Computer-Mediated Negotiated Interactions:

How is Meaning Negotiated in Discussion

Boards, Text Chat and Videoconferencing?

Cédric Sarré*

Abstract

D espite the amount of published research on the use of text-based

computer-mediated communication (CMC) in second language
acquisition (SLA), very little attention has been paid to voice-based CMC
(audioconferencing and videoconferencing) and to how it compares with
the better known text-based CMC modes. This chapter investigates and
compares the potential of three different CMC modes (discussion board, text
chat and videoconferencing) to foster negotiated interactions (negotiation
of meaning routines and negative feedback), as well as the influence of task
type on such interactions. From the analysis of the interactions generated
by the completion of meaning-focused tasks as part of an online module
of English for specific purposes (ESP) aimed at first year Master’s Degree
Biology students (French non native speakers (NNSs) of English), this study
demonstrates that closed tasks fostered more negotiation work than open
tasks, and that all three CMC modes gave rise to negotiation of meaning.
However, significant differences were highlighted between the three CMC
modes under study: overall, videoconferencing was conducive to a lot more
negotiation of meaning than the other two CMC modes, and discussion
board interactions did not generate any corrective feedback.

Keywords: negotiation of meaning, negative feedback, CMC, text chat, discussion

board, videoconferencing, meaning-focused tasks.

*Centre Interdisciplinaire de Recherche sur les Mobilités (CIRTAI). Université du Havre, France. Laboratoire
Ligérien de Linguistique (LLL). Université d’Orléans, France. e-mail: [email protected]

How to cite this chapter: Sarré, C. (2011). Computer-mediated negotiated interactions: how is meaning negotiated in
discussion boards, text chat and videoconferencing? In S. Thouësny & L. Bradley (Eds.), Second language teaching
and learning with technology: views of emergent researchers (pp. 189-210). Dublin: Research-publishing.net.

189
Chapter 10

1. Introduction

As was noted by Henri and Lundgren-Cayrol (2001), distance learning courses

are often limited to individual learning situations and therefore lacking in
connectivity between learners (p. 5). This criticism has, however, slowly been
tackled by the growing use of information and communication technology (ICT)
to enable such interactions to take place in what has become known as computer-
mediated communication. CMC encompasses a wide range of activities such
as sending e-mails, posting topics on a discussion board, chatting or talking to
someone on the internet, and can thus be either synchronous (taking place in
real time) or asynchronous, text-based (when communication occurs through
the written medium) or voice-based, and one-to-one (one person communicates
with another person) or one-to-many. Since the early 1990s, growing interest has
been shown in CMC for language learning and teaching in the field of second
language acquisition, which has now become a field of research in itself known as
network-based language teaching (NBLT). Indeed, as CMC supports interaction,
it has potential for interlanguage development (Kelm, 1996). If most published
research in the field of NBLT originally dealt with the use of discussion boards,
more varied modes of CMC for language learning and teaching are now being
investigated, even though text-based CMC (especially text chat) still prevails.
Research on voice-based CMC for language learning, however, remains
confidential. The aim of this chapter is to fill part of this gap by examining
the impact three different CMC modes (discussion board, text chat and desktop
videoconferencing) can have on the type of interaction learners engage in, as
interaction is believed to be beneficial to interlanguage development.

1.1. Interaction in SLA

The potential of interaction to interlanguage development has given rise to much

published research (Gass, 1997; Gass, Mackey, & Pica, 1998; Long, 1983, 1996;
Pica, 1994; Py, 1990). Building on Krashen’s (1981) input hypothesis, which claims
that learners acquire a second language (L2) when they are able to understand the
input they are exposed to, these researchers put forward the idea that interaction
can be considered as an essential source of comprehensible input. However,

190
 Cédric Sarré

Long’s (1983) interaction hypothesis takes Krashen’s (1981) theory one step
further as it states that exposure to comprehensible input is not enough to ensure
acquisition and that it should be complemented by social interaction. Indeed, Long
(1983) considers interaction between learners as the best type of input for language
acquisition as it gives learners exposure to more accessible input thanks to all the
adjustments their interlocutors are able to make on their request. In this way, a
more competent speaker will be able to provide a more comprehensible input to
his less competent interlocutor and thus help their interlanguage develop. These
adjustments to the interaction, which occur whenever one of the interlocutors
experiences problems to understand what is being said and which increase input
comprehensibility, are termed negotiation of meaning (Pica, 1994).

1.2. Negotiation of meaning

Negotiation of meaning is claimed to play an important part in SLA as it is

supposed to offer a connection between input, internal learner capacities and
output in productive ways (Long, 1996). In other words, being engaged in
interpersonal interaction during which comprehension problems that can be
negotiated arise supports acquisition (Ellis, 1999, p. 4). Pica (1994) showed that
negotiation routines give rise to a lot more input modification than any other part
of the interaction between learners. The first attempt at modeling negotiation
routines was made by Varonis and Gass (1985) who designed a four-phase
model: first, a trigger (which can be lexis-based, grammar-based, syntax-based
or content-based) sets off the negotiation routine. Phase two consists of a signal
from the interlocutor aimed at showing their non-comprehension. This signal can
take the form of a clarification request, a confirmation check or a comprehension
check (Long, 1983). The next phase is the response to the signal, which can be
a self-repetition, a paraphrase or an incorporation (Long, 1983). The last phase
–which is optional– is the reaction to the response given, its objective being to
signal the end of the negotiation routine and to show that the interactants are
ready to resume their conversation.

In addition to his original theory, Long (1996) later claimed that negotiation
of meaning can also contribute to L2 acquisition through negative feedback:

191
Chapter 10

this form of corrective feedback that learners receive from their interlocutors,
and the opportunities to repair their own utterances that stem from it, are also
suggested to be facilitative to interlanguage development. Negative feedback
can either be explicit (explicit correction by the interlocutor, question asked by
the interlocutor to prompt correction) or implicit (recast, i.e., implicit correction
of the speaker’s utterance by repeating it in its correct form), and can also give
rise to self-correction. It is beneficial to L2 acquisition as it encourages learners
to focus on form while completing a learning task.

1.3. Focus on form and meaning-focused tasks

Long (1983) emphasises the importance for learners to focus on form while they
are processing meaning, which is one of the possible outcomes of negotiation
of meaning as learners’ attention can temporarily shift from meaning to form as
comprehension problems arise (Long & Robinson, 1998). Focus on form through
negotiation of meaning occurs during the completion of a meaning-focused task
as negotiation of meaning and modified output are claimed to be more prevalent
in goal-oriented tasks than in casual conversation (Pellettieri, 2000). According
to Ellis’ (2003) typology, tasks can be either open or closed. Open tasks include
opinion gap tasks, a good example of which could be a debate. Problem-solving
tasks can be either open or closed tasks, depending on how contrived they are.
Following Pica, Kanagy and Falodun (1993) who claimed that closed tasks
are likely to give rise to more negotiation of meaning, Pellettieri (2000) thus
recommends to set up goal-oriented tasks with a limited number of possible
outcomes to encourage negotiation of meaning.

1.4. Research questions

This chapter is based on previously published research on negotiated interaction

through CMC and aims at exploring potential differences and similarities
in the use of negotiation of meaning routines and negative feedback between
three modes of CMC: asynchronous text-based CMC (discussion board),
synchronous text-based CMC (chat) and synchronous voice-based CMC
(desktop videoconferencing). It reports on a study we carried out as part of an

192
 Cédric Sarré

action-research project whose starting point was a problem identified in the field:
the interactional competence –which we see as the “fifth element”, following
Kramsch (1986) and He and Young (1998)– in English of a group of French
Master’s degree students specialising in biology was considered to be under-
developed compared to the other four skills (written and oral comprehension,
written and oral production). Consequently, the solution envisaged was to set
up an online English for specific purposes course following the action-oriented
approach encouraged by the Common European Framework (CEF) for languages
and thus promoting interactions (more appropriately termed “co-actions” in
Puren, 2002) between learners. This was thought to be a way of giving learners
more opportunities to interact in English about subject-specific topics outside the
English classroom. Indeed, the main objective of the course was to help students
develop their interactional competence through computer-mediated collaborative
work, with the technical support of the experimental virtual learning environment
called CLADUO (Centre de Langues A Distance de l’Université d’Orléans).
Negotiation of meaning is an important component of interactional competence
(Kramsch, 1986) and was thus the focus of part of our research project. This
chapter aims at answering the following research questions:

1. Does negotiation of meaning take place in all three CMC modes?

2. Which types of negotiation of meaning occur in the three different

CMC modes?

3. Which types of negotiation of meaning occur during completion of the

two different task types?

2. Method

2.1. Context of the study

This study was conducted in 2008 with a whole class of first year Master’s degree
students specialising in biology at a French university (Université d’Orléans).

193
Chapter 10

As part of their course requirements, all students had to follow a 55h English
class consisting of:

• 24h-face-to-face class aiming at developing their skills in oral scientific

English with a view to making subject-specific presentations in English;

• 6h mini-conference during which all students had to present a paper in

English based on a review article;

• 25h English module online (estimated time) following the CEF’s action-
oriented approach aiming at developing all five skills, with a strong emphasis
on collaborative work and interactional competence development.

All three elements of the class were assessed on a continuous assessment basis. This
chapter only reports on the research carried out about the online part of the class.

2.2. Participants

Between January and April 2008, 15 groups of four non native speakers of
English took part in the class. They were all biology majors enrolled on a
Master’s degree programme. Prior to their participation in the English class, a
computerised test in English was administered to all students using DIALANG,
a language diagnosis system developed by several European higher education
institutions and based on the CEF’s common levels (A1 to C2). In written
comprehension, over a third of the students were assessed at levels A (A1: 7%,
A2: 29%), over half were B levels (B1: 33% and B2: 23%), and the remaining
8% were C levels (C1: 7% and C2: 1%). As for listening comprehension results,
they showed that just under three quarters of the students were levels A (A1:
39% and A2: 33%), while the other quarter was mostly B levels (B1: 20%, B2:
5%). As they were clearly more discriminating than the reading comprehension
results, the listening comprehension results of the test were used to organise
students in mixed-ability groups of four students with a view to ensuring that
less competent students would make the most of peer scaffolding, hence setting
the stage for potential negotiated interactions.

194
 Cédric Sarré

The students were split into three meta-groups according to the CMC mode they
had to use to complete the collaborative tasks: a chat group, a videoconferencing
group and a discussion board group. Four groups of four students were assigned
to each meta-group (that is 16 students for each CMC mode), which makes a
total of 48 participants.

2.3. Tasks

The online part of the class was organised around five different subject-specific
scenarios (as shown in Table 1) whose main characteristics were as follows:

• they were composed of several macro-tasks (corresponding to Ellis’s

(2003) definition of tasks as “real-world activities”, that is meaning-focused
tasks) and micro-tasks (as termed by Guichon (2006) and described by
Bertin, Gravé and Narcy-Combes (2010) as language-oriented tasks
meant to fill language and communication gaps, that is, form-focused
tasks);

• learners were put in realistic situations and were given a main mission to
complete;

• the outcome of each scenario was a written language product that all the
previous micro-tasks were geared towards completing;

• the different tasks were either individual or collaborative (though the final
written production was always an individual task);

• the different tasks were organised in six parts: background and

objectives, getting started, reading time, listening time, sharing time and
writing time;

• the input (written and oral) that was provided to learners was progressively
more complex throughout the different scenarios, as was the type of
written production they were expected to complete;

195
Chapter 10

• two types of collaborative tasks were set up: problem-solving (with a

limited number of possible solutions) and opinion gap tasks (with more
possible outcomes); all collaborative tasks were meaning-focused.

Table 1. Scenario characteristics

Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5
Topic Studies and Safety in The Phytore- GM food
careers in the biology genetics mediation
biological lab of cancer
sciences
Written input Text Scientific Biology Scientific
Scientific
from the article university article abstracts
American textbook from
ministry different
of labor articles
Oral input Interviews Specialised Specialised Specialised Conference
of biological video video video presenta-
scientists tion extract
Collaborative task 1 Opinion Individual Opinion Opinion Opinion
gap (CT1) gap (CT4) gap (CT6) gap (CT8)
Collaborative task 2 Decision- Problem- Decision- Problem- Decision-
making solving making solving making
(CT2) (CT3) (CT5) (CT7) (CT9)
Outcome (written A brochure A poster Course A guide for A brochure
language product) material the general
public

2.4. Equipment and materials

The online module was based on the technical support of Dokeos, a courseware
management system (CMS) which was carefully selected after assessing over
40 different open-source CMSs (Sarré, 2008). The virtual learning environment
thus set up included, among other tools, a chat tool and a discussion board.
As no desktop videoconferencing tool was included in the CMS, an external
application (Flashmeeting, developed for the British Open University by the
Knowledge Media Institute (KMI)) was also selected to complement existing
CMC tools. For research and feedback purposes, all three CMC tools made it
possible for the tutor to have access to learners’ interactions after they had taken

196
 Cédric Sarré

place, which is fairly standard in the case of discussion boards, but not so much so
for the other two modes of CMC (the chat tool automatically created a log file of
the interactions, and Flashmeeting automatically recorded the videoconferencing
sessions which could then be played back using the integrated playback function).
The videoconferencing sessions then had to be recorded using Camstudio, open-
source screen recording software, which captured the sessions in a video file (avi)
for easier data processing. All tools were internet-based and did not require any set
up procedure by students as most of them are JAVA applications.

2.5. Procedure

All 12 groups had to complete the five scenarios over a ten-week period,
each scenario needing to be completed in no more than two weeks. The six
collaborative tasks (three open tasks: CT4, CT6 and CT8, and three closed tasks:
CT3, CT5 and CT7) gave rise to interactions in all 12 groups. It was decided
not to use data from the first scenario in order to give students time to adapt
to the virtual learning environment and format of the scenarios. No instruction
was given as to how long interactions had to be, the main objective being to
complete the task effectively (although videoconferencing sessions were limited
to 30 minutes each).

The data collected included 24 chat log files (text files), 24 discussion board files
(the discussions were copied from the webpage and pasted in a text file) and
24 video files (corresponding to the 24 videoconferencing sessions). The video
files were transcribed, time-aligned and annotated using EXMARaLDA Partitur
Editor, a tool originally designed for the transcription and annotation of spoken
language. What makes EXMARaLDA different from most other transcription
tools is that it can also be used for the annotation of written language (imported
from text files), thus making it possible to research different data types with
the same tool. Chat log files and discussion board files were also imported in
EXMARaLDA Partitur Editor and annotated.

Following Long (1983), Varonis and Gass (1985) and Pellettieri (2000),
negotiation routines were annotated as shown in Table 2.

197
Chapter 10

Table 2. Annotation of negotiation routines

Phase Tag Description
TRIGGER TLEX Lexical trigger
TSYNT Syntactic trigger
TCONT Content-related trigger
SIGNAL SCR Clarification request (What does that mean?)
SCC Confirmation check (Did you mean that…?)
SST Statement of non-understanding (I didn’t get that.)
RESPONSE RMIN Minimal response (Yes/No.)
RSR Self-repetition
RPAR Paraphrase (with lexical elaboration)
RSC Self-correction
RCOMPC Comprehension check (Is it ok?)
REACTION REAC Reaction to response

Following the annotation scheme, all negotiation work was annotated on a

specific speaker-dependent tier (coded [NOM] for Negotiation Of Meaning),
while a separate speaker-dependent tier was devoted to the verbal data itself
(coded [v]), as shown in Figure 1.

Figure 1. Example of annotated data

198
 Cédric Sarré

As for negative feedback (Long, 1996), it was annotated using the scheme
shown in Table 3.

Table 3. Annotation of negative feedback

Tag Description
EXCO Explicit correction: The interlocutor explicitly
corrects the speaker’s mistake (It’s not X, it’s Y.).
QUES Question: The interlocutor prompts the
speaker to self-correct with a question
(Could you say that again?).
RECA Recast: The interlocutor corrects the
speaker’s mistake by repeating the
utterance in its correct form.
INC Incorporation: The speaker repeats
his/her utterance in its correct form
following interlocutor’s feedback.
SELCO Self-correction: The speaker corrects
his/her own mistakes without prompting
from his/her interlocutor.

All negotiation of meaning routines and negative feedback were annotated on

a speaker-assigned tier that was separate to the orthographic transcription tier.
EXAKT, EXMARaLDA’s analysis and concordancing tool, was then used to
count and analyse tagged negotiation work and negative feedback.

3. Results and discussion

3.1. Global results

The very first conclusion that can be drawn from the analysis of the data
collected is the fact that the completion of meaning-focused tasks through
computer-mediated communication does foster negotiated interactions. This
result is evidenced in Table 4 and Table 5, and is in line with previous studies
on text chat (Pellettieri, 2000; Shekary & Tahririan, 2006; Smith, 2003), on
audioconferencing (Jepson, 2005) and, more recently, on videoconferencing
(Zhao & Angelova, 2010). No such study has been found on discussion boards.

199
Chapter 10

Table 4. Negotiation of meaning routines

Phase Tag Number
TRIGGER TLEX 13
TSYNT 7
TCONT 30
TOTAL 50
SIGNAL SCR 15
SCC 15
SST 28
TOTAL 58
RESPONSE RMIN 0
RSR 5
RPAR 34
RSC 3
RCOMPC 3
TOTAL 45
REACTION REAC 17

Our data shows that most negotiation routines were triggered by content-related
problems (30 such triggers were counted), while very few were triggered by
syntactic problems (7 in total), which can be explained by the fact that syntax has
a low communicative load and thus does not foster much negotiation (Pellettieri,
2000). It should also be noted that lexical triggers, which seem to be the main
cause of negotiation routines in text chat (Pellettieri, 2000), were present but not
as the main type of trigger in our mixed-CMC-mode data. However, the variety
of trigger types tends to show that learners engaged in negotiation work on both
form (lexis, syntax) and meaning (content), as illustrated in examples 1 to 3
(Table 5 below). It should be noted that triggers were always clearly identified
as being either lexical, syntactic or content-related. Although combined trigger
types (i.e., a lexical trigger combined with a syntactic trigger, for example)
would not be a problem in the annotation process, no such combined triggers
were found in our data.

Example 1, in Table 5, shows a lexical trigger (the word greenhouse), followed by

a clarification request from the interlocutor (FAN, on lines 6 and 7), the response
to which is a paraphrase that explains what a greenhouse is (CAR, on lines 10 and

200
 Cédric Sarré

11). In example 2, it seems to be CHR’s syntactically deficient sentence (line 3)

which calls for a non-understanding signal from her interlocutor (CAR, on lines
6 and 7), who then repeats her question and self-corrects her mistake (line 10). In
example 3, LAU signals her non-understanding with a clarification request (on
lines 5 and 6), but the problem cannot be attributed to either lexis, or syntax. This
time, the trigger is content-related, as LAU does not understand why MAM thinks
that the biosafety level of the laboratory they are carrying out their training period
in should be reassessed in the near future. MAM then explains (on lines 7 to 11)
that a recent incident that occurred in the lab (a man walking in a corridor carrying
dangerous cell cultures with no specific protection dropped them on the floor and
sprayed them onto someone who happened to be in the corridor at that time) is the
reason why she thinks the biosafety level should be reassessed. LAU then reacts to
the response and shows comprehension.

Table 5. Examples 1, 2 and 3

Example 1: TLEX Example 2: TSYNT Example 3: TCONT
1 CAR: Euh, moreover we 1 CHR: The last question 1 MAM: So, in a few months,
2 have to play safe and euh 2 is what do you think can 2 we will determine a new
3 perhaps build a greenhouse 3 be done the correct racial 3 biosafety level. Do you
4 in order to suppress the risk 4 disparity in the survival 4 agree?
5 of contamination by plants. 5 and treatment of cancer? 5 LAU: Can you explain,
6 FAN: Sorry, CAR, could 6 CAR: Sorry CHR, could 6 please?
7 you repeat please? 7 you repeat? 7 MAM: Yes, a mistake like
8 CAR: I said that euh we 8 CHR: Yes euh, what do 8 this should not occur. When
9 should perhaps build a 9 you think can be done 9 a man walks in a corridor
10 greenhouse. Euh it’s a place 10 the = to correct racial 10 with dangerous cultures, I
11 where we put plants. 11 disparity in the survival 11 think there is a big problem.
12 and treatment of cancer? 12 LAU: Ah, yes, OK, I
13 understand.

Concerning the nature of signals, our data demonstrates that all three types were used
by learners, the most commonly used one being statements of non-understanding
(SST). Confirmation checks (SCC) and clarification requests (SCR, as illustrated in
examples 1 to 3) were equally used. It should be noted that the difference between
the number of signals and the number of triggers (there are more signals than
triggers) can be accounted for by the fact that different signals can be attributed
to the same trigger (Table 6, example 4) and that signals produced by different
speakers can also be attributed to the same trigger (Table 6, example 5).

201
Chapter 10

Table 6. Examples 4 and 5

Example 4 Example 5
1 FAN: So, let’s continue with the 1 CHR: Yes, euh, what do you think
2 last question. What measures 2 can be done the = to correct racial
3 should be taken on a long term 3 disparity in the survival and
4 basis? 4 treatment of cancer?
5 LAU: Euh, in fact I don’t really 5 LAU: Euh, for me, euh. I don’t know
6 understand the question. Can you 6 what to say about that because in fact
7 help me? It seems that we have to 7 I don’t really understand the
8 take measures to avoid this 8 problem. So, I don’t know.
9 problem in the future? 9 CHR: Yes, I totally agree with you.
10 I don’t understand too.

In example 4, LAU signals her non-understanding by successively using a

statement (on lines 5 and 6), a clarification request (on lines 6 and 7) and a
confirmation check (on lines 7 to 9). In example 5, we can see that the same
content-based trigger (the question asked by CHR on lines 1 to 4) can give rise to
signals from both LAU, who uses a statement on lines 7 and 8, and CHR herself,
who also uses a statement (line 10).

As for responses to signals, they are mainly paraphrases (34 such responses
were used), as illustrated in example 1 above. Very few self-repetitions, self-
corrections (as illustrated in example 2 above) and comprehension checks were
found in our data. Occurrences of negative feedback were more modest, as
shown in Table 7, with the exception of unprompted self-corrections (SELCO).

Table 7. Negative feedback

Tag Number
EXCO 1
QUES 0
RECA 6
INC 0
SELCO 112
TOTAL 119

Table 8 shows that only one occurrence of explicit correction was noted (example
6), very few recasts (example 7) were produced (6 in total), and no incorporation
could be found, while over a hundred self-corrections occurred (example 8).

202
 Cédric Sarré

In addition, no occurrence of question was noted: this can be explained by the

very pedagogical nature of such feedback which is not naturally used by NNSs
communicating with each other (Long & Sato, 1983).

Table 8. Examples 6, 7 and 8

Example 6: EXCO Example 7: RECA Example 8: SELCO
1 NAB: Euh, it’s good to 1 GAE: Moreover, we need 1 LAU: I agree with you.
2 think to ameliorate always 2 a washer and bottled 2 Cancer begin by qualitative
3 boats and euh maritime 3 water because imagine if 3 and quantitative
4 transport in order to evitate 4 a solution arrives in your 4 modifications of genes.
5 euh… 5 eyes. Euh, it’s very 5 Sorry, cancer begins.
6 NAO: Euh, excuse me. It’s 6 dangerous.
7 not ameliorate but improve 7 NEL: Euh, GAE, it’s true
8 and not evitate but to avoid. 8 that an eyewash or a
9 shower would be a first
10 step.

In example 6, NAO explicitly corrects NAB’s mistakes (on lines 6 to 8), whereas
in example 7, NEL implicitly corrects GAE’s utterances with a recast of the
word eyewash (on line 8) inappropriately called a washer by GAE on line 2.
Example 8 shows LAU self-correcting a grammar mistake (on line 5) without
any prompting from her interlocutors. Generally speaking, we can say that the
more explicit the correction technique is, the more threatening for the speaker’s
face it becomes: learners thus prefer to use self-initiated, self-completed
repairs because they are less face-threatening acts than explicit repairs of an
interlocutor’s utterances (Schegloff, Jefferson, & Sacks, 1977).

3.2. Results per task type

The distribution of negotiation routines and negative feedback among the two
different task types is clearly irregular, as shown in Table 9 below. With the
exception of the optional phase (phase 4 – Reaction to response), negotiation
routine phases are consistently more numerous in interactions produced during
the completion of closed tasks, the total number of all four phases being 50%
higher in such tasks than in open tasks. The same conclusion can be drawn from
the analysis of the distribution of negative feedback: all three types of negative
feedback are consistently more numerous during the completion of closed

203
Chapter 10

tasks than that of open tasks. These results corroborate findings from previous
research both in traditional settings (Pica et al., 1993) and network-based settings
(Pellettieri, 2000): meaning-focused closed tasks, namely tasks with a limited
number of possible outcomes, completed through CMC are conducive to a lot
more negotiation of meaning and corrective feedback than open tasks. Research
question number three has thus been answered.

Table 9. Negotiation routines and negative feedback per task type

Type Open tasks Closed tasks
Trigger 19 31
NEGOTIATION Signal 22 36
ROUTINES Response 17 28
Reaction 10 7
TOTAL 68 102
Explicit correction 0 1
NEGATIVE Recast 2 4
FEEDBACK Self-correction 51 61
TOTAL 53 66

3.3. Results per CMC mode

The analysis of the distribution of negotiation routines among the three CMC modes
under study shows significant differences, as displayed in Table 10. As the amount
of output produced during interaction was considerably different between the three
modes (both in total number of words and in number of turns), the proportion of
negotiated turns has been calculated and included in the table, in addition to raw
numbers, for the sake of comparison. These results are in line with Zhao and
Angelova’s (2010) recent findings: overall, videoconferencing was conducive to a lot
more negotiation of meaning than text chat. However, despite the low raw number of
negotiation phases found in discussion boards, the proportion of negotiated turns in
discussion board interactions is also superior to that found in text chat. Still, the small
raw number of routine phases generated in discussion board interactions is probably
due to their asynchronous nature, which means that a reply can occur several days
after the original post was first written, thus not really encouraging question/answer
moves as they could take a long time to be completed.

204
 Cédric Sarré

Table 10. Negotiation routines per CMC mode

Phase Tag Text chat Videoconferencing Discussion board
Number % of turns Number % of turns Number % of turns
Trigger TLEX 3 0,1% 9 1% 1 0,5%
TSYNT 0 - 6 0,6% 1 0,5%
TCONT 12 0,5% 16 1,8% 2 0,9%
TOTAL 15 0,6% 31 3,4% 4 1,9%
Signal SCR 8 0,3% 4 0,4% 3 1,5%
SCC 3 0,1% 10 1,1% 2 0,9%
SST 6 0,3% 21 2,3% 1 0,5%
TOTAL 17 0,7% 35 3,8% 6 2,9%
Re- RSR 0 - 5 0,5% 0 -
sponse RPAR 12 0,5% 18 2% 4 1,9%
RSC 2 0,1% 1 0,1% 0 -
RCOM- 0 - 3 0,3% 0 -
PC
TOTAL 14 0,6% 27 2,9% 4 1,9%
Reaction REAC 8 0,3% 9 1% 0 -

If we now take a closer look at the nature of the triggers, we can see that no
syntactic trigger was found in text chat: this can be explained by the fact that
text chat interaction generates shorter turns (9 words per turn on average, versus
55 for videoconferencing sessions and 130 for discussion board interactions),
which means less complex sentences, thus greatly limiting potential syntactic
problems. Content-related triggers were the most numerous trigger types
found in our data.

The most widely used signal type in both synchronous CMC modes is statements
of non-understanding: this could be explained by the fact that statements are
less syntactically complex (i.e., easier to formulate for NNSs) than both other
types (clarification requests and comprehension checks) which are questions,
thus considered to be more difficult to formulate on the spot by many learners.
Not surprisingly though, this is not the case in discussion board interactions
which generate more clarification requests and comprehension checks than
statements. This could be due to the fact that asynchronous contributions can
be more easily thought out, making it easier for learners to produce more
complex utterances.

205
Chapter 10

A closer analysis of response types shows that self-repetitions (RSR) and

comprehension checks (RCOMPC) are absent from both text chat and discussion
board interactions, but not from videoconferencing sessions, which is consistent
with Jepson’s (2005) findings about audioconference and text chat. Self-
repetition seems fairly unnecessary in text chat and discussion board interactions
since interactants still have access to their interlocutors’ previous contributions,
which could explain why no such response could be found in these two CMC
modes. As for comprehension checks, they are sometimes considered to be too
pedagogical (Long & Sato, 1983) to be used naturally by learners. Finally, the
most widely used response type is paraphrases, possibly because this is seen as
being the most effective way of making oneself understood.

The distribution of negative feedback among the three CMC modes is displayed
in Table 11.

Table 11. Negative feedback per CMC mode

Tag Text chat Videoconferencing Discussion board
Number % of turns Number % of turns Number % of turns
EXCO 0 - 1 0,1% 0 -
RECA 2 0,08% 4 0,4% 0 -
SELCO 59 2,6% 53 5,8% 0 -
TOTAL 61 2,7% 58 6,4% 0 -

As was the case for negotiation routines, videoconferencing generated more

negative feedback (in proportion of turns) than the other two CMC modes,
even if text chat globally gave rise to more negative feedback in raw number.
The fact that more occurrences of self-correction were found in text chat and
videoconferencing is not surprising though: due to their asynchronous nature,
discussion board interactions can be carefully thought out and checked for
language, which is not the case in text chat and videoconferencing sessions.
It is thus suggested that the nature itself of discussion board interaction is
accountable for the total absence of negative feedback, as synchrony seems to
be required to foster all types of negative feedback, namely explicit correction,
recasts and self-correction.

206
 Cédric Sarré

4. Conclusion

The analysis of our data has demonstrated that negotiation work occurs
using all three CMC modes, thus providing an answer to research question
number one. However, it has also shown that the distribution of negotiation
routines and negative feedback among these modes is significantly different,
thus answering question number two: videoconferencing sessions contained
more negotiated turns (negotiation routines and negative feedback) than
both text chat and discussion board interactions. In terms of negotiation
routines, our data has also demonstrated that discussion board interactions
contain a higher proportion of negotiated turns than text chat. Nevertheless,
it should also be noted that discussion board interactions generated no
negative feedback at all, which counterbalances their superiority in terms of
negotiation routines.

From a more qualitative point of view, our study reports differences in the
routine and negative feedback types generated with the three different CMC
modes. It supports previous research findings about the importance of task type
as regards the quantity of negotiation work generated (Pellettieri, 2000): closed
tasks were shown to foster more negotiation of meaning (routines and negative
feedback) than open tasks.

Our study also contributes to the discussion of the potential benefits of

CMC-negotiated interactions to interlanguage development. Although our
objective was to better characterise and compare negotiation work in text chat,
videoconferencing sessions and discussion board interactions, our results should
not be generalised to other settings without great caution as many variables
could have influenced our findings. For example, Flashmeeting, the desktop
videoconferencing tool used, does not allow for multiple speakers to talk at the
same time: a queuing system enabling interactants to ask for the floor has to be
used. Our results might have been slightly different if the videoconferencing
system had given different speakers the opportunity to talk at the same time, thus
making interactions even more synchronous (for those who want the floor but
have to wait their turn) than they were with Flashmeeting.

207
Chapter 10

More empirical research is needed to explore the potential of CMC for

negotiation of meaning between NNSs, especially the role of videoconferencing,
which remains very uncommon in such research, and its potential differences
and similarities with audioconferencing (voice chat).

References

Bertin, J.-C., Gravé, P., & Narcy-Combes, J.-P. (2010). Second-language distance learning
and teaching: theoretical perspectives and didactic ergonomics. Hershey: IGI Global.
Ellis, R. (1999). Learning a second language through interaction. Amsterdam: John
Benjamins.
Ellis, R. (2003). Task-based language learning and teaching. Oxford: Oxford University
Press.
Gass, S. M. (1997). Input, interaction, and the second language learner. Mahwah: Lawrence
Erlbaum Associates.
Gass, S. M., Mackey, A., & Pica, T. (1998). The role of input and interaction in second
language acquisition. The Modern Language Journal, 82(3), 299-307. Retrieved from
http://www.jstor.org/stable/329956
Guichon, N. (2006). Langues et TICE: méthodologie de conception multimédia. Paris:
Ophrys.
He, A., & Young, R. (1998). Language proficiency interviews: a discourse approach. In R.
Young & A. He (Eds.), Talking and testing: discourse approaches to the assessment of
oral proficiency (pp. 1-24). Amsterdam: John Benjamins.
Henri, F., & Lundgren-Cayrol, K. (2001). Apprentissage collaboratif à distance. Sainte Foy:
Presses de l’Université du Québec.
Jepson, K. (2005). Conversations – and negotiated interactions – in text and voice chat
rooms. Language Learning & Technology, 9(3), 79-98. Retrieved from http://llt.msu.edu/
vol9num3/pdf/jepson.pdf
Kelm, O. (1996). The application of computer networking in foreign language education: Focusing
on principles of second language acquisition. In M. Warschauer (Ed.), Telecollaboration in
Foreign Language Learning (pp. 19-28). Honolulu: University of Hawai’i Press.
Kramsch, C. (1986). From language proficiency to interactional competence. The Modern
Language Journal, 70(4), 366-372. Retrieved from http://www.jstor.org/stable/326815

208
 Cédric Sarré

Krashen, S. D. (1981). Second language acquisition and second language learning. Oxford:
Pergamon.
Long, M. H. (1983). Native speaker/non-native speaker conversation and the negotiation of
comprehensible input. Applied Linguistics, 4(2), 126-141. doi:10.1093/applin/4.2.126
Long, M. H. (1996). The role of the linguistic environment in second language acquisition.
In W. C. Ritchie & T. K. Bhatia (Eds.), Handbook of second language acquisition. San
Diego: Academic Press Inc.
Long, M. H., & Robinson, P. (1998). Focus on form: theory, research and practice. In C.
Doughty & J. Williams (Eds.), Focus on form in classroom second language acquisition
(pp. 15-41). Cambridge: Cambridge University Press.
Long, M. H., & Sato, C. J. (1983). Classroom foreigner talk discourse: forms and functions of
teachers’ questions. In H. W. Seliger & M. H. Long (Eds.), Classroom-oriented research
in second languages (pp. 268-285). Rowley: Newbury House.
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: Cambridge
University Press.
Pica, T. (1994). Research on negotiation: what does it reveal about second-language
learning conditions, processes, and outcomes? Language Learning, 44(3), 493-527.
doi:10.1111/j.1467-1770.1994.tb01115.x
Pica, T., Kanagy, R., & Falodun, J. (1993). Choosing and using communication tasks for
second language instruction. In G. Crookes & S. Gass (Eds.), Tasks and language
learning: integrating theory and practice (pp. 9-34). Clevedon: Multilingual Matters.
Puren, C. (2002). Perspectives actionnelles et perspectives culturelles en didactique des
langues-cultures : vers une perspective co-actionnelle-co-culturelle. Les Langues
Modernes, 96(3), 55-71.
Py, B. (1990). Les stratégies d’acquisition en situation d’interaction. Acquisition et utilisation
d’une langue étrangère – L’approche cognitive. Le français dans le monde, numéro
spécial, 81-88.
Sarré, C. (2008). Les plates-formes de téléformation dans l’enseignement-apprentissage des
langues : pour un choix raisonné. Les Cahiers de l’APLIUT, XXVII(3), 48-69.
Schegloff, E. A., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the
organization of repair in conversation. Language, 53(2), 361-382. Retrieved from http://
www.jstor.org/stable/413107

209
Chapter 10

Shekary, M., & Tahririan, M. H. (2006). Negotiation of meaning and noticing in text-based
online chat. The Modern Language Journal, 90(4), 557-573. Retrieved from http://www.
jstor.org/stable/4127043
Smith, B. (2003). Computer–mediated negotiated interaction: an expanded model. The
Modern Language Journal, 87(1), 38-57. doi:10.1111/1540-4781.00177
Varonis, E. M., & Gass, S. M. (1985). Non-native/non-native conversations: a model for
negotiation of meaning. Applied Linguistics, 6(1), 71-90. doi:10.1093/applin/6.1.71
Zhao, Y., & Angelova, M. (2010). Negotiation of meaning between non-native speakers
in text-based chat and videoconferencing. US-China Education Review, 7(5), 12-26.
Retrieved from http://goo.gl/XmwkZ

Websites

CAMSTUDIO: http://camstudio.org
CLADUO: http://www.departementdeslangues.com/claduo
DIALANG: http://www.dialang.org
DOKEOS: http://www.dokeos.com
EXMARaLDA: http://www.exmaralda.org
FLASHMEETING: http://flashmeeting.open.ac.uk

210
Published by Research-publishing.net, not-for-profit association
Dublin, Ireland; Voillans, France, [email protected]

© 2011 by Research-publishing.net (collective work)

© 2011 by Authors (individual work)

Second Language Teaching and Learning with Technology: Views of Emergent Researchers
Edited by Sylvie Thouësny and Linda Bradley

Rights: This article is published under the Attribution-NonCommercial -NoDerivatives 4.0 International (CC
BY-NC-ND 4.0) licence. Under this licence, the contents are freely available online as PDF files (http://dx.doi.
org/10.14705/rpnet.2011.9781908416001) for anybody to read, download, copy, and redistribute provided that
the author(s), editorial team, and publisher are properly cited. Commercial use and derivative works are,
however, not permitted.

Disclaimer: Research-publishing.net does not take any responsibility for the content of the pages written by the
authors of this book. The authors have recognised that the work described was not published before, or that it
was not under consideration for publication elsewhere. While the information in this book are believed to be true
and accurate on the date of its going to press, neither the editorial team, nor the publisher can accept any legal
responsibility for any errors or omissions that may be made. The publisher makes no warranty, expressed or
implied, with respect to the material contained herein. While Research-publishing.net is committed to publishing
works of integrity, the words are the authors’ alone.

Trademark notice: product or corporate names may be trademarks or registered trademarks, and are used only
for identification and explanation without intent to infringe.

Copyrighted material: every effort has been made by the editorial team to trace copyright holders and to obtain
their permission for the use of copyrighted material in this book. In the event of errors or omissions, please notify
the publisher of any corrections that will need to be incorporated in future editions of this book.

Typeset by Research-publishing.net
Cover design by © Raphaël Savina ([email protected])

ISBN13: 978-1-908416-01-8 (Paperback - Print on demand, black and white)

Print on demand technology is a high-quality, innovative and ecological printing method; with which the book is
never ‘out of stock’ or ‘out of print’.

ISBN13: 978-1-908416-00-1 (Ebook, PDF, colour)

British Library Cataloguing-in-Publication Data.

A cataloguing record for this book is available from the British Library.