainstream education. As for the question ‘how much contact do you have with the
language outside of your school hours,’ students were asked to elaborate on their multiple
choice answers in order for the administrator to check the validity of their choice. The
numbers presented in Table 3 below are the mean ratings and standard deviations. Level 3
mostly involved watching English films, playing games in English, or reading English books.
Level 4 also included students watching the English news regularly as well as students
producing English at home in their spare time; this involved such things as chatting in
English, writing poems, speaking English to parents or family members or helping them with
their English, etcetera.
Table 3: Demographic background variables (and standard deviations) for all groups under
investigation
Enjoy
Difficult
SelfCulture
assessed
interest
proficiency
Contact
outside
school
Grade
TVWO
better?
VWO1
(n=22)
2.95
(0.90)
2.82
(1.37)
3.14
2.36
3.14
7.70
1.00
(1.04)
(0.90)
(0.47)
( 0.77)
(0.54)
TVWO1
(n=29)
3.88
3.07
3.86
3.34
4.17
7.55
1.38
(0.73)
(1.13)
(0.52)
(1.04)
(3.25)
(0.65)
(0.56)
VWO6
2.94
2.56
2.81
2.87
3.25
6.59
1.38
(n=16)
(0.85)
(1.26)
(0.91)
(1.09)
(0.45)
(0.76)
(0.50)
TVWO6
(n=13)
3.92
2.54
3.62
3.62
4.08
8.04
1.62
(0.28)
(0.88)
(0.51)
(0.51)
(0.28)
(0.75)
(0.51)
What is striking in these results is that, already at the first grade level, TVWO students
indicated that they enjoyed their English classes significantly more than their VWO peers,
33
assessed by an independent samples t-test: t(49) = -4.061, p < .000. This difference in the
extent the students enjoy their English classes remains visible throughout the years, rendering
a significant difference between TVWO6 and VWO6 as well: t(18,753) = -4,343, p < .000.
Similarly, both 1TVWO and 6TVWO students are significantly more interested in
Anglophone culture in general than their mainstream peers (t(28,856) = -3.011, p < .005 for
year 1 and t(22,103 = -2.419, p < .05 for year 6) and have significantly more contact with the
language outside of their school environment: t(49) = -8.682, p < .000 and t(25,432) = -6.093,
p < .000 for year 1 and year 6 respectively.
Interestingly, based on the results in Table 3, it can be concluded that neither the firstyears nor the sixth formers think that English is easy, as compared to their mainstream peers.
Only 1TVWO students consider English to be slightly easier than their mainstream
counterparts. However, it needs to be pointed out that this difference was not significant.
Despite the minimal differences in how difficult students perceive English to be, the first-year
TVWO students did believe they were better at English than their VWO peers: t(46,458) = 2,456, p < .05. Interestingly, that all groups think that TVWO students in general have a better
command of English than mainstream VWO students, although this is not significant. Over
the years, this tendency does, however, grow slightly stronger, as indicated by the results in
Table 2. What is also interesting is that TVWO students in all years believe TVWO students
in general to be better in English than their mainstream peers believe. What is more, both the
first and the sixth year TVWO students significantly rated their own English language
proficiency higher than the VWO students did: t(48,113) = -3.011, p < .005 for year 1 and
t(27) = -2.837, p < .01 for year 6. Objectively, the first-year TVWO students do not yet show
this superior language proficiency in the sense that their average grade for English is not
significantly high (in fact, it is a little lower on average) compared to the grade obtained in the
mainstream parallel group. However, by the sixth form, the TVWO students are markedly
34
better at English than their mainstream counterparts, as evidenced by a substantially higher
mark for English on average: t(27) = -5,133, p < .000. It is finally interesting to point out that,
while initially quite high in the first form, the mark for English in the mainstream education
group drops considerably from 1VWO to 6VWO. In short, as opposed to the TVWO group
the mark for English does increase throughout the years, but decreases in the mainstream
VWO classes under investigation here.
6.2 GJT Results
The data in Table 4 below presents the students’ performance with respect to prepositions; the
table lists the mean scores (and standard deviations) on the Grammaticality Judgment Task. It
should be noted that the maximally obtainable score in this task was 30, discarding the 10
fillers that were also included in the test. For the sake of this analysis, only the total score of
recoded 0 (wrong answer) and 1 (correct answer) variables was taken into consideration (see
section 4.4).
Table 4: Mean scores (and standard deviations) on the grammaticality judgment task
Mean score
SD
(max=30)
VWO1 (n=13)
11,62
2,53
TVWO1 (n=26)
13,42
2,59
VWO6 (n=13)
16,69
3,50
TVWO6 (n=10)
25,50
1,72
Based on the results in Table 4, it can be concluded that both the VWO and TVWO students
improve significantly from their first to their final year. In mainstream VWO alone, the sixth
35
formers outperformed their younger peers, as assessed by an independent samples t-test: t(24)
= -4.238, p < .000 and the same picture emerged for the TVWO students: t(34) = -13.558, p <
.000. What is more, the TVWO students were significantly better at evaluating the
grammatical content of the sentences in this task than the VWO students. While this is
perhaps not surprising in the case of the sixth formers: t(18,298) = -7.925, p < .000, it is
remarkable that already in the first year the TVWO students outperform their mainstream
peers: t(37) = -2.066, p < .05.
6.2.1 Instances of Insecurity
As described in the method section, students could choose between correct, incorrect, or I
don’t know when evaluating the sentences in the GJT on their grammaticality. A t-test was
carried out in order to determine whether there was a significant difference between the
frequency of the ‘I don’t knows’ used by the groups under investigation. Table 5 below
details the number of times students in all years and all types of education circled the ‘I don’t
know’ option on the grammaticality judgment task.
Table 5: mean number of times (and standard deviation) the ‘I don’t know’ option was
selected in the GJT.
Mean number of times
SD
VWO 1 (n=22)
3.14
3.68
TVWO 1 (n=29)
1.41
2.16
VWO 6 (n=16)
2.50
3.46
TVWO 6 (n=13)
0.85
0.80
36
Several trends can be distilled from this table. First of all, all students appear to become more
confident through the years: the 1VWO students used the ‘I don’t know’ option more than
their 6VWO peers and the same is true for the two cohorts of TVWO students. At no point,
however, was this decrease in use of the ‘I don’t know’ option significant. Within one year
cohort, the TVWO students used the ‘I don’t know’ option less often than the VWO students.
Although this difference was not actually significant in the sixth form, it was near significant
in the first-years (established by means of an independent samples t-test): t(31,807) = 1.952, p
= .060.
6.2.2 Disregarding Prepositions
It was evidenced that students often rejected a sentence in the GJT on ground other than the
preposition (see section 5.4). This comprised students who did not recognise the incorrect
prepositions and focuses on another aspect instead (which was actually correct) and students
who understood that the preposition was correct, but focused on another aspect (which was
actually correct as well). Table 6 below presents the mean number of times students rejected a
sentence in the GJT on grounds other than the preposition.
Table 6: mean number of times (and standard deviations) students rejected a sentence in the
GJT on ground other than the preposition
Mean number of times
SD
VWO 1 (n=22)
4.09
2.49
TVWO 1 (n=29)
3.69
1.89
VWO 6 (n=16)
3.19
2.34
TVWO 6 (n=13)
3.38
1.56
37
As can be seen in Table 6, first-year TVWO students were less quick to reject a sentence on
grounds other than the preposition it contained, but only marginally so. In other words, the
difference between the VWO and TVWO students in their first year was not significant. In the
6th form the scores were in fact so close together that the VWO and TVWO groups can even
be said not to be different at all.
When investigating the occurrences of students rejecting a sentence on ground other
than the prepositions, some interesting findings can be reported. It should, however, be noted
that no statistics were run on these figures. The findings reported below are based on careful
observation: occurrences were tallied and divided by the number of students in each group.
First of all, it can be concluded that across the four groups and years, students focused mostly
on tense, vocabulary (which entailed nouns as well as verbs), and the insertion or deletion of
words, phrases, or additional prepositions. Within the first year, 1VWO students focused
considerable more on word order than 1TVWO students. 1TVWO students, on the other hand,
focused considerably more on vocabulary as well as spelling than their mainstream peers.
Both VWO and TVWO students focused equally on tense. On the whole, numbers declined
over the years. Remarkably, 6TVWO students paid noticeably more attention to vocabulary
compared to their 1TVWO as well as 6VWO peers. Another great difference can be found in
students focusing on tense: 6VWO students focused considerably more on tense than their
6TVWO peers.
6.2.3 Students Unable to Correct Incorrect Trials
As a part of the GJT, students had to correct incorrect trials. Table 7 below presents the mean
number of times students knew something was wrong with the preposition but did not know
how to correct it or corrected it wrongly.
38
Table 7: mean number of times (and standard deviation) students knew something was wrong
with the preposition but did not know how to correct that (left that open) or corrected it but
the correction was wrong
Mean number of times
SD
VWO 1 (n=22)
0.86
1.55
TVWO 1 (n=29)
1.27
0.96
VWO 6 (n=16)
1.12
1.59
TVWO 6 (n=13)
0.23
0.60
As shown in Table 7 above, 1TVWO students more often found themselves in a situation
where they did know something was wrong with the preposition (evidenced by them
underlining the incorrect preposition in the sentence), but did not know how to correct it or
produced an incorrect alternative. The difference in scores was, however, not statistically
significant. Also interesting to note is that the situation was reversed in the sixth formers: the
TVWO students were less often found to be able to point to the error but unable to correct it
than their mainstream peers, even significantly so: t(22,363) = 2.514, p < .05.
6.3 Elicited Production Task
Table 8 below presents the findings of the free elicited writing assignment (see section 5.2): it
indicates the raw index of preposition use in these assignments. This index was arrived at by
subtracting the number of errors made in preposition use in this assignment from the total
number of prepositions that were used. As such, there is no maximum score, but a higher
figure does indicate a better command of English prepositions. To provide more information
regarding this index, Table 9 below shows the (raw) mean number of times prepositions were
used in the production task by both groups of learners, followed by the mean number of errors
39
that were produced. The third column, finally, translates this mean number of errors to
percentage of errors vis-à-vis the mean number of preposition usages.
Table 8: mean preposition index (and standard deviations) on the free elicited writing
assignment
Mean preposition index
SD
VWO 1 (n=22)
2.73
1.49
TVWO 1 (n=29)
4.86
2.46
VWO 6 (n=16)
5.00
1.67
TVWO 6 (n=13)
7.08
2.47
Table 9: mean number of times (and standard deviations) prepositions were used by learners,
mean number of errors (and standard deviations) produced, percentage of erroneously
produced prepositions.
Number of
Number of errors
Percentage of
preposition used
produced
erroneously
produced
prepositions on the
production task
VWO1 (n=22)
4.73 (1.78)
2.00 (1.11)
42.28%
TVWO 1 (n=29)
6.28 (2.69)
1.45 (1.09)
23.09%
VWO6 (n=16)
5.50 (1.46)
0.44 (0.63)
8%
TVWO6 (n=13)
7.54 (2.60)
0.46 (0.66)
6.1%
The trend that emerges in Table 8 clearly resembles that of the GJT: all students show a
marked progression, with VWO6 students outperforming students in VWO1: t(36) = -4.415, p
<.001 and TVWO6 students being significantly better than TVWO1 students: t(40) = -2.696,
p < .05. Furthermore, the TVWO students outperform their mainstream peers, both at the
40
sixth form level (t(27) = -2.696, p < .05) but also again in the first grade: t(49) = -3.598, p <
.005.
In general, first formers clearly translated prepositions directly from Dutch; they wrote
down phrases such as “*with the car” instead of “by car” and “*by the Sting” instead of “at
the Sting” in both the mainstream group and the CLIL group. 1TVWO students, however,
elaborated more on the details of their shopping experience and, therefore, used more
prepositions. In comparison with their first-grade peers, most sixth formers used prepositions
such at “by” and “at” described above correctly; in other words, sixth formers did no longer
negatively transfer prepositions from Dutch to English. Sixth formers in general used a
variety of prepositions and these were rather the same across both groups: mostly “in,” “at,”
“with,” and “by” occurred. 6TVWO students did use more prepositions of time, such as
“before,” “after,” and “during” as they elaborated more on their shopping experience than
their mainstream counterparts as they described how their day proceeded more detailed.
Whereas 6VWO students were clearly answering the prompts given in the assignment (see
section 5.2) and wrote down rather short sentences, 6TVWO students wrote more freely and
used more complicated structures and linkers, which is why their writing appeared more
authentic in general. Although they did not use different prepositions than their mainstream
peers, neither 6VWO students nor 6TVWO students used noteworthy collocations including
prepositions, they did use more prepositions and their prepositions were more mixed in their
writing.
41
Chapter 7 – Discussion
The present study was designed to investigate the immersion effect of CLIL teaching on
acquiring prepositions in an ESL classroom by Dutch secondary school pupils. Prepositions
are interesting to look at since using them fluently without much deliberation makes speech
more authentic and natural. Interestingly, Dutch and English have very different prepositions
and prepositions in English are often similar in meaning in some circumstances whereas they
are not interchangeable in others, which might be very confusing for learners of English. In
short, there are hardly any rules as to when to use which preposition the only way to learn
them is to learn collocations by heart or to simply look up prepositions in a dictionary.
Although a large body of studies has reported and commented on the positive effects of CLIL,
most of these studies have tended to focus on general language proficiency skills such as
speaking or writing. In reviewing the literature, no data was found on the effect of CLIL
teaching on acquiring specific grammatical features. This study was, therefore, designed to
investigate the effect of CLIL on acquiring prepositions. Through a data-driven approach, the
present study investigated whether CLIL education is more effective than mainstream
education in acquiring English prepositions.
The present results are significant in at least two respects. Regarding the question of
whether CLIL education is more effective in teaching English prepositions than mainstream
education, this study first of all found that CLIL students in both the first and the sixth year
were significantly better at evaluating the grammatical content of sentences in the GJT than
mainstream students (see section 6.2). This finding could also suggest that CLIL students
acquire prepositions quicker than mainstream education students since CLIL students score
significantly higher on the GJT in the first grade already. However, this explanation should be
interpreted with caution since in order to be admitted to TVWO, CLIL students already need
to have a quite high CITO score, including the language component. In other words, it may
42
well be that the CLIL students under investigation were always better from the outset and
they did not acquire anything fast anyway. It can, thus, only be suggested that the acquisition
process of CLIL students goes faster than mainstream education students based on students’
knowledge at the end of their first year. What is more, this study found that CLIL students
were significantly better than VWO students at the Elicited Production Task, both in the sixth
form level as in the first grade (see section 6.3). In other words, TVWO students produced
more accurate prepositions than VWO students. There are several possible explanations for
these results.
First of all, the results of this study (see section 6.1) showed that TVWO students in
the first year as well as in the sixth year enjoyed their English classes significantly more than
their mainstream education peers. In addition, the TVWO students under investigation
showed a significantly higher interest in Anglophone culture compared to their mainstream
peers (see section 6.1). Another finding was that TVWO students had more contact with
English outside their school hours than their mainstream education peers (see section 6.1).
One of the issues that emerges from these findings is that TVWO students might have been
more motivated to participate in the study which might have influenced the results.
Interestingly, there are similarities between the attitudes expressed by the students in this
study and those described by De Bot and Maljers. As De Bot and Maljers point out in their
article, CLIL education began to appeal to students in the latter half of the 90s (In De Graaff
and Tuin 140-141); the results described above show that this trend is still visible among
students today. This study confirms that motivation and interest are correlated with the
difference in educational types. There are several possible explanations for this outcome. First
of all, because of the specific goals set for CLIL teaching in the Netherlands (recall section
2.3.1), students may well experience English classes differently than their peers. In addition,
the first-year CLIL students under investigation in this study indicated to particularly like
43
their native teacher, which contributes to these results as well. Naturally, it might be expected
that CLIL students in general are more interested in the English-speaking world since they
deliberately chose to attend the CLIL programme, rendering a higher proficiency level almost
by definition. It could be questioned to what extent the higher level of motivation and interest
plays a role in the acquisition of prepositions by students; however, it could be argued that it
contributes to the higher levels of speaking and writing skills in general as focused on by prior
studies (Admiraal, Westhoff, and De Bot 61; Verspoor and Edelenbos 11).
Another important finding was that CLIL and mainstream students considered English
to be as difficult; only first-year CLIL students considered English to be slightly easier than
other students, although this difference was not significant (see section 6.1). Perhaps related
to this issue, but distinct at the same time, is that CLIL students in both the first year and the
sixth year significantly rated their own English language proficiency to be higher than the
mainstream students did (see section 7.1). It seems possible that these results are due to
TVWO students enjoying English more than mainstream students. Interestingly, however, is
that this superior language proficiency is not reflected in the average grades for English of the
first-year CLIL students as these were not significantly high compared to those of the
mainstream education students (see section 7.1). In other words, first-year CLIL students
consider themselves to be better than mainstream education students although their average
grade is in fact lower. This rather contradictory result can be explained by the CLIL standards
being much stricter than the mainstream education standards; in other words, CLIL students
have to put considerably more effort into their work in order to be awarded with the same
grade mainstream students are being awarded with. The result may, however, also be
explained by first-year CLIL students, unjustly, having a relatively high opinion about
themselves. It could also be explained by the fact that first-year students are still rather young
and simply find it more difficult to provide an accurate self-rating of their own language
44
proficiency compared to sixth formers. What is more, CLIL students under investigation
believed that TVWO students in general have a better command of English than mainstream
education students think they have (see section 7.1). In any case, mainstream VWO students
still believed that TVWO students have a better command of English, which is in agreement
with earlier studies that described the higher level of proficiency by CLIL students (see
section 6.1) (Verspoor and Edelenbos 11; Verspoor et al. 61; Admiraal, Westhoff, and De Bot
86-89; Huibregtse in De Graaff et al. 605). Interestingly, students’ self-assessment is in
agreement with the actual outcomes of this study. What is surprising is that the average mark
for English in the mainstream education group dropped considerably over the years (see
section 7.1), which could indicate a cohort effect in that the 6VWO class under investigation
consisted of relatively weak (language) students. On the other hand, it should be noted that
first graders overall tend to be more studious pupils than sixth formers, which might explain
why the mark dropped for English: first graders still spend more time on their (English)
homework than sixth formers; in addition, coursework generally increases in 6VWO, which
means less time is available for each individual subject. The fact that this is not true is for the
6TVWO group is because of their overall greater quantities of English input.
Since English prepositions are rather hard for Dutch pupils due to various reasons
(recall Chapter 2), it is striking that first-year CLIL students already have a significantly
better command of prepositions than mainstream students. This result could be explained by
the fact that the input CLIL students receive plays an important role in their language
acquisition process (Kimbrough and Eilers 286; Krashen 21). In agreement with Kimbrough
and Eilers (286), this study suggests that learners of English who are exposed to more
language input do better in language courses. However, whereas Kimbrough and Eilers
argued that this effect diminishes over time (286), the present study suggests that the effect of
input remains to exist over time. Compared to mainstream education, at least fifty per cent of
45
the subjects are taught in English in the lower years (recall section 2.3.2). Although in the
higher years, this percentage might drop, CLIL students still receive a large quantity of
English input. A large quantity of input is, however, not enough to explain the higher levels of
proficiency of CLIL students (Verspoor and Edelenbos 12; Marsh “Adding Language,”
par.6). The quality of input also plays an important role in explaining the results. Krashen
argued that language acquisition takes place when the learner receives input that is natural and
is somewhat beyond his or her current level (21); since the students in this study were taught
by native speakers, the quality of input can be expected to meet Krashen’s standards. CLIL
students in general are taught by at least two native speakers; in addition, the level of nonnative speakers is at least B2 for all language skills (recall section 2.3.2). What is more, CLIL
students are exposed to and have access to authentic materials in English at all times (recall
section 2.3.2). Taken together, these findings support strong recommendations for CLIL
education teachers to safeguard the quality and quantity of input and for mainstream
education teachers to improve both the quality and quantity of input.
Implicit language learning could also be a major factor causing the lead of CLIL
students. Verspoor and Edelenbos attributed the rate of acquisition by CLIL students to the
fact that they learn the foreign language implicitly (11). In addition, Verspoor et al.
concluded that the foreign language acquisition process by CLIL students mainly occurs
implicitly since they receive only little explicit attention to grammar (62). As De Graaff et al.
argued, fluency in the foreign language can best be achieved by using the language as a
“functional medium of communication and information” rather than by “making it the object
of analysis in class” (606). The latter is the case in mainstream education: students mainly
focus on the form of language and learn grammatical structures explicitly. Since CLIL
students learn a language mostly implicit as a result of a sufficient level of quality and
quantity of input, they acquire unconscious knowledge. According to Krashen, this
46
unconscious knowledge plays a crucial role in foreign language proficiency, whereas learned
knowledge only plays a limited role in language performance (16). It could thus be assumed
that TVWO students are indeed faster to reach higher proficiency levels in English, as
Verspoor and Edelenbos argue (11), because implicit language learning accelerates the
acquisition process. This explains why mainstream education pupils, who learn a language
explicitly and acquire mostly learned knowledge rather than acquired knowledge, show a
reduced command of prepositions.
The findings of this study thus suggest that CLIL students have a higher command of
English prepositions. What is more, a qualitative analysis of the Elicited Production Task
suggest that the language use of TVWO students is more authentic in the sense that it
resembles more closely that of native speakers than mainstream education students. First of
all, the TVWO students made more use of (correct) prepositions in their writing (recall
section 6.3), which might indicate that they take more risks in their writing as suggested by
Verspoor et al. (47). Especially 6TVWO students elaborated on the details of their shopping
experience and, as a result, used more prepositions in general and also more prepositions on
time in contrast to mainstream students. Interestingly, whereas the CLIL students under
investigation used more prepositions, they did not make more errors than the mainstream
students. The present results differ, therefore, from the findings of Verspoor et al., which
showed that CLIL students make more errors in their language output because they take more
risks and their output is more substantial (47). It should be noted that in case of the first-year
TVWO students under investigation, the amount of prepositions might have been related to
the fact that they had recently gone on a school trip to Newcastle and visited shops there.
Students were clearly motivated to write about this experience. Mainstream education
students, on the other hand, merely responded to the prompts they were given (see Appendix
C). The higher quantity of prepositions could be ascribed to the authenticity of input CLIL
47
students receive. In addition, this might also explain the more natural language use of CLIL
students as they entwined prepositions more freely in their writing. However, whereas prior
studies of Verspoor and Edelenbos as well as Verspoor et al. found that CLIL students make
more use of collocations (11; 47), this study did not report such an effect. In addition, CLIL
students used mostly the same prepositions as mainstream education students. Since all
students had to write about the same topic and were given prompts, it is to be expected that
they use fairly the same prepositions. The fact that the TVWO students under investigation
did not produce more errors than mainstream education students provide support for the
hypothesis that focusing on fluency instead of accuracy is seminal (Verspoor et al. 62,
Krashen 22): they show that CLIL students, who have not received explicit attention to errors
all the time, produce significantly more accurate prepositions than their mainstream peers. As
pointed out by Verspoor et al., making mistakes is natural, especially in the initial stages of
the language acquisition process (47). Mainstream education teachers should, therefore, not
focus purely on accuracy.
This study also found that students become more confident through the years: as
6TVWO and 6VWO students used the option ‘I don’t know’ less than 1TVWO and 1VWO
students respectively, although not significantly so (see section 6.2.1). This could be
explained by the fact that first-year students opted ‘I don’t know’ more often since the
sentences in the GJT were rather difficult for them. They often did not understand parts of a
sentence which lead them to choose ‘I don’t know.’ This study also found that TVWO
students were more sure of themselves than their mainstream peers, although this was not
significantly different in the sixth form but near significant in the first grade. These findings
must be interpreted with caution: whereas the number of ‘I don’t knows’ could reflect genuine
ignorance, it could also reflect insecurity or a lack of interest as it is very easy to say ‘I don’t
know.’ First of all, it could be argued that the lower usage of the option ‘I don’t know’ by
48
TVWO students reflect a higher knowledge of the language. This accords with earlier
observations, which showed that TVWO are better at assessing the grammaticality of
sentences and produced more correct prepositions in their writing (see section 6.2 and 6.3). It
could be argued that because of their acquired unconscious knowledge, CLIL students were
able to evaluate sentences more easily in contrast to mainstream students who acquired
primarily learned knowledge. Since this learned knowledge is believed to only play a
significant role (Krashen 15), mainstream students might have more difficulty evaluating the
sentences. The present finding could, however, also reflect the lower motivation level of
mainstream students to participate in this study: since the tests were administered on their
penultimate day of school, mainstream students could have been less motivated to participate
actively and could therefore have opted ‘I don’t know’ more often.
As to the ability of students to correct incorrect trials in the GJT, the results of this
study surprisingly showed that 1TVWO students more often found themselves in situations
where they were able to underline the problem in the sentence, but did not know how to
correct it or produced an incorrect alternative (see section 6.2.3). This seems to suggest that
first-year mainstream pupils are better at correcting incorrect trials than their TVWO peers.
Although this result may seem counterintuitive and contradictory to previous results at first,
the TVWO students in general were better at picking out incorrect prepositions than their
mainstream VWO peers. Strikingly, this situation reversed in the sixth formers. In the sixth
year, mainstream students knew significantly more often that something was wrong with the
sentence but did not know how to correct it or corrected it wrongly. Although previous
findings (recall section 6.2) reflected the increased knowledge base of the 6TVWO students
as well as that of the 6VWO students since they may well notice more incorrect prepositions
in their final year than in their first, 6VWO students are more often unable to correct these
errors as opposed to 6TVWO students. This further supports the idea that grammatical errors
49
cease to exist (Verspoor et al. 62) and accuracy develops over time (Krashen 22) in a CLIL
learning environment. When interpreting these results, it should, however, be noted that the
GJT in the 6VWO class was administered on their penultimate day of school; in other words,
students might have been less motivated in this group to write down the correction at all
although they might have known the answer.
This study also found that VWO and TVWO students in all years rejected a sentence
on grounds other than prepositions roughly as much (see section 6.2.2). Only TVWO students
were a little less quick to do this, although only marginally so. Although it is difficult to
explain this result, it might be related to the fact that students often tend to think that
something has to be wrong with a sentence when set a task like a Grammaticality Judgement
Task and therefore look for things that are in fact correct. For instance, practice has showed
the administrator that students begin to hesitate when successive sentences all turn out to be
correct; as a result, students start thinking that one of the sentences has to be incorrect or an
incorrect sentence must show up soon. On the other hand, the 6TVWO students in this study
also often underlined aspects in the sentence different from the prepositions as a result of finetuning the sentence. These sixth formers often replaced or added words in order to make the
sentence sound more authentic; for instances, “*We strolled over the beach while watching
the sunset” instead of “*We strolled over the beach watching the sunset.” It should, however,
be noted that the rather difficult words in the GJT might have caused especially first-year
students to reject sentences on grounds other than the prepositions. Although it was expected
that first-year students would focus on aspects other than the prepositions more often than
sixth formers given the fact that they had more difficulty understanding some sentences, this
was not reported. It could be argued that first graders and sixth formers focused on other
elements equally as much, although for different reasons. This study found that 1VWO
students in general focused more on word order than 1 TVWO students; this finding reflects
50
the fact that in the first year of mainstream education, grammar teaching focuses greatly on
word order (recall Chapter 3) and is, furthermore, taught explicitly in contrast to TVWO
where there is not as much as an emphasis on explicitly teaching grammar. As a result, VWO
students were more conscious of word order than TVWO students. First-year TVWO
students, on the other hand, focused more on vocabulary and spelling. The finding could be
explained by the fact that TWVO students are exposed to more authentic input and, therefore,
have a greater vocabulary knowledge. This would be consistent with research of Verspoor et
al. which showed that CLIL students have a significant larger vocabulary size (61). It should
be noted that the high number of students focusing on spelling can be attributed to sentence 7
in the GJT (see Appendix B). The spelling of the word ‘per cent’ seized almost every
student’s attention. It was also found that 6VWO students paid much more attention to tense
in contrast to TVWO students. Once again, this can be explained by the approach towards
grammar teaching. Explicit grammar teaching causes mainstream students to focus explicitly
on tenses.
Overall, the findings of this study suggest that, although Dutch pupils find English
prepositions difficult, their knowledge improves significantly over the years. CLIL teaching
proved to be significantly effective in acquiring prepositions by Dutch pupils: already in the
first grade, CLIL students have a higher level of proficiency than mainstream education pupils
and they retain this lead until their final year. Although the results of this study proved to be
significant, caution must be applied because of the small sample size. In addition, it is unclear
to what extent course books teach prepositions to students. Although the high level of English
prepositions by CLIL students can be explained by these students acquiring a foreign
language implicitly as a result of the quantity and quality of English input they receive, it
could well be the case that, in addition to this, CLIL course books focus more on prepositions
than mainstream education course books in order to promote authentic language use by CLIL
51
students, in which case CLIL students would have a lead on their mainstream education peers.
A further study with more focus on the differences in content between CLIL course books and
mainstream education course books is therefore suggested. The results of this study, however,
do lend further support for the hypothesis that immersion in CLIL teaching plays a positive,
significant role in the foreign language acquisition process. Mainstream education teachers
should, therefore, attempt to adopt CLIL teaching practices into their teaching. Instead of
explicit grammar teaching, mainstream teachers should try to draw implicit attention to
grammar; in addition, they should speak the target language as much as possible, give
students the opportunity to produce the foreign language actively, and test communicative
competence rather than grammatical correctness as a goal in itself.
52
Chapter 8 – Conclusion
This thesis has investigated the immersion effect of CLIL teaching on acquiring English
prepositions. CLIL teaching integrates language learning and content learning, as the foreign
language is used as a tool for learning a content-based subject but is viewed as a goal of the
learning process itself at the same time. The present study was designed to determine whether
CLIL education is more effective than mainstream education in acquiring English
prepositions in the context of Dutch secondary school students learning English.
This study showed that the first-year as well as sixth-year CLIL students under
investigation have a significantly better command of English prepositions than their
mainstream peers. First of all, CLIL students proved to be significantly better at evaluating
the grammatical content of sentences in the Grammaticality Judgement Task than their
mainstream peers; in addition, results showed that first-year as well as sixth-year CLIL
students used significantly more correct prepositions in the Elicited Production Task. The
following conclusion can thus be drawn from the present study: CLIL teaching is significantly
more effective in acquiring prepositions by Dutch pupils than mainstream education.
The results of this study support the idea that implicit language learning occurs in the
CLIL classroom as a result of immersion. Although the first graders under investigation had
only received eight months of foreign language education, they already showed a significantly
better command of English prepositions than mainstream students. These findings suggest
that the quality and quantity of input first-year CLIL students receive facilitate the acquisition
process of English prepositions. Immersion assists students to acquire the language implicitly;
this implicit knowledge helps them to gain a lead on their mainstream peers in the first-year as
well as in the sixth-year. As argued by Krashen and Ellis, competence in a foreign language is
53
a matter of implicit language knowledge rather than learned knowledge (15; 95). In agreement
with this, De Graaff et al. have argued that fluency in the foreign language can best be
achieved by using the language as a “functional medium of communication and information”
instead of by “making it the object of analysis in class” (606) as is often done in the
mainstream education teaching practice. Because of the quantity of authentic and natural
input CLIL students receive, it could be suggested that their language use is slightly more
authentic and natural, as is reflected in the use of correct prepositions in their free writing.
Although CLIL students used fairly the same prepositions as mainstream students, they did
include more prepositions in their writing and mixed prepositions more in their writing, which
made it sound more native. CLIL teaching did, however, not lead to a higher use of
collocations or a greater variety of prepositions by CLIL students. As has been pointed out by
Verspoor and Edelenbos as well as Marsh, the positive learning outcomes of CLIL should not
be ascribed to the quantity and quality of input alone (12; “Adding Language,” par.6). CLIL
students also tend to benefit from the large quantity of opportunities to produce the foreign
language: for instance, they also produce English in courses other than English. In short,
CLIL renders positive learning outcomes: in their first year of higher education as well as in
their final year, TVWO students under investigation in this study retain a significant lead on
their mainstream peers with regards to their knowledge of English prepositions.
It must, however, be pointed out that the findings in this thesis are subject to at least
three limitations. First of all, an issue that was not tested in this study was the general
language proficiency of students. It might have been the case that the TVWO students under
investigation possessed greater levels of language aptitude in general, which could in turn
have explained their higher scores on the GJT and Elicited Production Task. What is more,
tests were administered in the 6VWO group on the penultimate day of school in the period
before the final exams. As a result, 6VWO students might have been less motivated to
54
participate in the study compared to the 6TVWO students, which might have influenced the
results. Lastly, since this thesis did not analyse the content of English course books, it is
unclear to what extent course books taught the students under investigation English
prepositions and whether this might have explained the higher level of command of English
prepositions by CLIL students.
Overall, the current findings add to a growing body of literature on the learning
outcomes of CLIL teaching. In addition, this thesis adds substantially to teachers’
understanding of language teaching. Taken together, this thesis has some implications for
teaching practices. In general, reading and writing are the skills most practiced in Dutch
secondary schools despite the fact that a second language is acquired mostly through oral
communication in informal contexts (Kwakernaak 29-30). CLIL teaching imitates the
conditions under which a second language is learned naturally and this has proved to be
significantly effective, and is in fact underscored by the results of this study. Although the
increased use of the target language as language of instruction in recent years tends to
promote oral skills, it is still widely believed that students learn grammar in order to
communicate outside the classroom (De Bot and Maljers in De Graaff and Tuin 135). In
short, the difference between CLIL and the mainstream education practice is too great. The
mainstream foreign language education teaching practice should learn from the CLIL
teaching practice. First of all, using the target language as a language of instruction should be
the rule rather than the exception. Students should be exposed to more authentic input and
should receive more opportunities to perform language output themselves in order to boost
implicit language learning. In addition, teachers should provide feedback on language
production but should not only be focused on accuracy; instead, they should focus more on
the communicative competence of students. When these suggestions are taken into
55
consideration, not only will the level of English increase by mainstream education students,
they will most likely become to enjoy the subject more (recall section 5.1).
This study has thrown up some questions in need for further investigation. Further
work needs to be done to establish how students’ knowledge of prepositions develops over the
years. The results of this study could suggest that CLIL students maintain a lead on their
mainstream education with regards to their command of English prepositions; however, it
would be informative to administer tests in at least two third-year classes as well in order to
examine this. It would hence also be interesting to examine whether the acquisition process of
prepositions perhaps levels out at the end of students’ lower years. It would, furthermore, be
interesting to investigate the difference between the best 10 per cent students in the VWO
classes with the weakest 10 per cent students in the TVWO class. It would also be interesting
to analyse the content of VWO as well as TVWO course books. What is more, it would be
interesting to investigate the role of individual learner characteristics, such as working
memory capacity and cognitive control in general, in the acquisition process of prepositions.
56
References
Admiraal, Westhoff, and De Bot. “Evaluation of Bilingual Secondary Education in
The Netherlands: Students’ Language Proficiency in English.” Educational
Research and Evaluation. 12.1 (2006): 75-93. Web.
“CLIL.” European Platform. Europeesplatform.nl, n.d. Web.
Coyle, Do. “Content and Language Integrated Learning: Motivating Learners and
Teachers.” Scottish Languages Review. 13 (2009): 1-18. Web.
De Bot, Kees and Anne Maljers. “De enige echte vernieuwing: tweetalig onderwijs.”
De Toekomst van het Talenonderwijs: Nodig? Anders? Beter? Eds. Rick de
Graaff and Dirk Tuin. Enschede/Utrecht: NaB-MVT/IVLOS, 2009. 131-145.
Print.
De Graaff, Rick, Gerrit Jan Koopman, Yulia Anikina, and Gerard Westhoff. “An
Observation Tool for Effective L2 Pedagogy in Content
and Language Integrated Learning (CLIL).” The International Journal of Bilingual
Education and Bilingualism. 10.5 (2007): 603-624. Web.
Ellis, Nick C. “Frequency Effects in Language Processing.” SSLA. 24 (2002): 143
188. Web.
Ellis, Nick C. “Optimizing the Input: Frequency and Sampling in Usage-Based and
Form- Focused Learning.” The Handbook of Language Teaching. Ed. Michael
H. Long and Catherine J. Doughty. West Sussex: Blackwell Publishing, 2009.
139-158. Web.
Ellis, Rod. “Grammar Teaching – Practice or Consciousness-Raising?” Methodology
in Language Teaching. Ed. Jack C. Richards. Cambridge: Cambridge
University Press, 2002. 167-174. Web.
Krashen, Stephen D. Principles and Practice in Second Language Acquisition. 1st
57
Internet Ed. Pergamon Press, 1982. Web.
Krashen, Stephen D. Introduction. Second Language Acquisition and Second
Language Learning. Pergamon Press, 1981. 1-11. Print.
Krashen, Stephen D. and Tracy T. Terrell. “The Natural Approach: Language
Acquisition in the Classroom.”Oxford: Pergamon, 1983.
Kwakernaak, Erik. Didactiek van het Vreemdetalenonderwijs. Bussum: Uitgeverij
Couthino, 2009. Print.
Lindstromberg, Seth. English Prepositions Explained. 1947. Amsterdam: John
Benjamins Publishing Company, 2010. Print.
Marsh, David. “Adding Language Without Taking Away.” The Guardian. 20 April
2005.Web.
Marsh, David. “CLIL/EMILE: The European Dimension. Actions, Trends, and
Foresight Potential.” Finland: University of Jyväskylä, 2002. Web.
McCarthy, Michael and Felicity O’Dell. English Collocations in Use. 2005.
Cambridge: Cambridge UP, 2008. Print.
Oller, Kimbrough D. and Rebecca C. Eilers. “Balancing Interpretations Regarding
Effects of Bilingualism: Empirical Outcomes and Theoretical Possibilities.”
Language and Literacy in Bilingual Children. Eds. Kimbrough D. Oller and
Rebecca C. Eilers. Clevedon: Multilingual Matters Ltd., 2002. 281-292. Print.
Richards, Jack C. and Theodore S. Rodgers. Approaches and Methods in Language
Teaching. New York: Cambridge University Press, 2001. Print.
Saville-Troike, Muriel. Introducing Second Language Acquisition. New York:
Cambridge University Press, 2006. Print.
“Standaard tweetalig onderwijs Engels havo/vwo.” European Platform.
Europeesplatform.nl, n.d. Web.
58
“Tto scholen.” European Platform (EP). Europeesplatform.nl, n.d. Web.
Verspoor, Marjolijn and Peter Edelenbos. “Tweetalig onderwijs zorgt voor een
duurzame voorsprong.” Levende Talen Tijdschrift. 12:4 (2011). Web.
Verspoor et al. “Tweetalig onderwijs: vormgeving en prestaties”. Europees Platform (EP).
Europeesplatform.nl, June 2012. Web.
Widodo, Handoyo Puji. “Approaches and Procedures for Teaching Grammar.”
English Teaching: Practice and Critique. 5.1 (2006): 122-141. Web.
59
Appendices
Appendix A: Questionnaire
Questionnaire
Naam:
man / vrouw
Leeftijd:
Klas:
tto / geen tto
1. Op een schaal van 1 (heel stom, moeilijk, helemaal niet) tot 5 (heel leuk,
makkelijk, heel veel)…..
1
2
3
4
Hoe leuk vind je het vak Engels?
Vind je het vak Engels moeilijk?
Hoe goed denk zelf dat je in het vak Engels bent?
Ben je geïnteresseerd in de Engelse taal en cultuur in het
algemeen? (Buiten het schoolvak om)
In welke mate kom je in aanraking met de Engelse taal buiten
school? (Denk aan TV, films, schrijven in het Engels, etc.).
Leg uit:
2. Wat is je gemiddelde cijfer voor Engels? (Indien bijvoorbeeld een 7,5: omcirkel dan
het streepje tussen 7 en 8)
1 – 2 – 3 – 4 – 5 – 6 – 7 – 8 – 9 – 10
3. Denk je dat leerlingen die tweetalig onderwijs volgen Engels op een hoger niveau
beheersen dan leerlingen die geen tweetalig onderwijs volgen?
a. Geen verschil
5
60
b. TTO hoger niveau
c. TTO veel hoger niveau
4. Heb je nog iets toe te voegen aan deze enquête? Wil je nog wat kwijt?
61
Appendix B: Grammaticality Judgement Task and Production Task
Is this sentence correct or incorrect?
In deze opdracht moet je de gegeven zinnen beoordelen op hun grammaticaliteit: zijn de
zinnen grammaticaal geformuleerd of niet? Je hoeft dus niet te letten op spelling en
punctuatie. Je hebt steeds 3 opties waaruit je kunt kiezen bij het beoordelen van de zinnen:
correct, incorrect, of ik weet het niet. Indien je denkt dat een (gedeelte van een) zin incorrect
is, onderstreep je dit gedeelte en verbeter je het. Meestal is je eerste antwoord goed, dus geef
na het lezen van iedere zin direct antwoord. Je krijgt 20 minuten voor de opdracht, maar
waarschijnlijk zul je niet zoveel tijd nodig hebben. Het is wel belangrijk dat je alleen werkt.
Natuurlijk krijg je hier geen cijfer voor.
Een voorbeeld:
She bought her dress at a bridal store.
Incorrect
Correct
Ik weet het niet
(She bought her dress) in a bridal store
1. The brown bears found on Kodiak Island are very large.
Incorrect
Correct
Ik weet het niet
2. George is leaving Monday at noon.
Incorrect
Correct
Ik weet het niet
3. Harry didn’t get any birthday presents yesterday.
Incorrect
Correct
Ik weet het niet
62
4. I stood in line for thirty minutes.
Incorrect
Correct
Ik weet het niet
5. I have never seen a dolphin in my life.
Incorrect
Correct
Ik weet het niet
6. The teacher asked us to comment to the story.
Incorrect
Correct
Ik weet het niet
7. Concert ticket prices have risen with 5 per cent.
Incorrect
Correct
Ik weet het niet
8. She is already a hour late.
Incorrect
Correct
Ik weet het niet
9. By six o’clock, I had finished my homework.
Incorrect
Correct
Ik weet het niet
10. They strolled over the beach watching the sunset.
Incorrect
Correct
Ik weet het niet
11. At the end, everything worked out.
Incorrect
Correct
Ik weet het niet
63
12. My mother is on a plane to Brussels.
Incorrect
Correct
Ik weet het niet
13. Lady Gaga is playing on the radio.
Incorrect
Correct
Ik weet het niet
14. The children were sitting behind the TV watching a show.
Incorrect
Correct
Ik weet het niet
15. I couldn’t stop laughing about him; he was such a fool.
Incorrect
Correct
Ik weet het niet
16. We spent over $200 on food this week.
Incorrect
Correct
Ik weet het niet
17. When did you gave up smoking?
Incorrect
Correct
Ik weet het niet
18. In our house, there is a large painting at the wall.
Incorrect
Correct
Ik weet het niet
19. My father shouted: “go into the kitchen immediately!”
Incorrect
Correct
Ik weet het niet
64
20. I speaked at my father’s retirement party last year.
Incorrect
Correct
Ik weet het niet
21. Before 1300, English was not seen as a prestigious language.
Incorrect
Correct
Ik weet het niet
22. At your left side, you can see the Big Ben.
Incorrect
Correct
Ik weet het niet
23. My aunt suffers from diabetes.
Incorrect
Correct
Ik weet het niet
24. As usual arrived James too late.
Incorrect
Correct
Ik weet het niet
25. Susan will arrive on the afternoon.
Incorrect
Correct
Ik weet het niet
26. Just wait a second; I’ll be there in a minute.
Incorrect
Correct
Ik weet het niet
27. I bought two pairs of jeans, because my old ones are too short.
Incorrect
Correct
Ik weet het niet
65
28. She apologised about the mistake.
Incorrect
Correct
Ik weet het niet
29. What are you doing over the weekend?
Incorrect
Correct
Ik weet het niet
30. Traditionally, people kiss each other at New Year’s Eve.
Incorrect
Correct
Ik weet het niet
31. I will be back by 8 o’clock.
Incorrect
Correct
Ik weet het niet
32. Because banks are closed at Saturdays, we couldn’t exchange money.
Incorrect
Correct
Ik weet het niet
33. Who are the people on this picture?
Incorrect
Correct
Ik weet het niet
34. Unfortunately, my train is delayed. I’ll see you over an hour.
Incorrect
Correct
Ik weet het niet
35. It was difficult to get on the train since there were so many people.
Incorrect
Correct
Ik weet het niet
66
36. Susan is engaged to Edgar.
Incorrect
Correct
Ik weet het niet
37. Apples are more healthy than chips.
Incorrect
Correct
Ik weet het niet
38. Unfortunately, that bag is too expensive. I can’t afford it.
Incorrect
Correct
Ik weet het niet
Correct
Ik weet het niet
39. I voted on Obama.
Incorrect
40. I’m comparing the book to the film.
Incorrect
Correct
Ik weet het niet
67
Appendix C: Elicited Production Task
Tell me about your last shopping experience
Je gaat nu in het Engels een stukje schrijven waarin je vertelt over de laatste keer dat je bent
gaan winkelen. Geef antwoord op de volgende vragen:
 Wanneer en waar ben je gaan winkelen?
 Met wie ben je gaan winkelen?
 Hoe ben je daar gekomen? (auto, fiets, bus)
 Wat kocht je? Waar kocht je dit? Waar lag deze winkel?
 Heb je wat gegeten? Waar?
Je hoeft niet in volledige zinnen de antwoorden; je mag ook deelzinnen opschrijven.
Een voorbeeld: Afgelopen weekend in Utrecht met m’n moeder. Met de trein. Een jurk bij
H&M gekocht. Aan de Oude Gracht. Bij de Douwe Egberts op het Neude koffie gedronken.
Download