Language learning effects through the integration of synchronous online communication: The case of video communication and Second Life

3.3.1 Sessions 2 and 5: People and adventure and People and cultures

The qualitative analysis of interactions across conditions comes particularly from the second and fifth tasks, People and adventure and People and cultures, of which we provide further details below.

Task 2, People and adventure, was divided in two parts: the preparation of a holiday together and the holiday itself. Participants had to discuss their preferences and reach an agreement on the destination. During the process they were encouraged to talk about their last holiday, any negative experiences with accommodation they might have had previously, make a list of things they might need for their chosen destination, and talk about the activities they could undertake upon arrival.

For the SL group, the interactions took place in an area within Second Life designated for the NIFLAR

project, ^[1] but it also made use of readily available locations as well as venues specially designed for particular tasks.

At the beginning of the task, the SL group was teleported to a pizzeria holodeck ^[2] where they discussed their preferences and reached an agreement on their destination. Here they had the opportunity to talk about their previous holidays and make the preparations. They were then teleported to their hotel and in a role-play manner they had to sort out any problems they might encounter (leaking shower, unmade bed, dirty floors, etc.). Finally, they headed to their destination where they could, for example, go sailing, surfing, horse riding, skiing or ice skating. They were provided with a taskHUD ^[3] with additional objects (ski outfits, receptionist uniform, bikini, etc.) that could be used during task performance.

The VC group was able to contextualize the task via images uploaded onto the video communication environment, Adobe Acrobat Connect Pro. The participants themselves uploaded their own in order to talk about past holidays and the researchers placed images that depicted possible bad accommodation experiences (dirty floors, broken windows, noisy hotel guests, etc.) and various types of sports that could be practiced during both types of holidays.

As for the face-to-face control group, each member was asked to bring photographs from a past holiday to the lesson.The rest of the visual material (accommodation problems and activity possibilities) was provided by the teacher during task peformance. The same non-personal pictures were used in the VC and face-to-face control groups, and an attempt was made to make them as similar as possible to the scenarios in Second Life.

Task 5, People and Cultures, was organised as a cultural quiz in the style of a television game-show. Dutch and Spanish teams would compete in terms of cultural knowledge of the other country. Questions for which the teams could score points were displayed on a screen and the opponents had to judge the adequacy of the answers. The face-to-face control group, where students were paired to compete against each other, used only the images of the SL and VC group related to the target culture, and the instructor, in an expert role, acted as adjudicator.

3.4.1 Oral tests

The pre- and post-test were the same for the three research conditions and administered via a computer in a language lab. The questions had been video recorded by a native speaker and students had to record (voice only) their answers after each question. The test consisted of 11 open-ended questions comprising personal information and course of study (3 items), previous experiences in Spanish-speaking countries (1 item), student life (3 items), weather and gastronomy (2 items), economy (1 item) and recommendations for foreigners visiting their country (1 item). Although the topics that came up in the oral tests were dealt with in one way or another during the tasks, the oral tests did not have the same structure as the task sessions.

Language learners were assessed on measures of communicative language competence based on descriptors that were inspired by the scales proposed by the Common European Framework of Reference for languages (CEFR). Five measures were assessed on a ten-point scale: (a) range of language, (b) grammatical accuracy, (c) fluency, (d) thematic development and (e) coherence (see Appendix). Two native speakers of Spanish separately rated the tests and since there was a high correlation between all five indicators (.89≤ r ≤.98), they were collapsed into one measure of oral skill. An estimate of inter-rater agreement between both raters was calculated and proved to be high (α=0.91).

3.4.2 Surveys

After the conclusion of the project a final survey was digitally distributed to all three groups of learners and native student teachers in order to evaluate participants’ experiences. The survey for VC and SL groups had 29 items with open and closed questions concerning the virtual environment, tasks, speech partner, learning potential of the sessions, organization, and global project evaluation. For the face-to-face control group, questions related to the usability of the networking tools and the interaction with native speakers were eliminated from the survey since these elements were not present during task performance. In all three surveys for the closed items a five-point Likert scale was used.

3.4.3 Interaction analysis

Three factors were taken into account regarding the selection of recordings for this study: (a) accessibility of the recordings, (b) representativeness, and (c) previous use in earlier analysis. The recordings database was sometimes difficult to navigate, since participants themselves were in charge of the recording process and file names were not always clear. Easily identifiable files and quality of the recording were primary selection criteria. Once the file had been selected we looked for representativeness of B1 language level. The recordings in Second Life were made by one participating student per group with the help of the screencasting and video editing software Screen Flow under the supervision of a university technician. For video communication, Adobe Acrobat Connect Pro had a recording feature and one student per group was in charge of recording each session. In the face-to-face control group, a camera on a tripod was set in the classroom to film the sessions.

In order to analyze the interactions from all groups we used the following categories: (1) negotiation of formal meaning (Pica 1994; Long 1985; Gass 1997); (2) negotiation of cultural meaning; (3) interactional features; and (4) task engagement and involvement.

(1) The negotiations examined here tend to follow Varonis and Gass’s (1985) model of negotiation of meaning. This model presents a set of primes that occur in negotiation of meaning sequences: a trigger utterance (T) that causes a comprehension problem and starts the negotiation sequence, an indicator (I) that signals the misunderstanding, a response (R) to address the misunderstanding and an optional reaction to that response (RR). The nature of the triggers can vary, from lexical and morphosyntactic items to sociocultural issues; and the indicators of misunderstandings can manifest themselves in the form of echos (rising or falling intonation), explicit statements of non-understanding, silence, inappropriate responses, summaries and surprised reactions. The response from the interlocutor can take the form of repetitions, expansions, rephrasings, acknowledgements or reductions (Varonis and Gass 1985; Fernández-García and Martínez-Arbelaiz 2002). Once the negotiations were identified, they were sorted according to Toyoda and Harrison’s (2002) trigger categories: word level (recognition of a new word, misuse/misunderstanding of a word, pronunciation); sentence level (grammatical errors); discourse level (sudden topic change, slow response). The negotiation strategies to bridge the communication breakdown were categorized according to those identified by Lee (2001): comprehension checks, clarification requests, confirmation checks, use of English to translate unknown meanings in Spanish, word invention, requests and use of approximation.

(2) The tasks designed for the study were seeded with cultural information gaps from the L1 and L2 cultures that were likely to result in negotiation of intercultural content. These tasks, designed with an intercultural focus, offered opportunities for interaction that would elicit information on everyday cultural customs and provide opportunities for contrasting and comparing beliefs and habits in order to elicit awareness and reflections on both cultures. In order to see if there was a difference in the negotiation of cultural meaning between the three groups we measured the negotiations not triggered at word level but by a cultural information gap that occurred in task 5, People and cultures.

(3) In order to establish other differences between tasks performed via networked interaction and those carried out in the classroom setting, we observed the interactional features that allowed learners to share their feelings and to demonstrate a sense of sociability (Chun 1994; Darhower 2002). These features included greetings and leave takings, assistance requests, use of humour and use of L1 or English.

(4) We considered task engagement to comprise learners’ efforts, participation, involvement and positive conduct during the execution of the learning activities. We broadly observed Littlewood’s (2004) definition of engagement as the learners’ active personal involvement with the task and viewed engagement as student motivated behavior, but task engagement was not operationalized as number of turns or number of words learners produced while engaged in the tasks. Instead, observation notes were used to spot learners’ attitudes towards the language learning situations and tasks.

4.2.1 Negotiation of meaning

As Table 2 shows, in the analysis of the three recordings observed (task 2 – VC group; SL group, face-to-face control group), we found instances of negotiation in all groups, although there was a substantial difference between the number of negotiations encountered when the task was performed by the experimental groups (VC and SL) and those found in the same task by the face-to-face control group: SL and VC groups accounted for 27 and 23 negotiations respectively and the face-to-face control group for 2.

Analysis of the recordings reveals that lexical difficulties appear to be the principal triggers as described in previous literature (Blake 2000; Smith 2003) as most of the negotiations between the participants are triggered by lexical confusions. Table 3 summarizes negotiation instances and since all triggers found in task 2 were at word level, they were first categorized following Toyoda and Harrison’s (2002) classification of this type of trigger: recognition of a new word, misuse/misunderstanding of a word and pronunciation. The indicators of these negotiations are then sorted according to the strategies used by the participants (Lee 2001): comprehension checks, clarification requests, confirmation checks, use of English, word invention, requests and use of approximation. Table 3 shows only the categories most used during the negotiation routines.

4.2.2 Negotiation of cultural meaning

Further observation of the recordings confirmed that negotiations occurred in all tasks but that they were not always triggered at word level and instead the trigger depended on the type of task they were assigned. Task 5 (People and cultures) was a clear example of this. The task had been seeded with multiple information gaps involving culture-specific situations from the L1 and L2 cultures that would call for intercultural negotiation and it is no surprise then that the majority of the triggers for negotiation routines were at a sociocultural level where the intercultural communication gap became apparent (see Table 4).

The task generated in the experimental groups a high level of curiosity towards the other culture and participants engaged in rich exchanges as example 1 (VC group) shows. This example had as trigger a photograph of a rucksack hanging from a Dutch flag, and the Spanish team had to guess its meaning. Having provided the wrong answer the Dutch team proceeds with the explanation: in The Netherlands it indicates that someone has graduated from secondary education. The native speaker signals her surprised reaction not only with an indicator of meaning recognition (¡Ah!) but also followed by a clarification request (¡Ah! ¿y entonces se quedan ahí las mochilas?/Ah! and do the rucksacks stay there?). The response of the language learner is an acknowledgement with an expansion (sí/fuera/por dos semanas o así – yes/outside/for two weeks or so) by the student, followed by another indicator of confirmation of understanding by the native speaker which triggers another expansion response (porque es la idea que nunca tenemos que usar la mochila – because it is the idea that you don’t have to use the rucksack/backpack anymore). The sequence is rounded off by the native speaker with a reaction to the response (¡Ah! ¡qué originales! – Ah! how original!).

Picture 1:

Photograph of a schoolbag hanging from a Dutch flag outside someone’s house.

In example 2 (SL group) the Dutch team had to explain what “un botellón” is, depicted in a photograph. Once the question had been answered both teams proceed to contrast information on this social activity popular among some young people. The example below also shows how the language learner shares understanding (contributing a remark about bottles of wine and beer to the native speaker’s explanation of the fact that they bring alcohol) and participates actively in meaning creation.

In the face-to-face control group the dynamics were different: there was no comparing of opinions between the members of the team to reach an agreement over the correct answer, and even when their answers were wrong there was little evidence of curiosity to find out more about the topic. In this case, the teacher was the expert and she was the one that provided the right answers, giving the language learners the opportunity to ask questions if they wanted. The impoverished exchange process can be seen in example 3 (C, SL and VC groups), where the language learners had to guess the process to apply for a job in the public sector (teacher, doctor, secretary, etc.) in Spain.

4.2.3 Interactional features

The interactional speech acts of interest that emerged from the data were greetings and leave-takings, assistance requests, the use of polite formulas, and humour. Observation confirmed that both experimental groups spent time greeting each other and talking about their current activities outside the task before beginning with task performance and that leave-takings were generally used to arrange the next task and review the task just performed. Example 4 (VC group) below was taken from the VC group during the first few minutes of the interaction.

The recording of the SL group starts with “¿estáis listas?/are you ready?”. When the participants arrive at the first destination, the pizzeria holodeck, and upon seeing the food on the tables, one of the language learners asks her partner “Fleur, ¿aún tienes hambre?/Fleur, are you still hungry?” which seems to indicate that in the minutes prior to activating the recording and the beginning sentence, an exchange had taken place around the topic of food.

The face-to-face control group on the other hand started the task very abruptly and lacking spontaneity, with one of the language learners stating “Me gustaría ir de vacaciones /I would like to go on holidays”. Closure was also abrupt.

From the interactional speech act types observed there was one in particular that appeared to register more instances in the SL group than in the other two: the use of humour. In the episode where the participants are getting ready to go to the beach it becomes apparent that the native speaker cannot make her avatar put a bikini on: all she manages to put on is the box containing the items. The situation prompts funny comments like: “así no puedes nadar /you can’t swim like that; con la caja tú no estás a la moda /with the box you are not fashionable; si bajamos a la playa con estas cajas vamos a causar sensación en la gente /if we go down to the beach with these boxes we are going to cause a sensation among the people”. Example 5 (SL group) draws on the reactions caused in the language learners when they encountered a hotel room they were not expecting (dirty and untidy).

Within the same situation, complaining about the hotel room, sarcasm also occurred (example 6 – SL group). One of the language learners, playing the role of hotel guest, and seeing the state of the room, questions herself whether they would have to pay for it, to which the receptionist (native speaker) responds with surprise.

Observation reveals that also in the VC group a friendly atmosphere prevailed with constant laughter, allowing learners to share their feelings and demonstrate a sense of sociability (Chun 1994; Darhower 2002).

The supportive nature of a substantial part of the interaction was also observed when in the VC and SL experimental groups participants helped each other with technical issues when assistance was requested. Example 7 (SL group) is only one of the many instances where a language learner is explaining how the environment works when another interlocutor runs into problems. The avatars are getting their attire ready for their holiday, although some of them seem to be having difficulties performing the correct action with the desired effect.

These instances where assistance was requested were not part of the task itself, they were environment triggered exchanges. This seems to contrast with similar situations observed in the face-to-face control group, where on occasion requests for language help went unanswered. Example 8 (C group) shows how a language learner was trying to find the word for ‘hat’ in order to describe her picture. She made gestures referring to the object and specifically asked for the word but received no response. Surprisingly enough, minutes later, one of the other language learners used that same word in her speech.

The video recording of this scene seemed to show that her interlocutors were concerned with their own interventions and once the learner requesting assistance received no apparent help, she decided to leave her sentence unfinished and carry on.

4.2.4 Task engagement

The high levels of task engagement reported in other studies (Deutschmann et al. 2009; Gardner 2010; Gámen-Gutiérrez 2014) were also noted in both experimental groups. When discussing their previous holidays participants would seek engagement and personal involvement with their interlocutors for example by asking them if they had been to that place themselves: “¿conocéis Bodrum? /have you been to Bodrum?” (VC). The language learners from the face-to-face control group, however, proceeded to talk about their holidays one by one, with no interruptions from their interlocutors and with only a couple of questions at the end of each other’s turns, questions that seemed to be prepared beforehand and did not appear to be spontaneous. When the same topic was being handled by VC and SL, participants engaged in frequent turn overlaps, which shows involvement with the task topic and which is characteristic for pragmatic interaction style in Spanish (Escandell Vidal 2004). In example 9 (VC group) two of the VC participants are showing some photographs of their holidays and the others engage in the interaction immediately showing their interest by asking questions.

This involvement, with the task and its participants, not only manifested itself in the form of questions, but also participants often volunteered comments on each other’s interventions, showing a positive attitude towards their interlocutors. In example 10 from the SL group one of the language learners is talking about the things she always takes on holidays with her. The list of items (compass, walking boots, bikini, towel, thick jacket) is probably not what the other language learner was expecting and she does not let the opportunity go to show her surprise.

The environments appeared to elicit a high degree of rich participation that led to personal involvement. We observed that in Second Life elements of the virtual world generated a variety of conversations. For example, in the pizzeria they chatted about the movies displayed in wall posters; being able to drive a boat once they had reached their holiday destination prompted conversations about seatbelts and drivers’ licenses (“ponte el cinturón que vamos muy rápido - ¿tú tienes el carné de conducir? /wear your seatbelt we’re going very fast – do you have a driver’s licence?”); and visiting the hotel where they were meant to be staying made possible the transaction of booking the room. The interactions from the VC group and face-to-face control group were characterized by a more descriptive language limited by the photographs being used. In Second Life actions triggered conversations and there was more topic switching and engagement opportunities enabled by in-world elements.