L2 learner age from a contextualised perspective

In this qualitative study the author focuses on age effects on young learners’ L2 development by comparing the L2 learning processes of six young learners in an instructed setting: three who had started learning English as L2 at age 6/7 and three who had started at age 9/10. Both earlier and later young beginners were followed for three years (during their second, third and fourth year of learning English). The participants’ L2 development was measured through their oral output elicited by a two-part speaking task administered each year. Results of the analyses are interpreted taking into account each learners’ individual characteristics (learning ability, attitudes and motivation, self-concept) and the characteristics of the context in which they were learning their L2 (attitudes of school staff and parents to early L2 learning, home support, in-class and out-of-class exposure to L2, socio-economic status). The findings show that earlier and later young beginners follow different trajectories in their L2 learning, which reflects different interactions which age enters into with the other variables.


Introduction
Although it has been one of the most interesting issues in SLA and L2 teaching research for a long time, the age factor is still the main focus of many researchers' interest.Most studies to date have discussed the age factor with reference to the critical period hypothesis (Lenneberg, 1967;Penfield & Roberts, 1959).The findings of these studies are not unanimous.Some (e.g., DeKeyser, 2003;DeKeyser & Larson-Hall, 2005;Hyltenstam & Abrahamsson, 2001, 2003) conclude that there is a critical period beyond which it is difficult or even impossible to master L2, others (e.g., Bialystok, 2001;Birdsong, 2005;Marinova-Todd, Marshall, & Snow, 2000, 2001;MacWhinney, 2005;Moyer, 2004) claim that the critical period does not really exist, while still others maintain a more balanced position (e.g., Muñoz, 2006;Scovel, 1988Scovel, , 2000;;Singleton, 1989Singleton, , 2001;;Singleton & Ryan, 2004).In recent overview chapters of the age factor (e.g., DeKeyser, 2012;Muñoz & Singleton, 2011;Nikolov & Mihaljevi Djigunovi , 2006, 2011) it is possible to discern some new trends in viewing the age factor.They reflect, on the one hand, a move away from the idea of biological maturation implied by the critical period hypothesis as the explanation of differences in SLA between younger and older learners and, on the other, a realization that age is difficult to disentangle from other variables and its impact on SLA can be better understood if we take into account its interactions with other factors.Thus, the term age effects is nowadays used increasingly more often than critical period in discussions of the age factor in SLA.
In this paper I will move away from the prevalent approaches to age effects.More specifically, I will not aim at finding out which age is the best to start foreign language learning in terms of the rate of acquisition or the ultimate attainment, as most studies currently do.Since early L2 programmes have mushroomed in pactically all corners of the world (Nikolov & Mihaljevi Djigunovi , 2011), the start age has become something of a given because education policy makers decide on the introduction of L2 at a particular age irrespective of what research findings suggest and, often, only because of strong parental pressure to start early.Thus, the urgent problem that research on age effects can help solve concerns the understanding by all parties involved of the L2 learning by young learners at different early ages.Findings of such research can contribute to broadening our theoretical insights about early L2 learning and, at the same time, raise the awareness of L2 teachers and other stakeholders about the relevance of different factors which impact early instructed L2 learning and teaching.
A number of relevant variables confounding the age factor have emerged in recent literature.Thus, Muñoz and Singleton (2011) mention young learners' attitudes and motivation; amount and quality of input; amount, intensity and diversity of L2 contact; frequency of out-of-school use of L2 and the like; stressing that natural and instructed settings imply differences in relevance of these variables.Moyer (2004), for example, draws attention to the potential importance of the quality of L2 experience as well as to the level of interactivity required during exposure to L2.
Mihaljevi Djigunovi and Lopriore (2011) suggest that learner self-concept plays a role in early L2 learning, as well as home support.The socio-economic status has also been pointed out as a relevant variable in early L2 learning (Muñoz & Lindgren, 2011).In instructed learning settings L2 learners' classroom behaviour has also been found as a valuable source of information on the L2 learning process (Mihaljevi Djigunovi , 2009).DeKeyser (2012) adds increasing role of L1 and decreasing role of schooling.The above mentioned variables do not present a full list of factors which interact with age and potentially determine age effects on the processes and products of early L2 learning.It is very likely that the list will increase with the increase in our understanding of the complex role of age in SLA.
As has often been pointed out (e.g., Enever & Lopriore, 2014;Nikolov, 2009), deeper insights into early L2 learning require a longitudinal approach to research.In spite of the fact that quite a few studies have been carried out in the field, there have not been many which followed this research paradigm.Some recent notable exceptions are longitudinal projects carried out in Spain (García Mayo & García Lecumberri, 2003;Muñoz, 2006), Ireland (Harris & Conway, 2002), Scotland (Low, Duffield, Brown, & Johnstone, 1993) and Croatia (Vilke & Vrhovac, 1993, 1995;Vrhovac, 2001), as well as the transnational Early Language Learning in Europe (EL-LiE) (Enever, 2011) study which involved seven European country contexts.

Context of the study
The study was carried out in the Croatian socio-educational context.It is characterised by a long tradition in L2 learning.For decades, the starting age used to be grade 4 of primary school (age: 9/10 years), and in 2003 it was lowered to the very beginning of education: grade 1 of primary school, when learners are at the age of 6/7 years.All pupils follow the same national curriculum, which lays emphasis on developing positive attitudes to learning second languages, motivation for language learning and oral skills.Explicit teaching of grammar is discouraged.
Two types of teachers are considered qualified to teach English: specialist English teachers who have a university degree in English language and literature and have been trained to teach English to any age and type of learners, and class teachers with a minor in English who have specialized in teaching English to young learners.

Aims
This study aimed at getting an insight into age effects on young learners' development in L2 English by contextualising age at the learner and learning context levels.Following the earlier and later young beginners' language development over three years I wished to paint as comprehensive a picture of age effects as possible by taking into account their attitudes and motivation, L2 self-concept and classroom language learning behaviour, as well as their L2 exposure in class and out of class, home support and socio-economic status.

Methodology
The sample in this qualitative study included 6 Croatian young learners of English as L2.Three of them started learning English in grade 1 (earlier beginners) and 3 started in grade 4 (later beginners).They were followed during their second, third and fourth years of learning.In terms of age this means that the ealier beginners were followed from the age of 7/8 years till 9/10, while later beginners were aged 10/11 when the study started and were 12/13 years old by the end of the study.
The participants' L2 development was measured by means of speaking tasks involving a picture description and a personalised interview related to the picture.In their second year of L2 learning the young learners were presented with a picture of a family house and could see its different rooms and family members doing different things.They were first asked to describe the picture and then to talk about the place where they lived and about their family.The third year speaking task involved decribing a picture of a living room and a dining room in which family members were watching TV, or eating, or studying.This was also followed by discussing the related aspects of the participants' family life.In the final year of the study (fourth year of L2 learning) the participants were shown a four-part picture depicting the hall, the living room, the bathroom and the bedroom of a house.They were asked to first describe everything they could see and were then interviewed about what they thought about the house, whether they would like to live in it, and about their favourite room at home.During piloting, these pictures were found to be good elicitation tools: They provided enough stimulus for the participants in both age groups to produce orally at their current level.The personalized interview part was included to see if there were differences in oral production based on a visually guided and more structured stimulus and a more authentic-like and free use of L2.
Information about the young learners' attitudes and motivation for learning English and their L2 self-concept was gathered through interviews with the learners and through the parents' questionnaire.These instruments were taken over from the ELLiE project (for details see Enever, 2011).The participants were asked each year what their favourite school subject was (and if they liked English in cases when they did not list it among the favourite subjects) and what they liked and/or disliked about their English classes.Data on the participants' language learning behaviour during English classes and on the exposure to English through classroom teaching was collected using the observation schedules also developed by the ELLiE team.The already mentioned parents' questionnaire supplied information about the participants' out-of-school exposure and their socio-economic status, which helped to triangulate the data obtained from the young learners themselves.
The participants were observed in their classrooms nine times during the three years of the study.The speaking tasks and the interviews were administered at the end of each grade.The parents filled in the questionnaire in the second and in the final year of the study.Before the start of the study the participants' teachers were asked to describe each learner in terms of their language learning ability (high-ability, average-ability, low-ability).
Two independent raters assessed the participants' oral production.They used a 5-point scale on which they assessed the participants' task achievement, vocabulary, accuracy and fluency in relation to the respective curricular aims for a particular year of study.Their inter-rater reliability was .94.

Results
In this section I will first describe our participants' L2 development as evidenced by their oral production over three years.Then I will try to build up a profile of each participant based on their individual learner characteristics and on the key features of the context they learned English in.This will be followed by looking into the interaction of the L2 development and the individual and contextual factors investigated.The earlier beginners will be referred to as EB1, EB2 and EB3.The three later beginners will be named LB1, LB2 and LB3.

EB1
This was a female learner who displayed average performance in the second year of learning and then performed well above average in years 3 and 4, as shown in Figure 1.

Figure 1 Overall oral performance of EB1 during three years
A closer look at her performance on individual subtasks over the three years depicted in Figure 2 shows that the observed trend in her oral performance did not take place in all aspects of oral production.In the picture description part of the oral task the command of vocabulary peaked in the third year and then dropped, while fluency was highest in the fourth year.In the personalized interview each of the investigated aspects of her oral performance developed in the same way: They were lowest in year 2 and then rose in year 3, remaining at the same increased level in year 4. If scores on the two subtasks are compared, it can be seen that task achievement and accuracy followed the same pattern (rising, then stabilizing) in both subtasks, while fluency showed a rising trend already from year 2 in personalized interview and only in year 3 in picture description; vocabulary showed greater variability in picture description compared to personalized interview, where no drop was observed.

Figure 2
Oral performance of EB1 broken down by criteria and by subtask EB1 attended a small town school which had a long tradition in early foreign language learning, was well equipped and in which both the school principal and teaching staff considered early beginning of a foreign language important.The school library boasted a nice collection of English books, picture books and dictionaries as well as DVD players, computers and an IWB, all of which were available to be used during English classes.This participant was taught by a generalist teacher with a minor in English, who was very enthusiatic about teaching young learners.Her teaching approach was communicative, and she used English close to 90 per cent of class time.The number of students in the class was 20.In the home, the participant had no English books or dictionaries but was exposed to the language through foreign TV series and music.Her father had secondary education, while her mother had finished only primary school.Only her mother spoke a little English, and it was her who often helped her with the homework.Both parents supported her learning of English; they reported she enjoyed speaking it and each year became more confident about her knowledge.Once, during summer, she met some foreign tourists at the seaside and managed to talk to them in English.This was a very pleasant experience for her.
At the start, EB1 was assessed by her English teacher as an average-ability learner.Throughout the three years this participant listed English as her favourite (years 2 and 3) or one of the favourite school subjects (year 4).She also reported enjoying everything about her English classes, and pointed out games, songs and learning new words as the most enjoyable activities in years 2 and 3, and stories and listening in year 4. Her self-concept developed in the positive direction: In year 2 she explained she was just as good as others in class, while in years 3 and 4 she claimed she was better than her classmates because she could understand everything in class and could remember new things faster than her friends.During English classed she showed high interst and was very attentive and engaged.

EB2
This male participant's overall oral performance, which is shown in Figure 3, was on the constant increase.It was not too high in year 2 but rose to average in year 3 and reached the above average level in year 4. When considering the different aspects of oral performance in both subtasks together, it can be seen that this participant's performance increased over the three years in all the aspects included except fluency.In picture description the most obvious rising trend could be observed in task achievement, while vocabulary and accuracy increased in the final year only.Accuracy was the weakest point in this learner's performance in this subtask.Similar trends could be observed in his performance in personalized interview, with somewhat lower scores overall.Interestingly, in years 2 and 3 this participant displayed higher accuracy in the interview than in the picture description subtask.
EB2 attended a village school.His class included 22 students.The attitudes towards early learning of foreign languages were generally positive among the school staff as well as parents.The teacher was a generalist teacher with a minor in English.She was very enthusiastic about teaching young learners and had high rapport with them.Her teaching methodology was age-appropriate.She used English approximately 60 per cent of class time.What characterised her use of L2 and L1 was that her code-switching was strategic: She would use L1 in order to make a particular L2 word or structure more salient.She often paid attention to the English her learners encountered out-of-school and did her best to integrate it into her classroom teaching.At the beginning of the study the participant was assessed by the teacher as a low ability learner.
Both of the participant's parents had primary education and neither could speak English.It was his older brother that the participant turned to for help when necessary.The parents supported the participant's learning of English and were very happy when he got good grades.They reported that their son was very proud about his English but at the beginning was too shy to use it in the presence of his family.He became more confident from year to year.EB2 did not have access to books in English at home or to the Internet.Most of his exposure to L2 was through English speaking TV programmes and music.
EB2 did not mention English as a favourite subject in year 2. In year 3 he listed it as one of his favourite subjects, while in year 4 it featured as the favourite school subject.He enjoyed all the classroom activities but from year 3 on he singled out learning new English words and doing listening tasks as the best.At the end of year 2, EB2 perceived himself as less good than his classmates, in year 3 he claimed he could learn as fast as everyone else, while in year 4 he said he was better than most of his classmates and offered his high grades as evidence.During English classes he regularly displayed high interest in all the tasks and was generally very attentive and engaged.

EB3
This participant was a boy whose oral performance, showed in Figure 5, although well above average, showed the most variability.The performance in year 3 displayed a drop compared to year 2, and rose again in year 4.

Figure 5
Overall oral performance of EB3 during three years If the production is analyzed according to the four criteria and separately for the two subtasks, shown in Figure 6, again different trends can be seen.As far as the picture description subtask is concerned, a drop in the scores on the vocabulary and accuracy scales could be observed in year 3, while fluency increased in the same year.In personalized interview, the same variability was observed for accuracy, while fluency remained at the same level; vocabulary showed a rising trend after year 3. Task achievement was high in all three years and in both subtasks.
Figure 6 Oral performance of EB3 broken down by criteria and by subtask EB3 attended a big city school with a tradition in early foreign language learning.His class was quite large and comprised 29 learners.Most staff in the school entertained positive attitudes to early foreign language learning.The school was averagely equipped: There was a collection of children's books in English in the school library, the teacher had access to a DVD player, a computer and a data projector.Several dictionaries were also stocked in the library.The teacher was a specialist English teacher who was not specifically trained to teach young learners.She liked teaching children and attended in-service workshops in order to upgrade her competences.Her teaching generally reflected the communicative approach to foreign language learning, but she insisted on accuracy more than the other teachers in this study.Since the class was rather large, there were sometimes discipline issues she had to cope with; she also had to accommodate her teaching to two pupils with special needs in the class.The teacher used English between 50 and 70 per cent of class time.In the home the participant's parents generally supported his learning of English.However, there were no books or dictionaries the young learner could use and, up untill year 4, he did not have access to the Internet.The father had secondary education and the mother had finished primary school.Both parents had learned English in school but reported poor knowledge of the language.All of their other three children learned English and helped one another when necessary.
EB3 was assessed by his teacher as a high ability learner.Both his motivation and self-concept showed variability over the three years of the study.English was one of his favourite subjects in year 2, was not mentioned at all in year 3, and turned into the one and only favourite subject in year 4. His self-concept followed similar changes: in years 2 and 4 the participant claimed he was better at English than his classmates, while in year 3 he said we was neither better nor worse than others.As far as classroom activities are concerned, his constant favourites were listening to stories, games, learning new words and role plays.In year 3 he seemed less enthusiastic about his English classes than in other years but still reported liking practically everything.Observation of his classroom language behaviour reflected his affective dispositions: In contrast to years 2 and 4, in year 3 EB3 displayed average interest but less attention and engagement.That year he also seemed distracted by his unruly classmates whose behaviour often caused the teacher to stop the class and deal with the discipline issues at hand.

LB1
This was a female learner.Her oral production showed variability from year to year as well as from one aspect to another within the same year.She perforned best in year 3. Figure 7 presents the observed overall variability.

Figure 7 Overall oral performance of LB1 during three years
As can be seen in Figure 8, in the picture description subtask LB1 performed best in year 3, after starting rather poorly in year 2. Except for task achievement, her performance displayed an inverted U-shape development.In the interview subtask, the same trend could be observed in all aspects except vocabulary.It is interesting to note that, overall, this participant performed better in personalized interview than in picture description.

Figure 8
Oral performance of LB1 broken down by criteria and by subtask LB1 attended a town school that can be described as a typical primary school for the Croatian socio-educational context: It was equipped with basic teaching aids (specialized foreign language classroom with CD and DVD players, no IWB, and a number of English books and dictionaries in the school library) with the staff entertaining positive attitudes to early foreign language learning.She was taught by a specialist English teacher who applied age appropriate methodology and mostly used English in class.Although the teacher believed early foreign language learning was highly beneficial, she said she preferred teaching more mature learners.
At home LB1 had DVDs in English and a dictionary.She got access to the Internet during year 3 and made use of her knowledge of English while surfing.Although her mother could speak English, she would mostly ask her sister for help w hen do ing ho m ew o rk o r preparing fo r a test.T here w as a m aj o r difference in her parents' educational level: Her mother had a university degree, while her father completed only primary education.LB1 took pride in her knowledge of English, liked to speak the language and was not shy to use it.By year 3 she had already experienced communicating in English with a foreigner and reported feeling happy about being able to use her English.
She did not list English as a favourite subject till year 3.At first she claimed she liked everything in her English classes, but in year 3 she liked role-playing the most and said she disliked having to wait for everyone to finish the tasks they were still doing after she had finished hers.Her classroom language behaviour suggested slightly negative trends in terms of all the aspects observed from year 3 on.
LB1 could not decide in year 2 whether she was just as good at English as her classmates or better than them, but by year 3 she was sure she was better than others in class.Her self-concept remained positive in year 4 as well, when she claimed she could understand new things faster than her classmates.Still, she perceived English as getting more difficult each year.

LB2
The second later beginner was a boy whose oral performance excelled in all three years and according to all the investigated criteria, as can be seen from Figure 9.

Figure 9
Overall oral performance of LB2 during three years LB2 attended a city school which offered primary education in English besides the regular national curricula in Croatian.Thus it attracted foreign pupils such as children of foreign diplomats or businessmen, which gave the local children ample opportunity to use English during the breaks.LB2 was exposed to English quite extensively in the home too through watching TV programmes in English, listening to music in English and using the Internet.He also had contacts with English speaking people who visited his parents or whom he met at the seaside during summer vacations.His experience in communicating in English was highly positive.According to his parents, he felt confident using English with foreigners.The parents had university level education and used English at work.They rarely helped their son with homework or studying because he needed no help.
LB2 was taught English by a qualified teacher who used English 50 per cent of class time on average.The participant did not think highly about his teacher's English and claimed that her pronunciation was not too good but that her grammar was excellent.His self-concept was very positive during all the three years: He perceived himself as the best pupil in class.Although English was never among his favourite subjects because it was too easy and not challenging enough, LB2 enjoyed himself when they read stories or worked with comic strips.Still, he was often bored during his English classes and wondered how his classmates could find English difficult.His teacher assessed his language learning ability as high.

LB3
This participant was a boy whose performance was generally not high and also varied over time, as shown in Figure 10.It was assessed as below average in year 2, then rose to above average in year 3 to drop slightly again, but remaining above average.

Figure 10 Overall oral performance of LB3 during three years
If this participant's performance is comapred along the four criteria in the two subtasks with the help of Figure 11, it can seen that his performance varied over the three years in both cases.In picture description, task achievement remained at the same level throughout the three years, while vocabulary, accuracy and fluency increased in year 3, and remained at that level in year 4. In the personalized interview subtask the observed trends were somewhat different: While task achievement and vocabulary scores increased in year 3 and then stabilized, accuracy was highest in year 3 and then dropped to the year 2 level; fluency first increased in year 3 and then dropped in year 4 to the lowest point.
Figure 11 Oral performance of LB3 broken down by criteria and by subtask LB3 attended the same town school as LB1 but had a different English teacher.It was a specialist teacher who had not been trained to teach young learners.The teacher assessed him as an average-ability learner.His exposure to English in class was average (the teacher used English about 60 per cent of class time), as was the out-of-class exposire: He would sometimes watch films in English on TV, listen to music and, in year 4, he gained access to the Internet and used English when using Facebook or playing games.LB3 got a chance to communicate in English with a foreigner only in year 4, and reported feeling highly anxious while trying to show the way to a German tourist visiting his town.His parents had secondary level education and could speak some English.When he needed help, he would turn to his older brother.The parents were happy about his learning English and would praise him whenever he would get a good grade.They claimed their son liked learning English but was not very confident when using it and found it rather hard to learn and use.
LB3 never listed English among his favourite subjects but, when prompted, he reported that he liked English too, stressing that he liked other subjects better.He generally preferred structed learning activities and enjoyed listening to English texts in class, but disliked writing and doing tests.During his English classes he generally showed high interest, but it was not always accompanied by engagement.LB3's self-concept changed over the three years.In year 2 it was negative, and he believed he was worse at English than his classmates; in years 3 and 4 he claimed he was just as good as others.

Discussion
EB1 displayed increase in her L2 development at the start, which then stabilized at the above average level.Interestingly, it was fluency that showed the most consistent rise over the three years.I connect this with the high L2 use by her teacher during English classes combined with the opportunity the participant had to use English in the home because her mother spoke English and encouraged her to use it.It is also interesting that vocabulary showed a downward development after year 3.I believe it reflected her new interest in other school subjects in year 4, when she said English was not anymore the favourite subject but one of the subjects she liked best; this perhaps made her devote less time to mastering vocabulary in the final year.
EB2 seems like a success story of a young learner.As a low-ability learner his oral performance seemed to be constantly on the rise in most aspects.This was reflected in the development of his self-concept, which turned increasingly more positive from year to year.The same trend was observed in his motivation, which also steadily grew over the years.This is consistent with the claim that self-concept is the best predictor of motivation (Mercer, 2011).Although he had strong family support for his learning of English, if it is taken into consideration that his socio-economic status was not high, it seems most likely that the key explanation of his success is connected to the high quality classroom teaching which he was exposed to.Age-appropriate methodology coupled with integration of out-of-school exposure to English into English classes was probably the trigger for this participant's L2 development.Although in a descriptive study like this we cannot be sure about the cause-effect relationships, I believe that it was this development that led to positive trends in both the young learner's self-concept and motivation.This reflects Moyer's (2004) claim that quality of L2 experience is a very important factor in language learning.L2 development of EB3, a high-ability and a relatively successful young learner, painted an interesting picture: While his task achievement was excellent during all the three years, he displayed the most variability of the three earlier beginners in the vocabulary, accuracy and fluency aspects.It seems that year 3 was a crisis year for him.This can be connected with a simultaneous drop in his motivation as well as his self-concept that same year.A look at what went on in the EB3's classroom that year suggests some possible explanations.Discipline problems in his large class, causing the teacher to frequently interrupt her teaching, may have made learning too demanding for EB3.Another possible explanation is that his specialist teacher, with no training in teaching young learners, made learning too demanding by insisting on accuracy.Teachers without an appropriate training in working with young learners may not be aware of the fact that younger children learn more slowly than older ones, and that linguistic accuracy is not the main aim in the early years of foreign language learning.His teacher's prevalent experience with older learners might have misled her to expect the same progress with younger children in all language learning aspects which she had witnessed in her older learners.Thanks to his constant interest in listening and learning new words, combined with interest in role plays and games which enabled him to apply his knowledge in the classroom setting, EB3 resumed the upward trend in his L2 development after year 3.The key role of the teacher and ageappropriacy of teaching methodology has been stressed in previous research (e.g., Nikolov, 1999;Vilke, 1995).Enever and Watts (2009) have also drawn attention to the importance of aligning classroom activities with learners' interests.
LB1 displayed just the opposite trend in her L2 development from the one observed in EB3.Her performance was, overall, best in year 3. What most likely contributed to that is the out-of-school exposure to English and home support.It was in year 3 that she could for the first time apply her knowledge of English while surfing on the Internet as well in real life communication with foreigners.This coincided with a rise in both her self-concept and motivation, which remained high till the end of the study.The classroom context did not seem to be too stimulating for her, especially after year 2. Her high interest, attention and engagement during English classes slightly decreased in intensity from year to year although they were never low.Relying on previous research (e.g., Harris & Conway, 2002), it seems to me that it was the opportunities to use L2 in real life situations that had the largest impact on this participant's L2 development.
LB2's English development suggests a strong interaction with some contextual variables and individual learner characteristics.The broader school context in which he learned English was excellent, but the classroom context was not too positive: The teacher's language competence was far from perfect, just as her classroom management skills.This seemed to be off-putting for LB2.However, his socio-educational context which supplied him with substantial amounts of exposure to English was very stimulating and, combined with his high language learning ability and genuine interest in languages as such, re-sulted in top oral performance.This is consistent with Murphy's (2014) conclusions on the impact of the out-of-school context based on her review of a number of studies on instructed L2 learning at the primary school level.
LB3's overall performance seems to have followed the same trend as that of LB1, only the results were lower.What stands out about this young learner is his lack of confidence, which seems to have had an important impact on his L2 development.The contextual background in which this young learner learned English was not unsupportive, but he obviously needed more support and encouragement both from his teacher and family to progress faster.The opportunity to use English while using the Internet probably helped him to maintain his above average level of performance in year 4. His lack of confidence, which was only reinforced by high anxiety during his first real life communication in English with a foreigner, probably had a crucial role in his L2 development, overriding the high ability and the generally positive sides of contextual factors.Mihaljevi Djigunovi ( 2006) has shown that L2 anxiety impacts oral language production and processing.Csizér and Dörnyei (2005) have also observed that experiences of language anxiety in early L2 learning can lead to young learners' lack of trust in their own abilities.
The present findings show that both groups of young beginners shared the generally positive attitudes of school staff and parents to early learning of foreign languages.Also, no matter what type and level of education their parents had, all young beginners enjoyed strong support for their learning of English.Murphy (2014) has highlighted the role of the parents by referring to their insistence on an early start as parentocracy.Another thing in common was a considerable exposure to English outside school, mostly through the media.However, the town and city settings, in contrast to the village setting, offered more L2 exposure in terms of names of shops, street advertisements and the like.
What the present analyses of earlier and later young beginners' profiles show is, first of all, that their L2 developed following different trajectories.Interand intra-learner differences could be observed at different points of the language learning process as well as in different aspects of language mastery.I would like to suggest that this was so because learner age interacted with other individual learner characteristics and with contextual factors in different ways.In both earlier and later young beginners in this study, L2 development seemed to enter into complex interactions with other variables, but the dynamics of these interactions was different for earlier and later young beginners.As Tragant (2006) has found in the case of young learners' motivation, biological age plays a more important role than, for example, hours of instruction.
According to our findings, in the case of the earlier young beginners, the role of the immediate learning environment seemed to override the other variables to a large extent.Similar conclusions were reached by Nikolov (2002) in her longitudinal study of Hungarian young learners of English as L2, which offered evidence that early beginners were more motivated by classroom practice than by motivational orientations connected with integrative or instrumental reasons.L2 development was associated more strongly with what was happening in the young learner classroom than with out-of-class factors.A good case in point is EB2, who was not as much exposed to L2 outside school as the other young learners living in nonrural contexts, whose socio-economic status was the lowest of the six participants and who, at the beginning of our study, had a negative self-concept as a language learner.The high quality L2 teaching he was exposed to must have significantly contributed to his increasing motivation and L2 self-concept, as well as L2 development throughout the three years of the study.His teacher's skillful and meaningful use of code-switching and integration of the English that her learners came in contact with outside the classroom was probably more beneficial for her learners' L2 development than much larger amounts of out-of-class exposure never made use of in class.The smaller size of the groups in which the earlier beginners were learning English was probably also a contributing factor to L2 development.This was confirmed in earlier research on Croatian young learners of English (Mihaljevi Djigunovi , 2009;Vilke, 1995).
The later beginners analysed in this study, on the other hand, presented a different picture.Their L2 development seemed to be more associated with individual characteristics and the out-of-school factors.LB2, for example, was put off by what was going on in his L2 classes but excelled in L2 oral performance anyway thanks to the high quality out-of-class L2 exposure (both in school through contact with foreign children and outside school) and the high socio-economic status of his family.As Nikolov (1999) observed, classroom teaching is associated with intrinsic motivation in young learners, but it becomes a less important source of motivation in a few years' time.Other motives emerge and, besides, negative attitudes to school in general tend to appear.Also, as LB1 illustrates, the number of out-of-class experiences in authentic use of L2 increases with age, and this slowly develops a more significant impact on young learners' motivation, self-concept and L2 development.Unfortunately, classroom teaching often does not take into consideration the language competences in English which young learners nowadays obtain out of school.Another distinguishing factor between earlier and later young beginners is the role of individual learner characteristics.The older the learner the stronger the impact of such characteristics as shyness and language anxiety.A case in point is paricipant LB3, whose oral performance reflected more his lack of self-confidence as an L2 learner than the relatively positive outlook of all the other variables I looked into.

Conclusions
The findings of this study suggest that age effects on L2 development do exist and are probably mediated by individual learner characteristics and contextual factors.I found that earlier and later young beginners of English as L2 may follow different trajectories in their L2 development.Generally, earlier beginners are more dependent in their L2 development on what goes on in the L2 classroom in the sense that their attitudes, motivation and self-concepts are based on the quality of the teacher and the teaching they are exposed to to a larger extent than is the case with later beginners.In contrast, later young beginners seem generally to be influenced less by classroom teaching than by out-of-class factors, especially L2 exposure and authentic experiences in using it, which emerge as mediating factors between age and their L2 development.
Our study provides evidence that in early L2 learning a couple of years may make a difference.Starting at age 6/7 as compared to 9/10 years results in a different impact of factors which drive L2 development.This may have important implications for the choice of predictors in future quantitative studies.

Figure 3 Figure 4
Figure 3 Overall oral performance of EB2 during three years