Table of contents
2. Theoretical context
2.1 Collaborative dialogue
2.2 Language related episodes
3 Previous studies
4 The study
4.5 Results and data analysis
5 Application and discussion
(Appendix A: Text-reconstruction task
(Appendix B: Achieved points in percentage of different groups
(Appendix C: Table 1: Level of engagement of LREs produced
(Appendix D: Anti-plagiarism statement
The problem addressed in this study is whether collaborative dialogue helps second language (henceforth: L2) learners of English raise their language awareness.
In this paper, I first give a brief overview about the theoretical context on collaborative dialogue and language related episodes (henceforth: LREs).
According to a number of different studies, which will be presented in this paper in more detail, dialogue can help us to become aware of different linguistic and pragmatic problems. Dialogue forces us to cognitively as well as socially reflect on our language used and produced (Swain, 2000).
I will then discuss the data deriving from a study focusing on the effects of collaborative dialogue. The study contains a text- reconstruction task which will be carried out and presented under consideration of the research question. Collaborative dialogue is a “joint construction of language or knowledge about language - by two or more individuals” (Swain, 1997, p. 115). It is generally an important part of the learning process. Learners can construct their linguistic knowledge and self- or other-repair.
The results of the study show that collaborative dialogue has a clear impact on language awareness of the students learning English: different LREs show the engagement and development of the student's language reflection or production.
Furthermore, this paper will point to the importance of collaborative dialogue in the L2 learning and will give a brief overview of what teachers are recommended to focus on while choosing different tasks.
2. Theoretical background
One of the first attempts to explain the acquisition of second language is Krashen's input hypothesis. Although the reception of input is little beyond the current individual level of competence, language learning is stimulated (Johnson, 2008). According to Krashen “negotiation leads to a greater comprehensibility of input” (Swain, 2000, p. 98). Reviewers stated that input is only made comprehensible though modified interaction. Therefore, Swain (1995) developed an output hypothesis, which claims that output pushes the learner to develop and process the language in a more accurate way by raising his or her awareness on language characteristics and realising how language as a system works. While producing language, learners are in control and able to notice their own limits, recognise language gaps and identify linguistic problems. In general, it can be said that learners improve their interlanguage. Swain states three functions of output: the noticing function, the hypothesis-testing function and the metalinguistic function. The noticing function describes the act of a learner noticing a gap in their interlanguage and filling this gap by social interaction or simply by looking it up in a dictionary. The function of hypothesis testing is simply the testing of hypothesis while producing output independent from the task type, which is either spoken or written. Errors are corrected by rephrasing the own text to achieve accuracy (Garcia Mayo, 2002). Due to this progress, students are cognitively able to compare their own produced output with own mental feedback (Williams, 2012). The last function of output according to Swain is, as already mentioned, the metalinguistic function. Peers, teachers or simply grammar books can help filling these gaps which learners notice about the target language produced (Garcia Mayo, 2002). For all three functions mentioned above, output serves to control and internalise linguistic knowledge.
Hence, Verbalization, which is inaugurated through social interaction with dialogue as its foundation, also has important consequences for the process of L2 learning. Learners pay attention to their problems and associate linguistic needs (Swain, 2000). They can easily “monitor their own language use, and evaluate their overall success” (Swain, 2000, p. 109). Verbalization is spontaneously evoked. By means of collaborative dialogue the students are able to “engage in knowledge building” (Swain, 2000, p.109). Nevertheless, Swain (2006) states that output or verbalization do not describe the whole process of language learning. She rather uses the term languaging to describe not only the auditive activity but also the cognitive process when producing a language. The term languaging was used by Lado and other researchers, but however not in the way Swain defines it. According to Swain, it is an activity of the mind and a “never ending process of language to make meaning” (Swain, 2006, p. 96). She states that during collaborative dialogue speakers engage in problem solving and knowledge building. It generally supplies the learners with opportunities to reflect their own language use. They sometimes work collaboratively to express their intended meaning and consequently carry out a task. In the following, this will be elaborated upon.
2.1 Collaborative dialogue
Many authors and researchers use the word dialogue in various different ways and senses, but it generally delineates a specific set of assumptions, attitudes and motivations. In the study I focused on collaborative dialogue between students. Therefore, I will put a special emphasis on the concept of collaborative dialogue according to Swain. In collaborative dialogue, two or more students work together on a given task, discuss about a theme or given problem. It is claimed to be “knowledge building dialogue” (Swain, 2000, p. 97) because students are able to reflect on their own or others language produced and improve their competences. This requires the learners to adapt socially and cognitively to their environment at the given time to successfully problem-solve. The concept of language learning is benefited through the social interaction of the dyads (Swain, 2000). In collaborative dialogue, a number of turn-taking points can be observed which are classified as LREs in the following.
2.2 Language related episodes
LREs occur during dialogues in which learners are unable to solve a linguistic or semantic problem by themselves. This may be any part in which learners question the language they are producing or talk about (Swain, Lapkin, 1995). Learners solve their problem through a short knowledge building exchange with their interlocutors, which is called LRE. They do not only question a linguistic but also a grammatical form of a word, the correctness of a specific spelling and pronunciation or explicitly correct their own or another's usage of a word, form or structure (Ismail, Samad, 2010). LREs include conversational turns, which represent the L2 learning in progress. More turns usually lead to a discussion and show the learner's engagement concerning the problem (Garcia Mayo, Azkarai, 2016). The learner's engagement as well as the episodes in general can be classified into different categories. According to Garcia Mayo and Azkarai (2016), the level of engagement is distinguished into three subcategories: Elaborate (henceforth: E), Limited (henceforth: L) and Limited + Limited (henceforth: L + L). E engagement occurs when both learners came up with engagement and ideas to solve the problem. L engagement appears when only one participant shows condolence towards an LRE and L + L engagement means that both participants ignore the gap collaboratively which they have while producing the language. Furthermore, the LREs are classified according to different categories regarding their nature: whether it is meaning- or form-focused and whether the outcome is resolved (henceforth: R) or not resolved (henceforth: NR). In their study Garcia Mayo and Azkarai (2016) defined these particular subcategories in more detail, which I will leave out in my study, however. The following graph illustrates this information:
Abbildung in dieser Leseprobe nicht enthalten
Figure 1. LRE categorization according to Garcia Mayo and Azkarai (2016)
3. Previous studies
Various studies by all different researchers exist concerning the topic of collaborative dialogue focusing on specific tasks effects. Three studies will be presented in the following:
Garcia Mayo and Azkarai (2016) have dealt with the question whether different tasks have an impact on the LREs produced. 44 Spanish Basque learners of English were asked to do different tasks which consisted of four different collaborative tasks. Garcia Mayo and Azkarai (2016) have taken into consideration the level of engagement from the material they used. Either task included two writing and two speaking assignments. Their findings show that the writing tasks lead the learners to produce a higher number of LREs. Likewise, writing tasks also lead learners to focus more on form than the oral tasks, in which learners mainly focused on the meaning. Therefore, a clear task modality effect is shows that the more structured a task is, the more attention to language it seeks (Garcia Mayo, Azkarai 2016).
Whether young adolescent learners become aware of their gaps in the linguistic knowledge whilst producing the L2 has been examined by Swain and Lapkin (1995): They tested 21 French immersion students during a writing task. The students were isolated in a room with a researcher and asked to think aloud while working on the task. Consequently, the learners were forced automatically into thinking about the form of their linguistic output. This showed that the learners also became aware of their language gaps even when external feedback was unavailable to them.
In 2002, Garcia Mayo has investigated the effectiveness of two form focused tasks by testing 14 English foreign learners over a period of one and a half months. The tasks that were chosen included the dictogloss and text-reconstruction task. In doing so, the leaners' attention towards language has been focused upon. The results demonstrate that the text-reconstruction task showed eight times more LREs than the dictogloss task. It has also been shown that more structured tasks encourage the learners to pay more attention to grammatical problem solving. When doing the dictogloss task, participants needed to understand the text and reconstruct it afterwards, whereas whilst doing the text-reconstruction task, learners only talked about formrelated issues (Garcia Mayo, 2002).
4. The study
The task used for the study was a text-reconstruction task. This task has been chosen because it is believed to be an effective form-focused task. The text-reconstruction task is considered to be a task which is more structured. It therefore would seem to elicit more attention concerning the language than their counterparts, like the dictogloss task, for example (Garcia Mayo, 2002). This makes it easier for learners to notice their own grammatical strengths and weaknesses. In the text- reconstruction task, learners usually have to insert function words like articles, prepositions, linking words and inflectional morphemes to produce an appropriate text. This task has been slightly changed as it was only asking for and focusing on the application of function words. The text used here was an article published in the Guardian called “France records all-time hottest temperature of 44.3C” (cf. Appendix A).
According to the results of previous studies, I certainly expected the dyads to perform better and produce a better outcome than the students who worked on the task alone. When looking at the transcribed dialogues, I expected the pairs to produce a high number of LREs. Concerning the paired students, I will also look at the level of engagement of the LREs produced. Due to time issues and resources, I only looked at the LREs produced by the pairs.
The participants were university students between the age of 19 and 26 and were chosen and paired up randomly. In total, 15 (ten female and five male) university students took part in this study. The participants were all students enrolled in different university degree courses. All of the participants were exposed to English as a foreign language until the end of their A levels. After that, they have not been exposed to English on a daily basis. Therefore, the participants have some differences regarding their knowledge of English as they have finished school in different years.
After participating in the study, the students were asked to give information about their age and gender. It is also important to know whether the students spent some time in an Englishspeaking country because these students were likely to show higher language feeling. All of these factors will have influenced their knowledge of English and consequently, the students show different levels of language competence.
The task used in this experiment is a text-reconstruction task. The learners were presented with the written stimulus, including the task instruction on the top of the page. The text was an article taken from the Guardian and included a 681 word-passage. I erased 37 function words which the participants had to fill in correctly. The focus here was mainly on prepositions, determiners and conjunctions. The text-reconstruction task has been claimed to be an effective form-focused task. Learners can pay their attention to form and not only to meaning (Garcia Mayo, 2002). While reconstruction the paragraphs, learners can easily skip some gaps, go along and come back to the point they left out when re-reading the text.
All participants were chosen randomly and no laboratory setting was provided. The students worked on the study in different settings but everyone in an own room, so they did not get distracted by others. The setting can be described as one where the participants feel comfortable and at home. The task was handed out and the participants read the instructions. The ones working in dyads were given only one copy of the task which reinforced collaborative work. Consequently, they were obligated to exchange their thoughts with their partners. The students had the chance to ask questions before beginning the experiment. After having started, I did not answer any questions anymore.
Five students, including three female and two males, worked on the task alone. The procedure was simple. The participants filled out the information required in the task and handed it in afterwards. Neither the thoughts, nor the level of engagement was analysed when looking at this group. Thereafter, I corrected the sheets and calculated the percentage of achieved points (cf. Appendix B).
The conversations, which occurred during the editing phase of the exercise concerning the dyads, were audiotaped and transcribed afterwards. The focus here was on the LREs, as explained before. An LRE started when a participant raised concern about a decision of the language produced. I also looked at the level of engagement concerning the students. In doing so, I followed the categorisation according to Garcia Mayo and Azkarai (2016), as explained and pointed out in section 2.2 and in Figure 1. When transcribing the dialogues, I also marked the level of engagement of the participants. Both the number of LREs and the level of engagement are important indicators for L2 reflection and L2 learning in progress (Garcia Mayo, Azkarai, 2016).
4.5 Results and data analysis
In the following abstract present findings regarding concerning whether collaborative dialogue raises L2 learners of English's language awareness are presented and the most significant findings are summarised. None of the students taking part in the study had spent some time in an English-speaking country.
Abbildung in dieser Leseprobe nicht enthalten
Table 1. Quantitative comparison between the two groups
In total, the individuals nearly made twice as many mistakes as the dyads did. Consequently, the pairs scored much better when looking at the percentage of the points achieved. This clearly shows the impact of collaborative work. In the following, I will analyse the different time frames used by both groups and the LREs made concerning the dyads.
The time used in total by all participants consists of a total time of one hour, six minutes and six seconds. When looking at the different times of the two groups, a clear difference is visible.