CALL AND TEACHING 7

Susana Sotillo 467
CALICO Journal, 22 (3), p-p 467-496. © 2005 CALICO Journal
Corrective Feedback via Instant Messenger
Learning Activities in NS-NNS
and NNS-NNS Dyads
SUSANA SOTILLO
Montclair State University

ABSTRACT

This exploratory study examines corrective feedback in native speaker-nonnative
speaker (NS-NNS) and NNS-NNS dyads while participants were engaged
in communicative and problem-solving activities via Yahoo! Instant Messenger
(YIM). As “negotiation of meaning” studies of the 1990s have shown, linguistic
items which learners negotiate in face-to-face (F2F) interaction seem to facilitate
their L2 development. This type of research has been successfully extended to
computer-mediated contexts. Recently, findings from second language acquisition
(SLA)-motivated research have shown that negative evidence or incidental
focus-on-form in communicative classrooms appears to encourage learner
awareness that often leads to successful uptake. Six research questions are addressed
in this study. They include the availability of error correction episodes in
NS-NNS and NNS-NNS online dyad work, type of corrective feedback provided
to L2 learners, category of errors, learner uptake or response, type of uptake, and
length of moves. Among the results of this exploratory study we find the following:
error correction episodes are available in an instant messaging context
and, in this context, more indirect corrective feedback that focuses primarily on
grammatical and lexical errors is provided to L2 learners. Also, simple moves
characterize these YIM error correction episodes, and there is evidence of successful
learner uptake.
KEYWORDS
Instant Messaging, Focus on Form, Focus on Meaning, Corrective Feedback, Learner Uptake
INPUT, NEGOTIATED INTERACTION, AND NOTICING IN SECOND LANGUAGE
ACQUISITION
Second language acquisition (SLA) research continues to expand as researchers
in the field focus on a wide range of topics using a variety of methodologies.
A brief survey of SLA shows how the field continues to evolve. For example, a
behaviorist perspective that regarded language acquisition as habit formation was
468 CALICO Journal, Vol. 22, No. 3
eventually replaced by generative theories of language and psycholinguistic theories
of learning. In the 1980s, SLA research shifted its focus from investigations
of morpheme acquisition and question formation to cognitive processes, input,
and interaction. Studies investigating input and interaction of the 1980s revealed
that both factors—learners’ interaction with their interlocutors and the type of input
directed at them—facilitated the acquisition of a second language (Long 1981,
1985; Pica, Holliday, Lewis, & Morgenthaler, 1989; Pica, Doughty, & Young,
1986). SLA research has shown that second language (L2) input is modified as
learners and their interlocutors negotiate for meaning through the use of clarification
requests, repetition, rephrasing, and confirmation checks. Subsequent investigations
of negotiation, or what are known as “negotiation of meaning” studies,
support previous findings that linguistic items which learners negotiate in interaction
seem to facilitate their L2 development (Gass & Varonis, 1994; Mackey,
1999; Pica, 1994; Polio & Gass, 1998; Gass, Mackey, & Pica, 1998).
More recent findings from cognitively motivated SLA research have shed light
on the mental processes of attention and noticing that allow learners to “notice
the gap,” depending on the communicative and cognitive demands of the situation,
between their own output and the more complex linguistic features of the
L2 (Schmidt & Frota, 1986). As Pica (2003, p. 9) points out, “cognitive theories
[from psychology] have had a highly productive impact on SLA research and its
applications.” A major thrust of recent SLA research has been to assist learners
in noticing L2 input by providing both positive and negative evidence that can
be used to restructure interlanguage grammars (Doughty & Varela, 1998; Oliver,
1995; Spada & Lightbown, 1993; VanPatten & Cadierno, 1993; Williams, 1999;
Williams & Evans, 1998). Positive evidence refers to the provision of grammatical
utterances or well formed statements in the learners’ linguistic environment,
whereas negative evidence implies the provision of either implicit or explicit corrective
feedback. According to Long (1996), the type of negative evidence that
brings to the attention of the L2 learners differences between their output and the
target language is known as negative feedback.
There are various methods for providing negative evidence to L2 learners. These
include task-based approaches that raise learners’ consciousness about grammar
and require them to communicate with each other about target language structures
(Fotos, 1994; Fotos & Ellis, 1991; Sheen, 1992). While recent ESL and EFL
classroom research on L2 development has been motivated by the focus-on-form
(FonF) approach, a reconsideration of what is know as focus on forms has been
proposed by Sheen (2002, 2003), who calls for comparative longitudinal research
between this approach and FonF classroom instruction.
From a sociocultural and collaborative perspective, research on SLA has shown
that learner output, or L2 production, can provide a basis for noticing form and
meaning relationships (Swain, 1985, 1998). Likewise, learners are encouraged
to jointly reproduce language from a text read by the instructor through the use
of collaborative output tasks such as the dictogloss (Swain, 2000). For example,
Sullivan and Caplan (2003) have shown that by utilizing an activity dubbed the
“dictowatch,” which is based on the dictogloss, learners were able to focus their
Susana Sotillo 469
attention on a rich variety of forms (e.g., verb inflections, determiners, plural inflection,
vocabulary, and spelling). This noticing of linguistic forms may contribute
to second language development. Notwithstanding the benefits of various
types of negative feedback demonstrated in recent studies, some language
researchers question its role in language acquisition (Grimshaw & Pinker, 1989;
Pinker, 1989).
FOCUS-ON-FORM STUDIES
Empirical evidence from research conducted primarily in communicative and
content-based classroom contexts appears to support focus-on-form (FonF) instruction,
whether planned or incidental, in negotiation episodes. Findings from
recent studies indicate that FonF in meaning-focused activities is an instructional
option that leads to instances of successful learner uptake. It is hypothesized that
successful student uptake may contribute to L2 acquisition by facilitating noticing
and accurate learner output (see Ellis, Basturkmen, & Loewen, 2001; Loewen,
2004; Lyster & Ranta, 1997). However, Smith (2005) recently examined the relationship
between negotiated interaction and learner uptake in task-based synchronous
computer-mediated communication (SCMC) and concluded that learner
uptake is less likely to occur in an SCMC environment than in F2F interaction.
Furthermore, Smith (2005, p. 53) argues that “learner uptake does not seem to be
a key factor in lexical acquisition in an SCMC setting.”
The use of recasts as an implicit form of negative feedback for L2 development
has been investigated in both classroom contexts and nonclassroom environments.
For example, Braidi (2002) investigated the role of recasts in native
speaker/nonnative speaker (NS/NNS) interactions in a nonclassroom setting and
found different patterns of use according to three different types of negotiations
and grammaticality of learner output. As she points out, the role of recasts in L2
acquisition is still under debate.
Age, interactional input, and incorporation of feedback have also been investigated
in NS-NNS and NNS-NNS adult and child dyads (Mackey, Oliver, &
Leeman, 2003; Morris, 2005). Mackey et al. (2003) found that age was significant
concerning the nature of feedback only among NNS-NNS pairs. In addition,
significant differences were found between NNS-NNS and NS-NNS dyads. Although
NSs provided significantly more negative feedback than NNSs in adult
dyads, the feedback that NNSs did provide allowed learners more opportunities
to modify their output. Concerning child-to-child computer-mediated interaction,
Morris (2005) found that children participating in a computer-mediated Spanish
immersion program provided over 50% of implicit negative feedback in the
form of recasts and negotiation moves (e.g., clarification requests, comments, and
questions) to their peers and that most of the corrections involved lexical items.
L2 ACQUISITION IN COMPUTER-MEDIATED ENVIRONMENTS
Research on computer-mediated communication and its potential relationship to
SLA is still evolving and there remain many challenges in adapting theoretical
470 CALICO Journal, Vol. 22, No. 3
frameworks and methodologies from classroom-based studies or quasi-experiments
that focus on F2F interaction between L2 learners or between native-speakers
of a particular language and their non-native speaker interlocutors. The following
section provides an overview of recent research in this new field of inquiry.
Computer-Mediated Communication and Negotiation of Meaning Research
SLA researchers working in CMC learning environments have investigated
whether or not interactive discourse features that are present in F2F communication
and have been shown to facilitate L2 development are also present in CMC
interactions. For example, recent studies of both synchronous and asynchronous
CMC have examined the quantity and quality of language produced by learners
in negotiation of meaning sequences in jigsaw and free-discussion tasks (Smith,
2003; Oskoz, 2004). Researchers investigating SLA in CMC environments find
that the new technologies support quasi-experimental as well as qualitative studies.
Data collection for research and pedagogical purposes is facilitated by the
“logging” capabilities of most CMC programs that allow instantaneous capture
and convenient access to chat data or chat scripts.
In the last decade, investigations of networked language learning have shown
that this type of environment provides more equitable learner participation, with
the instructor becoming part of the ongoing communicative process (Beauvois,
1992; Kelm, 1992; Kern, 1995; Warschauer, 1996), and offers learners opportunities
to use more complex language than F2F communicative contexts (Chun,
1994; Kern, 1995). Nevertheless, differences have been reported between F2F
and CMC interactional patterns (Blake, 2000; Fernández-García & Martínez-Arbelaiz,
2002). For example, in order to ascertain whether or not features relevant
to the SLA process were present in CMC, Fernández-García & Martínez-Arbelaiz
(2002), investigated interaction among learners of Spanish as a foreign language
and found evidence of negotiation sequences. However, the tendency among participants
to resort to the use of the native language to resolve communicative difficulties
did not result in modified L2 output or “pushed output,” a factor Swain
(1985) considers crucial for SLA. The use of the L1 has also been reported in
child-to-child computer-mediated interaction (Smith, 2005).
Several studies have also attempted to gauge language use and development in
CMC environments, including Abrams (2003), Davis and Thiede (2000), Schultz
(2000), and Sotillo (2000). Abrams (2003) compared the oral performance in
German of three groups in synchronous and asynchronous CMC and confirmed
previously reported increases in the quantity of language generated by students in
synchronous CMC environments. However, statistical analyses of the quality of
language indicated no significant differences among the three groups with respect
to lexical richness, lexical density, or syntactic complexity. Sotillo (2000) compared
the discourse functions and syntactic complexity of 25 ESL students’ writing
and found that whereas synchronous discussions elicited conversation that
was similar to F2F communication in terms of discourse functions, asynchronous
Susana Sotillo 471
writing promoted more sustained interactions and greater syntactic complexity. In
her study comparing how L2 learners use peer editing feedback in CMC and oral
discussion, Schultz (2000) found a complex interrelationship among three factors:
students’ level, activity, and medium.
Using an expanded version of the Varonis and Gass (1985) model of negotiated
interaction, Smith (2003) carried out an investigation of task-based SCMC
among intermediate-level learners of English. His findings revealed that learners
spent at least one third of the total turns negotiating meaning, and that most
instances of nonunderstanding were triggered by lexical difficulty. Smith (2003)
also found that decision-making tasks yielded more negotiation sequences than
jigsaw tasks.
CMC has also been investigated from a sociocultural perspective. Darhower
(2002) examined the interactive features of synchronous in-class chat sessions
over a period of nine weeks. Analyses of the discourse generated by participants
showed that learners constructed “a dynamic, learner-centered discourse community”
(p. 272). That is, they used their L2 to express solidarity and practice
their evolving sociolinguistic competence. Lately, the inclusion of aural and visual
resources have begun to alter the CMC medium. For example, using a technique
called “stimulated reflection,” Levy and Kennedy (2004) discuss their use
of a task-cycling pedagogy for language learning. Learners of Italian as a foreign
language completed FonF tasks relevant to their own needs and engaged in stimulated
reflection through audiovisual recordings. This type of pedagogy allowed
them to focus on language forms they knew and could easily reproduce in online
conversations. It also helped them pay attention to forms they did not know or
could not imitate. As Kern, Ware, and Warschauer (2004, p. 246) assert, the new
resources in CMC increasingly complexify and problematize “current notions of
meaning negotiation.” They predict that research using these new modalities will
allow language educators to use the Internet for novel forms of global or transnational
collaborative inquiry, intercultural communication, knowledge construction,
and the negotiation of new roles and identities.
THE PRESENT STUDY
This article describes a small-scale descriptive study of computer-mediated negative
feedback or error correction between NS-NNS and NNS-NNS dyad pairs
working collaboratively on five learning activities (four communicative and one
problem-solving activity) in an instant messaging environment, Yahoo! Instant
Messenger (YIM). Since this is a descriptive study, it must be viewed as an exploratory
investigation of the availability and form of corrective feedback provided
to L2 learners, types of learner errors corrected, learner response or uptake, types
of uptake, and move lengths in the YIM environment. Its results cannot necessarily
be generalized. The study specifically addresses the nature of corrective
error feedback to L2 learners in meaningful communicative and problem-solving
activities via instant messaging.
This investigation is motivated by recent studies of FonF episodes in the communicative
classroom, the type of corrective feedback provided, and the role of
472 CALICO Journal, Vol. 22, No. 3
uptake in allowing learners to notice the gap between their output and the intended
target language forms. Results of studies by Ellis et al. (2001) and Loewen
(2004) have shown that certain characteristics of FonF, type of feedback, and
instructional context can lead to high levels of learner uptake, much of which is
successful. Unlike Loewen’s (2004) recent study that investigated causal factors,
this is an exploratory study that seeks to address six major research questions. The
categories used in the methodology section are adapted from previous F2F and
classroom-based SLA studies that focus on the nature of corrective feedback in
FonF episodes (Loewen, 2004).
Research Questions
1. Are error correction opportunities and actual episodes available to L2 learners
in a instant-messaging environment as they engage in communicative
and problem-solving activities with NS and NNS partners?
2. Who provides more corrective feedback to learners in an instant messaging
environment, advanced NNSs or NSs?
3. What type of corrective feedback (e.g., explicit/direct or implicit/indirect)
is more readily available to L2 learners in this type of online environment?
4. What aspect of language (e.g., vocabulary, grammar, pronunciation, spelling)
is targeted in this specific type of synchronous computer-mediated
error correction episode (ECE)?
5. Is there evidence of learner uptake, successful uptake, or is topic continuation
the norm in these computer-mediated ECEs?
6. What types of moves or turns (e.g., simple vs. complex) are more prevalent
in this type of environment?
Methodology
In order to investigate the nature of corrective feedback and its possible influence
on language learning in this online environment, five 45-minute collaborative
learning activities were originally designed for this pilot project with two main
objectives: (a) to encourage the exchange of information via a specific type of
synchronous communication, YIM, and (b) to provide L2 learners opportunities
to receive and request corrective feedback from NSs or more advanced learners
of English concerning some aspect of the language. The communicative learning
activities consisted of the following: (a) jointly filling out a goals and objectives
questionnaire, (b) synthesizing information inferred from reading newspaper and
magazine articles, (c) negotiating individual perceptions about the content of a
movie or documentary each student had seen separately, and (d) discussing and
evaluating the usefulness of instant messaging as a learning tool. The problemsolving
activity consisted of working out technical problems regarding the installation
and use of Web cams and setting up the audio or “Talk” feature in YIM.
Susana Sotillo 473
Participants
Three NSs of English and three advanced or highly competent NNSs volunteered
to participate in this study. These students were enrolled in an undergraduate
course required for ESL teacher certification, Principles of Second Language
Learning (LNGN 325). The participants, two males and four females, ranged in
age from 24 to 32. Two of the participants were native speakers of Latin American
Spanish and one was a native speaker of Brazilian Portuguese. These teachers in
training were familiar with SLA research that addressed the relevance of both
negative and positive feedback to L2 learners. Prior to their participation in the
study, they had become familiar with the conceptual distinction made in the SLA
literature between direct or explicit negative (corrective) feedback and implicit or
indirect negative feedback when focusing on specific lexical items or grammatical
forms while working on communicative tasks with NNSs. However, they were
not instructed to focus exclusively on error correction or to favor one or the other
type of negative feedback. Participants were encouraged to provide meaningful or
comprehensible input to their ESL partners via YIM.
Interlocutor feedback to L2 learners could focus on some linguistic aspect of
the language such as word order, verb inflections, or lexis. Since participants were
using the audio and visual modalities of YIM, feedback could also target pronunciation
errors and the mechanics of writing (e.g., spelling and punctuation). Their
six ESL counterparts were all volunteers who were not enrolled in ESL classes at
the time. One of those recruited for this study was a male student who lived in Sao
Paolo, Brazil. He and his advanced NNS partner spoke Brazilian Portuguese as a
first language. Three of the other ESL learners spoke Spanish as a first language.
One NNS spoke Vietnamese as a first language and another one Korean. There
were four females and two males, ranging in age from 18 to 36. Five of the ESL
participants had recently completed the high-intermediate-level ESL reading and
writing courses (ESOL 162 and 163) at a large suburban university. They had
responded to a call for volunteers that was emailed to all current and former ESL
students. Their ESL instructors had encouraged them to participate in this 9-week
study as a means of continuing their exposure to English by receiving ESL tutoring
via YIM. In exchange for their participation in this study, the six volunteers
enrolled in LNGN 325 and their six ESL counterparts were all loaned Web cams
and the accompanying software for 5 months, which they could use for a variety
of language learning purposes, including accessing information from the Web and
sharing it with other English-speaking classmates or co-workers. They were also
free to communicate with family members living in the continental United States
or abroad.
Once the six dyads were set up, NS-NNS and advanced NNS-NNS participants
were shown how to use the equipment and software and encouraged to complete
all five collaborative learning activities. Five of the six ESL learners were
randomly assigned to one of five dyads, and the NNS learner from Brazil was
assigned to a more advanced learner of English whose first language was also
Brazilian Portuguese.
474 CALICO Journal, Vol. 22, No. 3
A member of the IT staff and the researcher spent an average of three hours for
each of two consecutive weeks working with members of each dyad. Students
enrolled in the Principles course and their ESL counterparts were encouraged to
contact the researcher via YIM in order to field test the aural and visual modalities
of this type of SCMC. They were taught how to initiate chat sessions and save the
transcripts for subsequent analysis. They were also shown how to use emoticons
and other available features. Technical difficulties experienced by some participants
while completing each of the learning activities were related to incompatible
systems. The visual (i.e., the Web cam) and aural components of YIM (Talk
feature) or of any instant messenger software program work best with a high
speed Internet (cable) connection. Many technical difficulties were experienced
by those using a dial up connection.1 While it was possible to send and receive
attached documents and have a lively text-mediated chat, some of the participants
could not see their partners with the Web cam or utilize the Talk feature.
Upon completion of the required training sessions, the participants were asked
to sign a consent form that indicated their willingness to participate in the 9-week
pilot project and complete five 45-minute activities with their ESL partners via
YIM. Participants were also made aware that their chat logs and Web pictures
would be used in the study. They were also asked to sign a contract with the IT
team that spelled out their responsibilities for the equipment they were borrowing
(the Web cam and associated software). The IT staff was readily available to
help individual members of each of the six dyads. They encouraged participants
to contact them via email or by calling the IT support team between 7:00 a.m. and
midnight, 5 days a week.
Data Analysis
Transcripts for each of the five learning activities were obtained for five of the
six dyads. Since members of one of the NS-NNS dyads were unable to complete
all five learning activities because of scheduling conflicts, the data collected for
this dyad were not included in the analysis. Using the Talk feature in YIM, spoken
data were obtained for four of the six dyads examined. These audio recordings
were approximately 180 minutes long. Although participants in this exploratory
study were taught how to use the Talk feature in YIM, it was used sporadically
and primarily by ESL learners who were experimenting with technology as they
worked collaboratively with their NS and NNS partners in both types of dyads.
Thus both the chat and oral data were combined. Participants were responsible
for setting up their own YIM sessions and solving technical problems that arose in
consultation with members of the IT team. The researcher did not join any of the
sessions under investigation.
Chat logs obtained from each of the five dyads that were investigated and the
transcribed spoken data were first read by the researcher and a graduate assistant.
Error correction episodes (similar to focus-on-form episodes) were identified
using the working definition put forth by Ellis et al. (2001) and subsequently
adopted by Loewen (2004). In their investigation, Ellis et al. (2001) define a foSusana
Sotillo 475
cus-on-form episode (FFE) as consisting of “the discourse from the point where
the attention to linguistic form starts to the point where it ends, due to a change
in topic back to message or sometimes another focus on form” (p. 294) Since
this is a unique type of CMC environment, only some of the categories utilized
in previous form-focused SLA research were suitable for coding and analyzing.
These categories included: type of corrective feedback (i.e., whether teacher-initiated
(reactive) or learner-initiated, error correction and type of learning activity,
focus of error correction, manner of corrective feedback (direct or indirect), total
uptake, instances of successful uptake (or incorporation of linguistic information
into learner output or production at some point during negotiated interaction),
topic continuation (no uptake), and types of moves. Error correction episodes
(ECEs) were first identified and examined in the chat logs by the researcher. Coding
of the data was carried out by the researcher; an experienced undergraduate
student who had been trained to transcribe, code, and analyze discourse data; and
a linguistics graduate assistant.2 The two assistants coded and analyzed transcript
samples separately. Once the coding had been completed, discrepancies found in
the coded data were discussed in order to reach consensus.
Identifying Error Correction Episodes
In the learning activities3 designed for the present study, error correction could be
initiated by either the NS or more advanced NNS in reaction to a linguistic error
produced by the ESL learner. Since reactive FFEs and student-initiated FFEs are
categories commonly used in current FonF research, these were also employed in
this study.
Table 1 below illustrates the coding categories and characteristics of corrective
feedback provided to the ESL learners by both NSs and advanced NNSs while
collaboratively completing various communicative and problem-solving activities
in the YIM environment. The corrective feedback characteristics identified are
based on classification schemes used primarily in previous classroom-based FonF
SLA research that are easily adapted to a CMC environment.
Table 1
Corrective Feedback Episodes and Coding Categoriesa in the YIM Environment
Coding category Definition
Total number of opportunities Instances or opportunities available for corrective
feedback in communicative and problem-solving
activities
Total number of ECEs Actual instances of error correction in dyad work
Reactive Error correction initiated by either NS partner or
advanced NNS interlocutor
Learner initiated Generally self-correction or a query about some
aspect of grammar or the lexicon
Linguistic focus Grammar (syntax and morphology, vocabulary, pronunciation,
mechanics (e.g., spelling and punctuation)
476 CALICO Journal, Vol. 22, No. 3
Communicative learning
activity
NS or advanced NNS reacts to the learner’s inaccurate
use of a lexical item or linguistic form while collaborating
on a communicative activity
Problem-solving activity NS or advanced NNS reacts to learner’s linguistic error
while solving a problem or technical challenge
Length of corrective feedback
moves or turns
Simple
Complex
only one response move
more than one response move
Manner or type of corrective
feedback
Direct
Indirect
explicit (e.g., immediate explicit correction or metalinguistic
explanation)
implicit (e.g., recast, exact repetition, clarification
request, or confirmation check)
Uptake or learner response general learner response to corrective feedback
No uptake/topic continuation learner does not respond and continues with the activity
at hand
Successful uptake learner incorporates targeted linguistic form or lexical
item into his/her output immediately after corrective
feedback or subsequently during negotiation work
aBased on Loewen’s (2004) characteristics of FFEs
Once the corrective feedback episodes were identified in the CMC interactions,
the researcher and the two trained raters coded them for the characteristics described
in Table 1. All chat logs and recorded oral interaction via YIM’s Talk
feature were coded separately by each of the raters; the agreement rate among the
raters was 87%. When lack of agreement existed with respect to the coding of a
learner’s error or of some other trait, the researcher and her two assistants re-examined
the data and worked on reaching consensus.
Illustration of the Coding System
The numbers in excerpt (1) indicate the turns taken by the participants.4 The
learner’s erroneous output is italicized, and his partner’s corrective feedback is
underlined (see also Figure 1).
Excerpt 1
1 K: Can you hear me?
2 R: speak to I can hear you
R: start your voice?
3 K: MY second question …
K: Are you an undergraduate or graduate student?
4 R: OK I can’t see you. Please send invitation for the
webcam.
Susana Sotillo 477
R: I’m student in binesses and I not finish my undergraduate class
5 K: YOU should say, Invite me to view your web cam.
6 R: ok
R: MY plans are to study on post graduation, after look for a good job
and live in a big city or live in a other
country …
R: sorry other
R: invite me to view your web cam?
7 K: That’s ok. Your plans are really interesting. You
should definitely try to move to a big city and have
different experiences. But why did you choose to learn
English?
8 R: what’s this choose?
R: can you see me?
9 K: Well, ‘choose’ means prefer. It’s the same when you say
why do you prefer English than other languages? Do you
understand that?
10 R: First I want learn English and after learn Spanish because
is important
11 K: I think so too, very important. What is your primary
language?
12 R: Portuguese
Figure 1
NNS-NNS Error Correction Episode
Note: K = advanced NNS, R = ESL learner.
In this exchange, Kay is trying to determine Roberto’s reasons for learning
English. She is also trying to solve technical problems by guiding him through
K: Yes. I can see you. Oh, well,
business, thatʼs really interesting.
K: When do you finish your
undergraduate course?
R: On July o: of 2003.
K: O.K. pretty good! So: (typing) what are your
future plans with respect to employment and
professional growth …
R: My plans are to study on post graduation
after look for a good job and live in a big place or
live in a other country … (P) Sorry other.
K: Thatʼs O.K. You donʼt have to say sorry
all the time. Your plans are really
interesting. You should definitely try to
move to a big city and have different
experiences. I had very good ones. But why did you
choose to learn English?
X
478 CALICO Journal, Vol. 22, No. 3
the Web cam activation process. In turn no. 4, Roberto asks Kay to invite him to
view her camera, but he does not manage to convey his message satisfactorily.
This triggers an ECE that focuses on the wording of Roberto’s request. Kay’s response,
turn 5, is a reaction to Roberto’s non-targetlike utterance, which is coded
as explicit corrective feedback. He acknowledges Kay’s corrective feedback in
the fourth line of turn 6. Errors in spelling and morphosyntax in turns 4 and 6 are
not noticed or addressed by Kay. There is an instance of self-correction (“sorry
other”) and evidence of successful uptake as Roberto incorporates Kay’s feedback
into his L2 output (“invite me to view your web cam?” - in bold).
Italicized words and phrases, which correspond to grammatical, lexical, or
spelling errors, are coded as opportunities for error correction. In turn 8, Roberto
asks for an explanation of the lexical item ‘to choose.’ This is coded as a learnerinitiated
error correction episode (query), which Kay attempts to address with a
very brief explanation. She promptly follows up with a comprehension check.
Roberto’s non-targetlike output, which includes the omission of to in infinitive
constructions in the above excerpt, is not detected by Kay. She fails to provide
immediate corrective feedback and continues the discussion that focuses on Roberto’s
career goals and reasons for learning English. This encourages both of
them to shift their focus to the semantic content of the message and thereby miss
an opportunity to focus on form.
Excerpt (2) illustrates the provision of direct or explicit error correction during
an ECE between a highly competent NNS and a less experienced ESL learner.
Roberto is an ESL distance learner chatting from Sao Paolo, Brazil with Kay,
who lives in Nutley, New Jersey. They are discussing the “Sixth Sense,” a popular
movie both of them have seen.
Excerpt 2
12 R: please, invite me to view your web cam.
13 K: sorry, I thought the web cam was on already...
K: ok...So, today we are going to talk about the movie “the sixth sense”
and then practice some pronunciation, OK?
14 R: May I invite some friend my to you married, about 5 friends
15 K: you have to use the word WEDDING. The word MARRIED is an
[adjective or a verb].
K: who do you want to invite?
16 R: ok...the class already start?
17 K: who are the friends you want to invite to the wedding?
18 R: Re, Taiza, Gustavo, Camila and maybe Juliana.
19 K: OK … Let’s start
(six intervening turns)
25 K: So, do you like the movie?
26 R: Let’s start the class?
R: Yes, I love it.
27 K: What do you think about Bruce Willis’ character?
Susana Sotillo 479
Following Smith’s (2003) definition, a turn was identified as a chat move “where
there was a transfer of the [chat] “floor” from one participant to the other” (p. 42),
as indicated by the line numbers in the excerpt above. An ECE in the NNS-NNS
exchange in line 14 is underlined. It begins when Roberto makes an error in his
choice of words (married instead of wedding). This is immediately corrected by
Kay, who explicitly tells Roberto in line 15 that the correct word is wedding.
Kay’s correction is explicit and is followed by a direct question that recycles some
of Roberto’s own words. The discussion is briefly interrupted by six intervening
turns because participants were sidetracked by polite “small talk” before returning
to the topic of discussion. Roberto omits the -ed past participle marker in turn 26,
second line, but Kay fails to notice the error and continues with her discussion of
the Sixth Sense.
In the ECE in excerpt (3), Lupe, the more experienced learner who is working
toward ESL teacher certification, provides corrective feedback to Ana, an intermediate-
advanced learner of English in line 69 (for a sample interaction between
Lupe and Ana, see Figure 2).
Excerpt 3
66 A: that movie it’s incredible
A: how do you write ‘ammaizing’
A: well, I was confuse who was the Matrix
67 L: Basic Comp helped me
68 A: about the movie, I was a little confuse of who was the Matrix
69 L: You have to pay attention to your past tense, okay.
70 A: yes
Figure 2
Sample Interaction Between Lupe (advanced NNS) and Ana (ESL learner)
18 L: Were you surprised to read
about MccCarthy?
19 A: mass jailing_, I donʼt know
what is that about,
20 A: I have to read about this or
21 A: maybe you can explain it to me
22 L: What do you think about that …
(missed opportunities to focus on L2
learnersʼ failure to invert)
• A: I understand the invation of the US to Iraq, a lot of lifes will be lost
• L: McCarthy was someone who thought that everyone
in the media was a communist.
• L: a lot of lives (Recast)
• L: yes
• L: It is very complicated, right?
• A: that is why in the article they mention__ Stalin, and 0
is one of the reason_ 0 US wants to protect cevilians …
• (No evidence of uptake)
480 CALICO Journal, Vol. 22, No. 3
Ana failed to use the past tense marker in explaining her impressions of The
Matrix (turn 66, third line, and 68). Although Lupe provides a metalinguistic explanation
in line 69, she fails to acknowledge Ana’s request for assistance with
the spelling of “amazing” (turn 66, second line) or to address problems with the
complement of ‘confused,’ as is apparent in turn 68. Learner uptake occurs in 70,
where Ana acknowledges Lupe’s suggestion that she pay attention to the “past
tense” participle.
A detailed analysis of the transcripts showed that there were many opportunities
available for both learners and their interlocutors to focus on a specific linguistic
form or lexical item. Numerous errors in morphosyntax were evident in learner
output. However, in two of the five remaining dyads, in which a NS had been
partnered with a NNS, the NSs failed to provide corrective feedback to learners on
many occasions when they produced non-targetlike output. They were primarily
focused on message meaning and kept the chat going without momentarily drawing
the learner’s attention to an incorrect linguistic form. These were coded as
missed opportunities to provide corrective feedback (or to focus on morphosyntax
or lexis). Excerpt (4) illustrates one of these exchanges between Dana and Tina.
Excerpt 4
38 D: You must have lots of opportunities to practice your English. Is there any
particular situation that gives you more difficulty than others? I find that
in Spanish, it’s difficult for me to talk about history because there is a lot
of vocabulary that I am not used to using. Does that happen to you?
39 T: You must have felt sad at the beginning, but I’m sure that you keep close
contact with the rest of the family as we do in Mexico. Maybe because
we’re very attach to them. At least, that’s what has happened to me. It’s
great to hear about your daughter because kids are the most authentic
thing we have. In fact, a great deal of my job back there was to work with
children. They’re just great!
T: I think that I need a lot of practice on everyday exchanges. I don’t feel
confortable enough in situations like going to the cafeteria while the ca
shier asks you extrange things like TO GO! in a very fast way and I just
get block!
40 D: I know what you mean. It seems like everyone is in a hurry and just
mumbles the speech. My brother-in-law gets nervous when he feels
rushed and just doesn’t say anything
41 T: I think that I’m always missing a huge part of the message because I’m
not acquainted with that.
42 D: I’ll do some research on common linguistic features that deal with Service
encounters like restaurants and stores. Maybe Dr. S. will have some
suggestions for us.
43 T: Maybe this is the cruel reality that we’re not taught to when you study a
foreign language in your home country, but anyway I’d like to overcome
part of this problem by working with you as well as with other people.
Susana Sotillo 481
44 D: I think that it is impossible for a foreign language class to incorporate
all aspects of culture into the class. A teacher can try and expose the stu
dents to some expressions that are common or familiar to the teacher, but
communicative competence (the natural use of the language) can only be
fully understood by becoming part of the culture. What do you think?
The missed opportunities for error correction or for the provision of corrective
feedback that focuses on some aspect of morphosyntax or lexis in these exchanges
are underlined above. Common errors present in learner output include
inappropriate choice of vocabulary (lexicon), omission of past participle marker
(-ed), and third person singular marker (-s). Tina appears to transfer her spelling
conventions from Spanish onto English (confortable, estraño) and uses non-targetlike
expressions, such as “I just get block--.” She consistently leaves out the
past participle marker for regular verbs (-ed), but Dana fails to point this out to her
either directly or indirectly as they collaborate in the completion of various communicative
and problem-solving learning activities. She does try to elicit learner
output by asking the learner’s opinion about communicative competence (What
do you think?).
An analysis of the chat and spoken data showed that NSs provided more indirect
corrective feedback to L2 learners. In Excerpt (5) below, the NS, Dana,
recasts Tina’s non-targetlike output and avoids explicit corrective feedback or
metalinguistic explanations:
Excerpt 5
18 D: Where’s your sister?
19 T: Well, the owner of the cat is there, she’s there. She’s Jessica. And, the
third one is in Guadalajara right now.
20 D: Aren’t you from Guadalajara?
21 T: Yes, we’re all from Guadalajara, but my father got a good job opportu
nity in Zacatecas which is four hours far away.
22 D: Four hours away.
23 T: Four hours away. You see, those are my problems. Four hours away
24 D: Just repeat and you’ll remember
25 T: And, well, they’ve been there for like nine years, but at this time I think
they are thinking of going back to Guadalajara. I think that’s part of the
reason of my sister starting her professional degree here.
Dana uses a recast, shown in bold, to bring the correct lexical expression to
Tina’s conscious attention. Tina, the L2 learner, repeats the targetlike utterance
and incorporates it into her output in line 23. In line 25, Tina incorrectly uses a
possessive phrase (of my sister), but Dana fails to notice it and continues to focus
instead on message meaning.
One of the NSs in the NS-NNS dyads was a young male of Chinese ancestry.
He often dominated both types of exchanges, text-mediated and oral, and had a
482 CALICO Journal, Vol. 22, No. 3
tendency to interrupt his NNS partner. Initially, this NS had been very patient and
used clarification requests and comprehension checks when jointly filling out the
needs assessment questionnaire, but examination of the chat logs showed that he
was often in a rush to complete the remaining activities (e.g., discussing a movie
both had seen, summarizing an article, or solving technical problems). This might
have prevented him from noticing the learner’s errors and providing appropriate
corrective feedback. Excerpt (6) illustrates a short exchange between Walt (the
NS) and Betsy (the NNS).
Excerpt 6
15 W: So, B., what do YOU plan to do after graduation?
16 B: I am thinking of going to graduate level
17 W: You’re a Linguistics major, no?
18 B: I am now and I want to continue of doing that
19 W: What were you studying before you switched to Linguistics?
20 B: well, liberal arts in Union College, Cranford ‘cause I didn’t really know
what to do
21 W: Heh, me too. I have an Associate’s Degree in Liberal Arts from Essex
County College in Newark
22 B: yeah?
B: how about you?
B: what are you going to do after you graduate?
23 W: I’m working for 1 year and I’ll pursue a Master’s of Fine Arts in English
Creative Writing
24 B: that’s nice
Betsy uses a non-targetlike expression in line 16, but Walt does not notice it
and continues his exchange with Betsy. Once again, in line 18, there is evidence
of non-targetlike phrasal expressions in Betsy’s output (continue of doing ...),
but Walt provides neither implicit nor explicit corrective feedback and focuses
instead on the semantic content of the message.
Data Analysis
One-way chi-square statistical tests5 were performed on frequency data for five
dependent variables related to the ECEs in both NS-NNS and NNS-NNS dyads,
including the total number of ECEs, types of learning tasks (communicative vs.
problem-solving), and form of corrective feedback (i.e., NS- or NNS-reactive direct,
NS- and NNS-reactive indirect, and learner-initiated error corrections). The
elimination of one of the NNS-NNS dyads prevented the statistical analysis of
differences between dyad types. A subsequent chi-square analysis was performed
on nine dependent variables, including learner error type, uptake, successful uptake,
topic continuation, and length or duration of corrective move (simple vs.
complex) in both NS-NNS and NNS-NNS dyads. The results of these chi-square
analyses are presented in the following section.
Susana Sotillo 483
RESULTS
Table 2 below shows raw numbers and percentages of the total number of error
correction episodes (ECEs), type of learning task triggering error correction, and
different types of error correction (reactive direct, reactive indirect, and learner
initiated).
Table 2
Error Correction Episodes, Type of Learning Activity, Reactive Indirect and Reactive
Direct Error Correction, and Learner-initiated Error Correction
Dyads Opportunities
for error
correction
Error
correction
episodes
(ECEs)
Error corrections
(communicative
learning
activities)
Error corrections
(problemsolving
activities)
Error corrections
(reactive
indirect)
Error corrections
(reactive
direct)
Learnerinitiated
ECEs
(query or
self-correction)
NNS-NNS
(2 dyads;
six hours)
95
(60%)
46
(70%)
36
(84%)
10
(45%)
13
(43%)
22
(92%)
11
(100%)
NS-NNS
(3 dyads;
five hours)
64
(40%)
19
(30%)
7
(16%)
12
(55%)
17
(57%)
2
(8%)
0
(0%)
Totals 159
(100%)
65
(100%)
43
(100%)
22
(100%)
30
(100%)
24
(100%)
11
(100%)
A one-way chi-square analysis was performed on five of the dependent variables
(error corrections in communicative learning activities, error corrections
in problem-solving learning activities, reactive indirect error corrections, reactive
direct error corrections, and learner-initiated error corrections) for each of
the dyad groups. The results indicated significant differences among these five
variables between the NNS-NNS dyads, χ2 (4, n = 92) = 59, p = .001. Likewise, a
one-way chi-square analysis of the frequencies of these five dependent variables
revealed statistically significant differences in the NS-NNS dyads, χ2 (4, n = 38)
= 25, p = .001.
In addition to the significant differences among variables in each group of dyads,
the differences between the two groups of dyads become apparent when the
frequencies and percentages displayed in Table 2 are examined. It is important to
point out that although there were two NNS-NNS dyads compared with three NSNNS
dyads, NNSs spent more time interacting and negotiating meaning and form
with their ESL partners (six hours vs. five hours). The more advanced learners or
NNSs in the NNS-NNS dyads had 95 opportunities for error correction as they
collaborated on communicative and problem-solving activities with their ESL
counterparts. They provided corrective feedback to learners 46 times out of a total
of 95 opportunities (48%). In contrast, the NSs in the NS-NNS dyads provided
corrective feedback to their ESL partners only 19 times out of a total of 64 opportunities
(29%). ECEs were identified and analyzed in aggregate text-based (chat
logs) and spoken (Talk) data. (When the two categories were analyzed separately,
484 CALICO Journal, Vol. 22, No. 3
no significant differences were found between them.) In the aggregate data, 70%
of the ECEs were found in NNS-NNS dyads and 30% in NS-NNS dyads. Most of
the error corrections provided by the more advanced NNSs occurred in communicative
learning activities ( e.g., filling out a needs-assessment questionnaire, discussing
the main topic of a movie, and synthesizing information) and accounted
for 84% of all the communicative learning activities in both types of dyads. With
respect to the type of corrective feedback provided to ESL learners (i.e., direct or
indirect), frequency counts and percentages revealed that the NNS partners provided
mainly direct or explicit corrective feedback to learners compared to that
provided by NS partners; 22 times out of a total of 65 ECEs (34%) versus 2 times
out of 65 ECEs (3%), respectively. Most of the corrective feedback provided to
ESL learners by the NSs was of an indirect or implicit nature. When compared
with the amount of indirect corrective feedback provided by the advanced NNSs,
frequency counts and percentages show that the NSs provided indirect feedback
17 times out of the total of 65 ECEs (26%) and the NNSs provided indirect feedback
13 times out of 65 ECEs (20%). Overall, 46% of all the error corrections
(30/65) provided to L2 learners in both types of dyads was of an indirect nature.
Learner-initiated error correction episodes were found primarily in the NNS-NNS
dyads (17% [11/65] of all ECEs).
Table 3 presents frequencies and percentages for the types of learner errors
addressed by NSs and NNSs in the ECEs, total learner response to corrective
feedback (i.e., uptake), successful uptake, topic continuation (i.e., unsuccessful
uptake), and length of error correction moves (simple and complex) following
learners’ non-targetlike output. The four different types of errors identified include:
vocabulary, grammar (morphosyntax), spelling, and pronunciation.
Table 3
Error Types in Learner Output, Total Occurrences of Uptake, Occurrences of Successful
Uptake, Occurrences of Unsuccessful Uptake (Topic Continuation), and
Length of Error Correction Moves (Simple vs. Complex)
Dyads Error Types
Total
uptake
Successful
uptake
Topic
continuation
Length of error
correction move
Vocab. Gram. Spell. Pron. Simple Comp.
NNSNNS
16
(64%)
18
(67%)
11
(92%)
1
(100%)
20
(63%)
14
(58%)
26
(79%)
37
(70%)
9
(75%)
NSNNS
9
(36%)
9
(33%)
1
(8%)
0
(0%)
12
(37%)
10
(42%)
7
(21%)
16
(30%)
3
(25%)
Totals 25 27 12 1 32 24 33 53 12
Grammatical, lexical, and spelling errors triggered most of the error correction
episodes in both NS-NNS and NNS-NNS dyads. Only one pronunciation error
was directly addressed by a competent NNS in the NNS-NNS dyads. Grammar
(morphosyntax) was explicitly addressed in L2 learner output by the more advanced
NNS partner in the NNS-NNS dyads. As shown in Table 3, 41% (27/65) of
Susana Sotillo 485
all error types identified and corrected in both types of dyads were grammatical.
Lexical errors represented 38% (25/65) of all error types identified in L2 learner
output in both NS-NNS and NNS-NNS dyads. Of the total number of instances
of uptake, 63% (20/32) occurred in the NNS-NNS dyads, of which 58% (14/24)
were successful. Uptake in NS-NNS dyads accounted for 37% of all uptake in
both types of dyads. Topic continuation (unsuccessful uptake) was more prevalent
in the NNS-NNS dyads than in the NS-NNS dyads (26 instances vs. 7 instances,
respectively). In the NS-NNS dyads, successful learner uptake accounted for 31%
of all instances of uptake in both groups (10/32) and for 42% of all instances of
successful uptake (10/24) in the NS-NNS dyads; successful uptake in the NNSNNS
dyads accounted for 44% of all instances of learner uptake (14/32) in both
groups and 58% (14/24) in the NNS-NNS dyads.
Corrective feedback moves in all NNS-NNS ECEs took place immediately
after the learners produced an error while working with their partner on a communicative
or problem-solving activity. Seventy percent of the simple moves (37
out of a total of 53) in both dyad groups occurred in the NNS-NNS dyads. These
moves represented 57% of all the types of moves in the 65 ECEs investigated. In
the NS-NNS dyads, simple moves accounted for 25% of all moves (16/65) and
30% (16/53) of all simple moves found in both groups of dyads. Complex move
lengths were more readily available in the NNS-NNS dyads than in the NS-NNS
dyads (9/12 vs. 3/12, respectively). However, complex moves in both dyad groups
accounted for only 18% (12/65) of all types of move lengths investigated.
In order to test whether significant differences existed among the dependent
variables, a one-way chi-square analysis was performed on nine dependent variables
(e.g., error types such as vocabulary, grammar, spelling, pronunciation, total
uptake, successful uptake, topic continuation, simple corrective moves, and
complex corrective moves) in the NNS-NNS dyads and yielded significant differences,
χ2 (8, n = 152) = 50.9, p = .001. Similarly, a one-way chi-square analysis
was performed on the same nine dependent variables in the NS-NNS dyads. Statistically
significant differences obtained among the nine dependent variables, χ2
(8, n = 67) = 21.98, p = .01.
DISCUSSION
The first research question addresses the extent to which error correction opportunities
and error correction episodes (ECEs) are available in computer-mediated
interaction in which learners work collaboratively with NSs or more advanced
NNSs on communicative and problem-solving activities. The results of this exploratory
study reveal that there were 159 opportunities for error correction in
11 hours of online interaction and negotiation between ESL learners and their
partners in NNS-NNS and NS-NNS dyads. Of the total number of opportunities
for error correction, 65 resulted in actual ECEs. More opportunities for error
correction were found in the YIM chat logs and spoken data obtained from
NNS-NNS dyad participants than from the data obtained from participants in the
NS-NNS dyads (95 vs. 64, respectively). Additionally, the number of actual ECEs
was higher in NNS-NNS dyads than in NS-NNS dyads (46 vs. 19, respectively).
486 CALICO Journal, Vol. 22, No. 3
Though this study involved NS-NNS and NNS-NNS dyads engaged in a specific
type of SCMC, namely, instant messaging, the results provide partial support for
the availability of corrective feedback that focuses on form in communicative or
meaning-focused learning activities as reported in the SLA literature. Results of
recent investigations by Ellis et al. (2001) and Loewen (2004) have shown that
FonF episodes are available in communicative and meaning-focused ESL lessons.
Though the context in which the FonF episodes occurred in previous studies is
different from the one-on-one highly focused interaction of this exploratory study,
analysis of the data shows that ECEs were readily available as ESL learners and
their partners jointly completed learning activities via YIM.
The second research question was “Who provides more corrective feedback to
learners in an instant messaging environment, advanced NNSs or NSs?” A major
finding of this study was that the more advanced NNSs who were enrolled in
ESL teacher certification courses provided corrective feedback on 46 occasions
to ESL learners in the NNS-NNS dyads, compared with the significantly fewer
instances (19) of corrective feedback provided to ESL learners by NSs in the
NS-NNS dyads. The more competent NNSs in each of the NNS-NNS dyads were
able to focus on learners’ output errors more often than the NSs in the NS-NNS
dyads, while, at the same time, processing their output for meaning. Analysis of
the aggregated data (i.e., chat logs and transcribed spoken interactions) shows that
NSs were more keenly focused on the message or information being conveyed by
their ESL partners and less so on lexical items or linguistic forms. They consistently
missed many opportunities available for error correction of learner output,
as exemplified by the exchanges in Excerpt (4). It may also be the case that NS
partners were following politeness forms of American culture that discourage the
correction of regional or foreign language use patterns.
The results of this exploratory study appear to support previous findings on
NS-NNS negotiation research. For example, in a study of NS-NNS interaction by
Gass and Varonis (1984), the researchers found that when NSs were familiar with
NNS speech, they were able to understand the meaning of the NNSs’ utterances
and thus fill in the gaps in learner output. Since the NS interlocutors in this study
were teachers in training, they were more likely to understand the text-mediated
messages of their NNS partners, as well as their utterances when using the Talk
feature in YIM. Also, Gass and Varonis (1985) pointed out that in situations where
there is no shared background between NSs and NNSs who are engaged in meaning
negotiations, the conversational flow is disrupted by incessant interruptions
by one or both of the participants. Another plausible explanation for the results
in the present study is that the NSs did not want to discourage learners from using
their second language in solving problems online or when trying to exchange
information needed to complete various learning activities by directly focusing on
a linguistic form.
The response to the third research question—”What type of corrective feedback
(e.g., explicit or direct or implicit or indirect) is more readily available in a YIM
context?—is that indirect corrective feedback (i.e., implicit negative evidence in
the form of recasts, confirmation checks, clarification requests, and comprehenSusana
Sotillo 487
sion checks) was more readily available to the ESL learners in the YIM environment
of this exploratory study. NSs provided more indirect corrective feedback to
L2 learners than NNSs (17 instances vs. 13 instances, respectively). In contrast, of
24 direct error corrections in both types of dyads, 92% (22/24) were found in the
exchanges between partners in the NNSs-NNSs. Direct error correction by NSs
in NS-NNS dyads accounted for only 8% (2/24) of all direct corrective feedback.
Though 11 instances of learner-initiated error correction episodes were found in
the interactions, these were primarily queries about lexical items (e.g., “What’s
the word? ... I mistake. Is that right?”) or examples of self-correction made by
ESL learners who were assigned to the NNS-NNS dyads. These instances of selfcorrection
are noteworthy given the fact that self-repair has been shown to exist
primarily in NS-NS interaction, where interactants are equally competent in the
language, but not in domains with not fully skilled interactants (Schegloff, Jefferson,
& Sacks, 1977), as was the case in the study described here.
Turning to the fourth research question, the aspect of language that is targeted in
these synchronous computer-mediated error correction episodes (e.g., vocabulary,
grammar, pronunciation, spelling), the results indicate that 28% (18/65) of all
error types corrected by NNSs focused on grammar, primarily morphosyntactic
errors such as past participle marker (-ed) omission, third person singular marker
(-s) omission, and inaccurate word order. However, of all error types corrected by
NSs, only 14% (9/65) focused on grammar. Vocabulary errors were also corrected
in L2 learner output by both NNSs and NSs and represent 38% (25/65) of all error
types targeted for correction. Since the online exchanges were text mediated,
spelling errors were often targeted for correction by advanced NNS partners but
mostly ignored by NS partners (17% vs. 2%, respectively).
With respect to the fifth research question posed—”Is there evidence of learner
uptake, successful uptake, or is topic continuation the norm in these computermediated
ECEs?”—the results of this study indicate that learner response or uptake
occurred in these online interactions. ESL learners responded to corrective
feedback provided by both NSs and NNSs. There were 32 occurrences of uptake
(learner response) found in both types of dyads (NS-NNS and NNS-NNS) in the
YIM environment, and 75% of these were instances of successful uptake (24/32).
ESL learners appeared to successfully incorporate corrective feedback provided
by their partners into their subsequent interlanguage production. Though the context
of this study is different from Loewen’s (2004) investigation, these results
seem to indicate that there is evidence of both general learner uptake, which in
many cases consists of a simple “yes” or “OK” response, and successful uptake
while working with partners in meaning-focused and problem-solving activities.
In discussing her findings, Loewen (2004) states that a variable that predicted both
uptake and successful uptake was response. That is, the responses were teacherinitiated
moves such as recasts or inform attempts in complex FFEs. In this exploratory
study of corrective feedback in communicative and problem-solving
learning activities in YIM, the NNSs and NSs provided both explicit and implicit
correction to learners as they reacted primarily to grammatical and lexical errors
in the ECEs.
488 CALICO Journal, Vol. 22, No. 3
In addition to successful uptake, 51% of the corrective feedback (33/65) was
neither acknowledged nor incorporated by the learners into their output following
the ECEs in NS-NNS and NNS-NNS dyads. Topic continuation often followed
these missed opportunities for error correction. More occurrences of topic continuation
moves by ESL learners were observed in the NNS-NNS dyads than in
the NS-NNS dyads. Following an error correction move by their partner, the chat
logs show that when negotiating information, ESL learners would often continue
to focus on the semantic content of their partner’s electronic message or utterances
when using the Talk feature. When engaged in problem-solving activities
(e.g., learning to activate their Web cams), they would carry on with the task at
hand rather than acknowledge or incorporate their partner’s lexical or grammatical
modifications into their output.
Though this study did not examine recasts as a separate category, instead combining
them with other indirect forms of corrective feedback such as clarification
requests, confirmation checks, and comprehension checks, the rates of successful
uptake in both NS-NNS and NNS-NNS dyads partially support some recent
findings by SLA researchers who have examined the role of NS recasts in different
types of NS-NNS negotiated interaction. For example, Braidi (2002) found
that recasts occur in NS-NNS interactions and that different types of negotiation
and NNSs’ level of grammaticality affect to some degree the occurrence of these
recasts (Braidi, 2002). In the present study, the NSs provided more indirect corrective
feedback to ESL learners, and 35% of this corrective feedback was in the
form of recasts; direct corrective feedback such as singling out the incorrect form
or providing metalinguistic explication accounted for 11% (2/19) of all error corrections
made by the NSs.
Concerning the sixth research question—the prevalence of specific types of
moves or turns (i.e., simple vs. complex) in an instant-messaging environment—
the results of this study show that simple moves, or moves immediately following
the learner’s erroneous output, characterized the length or duration of most of
the corrective moves. Simple moves occurred 82% of the time (53/65). In other
words, advanced NNSs and NSs collaborating with less experienced L2 learners
in the online learning activities provided immediate feedback to the learners.
Moves of a more complex nature (i.e., error correction or negative feedback
which is spread out beyond one or two moves) accounted for only 18% (12/65) of
the total length or duration of corrective moves.
The provision of corrective feedback in a computer-mediated environment, especially
one involving instant messaging, should ideally involve simple moves. In
examining the chat logs and the Talk feature data, it became apparent that when
either the NS or the more advanced NNS partner extended the provision of corrective
feedback beyond two or three moves, the learner often chose to ignore it. One
may speculate that failure to provide immediate corrective feedback sends a signal
to the L2 learner that the focus of the interaction is primarily on the semantic
content of the message. Not taking advantage of opportunities for error correction
or delaying error correction in this type of closely focused online interaction diminishes
its effectiveness in encouraging learners to incorporate the feedback into
Susana Sotillo 489
their subsequent output. These findings may be in keeping with previous results
concerning NS-NNS negotiated interaction in the SLA literature. Pica (1988),
who investigated interlanguage adjustments in NS-NNS negotiated interaction,
found that in one-signal negotiations with their NS interlocutors, L2 learners
modified approximately 50% of their initial utterances incorporating morphological,
phonological, semantic, and syntactic changes into their output. However,
when L2 learners engaged in extended negotiation with NSs, they produced very
few modifications. A plausible explanation offered by Pica (1988) is that since the
NS correctly understood and recast the NNSs’ intended meaning, the NNSs had
very little motivation to modify their output.
It should be pointed out that research by Smith (2003) on computer-mediated
negotiated interaction suggests that the lack of turn adjacency commonly found
in F2F interaction affects the rate of response to triggers in the negotiated interaction.
Smith (2003) urges that a delay feature be factored into future models of
SCMC.
Figure 3 presents a proposed model of corrective feedback in focused one-onone
exchanges between NS-NNS and NNS-NNS dyad partners working together
on learning activities in an instant-messaging environment.
Figure 1
Proposed Model of Corrective Feedback to Learners by NS-NNS and NNS-NNS
Dyad Partners in Learning Activities in an Instant-messaging Environment
The provision of comprehensible input is feasible in SCMC and instant messag-
NS or more
advanced NNS
engaged in
communicative/
problem-solving
activities with
L2 learner in an
instant-messaging
environment
Number of
error
correction
episodes
(ECEs) while
participants
are engaged
in online
learning
activities
NS or NNS
partner reacts
to L2
learnerʼs
erroneous
output or
utterance
L2 learner
elicits
metalinguistic
explanation,
initiates selfcorrection
or
requests error
correction/
clarification
Learner uptake:
learner responds
to corrective
feedback or
notices linguistic
form/lexical item
Successful uptake:
learner repairs
erroneous
utterance or
incorporates
corrective
feedback into
his/her output
No uptake/topic
continuation
L2 learner notices or chooses
to ignore corrective feedback
Error correction
episodes
(negotiation
of form)
Comprehensible
input + negotiation
of meaning and
form
490 CALICO Journal, Vol. 22, No. 3
ing. In carefully designed communicative and problem-solving activities, meaning
and form can be negotiated through dyad work. It is possible for NSs and
competent NNSs to provide two types of negative or corrective feedback (direct
and indirect) to L2 learners as they engage in the negotiation of both meaning
and form. When NS or NNS partners react to L2 learners’ erroneous output, that
reaction brings the non-targetlike lexical feature or linguistic form to their attention.
Learners can then respond to corrective feedback by either ignoring these
modifications or successfully incorporating them into their output. The absence
of uptake or opportunities for uptake is also a possible outcome of NS-NNS and
NNS-NNS online negotiated interaction.
Limitations of the Current Study
Several limitations of the study described here should be noted. First, although 14
participants (seven dyads, including one NS-NS dyad for baseline data collection)
originally formed the basis of this pilot study, only 10 participants completed all
of the communicative and problem-solving activities. A carefully controlled study
of negotiated interaction between members of at least 30 dyads evenly divided
between NS-NNS and NNS-NNS pairs would generate more data on error correction
episodes and various features of corrective feedback in an instant-messaging
environment, preferably one making use of at least two or three modalities (e.g.,
text, speech, Web cams). Second, a more detailed analysis of learner errors would
have been helpful. Following the careful identification, coding, and tagging of
each type of error, a detailed analysis of grammatical errors could be undertaken
in other projects of this kind with text analysis software such as MonoConc or
WordSmith that would help clarify the type of error correction response provided
by NSs and NNSs and the duration or length of moves leading to successful learner
uptake. Third, an equal number of communicative and problem-solving activities
or tasks would have allowed the researcher to determine which type of learning
activity significantly accounted for most of the error correction episodes. An appropriate
balance of learning activities should be considered in future studies. Finally,
this study examined text-mediated interactions and a limited number of oral
exchanges between NSs-NNSs and NNSs-NNSs. Since the use of the visual and
aural components in YIM (as well as of any desktop videoconferencing software)
work best with high-speed Internet connections, the learners and some of their
NNS counterparts who used dial up connections experienced various technical
difficulties while completing the online learning activities. Technical difficulties
would need to be solved in larger-scale projects of a similar nature.
CONCLUSION AND IMPLICATIONS
This exploratory study has addressed questions concerning the availability of
error correction in computer-mediated communicative and problem-solving activities
between members of different types of dyads (NNS-NNS and NS-NNS).
When communicative and problem-solving activities are carefully orchestrated,
corrective feedback is readily available to learners. Although chi-square was used
Susana Sotillo 491
to investigate the association among several dependent variables in each of the
two dyad types, it did not indicate causality. It would be desirable if future investigations
of synchronous computer-mediated negotiated interaction would isolate
the performance of each dependent variable in different types of dyads. For this
purpose, a binary logistic regression analysis would need to be performed on a
larger database. The findings of this exploratory study also show that NNSs provide
more corrective feedback of an explicit nature to ESL learners than NSs.
One can only speculate as to the motives of NNSs in providing explicit corrective
feedback when, as teachers in training, they were aware that indirect negative
feedback in the form of recasts or negotiation moves such as clarification requests
has been shown to facilitated learner uptake. It is possible that NNSs who have
themselves undergone the process of language learning, which includes interlanguage
restructuring, feel compelled to overtly point out the erroneous features
in their partners’ output. The findings of this study also show that both NSs and
NNSs reacted primarily to learners’ grammatical and lexical errors. This is an
important finding since interaction among participants collaborating on various
learning activities was computer-mediated. The YIM chat logs provided instant
access to learner output data, and the use of this kind of tool has implications for
language learning. Readily available learner output data could be a useful resource
for planning lessons in communicative and content-based language classrooms.
L2 learners could also benefit from carefully examining chat logs or listening to
audio recorded exchanges.
Since most instances of corrective feedback moves occurred immediately following
learners’ erroneous output, successful learner uptake might have been
affected by this fact. The effect of length or duration of corrective moves on
successful learner uptake should be investigated in future studies of SCMC or
interaction in an instant-messaging environment between partners in NS-NNS
and NNS-NNS dyads. Successful learner uptake was evident in the computer-mediated
NS-NNS and NNS-NNS error correction episodes in the study. However,
as Gass et al. (1998) have pointed out, researchers need to be cautious concerning
categorical claims for the role of the context in which interaction takes place
since both input and interaction have been shown to facilitate but not cause L2
development. The findings of this exploratory study have implications for the design
of learning activities to be completed using instant messaging or SCMC that
facilitate language learning. As L2 learners collaborate in online communicative
and problem-solving activities with their NS or more advanced NNS partners,
focused corrective feedback could be made available that would eventually lead
to successful learner uptake. These constitute supplemental learning activities that
take place outside the traditional classroom environment. New synchronous computer-
mediated modalities such as instant messaging that encourage one-on-one
interaction between NSs or more experienced interactants and language learners
clearly have potential benefits for the language learning enterprise.
492 CALICO Journal, Vol. 22, No. 3
NOTES
1 During the trial period, chatting via YIM and using the Web Cams presented problems
because of system requirements for the use of audio and video components. For instance,
when one of the participants in the study wanted to chat with the researcher or use the Web
cam and YIM’s Talk feature, the system crashed in the middle of a conversation because she
was using a dial up connection while the researcher was using cable Internet access. The
participant was very frustrated at the beginning and decided to use only the text-mediated
or regular chat features of YIM. The Talk feature was unstable; she used it a few times but
did not save the data.
2 I am grateful to Allison Nazimek and Jessica Castillion for their assistance with the coding
and analysis of chat data.
3 A learning activity in this study is defined as an endeavor or undertaking that is not limited
to a specialized purpose such as commercially available jigsaw or decision-making tasks
commonly used by SLA researchers in the communicative language classroom or in quasiexperimental
settings.
4 The participants’ names cited in the article are pseudonyms.
5 Professor Longxing Wei provided expert assistance with the statistical tests used to analyze
the data in this project. Since chi-square tests only reveal significant differences in
frequency data and test the relationship between variables, no assumptions or suggestions
regarding causality are possible. A Yates correction factor was applied in order to compensate
for the discrepancy in cell frequency (expected less than 5) between the chi-squared
distribution and the observed chi-squared value.
REFERENCES
Abrams, Z. I. (2003). The effect of synchronous and asynchronous CMC on oral performance
in German. Modern Language Journal, 87 (2), 157-167.
Beauvois, M. H. (1992). Computer-assisted classroom discussion in the foreign language
classroom: Conversation in slow motion. Foreign Language Annals, 25 (5), 455-
464.
Blake, R. (2000). Computer mediated communication: A window on L2 Spanish interlanguage.
Language Learning & Technology, 4 (1), 120-136. Retrieved May 14,
2005 from http://llt.msu.edu/vol4num1/blake/default.html
Braidi, S. M. (2002). Reexamining the role of recasts in native-speaker/non-native-speaker
interactions. Language Learning, 52 (1), 1-42.
Chun, D. M. (1994). Using computer networking to facilitate the acquisition of interactive
competence. System, 22 (1), 17-31.
Darhower, M. (2002). Instructional features of synchronous computer-mediated communication
in the L2 class: A sociocultural case study. CALICO Journal, 19 (2),
249-277.
Davies, B., & Thiede, R. (2000). Writing into change: Style-shifting in asynchronous electronic
discourse. In M. Warschauer & R. Kern (Eds.), Network-based language
teaching: Concepts and practice (pp. 87-120). Cambridge: Cambridge University
Press.
Susana Sotillo 493
Doughty, C., & Williams, J. (1998). Focus on form in classroom second language acquisition.
Cambridge: Cambridge University Press.
Doughty, C., & Varela, E. (1998). Communicative focus on form. In C. Doughty & J.
Williams (Eds.), Focus on form in classroom second language acquisition (pp.
114-138). Cambridge: Cambridge University Press.
Ellis, R., Basturkmen, H., & Loewen, S. (2001). Learner uptake in communicative ESL
lessons. Language Learning, 51 (2), 281-318.
Fernández-García, M., & Martínez-Arbelais, A. (2002). Negotiation of meaning in nonnative
speaker-nonnative speaker synchronous discussions. CALICO Journal, 19
(2), 279-294.
Fotos, S. (1994). Integrating grammar instruction and communicative language use through
grammar consciousness-raising tasks. TESOL Quarterly, 28, 323-351.
Fotos, S., & Ellis, R. (1991). Communicating about grammar: A task-based approach. TESOL
Quarterly, 25, 605-628.
Gass, S. M., Mackey, A., & Pica, T. (1998). The role of input and interaction in second language
acquisition: Introduction to the special issue. Modern Language Journal,
82 (3), 299-307.
Gass, S. M., & Selinker, L. (2001). Second language acquisition (2nd ed.). Mahwah, NJ:
Lawrence Erlbaum Associates.
Gass. S. M., & Varonis, E. (1984). The effect of familiarity on the comprehension of nonnative
speakers. Studies in Second Language Acquisition, 7 (1), 37-57.
Gass, S. M., & Varonis, E. (1994). Input, interaction, and second language production.
Studies in Second Language Acquisition, 16 (3), 283-302.
Grimshaw, J. & Pinker, S. (1989). Positive and negative evidence in language acquisition.
Behavioral and Brain Science, 12 (2), 341-342.
Kelm, O. R. (1992). The use of synchronous computer networks in second language instruction:
A preliminary report. Foreign Language Annals, 25 (5), 441-454.
Kern, R. (1995). Restructuring classroom interaction with networked computers: Effects
on quantity and quality of language production. Modern Language Journal, 79
(4), 457-476.
Kern, R., Ware, P., & Warschauer, M. (2004). Crossing frontiers: New directions in online
pedagogy and research. Annual Review of Applied Linguistics, 24, 243-260.
Levy, M., & Kennedy, C. (2004). A task-cycling pedagogy using stimulated reflection and
audio-conferencing in foreign language learning. Language Learning & Technology,
8 (2), 50-68. Retrieved May 14, 2005, from http://llt.msu.edu/vol8num2/
pdf/levy.pdf
Loewen, S. (2004). Uptake in incidental focus on form in meaning-focused ESL lessons.
Language Learning, 54 (1), 153-188.
Long, M. (1981). Input, interaction, and second language acquisition. In H. Winitz (Ed.),
Native and foreign language acquisition (pp. 259-278). Annals of the New York
Academy of Sciences.
Long, M. (1985). Input and second language acquisition theory. In S. Gass & C. Madden
(Eds.), Input in second language acquisition (pp. 377-393). Rowley, MA: Newbury
House.
494 CALICO Journal, Vol. 22, No. 3
Long, M. (1996). The role of the linguistic environment in second language acquisition. In
W. Ritchie, & T. Bhatia (Eds.), Handbook of second language acquisition (pp.
413-468). New York: Academic Press.
Long, M., & 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.
Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake. Negotiation of
form in communicative classrooms. Studies in Second Language Acquisition,
19, 37-66.
Mackey, A. (1999). Input, interaction, and second language development: An empirical
study of question formation. Studies in Second Language Acquisition, 21, 557-
587.
Mackey, A., Gass, S., & McDonough, K. (2000). How do learners perceive interactional
feedback? Studies in Second Language Acquisition, 22, 471-497.
Mackey, A., Oliver, R., & Leeman, J. (2003). Interactional input and the incorporation
of feedback: An exploration of NS-NNS and NNS-NNS adult and child dyads.
Language Learning, 53 (1), 35-66.
Morris, F. (2005). Child-to-child interaction and corrective feedback in a computer-mediated
L2 class. Language Learning & Technology, 9 (1), 29-45. Retrieved May
14, 2005, from http://llt.msu.edu/vol9num1/morris/default.html
Nassaji, H., & Fotos, S. (2004). Current developments in research on the teaching of grammar.
Annual Review of Applied Linguistics, 24, 126-145.
Oliver, R. (1995). Negative feedback in child NS-NNS conversation. Studies in Second
Language Acquisition, 17 (4), 459-481.
Oskoz, A. (2004, June). Negotiation of meaning in jigsaw and free discussion in synchronous
computer-mediated communication (S-CMC). Paper presented at CALICO
2004, Carnegie Mellon University, Pittsburgh, PA.
Pica, T. (1988). Interlanguage adjustments as outcome of NS-NNS negotiated interaction.
Language Learning, 38 (1), 45-73
Pica, T. (1994). Research on negotiation: What does it reveal about second-language learning
conditions, processes, and outcomes? Language Learning, 44, 429-527.
Pica, T. (2003). Second language acquisition research and applied linguistics. Working
Papers in Educational Linguistics, 18 (2), 1-26.
Pica, T., Doughty, C., & Young, R. (1986). Making input comprehensible: Do interactional
modifications help? I.T.L. Review of Applied Linguistics, 72, 1-25.
Pica, T., Holliday, L., Lewis, N., & Morgenthaler, L. (1989). Comprehensible output as
an outcome of linguistic demands on the learner. Studies in Second Language
Acquisition, 11 (1), 63-90.
Pinker, S. (1989). Resolving a learnability paradox in the acquisition of the verb lexicon.
In M. L. Rice & R. L. Schiefelbusch (Eds.), The teachability of language (pp.
13-62). Baltimore, MD: P.H. Brookes.
Polio, C., & Gass, S. M. (1998). The role of interaction in native speaker comprehension of
non-native speaker speech. Modern Language Journal, 82 (3), 308-319.
Susana Sotillo 495
Schegloff, E. A., Jefferson. G., & Sacks, H. (1997). The preference for self-correction in
the organization of repair in conversation. Language, 53 (2), 361-382.
Schmidt, R., & Frota, S. (1986). Developing basic conversational ability in a second language:
A case study of an adult learner of Portuguese. In R. R. Day (Ed.), Talking
to learn: Conversation in second language acquisition (pp. 237-326). Rowley,
MA: Newbury House Press.
Schultz, J. M. (2000). Computers and collaborative writing in the foreign language curriculum.
In M. Warschauer & R. Kern (Eds.), Network-based language teaching:
Concepts and practice (pp. 121-150). Cambridge: Cambridge University Press.
Sheen, R. (1992). Problem solving brought to task. TELC Journal, 23, 44-59.
Sheen, R. (2002). ‘Focus on form’ and ‘focus on forms.’ ELT Journal, 56, 303-304.
Sheen, R. (2003). Focus on forms as a means of improving accurate oral production. Unpublished
manuscript.
Smith, B. (2003). Computer-mediated negotiated interaction: An expanded model. Modern
Language Journal, 87 (1), 38-54.
Smith, B. (2005). The relationship between negotiated interaction, learner uptake, and lexical
acquisition in task-based computer-mediated communication. TESOL Quarterly,
39, 33-58.
Sotillo, S. (2000). Discourse functions and syntactic complexity in synchronous and asynchronous
communication. Language Learning & Technology, 4 (1), 82-119. Retrieved
May 14, 2005 from http://llt.msu.edu/vol4num1/sotillo/default.html
Spada, N., & Lightbown, P. (1993). Instruction and the development of questions in L2
classrooms. Studies in Second Language Acquisition, 15 (2), 205-224.
Sullivan, J., & Caplan, N. A. (2003). Beyond the dictogloss. Leaner-generated attention to
form in a collaborative communicative classroom activity. Working Papers in
Educational Linguistics, 19 (1), 65-89.
Swain, M. (1985). Communicative competence: Some roles of comprehensible input and
comprehensible output in its development. In S. M. Gass & C. Madden (Eds.),
Input in second language acquisition (pp. 235-253). Rowley, MA: Newbury
House.
Swain, M. (1998). Focus on form through conscious reflection. In C. Doughty & J. Williams
(Eds.), Focus on form in classroom second language acquisition, (pp. 64-
82). Cambridge: Cambridge University Press.
Swain, M. (2000). The output hypothesis and beyond: Mediating acquisition through collaborative
dialogue. In J. P. Lantolf (Ed.), Sociocultural theory and second language
learning (pp. 97-114). Oxford: Oxford University Press.
VanPatten, W., & Cadierno, T. (1993). Explicit instruction and input processing. Studies in
Second Language Acquisition, 15 (2), 225-243.
Varonis, E., & Gass, S. M. (1985). Non-native/non-native conversations: A model for negotiation
of meaning. Applied Linguistics, 6 (1), 71-90.
Warschauer, M. (1996). Comparing face-to-face and electronic communication in the second
language classroom. CALICO Journal, 13 (2), 7-16.
Williams, J. (1999). Learner-generated attention to form. Language Learning, 49, 583-
625.
496 CALICO Journal, Vol. 22, No. 3
Williams, J., & Evans, J. (1998). Which kind of focus and on which forms? In C. Doughty
& J. Williams (Eds.), Focus on form in classroom second language acquisition
(pp. 139-155). Cambridge: Cambridge University Press.
AUTHOR’S BIODATA
Susana M. Sotillo is Associate Professor of Linguistics and current Department
Chair at Montclair State University. Her areas of interest include computer-mediated
communication and second language acquisition, critical discourse analysis,
and corpus linguistics. She has published book chapters on political discourse,
multicultural and bilingual education, and articles in Discourse & Society, Language
Learning & Technology, TESOL Journal, and The Technology Source.
AUTHOR’S ADDRESS
Susana M. Sotillo,
Associate Professor of Linguistics
Chair 2004-2005
Montclair State University
Montclair, NJ 07003
Phone: 973/655-7377
EMAIL: sotillos@mail.montclair.edu
http://chss.montclair.edu/~sotillos/susana.html