Transfer in Second Language Acquisition - Turkish Students speaking English

Seminar Paper 2008 43 Pages

English - Pedagogy, Didactics, Literature Studies


Table of Content

1. Abstract

2. Introduction
2.1 Introduction and theoretical background
2.2. Hypothesis

3. Study
3.1. Participants/ Procedure
3.2. Materials

4. Results
4.1. Results referring SVO-constructions
4.2. Results referring the usage of auxiliaries
4.3. Results referring negations
4.4. Results referring the filler items for German-Turks
4.5. Results referring the filler items for Turkish-Turks

5. Discussion
5.1. Discussion of the results referring SVO-constructions
5.2. Discussion of the results referring the usage of auxiliaries
5.3. Discussion of the results referring negations
5.4. Discussion of the results referring the filler items
5.5. Other observations

6. Conclusion

7. Appendices

8. References

1. Abstract

This study deals with the issue of language transfer. Thus, we compared the performance of English by two different groups. 20 students participated in the study, 10 live in Germany and 10 in Turkey. All 20 are L1 speakers of Turkish. The students from Turkey are L2 speakers of English. The other 10 students are L2 speakers of German with English as their L3. The participants translated sentences from Turkish into English. We examined whether the participants from Germany outperformed students from Turkey. If language transfer took place students from Germany should indeed perform better. Our assumption is based on their knowledge of German, which is similar to English in some aspects. Performance relating to word order, auxiliaries and negations were appraised. However, the results show that our assumption cannot be affirmed. Students from Turkey were as good as or even better than the students from Germany in all three categories. Nevertheless, we made observations supporting the idea of language transfer. For instance, students from Germany were definitely better in formulating relative clauses, which do not occur in Turkish. In summary, there are no conclusive results to confirm the existence of language transfer.

2. Introduction

2.1. Introduction and theoretical background

We concerned ourselves with theories and studies that dealt with the issue of language transfer. Afterwards we felt disposed to conduct this study on language transfer. Both the study by Bohnacker (2006) and the study by Hakkonson, Gisela, Pienemann, Manfred & Sayehli, Susan (2002) were decisive factors for the design of this study. Hakkonson et al. examined cross linguistic influences in their study “Transfer and typological proximity in the context of second language processing”. The results showed that Swedish learners of German do not transfer the syntactic feature V2 to German from their L1. On account of the fact that the participants did not start off with V2 structures in German, Hakkonson, et al. assume that no transfer took place. The acquisition of German developed along certain stages instead. The developmental stages are believed to be predicted by the processability hierarchy according to the processability Theory (Pienemann, 1998). However, the study was criticised because the participants were L2 speakers of English and L3 speakers of German. English is a non-V2 language. Thus a transfer from English to German could not be denied. However, Hakkonson et al. refuse this option termed “English illness” in their study.

Four years later Bohnacker (2006) conducted another study “When Swedes begin to learn German”. She took “English illness” into consideration, since she divided the participants into two groups. On the one hand L1 speakers of Swedish and L2 of German and on the other hand L1 speakers of Swedish, L2 speakers of English and L3 speakers of German. The results showed that participants do transfer the V2-syntax from their L1. If, however, their L2 is a non-V2 language like English, the transfer might be obscured.

The current study is similar to the study by Bohnacker (2006). It is structured as follows; in section 2 we will outline theoretical approaches to language transfer and introduce the hypothesis of the current study. Section 3 deals with the survey in general comprising procedure, method, participants and linguistic analysis of exemplary sentences from the questionnaire. In the next section the results will be presented. The interpretation of the results in regard to the hypothesis is the subject of section 5. In the following we refer to possible rationales behind the results. In addition, we give suggestions for improvements in efficiency of this study. Section 7 contains further information about both the questionnaires and the participants.

2.2. Hypothesis

We were especially enthused by the above-mentioned studies conducted by Bohnacker (2006) and Hakkonson, et al. (2002). The inspiration for this study is based on own experiences as well. On previous vacations we spent time with our relatives in Turkey. We noticed that their English was often inaccurate. It seemed that they made word-by-word translations from Turkish into English. Relating to the acquisition of English, our knowledge of German is probably of benefit to us. As German and English share regularities in terms of vocabulary, word-order etc. speakers of German might enjoy the advantage of language transfer. Yet, these thoughts are not grounded on substantiated data but on assumptions. On account of previous research as well as own experiences, we formulated our hypothesis as the following:

- L3 speakers of English perform better in English if they are L2 speakers of German.

In order to lean towards the term “English illness” introduced by Hakkanson, et al., German is supposed to function as “German illness” but with a positive effect on speakers of German.

3. Study

To test our hypothesis we constructed a questionnaire with 25 Turkish sentences that had to be translated into English within 30 minutes. Besides the translation task the participants were to fill out a background questionnaire including questions about their age, gender, language skills and time spent abroad. The complete questionnaire is given in appendix A and B in section 8.

3.1. Participants/ Procedure

A total of 20 Turkish students participated in this survey. Half of the data was collected in Germany from Turkish students at the age of 14-15 attending the “Gymnasium” in class eight. They are L1 speakers of Turkish, L2 speakers of German and L3 speakers of English. In Germany, the study was conducted in “Türkisch-Deutscher Bildungsverein Mannheim e.V.” In the following this group is termed “German-Turks.”

The other ten students live in Turkey and are attending the eighth class of1 “Ilk-Öğretim” in Denizli, Turkey. Their L1 is Turkish and L2 is English. They will be called “Turkish-Turks,” throughout this paper. Both countries involved differ in their system of education, but both groups have had English lessons for four years. Furthermore, the curricula were reviewed to ensure that both groups are comparable in terms of their educational level.

In this paragraph we will mention some assumptions in regard to the hypothesis. We have hypothesized that L3 speakers of English will perform better in English if they are L2 speaker of German. In the case of this study, German-Turks are supposed to be better in the translation task than TurkishTurks. We presume that the first-mentioned group translate the Turkish sentences into German first and then into English. Due to the similarities of English and German this approach to the task should be more successful. We believe that this sequence of translating will be obeyed on the part of GermanTurks, because of three imaginable reasons:

a) Turkish students living in Germany attend a German school. Thus, they are usually confronted with translation tasks from German into English or the other way around. Consequently, they would probably translate the sentences to German first in order to get tasks they are more familiar with.
b) We think that German is the dominant language for the majority of students. The prevailing language at school, in everyday life and in peer groups is probably German rather than Turkish.
c) German-Turks usually get formal Turkish education neither at school nor at home. By formal education we mean traditional lessons including grammar, writing and reading. It is common to learn Turkish by being exposed to a Turkish speaking environment (family, relatives, television programs…). Thus, their Turkish skills are good as far as conversations are concerned. But students seem to have problems with Turkish in written forms. Our observations during the data collection confirmed our assumption. Students complained about the questionnaire being Turkish.

3.2. Materials

We focused on three characteristics of Turkish that differ from both English and German. These are (1) word order, (2) auxiliaries and (3) negations. The questionnaire is composed of five sentences for each characteristic and ten filler-sentences. In this section we are going to outline the differences of Turkish compared to English and German in more detail and give a sample for each category.

(1) Word order:

Concerning main clauses both English and German follow a SVO pattern, whereas Turkish has a SOV word order.

illustration not visible in this excerpt

Linguistic analysis of the English sentence

illustration not visible in this excerpt

Linguistic analysis of the German sentence

illustration not visible in this excerpt

Linguistic analysis of the Turkish sentence

(2) Auxiliaries:

In contrast to English and German there are no auxiliaries in Turkish. Example from the questionnaire:

illustration not visible in this excerpt

(3) Negations:

In negations it is usual to add a “not” in English and “nicht” in German to the (first) auxiliary if there is one. In sentences with “be” in Simple Present and Simple Past and “sein” in Präsens and Präteritum it is necessary to add a “not” or “nicht” in order to negate in English and German, respectively.

In Turkish, on the contrary, the lack of auxiliaries as well as the word order being head-final plays a role in negations. There are two ways of negating. First, the negative morphemes -me, -ma, - mi, -mı, -mu, -mü are attached to the verb stem according to the vowel harmony. Second, nouns, adjectives and the verb “to be” (existentials) are immediately followed by the negation words “değil” or “yok”.

Example from the questionnaire:

Translation task: Teyzem genç değil, ama çok dinç.

Correct answer: My aunt is not young but very fit.

German: Meine Tante ist nicht jung, aber sehr fit.

illustration not visible in this excerpt

Linguistic analysis of the English negation

illustration not visible in this excerpt

Linguistic analysis of the German negation

illustration not visible in this excerpt

Attempt of linguistic analysis of the Turkish negation

The questionnaire includes ten additional filler-items. “Filler items for Turkish-Turks” are primarily meant to be difficult for Turkish-Turks. We designed two fillers with relative-clauses (4.1), which do not exist in Turkish in comparison to English and German. In (4.1) words in italics indicate equivalent meaning in each language. Furthermore, there are also fillers which make a word-by-word translation more likely (4.2). Finally, there are fillers without an overt form of “be” in Turkish (4.3). Neither in English nor in German is such a construction possible. The filler items might induce “Turkish-Turks” to use different answers than the “German-Turks”.

Examples from the questionnaire:

(4.1) Translation task: Satılmayan araba eski idi.

Correct answer: The car that was not sold was old.

German: Das Auto, das nicht verkauft wurde, war alt.

(4.2) Translation Task: Hiç sigara içtinmi? (içmek= trinken, to drink)

Correct answer: Have you ever smoked?

German: Hast du je geraucht?

(4.3) Translation Task: Kalem masanın üstünde [ ].

Correct answer: The pen is on the table.

German: Der Stift ist auf dem Tisch.

On the other hand there are “filler items for German-Turks” that were supposed to be tricky for especially German-Turks. It is assumed that German- Turks give an inappropriate translation on account of misleading similarities between German and English. The pronunciation of words may be alike although they differ in meaning. Consider, for instance, “become” and “bekommen”. An example relating this word-couple is given in (4.4):

(4.4) Task: Bugün bana yeni bisiklet alınacak.

German: Ich bekomme ein neues Fahrrad heute.

English: I will get a new bike today.

If German-Turks translate the task into German first they probably tend to use “become” in their translation.

4. Results

In the following paragraphs we are going to present the results of the translation tasks of the two groups by comparison. The results for all categories are each visualized in a bar chart. German-Turks are abbreviated to Ger-Tr and Turkish-Turks to Tr-Tr, respectively. The percentages for right and wrong translations of both groups are shown on the vertical axis. In addition, the accurate value of percentages are given within the bar diagram.

4.1. Results referring SVO-constructions

Unanswered tasks were treated as a mistake throughout the study. Conforming with the appropriate word order was the pivotal criteria for translations to be categorised as correct. The usage of wrong vocabulary as well as wrong tense was neglected. An example that was counted as a correct translation is quoted in (5). In sections 4.1 to 4.3 words in italics will indicate derivation from the expected correct answer.

(5) Translation task: Hediyesini kapının arkasında buldu.

Expected answer: He/ She found his/ her present behind the door.

Answer: She found her gift back the door.

The word order was correct in this example. For this reason the answer was accepted as being correct, albeit “back” was an unsuitable vocabulary.


1 After the reform of the Turkish school system “Ilk okul” and “orta okul” were joined to “Ilk ögretim okulu”. Students attend “Ilk ögretim okulu” for 8 years. At the end of the last term students have to sit an exam called “Ortaöğretim Kurumları Öğrenci Seçme ve Yerleştirme Sınavı”, which is abbreviated to ÖKS. The results achieved at ÖKS determine which “lise” the students are allowed to study at, since there are different types of “lises”. Students spend 4 years at “lise” regardless of which “lise” they are attending. After 4 years at lise students have to take another exam named “Öğrenci seçme ve yerleştirme sınavını” shortened to “ÖSS”. The results of “ÖSS” and the type of the attended “lise” together determine both future subject and university of the students. Universities differ in duration of study, which is either 2 or 4 years.


ISBN (eBook)
ISBN (Book)
File size
697 KB
Catalog Number
Institution / College
University of Mannheim
transfer second language acquisition turkish students english



Title: Transfer in Second Language Acquisition - Turkish Students speaking English