No language barriers for neighbouring languages
Does a short teaching intervention change the attitudes and receptive skills of young Dutch and German speakers?
Aylin Bedir 3417506 MA in Multilingualism Faculty of Arts University of Groningen Supervisors: Dr Charlotte Gooskens Dr Joana Duarte 11.07.2018 Words: 12103
Table of Contents
Page
Abstract………1
1.0 Introduction………...2
2.0 Background………..3
2.1 Historical excurse in attitudes between the Netherlands and Germany………..3
2.2 English as lingua franca………...………...5
2.3 Receptive multilingualism as means of communication………..6
2.4 Mutual intelligibility………...………...………...8 3. 0 Statement of purpose………...………...………...10 4.0 Method………..11 4.1 Subjects………...………...………...11 4.2 Materials………...………...………...12 4.3 Procedure………...………...13
4.4 Design and Analyses………...………...16
5.0 Results………...………...………...17
5.1Attitudes………...………...………...17
5.2Word Translation Task………...………...………...19
5.3 Smurf Task………...………...………...21
6.0 Discussion………...………...………...22
6.1Attitudes………...………...………...……….22
6.2 Word Translation Task………...………...……….24
6.3Comparison to and with Gooskens et al. (2015): Mutual intelligibility of Dutch-German cognates by children: The devil is in the detail………...………...……….27
6.4 Smurf Task………...………...………...29
6.5 Returning to the Hypotheses……...………...………...30
6.6 A critical note on the conducted research………...31
6.7 Further research………...………...………...32
7.0 Conclusion………...………...………...32
Appendix A: Word Translation Task – Alternative Answers……….34
Appendix B: Word Translation Task - Results………...35
Appendix C: Test Material – Questionnaire in Dutch and German...39
Appendix E: Test Material – Smurf Task in Dutch and German………45 Appendix F: Teaching Intervention Material – Power Point in Dutch and German…………49 Appendix G: Teaching Intervention Material –Colouring picture in Dutch and German…57
Abstract
What do young Dutch and German speakers think of each other? Do they like each other’s language? Would they like to learn it? Could a teaching intervention change their attitudes and degree of understanding?
In this study, two primary school classes, one from the Netherlands and one from Germany participated. For the data collection, a pre and posttest design was used, in which the experimental groups received a teaching intervention in between in German or Dutch respectively which was not the case for the control groups. It aimed at enhancing the participants’ awareness of the similarities and proximity to the opposing language. During the pre- and posttests, the participants were questioned about their attitudes towards each other. Furthermore, they were tested on their receptive skills in the other language. The results showed that the attitudes of all participants were relatively positive from the beginning. The experimental group's results did not change remarkably after the teaching intervention whereas the German control group rated their perceived understanding of Dutch significantly higher in the posttest (z=-2,236, p<0,025). Overall the Dutch subjects were better at understanding German.
1. Introduction
“Tussen buurlanden is altijd wel wat“ (Engl. There is always something going on between neighbouring countries) states Janssen (as cited in Ház, 2005:51). This quote refers to the circumstance that the relations between Germany and the Netherlands have always been unsettled. Their shared past which consisted of both peace and war shaped their attitudes. Since their languages are closely related, they could easily communicate with each other by either acquiring the other language or by using a mode of communication where they keep speaking their language. However, English is often used instead. German is a school subject widely used in secondary education in the Netherlands whereas Dutch is not as easy to find in German school curricula. Regarding their language, although they are neighbouring countries, neither the Dutch language has derived from the German nor the German language from the Dutch. Two languages with their own cultures came into existence. The Germans have a more extensive territory which, however, does not give reasons to not treat both as equals (Ház, 2005:44). However, an imbalance between the two neighbouring countries is still a subject matter nowadays. Therefore, it is valuable to research if these attitudes are already acquired at an early age as primary school education. Moreover, it is interesting to see if such young learners can understand the language better when they were taught about the proximity between them.
In the current research, first, more insight into the attitudes of Dutchmen and Germans towards each other will be given by also referring to past event. Followed by that English as a lingua franca will be presented as well as receptive multilingualism as a mode for communication for closely related languages. Afterwards, the concept of mutual intelligibility will be explained. In the next section a detailed description of the methods used for this study will be given, followed by the results and discussion. Finally, a conclusion will be drawn.
2. Background
2.1 Historical excurse in attitudes between the Netherlands and Germany
“Vooroordelen afschafen is niet mogelijk; men zou tegen windmolens vechten” (Engl. “It is not possible to eliminate prejudices; men would fight against windmills” stated by Jansen (as cited in Ház, 2005:52)
Both the Netherlands and Germany have been through episodes of war and peace in the past and developed independently. This process also shaped the image that the nations have of themselves and their neighbour. Ház (2005:51) argues that it is normal to have prejudices against the neighbour, however in the case of Germany and the Netherlands they often do not help but rather hinder their communication. Since the languages are linguistically so close, methods such as receptive multilingualism would offer a great possibility. Receptive multilingualism entails that two people communicate with each other however each one speaks their language, and due to the similarity of the languages they will still be able to understand each other (Gooskens 2007: 445; Bergsma, Swarte, & Gooskens, 2014:45-46) Ház (2005:52) argues that the stereotypes and prejudices often get in the way of possible communication. An individual then only functions as a representation of a whole nation that one was taught to dislike. She highlights however that the situation nowadays is better than its reputation and has gradually improved since the end of the Second World War. The two neighbours cooperate more, and cultural exchanges are very frequent. However, the antipathy has been reoccurring. Ház (2005:53) states that regarding the attitudes there is still the asymmetry that Germans think rather positively of the Dutchmen whereas the Dutchmen regard the Germans less positively. She presents three reasons for this circumstance. Firstly, she mentions the difference in size of the countries. Inhabitants of smaller countries are more likely to look at the bigger one with a negative attitude and perhaps even a bit of dread. Since Germany is a much bigger country with more inhabitants, she argues that Dutchmen would trust a Belgian person much more because they are equals regarding the sizes of their countries. Secondly, she mentions the conditions that with more inhabitants the number of the language’s speaker is also greater and therefore more used, also when visiting the neighbouring country. According to her, Germans often assume that the Dutchmen can understand them which is considered as very impolite and arrogant by the Dutchmen who either switch to English or pretend not to know German at all. Thirdly she argues that the degree of knowledge of the neighbouring country is also a critical factor (Ház, 2005:53).
Dutchmen know more about Germans than the other way around. Because German is a subject in Dutch school curricula and through the availability of German TV shows and radio stations they are naturally more exposed to German (Gooskens, Van Bezooijen, & Van Heuven, 2015:256).
The Germans, however, know little about their neighbour. Aside from football championships or royal events they are not exposed to their neighbouring culture or language. Dutch is also not often a subject in school except for schools in the provinces Lower Saxony and North Rhine-Westphalia which are bordering federal states. The Netherlands is often connected to ideals such as tolerance or cosmopolitanism. The degree of knowledge Germans have of the Netherlands becomes visible when they naturally refer to the Netherlands as Holland. They are not aware that not the whole of the Netherlands is Holland but only the two provinces along the coast: North Holland and South Holland (Ház, 2005:53-54).
Moreover, also the attitude towards (the sound) of a language is mostly based on stereotypes that have been established over time and are accepted without critical thinking of the individual (Schüppert, Hilton, & Gooskens, 2015:376). Personal experiences can also form personal connotations for a particular person. This is not only true for Germans and Dutchmen. Schüppert et al. (2015) aimed to find out if “asymmetric” attitudes held by Scandinavians towards neighbouring, closely related Scandinavian languages are linked to “asymmetric” intelligibility of these neighbouring languages” (Schüppert et al.,2015:376-378). They found a low however significant correlation between the factors attitude and intelligibility in their word recognition experiment since the participants with a more positive attitude also performed better (Schüppert et al., 2015:399).
The condition that attitude is an important factor when it comes to language acquisition was also looked at in more in detail by Kuhlemeier, Van Den Bergh, & Melse (1996) study: “Attitudes and Achievements in the First Year of German Language Instruction in Dutch Secondary Education”. Looking at the younger generation, Kuhlemeier et al. (1996) focused in their study on German as a foreign language and the possible relationship “between students’ attitudes toward the subject of German, the course material, and the teacher and students’ achievements” (Kuhlemeier et al., 1996:498). Thus, they investigated the attitudes and achievements at the beginning and end of the first year of German in the curriculum for the participants. They stated that especially achievements made in the field of languages have a considerable influence on the language attitude and that the reversed effect is not as strong
(Kuhlemeier et al., 1996:495). Their results show that a positive attitude towards German had a small positive effect on the writing skills and spelling. A positive attitude towards the teacher also influenced the writing achievements significantly (Kuhlemeier et al., 1996:500). Furthermore, they also demonstrated that an existing positive attitude at the beginning of the school year was retained when they measured it again at the end (Kuhlemeier et al., 1996:500). Moreover, Baker (as cited in Zhao, 2015) states that since attitudes are learnt and not inherited, they are a subject of influence and can, therefore, be changed. Overall the students that had a German course which focused on communication rather than grammar acknowledged their course more (Kuhlemeier et al., 1996:505).
The determining factor attitude is also present in other neighbouring nations such as the Catalan/Spanish bilingual communities. Huguet & Llurda (2001:267) investigated the attitudes towards the other language which were characterised by their shared history. The study was conducted at 267 schools, and their participants were asked to fill in questionnaires with multiple choice questions as well as giving their opinion on statements over the Catalan and Spanish communities. The results showed that the subjects attending Catalan classes had a more positive attitude compared to those who did not. Getting involved with a language also means accepting elements of the other nation and can be a complicated matter. However, they concluded that more research is also needed in this field (Huguet & Llurda, 2001).
2.2 English as lingua franca
As previously mentioned English for German and Dutchmen often functions as a bridge for communication when they want to avoid speaking the other language. However, in a world where English is the lingua franca (common language for people with different native languages) for most people, neighbouring languages or any other language have a high chance of being neglected because of a more prominent language.
Blees, Mak, & ten Thije (2014) conducted a study in the context of the university on the use of English as a lingua franca compared to the use of lingua receptiva (another term for receptive multilingualism) between eight pairs of Dutch-German students. They asked half their participants to solve a maze puzzle in English and the other half to remain speaking in their mother tongues (either Dutch or German). Their motivation was the circumstance that the increasing use of English in academics has disadvantages. These can be an absence of linguistic diversity and therefore also cultural diversity since globalisation is emphasised. Their
participants found it much easier to solve the task in English than using receptive multilingualism. Additionally understanding each other and producing the own first language during the experiment was also more challenging. They offered alignment problems and processing costs in the brain for switching between languages as a possible explanation (Blees et al., 2014:188). Alignment problems are caused by the difficulty to “align […] so-called situation models: mental representations of the situations” Zwaan and Radvansky (as cited in Blees et all., 2014:177). The processing costs refer to the effort that bilinguals have to make in order to “suppress competing lexical items from the other language” (Blees et al., 2014:177). Consequently, in solving the task by speaking English, they were significantly more effective than in the other mode. However, it has to be noted that the participants were also more proficient in English than in the other language respectively and that the results were significantly influenced by this circumstance (Blees et al., 2014:188-89). This study shows that especially when German or Dutch speakers are more proficient in English, they tend to use that instead of using other modes of communication such as receptive multilingualism which would also be possible for them.
2.3 Receptive multilingualism as means of communication
For linguistically close languages such as German and Dutch, receptive multilingualism could be a well-chosen mode of communication since it does not expect the interlocutor to adjust to the other language and it also reduces the need for a lingua franca. Additionally, Gooskens et al. (2018) highlight that receptive multilingualism does not require prior instruction in the other languages.
Ház (2005) states that Kloss (as cited in Ház, 2005:39) has an impressive view on this topic as well. He made three statements regarding the acquisition of a closely related language. Firstly, he says, that it does not make sense to learn a closely related language with the same method as one that is entirely different from the speaker's mother tongue. It would suffice, e.g. for a German speaker only to learn these parts of grammar and vocabulary that are different to the language he or she wants to learn. Secondly, the aspiration to learn a language to its fullest, that means to know every word, is not necessary, it is already enough to know the language passively by training reading or listening comprehension. This is even more valuable when the other speaker of the closely related language is doing the same.
Thirdly, Kloss (as cited in Ház, 2005:39) highlights that it could be more profitable only to learn a closely related language passively. While the knowledge of the mother tongue can be used as a guideline when starting to learn the closely related language, it can change into a hindrance later on. When actively using the other language constructions, the mother tongue can interfere while trying to produce the other language.
Receptive multilingualism also has other benefits such as improving creativity and cognitive processes, especially problem solving and communicating. Naturally, the linguistic awareness is also higher, making it easier for them to acquire new languages since they have more experience and know about their best strategies (Gooskens & Heeringa, 2014:249). In this study, Gooskens & Heeringa (2014) wanted to prove that Norwegians “are better at understanding closely related varieties than Danes due to more experience with their dialects” (Gooskens & Heeringa, 2014:250). Since many dialects are still prominent and widely used in Norway, one could expect the Norwegians to have more experience and therefore perform better than their neighbours the Danes who barely have dialects left, and who only encounter regionally accepted speech (Gooskens & Heeringa, 2014:250). However, they did not find a significant difference between the performances. Since the experiment was auditory, they assume that multilingual subjects might have a more significant advantage at reading comprehension.
However, receptive multilingualism it is not used much by Germans and Dutchmen except for the border regions. Ribbert & ten Thije (2007:73) tried to answer the question as to why Dutchmen and Germans do not tend to use receptive multilingualism as a mode of interaction. Theirstudy investigated German-Dutch team cooperation. The participants had moderate knowledge of the other speaker’s language but still restrained from speaking it. According to Ribbert & ten Thije (2007:76) “[r]eceptive multilingualism is the equal mode of communication”. However, there is a strong tendency to speak the language with the higher status in economics and politics, which is also represented in the school curriculum (Ribbert & ten Thije, 2007:76). They highlight that traditionally there has been an asymmetry; as a Dutch person is more expected to know some basic German than vice versa. However, this attitude is changing as German is becoming less prominent among Dutch schools and the opposite is occurring in Germany since the demand for Dutch at German high schools and universities in increasing. In order to avoid the unbalanced situation where one speaker would
feel the urge to switch to the more dominant language, receptive multilingualism offers a solution.
2.4 Mutual intelligibility
According to Ház (2005) “the standard varieties of modern German and Dutch are mutually intelligible only to a rather limited degree” because German changed through the New High German Sound shift significantly.
Bergsma et al. (2014) have researched in the field of improving mutual intelligibility between closely related languages. Mutually intelligibility means that speakers of close varieties can understand each other without prior knowledge of the other language. They support the idea that English as a lingua franca is getting in the way because “only 38% of EU Citizens state that they have sufficient knowledge of English to have a conversation” European Commission, 2006 (as cited in Bergsma et al., 2014:45-46). Adding to that, only a small amount feels as if they can communicate in English on the same level as their native tongue. Therefore, they suggest the idea of using receptive multilingualism as a helpful tool. They highlight however that “communication problems are encountered using this strategy” (Bergsma et al., 2014: 46). The nature of the problems is said to be connected to “extra-linguistic factors such as language attitude, as well as linguistic factors, such as lexical, phonetic and orthographic distances” (Bergsma et al., 2014:46). The study aimed to teach Dutch people a way to understand the closely related language Frisian. They instructed them on phonetic and lexical differences which would help improve their intelligibility (Bergsma et al., 2014:47). Their teaching intervention which aimed at making them familiar with “frequent phonological correspondences” Bergsma et al. (2014:58) was not able to achieve a significant effect in a positive change for their intelligibility.
Similarly, Gooskens & Swarte (2017) state that the intelligibility is determined by the linguistic features languages share. In order to determine the lexical distance of two languages, they explain that it is a conventional method to “count the number of non-cognates (words that are not historically related) in the native language of the participant and the test language” (Gooskens & Swarte, 2017:131-32). Generally, the higher the number of cognates that two languages share, the better one can expect the mutual intelligibility to be. However, they highlight that this can also show an asymmetric relationship since “[a] word in language A may have a cognate in language B, but a synonym in B need not have a cognate
synonym in A” (Gooskens & Swarte, 2017:132). Since they also included attitude as a factor they wanted to see if there was a correlation between it and the overall intelligibility, which was not the case. They also found that the written form of a closely related language is much easier to comprehend than testing auditorily.
Another study by Golubović & Gooskens (2015) investigated mutual intelligibility between three West Slavic and three South Slavic languages. Within the aforementioned groups, receptive multilingualism is also a means of communication. They elaborate on different methods of measuring mutual intelligibility, distinguishing between opinion and functional testing. The former asks for self-evaluation on, e.g. level of understanding whereas the latter wants the participant to prove it. Functional testing includes sentence translation tasks which ask the participant to translate each word. The same procedure exists for word translation tasks which focus on the isolated word. Moreover, multiple choice questions are a useful method for data collection as well. Once the obstacle of creating the tests is negotiated, it can be a valuable tool (Golubović & Gooskens, 2015:355). For their purpose, they decided on a digital time limited word translation task where the participant had 10 seconds to translate each word. The test consisted of 50 words randomly chosen from a list of 100 words. They also used a written and spoken form of this test (Golubović & Gooskens, 2015:357).
On a larger scale, Gooskens et al. (2018) focused on investigating the cross-language intelligibility in three European language families (Germanic, Slavic and Romanic) in young, educated Europeans in a cloze test (Gooskens et al., 2018:170-71). In a cloze test, a participant has to fill in blanks in a text. They highlight that other studies have illustrated an asymmetry in mutual intelligibility.Explanations for this are the degree of exposure to the other language, the linguistic distance as well as an asymmetric relationship in attitudes as well (Gooskens et al., 2018:171). From their results, it can be seen that Germans can understand Dutch to a certain extent (25%) whereas the percentage is over 70% for the Dutch speakers when confronted with German. The difference in exposure could explain these results (Gooskens et al., 2018:179). They explain that it is common to use the mode of lingua receptiva in the border region. However, this communication will likely occur only on a fundamental level due to this asymmetry (Gooskens et al., 2018:180). Results that support this asymmetry were also found in Gooskens et al. (2015) study which “presented highly frequent Dutch and German cognate nouns […] to German and Dutch speakers in a word translation task” (Gooskens et al., 2015:255). They eliminated bias by removing extralinguistic factors such as exposure as much
as possible, thereby creating two highly comparable groups. The subjects showed similar positive attitudes towards the other language. The “results revealed that the Dutch subjects were significantly better at understanding the German cognates (50.2% correct translations) than the German subjects were at understanding the Dutch cognates (41.9%)” (Gooskens et al., 2015:255).
3.0 Statement of purpose
The research questions which will be answered in this thesis are:
1. Does a short teaching intervention change the attitudes and receptive skills to a closely related language?
2. Are speakers of Dutch better at understanding German than vice versa?
3. How do the results of this research compare to the study of Gooskens et al. (2015)? The corresponding hypotheses are:
Hypothesis 1: The intervention improves the attitudes and the receptive skills for the experimental groups.
Hypothesis 2: The Dutch speakers are better at understanding German.
4. Method
4.1 Subjects
The subjects were pupils from the primary school level in Germany and the Netherlands. In total 38 subjects participated in the study, 13 were Germans and 25 Dutchmen. The German and Dutch subjects were divided into two groups each. Group A functioned as the experimental group which received the teaching intervention in either German or Dutch respectively and Group B was the control group. The Dutch and German subjects were both in the second to last year of primary school education which however differs in the two countries. Therefore, the Dutch students were in grade 7 (which equals grade 5 in Germany) and the German students in grade 3 (which equals grade 5 in the Netherlands). The pupils were chosen so that they already gained the ability to read and write but were not at an age where they had had more chance to get exposed to the German or Dutch language through education. Both schools did not have the other language in their curriculum or after-school activities. Furthermore, by choosing an age group that would not have been exposed to German or Dutch extensively, they would also not have built strong positive or negative attitudes towards the other country. Naturally, every child’s upbringing influences this factor, but the older the child, the more solidified are these attitudes.
The German subjects did not know any dialect such as Plattdeutsch which might have helped them understand Dutch. However, it must be noted that 8 out of the 13 (61,5%) Dutch subjects in the experimental group and 8 out of the 12 (66,6%) subjects in the control group knew Gronings and therefore might have had an advantage understanding German since this Dutch variety is close to the German language. All the Dutch subjects had Dutch as their mother tongue except for one participant who stated that the second mother tongue was South African but since this language is relatively close to Dutch this circumstance was accepted, and the participant kept for the experiment. In the German experimental group was one subject with two mother tongues which were German and Lebanese. Similarly, in the German control group, one subject stated to have German and Albanese as mother tongue and another who said to have grown up with German and Turkish. These subjects were asked which language they would consider their dominant one. They went for German and were therefore also kept in the experiment.
The German subjects went to school in the small town of Syke in the Bundesland Niedersachsen, which is about 130 km from the Dutch border. The Dutch subjects went to
school in Onstwedde in the province of Groningen which is about 14km from the German border.I visited the schools for two days each during which I conducted the pretest on one day and the teaching intervention and posttest on another. The subjects were picked randomly from their class list but with the aim to keep the number of males and females as equal as possible. The children were divided into the experimental and the control group.
German subjects (N=17) Dutch subjects (N=25) Experimental group control group experimental group control group (N=7) (N=6) (N=13) (N=12)______________ Age range 8-9 8-10 10-11 10-12
mean 8.85 9 10.61 10.75
sex female 42.8% 66.6% 46.2% 41.7% male 57.2% 33.3% 53.8% 58.3% Table 1: Age and Sex of the participants
4.2 Materials
The participants received a questionnaire regarding their biographic data (sex, age), their exposure to German/Dutch by answering questions that asked for familiarity with the language and their mother tongue(s). Beneath that, they were asked to fill in a five-point differential scale concerning their attitudes towards German/Dutch culture, language and people. Furthermore, they were also asked about their self-rated ability to understand the language. In this case, 1 stood for “not good“ and 5 for “very good”. They were asked if they would like the idea of being taught German/Dutch at school. All these questions gave an impression of the subjects' attitudes and experience with the other nation.
After that followed a word translation task, for which the participants received a list of twenty words in the foreign language and were asked to translate them. These were taken with a previous agreement from Gooskens et al. (2015) study that was already elaborated on in the literature section. For this research, the 40 cognates were used and divided into two groups. This was done so that 20 they could be used for the pretest and 20 for the posttest. The word lists can be found in Appendix D.
During the listening comprehension test (called the Smurf task), each experimental and control group received a questionnaire with ten multiple choice questions to be answered
while watching the corresponding episode. These questions dealt with the content of the episodes by the Smurfs in either Dutch or German. The episodes were called Zuster Smurf/Schlumpf Schwester (Engl. Sister Smurf) and Baby Smurf/Das Schlumpf Baby (Engl. Baby Smurf). The questionnaires can be found in Appendix C.
The teaching intervention consisted of a PowerPoint presentation about basic sentences for introduction and colours in the foreign language. The slides can also be found in Appendix F.
This PowerPoint presentation followed a game for which a few songs were used (e.g. German: Hab ne Tante aus Marokko Dutch: Tante uit Marroko en die komt and German: Alle
meine Entchen Dutch: Alle Eendjes zwemmen in het water. Both authors are not mentioned
here since they are not known; the songs are said to be passed down orally.) For this game carton paper in the colours which were subject of the power point presentation were used as well. In the end, the participants got some colouring pencils (only the colours that we discussed) and a picture with a clown which was divided into small sections. Each section had the name of a colour in it, telling the participant which colour to use for that field. Next to the picture were the colours with their respective names depicted. This colouring picture was taken from a website for teaching materials and adjusted to Dutch for the German participants. The two versions can be found in the Appendix G.
4.3 Procedures
The pretests, posttests and teaching interventions were conducted on regular school days. Since the classes were divided in half, one being the experimental and the other the control group, I would take the group I needed out of their classroom and bring them to a previously organised free room with a Smartboard while the others stayed with their teacher. The children were asked to only take a pen with them. During all pre- and posttests all subjects had a separate table to themselves so that they could not cheat or be influenced by someone else’s answers to the questions. During the test phases all subjects stayed in their respective groups; either group A (experimental group) or B (control group). The other half stayed with the teacher so that the subjects did not receive a chance to interact with the other group during the tests. When the groups switched, one group had to first go out of the testing room, and the other group would be already waiting outside to come in. The teachers were very organised and supported this method. The share of the participants that did not belong to the group that was tested at that moment stayed with the teacher and worked on something
related to their current school lesson. These conditions were the same for the German and the Dutch class. If subjects finished earlier than the rest, their paper was collected, and they had to remain seated quietly.
Figure 1: Overview of the experimental design (top: the experimental group, bottom: control group without intervention) which was the same for the Dutchmen and the Germans
On the first of the two days, the pretest was conducted which consisted of the questionnaires regarding the participant’s attitudes and the word translation task. They were asked to fill them in on their own. The participants were encouraged to be creative and guess if they did not know the meaning of individual words.
For the Smurf task, the participants watched an episode in the foreign language but received the ten multiple choice questions in their native language. The questions were asked in the order in which the events occurred in the episodes to make it easier for the subjects and they were also allowed to answer them while watching the episode and had some extra time afterwards as well. Furthermore, the questions specifically focused on what was said so that they could not answer the questions correctly by just watching. Thus, the subjects had to pay attention and listen. An effort was made to make some questions a little more evident than other ones so that the subjects were not discouraged already at the beginning of the experiments. The Smurfs was chosen as it is a children’s TV series that is famous amongst children of the tested age groups and the specific episodes because they were available in both languages. The multiple-choice questions in both languages can be found in Appendix E. There was of course always the chance that the children would get the right answers by guessing. They were encouraged to listen carefully and work by themselves.
On the second day of the experiment (which was about one week after the conduction of the pretest), the experimental groups (either German or Dutch) received a 45-minute long
teaching intervention. The content of the lesson did not focus on the events in the episodes of the pre or posttest; neither did it teach the children words that included in the word list translation task. It only aimed to give the participants a first impression of the other language and make them aware of their proximity by showing them examples for the similarity Furthermore, it aimed at awakening the subjects’ curiosity by using a diverse structure of the lesson.
At first, the participants learned some introductory sentences. These sentences are quite similar in both languages, so this task was designed to be relatively easy for the subjects. The participants were at first asked if they knew some of the introductory sentences. After that, I read the sentences out loud, translated them into Dutch or German and let them repeat the foreign version. Individuals were also asked to do a full presentation (which included sentences stating their name, where they are from, how old they are and what they like to do in their free time. They received assistance when they, e.g. did not know a word. Secondly, I taught them the colours were. These included white, blue, red, yellow, green, brown, orange, pink and purple. Here, I used the same procedure and let them tell me what they knew first. After that, I told them the right phrases and let them repeat those. I addressed the subjects mostly in their foreign language, only for clarification or when presenting, e.g. the introductory sentences for the first time I spoke in their language.
The vocabulary for introduction and colours are very similar in both languages which made it easier for the children to understand and learn in such a short amount of time. The colours that were learned were also provided as carton papers which were used to do an activity with the children to make them memorise the colours with the correct pronunciation. During this activity, the children listened to the songs mentioned in the previous section. These were also carefully selected so that the participants could have heard them before in their language due to their popularity. The carton papers were lied out in a circle and stuck to the ground with tape. The subjects were asked to stand next to one colour and dance to the music however they preferred. Once I stopped the music, they had to jump on a colour carton and say it out loud in the foreign language. The power point presentation was still on the screen of the smartboard for the first few rounds so that the subjects could look at them. They were encouraged to form a full sentence saying “I stand on (colour)” (German: Ich stehe auf (Farbe), Dutch: Ik sta op (kleur)).
Lastly, the children received colour pens and a picture of a clown for them to colour. The picture was divided into small sections that had the name of a colour in them, telling the children in which colour to use for the section. An overview of the colours was provided on the side of the picture.
After the intervention followed the conduction of the posttest. This included all groups to watch the episode of the Smurf which they had not seen yet and answer multiple choice questions again. Next, they were asked to translate the twenty other words they had not translated last time into their language. After that, they were asked to fill in the same questionnaire asking about their attitudes towards the other language, culture and people. The pretest and posttest phases took about 30 Minutes per group per day. The teaching intervention was scheduled for 45 Minutes only for the experimental group of the German and Dutch classes.
4.4 Design and Analyses
The data was arranged in an excel file so that the results for each participant per group and per test were visible. This allowed a decent overview that made distinguishing the data easier. For the attitude questions (questionnaire) and the Smurfs task t-tests functioned as the statistical tests since it was mostly about group comparison.
I conducted paired sample t-tests for comparing the scores of the experimental group in their pretest and posttest phases. The same was done for the control group. After that independent sample t-tests were used to compare the means of the between groups (e.g. German experimental pretest and German control pretest) to see how the groups scored at the beginning and end of the experiment and if the teaching intervention that only the experimental group received caused a significant change.
Furthermore, independent sample t-tests were used to compare the means between the Dutch and German experiment/control groups in their pre-and posttest phases in order to see if the Germans or Dutchmen did better. An alpha level of 0,05 was chosen.
When after testing the normality the results were not normally distributed, I used the equivalent non-parametric tests, either the Wilcoxon signed rank instead of the paired sample t-test for within groups or the Mann-Whitney-U-test instead of the independent sample t-test. For the wordlist, all results were arranged manually. All answers of the participants were noted and checked. The answers were distinguished between right for responses which were correct or spelling mistakes which only differed in one letter without changing the meaning of
the expected word. This method was also applied when the “stimulus word has two meanings in the stimulus language with two corresponding cognate forms in the response language” (Gooskens et al., 2015) “For example, Du. zijde has two meanings, namely ‘side’ or ‘silk’. Du.
Zijde may, therefore, be responded to with either Ge. Seite ‘side’ or Ge. Seide ‘silk’.” (Gooskens
et al., 2015). For each correct response, the participants received one point. Missing responses or wrong responses, including plurals of words, were counted as incorrect and gave 0 points. Those words with an intelligibility difference of more than 20% between the Dutch and German groups were visualised in a Figure 1 to 4 in Appendix B. The individual words that had more than a 50% difference were discussed in more detail to give insight into the thought processes of the participants. The results were compared to those of Gooskens et al. study.
In order to find out if the teaching intervention affected the experimental group, more t-tests were conducted in the same manner as for the questionnaire and the Smurfs – Task.
5.0 Results 5.1 Attitudes
Neither for the German nor the Dutch groups, significant effects could be found that would have proven a change in attitude towards the other people, language or the sound of the other language.
Table 2: Means for the three attitude question per group and per phase. The Likert scale went from 1 (very bad) to 5 (very good).
Attitude to people Attitude to language Attitude to the sound of the other language
Pretest Posttest Pretest Posttest Pretest Posttest
German experimental group A 3,71 3,71 3,42 3,57 3,28 3,57 Dutch experimental group A 3,92 3,76 3,92 3,76 3,53 3,46 German control group B 3,33 3,66 3,16 4,33 3,16 3,16 Dutch control group B 3,66 3,25 3,33 3,66 3,16 3,25
All groups were either relatively positive from the beginning or just went for the middle of the scale. The only greater difference that could be found was in the German control group which showed a much more positive attitude towards the Dutch language in the posttest compared to their pretest results.
The questions regarding their perceived understanding of the other language, however, showed some noticeable results on the German side. The experimental group that received the teaching intervention stated in the pretest that they had no knowledge or understanding of the Dutch language; in the posttest phase, they rated their understanding higher. Naturally, they also rated the number of words higher in the posttest compared to the pretest.
Furthermore, the German control group also rated their perceived understanding of the language remarkably higher. This effect was significant (z= -2,236, p<0,025).
The Dutch subjects had high means for perceived understanding and estimation of their knowledge of German words in both the pretest and posttest in both groups.
The percentages of the people per group who would like to learn the other language in school also did not change noticeably. For the German experimental group and control group, the values stayed precisely the same for the pretest and posttest. This was different for the Dutch experimental group whose percentage dropped by around 8% from 100% to 92,31%. Similarly, the Dutch control group’s percentage also decreased by about 8% from 75% to 66,67%.
Table 3: Means for the three questions regarding the participants understanding of the foreign language, knowledge of words and opinion on wanting the foreign language as a subject in school
5.2 Word Translation Task
T-tests were conducted to compare individual groups. A significant difference was found for the German experimental group which did worse in the posttest compared to their pretest (t=3,35, df=6, p<0.015).
The German control group however improved from their pretest to the posttest (t=-5,937, df=5, p<0,002). This effect could not be found for the Dutch groups; both tests did not reach significance.
Moreover, there was a significant difference in the pretest and posttest scores of the German experimental group compared to those of the Dutch experimental group. The Dutch group scored significantly higher (U=20,5, p<0,046) in the pretest and the posttest (U=11, p<0,006).
This effect could also be found for the control groups. The Dutch control group scored significantly higher in the pretest phase (U=0, p<0.001) and the posttest (U=4, p<0.002)
Perceived understanding of language
Perceived: how many words
Wish to learn other language in school [%]
Pretest Posttest Pretest Posttest Pretest Posttest
German experimental group A 1,28 1,85 1 2 71,43% 71,43% Dutch experiment group A 3,46 3,23 2,69 3,07 100% 92,31% German control group B 1,33 2,16 1,33 1,83 50% 50% Dutch control group B 3,58 3,50 3,08 3,33 75,00% 66,67%
compared to the German control group. The results are also represented in Table 4 below with their descriptive data.
Wordlist Test Mean Standard
Deviation
Median Minimum Maximum
German Exp. Group Pretest 7.08 1.74 6.75 5.0 10.0 Posttest 5.16 1.6 4.5 4.0 8.0 Dutch Exp. Group Pretest 9.25 2.44 9.5 5.0 12.0 Posttest 9.33 2.89 10.0 4.0 13.0 German Control Group Pretest 4.3 3.7 5.0 2.0 6.0 Posttest 7.1 2.19 8.0 4.5 9.0 Dutch Control Group Pretest 9.59 1.65 10.0 7.0 11.5 Posttest 10.18 1.48 10.0 8.0 12.5
Table 4: Descriptive data of all Wordlist tests
There were also three words that neither the Dutch nor the German subjects could translate correctly at all in any of the groups or phases. These were (summarised from all groups across tests): time (Du. Tijd, Ge. Zeit), part (Du. Deel, Ge. Teil) and heart (Du.hart, Ge. Herz).
The alternative answers of the individual groups can be found in Appendix A and will also be subjects of discussion in the following section.
There were also cognates which were better understood by one subject group compared to the other. These with a difference of at least 20% are presented in the Figure 1 till 4 in Appendix B. For this comparison, the experimental and control groups were put together. In the pretest phase, the Dutchmen knew 16 cognates better compared than the Germans. They did not know any cognates better in that phase. However, in the posttest, the Dutchmen knew 14 cognates at least 20% better than the Germans whereas the latter knew three better than the Dutchmen. When looking at the difference exceeding 50%, the Dutchmen did 4 better than the Germans who scored 0 higher than 50%. In the posttest the Dutchmen knew 9 cognates better than the Germans, who again scored 0 when using this measurement. Since the number of cognates exceeding 20% in intelligibility was so high, only the cognates exceeding 50% in difference will be the subject of discussion in the following section. The three cognates better understood by the Germans did not exceed 50% but will
be discussed in the following section to give possible explanations as to why they scored higher than the Dutchmen for these three.
5.3 Smurf – Task
Due to mistakes in the conduction, the Smurf Task could only be evaluated for the between groups and not the individual evaluation within the groups. None of the tests for the Dutch or German within groups reached statistical significance which means that there was no noticeable difference between each group pair (Dutch experimental/Dutch control and German experimental/German control). Likewise, none of the tests for comparing the German pretest and posttests groups with the corresponding Dutch group reached remarkable results either (e.g. German experimental pretest/Dutch experimental pretest). The German experimental group did worse in the posttest than in the pretest whereas this effect was reversed for the German control group. The results of the Dutch subjects in the experimental and control groups overall stayed between 7,09 and 8,5 out of 10 for both the pre and the posttests.
However, as can be seen from the descriptive data the asymmetric trend is reoccurring.
Smurf Task Test Mean Standard
Deviation
Median Minimum Maximum
German Exp. Group Pretest 7.33 2.73 8.5 2.0 9.0 Posttest 6.50 3.08 7.5 1.0 9.0 Dutch Exp. Group Pretest 8.33 0.78 8.5 7.0 9.0 Posttest 8.5 1.08 8.0 7.0 10.0 German Control Group Pretest 6.20 2.58 7.0 2.0 9.0 Posttest 8.0 1.58 8.0 6.0 10.0 Dutch Control Group Pretest 7.72 1.95 8.0 4.0 10.0 Posttest 7.09 1.51 7.0 4.0 9.0
Table 5: Descpritive data of the Smurf Task
Both the Dutch experimental and control group scored higher in the pretest compared to the corresponding German groups. Similarly, this can be found in the posttest of the German and
Dutch experimental group. However, the German control group did a little better in the posttest than the Dutch control group.
Table 6: Summary of relevant results
6.0 Discussion
Note: The sometimes unexpected results may be explained by the fact that the experimental and control groups did not receive the same tasks in the same phases. A crossed design was used.
6.1 Attitudes
The German control group showed a noticeable more positive attitude towards the Dutch language in the posttest although they did not receive the Dutch lesson. This result is all the
Attitudes Wordlist Smurf Task
Dutch and German groups were relatively positive from the beginning
German control group showed the more positive attitude towards the language in posttest (no significance) Significant effect or rather trend could
be found for the perceived understanding of the German control
group (N=6) rated it considerably higher
Almost all participants were more than enthusiastic about learning the other
language at school (over 50%)
German experimental group did significantly worse in the posttest
compared to their pretest Effect reversed for the German control
group → German subjects seemed to find Wordlist B more difficult Dutch experimental group scored significantly higher than the German experimental group in the pretest and
the posttest Dutch control group scored significantly higher in the pretest phase and the posttest compared to
the German control group Cognates exceeding 20% of the difference for the pre and posttest (put together) of the experimental groups was 12 to 2 and for the control
groups: 18 to 1, both in favour of the Dutch speakers
No remarkable results for the within groups tests, as well as comparing German pre- and posttest groups with
the respective Dutch groups German experimental groups did
worse in the posttest than in the pretest whereas this effect was
reversed for the control group Dutch subjects in the experimental and
control groups overall stayed between 7,09 and 8,5 out of 10 for both the pre and the posttests compared to the
Germans with 6,2 and 8,0 out of 10
German and Dutch subjects got on average at least half of the questions
right
Shows trend of asymmetric relationship between Dutch and German subjects however extralinguistic factors such as exposure and previous knowledge of Gronings on the part of the Dutch subjects have to be considered
more surprising since none of the participants has been to the Netherlands in both the experimental and the control group, nor do they have family in the Netherlands. So it seems that what they learned during the pretest phase must have made them more open and confident. The circumstance that the German experimental rated their understanding and the number of words they know higher in the posttest could be explained by the fact that many participants did not know the language before, some of them had never heard the Dutch language. Thus, they would rate their understanding very low in the beginning. Since all participants stated that they had never been to the Netherlands or have family there, this rating seems to reflect that after the teaching intervention they had a better understanding of the Dutch language and could better evaluate their perceptions.
Similarly, the German control group also rated their perceived understanding of Dutch higher in the posttest, although they did not receive a teaching intervention. However, similar to their peers in the experimental group, they had no or very little knowledge of the Dutch language and supposedly learned about the familiarity of the languages through the wordlist they filled in and the episodes of the Smurfs that they watched.
Looking at the Dutch subjects, they had high means for rating their perceived understanding and knowledge of words in both phases of the testing which could be traced back to the fact that all of the Dutch participants had been to Germany at least once and therefore had had contact with the German language before.
The percentage of subjects per group who would like to learn the other language in school also did not change noticeably. For the German experimental group and control group, the values stayed precisely the same for the pretest and posttest. The German experimental group’s percentage did not rise after the teaching intervention.
This was different for the Dutch experimental group whose percentage dropped by around 8% from 100% to 92,31%. Similarly, the Dutch control group’s percentage also decreased by about 8% from 75% to 66,67%.
The fact that the percentages for wanting to learn the language in school dropped slightly for both Dutch groups was also not expected. However, compared to the German groups their percentages were much higher, to begin with so it can be concluded that even after some intense contact with the language, the motivation to learn it was still high.
6.2 Word Translation Task
The reader should be reminded here that the conduction method of Gooskens at al. (2015) study and this one differ in that that the words were not read out loud to the participants. However, instead, they had to read the words themselves quietly and use their knowledge or ideas of how the words could look or sound like.
The unexpected effect that the German control group improved from their pretest to the posttest compared to the opposite effect in the experimental group might indicate that the German subjects found the selection of words (Wordlist B in Appendix D) much harder than the other one (Wordlist A in Appendix D). This is visible since both groups did worse in Wordlist B (which the experimental group received in the pretest and the control group in the posttest).This wordlist might have caused the subjects more confusion and problems than the other.
In the following, the three cognates exceeding 20% in intelligibility in favour of the Germans will be discussed followed by the cognates exceeding 50% in favour of the Dutch speakers. The focus was put on these instead of the ones exceeding 20% because elaborating on them would have taken too much room. Since there were no cognates in favour of the Germans for the 50% limit, I wanted to discuss at least the three they could translate better than the Dutchmen.
The pronunciation is only put down here to give indications on how the participants might have read the word aloud (or in their head). The assumptions were made based on the answers of the participants.
Mal – maal (Engl. meal): This cognate is spelt differently in both languages but pronounced the same /ma:l/. It should have been relatively easy for both the Germans and the Dutch to decode this one but was not the case for the Dutchmen. They presented alternative answers such as mail (Engl. mail), meel (Engl. flour) or winkel (Engl. shop). These indicate that they did not know about the similarity of the words as the German word is also not used very often nowadays.
Tag – dag (Engl. day): The Dutch speakers might have read the word as /taɣ/ however their answers indicate that they knew that in the German the g is not pronounced as <ɣ> but as <k> at the end of a word. Therefore, they came up with words such as tak (Engl. branch) or tafel (Engl. table).
Buch – boek (Engl. book): In this case, the Dutch might have read the German word /bu:x/ as /buk/ or did not know how to read that word at all. Their answers indicate that they did not know that the German <u> is the Dutch <oe> because they gave answers such as buik (Engl. stomach) or brug (Engl. bridge).
In the following the cognates exceeding 50% in the difference in favour of the Dutch speakers will be discussed:
Frau – vrouw (Engl. woman): The Dutch subjects did very likely do better on this (84,6% correct answers in the pretest and 83,3% in the posttest) one due to their exposure to German during their vacation. The word Frau /frʌu/ is frequently heard when someone, for example, addressed their mother. Also, the grapheme <au> exists in Dutch and is similar to <auw> or <ouw> whereas the German language does not have a grapheme similar to <ou>. They would just have read it as /vrɔʊv/ which could explain the creative answers such as Wurst (Engl. sausage) or Hund (Engl. dog).
Grund – grond (Engl. ground): The Dutch subjects most likely read the cognate as /ɣrʏnt/ and saw the similarity to the Dutch word whereas the Germans would have understood it as /grɔnt/ which of course resembles the word Grund but maybe just not came to their minds. A possible explanation is that they were not aware that some words in German and Dutch share the same meaning and are also spelt the same.
Ende – einde (Engl. end): /ɛndə/ This cognate also seems to be easier to decipher for the Dutch subjects than for the German. They most likely thought of the synonym /ɛnt/ (Engl. end). The Germans, however, were misled and read it as /aɪndə/ resulting in them thinking of words starting with /aɪ/ e.g. eine (Engl. one).
Haus – huis (Engl. house): /hʌus/ In this case, similarly to the previously mentioned cognate
Frau (Engl. woman), the Dutch subjects most likely did better due to their familiarity with
words consisting of sounds similar to <au>. The German subjects read it as /huːiːs/ or /hu:ɪs/. There is no equivalent in German for /ui:/ or /u:ɪ/. Therefore, they just went for any words starting with a H.
Jahr – jaar (Engl. year): /jɑr/ In the Dutch control group the subjects most likely read the German word Jahr with a short /ɑ/ but were still able to connect it to the Dutch word /ja:r/ with a long /a/. The German subjects, however, seem to have read the Dutch cognate as /jɐ̯/ since for many German words the /r/ at the end of words is not pronounced but changed it to /ɐ̯/. Therefore, the German subjects went with ja.
Bad – bad (Engl. bath): Here the words are spelt the same in both languages. However, they are pronounced a little differently. Dutch people would read /bɑt/ whereas German would go for a longer /ba:t/. It seems that the German subjects were just confused, and many did not think that the words could be identical in both languages.
Stadt – stad (Engl. city): The Dutch subjects seem to have read the German cognate /stɑt/ just the way it is in Dutch and therefore recognised the word whereas the German subjects did not see the resemblance and thought of other options, e.g. Start (Engl. Start) or Schild (Engl. Sign).
Mensch – mens (Engl. human): For this cognate, the Dutch subjects had an advantage as they could have read the German word Mensch old Dutch spelling as /mɛns/ which is almost the same to the cognate in their language. However, the Germans most likely read it as /mɛns/ which makes the reversed connection much harder.
Stück – stuk (Engl. piece): This one was more challenging for the Dutch subjects on the one hand since there is no such as ü letter in Dutch. They presumably read it as /stʏk/ which on the other hand could have helped them think of the right word. The Germans would have read it as /stu:k/ or / ʃtʊk/, making them think for the former of similar words, e.g. Stuhl (Engl. chair) or for the latter they thought of Stock (Engl. branch) since they went from /ʊ/ to /ɔ/.
The three cognate pairs that none of the Germans nor the Dutch subjects had correct:
The cognate pair Zeit – tijd (Engl. time) posed problems for the German and Dutch subjects alike. Comparatively, to (Gooskens et al., 2015) study in which the Dutch subjects did considerably better, the Dutch subjects of this study did not read the German word as /tsait/ but /zɛit/. Therefore, they did not consider the sound /ts/ at all (which in combination with /
ɛi/ also does not exist in the Dutch language (Gooskens et al., 2015)) and went for words starting with /z/ such as zijn or zout. The Germans read the Dutch word supposedly as /tɪt/ or /ti:jt/ which caused them to look for words starting with Ti. Therefore, they went for e.g. Tisch (Engl. table). However, it seems that some of them also guessed that <ij> is pronounced /ɛi/ and thought of Teig (Engl. dough).
Teil – deel (Engl. part): This cognate has a false friend in Dutch since the German word Teil (Engl. part) also exists in Dutch. However, for this cognate pair deel is correct since Dutch /tɛil/ means Wanne in German (Engl.: tub). The German subjects read the cognate as /deːl/ and thought of similar words in German starting with a <d>, e.g. Diele (Engl. hall) or Delle (Engl.dent).
Herz – hart (Engl. heart): The Dutch subjects supposedly read the cognate as /hɛrz/. They went for words which are similar such as hert (Engl. deer) and did not think of /a/ when reading /ɛ/. On the German side, this cognate caused problems since the word /hart/ exists in German as well. However, it means “hard” and not “heart”. Thus, most of them went for the wrong word. If this test were conducted auditorily, the word would have been right as they sound the same.
Overall, in the current study many times the Germans did not seem to know that some words have the same meaning and spelling in Dutch and German or at least that there is the possibility of opting for the correct answer just by copying the given word. Since both the German and Dutch subjects barely had English as a subject for about a year they might not have had the awareness yet that many words resemble each other or are even translated the same in multiple languages. Therefore, they did not always consider writing down a German word that looks the same as the Dutch one or vice versa.
6.3 Comparison with Gooskens at al. (2015) study: Mutual intelligibility of Dutch-German cognates by children: The devil is in the detail
The results of this study show clearly that the Dutch group did better in all wordlist tests and therefore might understand German words better than German pupils understand Dutch words. This asymmetric relationship is very similar to the results of Gooskens et al. (2015) study. However, they were able to eliminate all extralinguistic factors such as language
experience which can have a significant influence on the subjects’ ability to understand the language. The effects in my study could have occurred because the Dutch subjects had had more exposure to German in the past than vice versa. It should also be mentioned, that in Gooskens et al. (2015) study, the Germans and Dutchmen both went to school rather far away from the borders whereas in this study the Dutchmen’s school was very close to the border compared to the German school.
The two studies can only be compared to a certain extent since the conduction methods differ. However, it is interesting to note that some results overlap. The German subjects in their study also did better. However, they did not have cognates in their study, that neither the Dutch nor the Germans could decode. In their pilot study, they removed the three cognates, Du. kerk Ge. Kirche (Engl. church), Du. zijde Ge. Seite or Seide (Engl. silk or page) and Du. Heer Ge. Herr (Engl. mister) because the amount of intra-language incorrect responses given by the subjects was too diverse (Gooskens et al. 2015:263). In the current study the cognate Du. zijde Ge. Seite/Seide caused problems for both the German and Dutch subjects.
When looking at their cognates that were better understood by the Dutch subjects and exceeded 20% in intelligibility, there were a few ones that overlapped with the current research. The cognates Engl. end and Engl. ground seemed to have been easier for the Dutch speakers both auditorily (Gooskens et al., 2015) and regarding their reading comprehension (current study). Another accordance could be found for one cognate that was better understood by the German subjects in both studies: Engl. book. This result is very interesting as this cognate was easier to translate both when it was heard and read. However, when they hear the word /buk/ it is much closer to the German equivalent as to when it is read /bɔek/. Supposedly the German subjects read it also as /bøːk/ since German words with an umlaut are often spelt with the vowel (without the dots) and then with an additional -e (e.g. oe would be ö, and ue would be ü).
Moreover, the three cognates that none of the subjects in the current study could decode did not seem to be that problematic for the participants in Gooskens et al. (2015) study. The cognate Engl. time was much better understood in their study (over 80%). The cognate Engl. part was slightly better understood by the Dutch subjects and the cognate Eng. heart a little better by the German subjects.
The vast difference in the number of cognates that were understood better or understood at all indicates that the Dutch and German cognates are easier to understand when they were heard compared to when they were read.
6.4 Smurf Task
It has to be mentioned that only the data for the Smurfs task was not carried out appropriately which means that there was no indication made as to who was who in the pre- and the posttest. However, it is known which participant was in which group, therefore the data can still be used, but the within groups tests data were treated as independent data. Naturally, this must not happen, and the subjects should be numbered in order to be able to link them to the data.
The German and Dutch subjects did well on the Smurfs tests. In both the pre and posttests their means indicate that they always got more than half of the questions right. The results show that both Germans groups found the multiple-choice questionnaire for the episode of the Babysmurf a little harder than the Dutch groups. Everyone received the same questions only in their language respectively. Therefore, this trend can only be explained again by the higher degree of exposure to German for the Dutch subjects or a more familiar theme in the episode. Although there was no significant improvement for the groups, it is still noticeable to see that the degree of understanding the other language seems to be moderate to high.
Looking at the individual questions, it is noticeable that many students on the German as well as on the Dutch side had difficulties especially answering question 9 and 10 for both episodes. They were asking for specific words or phrases in the other language which were only used once in the episodes. For question nine of the episode “Zuster Smurf/Schwester Schlumpf” (Engl. Sister Smurf) the participants were asked what the woman gave Smurfette and her friend to eat. However, their plates were already empty, and so the participants had to understand that the woman was telling them that she would give them more bread and water if they wanted. Similarly, for question ten the participants were asked what the girl calls her two brothers. She only says it once in the episode. For the Baby Smurf episode, the participants were asked for question nine when the bomb is supposed to explode. They could tell that from watching alone since Baby Smurf uses an alarm clock to set the alarm for seven
o clock, however, a little later, he also says that the sun is going to set soon. Thus, if the participants did not understand that he was talking about setting the alarm for the bomb previously, they could assume that it is going to explode in the middle of the night since the Baby Smurf is waiting for the sun to go down. The tenth question asked the participants what the Smurf in the barrel tells the other Smurfs. He was previously introduced as the Smurf that is always tired and sleepy. Therefore, the participants could assume that he tells them how he is feeling. However, he was able to watch what the baby Smurf was planning and could, therefore, warn them about the bomb. For this question, the participants had to listen carefully again. This kind of questions might have caused problems to the subjects since they had to rely on their listening comprehension entirely. This outcome can be explained due to supposedly more complicated vocabulary particularly in the “Baby Smurf” episode. It was tried to keep the questions as simple as possible without giving to much away. It would have been beneficial to have run a pilot study to eliminate these questions and to avoid having material that does not have the same difficulty level although this effect was only noticeable for the German subjects.
6.5 Returning to the Hypotheses
The first hypothesis: There will be a positive difference in attitudes, and the receptive skills for the experimental groups could only be partly confirmed. The Dutch and German groups were somewhat positive from the beginning. Although not significant, the result that the German control group showed a more positive attitude towards the language in the posttest should be mentioned. Similarly, the same group rated their perceived understanding of Dutch significantly higher. However, the experimental group did not achieve such results. It was rather the opposite since they did worse in their posttest for the wordlist as well as the Smurf task. The Dutch experimental group also did not improve remarkably. However, they already had good results in their pretest phase. Due to the mistake regarding the test material for each group, it cannot be concluded that the experimental group had a learning effect.
The second hypothesis: The Dutch speakers are better at understanding German could be confirmed since their experimental group scored significantly higher than the German experimental group in the pretest and in the posttest of the wordlist task. This was also true for the Dutch control group compared to the German control group. The cognates exceeding
