Language Learning & Technology
Vol. 7, No. 3, September 2003, pp. 141-159
USING NATIVE SPEAKERS IN CHAT
Paginated PDF version
University of South Australia
SLA research indicates that negotiation promotes interlanguage development and that learners are most likely to negotiate if opportunities for oral interaction are provided. In the case of campus-based students, learners' progress is supported and monitored mainly through classroom interactions. If students do not attend classes on campus, how do they gain the reported benefits of oral interaction? Recent studies indicate that chatting provides opportunities for the negotiation of meaning, as occurs in oral interaction. However, most of these have been conducted on interactions between learners, with teacher supervision, often in task-based instructional settings. This study considers implications for distance language learning of negotiations by a group of intermediate learners of Italian interacting in dyads on a Web based Italian native speaker (NS) chat program. The research specifically explores (a) whether live chat with native speakers offers opportunities for negotiation of meaning in open ended tasks carried out in single session interactions with unfamiliar NS without teacher supervision, (b) the principal triggers for negotiation and modification of interlanguage in these interactions, and (c) whether public NS chat rooms are likely to offer an optimal environment for SLA, even for learners studying at a distance who need to chat without supervision. Chat logs indicate that learners do in fact negotiate for meaning and modify their interlanguage when engaged in open ended conversational tasks with unfamiliar interlocutors, with lexical and structural difficulties triggering most negotiations. Though further research needs to probe whether these negotiations and modifications lead to acquisition in the longer term, they would be particularly valuable for distance learners who need opportunities to negotiate within authentic target language contexts.
It is essential that distance language courses provide learners with opportunities for oral interaction since it is within this context that negotiation of meaning and interlanguage development are most likely to occur. Second Language Acquisition (SLA) research findings are in fact based mainly on face-to-face conversational interactions:Negotiation for meaning is the process in which, in an effort to communicate, learners and competent speakers provide and interpret signals of their own and their interlocutor's perceived comprehension, thus provoking adjustments to linguistic form, conversational structure, message content, or all three, until an acceptable level of understanding is achieved. (Long, 1996, p. 418)
Though not conclusive, there is evidence of a connection between conversation, negotiation and interlanguage development, as articulated by Long's (1996) Interaction Hypothesis:...negotiation for meaning, and especially negotiation work that triggers interactional adjustments by the NS or more competent interlocutor, facilitates acquisition because it connects input, internal learner capacities, particularly selective attention, and output in productive ways (pp. 451-452)
In campus-based courses, classroom interactions provide opportunities for speaking practice while at the same time allowing the language teacher to support and monitor learner progress. Even though distance language courses enjoy an increased popularity, many potential students believe they lack the same levels of interactivity and teacher support as campus-based courses. In particular, many would-be distance language learners do not perceive how they can learn to speak the language. This is despite the increased integration of computer mediated communication (CMC) tools in language courses being delivered in combined distance/online mode (Goodfellow, Manning, & Lamy, 1999; Kotter, 2001; Shield & Hewer, 1999). As part of the Cassamarca Foundation project Italian Online at the University of South Australia, the potential of online chatting as a bridge to face-to-face interaction and which is conducive to negotiation in Italian is being explored.
In particular, this article reports on the first stage of a study which aims to fill a gap in the research on the pedagogical uses of live chat and contribute to the debate in this area by exploring the following questions:
- Does chatting with unfamiliar native speakers (NS) in single session exchanges offer opportunities for negotiation of meaning in open-ended tasks, carried out without teacher supervision?
- What are the principal triggers for negotiation and modification of interlanguage in these interactions?
- Are NS chat rooms likely to offer an optimal environment for SLA, even for learners studying at a distance?
The main focus of the analysis is to ascertain whether negotiation is a feature of NS-learner chatting, as occurs in conversation, and hence whether it can fill a gap in distance language courses, which may lack the SLA benefits that oral interaction and negotiation provide. Based on findings regarding the above, this study draws some conclusions regarding NS chat rooms as pedagogical tools for distance learners and considers further research directions.
Previous Research on Synchronous CMC
A number of researchers of Computer Mediated Communication (CMC) have uncovered similarities between text based interactions via computer and face-to-face interactions. These studies (e.g., Chun 1994; Kern 1995; Pellettieri 2000; Smith 2003) have been carried out mainly with campus-based learners working under teacher supervision in computer laboratories but have implications for distance language study.
In particular, two aspects of chatting have been reported in the research as especially effective in promoting language learning: it provides a bridge to face-to-face interaction and is an optimal environment for SLA. On the issue of chatting as a bridge to face-to-face interaction, for example, Sotillo (2000) compared synchronous and asynchronous CMC and identified a much stronger resemblance to spoken language in the former. She found that synchronous CMC presented discourse functions "similar to the types of interactional modifications found in face-to-face conversations that are deemed necessary for second language acquisition" (2000, p. 82).
Chun (1994) investigated the language production of on-campus learners of German who had regular 15-20 minute chats over two semesters, and found that computer-assisted class discussions appeared to facilitate the acquisition of interactive competence, since learners tended to engage in many types of discourse initiation. The electronic discussion nonetheless tended to be similar to written language in terms of linguistic complexity. Chun also emphasised that the displacement of the instructor from a central role made a difference in these discussions, since it gave the learners a greater role in managing the discourse. She suggests that text-based CMC is therefore a useful bridge between written and spoken skills for learners.
Also, Tudini (2002a, 2002b, 2003) found that learner chat discourse displays features that, according to selected indicators of spoken discourse, bring it closer to the oral than written medium. These indicators include repairs and incorporation of target forms, variety of speech acts, discourse markers, and feedback tokens. Gastaldi (2002) even argues that Italian NS chat discourse is "italiano parlato digitato" (typed spoken Italian) and Beauvois (1992) describes chatting as conversation in slow motion due to the slowing down of real time interaction which the text based medium promotes. The "orality" described in these studies points to possible use of live chat as a bridge to face-to-face interaction and the likely presence of negotiations in live chat interactions involving learners.
Researchers have also noted aspects of chat discourse which differentiate it from face-to-face. Of particular interest is Negretti's (1999) analysis of nonnative speakers' chat sessions in English, which addressed differences between chatting and face-to-face interaction from a conversation-analysis perspective. She described structures and patterns which were peculiar to interlanguage produced in a CMC context, even though they were based on conversational strategies. Her analysis thus focused on the distinctiveness of typically oral structures and patterns produced in the CMC context with respect to face-to-face interaction.
The main elements of oral interaction analysed by Negretti (1999) in chat sessions were overall structure of interaction and sequence organisation, turn-taking organisation (especially openings and closings), turn design, expression of paralinguistic features and some pragmatic variables. Her results showed that sequencing and timing were dealt with very differently in the chat sessions compared with conversational settings. These observations were based on chats carried out in a group setting, with group postings and one-to-one postings intermingling, as the Web chat software used in the study permitted. Such a difference might have been less marked in a chat restricted to two participants, where learner postings tend to be more orderly than in group sessions. This is also the case with turns, which were constantly disrupted or overlapping in the Web chat setting described by Negretti.
Regarding the second aspect of chatting reported in the research, there is evidence that chat rooms can represent an optimal environment for SLA, as occurs in conversation, in that they provide opportunities for negotiation of meaning, thus promoting language acquisition according to Long's Interaction Hypothesis, as summarized by Hegelheimer and Chapelle (2000):The most useful interactions are those which help learners comprehend the semantics and syntax of input and which help learners to improve the comprehensibility of their own output ... In face-to-face conversation, comprehension can be achieved through negotiation of meaning ... one reason that negotiation of meaning is valuable is that it can result in modified input -- input which is better tuned to the learner's level of ability. (p. 42)
Several studies on synchronous CMC have found that comprehensible input and modified output results from negotiation of meaning (Blake, 2000; Iwasaki & Oliver, 2003; Pellettieri, 2000; Smith, 2003; Toyoda & Harrison, 2002).
Blake's (2000) study tests the Interaction Hypothesis in a CMC context with learners of Spanish who work in pairs and finds that most of the negotiations between students are triggered by lexical confusions rather than morphological or syntactical ones. He also ascertains the importance of task design in eliciting negotiations, with jigsaw type tasks containing the greatest number of negotiations.
Iwasaki and Oliver (2003) found that implicit negative feedback, with subsequent modification of linguistic production, was a feature of chat interactions of NNS (nonnative speakers) of Japanese interacting with NS partners. NNS used more than a quarter of NS negative feedback in their subsequent production, a figure which was considered useful but lower than in previous face-to-face studies. Pellettieri's (2000) study focuses on the issue of grammatical competence in a study of chatting as a tool for the negotiation of meaning. In her analysis, based on Varonis and Gass' (1985) model for NNS negotiation, she finds that both implicit and explicit feedback leads to repair of errors and incorporation of "target" forms during learners' one-to-one sessions, with 34% of total turns dedicated to negotiation.
While negotiations reported in the synchronous CMC literature apparently tend to follow a similar pattern to those evidenced in face-to-face literature, Smith (2003) builds on Varonis and Gass' (1985) model of negotiation within the CMC context and proposes an expanded model of "computer-mediated negotiated interaction." His study of negotiation sequences is based on the chat negotiations of learners of English working in pairs on jigsaw and decision making tasks and finds that around one third of total turns are composed of negotiation sequences, confirming previous research by Pellettieri (2000). He also finds that lexical problems are the main triggers for negotiation, as in previous face-to-face literature.
Toyoda and Harrison's (2002) study, based on discourse analysis methods sorted negotiations between Japanese NS and NNS into nine categories according to cause of difficulty in communication: recognition of new word, misuse of word, pronunciation error, grammatical error, inappropriate segmentation, abbreviated sentence, sudden topic change, slow response, and inter-cultural communication gap. Many of these communication difficulties led to modified output by both NS and NNS.
Chat environments reputedly even present some advantages over face-to-face interactions that are conducive to SLA. For example, learners are more likely to monitor and edit their language production since they can view their language as they produce it and after the chat sessions by examining chat logs (Kern, 1995; Ortega, 1997; Pellettieri, 2000). This aspect of chatting may promote "noticing," an important principle of SLA research which is described by Swain and Lapkin (1995):In producing the L2, a learner will on occasion become aware of (i.e., notice) a linguistic problem (brought to his/her attention either by external feedback (e.g., clarification requests) or internal feedback). Noticing a problem "pushes" the learner to modify his/her output. In doing so, the learner may sometimes be forced into a more syntactic processing mode than might occur in comprehension. (p. 373)
Also in favour of chat interactions is their equalizing effect which may facilitate SLA. Kern (1995), Chun (1994) and Warschauer (1996) noted the decentralization of the instructor and the greater role of learners in managing the discourse. Kern (1995) observed that the direction of learners' discourse is different from face-to-face: there are more student to student exchanges rather than teacher-student. These studies indicate that domination of discussions by confident students and the teacher is dramatically reduced since even shy students can easily take the floor when chatting.
These are promising findings for learners but require further investigation in the NS-learner chat context. In fact, most of the synchronous CMC studies referred to above are based on interactions that take place between language learners, whereas only few explore interactions that take place between NS and learners (Iwasaki & Oliver, 2003; Negretti, 1999; Toyoda & Harrison, 2002). Even the studies that do analyse interactions between NS and learners sometimes do so in a non systematic way. For example, Negretti's study, based on conversation analysis theory, includes both interactions between learners and those between learners and NS, dyad and group chats, with no differentiation between the two. This is despite the fact that she claims to have observed improvements in the oral proficiency of her participants after two months of chat activities.
Such a scarcity of data is surprising, as there is evidence that learners imitate NS discourse (St. John, 1995) and, according to Toyoda and Harrison (2002), need to have the opportunity to chat with NS since certain aspects of language which come to the fore in chat sessions are normally neglected in teaching.
Also, if chatting is to be considered as a negotiation/speaking tool for distance learners, features of NS-learner interactions have implications for task design and assessment as well as contributing to research on SLA. As noted earlier, Long (1996) cites interactions with NS as those which are most likely to promote acquisition, according to previous studies.
Furthermore, most research on text chat communication tools has been carried out with learners who are chatting under supervised conditions. In order to test the suitability of chat tasks for distance learners, participants in the present study carry out simple, easy to follow chat tasks in their own time. Also, previous research on NS chats is based on group chats (Chun, 1994; Kern, 1995; Negretti, 1999) or dyads where the NS partners are known to the learner (Iwasaki & Oliver, 2003; Toyoda & Harrison, 2002). This study is based only on one to one interactions where the NS is unfamiliar to the learner, mostly has no specific interest in language learning, and interacts with the learner only on one occasion.
More recent CMC studies focusing on negotiation (Blake, 2000; Pellettieri, 2000; Smith, 2003) are based on a task-based instructional setting. The use of communicative tasks dealing with particular aspects of the target language were not appropriate or feasible in the public NS chat setting proposed in this study, which is based on an open ended conversational task, as occurs in the studies by Iwasaki and Oliver (2003) and Toyoda and Harrison (2002).
As articulated in the research questions, this study provides a qualitative assessment of whether public NS chat rooms are likely to offer optimal environmental conditions for SLA by providing opportunities for the negotiation of meaning in an open ended conversational task, in unsupervised conditions which reflect the situation of the distance learner. The main triggers for negotiation and instances of modified output are also described as the first stage to documentation of key linguistic concerns of learners of Italian and Italian NS in such a setting. Based on such findings, this study also aims to draw conclusions regarding
- the usefulness of public NS chat rooms as pedagogical tools in the distance education context, where oral-like activities which can be monitored are most urgently required; and
- further research directions.
The Teaching and Learning Context: Opportunities for Oral Interaction at a Distance
In the case of campus-based students, learners' progress in speaking the target language is supported and monitored mainly in the classroom. As far as external students at the University of South Australia are concerned, metropolitan area (city-based) students are encouraged to attend the language and small group classes which focus on oral interaction. Nonetheless, many do not attend classes, especially non-metropolitan students who are dispersed all over Australia or abroad. Both conventional and new technologies are of assistance in the delivery of distance language programs which seek to address the issue of competence in oral interaction.
Table 1 describes some strategies for the development of speaking skills within distance education programs at the University of South Australia, where students cannot attend intensive courses or regular conversation classes.
Table 1. Tasks and Technological Tools for the Development of Speaking Skills by Students of Italian Enrolled by Distance
Level of compulsion
Monitorability & assessability
Level of interactivity
Conversations with lecturer (by phone or in person)
Multimedia resources (audiotapes, videotapes, CDs)
Italian community radio
Italian government funded conversation classes
Italian national TV via cable/satellite
Italian movies and news broadcasts via the state funded SBS (Special Broadcasting Service)
Projects requiring interviews with the local Italian community
Conversations with Italian neighbours, friends or relatives
Chat conversations (students only)
Chat conversations with native speakers
Voice (audio) emails and forums
A rating from low to high has been provided to assess both the level of compulsion and the degree of monitoring and assessment which the language lecturer can realistically provide for each of the cited activities.1 The level of compulsion concerns the extent to which the listed oral or oral-like activities can be classified as compulsory, as opposed to an optional component of the language program. Unless tasks are a compulsory part of the assessment, students are likely to perceive them as marginal. The degree of compulsion depends on whether the activities are accessible to learners. For example, conversations with NS Italians in the community are subject to NS linguistic knowledge and availability in a particular area. The lecturer has little control on a learner's choice of interactant, who may speak a non standard variety of Italian such as a dialect or may be heavily influenced by the dominant language (in this case English). The learner does not always have the ability to recognize non standard forms. Also, students in remote areas or in predominantly English speaking background areas are unlikely to be able to find NS of Italian. While the "low compulsion" activities can obviously be made compulsory in distance education language programs, alternative more reliable activities are often preferred. The need to provide alternate activities reduces their compulsory nature.
The degree to which activities can be monitored and assessed by the lecturer has also been rated, though again this varies according to the learner's location and level of access to the cited activities. It is obviously difficult for the lecturer to personally monitor and assess real life conversations though they can be recorded and submitted by the learners. Listening comprehension tasks can be problematic when based on unseen subtitled TV programs or live radio broadcasts, unless specific broadcasts are known to the lecturer. Also, while voice forums appear to promote verbal interaction, it is unclear whether this can be reliably considered an "oral" activity given that learners tend to write their contributions before recording them, as might occur in audiotape recordings. Thus this activity is rated as low under the monitorability/assessability column since a spontaneous oral rendition by learners cannot be guaranteed. Synchronous audio is clearly a better option from the point of view of monitoring and assessment.
The level of interactivity has also been rated. Interactivity is defined as permitting negotiation and input by the learner. For this reason, the task of listening to Italian radio or TV programs has been rated at a low level of interactivity.
The above strategies have advantages and disadvantages for the teacher and learner, which is not the purpose of this paper to explicate. Suffice to say that despite the various listening comprehension activities, instructions, tasks and assessment procedures in place to promote oral interaction, the weighting of assessment in relation to this ability is very low (20-30%) and reflects the degree of monitoring that can realistically be provided to external students, especially since some of the most interactive and monitorable activities require lecturers to spend time one on one with students. For this reason, the use of learner only and NS chat has been introduced as an assessed component (5-10%) of both internal (campus-based) and external (distance) courses.
Nine learners of Italian and 49 NS participated in this study. Learner ages ranged from 19 to 40, with a median age of 19.5. The age of NS participants was impossible to ascertain given that NS chat participants chat under cover of a nickname and are under no obligation to provide their actual age. Their socioeconomic background, as declared to learners, was extremely varied, ranging from accountants and lawyers to policemen and factory workers. Learners had previously completed secondary school, final year Italian examinations, and one year of university study before commencing the Italian 3A and 3B courses. Six learners had at least one parent of Italian background (third generation) while three were of Anglo-Celtic background.
This study was conducted over 1 year (March to November 2002) or two semesters at the University of South Australia.
Before introducing the NS chat tool "C6" described in this study, students exclusively used a learner only password protected chat tool available on the university's online learning environment. However, some students complained that they wanted to chat with "someone they could learn from" so a suitable NS Web-based chat tool was introduced as an assessed component of the program.
After a brief introductory session during class time in a networked computer laboratory, internal students were asked to interact with NS of Italian on the selected chat room in their own time so that they could evaluate it as a teaching and learning tool. These students were offered the opportunity of submitting these sessions as part of their assessment, in place of the learner only chat sessions which they were required to submit. Nine students accepted this opportunity in semester 1, with four of these continuing to use the C6 chat tool in semester 2 from their home computers, due to the removal of the chat software from the university computer pools during an upgrade period.
Students were surveyed by means of a questionnaire at the end of semester 1 and semester 2. Since students were involved in both the learner only and the NS-learner chatl, their opinion was sought on perceived advantages or disadvantages of both chat tools. Course evaluation instruments also included questions that invited students to provide feedback on live chat as a language learning tool. In semester 1, all chat sessions were worth 5% of the total assessment, in semester 2 this was increased to 10%.
The corpus of data analysed for this study consists of 49 one to one chat sessions between 49 NS and 9 learners of Italian, recorded during April to November 2002. Session length varied from 14 to 237 turns. A turn was counted each time the "floor" was transferred from one participant to the other, regardless of its length. A total 3,687 turns were recorded and submitted by students as part of this project. Other sessions may have been lost since students were asked to submit only their best sessions. One session was also excluded as it was a group chat rather than one to one.
In order to assess the inclusion of Italian NS chat in both internal/online and distance/online courses, an open ended "conversational" task was set. Students were simply asked to chat with NS with a view to evaluating the live chat as a possible teaching and learning tool. It was considered inappropriate to set more prescriptive goal oriented tasks such as those used in previous CMC literature (Blake, 2000; Pellettieri, 2000; Smith, 2003) during this exploratory stage, especially since the NS involved in the interactions were unknown to both teacher and student. It was also hoped that these interactions would generate ideas for suitable tasks to be carried out in this context.
Criteria for Selection of NS Chat Tool
In selecting a suitable Web based chat tool for student use, the following criteria were considered of paramount importance if the chat software was to be of pedagogical value:
- ease of subscription and use;
- exclusive use of Italian;
- wide range of topics;
- potential for interesting discussions;
- option to save chat session;
- option to chat one on one as well as in group; and
- flexibility in choice of numbers of chat participants (small groups versus large groups).
The chat tool that was selected for this study is C6 (Ci sei, Are you there?) since it matches most of the above criteria and therefore appears to be the most promising from the point of view of teaching and learning. It is important that it offers participants the choice of chatting either one to one or in a group, as there is evidence that one to one chatting better reflects face-to-face conversation. According to Pellettieri (2000) and Blake (2000), dyads are the preferred set up for learners' CMC sessions to promote negotiation of meaning. Furthermore, as an added bonus, C6 offers access to audio and webcam communication options. Other features include the participant profiles section, which provides access to information provided by participants about themselves, including professional and personal interests. Also, when in one to one chat mode, the "solo netfriend" (only netfriends) option and other buttons allow chat participants to signal whether they are available or engaged in conversation, thus permitting a degree of filtering of unwelcome interlocutors. Blacklisting is also possible to prevent contact of offensive interlocutors.
Criteria for Linguistic Analysis
In order to verify whether chat rooms offer optimal environmental conditions for SLA by providing opportunities for negotiation, the linguistic analysis identifies sequences that show evidence of negotiation of meaning and modified output. Identification of negotiation sequences is based on Long's (1996) definition of negotiation:
denser than usual frequencies of semantically contingent speech of various kinds (i.e. utterances by a competent speaker, such as repetitions, extensions, reformulations, rephrasings, expansions and recasts) which immediately follow learner utterances and maintain reference to their meaning ... (p. 452).
According to Varonis and Gass (1985), the conversational flow is interrupted by the need to question particular utterances and request conversational help. Sequences of turns which include instances of both implicit and explicit feedback (including recasts) by NS interlocutors are included as examples of negotiation since research indicates that both types of feedback promote the incorporation of target language forms by learners (Gass & Varonis, 1989). Comprehension checks and clarification requests initiated by either NS or learners are also included as signals2 of negotiation sequences since they point to difficulties in communication and consequent attention to form which promote the modification of learner discourse. The following example of a negotiation sequence cited by Long (1996) includes negative feedback (implicit correction immediately following an ungrammatical learner utterance) and is signaled by a NS clarification request:NNS: Uh, yes...a woman drinking (and bottle) wine, uh, bottle and man drinking (a) beer
NS: Yes and she's drinking a glass or a bottle of wine?
NNS: No, uh, she? She's drinking in (no) glass. (p. 429)
Instances of self repair have also been included as examples of negotiated interaction leading to adjustment of learner output. However, self-repairs due to typographical errors have been excluded as examples of negotiation even though it could be argued that these signal learners' attempt at achieving successful communication with an interlocutor through the noticing of errors. This type of error and modification of output is very common to the chat discourse of both learners and NS given the need to type quickly. However it is not always clear whether learners are correcting a typing, spelling or pronunciation error. Such repairs are therefore only included where they are likely to cause misunderstanding. Total turns which were part of negotiation sequences are also counted for comparison with previous studies.
Instances of modified interlanguage are categorized according to triggers which provoke them, as described by Varonis and Gass (1985). Triggers are operationalized in this study as the problematic part of a phrase which provokes a negotiation routine with the native speaker and/or an adjustment of a learner to his/her production (self repair). These triggers vary in nature and may include lexical, morphosyntactical or semantic problems. In the example cited above, the trigger can be categorized as morphosyntactic and leads to an attempt at repair (modified output) by the NNS.
Sequences which lead to conversational modifications by NS to their own production are not included in the summary of instances of modified output since the focus of this study is learner interlanguage development. A sample of 25% of the data was coded by an independent rater (inter-rater reliability: negotiation sequences = 99%, types of modified output = 100%).3
Negotiation Sequences and Modified Output
Negotiations were a feature of most NS-learner chat sessions. In a total 3,687 turns submitted by students participating in this project, 61 negotiations and 11 examples of learner modified output were evident. However, the total percentage of turns spent in negotiation sequences and adjustment of linguistic output was 9% (328 turns). Table 2 provides a breakdown of the total number of chat sessions, turns, negotiations and instances of modified output per learner.
Table 2. Information on Individual Learners' Chat Sessions with NS
Cases of modified output
There were various triggers for learners to modify their language. Nine of these involved either implicit or explicit input by the NS which led to reformulation to more target like forms by learners to make themselves understood. There were also two self repairs by learners.
Table 3 provides a breakdown of the various types of triggers which provoked modifications by learners to their own linguistic production as they interacted with NS on C6. While misuse of word might rightfully be subsumed under the "lexicon" heading, it is here categorized separately since it refers to lexicon which is known to the learner but which is used incorrectly in a particular context, as in the case of ancora described in Example 1. Lexical problems are otherwise considered to trigger negotiations / modifications over unknown vocabulary.
Table 3. Triggers for Learner Modified Output in NS-Learner Sessions
self repairs (lexical)
misuse of word
As shown in Table 4 below, triggers for negotiation varied considerably in nature but were concerned mainly with lexical issues (30 instances), followed by morphosyntactical problems (14). In most cases they involved the unsolicited implicit or explicit input of the NS, although there were also instances of learners requesting assistance of the NS when they did not understand a word.
Table 4. Triggers for Negotiation in NS-Learner Sessions
misuse of word
Types of Negotiations
In one single session (174 turns) there were six negotiations including two requests for clarification initiated by the learner, one clarification request initiated by the NS and three examples of explicit corrective feedback (recasts) by the NS. The three requests for clarification were semantic in nature. Example 1 presents two linked negotiations. In the first, the NS provides the learner with unsolicited explicit corrective feedback (recasts), while the second negotiation is triggered by the learner's clarification request over a semantic issue, which the NS elaborates on.
ok, che fai nella vita?
ok, what do you do in life?
I study at university
il giornalismo e anche un corso dell'italiano
journalism and also a course in the italian
ancora sono stanca
I'm still tired
ci sono abituato
I'm used to it
I don't understand
avevo la ragazza americana e all'inizio la correggevo
I had an American girlfriend and at the beginning I used to correct her
ah ok, ho capito
oh ok, I understand
In this particular learner's session, the cases of explicit corrective feedback were mainly morphosyntactic in nature. However, only one of the three non-target forms could have potentially triggered misunderstanding. Nonetheless, the NS did not appear to tolerate them and provided corrective feedback as illustrated in Example 1. Example 2 includes an example of a negotiation triggered by misuse of the polysemic ancora (yet / still / even / more). In this case the NS provides an explanation to ensure that the learner has understood his correction.
Example 2. Example of Explicit Corrective Feedback (Recast): NS to Learner
dove vivi in australia?sidney,camberra,perth?
where do you live in australia
do you know it?
si,me ne ero dimenticato,ma non ricordo bene l'ubicazione
yes, I had forgotten it, but I don't quite recall the location
facevano la formula 1 una volta o ancora adesso?
did they have the formula 1 once or even now
north, south, east, west?
lo fanno a melbourne adesso
they have it in melbourne now
non si dice non ancora,non +[più],ha un significato totalmente diverso quello che hai detto
you don't say not yet, no more (+), it has a totally different meaning from what you said
quello che hai detto tu significa che tra un pò la faranno
what you said means that they're having it shortly
ah ho capito
oh I see
Example 3. Recast by NS
mi dispiace se non rispondo cosi veloce, cerco qualche cose sul web per universita'
G: I'm sorry if I don't answer so quickly, I'm looking for something(s) on the web for university
cerco qualche cosa sul web per l'universita'
I'm looking for something on the web for (the) university
Example 4. Recast by NS Followed by Learner Acknowledgement
io devo vado adesso
I have I go now
io devo andare adesso
I have to go now
mi dispiace si tu hai ragione andare
Sorry you're right to go
A total of 15 explicit and implicit recasts by the NS were recorded. In contrast to previous examples, Example 5 shows a less critical NS who requested clarification rather than explicitly correcting the learner, who subsequently repaired her discourse to get her message across when she perceived difficulty in communication:
Example 5. Clarification Requests from NS with Subsequent Modification of Output by Learner
t va di scrivere in pvt?
do you feel like writing in pvt [private]
si, se vuole
if you (polite form) or she wants
se vuole chi??
if who wants??
mi dispiace, voglio dire "se vuoi"
I'm sorry, I mean to say "if you want"
studio l'italiano all'universita
I study Italian at university
com e' l Australia?
what's Australia like?
e' buona, e' bella
it's good, it's beautiful
perche' -- mi dici primo
why-tell me first
cosa vuoi dire
what do you mean
Voglio che mi dici quanti anni hai
I want you to tell me how old you are
io ne ho 23
ok, ho 30, sono vecchia
ok, I'm 30, I'm old
The first question posed by this study was, Does chatting with unfamiliar NS in single session exchanges offer opportunities for negotiation of meaning in open ended tasks, carried out without teacher supervision? Findings appear to confirm previous research which indicates that negotiation is a feature of synchronous CMC, even in an unsupervised setting where the NS is unknown to the learner, with 9% of turns involving negotiation. This figure is lower than in previous learner only chat studies which cite a figure of around one third (Pellettieri, 2000; Smith, 2003). It is also lower than the Iwasaki & Oliver (2003) NS-learner study which mentions a figure of around 25%. Various reasons for this might include:
- educational background of NS;
- NS disposition towards learners of Italian (as compared to teacher);
- level of proficiency of learners involved in learner-NS chat studies;
- personality of NS; and
The C6 NS chatters are from various walks of life, not only university students as in previous studies. There is therefore a great deal of variation in the extent of feedback and sensitivity to learner errors. Some evidence of the importance of this factor is that one of the NS who provided continuous feedback to the learner claimed he had experience as a teacher of Italian. Clearly, the level of training, personality and gender of the NS are factors in the NS-learner chat context. Also, though there is insufficient information in the studies concerned, an intermediate learner of Italian may have a higher level of proficiency than a learner of Japanese at the same or more advanced levels. Comparison with NS-learner CMC studies on learners of non character based languages would be more accurate.
The second question investigated the principal triggers for negotiation and modification of interlanguage in NS-learner interactions. The fact that lexicon is the main trigger for negotiation (30 instances out of 61) confirms previous synchronous CMC research (Blake, 2000; Pellettieri, 2000; Smith, 2003) and face-to-face research, as reported by Smith. While lexicon was the principal trigger for negotiation, morphosyntax was also an important factor with 14 out of 61 instances concerned with grammar and syntax. This points to a considerable attention to form in the NS-learner chat environment.
The third question posed by this study asked, Are NS chat rooms likely to offer an optimal environment for SLA, even for learners studying at a distance? The context of this study was a public Italian NS chat tool which, after downloading procedures on the home computer, can be accessed by learners at all hours, allowing them to interact without teacher intervention, either in a one to one or group chat setting. As pointed out above, negotiation and modified output are a feature of learners' one to one sessions although the extent of this is dependent on the NS. It is also important that recasts (negative feedback) are featured in these negotiations since these are considered likely facilitators of SLA (Long, 1996; Mackey, Gass, & McDonough, 2000).
However, in the sample of NS-learner discourse presented in Tables 1, 2, 3, and 4, the NS displayed intolerance of morphosyntactic errors made by the learners and provided explicit corrective feedback, which did not always lead to immediate modification by the learner. When providing the learner with feedback, it may in the long term be counter productive if the learner is not a confident speaker/writer of the target language. Nonetheless, learners expressed appreciation for the helpfulness of some NS in correcting their Italian. Many NS were very encouraging and praised learners for their excellent Italian. These NS tended not to correct non target forms or negotiate meaning in any way, despite some instances of barely comprehensible language.
However, the presence of negotiations, often including both implicit and explicit feedback by the unfamiliar NS chatter, suggest that the public NS chat environment could facilitate the distance learner's SLA.
NS CHAT ROOMS AS PEDAGOGICAL TOOLS
The chat experience provided some ideas for authentic tasks which can be carried out on C6, since these appeared to occur spontaneously during learners' virtual encounters. These tasks include:
- the setting up of e-mail and SMS text message (mobile phone) exchanges;
- the organization of an exchange to permit learners to visit Italy in exchange for supporting the visit of a NS to Australia; and
- the exchange and discussion of photographs.
The most discussed topics were professional interests, hobbies, and relationships. Some sessions were however quite brief and superficial in content. In order to encourage more goal oriented, cross-culturally useful and lengthier discussions with NS, a focus on intercultural issues has been introduced as an element of chat tasks for both internal and external students. For example, the C6 chatline is an ideal forum to learn more about differences and similarities between modern Italian and Australian culture, especially youth culture. So in the second stage of this project, learners have been asked to quiz chat participants more assertively to investigate Italian chatters' views on various topics, including differences between Italian, Australian and Italo-Australian perceptions of family. Chat sessions are accompanied by a report on learners' investigation on their chosen topic. Learners also currently use the chat experience as a forum for learners to reflect on second language acquisition issues so that they become aware of their own learning processes during chat sessions.
Student evaluations endorsed the use of C6 in the teaching and learning of Italian for the following reasons:
- it gives practice in thinking on your feet;
- it provides practice in speaking with "real" Italians; and
- it avoids problems in arranging appointments to chat with other students, as occurs in learner only interactions.
The only disadvantage cited was the issue of interlocutors seeking virtual sex but they found the chat software made it easy to identify and avoid these people.
From an assessment point of view, the ability to print out the logs of learners' interactions is a useful monitoring and assessment tool for students studying at a distance. Unlike tape recordings or contributions to voice forums, where students can write responses before recording them, the immediacy of real time interactions via computer provides a snapshot of learners' interlanguage as it might occur in an oral setting. In the case of learner only chat sessions, it is also more difficult for students to submit work which is not their own when chatting for assessment points since their password protected chat sessions are recorded in real time and can be accessed by the lecturer and other students enrolled in their course. When using NS chat tools, however, students operate under cover of a nickname and are not easy to identify. They can also submit their logs after they have been corrected by a more proficient speaker. However, if assessment criteria emphasize negotiation and modified output rather than accuracy, students are less likely to change their logs and may be encouraged to negotiate and improve their language in real time. These instructions and criteria have been adopted for the next stage of the study:
Chat sessions will not be corrected word for word since chat language is considered closer to oral than written communication. A high percentage of typing errors is also expected given that chat contributions are being typed in real time and at high speed. The following criteria will however be taken into account:
- fluency (ability to keep conversation flowing);
- richness of vocabulary;
- use of correct grammar;
- idiomatic use of language (showing awareness of possible transfer from structures/expressions of mother tongue); and
- ability to make use of native speaker's knowledge by improving language during chat session.
NS Versus Learner Chat Partners
While the introduction of C6 was a success, it is worth including both learner only and NS chat interactions in courses. Negotiation with other learners appears to serve a different purpose from negotiating with NS. While negotiations generally follow patterns described in face-to-face literature when engaged with NS, open ended tasks elicit mainly self repairs in the learner only interactions which were carried out during the period of this study. This confirms previous literature which indicates that specially designed two way tasks are required to elicit interactive negotiation in learner only chat sessions. It may also be the case however, that some benefit can be derived from the modification of interlanguage which occurs through self repairs in the learner only interactions. A combination of open ended and more prescriptive communicative tasks need therefore to be considered for the distance learner according to the chat tool they engage with and in accordance with the aims of the curriculum. The learner chat tool is purposeful for distance learners since it provides them with an opportunity to establish connections with each other in a colloquial non threatening context. On the other hand, NS chat tools are a valuable connection to the target language and culture which can provide learners with the opportunity to develop colloquial interactive language which is rarely found in textbooks.
LIMITATIONS OF STUDY
Findings of this study are tied to the type of task that learners carried out and are likely to be quite different with a more goal oriented task that is designed to encourage negotiation, although such tasks are difficult to set up when dealing with unfamiliar NS. Results are also tentative due to the small number of learner participants. More work is needed on a larger corpus of chat interactions to provide a more accurate picture of the type and extent of negotiation and modified output in the synchronous CMC context.
Also, more negotiations and cases of modified output may have occurred in unsubmitted C6 sessions since the corpus is based only on what students may consider their "best" sessions for assessment purposes. Learners need to be quizzed on whether they consider the presence of negotiations to be detrimental to their overall grades. A separation of the research project from assessment requirements may yield a more accurate picture of the extent of negotiations within NS-learner sessions.
FURTHER RESEARCH DIRECTIONS
As described in face-to-face literature (Long, 1996), the phenomenon of foreigner talk is obviously an issue in NS chat rooms since NS are likely to adapt their language to the foreign interlocutor's level and avoid typical group chat discourse, as described in Gastaldi (2002) and a separate study on C6 chat discourse by the author (Tudini, 2002). This includes the avoidance of regional varieties of Italian, colloquialisms, and chat jargon, which may be perceived to be more difficult for NNS. In fact learners tended to reveal their learner status within the first ten turns of chat conversation, as if to encourage NS understanding. This is evident at the end of the first negotiation in Example 5 where the learner almost tries to excuse her non target language by stating "studio l'italiano all'università" (I study Italian at university). A research focus on foreigner talk in NS chat tools such as C6 is an area worthy of investigation.
Comparison with open ended learner only interactions by the same group of learners who used both learner and C6 chat tools suggest that further research on differences between learner to learner and NS-learner interactions may yield further insights on suitable tasks for distance learners in the chat context. Negotiations and modifications of learner output appear to present different characteristics according to whether learners are engaging with their peers or with NS. Self repairs seem to characterize learner only interactions while NS feedback, both implicit and explicit, seems to be a feature of NS-learner interactions. This is evidenced by the fact that 50% of learner only negotiations were in fact self repairs. These self repairs constitute the bulk of modifications which were carried out by learners in this context. On the other hand, in NS-learner chat interactions, only two self repairs by learners were present and a more varied picture of negotiation and modified output was present, with the NS as a key figure for both solicited and unsolicited feedback.
While the unfamiliar NS chat interlocutor has a potentially significant role to carry out in providing learners with opportunities for negotiation and SLA, interaction with peers in chat rooms seems to promote a type of thinking aloud or self monitoring. This need to think aloud and evidence of noticing is exemplified by Example 6 where a learner declared three times that her previous turn was full of errors but did not rewrite it:
Example 6. Learner Openly Noticing Errors (3 instances)
Beh, tanti errori, mi dispiace
Well, lots of mistakes, sorry
Interessante, dopo scrivere, vedo i miei errori. Mamma mia! Devo scrivere in più
Interesting, after writing, I see my mistakes. My goodness me! I have to write more
Boh. Molti errori quella volta. Scusa.
I don't know. Lots of mistakes that time. Sorry.
Attention to form is a concern for learners interacting in NS chat rooms. Lexicon is the key concern, followed by structural issues. According to the hard copy printouts of chat logs handed up for assessment, this attention to form and noticing of linguistic problems continues when learners are able to see their chat log printouts, as mainly morphosyntactic corrections are made by hand. As suggested by Ortega (1997), further research needs therefore to probe the issue of salience of errors in chat interactions as it is likely that the noticing of errors is enhanced by the chat medium which allows learners to view their own interactions in written form.
This study set out to evaluate the usefulness of public NS chat rooms as pedagogical tools in the distance education context, where oral-like activities which can be monitored are most urgently required. The evaluation was based on exchanges between a group of learners and unknown NS on the public Italian NS chat tool C6, who interacted as individuals without any teacher support or intervention. The main focus of the analysis was to ascertain whether negotiation is a feature of such chatting, as occurs in conversation, and hence whether it can fill a gap in distance language courses, which may lack the SLA benefits that oral interaction and negotiation provide. The research focussed on the following questions:
- Does chatting with unfamiliar NS in single session exchanges offer opportunities for negotiation of meaning in open-ended tasks, carried out without teacher supervision?
- What are the principal triggers for negotiation and modification of interlanguage in these interactions?
- Are NS chat rooms likely to offer an optimal environment for SLA, even for learners studying at a distance?
Analysis of chat interactions with unfamiliar NS interlocutors indicate that they offer learners a type of informal conversational practice which includes a central component, namely, negotiation of meaning. Negotiation sequences occur over 9% of total turns, in single session exchanges within open ended tasks, carried out without teacher supervision. The data also provide some evidence of the ability of chat discourse to promote learner noticing of errors and attention to form: lexical and structural difficulties appear to be the principal triggers for negotiation and modification of interlanguage in these interactions, as described in previous synchronous CMC literature (Blake, 2000; Pellettieri, 2000; Smith, 2003). Given that learners do negotiate with NS, receiving both implicit and explicit feedback, possibly in a less threatening context than the classroom, it appears that this chat environment is likely to facilitate SLA for the distance learner.
Chatting with NS in chat rooms cannot replace oral interaction in real life contexts, nor can it provide the physical aspects of oral discourse such as pronunciation and other non verbal features. However, the opportunity to negotiate would be of particular use to the external student who aims to become a competent speaker of the target language. Chatting with NS in a chat room where only the target language is spoken provides an authentic and purposeful cross-cultural experience which is otherwise limited to the language teacher, members of the local community or other learners. This is an opportunity which should not be restricted only to external students, according to evaluations of this project. The provision of linguistic feedback does however appear to depend to some degree on the educational background, disposition towards learners, personality and gender of the NS.
Further research is of course required to assess whether gains in SLA are short term or long term and whether learner efforts to improve their language transfer from the chat log to the spoken word, especially where distance learners are concerned.
1. Teleconferencing via computer is worth considering in future as a conversation option for distance learners but is currently limited by bandwidth and low take up issues. E-mail exchanges and written discussion forums have not been included in this list of "conversational" activities, despite their use in distance language programs and potentially high level of interactivity, as described by Mondada (1999). Audioconferencing also had strong potential to fill a gap in the distance learner's oral interaction options. It should also be noted that the strong presence of an Italian community in Australia and other countries, provides some additional options for the distance learner of Italian, which are not necessarily available to all other languages.
3. During a training session, the rater (a NS of Italian) was provided with criteria for analysis described above and examples of negotiation sequences and modified output from the relevant face-to-face literature.
4. The chat discourse is reported verbatim, including the use of quotation marks in place of accents. Student identity numbers have been replaced with the letter L to indicate learner and NS to indicate native speaker. Square brackets are used for explanations and emphasis of non target forms. The translation is literal and does not reflect non target forms unless these are emphasized.
I am grateful for the extremely helpful advice of the anonymous LLT reviewers. Thank you also to Dr. Leo Papademetre for insightful comments on an earlier version.ABOUT THE AUTHOR
Dr. Vincenza (Enza) Tudini is a Senior Lecturer in Italian and is responsible for the development of computer assisted language learning (CALL) at the University of South Australia. Her research interests and publications in the CALL field include multimedia computer software, and synchronous and asynchronous computer mediated communication. With Dr. Giancarlo Chiro and Dr. Andrea Rizzi, she is currently working on a Cassamarca Foundation funded project on the development of online teaching resources for students of Italian who are either campus-based or studying by distance in the School of International Studies, University of South Australia.
Beauvois, M. H. (1992). Computer-assisted classroom discussion in the foreign language classroom: Conversation in slow motion. Foreign Language Annals, 25, 455-464.
Blake, R. (2000). Computer mediated communication: a window on L2 Spanish interlanguage. Language Learning & Technology, 4(1), 120-136. Retrieved May 26, 2003, from http://llt.msu.edu/vol4num1/blake/
Chun, D. (1994). Using computer networking to facilitate the acquisition of interactive competence. System, 22(1), 17-31.
Gass, S., & Varonis, E. (1989). Incorporated repairs in nonnative discourse. In M. Eisenstein (Ed.), The dynamic interlanguage: Empirical studies in second language variation (pp. 71-86). New York: Plenum Press.
Gastaldi, E. (2002). Italiano digitato. Italiano e Oltre 3, 134-137.
Goodfellow, R., Manning, P., & Lamy, M. (1999). Building an online open and distance language learning environment. In R. Debski & M. Levy (Eds.), WORLDCALL: Global perspectives on computer-assisted language learning (pp. 267-285). Lisse, The Netherlands: Swets & Zeitlinger.
Hegelheimer, V., & Chapelle, C. (2000). Methodological issues in research on learner-computer interactions in CALL. Language Learning & Technology, 4(1), 41-59. Retrieved May 26, 2003, from http://llt.msu.edu/vol4num1/hegchap/
Kern, R. (1995). Restructuring classroom interaction with networked computers: Effects on quantity and characteristics of language production. The Modern Language Journal, 79, 457-476.
Iwasaki, J., & Oliver, R. (2003). Chatline interaction and negative feedback. Australian Review of Applied Linguistics, 17, 60-73.
Kotter, M. (2001). Developing distance language learners' interactive competence - can synchronous audio do the trick? International Journal of Educational Telecommunications, 7(4), 327
Long, M. (1996). The role of linguistic environment in second language acquisition. In W. C. Richie & T. K. Bhatia (Eds.), Handbook of research on language acquisition, Vol.2: Second language acquisition, (pp. 413-468). New York: Academic Press.
Mackey, A., Gass, S., & McDonough, K. (2000). How do learners perceive interactional feedback? Studies in Second Language Acquisition, 22, 471-497.
Mondada, L. (1999). Formes de séquentialité dans les courriels et les forums de discussion:une approche conversationnelle de l'interaction sur Internet. Apprentissage des Langues et des Systèmes d'Information et de Communication, 2(1), 3-25.
Negretti, R. (1999). Web-based activities and SLA: a conversation analysis research approach. Language Learning & Technology,3(1), 75-87. Retrieved May 26, 2003, from http://llt.msu.edu/vol3num1/negretti/
Ortega, L. (1997). Processes and outcomes in networked classroom interaction: defining the research agenda for L2 computer-assisted classroom discussion. Language Learning & Technology, 1(1), 82-93. Retrieved May 26, 2003, from http://llt.msu.edu/vol1num1/ortega/
Pellettieri, J. (2000). Negotiation in cyberspace: The role of chatting in the development of grammatical competence. In M. Warschauer & R. Kern (Eds.), Network-based Language Teaching: Concepts and Practice (pp. 59-86). Cambridge, England: Cambridge University Press.
Shield, L., & Hewer, S. (1999). A synchronous learning environment to support distance language learners. In K. Cameron (Ed.), CALL and the learning community (pp. 379-390). Exeter, England: Elm Bank.
Smith, B. (2003). Computer-Mediated Negotiated Interaction: An Expanded Model. The Modern Language Journal, 87(1), 38-57.
Sotillo, S. (2000). Discourse functions and syntactic complexity in synchronous and asynchronous communication. Language Learning & Technology, 4(1), 82-119. Retrieved May 26, 2003, from http://llt.msu.edu/vol4num1/sotillo/
St. John, E. (1995). Language learning via email: Demonstrable success with German. In M. Warschauer (Ed.), Virtual connections: Online activities and projects for networking language learners (pp. 191-197). Honolulu: University of Hawaii.
Swain, M., & Lapkin, S. (1995). Problems in output and the cognitive processes they generate: A step towards second language learning. Applied Linguistics, 16, 371-391.
Toyoda, E., & Harrison, R. (2002). Categorization of text chat communication between learners and native speakers of Japanese. Language Learning & Technology6(1), 82-99. Retrieved May 26, 2003, from http://llt.msu.edu/vol6num1/TOYODA/
Tudini, V. (2002). The role of online chatting in the development of competence in oral interaction. In the Proceedings of the Innovations in Italian Workshop, Griffith University. Retrieved May 26, 2003, from http://www.gu.edu.au/centre/italian/
Tudini, V. (2003). Virtual immersion: native speaker chats as a bridge to oral interaction. Manuscript submitted for publication.
Tudini, V. (in press). Conversational elements of online chatting: Speaking practice for distance language learners? Apprentissage des Langues et Systemes d'Information e de Communication.
Varonis, E., & Gass, S. (1985). Non-native/non-native conversations: A model for negotiation. Applied Linguistics, 6, 71-90.
Warschauer, M. (1996). Comparing face-to-face and electronic communication in the second language classroom. CALICO Journal, 13, 7-26.
Home |About LLT | Subscribe | Information for Contributors | Masthead | Archives