Early and Late Language Start at Private Schools in Turkey*

Educational Sciences: Theory & Practice - Special Issue • Autumn • 3199-3209 © 2012 Educational Consultancy and Research Center www.edam.com.tr/estp Early and Late Language Start at Private Schools in Turkey* a Saban CEPIK Hedayat SARANDI Zirve University Zirve University Abstract This study examines the interaction effect of age in L2 attainment. It explores whether success in foreign language learning at early childhood grades varies depending on age. It also addresses the beliefs of foreign language teachers regarding the variables under review. Eighty-three 11 year-old language learners who started learning English at different ages were placed into two groups. The initial exposure of the early starters was 5-6 and of the late starters, 9-10. A set of language proficiency tests covering all four-language skills were given to the participants to determine the possible differences in the proficiency of the two groups. Also, qualitative data was collected from 6 teachers through a questionnaire that aimed to elicit their beliefs regarding the effect of age on L2 attainment. The findings showed that the early starters did not perform significantly better than the late starters in any of measures. The teachers, however, indicted that the early language learners had more positive attitude towards English compared to the late starters. Findings underscore that language attainment may involve a lot of variables and that early age may not take account of the whole issue. Key Words Early Childhood Language Instruction, Age Period, Length of Instruction, Early Starters, Late Starters. he issue regarding the relationship between age and language learning has both theoretical and practical value. On theoretical ground, it relates to * This paper was revised after being presented at the International Conference on Global Issues of Early Childhood Education and Children’s Rights, Gaziantep, Turkey, 27-29 April 2012. a Saban CEPIK, Ph.D., is an assistant professor and director of the English Preparatory School and Chair of the English Language Teacher Education (ELTE) department at Zirve University in Gaziantep/Turkey. Dr. Cepik’s research interests include teaching methodology, comparative study in language learning and teaching, pedagogical beliefs of language teachers, curricula frameworks in language teaching at schools and in higher education: Correspondence: Assist. Prof. Saban CEPİK, Zirve University, Faculty of Education, Department of English Language Teaching, Kızılhisar Campus 27260 Gaziantep/Turkey. E-mail: saban.cepik@ zirve.edu.tr Tel: +90 342 211 6830/6952. one of the key issues in language acquisition: whether there is a diference between adult and child language acquisition or if Universal Grammar (UG) is still functional for adult second language learners (Mayo & Lecumberri, 2003). Questions like these are of the utmost importance for the researcher of SLA (Long, 2007). On the practical side, however, the age-related issue involves the key decision of when to introduce second language into the classroom settings (Mayo & Lecumberri, 2003), and whether or not it is possible to attain native-like proficiency ater a certain age. Such questions fall into the interest range of a large number of people including language planner, language teachers, parents and language learners themselves. he issue has gained an added importance in Turkey owing to Turkey’s relatively unsuccessful history of English language teaching. Koru and Akessson (2011), for example, in a study conducted by TEPAV (Economic Policy Research Foundation of Turkey) reported the results of the English Proficiency Index in which Turkey ranks 43 among 44 countries. he researchers argued that the so- EDUCATIONAL SCIENCES: THEORY & PRACTICE called late introduction of English into language classes is one of the main reasons, among others, for the lack of success in language learning in Turkey. he present research study aims to explore this issue hoping that its findings may cast light on the long-running argument regarding the relationship between age and language learning. he Earlier, the Better? he idea that an early introduction of a second language leads to a higher level of language proficiency appears to be accepted around the world (Pufahl, Rhodes, & Christian, 2000). It is assumed that an early start of second language learning provides the time that learners need for the daunting task of learning a second language. Furthermore, early starters are arguably endowed with the ability to “sponge” the new language and become more proficient (Cenoz, 2003; Heighington, 1996). In fact as Torras, Tragant, and García (1997, p. 142) put it, “he younger they are, the more they are like sponges, the more they absorb, the more they retain.” heoretically, the idea of the supremacy of young learners over late learners is fueled by the Critical Period Hypothesis (CPH). Originally proposed for the first language, CPH claims that there is a certain time restriction outside which the learning of a new language becomes dificult, if not impossible. It is argued that the facility for the initial unconscious acquisition of a certain language through exposure disappears ater puberty and that learning a foreign language becomes more of a conscious and efortful endeavor aterwards that does not bring the same results (Lennenberg, 1967). One reason for this is the maturation changes, such as lateralization and plasticity-loss, that brain goes through which in turn afects the processing of a new language. In lateralization, diferent hemispheres of the brain take up separate responsibilities, allocating language-related activities to the let hemisphere. In terms of plasticity, a process called myelination is reduced resulting in the hardening of neural cords and the transportation of neural impulses (Marinova-Todd, Marshall, & Snow, 2000). Maturational factors are only one of the reasons that accounts for the superior performance of young language learners. Gass and Selinker (2008) offer social, psychological, cognitive and neurological reasons to account for why adults are not as good as children in language learning. In terms of social and psychological factors, adults’ willingness to adhere to their identity and ego may prevent them 3200 from picking up a new accent and approaching native-like mastery of the target language. Furthermore, adults have more developed cognitive skills compared to children, and use analytical approaches to learning a second language, which are less productive compared to children who rely more on their language acquisition devices and implicit learning. Finally, compared to adults, children have access to more modified language, which results in a better language input. Other reasons are ofered for the low performance of adults. Herschensohn (2008) attributes the dificulty of learning a second language to the learners’ L1. She argues that it is the perfection of the L1 language system that somehow prevents the older learners accessing the internal language learning mechanism. Bridsong (1999) considers the loss of UG ater puberty, and changes in memory capacity (and processing) as two reasons, among others, for the poor performance of adults in terms of learning a second language. he bulk of studies appear to show that in the ESL context younger learners are better than older ones, when considering the ultimate proficiency attainment. his is mainly so for phonology although there is evidence for the success of young leaners in other language areas as well (Gass & Selinker, 2008). In pronunciation, for example, several studies show that the age at which second language learning started has a great bearing on the foreign accent (Yeni-Komshian, Feldge, & Liu, 2000; Yeni-Komshian Robbin, & Flege 2001). Tahta, Wood, and Loewenthal (1981) found that the ability of American children to imitate Armenian and French intonational pattern reduced significantly ater the age of 8. Moyer (1999) found that German adult students, despite ample exposure to English both in and out of classroom and having high motivation, failed to attain a native-like accent (see Gass & Selinker). Similar findings were reported for syntax. Patkowski (1980) found that out of 33 people who started learning a second language before the age of 14, all except one achieved native or near-native mastery of syntax whereas the majority of the people who started their second language ater 14 remain at a comparatively lower proficiency level. Likewise, Johnson and Newport’s study (1989) examined the syntactic knowledge of a group of Chinese and Japanese immigrants who arrived at the US at diferent times and found that those who arrived earlier achieved native-like mastery in grammatical tests, whereas late arrivals varied widely in this regard. CEPIK, SARANDI / Early and Late Language Start at Private Schools in Turkey Voices of Concern! While CPH is mainly accepted in first language acquisition, its existence and role in L2 is still debated. For one thing, it is argued that the decline in the acquisition of diferent language areas through age is not as abrupt as predicted by CPH. In fact it is rather gradual. his has caused some to use the term “sensitive” period rather than critical (see Long 1990, for example). Second, puberty may not be the right borderline as some language areas may start to deteriorate earlier. Foreign accent, for example, may start to reveal itself at any time ater the age of 6 (Yeni-Komshian et. al., 2000; Yeni-Komshian et. al., 2001). here are others who argue that CPH fails to take full account of the issue. Singleton and Ryan (2004, p. 18), for example, raises the point that age-related decline in learning ability may be the result of multiple factors, and that to attribute it solely to maturation and neurobiological phenomena as theorized by CPH is not justified by the latest data. He concludes, “he idea of a critical period for language development may well have had its day.” Perhaps the most severe criticism against CPH comes from the records of successful adult language learners. Some researchers, drawing upon studies whose findings revealed variations in the performance of late language learners, show that at least some of the late language learners managed to reach a native or native-like proficiency level. hey argue that if CPH holds true, then all late learners should perform at a level lower than that of both young learners and native speakers (MarinovaTodd, 2003). he results of research show, however, that this is not necessarily the case. Marinova-Todd reports a surprising study by Birdsong (1992), where some late-learner near-native speakers performed better than native speakers. here are other studies that show that native-like proficiency, at least for some adults, is possible (see Juf & Harrington, 1995; White & Genesee, 1996 in syntax; Bongarets, 1999 in pronunciation). Another issue is that if the argument in favor of CPH is valid in terms of final achievement, it is not so in terms of the speed of acquisition. In fact, with regard to the rate of acquisition, the opposite holds true, and adults become better (faster) language learners, though young learners usually catch up with them in the long run (Snow & HoefnagleHohle, 1978). here are of course other fallacies regarding the ineficiency of adult language learners. Marinova-Todd et al. (2000, p. 27), for example, argue that children learning a second language go through a similarly efortful endeavor as adults. hey also argue that it is premature to attribute the diferences in language proficiency between adults and children to neurobiological factors such as lateralization and plasticity. Scovel (2000), voiced similar concerns, making the point that biological explanations such as lateralization can only account for the speech diferences between adults and children, as speech has neuro-motor etiology that can be afected by the loss of plasticity. However, it is not yet clear how lack of plasticity in adults may result in poor morphosyntactic knowledge. Furthermore, the researchers also contended that the relatively lower achievement of adults compared to children may not be due to CPH but to factors such as lack of motivation, commitment and lack of support from the context in which the new language is learned (Marinova-Todd et al.). Several studies’ findings cast doubts on the assumption that late starters are not good language learners. White and Genesee (1996) found that some French adult English-language learners were able to perform at a level not diferent from native speakers on certain language structures that tap UG access. he researcher concluded that even older language learners can achieve native-like competency, and that access to UG is not reduced through age, which they consider as evidence against CPH. Bialystock (1997) examined two studies that looked at the acquisition of French gender marking and the acquisition of English syntax by non-native speakers, and found that it was the length of residence, not the age of arrival, that accounted for the success of second language learners leading the researcher to argue against the role of maturational factors. Slavof and Johnson (1995) showed that the idea of early starters eventually overtaking late starters is not necessarily true. In their study, they compared the grammatical performances of young language learners who arrived at the US at diferent ages (7-12) and found no meaningful diferences between early and late starter. Even at the level of pronunciation, the findings of several studies show that some language learners who started their education late can attain native-like pronunciation (see for example cited in Bongaerts, 1997; Bongaerts, Mennen, & Van der Slik, 2000; Marinova-Todd, 2003; Palmen, Bongaerts, & Schils, 1997). What these studies show is that late starters are not necessarily doomed to failure. Overall, while the bulk of the studies show that early starters are generally more successful, some variation in the performance of adult language learners is also observed. 3201 EDUCATIONAL SCIENCES: THEORY & PRACTICE Such variations, while casting doubts on the strong version of CPH (Nikolov & Mihaljevic-Djigunoic, 2006), do not refute it completely. he key question is perhaps not whether or not age diferences exist, but what could be the source of the discrepancy between early and late language learning (Scovel, 2000). A lot of non-maturational factors could work alongside the biological ones afecting the learning abilities of language learners as they grow up Nikolov & Mihaljevic-Djigunoic (2006). EFL vs. ESL Settings One of the important features of the studies that have examined the relationship between age and language learning is that they are mainly conducted in ESL settings. he findings of these studies are later extrapolated to the EFL environment though these two settings have striking diferences (Muñoz, 2008, 2010). Muñoz (2008), for example, criticizes the hasty generalization of studies carried out in naturalistic learning contexts to formal learning contexts since each learning context has its own unique features. he two contexts are diferent on several grounds including quality and quantity of the input that students receive, the objective of language classes, and the proficiency level of language teachers (see Muñoz, 2008, 2010; Nikolov & Mihaljevic-Djigunoic, 2006). Several studies that compared late and early starters of English in EFL setting in Spain found that in a great number of language measures older learners outperformed the younger ones (Cenoz, 2003; Lecumberri & Gallardo, 2003; Mayo, 2003; Lasagabaster & Doiz, 2003; Muñoz, 2003). Cenoz, for example, compared the performance of three groups of language learners on diferent language proficiency tests (oral production, writing, cloze test, reading) and found that overall the general trend observed ater 600 hours of exposure was that the oldest group (fith year of secondary school, age 16 years) presents the highest level of proficiency in English followed by the intermediate group (second year of secondary school, age 12). Of all three groups, the lowest scores correspond to the youngest group (fith year of primary school, age 10). García Mayo (2003) examined Basque/Spanish language learners learning English as a foreign language to find out if the length of exposure to English has any efect on their performance in grammaticality judgment tests (GJT) and metalinguistic awareness test. he findings showed that the length of exposure resulted in the better performance of 3202 the participants in the grammaticality judgment test. Moreover, once the length of exposure held constant, it was the older learners who significantly outperformed the younger group both on GJT and metalinguistic tests. he researcher interpreted the findings as evidence that early language introduction is efective only when instruction hours are used in an eficient way. Lasagabaster and Doiz (2003) explored whether the written proficiency of Spanish students who were learning English was related to their age. he results showed that the writing scores of the participants increased with their age group. Of three diferent age groups, the oldest group outperformed the other two groups and the youngest group was significantly outperformed by both of the older learners’ groups, leading the researchers to suggest that cognitive development had a positive impact on the writing performance of language learners. Muñoz (2003) compared early and late starter Catalan-speaking Spanish students who were learning English, in terms of listening comprehension performance and the comprehension and production parts of oral interviews at two diferent times. he results show that the older group statistically outperformed the younger one on both productive and receptive measures of the interview but not in the listening comprehension tests. he research attributed the significant performance of older learners in the interview to the active role and successful negotiation that they used during the interview, while such an interactive role may not have been possible in the listening comprehension where their role was more of input recipients. here are also some studies that examined the effects of both age and length of instruction on the performance of language learners in EFL settings. Oller and Nagata (1974), at several stages, explored the performance of Japanese students who were exposed to English for diferent lengths of time, using a cloze test. In their study, students in three diferent grades were mixed. here were students who had received early exposure to English and students who had not been. heir findings showed that the diference between the early and late starters, though significant at initial stages, became insignificant in the upper levels. Similar findings were reported in Burstall, Jamieson, Cohen, and Hargreaves (1974). In their study English students learning French had diferent experiences in terms of prior exposure to the target language. While some of them had received primary school instruction in French, others had not. CEPIK, SARANDI / Early and Late Language Start at Private Schools in Turkey he results showed that the superior performance of students with prior French knowledge diminished through time in all measures except for the test of listening. Similar to Oller and Nagata (1974) these findings suggest that the advantage that early language learners enjoy at the beginning of language instruction may not last long. A diferent result, however, emerged from a study conducted recently. Domínguez and Pessoa (2005) compared 32 English native speakers studying Spanish in the sixth grade. 27 of these students had been receiving Spanish lessons since they were in Kindergarten and 5 others had only been learning Spanish for a year when the study was conducted. he results showed that except for reading tasks, the earlier starters outperformed the late starters in oral communication (listening and speaking) as well as writing. he early starters also proved to be more confident in using Spanish compared to the late starters. he researchers interpreted the findings as a sign of the potential advantages of an early start. Overall, the bulk of the studies show that when the length of instruction is constant, the late starters are at an advantage. Not many studies, however, examined the efect of diferent amount of instruction for the students who were in the same age group, and the results gained from these studies were inconclusive. While Burstall et al. (1974) and Oller and Nagata (1974) found that the earlier positive efect of longer instruction disappears through time, Domínguez and Pessoa’s (2005) findings contradicted this. Further studies may cast light in this issue. Indeed, this is the objective of the present study. It is also noteworthy that the recent history of English as a foreign language in Turkey began with the Foreign Language Education and Teaching Act of 1983 and the Primary Education Act of 1997, which paved the way for radical changes in ELT curriculum in Turkey. With respect to the foreign language education in Turkey, the most significant innovation brought by the above-mentioned laws was the integration of foreign language education into the national curriculum starting from the 6th grade. he starting age was subsequently further reduced to as low as the 4th grade, the aim being to continually adjust and standardize English language teaching in line with the norms of the EU, later the Common European Framework of Reference for Languages (CEFR). Ever since, English as a foreign language has been introduced in grade four in primary schools run by the state though some private schools may start it earlier. Method Research Questions In the lights of the issues raised above, the present study was conducted to find answers to the following research questions 1. Do early starters perform better than late starters in diferent measures of language proficiency in EFL context of Turkey? he study also intends to triangulate the findings of the research question mentioned above with the insights and opinions of language teachers about diferent types of language learners. More specifically, it intends to find the answer to the following second research question. 2. What is the perception of language teachers regarding the early and late starter students? Which groups of learners do language teachers perceive as successful? Research Model Quantitative and qualitative research models were used to collect data. Quantitative data were collected using quasi-experimental method where language learners with diferent amount of exposure to English language were selected from intact groups and their language proficiency was measured. In this sense, the length of instruction was independent variable and the proficiency level of language learners was dependent variable. Qualitative data were collected using an open-ended questionnaire through which teachers’ opinions about the success of language learners with diferent starting age for a foreign language were elicited. Context and Participants he participants of the study were 83 male and female sixth grade students at a private elementary school in Gaziantep, Turkey. he ages of the students at the time of testing were 11 and 12. he study was carried out at the end of the first semester, academic year 2011-2012. Two groups of the students were selected according to the age at which they started to learn the English. he first group represented the early beginners who had started learning English in the first grade and in kindergarten, and the second group represented the late beginners who had started to learn English language when they were in the fourth grade. he two groups were later mixed and both the late and early starters received two years of English lan- 3203 EDUCATIONAL SCIENCES: THEORY & PRACTICE guage instruction until they reached grade 6. he English instruction hours at school were 6 hours till the fourth grade and 7 hours for the fith grade and above per week. he age and estimated amount of instruction of the two diferent age groups was shown in the Table 1 below. All the participants were Turkish and started to learn English in Turkey. he data of those students who received foreign language instruction outside the school were excluded from the data analysis. Table 1. Characteristics of the Students Participating in the Study Age of first exposure Age at the time of testing Estimated amount of instruction Early age group 5-6 years old 11-12yrs old 992 hrs Late age group 9-10 years old 11-12 yrs old 406 hrs Six language teachers who had the experience of teaching the late and early starters in the same school were also selected to respond to a questionnaire aiming to elicit the teachers view over the efectiveness of instruction and early language learning. Data Collection he data was collected in two diferent ways; first, a questionnaire sent to 6 English language teachers of the school, second, a full language test given to the students. he interview posed the following questions: 1. Are there any diferences between the students who have had previous experience of learning English and those who were exposed to English in 4th grades (the late starters)? If yes, in which language areas are the diferences more noticeable? Do the diferences create any problems in terms of teaching? How do you usually deal with these problems? 2. Does the diference between these students carry over to the next levels (later classes)? If yes, how is the existing proficiency gap filled, if at all? Peer debriefing, avoiding researcher’s bias and negative case analysis were used to take care of the validity issues of the qualitative data. First, diferent stages of data collection and analyses were monitored by two researchers who had a considerable experience in qualitative research. Second, since the data for this part was collected on-line where the researchers had no direct contact with the 3204 teachers, it is believed that teachers could not be afected by the possible bias of researchers. hird, before drawing any conclusion in either direction, the data was meticulously monitored for the counter evidence. he second data was collected from a series of tests administered to the students to measure their language proficiency in all skills; first, they took a computer-based test that consisted of a cloze test, vocabulary, and reading tests. his was followed by a writing test and speaking test. he computerbased test was taken from a web 2.0-based program called 7/24 for teaching and testing language learners on line. A selection of elementary and pre-intermediate level test material was made to suit the level of the participants. he program included diferent sections mainly taping the students’ vocabulary, grammar and reading knowledge. he cloze test included some multiple-choice grammar questions, some gap-filling questions (asking prepositions, adjectives, and auxiliary verbs), picture-vocabulary matching (asking professions) and daily conversations (asking hobbies, ages, families, and parents’ profession). he vocabulary part included questions about colors, animals and some daily life activities. he reading part consisted of two short passages of beginning level followed by some comprehension questions. In the writing test, students were asked to write an imaginary story to a family in England where they were supposedly going to stay for a summer vacation, explaining about themselves, their daily life, their hobbies, their interest etc. he task was completed in a classroom with no time limit allotted. Two assessors graded each writing, and an average grade was taken as the final score. To ensure the reliability of the scores, a holistic approach suggested by Jacobs, Zinkgraf, Wormuth, Hartfiel, and Hughey (1981) was applied to grade students writing. Scales were used to measure diferent aspects of writing including content/ organization/ vocabulary/ language usage and mechanics. Following Lasagabaster and Doiz (2003), these criteria carried the weight of 30, 20, 20, 25, 5 points respectively culminating to a total score of 100. In cases where more than 25 points discrepancy between scores were observed a third assessor graded the paper. In this case, the grade of the assessor who showed the highest variance was removed and the average of two closest grades was calculated to represent the score of the participants. he Pearson correlation coeficient between the final judgments of the raters was 0.80 (p= 0.000). CEPIK, SARANDI / Early and Late Language Start at Private Schools in Turkey In the speaking test, pupils were asked to answer several questions to elicit mainly the use of simple present tense. he last question addressed the use of simple past tense. hese questions included “what is your name? where are you from? where do you live? how many people are there in your family? what does your father do? what are your hobbies? what do you do in your free time? what did you do last weekend? he performance of the participants was graded based on luency, communication, accuracy, and vocabulary choice. Similar to writing test, to guarantee reliability, two independent graders assessed the oral performance and an average grade was calculated. A similar process to writing was applied for the speaking once a 25 percent of discrepancy was observed between the scores. he Pearson correlation coeficient between the final judgments of the raters was 0.85 (p= 0.000). Results Results of Interview with the school Language Teachers he irst item of questionnaire read as 1. Are there any diferences between the students who have had previous experience of learning English and those who were exposed to English in 4th grades (the late starters)? If yes, in which language areas are the diferences more noticeable? Does the diference create any problems in terms of teaching? How do you usually deal with it? In terms of language teachers’ perception, the majority of language teachers believed that early language learners performed better in language classes. However, 5 out of 6 teachers mentioned that the superior performance of early starters were in oral communication, mainly pronunciation. Some also mentioned that earlier starters were better in terms of vocabulary. However, only two teachers out of 6 mentioned, surprisingly, that late starters were better at grammar and cognitively demanding tasks and that they were very eager at language learning. Also, all of them agreed that the early starters have the self-confidence to express themselves easily. For example, one of the language teachers wrote: I have experience working with second and fourth grade students. I think early starters are better than late starters. hey are more interested in new languages. As a child, they want to learn everything. here are some diferences between early learners and late learners in terms of their language skills. For example; early learners are better at listening and speaking. heir pronunciation is much better than late learners. hey can pronounce the words better. Most of them acquire the language because they cannot write and read properly. hey memorize what they hear. heir brains are like an empty disc. hey like repetition. On the other hand, late learners are better at grammar. Because they start to learn a new language ater they learn their mother tongues’ grammatical rules. Knowing grammatical rules makes their job easier. hey don’t want to speak in English because they are afraid of making mistakes, but early learners say what they want. hey are not afraid of making mistakes. he interview data suggested that the teachers consistently argued that “the level system” through which students are assigned to their appropriate levels should be applied, so as not to face the expected and predicted problems in the language classes in the schools. Here, two teachers comment on the diferences between the two groups of the students and mention how they deal with the problems; ‘he students who start learning English from the 1st grade use English better than those who start later. hey practice English more actively and willingly. hey have confidence and their pronunciation is better. But those who start learning English at 4th grade are more enthusiastic and dynamic. hey are curious and have pleasure to use English. And, of course, these diferences create problems. But we deal with them by putting students into diferent level and teaching them according to their knowledge.’ he second question asked how persistent the differences between students’ proficiency levels were. It read: 2. Does the diference between these students carry over to the next levels (later classes)? If yes, how is the existing proficiency gap filled, if at all? Generally teachers believe that the gap can be filled if the students are well motivated and supported by their teachers. But the most significant thing is to group the students according to their level. Another language teacher expressed his ideas as follows: As long as the English teacher who is in charge of the new learners of English can manage to motivate these students, the 3205 EDUCATIONAL SCIENCES: THEORY & PRACTICE gap between the old comers and the new comers close approximately in one educational year. But this may happen if the setting system is applied. If there is no setting system and the students are mixed in one class, then the new comers need to be given extra courses to fill the gap. Some teachers say that it is inevitable to experience the problem at the later classes, but teacher support may minimize the problem. Here is how an experienced teacher commented with this regard: As far as I have experienced during my teaching life, students who have problem in the early years of learning English carry over all these problems to their further learning. But if the teacher gives more exposure to the late starters, they may solve the problem. Results of Proficiency Language Tests Below are the descriptive statistics collected for the diferent groups of the students in diferent language measures. Table 2. Descriptive Statistic of Late Early Starters in Diferent Language Proficiency Tests Speaking Groups N Mean Early starter 47 67 SD 19 Later starter 33 65 17 Groups N Mean SD Early starter 49 79 15 Later starter 34 78 14 Groups N Mean SD Early starter 19 15 23 Later starter 29 13 25 Groups N Mean SD Early starter 19 34 39 Later starter 29 39 43 Groups N Mean SD Early starter 19 61 40 Later starter 29 64 41 Writing Reading Cloze test Vocabulary In order to detect whether the diferent amount of instruction that late and early starters had received 3206 resulted in diferent performance of the language learners in diferent language measures independent T-tests (SPSS 16) were run on data. he results of data analysis revealed that in none of the measures used in the study the diference between two groups reached a significant level (Speaking: F=0.33 p=0.855, Writing: F=0.104 p=0.748, Reading: F=0.30 p=0.864, Close test: F=0.238 p=0.628, Vocabulary: F=0.002 p=0.965). Discussion he first research question asked if the students who started language learning earlier had any advantages over the language learners who started later. he findings show that early starters had no advantage over late ones in any of the measures used in the study. hus our findings confirm earlier studies of Burstall et al. (1974) and Oller and Nagata (1974) and contradict those of Domínguez and Pessoa (2005). It is also in line with the majority of studies conducted in EFL settings whose findings did not reveal any advantage for an early start (Cenoz, 2003; Lecumberri & Gallardo, 2003; Mayo, 2003; Lasagabaster & Doiz, 2003; Muñoz, 2003). he first reason to account for the findings is the learning context in which the study was conducted. As mentioned earlier, the exposure received in EFL contexts both in terms of quality and quantity is not significant compared to ESL settings (Muñoz, 2010). Hence, it could be argued that the amount and intensity of the instruction that younger learners received may not have been suficient enough to create a significant diference. he diference between the early and late starters was about 586 hours. One may argue that this amount of extra time may not to be enough to create any noticeable change in an EFL setting. Another possibility is that the types of tasks that young learners had dealt with may not be cognitively challenging, and as such might have failed to stretch their inter-language beyond rudimentary levels (see Nikolov & Mialjevic-Djigunoic, 2006 for a similar argument). he efectiveness of instruction ofered to language learners might also depend on the types of processing employed by the language learners. Young learners learn mainly implicitly while adults’ learning is mainly explicit (DeKeyser, 2010). In this sense, the mechanism used by older language learners fit the formal instruction used in the great majority of language classes in EFL settings (Muñoz, 2008) including Turkey. In other words, the late starters’ learning mechanism might have allowed them to CEPIK, SARANDI / Early and Late Language Start at Private Schools in Turkey benefit from form-focused language classes and hence they managed to cover the gap. Another reason for the present findings is that the early and late starters were mixed in the classes. As mentioned previously, early starters who had had 3 years of instruction beginning from the first grade were mixed with the other language learners who had no former instruction upon their arrival into the new language setting. It could rightly be argued that school system worked in such a way that the preliminary higher proficiency of early starters was leveled so that the instruction could be conducted in a smooth manner (see Stern, 1976; Muñoz, 2008 for a similar argument). he following statement from one of the teachers relects this point: When I ask a question, early learners reply it in few seconds. his situation went on a few months, then, I started to give extra homework to the late learners. We memorized new songs with them…. Now they are good at English especially in grammar. his statement shows that the instruction may not have been distributed fairly between the diferent age groups and while the earlier language learners were not challenged, the other group received some extra instruction. Psychological variables may also afect the efectiveness of instruction. Moyer’s (2004) study on successful immigrants in Germany showed that age factor and length of residence contributed little to the final attainment in learning the German language. 74% of the variance in attainment was related to psychological factors. Marinova-Todd (2003) argues that the right time for the introduction of high quality instruction is when children are motivated, not when they are necessarily young. In our case, it is possible that at least some of the older students were motivated to cover the existing gap and catch up with the earlier language learners up with the earlier language learners. he second research question asked about theperception of language teachers in respect of the efectiveness of early language learning. As the results of the study showed the majority of language teachers believed that early language learners were better at pronunciation and vocabulary. However, the findings of our study showed no diference in language proficiency of the two groups in terms of any measures used in the study. One reason to account for the seemingly contradicting results as far as pronunciation was concerned was that in speaking tests that we applied we did not include a component to assess the pronunciation of the students. his was mainly so because the proficiency level of the students was low and the diference between the accents of the language learners did not appear noticeable at the time of the testing. he limited range of vocabulary chosen for the present test compared to what students usually use in the classroom might be another reason for the existing discrepancy. he vocabulary used in the study was limited to some colors, animals and daily life activity, which may not be broad enough to relect the language areas that students encounter in their language classes. However, it is interesting to note that the language teachers believed that the late starters were good at grammar, and cognitively challenging skills such as reading and writing. Obviously, these skills and language areas match the brain maturation that these learners possessed at the time of language testing. Conclusion and Implications hough the common belief is that early language starters have some advantages over the later ones, the results of the present study, as far as the students’ language proficiency is concerned, do not indicate that this is necessarily the case. he early language learners in our study failed to perform better in any of language measures. he lack of discernable proficiency diferences in the measures applied implies that intensity and quality of programs used for the early starters may not be efective enough to create a profound change in their linguistic competence. his underscores the need for the reevaluation of the materials and syllabi used for early starters so that the time and resources dedicated to this group are used in a more eficient way. However, one should be careful not to misinterpret the findings of the present study as evidence that starting age has no efect on language learning. For one thing, the testing instruments applied in the present study did not allow us to measure the possible pronunciation supremacy (and to a lesser extend the vocabulary advantage) of the early starters, as suggested in the comments of language teachers. Furthermore, there was no measurement of the willingness of early language learners to initiate conversations and interactions in the classrooms as the teachers ascribed to early starters. As Cenoz (2003) points out, early language learners has more positive attitude towards language learning, which in the long-run may result in a better language development. 3207 EDUCATIONAL SCIENCES: THEORY & PRACTICE Moreover, one may argue that early language instruction provides the opportunity to handle the difficult process of learning a new language in an extended time as opposed to an intensive one. In this sense, the earlier starters have more time to process the language implicitly, which may result in a more profound and engrained learning, even though this is not readily relected in their performance in the language tests they take in their classes. It also introduces children to a new culture and receives the positive attitudes of students (see also Lasagabaster & Doiz, 2003). Overall, further studies are required to examine if early starters become more successful language learners in the later stages of their education, in the secondary school for example. Finally, it should be born in mind that age is only one of the factors involved in language learning. here are other variables at work, alongside age, such as the length of the instruction, methods of teaching, students’ motivation, ability to simulate a positive learning environment. that could afect the final linguistic attainment of students. he introduction of the new language in the early ages can be more efective if it is accompanied by professional teaching instructors, appropriate materials, right methodology, suficient time, etc. Otherwise, to expect that an early start on its own would lead to ultimate success in foreign language learning is too unrealistic. References Birdsong, D. (1992). Ultimate attainment in second language acquisition. Language, 68 (4), 706-755. Birdsong, D. (1999). Second language acquisition and the critical period hypothesis. Lawrence Erlbaum. Bongaerts, T. (1999). Ultimate attainment in L2 pronunciation: he case of very advanced late L2 learners. In D. Birdsong (Ed.), Second language acquisition and the critical period hypothesis (pp. 133-159). Mahway, NJ: Erlbaum. Bongaerts, T., Mennen, S., & van der Slik, F. (2000). Authenticity of pronunciation in naturalistic second language acquisition: he case of very advanced late learners of Dutch as a second language. Studia Linguistica, 54 (2), 298-308. Burstall, C., & Jamieson, M., Cohen, S., & Hargreaves, M. (1974). Primary French in the balance. NFER Slough. Cenoz, J. (2003). he inluence of age on the acquisition of English: General proficiency, attitudes and code mixing. In M. P. G. Mayo and M. L. G. Lecumberri (Eds.), Age and the acquisition of English as a foreign language (pp. 77-93). Clevedon: Multilingual Matters. DeKeyser, R. M. (2000). he robustness of critical period effects in second language acquisition. Studies in Second Language Acquisition, 22 (4), 499-534. 3208 Domínguez, R., & Pessoa, S. (2005). Early versus late start in foreign language education: Documenting achievements. Foreign Language Annals, 38 (4), 473-483. Lecumberri, M. L. G., & Gallardo, F. (2003). English FL sounds in school learners of diferent ages. In M. P. G. Mayo & M. L. G. Lecumberri (Eds.), Age and the Acquisition of English as a foreign language (pp. 115–35). Clevedon: Multilingual Matters. Mayo, M. P. G. (2003). Age, length of exposure and grammaticality judgments in the acquisition of English as a foreign language. In M. P. G. Mayo & M. L. G. Lecumberri (Eds.), Age and the acquisition of English as a foreign language (pp. 94-114). Clevedon: Multilingual Matters. Gass, S. M., & Selinker, L. (2008). Second language acquisition: An introductory course. Taylor & Francis. Heighington, S. (1996). Case study 5. Taking up the challenge. In A. Hurrell & P. Satchwell (Eds.), Relections on Modern Languages in Primary Education (pp. 55-61). London: CILT. Herschensohn, J. (2008). Developing I-language in L1 and L2. In P. Masullo & C-H. Huang (Eds.), Romance Languages: Structure, interfaces, and microparametric variation (pp. 157-174). Amsterdam/Philadelphia: J. Benjamins. Jacobs, H. L., Zinkgraf, S. A., Wormuth, D. R., Hartfiel, V. F., & Hughey, J. B. (1981). Testing ESL composition. Newbury: Rowley. Johnson, J. S., & Newport, E. L. (1989). Critical period efects in second language learning: he inluence of maturational state on the acquisition of English as a second language. Cognitive psychology, 21(1), 60-99. Jufs, A., & Harrington, M. (1995). Parsing efects in second language sentence processing. Studies in Second Language Acquisition, 17 (04), 483-516. Koru, S., & Akesson, J. (2011). Turkey’s English deficit. Retrieved June 1, 2012 from http://www.tepav.org.tr/upload/ files/1324458212 1.Turkey_s_English_Deficit.pdf. Lasagabaster, D., & Doiz, A. (2003). Maturational constraints on foreign language written production. In M. P. M. Mayo & L. G. Lecumberri (Eds.), Age and the acquisition of English as a foreign language (p.136-160). Clevedon: Multilingual Matters, Long, M. H. (1990). Maturational constraints on language development. Studies in Second Language Acquisition, 12 (03), 251-285. Long, M. H. (2007). Problems in SLA. Lawrence Erlbaum. Marinova-Todd, S. H. (2003). Comprehensive analysis of ultimate attainment in adult second language acquisition. Unpublished doctoral dissertation, Harvard University. Marinova-Todd, S. H., Marshall, D. B., & Snow, C. E. (2000). hree misconceptions about age and L2 learning. TESOL quarterly, 34 (1), 9-34. Mayo, M. P. G., & Lecumberri, M. L. G. (2003). Age and the Acquisition of English as a Foreign Language (4). Multilingual Matters Ltd. Moyer, A. (1999). Ultimate attainment in L2 phonology. Studies in Second Language Acquisition, 21 (1), 81-108. CEPIK, SARANDI / Early and Late Language Start at Private Schools in Turkey Moyer, A. (2004). Age, accent, and experience in second language acquisition: An integrated approach to critical period inquiry (7). Multilingual Matters Ltd. Muñoz, C. (2008). Symmetries and asymmetries of age efects in naturalistic and instructed L2 learning. Applied Linguistics, 29 (4), 578-596. Muñoz, C. (2010). On how age afects foreign language learning. Retrieved June 5, 2012, from http://www.enl.auth.gr/ gala/14th/Papers/Invited%20Speakers/Munoz.pdf. Nikolov, M., & Mihaljevic-Djigunovic, J. (2006). Recent research on age, second language acquisition, and early foreign language learning. Annual Review of Applied Linguistics, 26, 234-260. Oller, J. W., & Nagato, N. (1974). he long-term efect of FLES: An experiment. Modern Language Journal, 58, 15–19. Palmen, M.-J., Bongaerts, T., & Schils, E. (1997). L’authenticité de la prononciation dans l’acquisition d’une langage étrangère au-delà de la période critique: Des apprenants néerlandais parvenus à un niveau trés avancé en français. Acquisition et interaction en langage etrangère, 9, 173–191. Patkowski, M. S. (1980). he sensitive period for the acquisition of syntax in a second language. Language Learning, 30 (2), 449-468. Pufahl, I., Rhodes, N. C., & Christian, D. (2001). Foreign language teaching: What the United States can learn from other countries. Foreign Language Annals, 34 (3), 281-283. Scovel, T. (2000). A critical review of the critical period research. Annual Review of Applied Linguistics, 20 (1), 213-223. Singleton, D. M., & Ryan, L. (2004). Language acquisition: he age factor (s). Multilingual Matters Ltd. Snow, C., & Hoefnagle-Hohle, M. (1978). he critical age for second language acquisition: Evidence from second language learning. Child Development, 49, 1114–1128. Stern, H. H. (1976). Optimal age: Myth or reality? Canadian Modern Language Review, 32 (3), 283-294. Tahta, S., Wood, M., & Loewenthal, K. (1981). Foreign accents: Factors relating to transfer of accent from the first language to a second language. Language and Speech, 24 (3), 265-272. Torras, M. R., Tragant, E., García, M. L. (1997). Croyances populaires sur l’apprentissage précoce d’une langue étrangère. In C. Muñoz, L. Nussbaum & M. Pujol (Eds.), Acquisition et interaction en langue étrangère 10 (pp. 127-158). White, L., & Genesee, F. (1996). How native is near-native? he issue of ultimate attainment in adult second language acquisition. Second Language Research, 12 (3), 233-265. Yeni-Komshian, G. H., Flege, J. E., & Liu, S. (2000). Pronunciation proficiency in the first and second languages of KoreanEnglish bilinguals. Bilingualism Language and Cognition, 3 (2), 131-150. Yeni-Komshian, G. H., Robbins, M., & Flege, J. E. (2001). Effects of word class diference on L2 pronunciation accuracy. Applied Psycholinguistics, 22, 283–299. 3209