Learning to Write in a Second Language : The Role of Guided Interaction in Promoting Children ’ s Noticing from Model Texts

DOI: https://doi.org/10.5565/rev/clil.22 e-ISSN : 2604-5613 Print ISSN: 2605-5893 Coyle, Y.; Cánovas Guirao, J. (2019). Learning to Write in a Second Language: The Role of Guided Interaction in Promoting Children’s Noticing from Model Texts. CLIL Journal of Innovation and Research in Plurilingual and Pluricultural Education, 2(1), 21-30. DOI: https:// doi.org/10.5565/rev/clil.22 To cite this article:


To cite this article:
T his paper illustrates the use of model texts as a written corrective feedback technique with young foreign language learners.The procedure used by the teacher to focus the learners' attention on grammatical, lexical and textual differences between a model story and a draft version written by the children is described, and implications are suggested for the role of feedback processing in promoting L2 learning.

I
n foreign language learning contexts such as Spain, children spend at least twelve years of their lives in primary and secondary education.During this time, they are exposed to English as a subject in the school curriculum for a few hours a week and often with little or no contact with the language beyond the classroom.In many schools, under the guise of a communicative approach to language learning, much of the time is dedicated to explicit focus on forms (FonFs) (Long, 1996) and vocabulary teaching in teacher-dominated classrooms.The development of competence in foreign language (FL) writing is rarely contemplated beyond the completion of textbook exercises.Young learners in mainstream and content classrooms frequently copy, match, underline, circle and fill in gaps in sentences with topic vocabulary, but they are seldom required to write different types of texts in English.As a result, they are denied important opportunities to try out their developing knowledge of the FL and to receive feedback on their writing.Against this backdrop, it seems that many English as a foreign language (EFL) teachers of younger learners in the Spanish educational context are unaware of the instrumental role that writing practice and written corrective feedback (WCF) can play in the linguistic development of their pupils.The idea that engaging in both of these activities might actively contribute to children's foreign language development does not normally form part of current pedagogical agendas.The aim of the present paper is to present and illustrate an instructional approach using model texts as a WCF technique with a group of young EFL learners that led them to engage in collaborative reflection on written narratives with their class teacher, an experience that was found over time to improve the quality of the children's written output.
A rationale for the language learning potential of written corrective feedback T his neglect of writing in young learner classrooms clashes with contemporary theory and research in the field of foreign language writing.The cognitive'writing-to-learn-language' strand of second language acquisition research (Manchón, 2011) highlights the potential of writing to contribute to foreign and second language learning.From this theoretical perspective, it is argued that language learners should be given frequent opportunities to write in the second language (L2) and provided with relevant feedback on their language errors in order to trigger important learning processes including noticing, hypothesis testing and metalinguistic reflection (Ferris, 2010;Williams, 2012).By allowing learners to try out their developing knowledge of the L2 in writing, and by scaffolding this process with corrective feedback, teachers may be able to help learners become more aware of what they know in the L2 and what they do not.From this perspective, raising learners' awareness of 'gaps' in their L2 knowledge can be facilitative of interlanguage development (Schmidt, 2001;Swain, 1985).
Sociocultural theory has similarly advanced our appreciation of the language learning potential of WCF through the importance attached to cognitive development as a socially constructed activity.Inherent to this idea is the Vygotskian premise that the individual appropriation of linguistic knowledge can be co-constructed through collaborative talk during problem-solving tasks.Therefore, when pairs share their L2 knowledge when writing a joint text and then discuss together the corrections or differences they notice between their own work and the teacher's feedback, or when teachers help learners to focus on errors in their own writing or on the positive qualities of model texts, they are engaging in an important learning activity.Seen from both cognitive and sociocultural perspectives, the scaffolding afforded by the pooling of linguistic resources during feedback analysis both in teacher-led whole class discussions or in pair work, would appear to be a useful starting point to promote the creation of new knowledge through a process of what Swain (2006) has referred to as 'languaging' or '…making meaning and shaping knowledge and experience through language' (p98).

Model texts as a written corrective feedback technique
W ritten corrective feedback can be delivered in a variety of different ways including (a) explicit error correction, (b) use of codes or underlining to signal that an error has been made or (c) the provision of metalinguistic explanations or rules.In recent years alternative techniques including reformulations or model texts have become popular.Both of these more discursive types of feedback consist of providing learners with whole texts rather than lists of errors or codes.Reformulation involves rewriting the learners' text but correcting any grammatical, lexical, spelling or stylistic mistakes while maintaining the original ideas.Models are complete, wellwritten texts created by teachers taking into consideration the content and the genre of the target text, as well as learners' age, proficiency level, etc., but without specifically referring to the learners' written output.Models provide learners with rich sets of appropriate L2 words and structures for a given context, which can help them both identify their own errors and become aware of the alternative ideas and content in the model.They are also a less time-consuming feedback technique for teachers than individual corrections or reformulations, as one or two models can be used in a given class and tailored to meet the proficiency levels of the children.
Studies of model texts with young learners (Cánovas, Roca de Larios & Coyle, 2015) have found them to be useful in allowing the children not only to identify their own errors, but also to incorporate new lexis and chunks of language and improve the overall structure of their writing.However, since working with models requires further analysis and reflection than when errors are explicitly highlighted by the teacher, researchers have also suggested a useful role for instruction in helping learners take advantage of models.In some classrooms, especially with older learners, a focus on form and writing activities are often the norm and learners are well accustomed to receiving and analysing feedback.
Learning to Write in a Second Language: The Role of Guided Interaction in Promoting Children's Noticing from Model Texts However, in young learner classrooms, writing tasks and the provision of feedback are often overlooked in favour of oral communication tasks.This means that children are not only unfamiliar with writing regularly in the L2, but also that they have very little experience of handling feedback on their texts.Consequently, teaching aimed specifically at helping learners to identify and understand the nature of errors in their written output would seem to be a worthwhile venture.This was what we set out to explore.

Multi-stage Writing and Written Corrective Feedback Tasks
T he experience we describe below took place in a small state school in a village in southeast Spain as part of a larger study on feedback processing.Sixteen children forming a total of eight pairs from two EFL classes in primary education, a grade 4 (aged 9-10 years old) and a grade 5 (aged 10-11 years old) took part in the research.The two EFL classes were divided into a teaching group (grade 5) and the non-teaching group (grade 4), since despite the age difference, the children in both classes had overall similarly low levels of L2 competence.Over the school year, the children took part in two multi-stage writing and feedback tasks.At Stage 1 (Composing stage), the children were asked to jointly write a story in response to a picture prompt.At Stage 2 (Comparison stage), two days later, the pairs were provided with the stories they had written and a model text written by their teacher and discussed or made a note of any differences they could find.At Stage 3 (Rewriting stage), one week after having completed the initial writing task, the children were given the pictures again and asked to rewrite their stories.Between each task and over a period of six weeks, the children in the teaching group devoted one weekly English lesson to writing stories and discussing model versions with their class teacher while the nonteaching group continued with their regular EFL lessons.The procedure followed will be described and illustrated below.
"Models provide learners with rich sets of appropriate L2 words and structures for a given context, which can help them both identify their own errors and become aware of the alternative ideas and content in the model." Teaching children with model texts T he six teaching lessons in which the children worked with model texts were divided into two parts.In every session, firstly the children jointly wrote a story using a set of pictures (Appendix A) to guide their narratives (20 min) and then participated in a whole class discussion activity in which the teacher helped them to identify differences between a model text (Appendix B) and one of their anonymous drafts (40 min) (see Figure 1).To begin with, after the children's drafts were collected, the learners were encouraged by the teacher to tell the story in Spanish so that the meaning of the drawings could be clarified collectively.During this discussion, the teacher used mainly English while the children relied mostly on their L1.The teacher then projected a model version on the blackboard and wrote alongside it one pair's original text (each week a different pair was chosen) without revealing the identity of the writers.The children were told that the two versions of the stories had been written by Spanish and English children respectively (Picture 1).Explaining the task 1. T: Ok.Well that's the story and here we have a story written by some English children and here is another one written by Spanish children.
As you can see, there are some differences between them.
3. T: Well that's what we have to do, find all the differences.The story written by the English children is perfect as English is their first language and the Spanish story, since the children are still learning English, has some mistakes.We have to find them.Is that clear?
The teacher then read both texts aloud and the children followed silently before being asked to identify any differences they could find between the model text and the Spanish children's version.

Excerpt 3
Reading the draft (2) Sentence structure O ften, the learners noticed that the models contained longer sentences than the original stories, which enabled the teacher to explain that longer and more complex sentences could be produced by using connectors such as 'when' or 'but'.In Excerpt 5, the teacher explicitly draws the children's attention to the combination in the model of two ideas into a single sentence and encourages them to attempt this in their own writing.
2. T: Very good.They haven't written it.What does 'but' mean?
3. PP: Pero 4. T: Very good.The children haven't used the connector to join the sentences.They've written shorter sentences right?Can you see how the English children have joined two ideas together: "She goes to school but at a quarter past nine she falls asleep?That's a nice way to write better stories.Try to use connectors to write long sentences.hen the children found a content-related difference, the teacher pointed out that the model could be used as a source of ideas to improve their own writing.For example, a common content difference noticed by the children was the spatial location of the story characters in the models.This led the children to make strategic comments on their future writing such as 'Next time, we should include the place'.In the following example, a child noticed that this particular model provided more detailed reference to the temporal setting of the story than the learners' draft, which began by describing the sequence of events.

Excerpt 4
Noticing differences in story content 1. P1: En la parte de los ingleses especifican un poco más las cosas que las de los españoles, por ejemplo: al empezar la historia dicen "It is Monday the first of May" y los otros escriben "Emily leaves school".T he teacher's interventions in drawing the children's attention to linguistic and textual differences in the imperfectly written texts of their peers in comparison to model texts proved valuable in helping them to make better use of the feedback.In the second multi-stage task, the pairs in the teaching group made fewer errors and wrote qualitatively better texts that the learners who had not been helped to analyse models (Table 1).

T: Supergood
The question is, then, what was it that enabled the children in the teaching group to produce better texts?The answer, we believe, lies in the social dynamics of the classroom discourse, which became a forum for teacher-led collective 'languaging'.As described above, the children were shown how to go about analysing diverse features of the model text while simultaneously supported in the development of metalinguistic knowledge through a process of dialogic interaction with the teacher.The teacher's role in scaffolding the children's noticing and in offering metalinguistic explanations for errors in the sample texts over a sustained six-week period seemed to have raised the children's awareness of form-function mappings and strengthened their grammatical, lexical and discursive knowledge, thus priming them to become more perceptive when handling feedback on their writing.
The findings of our study suggest a role in the classroom for consciousness-raising activities using model texts to help younger, less proficient learners improve their foreign language writing and develop their meta awareness of language as a system.Without the experience and knowledge accumulated in the teaching sessions, children in the nonteaching group struggled to notice and understand linguistic features in the input.However, by actively directing learners' noticing and filling in gaps in their L2 knowledge, the teacher helped the children in the teaching group to improve the quality of their noticing from the model and improve their written output.It seems important, therefore, Pluricultural Education, 2(1), 2019: 21-30 The Crazy Scientific Once upon a time a one scientific does a potion.They are a dog sleeping.The scientific drink a potion, the scientific it's crazy.The head scientific bumm, bumm!The dog gets up.The scientific turns into a cat!The dog looks angry a cat.The dog jump the cat.

The cientiffic and dog
The cientiffic is create poccy and dog is sleeping.The cientiffic is drink poccy and dog sleeping.He dolor the gargant.An cientiffic explossion is hear and dog is week up.He cientiffic is convertic a cat and dog the look is roon.An dog is attacks at cientiffic.The end.
Learning to Write in a Second Language: The Role of Guided Interaction in Promoting Children's Noticing from Model Texts for teachers and future teachers to become more aware of the benefits to be gained from regular writing practice and guided feedback processing, as well as of the theoretical implications involved in the impact of writing on children's second language development (Manchón, 2011).
Further practical considerations concerning the nature of feedback, the pairing of learners and the integration of reading and writing activities deserve special attention.Models are not a personalized form of feedback directed at individual errors in children's writing.Therefore, for younger learners at different proficiency levels, the uniformity of model texts as a 'one size fits all' technique may not always be the most appropriate solution.The idea of preparing two or three models to cater for the different proficiency levels within EFL classes might be a worthwhile option to explore.In practice, this would mean elaborating various models at different levels of difficulty to match the diversity that characterizes mixed-ability classrooms in primary education.Teachers might also alternate models with other forms of WCF, including error correction, reformulation or metalinguistic explanations, which might make specific errors more perceptually salient.Becoming aware of the advantages afforded by different types of feedback and selecting suitable techniques for different language learning purposes should be a central part of the teaching agenda of EFL teachers and trainees.
It is also true that some thought should be given to assigning learners in pairs.Placing two children together does not necessarily guarantee that they will collaborate effectively.The children in this study worked well together, although lower proficiency learners had greater difficulties in providing support for each other given their own limited knowledge of the L2.In this respect, the idea of 'paired writing' (Yarrow & Topping, 2001), in which more experienced and knowledgeable learners actively assist less proficient peers in producing a collaborative text using a Coyle, Y.; Cánovas Guirao, J.
Learning to Write in a Second Language: The Role of Guided Interaction in Promoting Children's Noticing from Model Texts

Figure 1 .
Figure 1.Structure of the teaching sessions

Picture 2 .
Whole class discussion of the draft and model text (1) Content W

"When pairs share their L2 knowledge when writing a joint text and then discuss together the corrections or differences they notice between their own work and the teacher's feedback, or when teachers help learners to focus on errors in their own writing or on the positive qualities of model texts, they are engaging in an important learning activity. " "Models are complete, well-written texts created by teachers taking into consideration the content and the genre of the target text, as well as learners' age, proficiency level, etc., but without specifically referring to the learners' written output."
Coyle, Y.; Cánovas Guirao, J.CLIL Journal of Innovation and Research in Plurilingual andPluricultural Education, 2(1), 2019: 21-30

"The question is, then, what was it that enabled the children in the teaching group to produce better texts? The answer, we believe, lies in the social dynamics of the classroom discourse, which became a forum for teacher-led collective 'languaging'." Pedagogical Implications
Learning to Write in a Second Language: The Role of Guided Interaction in Promoting Children's Noticing from Model Texts 4. T: Very good.Exactly.Theyonly saywhat Emily is doing.Good difference.Coyle, Y.; Cánovas Guirao, J.

Table 1 .
Sample stories written after the instructional intervention