November 2021 – Volume 25, Number 3
Gökhan Yigit
Kafkas University, Turkey
<gokhanzelhotmail.com>
Abstract
This study aimed at investigating the influence of L1 (Turkish) on the acquisition of L2 (English) collocations by comparing the scores of two Turkish ELT student groups with a phrase-judgment test. The test included both congruent and incongruent collocations as well as unrelated word combinations. The participants whose vocabulary levels were defined as 3000-word level (low level) and 5000-word level (high level) according to Vocabulary Level Test (VLT), which was developed by Schmitt et al. (2001) were asked to judge whether the expressions were acceptable in English. The scores of the low-level and high-level students were compared in terms of their error rates and reaction times to the collocations. The overall results indicated that participants in both groups made more errors with incongruent collocations than congruent collocations. Their reaction times were also longer with incongruent collocations than with congruent collocations. However, the data yielded non-significant results in terms of the vocabulary levels of the participants for both error rates and reaction times.
Keywords: collocations, L1 influence, congruency
There is a growing interest in research on the issue of multi-word units from various perspectives and hence researchers (e.g., Lee, 2016; Wolter & Yamashita, 2015; Yamashita & Jiang, 2010; Zareva & Sheheta, 2015) call these expressions with different names such as multi-word units, fixed expressions, formulaic expressions, and so on. The researchers have agreed on the idea that multi-word units are one of the most important aspects of language learning. These units such as collocations, idioms, proverbs, and fixed expressions are widely used by the users of a language (Yamashita & Jiang, 2010). It can even be difficult for native speakers to communicate without these expressions since languages are full of multi-word units that are used in both spoken and written discourse. In a similar vein, Sinclair’s (1991) idiom principle, which emphasizes the use of pre-constructed phrases instead of relying on the creativity of the language, shows how important it is to use such expressions in language learning.
Collocations, as a type of multi-word units, consist of a large proportion of language production. Collocations can be defined as recurrent word combinations although there is no definition of collocation agreed upon in the literature (Vural, 2010). This is stemmed from the researchers’ approaches to collocation studies in that collocations are examined from two basic approaches: the phraseology-based approach and the frequency-based approach. In other words, the definition of collocation is shaped concerning these approaches since both approaches deal with the notion of collocation from different perspectives. That is, a frequency-based approach determines how frequently a word combination is used. Corpus-based studies on the frequency of the expressions are samples of this approach. Different from the frequency-based approach, the phraseology-based approach focuses on the syntactic and semantic relations of the words and therefore it examines semantic transparency (Gyllstad & Wolter, 2016) of collocations as well as their being restricted (Nesselhauf, 2003). According to this view, collocations locate between fixed expressions and free variations (Sheheta, 2008). When compared to the fixed expressions such as “in order to, in front of”, collocations are looser combinations in that a word can collocate with different words such as “fall asleep” and “fall apart”. They are also restricted in a way that every word may not come together as a multi-word unit although they can be accepted as grammatically correct (e.g., perform a task vs. *perform a survey) (Nesselhauf, 2003). This can confirm why collocations and free variations are different from each other.
Considering their specific features, collocations are valuable for language learning and research on language processing since the focus is on whether multi-word units such as collocations are stored as if they were single words (Yamashita & Jiang, 2010) and hence they can easily be retrieved as a whole. Therefore, collocation learning may facilitate the comprehension and production of the language. Many researchers state that collocations can enhance accuracy and fluency (Webb & Kagimoto, 2010). It is also accepted that these expressions help learners to improve their fluency in written and spoken language and to attain a native-like mastery of a language (Laufer & Waldman, 2011). In this regard, collocations are crucial in having native-like language competence, and thus they should be integrated into foreign/second language learning (Nesselhauf, 2003).
Many researchers (e.g., Chorbwhan & McLennan, 2016; Gllystad & Wolter, 2016) maintain that the acquisition of collocations can be problematic even for advanced foreign/second language learners. The reason why it is difficult for them, for instance, is that collocations are arbitrarily and semantically restricted although they can be grammatically correct (Yamashita & Jiang, 2010). For example, “heavy traffic” is acceptable as a collocation; however, there is no such expression as “strong traffic” used as a collocation unlike any grammatical restriction for these words. Another reason is that collocations are different from free variations because collocations are semi-transparent. That is, one of the words that constitutes collocations loses its basic meaning and becomes semantically less transparent.
Many types of research on collocations confirm that native languages of learners have a strong influence on the acquisition of collocations (Lee, 2016; Phoocharoensil, 2013). That is, cross-linguistic influence is an important concept for second language acquisition and similarities and differences between the native language and the target language may be facilitative or inhibitory for language learners. Empirical studies on cross-linguistic influence (e.g., Sheheta, 2008) also reveal that the relationship between two languages (L1-L2) may explain in what ways collocations are acquired. For example, Sheheta (2008) examines the influence of L1 (Arabic) on both reception and production of English collocations by advanced Arabic-speaking learners of English in an ESL setting in comparison with EFL learners. The findings of this study reveal that learners’ L1, as well as their learning environment, has a strong influence on the acquisition of L2 collocations. One of the reasons for this is that collocations are not restricted like idioms and different languages use different word combinations for collocations with the same meaning. For example, the expression of “black coffee” is translated into Turkish as “sade kahve”; however, the literal meaning of the word “black” is “siyah” not “sade” in Turkish. Wolter and Gllystad (2011) explain that there is no semantic reason for the languages to use different words to constitute collocations and hence a possible transfer from L1 to L2 occurs resulting in learners making mistakes in the target language. Another reason is that collocations are partially restricted and any word may collocate with different words. As a result of this flexibility, learners may not be aware of collocations as multi-word units and they can use a variety of words to make collocations (Yamashita & Jiang, 2010) although they are non-collocations.
The differences between native speakers and L2 learners regarding mental lexicon, which can be defined as a dynamic representation of words in the memories of language users (Cangır, 2018) can also be influential in the acquisition of collocations since native speakers and L2 learners have different experiences in language processing. That is, L2 learners may easily learn a lexical item that already exists in both languages since this item is already processed in the mental lexicon of the learner. From a similar perspective, collocations with the same literal meanings in both languages are called congruent collocations whereas collocations that are not translated word by word from one language to another are called incongruent collocations. For example, “break a heart” has a word by word translation in Turkish as “kalp kırmak” (break= kırmak, heart= kalp) and it is regarded as a congruent collocation; however, there is no word by word translation of “fall asleep” in Turkish and hence it is called an incongruent collocation.
The present study aims to examine the error rates and the reaction times of first-year Turkish ELT students in the processing of collocations in terms of congruency and vocabulary levels. The current study may shed light on the L1 influence in the processing of L2 collocations in the Turkish context since the studies testing the L1 influence on the acquisition of collocations are very limited in the Turkish context (Cangır, 2018). Wolter and Gyllstad (2011) also state that studies that investigate the psychological reality of L2 collocations are rare and the studies that explore the influence of L1 on the processing of L2 collocations are few. Considering the learners’ avoidance strategies which can be seen in corpus-based data (Gass & Selinker, 2008), the current study is not likely to be affected by these strategies since a phrase-acceptability judgment test is used to examine the error rates and the reaction times of the participants in the processing of English collocations in this study. Therefore, the present study can contribute to the pedagogy regarding the acquisition of L2 collocations in the Turkish context, for instance, by making learners aware of cross-linguistic differences. In this respect, the following research questions are to be answered:
- Does L1 of the first-year Turkish ELT students influence the acquisition of L2 collocations in terms of their (a) error rates and (b) reaction times concerning the distinction between congruent and incongruent collocations?
- Do vocabulary levels of the first-year Turkish ELT students influence the acquisition of L2 collocations in terms of (a) error rates and (b) reaction times concerning the distinction between congruent and incongruent collocations?
Review of Literature
Collocation studies have recently become prominent in foreign/second language learning. However, cross-linguistic influence on the processing of collocations can be regarded as one of the neglected aspects of collocation learning. Since this study tries to examine the influence of L1 on the acquisition of L2 collocations, the studies, which explore the cross-linguistic differences in terms of collocation learning, are voiced in this part.
Nesselhauf (2003) investigated the problems the advanced learners of English faced in the production of L2 collocations. He investigated 32 written essays of German speakers of English in terms of verb-noun collocations. The collocations found in these essays were checked whether they were correct or not. The researcher realized that nearly 25% of the collocations produced by the learners were wrong. The most important reason for erroneous collocation production was the influence of L1 (52%). He also investigated the types of errors in collocations and found no difference among them with regard to the L1 influence.
In a similar study, Mahmoud (2005) investigated the collocation errors in 42 essays written by Arab learners of English. The analysis of the data indicated that most of the collocations identified were lexical collocations. Nearly 61% of the collocations found in the essays of the participants were incorrect. The native language of the learners, that is a negative transfer from Arabic to English negatively influenced the usage of collocations.
In another context, Kuo (2009) investigated the collocation errors of Taiwanese learners of English. The analysis of 98 essays of the learners showed that students made more errors with the V+N type of collocations than with the adj+N type of collocations. According to the writer, the sources of collocation errors were the negative transfer from L1 to L2, the use of synonymy and approximation, that is, the closeness of the words, and negative inference.
Different from corpus-based studies, Yamashita and Jiang (2010) used a phrase-acceptability judgment test to investigate the role of L1 on the processing of L2 collocations. The researchers also compared the error rates and the reaction times of the learners concerning congruent and incongruent collocations since their participants had been living in different language settings (ESL, EFL, and NS). The results indicated that ESL learners were more successful than EFL learners in both congruent and incongruent collocations in terms of their error rates and reaction times. Besides, ESL learners were as fast as NS; however, EFL learners made more errors with incongruent collocations than congruent collocations although they were regarded as advanced learners of English. The results of this study indicated that both congruency and L2 proficiency were important predictors of L2 collocation learning.
From a similar perspective, Wolter and Gyllstad (2011) investigated verb+noun collocations based on the types of collocations, congruent and incongruent. The participants were native speakers of English and Swedish learners of English. The researchers used a Lexical Decision Task, which measures the accuracy and speed of the participants, to examine whether congruency and English proficiency affected the acquisition of L2 collocations. The results indicated that native speakers were more accurate and faster than Swedish learners of English were. Secondly, Swedish participants did statistically better in congruent collocations than in incongruent collocations although there was not any statistically significant difference in the scores of native speakers in terms of congruent and incongruent collocations.
Yamashita (2014) experimented with the influence of L1 on L2 collocations with his study and the participants were 144 Japanese students and 24 native speakers. The researcher constructed four types of collocations: (1) the occurrence of the collocations word by word in two languages, (2) collocations which are felicitous in English and infelicitous in Japanese, (3) collocations which are infelicitous in English and felicitous in Japanese and, (4) unrelated items. The results indicated that the native speakers were more accurate in all collocations types whereas there was a fluctuation among the NS considering the accuracy rates of collocation types. Japanese students did only better with infelicitous collocations, showing that there was a great influence of L1 on the acquisition of L2 collocations.
In the Turkish context, although numerous studies related to collocations focused on different aspects of the issue, and, to the best of our knowledge, only one study (Cangır, 2018) examined the acquisition of collocations from the cross-linguistic perspective. He investigated the cross-linguistic nature of the bilingual mental lexicon by comparing the collocations in two languages, that is Turkish and English. The participants of the study were Turkish ESL and EFL learners. A lexical decision task was used to identify the collocation knowledge of the participants concerning Adj-N and V-N collocations. The results indicated that the participants responded faster to the collocations than unrelated items, to the Adj-N collocations than the V-N collocations. They also responded faster to the congruent collocations than incongruent collocations.
As understood from the empirical studies on the acquisition of L2 collocations, congruency has a strong influence on the processing of L2 collocations for foreign language learners. Therefore, this study mainly attempts to investigate the influence of L1 on the acquisition of L2 collocations by comparing two typologically different languages, Turkish (L1) and English (L2) in terms of the error rates and the reaction times of Turkish ELT students to congruent and incongruent collocations.
Methodology
Participants
At the beginning of the study, 61 first-year Turkish ELT students in a state university of Turkey participated in this study. We asked questions to get information about their backgrounds just before the data collection process. Their ages ranged from 18 to 26 (average: 19) and 30 of the participants were female and 31 of them were male. Their first language was Turkish except for one student who informed us that her native language was Persian. Our aim was also to make a homogenous group in terms of exposure to the target language and therefore we asked whether the participants lived in an English-speaking country before. All of them acknowledged that they have not had such an experience in their lives. We chose the participants from the first-year ELT students for this study because they passed the Turkish National Foreign language exam, which is called “YDS”, a test, which includes vocabulary, grammar, and reading skills, to be enrolled in ELT departments. They also passed the placement test of English, which is prepared by the School of Foreign Languages of the university and this test includes all basic skills of language learning. Therefore, they can be regarded as having a certain level of proficiency in English. Then, Vocabulary Levels Test (VLT), which Schmitt et al. (2001) developed was applied to test their vocabulary levels and to see whether the participants had similar vocabulary knowledge. VLT scores indicated that 13 of the students were at the 2000 word level, 28 of them were at the 3000-word level, and 20 of them were at the 5000-word level. 13 students whose vocabulary level was 2000 and 1 student whose native language was Persian were excluded from the study. There were only 47 students (21 females, 26 males) remaining for the present study.
Item Development
For the item development, we identified 100 adjective-noun collocations from “Dictionary of Collocations Idioms and Phrasal Verbs for Turkish Students”, which was written by Metin Yurtbaşı. Adjective-noun collocations were selected for this research because adjectives precede a noun in both English and Turkish languages. For example, in the expression of “bad temper” the adjective precedes the noun as in “kötü huy”, that is the adjective “kötü” precedes the noun “huy” in Turkish. Therefore, learning adjective-noun collocations can be easier for Turkish learners of English. Then, every word that constitutes collocations was checked out in the Oxford Advanced Learners Dictionary to see whether each word is included in the most frequent 3000-word list. When any word of collocations was not included in the most frequent 3000-word list, this collocation was excluded from the list and a new collocation meeting the condition of 3000 frequent word list was added to the collocation list. The selected collocations were assessed by five experts who had at least a Master’s degree in English language teaching (ELT) and work in this department to determine whether they were congruent, that is, a direct translation of collocations was available in L1, or incongruent, that is, there was no direct translation of collocation in L1. Collocations were defined as congruent and incongruent according to these criteria:
- At least four experts out of five define collocations as congruent for the congruent collocations
- At least four experts out of five translate collocations into Turkish with the same words for the congruent collocations
- At least four experts out of five define collocations as incongruent for the incongruent collocations
- At least four experts out of five translate collocations into Turkish by using different words for the incongruent collocations
Considering the experts’ judgments about collocations, the list consisted of two types of collocations: Congruent list and incongruent list. We also checked the collocations in both lists from Corpus of Contemporary American English (COCA) in terms of their frequencies and Mutual Information (MI) scores. Collocations that are in the 0-1000 frequency band (mean: 405 for congruent collocations, 220 for incongruent collocations) were chosen as items. The MI scores were also checked and collocations whose MI scores were over 3.0 (mean: 4.80 for congruent collocations, 5.38 for incongruent collocation) were chosen as the items for this study. Both frequencies and MI scores of congruent and incongruent collocations were compared to see whether collocations in both groups are similar in difficulty and comparability. The results of the T-test indicated that there was not any statistically significant difference between congruent and incongruent collocations in terms of frequency and MI score. Therefore, we chose 10 congruent and 10 incongruent collocations as the items of this research. Besides, 20 unrelated word combinations were selected from the study of Wolter and Gllystad (2013) as the baseline data. In the end, we defined 40 items (10 congruent, 10 incongruent, and 20 unrelated word combinations) as the items of the research.
Data Collection Procedure
Before the data collection, the first step was to ask the students whether they voluntarily wanted to participate in the study. Considering ethical issues, a consent form was given to the students and only those who signed the consent form were accepted for the study. Then, we asked the participants about their background information such as their gender, age, and whether they lived in English-speaking countries or not. Then, the VLT test was applied to define their vocabulary levels. Based on the results of VLT, the participants were grouped as High level (whose level is 5000) and Low level (whose level is 3000). Of the 61 students, 13 of them were at the 2000 word level and one student was Persian; therefore, they were excluded from the study. Of the remaining 47 students, 27 of them were at the 3000 word-level and 20 of them were at the 5000 word-level.
A software program, PsychoPy, which is free to use, was downloaded, and then the phrase acceptability test was prepared in this program. The participants were asked whether the expressions were acceptable in English. When they believed that the expressions they saw on the screen were acceptable, we asked them to press the key “y” meaning “yes”. If not, they pressed the key “n” meaning “no”. The program presented the items randomly one by one in the middle of the screen. The program was adjusted in a way that the participants could only press the keys 500 milliseconds after they saw the items to make sure that they saw the items first and then pressed the key.
47 students were asked to come to the researcher’s office to participate in the test since it was the room in which noise or other things did not distract the participants. Every participant took the test individually. Before the test, participants were instructed about how the program worked and a trial session was applied to relieve the participants about the test and see whether they understood the instruction. After the trial session, the real experiment was applied to the participants. The data collection lasted nearly 10 to 15 minutes including the instruction of the program, trial session, and experiment on each participant. For the experiment, the program automatically recorded the answers of the participants whether the students knew the correct answer and their reaction times to the items shown on the screen of the computer.
Results
After applying the phrase-acceptability test, we computed the descriptive statistics of the data based on the collocation types, that is congruent collocations and incongruent collocations. The results showed that the students’ error rates were significantly lower in congruent collocations (M: 1.85 SD: 1.231 for low level; M: 1.35 SD: 0.933 for high level) than in incongruent collocations (M: 5.26, SD: 1.701 for low level; M: 5.40, SD: 1,789 for high level) for both groups. The results were also similar in terms of the reaction times in that first year Turkish ELT students responded significantly faster in congruent collocations (M: 867, SD:398 for low level; M: 804, SD: 406 for high level) than in incongruent collocations (M: 1238, SD:539 for low level; M: 1070, SD: 563 for high level) for both groups.
Table 1. Descriptive Statistics related to L2 Collocations in Terms of Error Rates and Reaction Times.
Low level | High level | ||||||
Mean | SD | Mean | SD | ||||
Error rates | Congruent | 1.85* | 1.231 | 1.35* | 0.933 | ||
Incongruent | 5.26* | 1.701 | 5.4* | 1.789 | |||
Reaction times | Congruent | 867* | 398 | 804* | 406 | ||
Incongruent | 1238* | 539 | 1070* | 563 | |||
Note: * p<0, 01 for congruent and incongruent collocations. Reaction times were computed in milliseconds
According to the first research question, we investigated whether the students’ L1 influenced the acquisition of L2 collocations in terms of the error rates and the reaction times concerning the distinction between congruent and incongruent collocations. Paired samples t-test scores indicated that congruent collocations were easier for the participants than incongruent collocations in terms of both error rates and reaction times. In other words, first-year Turkish ELT students significantly outperformed the congruent collocations compared to the incongruent collocations with regard to the error rates at p<0,01 level. Similarly, the participants reacted faster to the congruent collocations than the incongruent collocations at p<0,01 level.
Table 2. Error Rates Results Based on 2×2 Mixed-Design ANOVA.
Tests of Between-Subjects Effects | |||||
Source | Type III Sum of Squares | df | Mean Square | F | Sig. |
Corrected Model | 321,515a | 3 | 107,172 | 50,252 | ,000 |
Intercept | 1103,728 | 1 | 1103,728 | 517,527 | ,000 |
Level | ,749 | 1 | ,749 | ,351 | ,555 |
Congruency | 319,479 | 1 | 319,479 | 149,800 | ,000 |
Level * Congruency | 2,372 | 1 | 2,372 | 1,112 | ,294 |
Error | 191,943 | 90 | 2,133 | ||
Total | 1651,000 | 94 | |||
Corrected Total | 513,457 | 93 | |||
a. R Squared = ,626 (Adjusted R Squared = ,614) |
The second research question aimed to investigate whether the participants’ vocabulary level influenced the acquisition of congruent and incongruent collocations in terms of the error rates and the reaction times concerning the distinction between congruent and incongruent collocations. A 2X2 mixed-design ANOVA was utilized to explore the effects of congruency and vocabulary level of the participants on the error rates and the reaction times to test items. The results in Table 2 regarding the interaction between the level of the students and collocation types, congruent and incongruent, indicated a non-significant value (F(1, 90) = 1.112, p= .294) in terms of the error rates. As for the main effects, the vocabulary levels of the students were found to be non-significant affecting the error rates of students. (F(1, 90) = .351, p= .555). On the other hand, there was a statistically significant main effect for the congruency in terms of error rates of the students (F(2,74)=149.800, p<.001).
Table 3. Reaction Times Results Based on 2×2 Mixed-Design ANOVA.
Tests of Between-Subjects Effects | ||||||
Source | Type III Sum of Squares | df | Mean Square | F | Sig. | |
Corrected Model | 2868018,696a | 3 | 956006,232 | 4,125 | ,009 | |
Intercept | 91046147,632 | 1 | 91046147,632 | 392,82 | ,000 | |
Level | 305618,611 | 1 | 305618,611 | 1,319 | ,254 | |
Congruency | 2330986,298 | 1 | 2330986,298 | 10,057 | ,002 | |
Level * Congruency | 61653,021 | 1 | 61653,021 | ,266 | ,607 | |
Error | 20859697,219 | 90 | 231774,414 | |||
Total | 118453110,000 | 94 | ||||
Corrected Total | 23727715,915 | 93 | ||||
a. R Squared = ,121 (Adjusted R Squared = ,092) |
The results in Table 3 also indicated that there was not a statistically significant interaction between the vocabulary level of the students and the collocation types, congruent and incongruent, (F(1, 90) = .266 p= .607) in terms of the reaction times. As for the main effects, the vocabulary levels of the students were not found to be significant in terms of reaction times of the students (F(1, 90) = 1.319, p= .254). On the other hand, there was a statistically significant main effect for the congruency in terms of reaction times of the students (F(2,74)=10.057, p=.002). As for the second research question, we can say that the vocabulary levels of the participants were not found to have a statistically significant effect on the acquisition of different types of collocations in terms of the error rates and the reaction times at p> .05 level.
Discussion
This study investigated whether L1 affected the acquisition of L2 collocations. In other words, the present research aimed to explore the cross-linguistic influence on collocation learning by comparing the students’ scores in terms of the error rates and the reaction times gathered from a phrase-accessibility test. The results revealed that congruency is a strong predictor of the learner’s success in the processing of collocations. That is, the participants were more successful with the congruent collocations than the incongruent collocations concerning the error rates of the participants. Similarly, the participants reacted to the congruent collocations faster than the incongruent collocations. It can be concluded that L1 has a strong influence on the processing of L2 collocations. We can say that positive transfer (Sheheta, 2008) influences collocation learning. In other words, when L1 and L2 become more familiar in terms of collocations, L1 may have a facilitative role in the acquisition of L2 collocations for EFL learners. However, collocation learning may be more difficult for learners of English when incongruent collocations are instructed since the mental processing of collocations can be different for EFL learners.
The results of the present study are also in line with the studies, which examine the effect of congruency in collocation acquisition. For example, Wolter and Yamashita (2018) found similar results in that NNS processed congruent collocations faster than incongruent collocations. Similarly, Yamashita and Jiang (2010) used a phrase-acceptability judgment test and compared the EFL and ESL learners whose native language is Japanese and found that both EFL and ESL students made significantly more errors in incongruent collocations than congruent collocations. Besides, Wolter and Gyllstad (2011) identified that Swedish learners had significantly higher error rates for incongruent collocations than congruent collocations. The findings of these researches claim that the influence of congruency is stronger for the processing of L2 collocations. Many researchers accept that congruent collocations are acquired faster than incongruent collocations; however, it still needs to be enlightened why this is so. According to Wolter and Gyllstad (2011), congruent collocations are directly copied from the native language of the learners to L2 collocation learning. That is, learners of the study copy the syntactic and semantic information of the collocations from L1 to L2 and this makes the processing of congruent collocations faster than incongruent collocations. Another reason for faster processing of congruent collocation can be related to the order of acquisition. Wolter and Yamashita (2018) report that congruent collocations are processed faster than incongruent collocations, which may be explained with the system of age/order of acquisition (AoA, OoA), as congruent collocations, which have equivalents in L2 are acquired in the first language, strengthen their positions in the process of language learning. When the learners become more experienced in language learning, this may probably lead to the exposure of L2 collocations and the acquisition of congruent collocations become more salient than the incongruent collocations since incongruent collocations are not transferred; and hence they become less salient.
This study also investigated the acquisition of collocations in terms of the vocabulary knowledge of the participants. In other words, the participants who were grouped as low level and high level according to the VLT test were compared based on the error rates and the reaction times. The results indicated that there was not any statistically significant difference between the low-level and high-level students concerning their error rates and reaction times. Although many studies (e.g., Yamashita & Jiang, 2010) found that the proficiency was influential on the processing of L2 collocations, the present study revealed that Turkish ELT students, regardless of their vocabulary levels, had similar success in both error rates and reaction times in terms of the congruency types. In other words, the present research did not approve that learners who were more knowledgeable and experienced in the target language were likely to be successful in the processing of L2 collocations. One of the basic reasons for this situation can be related to the conditions of learners in that all the participants informed that they have not lived in an English-speaking country before. Most of the studies (e.g., Wolter & Gllystad, 2013) compared the scores of native speakers and non-native speakers or compared the scores of those who live in the ESL and EFL contexts (e.g., Lee, 2016). In these studies, the participants have different exposure rates to the target language and this can be the reason why the participants in the ESL context were faster and responded more accurately than the participants in EFL contexts. Cangır (2018) also informed that the links between congruent collocations in L1 and L2 might fade in time since the language users become more proficient in an ESL context with the help of exposure to the target language. He also asserted that the activation of congruent collocations may not be linked to the native languages of the learners since collocations are directly stored in the mental lexicon regardless of the native language of the learners. However, the participants of this study have similar educational backgrounds and there is not immense exposure to the target language although their vocabulary levels differed from each other. That is, the conditions of the participants in this study and the participants in the other studies are not comparable with regard to the exposure rates to the target language.
Conclusion and Implications
This study investigated the processing of L2 collocations concerning the influence of L1 and the vocabulary knowledge of the participants. The findings revealed that congruency had a significant effect on the acquisition of collocations whereas there was not a statistically significant difference between the groups of participants in terms of their vocabulary knowledge. Considering the results of this study, although the participants are advanced learners of English, L1 has a strong influence on the acquisition of L2 collocations since they have made many errors in the incongruent collocations. Therefore, we can suggest that special attention be paid to the teaching of collocations in the syllabi of foreign language teaching since it seems problematic even for the advanced learners of English.
Concerning the results of the study, we can conclude that incongruent collocations are harder to grasp. Providing activities that include incongruent collocations can be useful for an English teacher to make his lesson effective. In this regard, foreign language teachers should be aware of how important collocations are in language learning and they should raise the awareness of the learners about incongruent collocations. Besides, teaching such collocations explicitly can be suggested to the teachers since the learners (Cangır, 2018) may not easily realize these expressions. Several researchers (e.g., Wolter & Yamashita, 2018) state that emergentist approaches that take the frequency of input as the basis should be considered in L2 collocation teaching. It is because we need to develop a broader understanding to uncover the influence of L1 not only on the processing of collocations but also on learning a foreign language.
Nesselhauf (2003) also states that collocations can be problematic since they are not restricted as much as fixed expressions although they are congruent collocations. More salient and frequently-used collocations can therefore be placed in the teaching materials since not all collocations are likely to be involved in teaching materials.
About the Author
Gökhan Yigit is a research assistant in English Language Teaching Department of Dede Korkut Faculty of Education, Kafkas University, Turkey. He is a Ph.D. Candidate in English Language Teaching. His research interests include English language teaching as a foreign language, second/foreign language teacher education, and affective factors in language teaching.
To cite this article
Yigit, G. (2021). The influence of L1 on the acquisition of l2 collocations: Turkish ELT students. Teaching English as a Second Language Electronic Journal (TESL-EJ), 25(3). https://tesl-ej.org/pdf/ej99/a2.pdf
References
Cangır (2018). Investigating the relationship between L1 and L2 collocational processing in the bilingual mental lexıcon. (Unpublished Ph.D. Dissertation). Hacettepe University.
Chorbwhan, R. & McLellan, J. (2016). First language transfer and the acquisition of English collocations by Thai learners. Southeast Asia: A Multidisciplinary Journal, 16, 16-27.
Gass, S. M. & Selinker, L. (2008). Second language acquisition: An introductory course. Routledge.
Gyllstad, H. & Wolter, B. (2016). Collocational processing in light of the phraseological continuum model: Does semantic transparency matter? Language Learning, 66(2), 296-323. https://doi.org/10.1111/lang.12143
Kuo, C. L. (2009). An analysis of the use of collocation by intermediate EFL college students in Taiwan. Arecls, 6, 141-155.
Lee, S. (2016). L1 influence on the processing of L2 collocation: An experimental study of Korean EFL learners. Linguistic Research, 33, 137- 163.
Mahmoud, A. (2005). Collocation errors made by Arab learners of English. Asian EFL Journal, 5(2), 117-126.
Nesselhauf, N. (2003). The use of collocations by advanced learners of English and some implications for teaching. Applied Linguistics, 24(2), 223-242. https://doi.org/10.1093/applin/24.2.223
Phoocharoensil, S. (2012). Cross-linguistic influence: Its impact on L2 English collocation production. English Language Teaching, 6(1), 1-10. http://doi.org/10.5539/elt.v6n1p1
Sheheta, A. K. (2008). L1 Influence on the Reception and Production of Collocations by Advanced ESL/EFL Arabic Learners of English. (Unpublished MA thesis). Ohio University.
Vural, E. (2010). Explicit and Incidental Teaching of English Verb+Noun Collocations in an EFL Context. (Unpublished Ph.D. Thesis). Anadolu University.
Webb, S. & Kagimoto, E. (2010). Learning collocations: Do the number of collocates, position of the node word, and synonymy affect learning? Applied Linguistics, 32(3), 259-276. https://doi.org/10.1093/applin/amq051
Wolter, B. & Gyllstad, H. (2011). Collocational links in the L2 mental lexicon and the influence of L1 intralexical knowledge. Applied Linguistics, 32(4), 430-449. https://doi.org/10.1093/applin/amr011
Wolter, B., & Gyllstad, H. (2013). Frequency of input and L2 collocational processing: A comparison of congruent and incongruent collocations. Studies in Second Language Acquisition, 35(3), 451-482. https://doi.org/10.1017/S0272263113000107
Wolter, B. & Yamashita, J. (2015). Processing collocations in a second language: A case of first language activation? Applied Psycholinguistics, 36(5), 1193-1221. https://doi.org/10.1017/S0142716414000113
Wolter, B. & Yamashita, J. (2018). Word Frequency, Collocational Frequency, L1 Congruency, and Proficiency in L2 Collocational Processing: What Accounts for L2 Performance? Studies in Second Language Acquisition, 40(2), 395-416. https://doi.org/10.1017/S0272263117000237
Yamashita, J. & Jiang, N. (2010). L1 influence on the acquisition of L2 collocations: Japanese ESL users and EFL learners acquiring English collocations. Tesol Quarterly, 44(4), 647-668. https://doi.org/10.5054/tq.2010.235998
Yamashita, (2014). Effects of the Instruction on Yes-No Responses to L2 Collocations. Vocabulary Learning and Instruction, 3(2), 31- 37. http://dx.doi.org/10.7820/vli.v03.2.yamashita
Zareva, A. & Shehata, A. (2015). At the Intersection of L1 Congruence and L2 Exposure: Collocational Knowledge of Advanced Arab Users of English. International Journal of Arabic-English Studies, 5, 77- 102.
Copyright rests with authors. Please cite TESL-EJ appropriately. Editor’s Note: The HTML version contains no page numbers. Please use the PDF version of this article for citations. |