Australia • Brazil • Japan • Korea • Mexico • Singapore • Spain • United Kingdom • United States
Julie A. Belz
Steven L.Thorne
Editors
AAUSC Issues in Language
Program Direction
Internet-mediated Intercultural Foreign
Language Education
© 2006 Heinle, Cengage Learning
ALL RIGHTS RESERVED. No part of this work covered by the 
copyright herein may be reproduced, transmitted, stored or used 
in any form or by any means graphic, electronic, or mechanical, 
including but not limited to photocopying, recording, scanning, 
digitizing, taping, Web distribution, information networks, or 
information storage and retrieval systems, except as permitted 
under Section 107 or 108 of the 1976 United States Copyright Act, 
without the prior written permission of the publisher.
Library of Congress Control Number: 2005935615
ISBN-13: 978-1-4130-2992-5
ISBN-10: 1-4130-2992-2
Heinle
20 Channel Center Street
Boston, MA 02210
USA
Cengage Learning is a leading provider of customized learning 
solutions with offi  ce locations around the globe, including 
Singapore, the United Kingdom, Australia, Mexico, Brazil, and Japan. 
Locate your local offi  ce at www.cengage.com/global
Cengage Learning products are represented in Canada by 
Nelson  Education, Ltd.
To learn more about Heinle, visit www.cengage.com/heinle  
Purchase any of our products at your local college store or at our 
 preferred online store www.cengagebrain.com
AAUSC Issues in Language 
Program Direction: Internet-
mediated Intercultural Foreign 
Language Education
Julie A. Belz, Steven L.Thorne
Executive Editor: Carrie Brandon
Senior Development Editor: 
 Joan M. Flaherty
Assistant Editor: Arlinda Shtuni
Editorial Assistant: Morgen Murphy
Technology Project Manager: 
 Sacha Laustsen
Production Director: Elise Kaiser
Marketing Manager: 
 Lindsey Richardson
Marketing Assistant: 
 Marla Nasser
Advertising Project Manager: 
 Stacey Purviance
Manufacturing Manager: 
 Marcia Locke
Compositor: GEX Publishing 
 Services
Project Manager: GEX 
 Publishing Services
Printed in the United States of America  
  2 3 4 5 6 7 14 13 12 11 10
For product information and 
technology assistance, contact us at Cengage Learning 
Customer & Sales Support, 1-800-354-9706
For permission to use material from this text or product, 
 submit all requests online at www.cengage.com/permissions 
Further permissions questions can be emailed to
permissionrequest@cengage.com
147
Chapter 6
A Study of Native and Nonnative Speakers’
Feedback and Responses in Spanish-American
Networked Collaborative Interaction 
Lina Lee
Abstract
Networked collaborative interaction (NCI) promotes the negotiation for
meaning and form that plays a crucial role in the development of language
competence. This chapter reports and discusses a study that focused on the
examination of relationships among error type, feedback types, and
responses in synchronous communication between native teachers and non-
native speakers (N = 26) working on two tasks—an open-ended question
and a goal-oriented activity.The results revealed that differences were found
not in the various types of negotiation moves but in the proportional use of
particular moves.The native speakers had an overwhelming tendency to use
recasts to provide corrective feedback. This feedback also had the positive
effect of drawing learners’attention to form, which led to the repair of errors.
Successful uptake, however, does not guarantee second language acquisition.
In addition, lexical rather than syntactical errors were the main triggers for
negotiation moves generated by both groups of interlocutors. NCI as a form
of written visual communication facilitated the response to corrective feed-
back.Learners generated high rates of repairs for both lexical and syntactical
errors. Overall, this study demonstrates that NCI is a powerful communica-
tion tool for the enrichment of language learning.
Introduction
In recent years, many foreign language (L2) programs have moved their curricu-
lum toward a communicative-based approach that emphasizes the importance of
genuine human interaction in L2 development. Central to such interaction is
negotiation of meaning that elicits negative feedback using various types of nego-
tiation moves (e.g., recasts, confirmation checks, and clarification requests) to
facilitate comprehension (Long 1996). Crucial to the feedback is a focus-on-form
procedure that brings learners’ attention to linguistic problems for output modifi-
cation (Long 1996; Long and Robinson 1998). Modified output, therefore, pro-
motes learners’ language accuracy and fluency through increased control over L2
forms (Swain 1995). Feedback through negotiation of meaning and form plays a
beneficial role in language acquisition and should be encouraged in language
classrooms. Given the limited time that many students spend in the regular class-
room and the lack of opportunities that they have to use the target language to
148 LEE
interact with others outside the classroom, how can we find an effective means of
providing them with acquisition-rich interaction through out-of-class commu-
nicative activities? Schmidt (1995) claims that conscious “noticing” of linguistic
forms is necessary in order for learning to take place. Through the use of net-
worked collaborative interaction (NCI), how can we make feedback available to L2
classroom learners, engage them in attention to form, and encourage uptake? 
NCI has been increasingly implemented in L2 instruction to extend learners’
communicative experience beyond the spatial and temporal limitations of the
classroom. NCI takes place in real time during which learners are exposed to
input, feedback, and output in a way similar to that which they would experience
through face-to-face communication. NCI, however, is different from spoken dis-
course in its textual representation, which relies heavily on writing and readings
skills; thus, learners may need more time to process input and output (Abrams
2003). Learners may also pay attention to certain aspects of written discourse on
the screen that facilitate input comprehension (Warschauer 1997) or edit the ini-
tial response using self-corrections (Blake and Zyzik 2003). This activity, in turn,
may lead them to provide and attend to feedback more frequently than during
face-to-face oral exchanges. To date, the studies of the potential role of feedback
and its effects on L2 development have been conducted mainly in the oral mode
(e.g., Ellis, Basturkmen, and Loewen 2001; Iwashita 2003; Leeman 2003; Lyster
and Ranta 1997; Mackey 1999; Morris 2002; Oliver 2002). In order to gain a better
understanding of the impact of interactive negotiation and feedback in L2, studies
must be conducted in different instructional contexts including NCI using both
native (NS) and nonnative speakers (NNS). 
Only a limited number of research studies have examined the effect of computer-
mediated communication (CMC) on learners’ interaction (e.g., Blake 2000; Chun
1994; Darhower 2002; Lee 2001, 2002b; Pellettieri 2000; Smith 2003; Tudini
2003). Although these studies are valuable, to the best of my knowledge, no
research within the interactionist tradition has explored closely the relationship
among error types, feedback types, and uptake moves that learners generate dur-
ing NCI. The current study is the first NCI research to include NS secondary
teachers and NNS learners of Spanish from two U.S. institutions. The distinct
roles of the interlocutors in the exchange (e.g., teacher-like vs. learner-like) com-
pound the proficiency differential (e.g., proficient vs. less proficient) and these two
interlocutor features together rather than one alone may affect the preference for
type of feedback, the nature of the negotiation move, and the amount and types of
responses. This paper examines such NS-NNS interaction by addressing three
major questions: (1) What types of feedback1 do NSs and NNSs provide to each
other during NCI?; (2) What types of linguistic triggers on the part of the NNSs
lead to what types of corrective feedback by the NSs?; and (3) Upon receiving feed-
back, do NSs and NNSs respond to each other? What types of NNS linguistic prob-
lems lead to what types of immediate uptake moves? The analysis of these issues
provides language professionals with important pedagogical information for
technology-based teaching. Furthermore, language educators will gain insight
into a reactive approach to focus-on-form involving the provision and use of feed-
back during teacher-learner interaction. 
Interaction, Negotiation of Meaning, 
and L2 Development
The early work of Evelyn Hatch in the 1970s led L2 researchers to develop and
explore the relationship between interaction and second language acquisition
(SLA). Hatch (1978) claims that learners acquire an L2 through interacting with
others rather than learning grammatical structures in order to interact. For years,
SLA researchers have developed theoretical arguments in support of oral interac-
tion and its effects on the development of learners’ interlanguage (e.g., Gass,
Mackey, and Pica 1998; Long 1996; Pica 1994; Swain and Lapkin 1998). One of the
most compelling theories is the Interaction Hypothesis (IH). The IH suggests that
learners not only need to have the opportunity to receive comprehensible input
(Krashen 1985) but also to negotiate for meaning to solve communication prob-
lems and further achieve mutual understanding (Long 1996). Crucial to the input
is how it is made comprehensible through interactive negotiation. Long (1996)
points out: 
. . . negotiation of meaning and especially negotiation work that triggers 
interactional adjustments by the NS or more competent interlocutor,
facilitates acquisition because it connects input,internal learner capacities,
particularly selective attention, and output in productive ways (pp.
451–452).
Similarly, Pica (1994) stresses the potential role of negotiated interaction in
allowing learners to attend to both meaning and form as they engage in trying to
understand and be understood. Learners’ attention to problematic discourse can arise
either from their own production or from what is produced by their interlocutors; the
noticing of a mismatch may lead to linguistic modifications (Gass and Varonis 1994).
In contrast to semantically processed input, modified output resulting from meaning
negotiation fosters L2 development because it pushes learners to reflect consciously
on linguistic forms as they engage in syntactic processing (Swain 1995). 
Existing SLA research suggests that oral interactive negotiation improves
comprehension and L2 development (e.g., Gass and Varonis 1994; Mackey 1999;
Pica 1994), draws learners’ attention to notice differences between their L1 and L2
(e.g., Gass 1997; Long 1996; Long and Robinson 1998; Schmidt 1995), and pushes
learners to produce modified output (Swain and Lapkin 1998). Similar results
regarding CMC have been found by L2 researchers as well (e.g., Blake 2000; Kern
1995; Lee 2001, 2002b; Pellettieri 2000; Smith 2003; Tudini 2003). In their stud-
ies, Chun (1994) and Kern (1995) found that learners produce a wide range of dis-
course structures in CMC and that it further fostered their sociolinguistic and
interactive competence. A study conducted by Fernández-García and Martínez-
Arbelaiz (2003) compared the negotiation generated by three different groups—
NNS-NNS, NS-NS, and NS-NNS in the oral and CMC modes. The findings revealed
that the NS-NNS group negotiated in the oral mode significantly more than in the
CMC mode and learners tended to employ L1-based strategies to solve the com-
munication problems. The burden of negotiation of meaning through written
NS AND NNS FEEDBACK AND RESPONSES 149
150 LEE
discourse without non-verbal cues pushed learners to employ a wide range of
negotiation moves to solve communication breakdowns in learner-learner syn-
chronous online exchanges (Lee 2001, 2002b). Both Pellettieri’s (2000) and
Smith’s (2003) studies of negotiation routines in CMC yielded similar results.
Their findings showed that approximately one-third of all turns were negotiated by
the learners. More importantly, learners produced lexical and syntactic output
modifications in response to learners’ feedback during CMC. 
Feedback, Negotiation Moves, Uptake 
and Second Language Acquisition 
The potential role of feedback through negotiation of meaning has been explored
extensively by L2 researchers (e.g., Iwashita 2003; Long 1996; Long, Inagaki, and
Ortega 1998; Mackey 2000; Mackey, Gass, and McDonough 2000; Oliver 2000).
Corrective feedback supplied by teachers or NSs is a response to learners’
problematic utterances. It is often considered a focus-on-form procedure that
overtly draws learners’ attention to L2 forms (Long 1996; Long and Robinson
1998), a process hypothesized to be necessary for SLA (Mackey 2000; Schmidt
1995; Williams 1999). Although learners cannot provide corrective feedback to
NSs, they often negotiate meaning or form using negotiation moves (e.g., confir-
mation checks or clarification requests) to overcome comprehension difficulties
that arise due to their own insufficient L2 knowledge or underdeveloped interlan-
guage structures. NS corrective feedback on learners’ linguistic errors and NNS
negotiation of meaning to comprehend NS discourse, therefore, play a fundamen-
tally distinct role during NS-NNS interactions. 
Corrective feedback viewed as negative evidence provides direct or indirect
evidence to learners about what is not possible in the L2 (e.g., Long 1996; White
1990). Oliver labels this “negative feedback” to indicate “differences between the
target language and a learner’s output” (2000, p. 120). Explicit negative feedback
indicates clearly that what the learner has said is incorrect (e.g., “You should say
Y but not X.” “X is incorrect. The correct form is Y.”). It is unlikely that direct cor-
rection is made during naturalistic conversations because it is viewed as impolite.
In contrast to explicit feedback, implicit feedback contains various types of negoti-
ation moves2 (e.g., confirmation checks or requests for help) that are a reaction to
non-target forms. Recent studies have focused on the role of implicit negative
feedback and its effect on language learning (e.g., Ayoun 2001; Iwashita 2003;
Mackey 2000; Mackey, Oliver, and Leeman 2003; Morris 2002, 2005). The findings
of these studies in different contexts have shown that implicit feedback is often
available and used by L2 learners. Importantly, both explicit and implicit correc-
tive feedback fosters learners’ increased awareness of forms and pushes them to
produce modified output that may lead to the development of learners’ interlan-
guage (Swain and Lapkin 1995). 
One of the most common ways of providing corrective feedback is the use of
recasts, especially within teacher-learner and NS-NNS interactional contexts.
Recasts are frequently available for learners in language classrooms but less often
NS AND NNS FEEDBACK AND RESPONSES 151
through conversational interaction (Mackey and Philp 1998). Recasts are reactive
rather than preemptive because they are given in response to the learner’s initial
non-target-like utterance. Recasts can involve a partial or full reformulation of an
utterance (see Long 1996; Oliver 1995, for a comprehensive discussion of recasts).
Long (1996) argued that confirmation checks can function as recasts using addi-
tional cues, such as rising or falling intonation. In CMC, recasts and confirmation
checks are distinguished by the written question mark (e.g., “on the right” vs. “on
the right?”). The written discourse displayed on the screen and self-paced setting
of CMC may help learners to notice these two distinct interactional features and
further respond to them. 
Research studies of the role of recasts have shown conflicting findings, with
some studies suggesting that some recasts are available and beneficial to L2 learn-
ers (e.g., Doughty and Williams 1998; Iwashita 2003; Leeman 2003), whereas
others are not (e.g., Izumi 1998; Lyster 2002, 2004). In a study of university-level
beginning French classes, Doughty (1993) argued that NS teachers tended to use
recasts when there was only a single error. Oliver (1995) found that in interaction
involving NS-NNS children, recasts were more likely to occur after subject-verb
agreement errors. Izumi (1998) followed Oliver’s (1995) work on the provision of
corrective feedback in task-based conversation between NS-NNS pairs. Her find-
ings revealed that learners did not use negative feedback, including recasts, nor
did they respond to feedback frequently. In a similar vein, Lyster and Ranta (1997)
found that recasts were not as effective as other types of feedback, such as clarifi-
cation requests, repetitions, or elicitations. In their study of the use of question
forms using two groups of adult learners, Mackey and Philp (1998) found that
recasts proved to be beneficial for advanced learners and facilitative of learners’
short-term L2 development. They argue, however, that learners need to have ade-
quate language proficiency to process the implicit feedback and they also need to
be oriented toward form in order for uptake to occur. 
The above research shows considerable variability and suggests that learners
may or may not respond to corrective feedback from interlocutors. Uptake as the
learner’s response to corrective feedback includes the use of self-repair or repeti-
tion of the correct form provided by the interlocutor. Topic continuation and
acknowledgment of recasts are viewed as “no response” to corrective feedback
because learners did not make the attempt to correct linguistic errors (Mackey and
Philp 1998). Noticing the linguistic gap between L1 and L2 during the interaction
is essential for uptake to occur. It is difficult for learners to pay attention simulta-
neously to meaning and form due to limited information processing capacity
(VanPatten 1990) unless they “are able to draw on automatized knowledge of the
L2” (Ellis 1999, p. 8). In other words, learners need to have sufficient linguistic
skills to repair errors. Additionally, it has been argued that learners may not have
a chance to respond to corrective feedback or may not repair errors because they
do not necessarily notice corrective feedback as deliberate (e.g., Ellis, Basturkmen,
and Loewen 2001; Lyster 1998a, 1998b, 2004; Oliver 2000). It is claimed that the
learner’s response to corrective feedback is facilitative of acquisition (Ellis,
Basturkmen, and Loewen 2001). Other L2 researchers, however, argue that uptake
does not guarantee acquisition because learning may take place without uptake
(e.g., Mackey and Philp 1998; Morris 2002). 
152 LEE
Studies to date show inconsistent findings with respect to the role of uptake
in SLA. Some previous research suggests that learners’ uptake rate is generally
low (Izumi 1998; Lyster and Ranta 1997; Mackey and Philp 1998; Morris 2002;
Oliver 1995, 1998). In contrast, an ESL study conducted by Ellis, Basturkmen, and
Loewen (2001) found that uptake was largely successful (74.1%) and the majority
of instances involved the negotiation of form as opposed to meaning. They suggest
that uptake is an optional move and learners may not have a chance to or may
choose not to respond to feedback. They further point out that uptake is considered
successful when the learner demonstrates the understanding of a linguistic element
or uses it correctly. In a study of relationships among error types, feedback types,
and learner repairs, both Lyster (1998b) and Morris (2002) concluded that learners
tended to use negotiation moves for lexical errors and recasts for syntactic errors
during the interactive negotiation. Morris (2002) and Pica (1994) found that lexi-
cal items were negotiated and repaired more frequently than syntactic elements.
The relatively few existing studies on CMC in the interactionist tradition
including both NSs and NNSs (e.g., Blake and Zyzik 2003; Toyoda and Harrison
2002; Tudini 2003) showed that corrective feedback provided by the NSs triggered
negotiation and output modification. González-Lloret (2003) found that clarifica-
tion requests were the preferred way to provide feedback among NNSs. In their
recent study of heritage speakers and NNSs of Spanish in online interactions,
Blake and Zyzik (2003) revealed that lexical negotiations through the use of
recasts, clarification requests, and expansions had a positive effect on vocabulary
use. Learners did not seem to notice or pay attention to morphological errors. As a
result, the rate of grammatical negotiations was low (four out of thirty instances)
in comparison to the rate of lexical negotiations (twenty-four times). Because lex-
ical elements are meaning-oriented and have more communicative value than
syntactic forms (VanPatten 1996), it is logical that learners solve semantic prob-
lems before they attend to syntactic errors. These findings are consistent with
those reported in Tudini’s (2003) and Smith’s (2003) studies of synchronous CMC. 
In summary, research has shown that the role of feedback including recasts,
negotiation moves, and uptake in both oral and CMC contexts is controversial, mixed,
and inconclusive. The current study, motivated by the ongoing discussions and
debates on the potential impact of feedback on L2 learning, explores closely the rela-
tionship among error types, feedback types, and immediate responses involving NS
teachers and NNS learners of Spanish engaged in NCI. The main goal of this study is
to examine how both NSs and NNSs generate feedback and to what extent feedback is
used to solve both linguistic and comprehension problems during NCI. (Refer to
“Introduction” at the beginning of this chapter for specific research questions.) 
Method
Participants
The participants for this study were thirteen NS teachers and thirteen NNS learn-
ers of Spanish from two U.S. institutions. All participants completed a question-
naire concerning their personal, educational, and linguistic backgrounds. The
NS AND NNS FEEDBACK AND RESPONSES 153
NNSs were enrolled in a third-year advanced Spanish course at a northeast U.S.
university. They were required to use the Blackboard as part of their course
requirement. Most of these students had three years of Spanish in high school and
two or three semesters of Spanish in college prior to the study. They did not have
opportunities to use Spanish outside the classroom. The NSs were secondary
teachers of Spanish who were enrolled in the course “Integrating Technology into
Language Teaching” at an eastern seaboard university. One of the course require-
ment for the NS teachers was to interact with the NNS students selected for this
study using one of the Blackboard communication tools—“Virtual Classroom.”
Thirteen groups were paired up with one NS and one NNS in each group. The par-
ticipants’ profiles are summarized in Table 1. 
Table 1
Participant Profiles
Nonnative Speakers Native Speakers
(N = 13) (N = 13)
Gender Female = 12 Male = 1 Female = 13
Age 19 to 20 (Mean = 19.2) 28 to 45 (Mean = 34.7)
Nationality Americans Hispanics living in U.S.
Education Undergraduate students Graduate students
Language Proficiency Intermediate level3 Native fluency
Tasks/Procedure 
As has been described in the literature, tasks should be designed in ways that
ensure a primary focus on meaning but also allow for incidental attention to form
(Ellis 2003; Long 1985; Swain and Lakpin 1998). Crucially, task-based activities
should provide conditions that allow learners to maintain the balance of fluency,
accuracy, and complexity of language development (Skehan 1998).4 Two types of
two-way tasks were chosen for the current study: a goal-oriented activity and an
open-ended question. Closed tasks with one possible outcome, a type of goal-
oriented activity, have been suggested to promote more interaction because learn-
ers need to exchange information in order to complete the task (Long 1989;
Skehan 1998). Open-ended questions are viewed as two-way information exchange
activities that focus on meaning and turn-taking using a variety of negotiation
moves, such as confirmation checks and clarification requests (Duff 1999). 
Within this study, a goal-oriented activity called “spot-the-differences” was
used as the first two-way exchange. Participants were expected to work collabora-
tively to identify fifteen differences between two drawings of Santa Claus. The open-
ended question involved the role of advanced technology in society. The specific
sub-questions were as follows: (1) Do the new information technologies contribute
to the creation of the so-called global village? (2) In what ways do the new tech-
nologies influence your professional life, your personal life, and the lives of your
154 LEE
classmates and family members? and (3) Do the new information technologies
allow the exchange of ideas in a more democratic way? Give examples for and/or
against. 
NCI was carried out in Spanish using the “Virtual Classroom,” one of the com-
munication tools supported by Blackboard.5 Twenty-six participants formed
thirteen NS-NNS pairs for two 50-minute chat sessions. Both sessions took place
at mid-semester after participants were familiar and comfortable with Blackboard.
No particular instructions were given to the two groups. However, they were
encouraged to focus on the topic at hand. The NS and NNS groups did not have the
opportunity to interact with one other prior to the study and participants were not
given time for pre-task planning. Discussions were automatically saved in
Blackboard’s archives and were retrieved later for data analysis. 
Data Analysis 
Transcripts were coded for error types, feedback types, and modification devices
using Varonis and Gass’ (1985) coding scheme. The researcher and a trained
instructor coded the data to ensure consistency. Each coder first worked individu-
ally and then both coders compared and discussed the discrepancies until they
reached the final agreement.6 Example (1), taken from the current study,7 illus-
trates the scheme during lexical negotiation: 
(1)
1. NS: Es la próxima etapa, no? [trigger]
(It is the next stage. no?)
2. NNS: Que es “etapa”? [indicator]
(What is “etapa”?)
3. NS:“stage” [response]
(“stage”)
4. NNS: O.K. Gracias! [reaction]
(O.K.Thanks!)
In the above example, the unfamiliar word etapa “stage” serves as the trigger
for negotiation. Upon encountering communication difficulty, the NNS expressed
the need for clarification of the unknown word by using the Spanish for “what is”
(line 2). The NS partner then made an attempt to respond to the incomprehensible
message from the NNS using L1 translation (line 3). Finally, the NNS brought
closure to the negotiation by acknowledging the response (line 4). In this case, the
question “What is x?” was coded as a “clarification request,” one of the negotiation
moves used to elicit feedback. The answer to this request was coded as a
“response” to the clarification request. 
NS AND NNS FEEDBACK AND RESPONSES 155
Two types of linguistic errors were coded (see Table 2). Inter-rater reliability
was high (r = .92). 
Table 2
Types, Definitions, and Examples of Error
Error Type Definition Example
1 Lexical Incorrect, inappropriate choices of —“abrazo” for “brazo”
lexical items including non-target (“hug” for “arm”) or
derivations of nouns, verbs and —“realizar” for “darse
adjectives cuenta de” (“carry out”
for “to realize”)
2 Syntactic Incorrect usage of articles, verb —“la sistema” instead of 
section, pronouns, prepositions, “el sistema”
gender, subject, verb and adjective —“fue nervioso.” (He
agreement, wrong verb conjugations was nervous.) instead
of “estaba nervioso.”
(He was nervous.)
Table 3 illustrates types of feedback with their definitions and examples.
Inter-rater reliability for coding feedback types was high (r = .95).
Table 3
Types, Definitions and Examples of Feedback
Feedback
Type Definition Example
1 Explicit To directly correct non-target form —El subjuntivo de 
to target form “sufrir”es “sufra”no
“sufre.”
(The subjunctive of
“sufrir” is “sufra”not
“sufre.”)
2 Implicit To provide a clue to non-target form —Quieres decir “cuento”
or incomprehensible message using no “cuenta?
modification devices or recasts. (Do you mean “cuento”
[story] not “cuenta”
[bill]?) 
Five types of modification devices were identified and coded for providing
implicit feedback. Table 4 shows each type and a definition of the type of modification
device along with an example. Inter-rater reliability for coding was high (r = .90). 
156 LEE
Table 4
Categories, Definitions and Examples of Modification Devices
as Implicit Feedback
Type of 
Modification
Device Definition Example
1 Confirmation To repeat parts of the statement to —La gente no tiene 
check ensure understanding derecha.
(People don’t have 
right hand side?)
—Derecha? 
(Right hand side?)
2 Clarification To express confusion or ask for help —Los precios han 
check8 due to unfamiliar words or  aumentado mucho.
incomprehensibility (Prices have increased 
a lot.)
—Qué es “aumentado”?
(What is “aumentado”?)
3 Recast To reformulate all or part  of  —No lo sabieron.
non-target form (They did not know it.)
—No lo sabían.
(They did not know it.)
4 Request for  To ask for help with non-understood —Hmm . . . No entiendo.
help message9 Puedes explicar mas? 
(Hmm . . . I don’t
understand. Can you
explain more?
5 Use of To signal uncertainty or confusion —Las piernas son 
keyboard of a lexical item or an idea pequenas.
symbol (The legs are small.)
—???? or :<
The information in Table 5 describes responses to feedback. The researcher
coded responses by either “response” or “no response” to feedback. Any learner’s
response to feedback was considered uptake, but this may not necessarily involve
using the correct form (see Ellis, Basturkmen, and Loewen 2001 for a review). It
should also be noted that “response” to the recast refers to “repeat the recast”
and “no response” is “continue on task” or “acknowledge the recast.” (For more
discussion, see Mackey and Philp 1998.) Inter-rater reliability for coding was high
(r = .94). Within the category of “response,” five types10 were identified and coded
(see Table 5), with high inter-rater reliability (r = .90). 
NS AND NNS FEEDBACK AND RESPONSES 157
Table 5
Categories, Definitions and Examples of Responses to Feedback 
Type of
Responses Definition Example
1 Repair11 To correct errors made on lexical —NNS: No pago el 
items or grammatical structure cuento.
(She did not pay the
story.)
—NS: El cuento?
(The story?)
—NNS: Es la cuenta. Lo
siento.
(It is the bill. I’m sorry.)
2 Elaboration To elaborate the original idea or —NNS: Los picos? Que 
concept in Spanish son?
(The points? What are
they?)
—NS: Son los puntos en
el sombrero. Es como
el punto del lápiz.
(They are points on
the hat. It is like the
point of the pencil.)
3 Use of L1 To use English to substitute words —NNS: No entiendo la 
or ideas in Spanish palabra “chistoso”
(I don’t understand the
word “chistoso”)
—NS:“chistoso”es funny.
(“chistoso” is funny.)
4 Repetition To repeat and confirm the —NNS: Las mujeres no 
reformulated target form tienen mucho libre.
(Women don’t have
much free.)
—NS: Quieres decir la
libertad?
(Do you mean the
freedom?)
—NNS: Si, si. la libertad.
(Yes, yes the freedom.)
5 Use of  To signal agreement, certainty or —NS: Quieres decir la 
keyboard uncertainty on the request pantalla no la ventana?
symbol —NNS: si :-)
158 LEE
Results and Discussion
Research Question 1
The first research question addresses the types of feedback both NSs and NNSs
provided for negotiation of meaning and form during NCI. The results shown in
Table 6 support those reported in previous synchronous CMC studies (Blake and
Zyzik 2003; Fernández-García and Arbelaiz 2002; Lee 2002b; Tudini 2003) reveal-
ing that both explicit and implicit feedback were provided by the NSs and NNSs. 
Table 6
Number and Percentage of Feedback Type Used by NSs and NNSs
Feedback Type NSs (N = 13) NNSs (N = 13)
Explicit 2 (2.5%) 0 (0%)
Implicit 79 (97.5%) 53 (100%)
Total 81 (100%) 53 (100%)
However, there was a difference in the degree to which each type of feedback
was employed by the NSs and NNSs. Both NSs and NNSs gave very little or no
explicit feedback. It is logical to assume that they would not need to provide any
explicit feedback to their partners since the latter are NSs who would not make
errors in Spanish. One NS made the attempt to explicitly correct her partner’s
errors twice; once for a lexical error involving realizar “to carry out” instead of
darse cuenta “to realize” and once for a morpho-syntactic error involving the pres-
ent subjunctive form of sufrir “to suffer.” Within a social context, it is unlikely that
a partner would directly correct the other’s mistakes, especially when they are not
well acquainted with each other. Therefore, the majority of interactional moves
involved implicit feedback.
The NSs used implicit feedback 79 times while the NNSs used implicit feed-
back 23 times. Recasts resulted in the highest rate of corrective feedback (53.1%)
initiated by the NSs, whereas the clarification check (49%) was the most frequent
modification device used by the NNSs. Both groups used other types of modifica-
tion devices as well (see Table 7). 
Table 7
Number and Percentage of Modification Devices Used as Implicit Feedback 
Type of Modification Device NSs (N = 13) NNSs (N = 13)
Confirmation check 11 (13.9%) 9 (17%)
Clarification check 12 (15.3%) 26 (49%)
Recast 42 (53.1%) 2 (3.8%)
Request for help 10 (12.6%) 8 (15.1%)
Use of keyboard symbol 4 (5.1%) 8 (15.1%)
Total 79 (100%) 53 (100%)
Similar to the results reported in González-Lloret (2003), the NNSs tended to
use clarification checks to provide implicit feedback. The majority of the feedback
initiated by the NNSs was triggered by unknown lexical items (50 times at 94.3%).
In many cases, when the input was incomprehensible to the NNSs, they asked for
clarification, such as “What is x?” “What do you mean by y?” or “I don’t understand
what you said.” The results are consistent with Lee (2002b) who found that clarifi-
cation checks prompt the NS to provide additional information to make the input
understandable as shown in example 2:
(2)
1. NS: La computadora es azul y la pantalla amarilla.
(Is the computer blue and the screen yellow?)
2. NNS: que es la pantalla? [Clarification check]
(what is pantalla?)
3. NS: Pantalla es el vidrio de la computadora. Es la ventana.
(Screen is the glass of the computer. It is the window.)
4. NNS: screen? (screen?) [Use of L1]
5. NS: correcto. (correct.)
6. NNS: La pantalla mia es roja.
(My screen is red.)
The unknown lexical item pantalla (line 2) was not understood by the NNS.
The NNS then negotiated the meaning of the unknown lexical item using a clarifi-
cation check. In the response to the NNS, the NS explained what pantalla is in
Spanish (line 3) and the NNS responded to the feedback by using the L1 to confirm
the meaning of the new word (line 4). The evidence from the current study also
suggests that, similar to the claim made by Lyster (1998a), the clarification check
is effective as it requires the partner to make input adjustments and more impor-
tantly, provides the learner with elaborated and modified input in the L2. In this
study, the clarification check appears to have a positive short-term effect on com-
prehension and acquisition of vocabulary as the NNSs often incorporated new
words into later interaction (see examples 2, 3, 5, 6). Although these findings cor-
roborate those found in Blake’s (2000) and Smith’s (2003) studies of learners’
negotiation in CMC, further research is needed in order to confirm the long-term
effect of clarification checks on the acquisition of L2 vocabulary. 
Unlike face-to-face interaction, in text-based NCI learners cannot use non-
verbal cues such as gestures or facial expressions. However, emoticons and
repeated question marks are used as nonverbal negotiation devices to facilitate the
negotiation of meaning. For instance, the NNSs used the sad face “” to show the
incomprehensibility, confusion, or dislike of an idea and multiple questions marks
“???” to signal uncertainly or to ask for help from their NS partners. In this sense,
NS AND NNS FEEDBACK AND RESPONSES 159
160 LEE
emoticons function as clarification checks, requests for help, or confirmation
checks. As can be seen in example 3, the use of keyboard symbols clarified the
meaning of the word involucrada (involved, in line 1), changing what initially
appeared to be incomprehensible into comprehensible input.
(3)
1. NS: Estas involucrada? (Are you involved?)
2. NNS: ??? [Keyboard symbol]
3. NS: Es como ayudar a la comunidad o participar en algunas actividades.
(It is like to help the community or to participate in some activities.) 
4. NNS: Ahora entiendo. No, no estoy involucrada porque no tengo
tiempo.
(Now I understand.No,I’m not involved because I don’t have time.)
In this example, multiple question marks functioned as a clarification check
that asked the NS to provide input modification. It is evident that this particular
negotiation improved mutual comprehension. However, it remains difficult to
ascertain whether emoticons and creative uses of punctuation marks make for a
more salient trigger of input modifications than text-based clarification requests
as the proportional use of such keyboard symbols is small (15.1%). It is possible
that the keyboard symbol was the quick and easy way to provide feedback as the
learners were engaged in synchronous exchanges. Lee (2001), in her study of
online interaction among the NNSs, has also confirmed that learners used sym-
bols to compensate for the lack of face-to-face interaction and to clarify meaning.
The use of keyboard symbols may be affected by learners’ preference and style of
negotiation strategies. 
To summarize, lexical items rather than syntactical elements were the main
trigger for negotiation moves generated by both NSs and NNSs. Almost half of
negotiation moves following unknown lexical items involved clarification checks
(49%) initiated by the NNSs. In contrast, a great majority of feedback moves pro-
vided by the NSs were triggered by learners’ linguistic errors (both lexical and syn-
tactic items). The use of recast (53.1%) was the most salient modification device as
corrective feedback provided by the NSs to respond to learners’ non-target-like
forms. The following discussion seeks to find out what types of NNSs’ linguistic
errors lead to what types of implicit feedback provided by the NSs. 
Research Question 2
The analysis yielded a total of 97 linguistic errors made by the NNSs, 66% were
syntactic errors and 34% were lexical errors (see Table 8).
Each error was made by a NNS student, which may or may not result in the pro-
vision of feedback by a NS teacher. Table 9 shows that 71 out of 97 linguistic errors
received implicit feedback from the NSs. The findings reveal that in some cases there
was no need for negotiation when errors did not cause communication problems. 
Table 8
Number and Percentage of Error Type 
Lexical 33 (34%)
Syntactic 64 (66%)
Total 97 (100%)
Table 9
Number and Percentage of Implicit Feedback per Error Type 
Lexical 29 (40.8%)
Syntactic 42 (59.2%)
Total 71 (100%)
It is important to note that there was a difference found in the distribution of
feedback provided by the NSs shown in Table 7 (79 times) and in Table 9 (71 times).
Of the 79 feedback moves, 8 were used for unclear messages which focused on the
meaning (content) rather than the form (linguistic error). In other words, some-
times the NNS’s utterance was grammatically correct but the NS did not under-
stand the intended meaning. Requests for help, confirmation checks, and keyboard
symbols, therefore, were used for meaning clarifications rather than linguistic
modifications.
Table 10 shows that 73.2% of NNS linguistic errors received implicit feedback
from their NS teachers during the NCI. In contrast to previous findings that showed
that NSs were disinclined to point out learners’ linguistic errors (Iwashita 2003),
both lexical (87.9%) and syntactic errors (65.6%) received a high rate of feedback
moves in the present study. It is possible that as teachers themselves, the NSs may
have less tolerance for the NNSs’ non-target-like utterances. Additionally, from the
pedagogical point of view, grammatical accuracy and lexical growth should be
equally important for developing learners’ L2 language competence. It may be diffi-
cult for the NS teachers to ignore errors and they may have felt obligated to implic-
itly correct learners’ imperfect utterances. It is also possible that the text-based
communication via a computer screen would draw interlocutors’ attention to focus
on form more easily than through the oral modality. The NSs, therefore, generated a
high frequency of corrective feedback to push learners to modify output. 
Table 10
Rate of Implicit Feedback per Error Type 
Lexical 29/33 (87.9%)
Syntactic 42/64 (65.6%)
Total 71/97 (73.2%)
NS AND NNS FEEDBACK AND RESPONSES 161
162 LEE
A comparison of the distribution of the feedback types provided by the NSs
across the error types produced by the NNSs appears in Table 11. 
Table 11
Number and Percentage of Errors Receiving Implicit Feedback across
Feedback Types and Error Types
Type of Modification Device Lexical Syntactic Total
Confirmation Check 6 (20.7%) 3 (7.1%) 9 (12.7%)
Clarification Check 11 (38%) 1 (2.4%) 12 (16.8%)
Recast 6 (20.7%) 36 (85.7%) 42 (59.2%)
Request for help 5 (17.2%) 1 (2.4%) 6 (8.5%)
Use of keyboard symbol 1 (3.4%) 1 (2.4%) 2 (2.8%)
Total 29 (100%) 42 (100%) 71 (100%)
The most striking type of implicit feedback was the recast (59.2%) used by the
NSs. During the negotiation routines, the NSs showed a strong tendency to use
recasts to provide corrective feedback about non-target-like forms (see example 4).
This study supports the findings by Lyster and Ranta (1997) that teachers have an
overwhelming tendency to use recasts. By resorting to the use of recasts, the NS
avoided embarrassing the learner. Directness of explicit feedback may result in a
partner who is too demoralized to participate in discussions. 
As mentioned, Doughty (1993) argues that the NS teachers in her study tend
to use recasts when there is only one error. In this study, the recast in response to
a single error appeared in several occasions, such as Mi computadora [*esta]
grande y hay dos lapices en la mesa “My computer is big and there are two pencils
on the table” and Tengo ropa [*amarillo] y zapatos negros “I have yellow clothes
and black shoes.”12 The NS immediately reformulated the wrong choice of the
verb está to es and the wrong gender of the adjective amarillo to amarilla. In these
cases, the recast was provided as an attempt to highlight the error to draw the
learner’s attention to focus on form. The learner was able to incorporate the
correct form into the follow-up turn, such as Tienes ropa amarilla? “Do you have
yellow clothes?” The evidence appears to support the claim made by Lyster (2004)
that learners benefit from less salient recasts which have been shortened to high-
light one particular error. By contrast, the NS did not use full or partial recasts
when multiple errors were present in sentences, such as Mi abuelo murió [*cinco
años pasados] pero mi abuela todavía [*es] viva “My grandfather died five years
ago but my grandmother is still alive” or Es posible que [*muchos tiempos] los
estudiantes no [*pueden] usar la computadora en [*algunas] lugares “It is possi-
ble that many times students cannot use the computer in some places.” In these
cases, the NSs continued the discussion as the errors did not cause comprehen-
sion problems and they may have felt reluctant to interrupt the flow of communi-
cation. Pedagogically, students need to write correctly to maintain a balance
between function, content, and accuracy. One of the corrective techniques is to
make students re-examine and revise the exchanges with guided instruction. 
In addition, Table 11 shows that the majority of implicit feedback moves fol-
lowing syntactic errors were recasts (85.7%), whereas 79.3% of lexical errors
incited other types of modification devices, in particular, the use of clarification
checks (38%) and confirmation checks (20.7%). Lyster (1998b) and Morris (2002)
also found that syntactic errors favored recasts and lexical problems invited nego-
tiation moves. Example 4 illustrates how a NS utilized the recast to correct a
syntactic error made by an NNS: 
(4)
1. NS: La pantalla en mi dibujo es amarilla. (The screen of my picture is
yellow.)
2. NNS: ¿Qué color son los zapatos? (What color are the shoes?)
3. NS: ¿De qué colores son los zapatos? (What color are the shoes?) 
4. NNS: Si. Los mios son rojos. (Yes. Mine are red.)
5. NS: Ahora el gorro. El mío tiene tres colores, rojo, verde y amarillo.
(Now the cap. Mine has three colors, red, green and yellow.)
6. NNS: ¿Es gorro como sombrero? (Is “gorro” like hat?) 
7. NS: Correcto. (Correct.) 
8. NNS: Tengo gorro azul y blanco. ¿De qué color es la bufanda?
(I have blue and black cap.What color is the scarf?)
In this particular case, the recast drew the learner’s attention to the form.
After reading the message corrected by the partner, the NNS noticed the mismatch
between input and output and repeated the same structure in line 3 (¿De qué
color? = “what color?”) several turns later (line 8). The recast is more salient in
CMC than in face-to-face interaction as the learner reads the correct written text
on the screen. Moreover, written discourse can easily be retrieved by the use of the
vertical scroll bar. While this study makes no claim on the positive effect of recasts
for L2 development, the data seems to suggest that the recast did reorient the
learner’s attention toward the accurate form. 
Although only 20.7% of lexical errors received recasts as implicit feedback
(see Table 11), the majority were triggered by the misuse of the definite article or
gender as is illustrated in example 5: 
(5)
1. NNS: Los dos tienen las vestidas verdes. (Two have green dresses.)
2. NS: Llevan vestidos verdes. (They wear green dresses.) 
3. NNS: si, los vestidos son verdes ☺!! (Yes, the dresses are green ☺!!)
NS AND NNS FEEDBACK AND RESPONSES 163
164 LEE
In this case, the NNS did not know the word “dress” in Spanish is masculine.
Instead, she wrote both the article la and the noun vesitda in the feminine form
(line 1). To avoid confusion, the NS teacher immediately used the recast to high-
light the syntactic problems that caused the lexical error13 (line 2). As learners at
the intermediate proficiency level have great need for interlanguage restructuring,
it is crucial for them to receive models of the target language that may draw their
attention to target-like-forms and further that encourage them to use correct
forms (as demonstrated in example 5). Lyster (1994) points out that it is likely that
learners benefit from focus on form to overcome incorrect target language fea-
tures. The long-term effect of recasts on the developmental levels of the learner is
a worthwhile issue for further research. 
This study found evidence that lexical errors tended to invite the use of modi-
fication devices (79.3%) rather than recasts. Example 6 illustrates how the NS
negotiated inappropriate lexical items using both clarification checks (line 2) and
confirmation checks (line 6). 
(6)
1. NNS: Son hombres pequenos. (They are small men.)
2. NS: ¿Quienes son hombres pequenos? [Clarification request]
(Who are the small men?)
3. NNS: Son las personas que ayudan Santa Claus.Personas pequenas, si?
(They are people who help Santa Claus.Small people,yes?)
4. NS: Oh si entiendo.Son duendes.(Oh yes I understand.They are elves.)
5. NNS: Mis duendes no tienen piernas pequenas. (My elves don´t have 
small legs.)
6. NS: No son estrechas? [Confirmation check]
(They are not skinny?)
7. NNS: si, tienen piernas estrechas. (yes, they have skinny legs.)
The example above demonstrates that approximation and circumlocution are
typical communication strategies that the NNSs used for negotiation of meaning
due to the lack of vocabulary knowledge, as Lee (2001) reported in a previous
study. The NNS used approximations twice to substitute the correct and appropri-
ate Spanish words, such as hombres pequeños “small men” for duendes “elves,”
and piernas pequeñas “small legs” for piernas estrechas “skinny legs.” As a result
of interactive negotiations, not only did both interlocutors achieve mutual com-
prehension, but they reinforced the appropriate usage of L2 lexical items.
In conclusion, in spite of the distinctly different language proficiency levels of
the two groups (NSs vs. NNSs), clarification and confirmation checks appeared to
be the two most frequent modification devices used for lexical negotiations.
Recasts were primarily triggered by morpho-syntactic errors made by the NNSs.
The question remains as to whether these feedback moves resulted in responses. If
so, what types of NNSs’ linguistic problems led to what types of immediate uptake
moves? The following discussion attempts to answer these questions. 
Research Question 3
As can be seen from Table 12, the NSs responded to more than 90% of their part-
ners’ feedback. The NSs were foreign language teachers and hence accustomed to
dealing with NNS discourse. Therefore, the tendency to respond to learners’ feed-
back was likely more probable when confusion arose. In contrast, the NNSs only
responded to 58.2% of the NS teachers’ feedback. This aligns with the findings
reported by Lyster and Ranta (1997) that 55% of teachers’ feedback moves resulted
in uptake moves.
Table 12
Number and Percentage of Responses to Implicit Feedback 
Implicit Feedback NSs (N = 13) NNSs (N = 13) 
Responses 48 (90.6%) 46 (58.2%)
No responses 5 (9.4%) 33 (41.8%)
Total  53 (100%) 79 (100%)
A high percentage of implicit feedback provided by the NSs (41.8%) was not
responded to by the NNSs. The following example demonstrates how a NNS did
not seem to know how to respond to the NS’s request for help (line 5) regarding a
problematic message (line 4).
(7) 
1. NS: Que sabes de la famosa aldea global? Tu crees que la tecnologia
esta contribuyendo a la creacion de esta aldea global? 
(What do you know about the famous global village? Do you believe
technology is contributing to the creation of this global village?)
2. NNS: que es la aldea global? es una idea del mundo a ser uno? 
(what is the global village? Is it an idea of the world to be one?)
3. NS: Bueno me imagino que por medio de la tecnologia de la infor-
macion podemos estar en contacto con todo el mundo y de esa
forma volvernos una aldea global. Si, creo que la idea principal es
que el mundo este unido por medio de la red. Me comprendes?
(Well I can image that through information technology we can be in
contact with everyone and by this way we become a global village.
Yes, I believe the principal idea is that the world is united by the
web. Do you understand me?)
NS AND NNS FEEDBACK AND RESPONSES 165
4. NNS: No se. Si, pienso que la nueva technologia esta ayudando este
causa pero no me gusta todo el mundo usa mucho computadora.No
es personal y no conoces bien la gente.
(I don’t know.Yes, I believe the new technology is helping this cause
but I don’t like everyone to use a lot computer. It is not personal and
you don’t know people well.)
5. NS: Hmm… ¿Me puedes explicar un poco más? [Request for help]
(Hmm… Can you explain a little bit more to me?)
6. NNS: Tienes un Palm Pilot? No puedo hacer nada sin eso.
(Do you have a Palm Pilot? I cannot do anything without that.)
Multiple errors found in line 4 prevented the NS from understanding the NNS’s
intended message. The NS provided feedback using a “request for help” (line 5) that
challenged the student to elaborate and articulate her thoughts, a step crucial to
building learners’ language proficiency at the advanced level (Lee 2002a). The
learner, however, failed to respond to the NS’s feedback. Rather, the student initiated
a new topic (line 6). 
Table 13 presents the distribution of responses used by the NSs and the NNSs
to respond to their partners’ feedback. 
Table 13
Number and Percentage of Responses to Feedback
Type of Responses NSs (N = 13) NNSs (N = 13)
Repair 0 (0%) 20 (43.5%)
Elaboration 33 (68.7%) 6 (13.1%)
Use of L1 10 (20.8%) 10 (21.7%)
Repetition  3 (6.3 %) 8 (17.4%)
Use of keyboard symbol 2 (4.2%) 2 (4.3%)
Total  48 (100%) 46 (100%)
Two salient responses shown in Table 13 were NS elaborations (68.7%)
and NNS repairs (43.5%). Interestingly, L1 was used in similar proportion by the
NSs (20.8%) and NNSs (21.7%) respectively, whereas the lowest rate of response
was in the use of keyboard symbols by both groups (4.2% by the NSs and 4.3% by
the NNSs). 
As mentioned previously, 94.3% of feedback moves initiated by NNSs were
triggered by unknown lexical items. Upon receiving the clarification requests,
such as “I don’t understand. What is . . .?” from the NNSs, the NSs tended to elab-
orate the new word to explain the meaning. Of the 33 instances of elaboration, the
majority were used for clarifying the meaning of the word. Elaboration often
166 LEE
occurred when the NNS made the clarification request, such as “I don’t under-
stand. What is . . .?” In most cases, elaboration was provided to clarify the meaning
of a particular lexical item as shown in example 8:
(8)
1. NS: Si. En el pico del zapato o en el borde? 
(Yes.At the tip of the shoe or at the edge?)
2. NNS: que es un pico en tus palabras? [Clarification check]
(what is un pico in your words?) 
3. NS: La punta (The tip) 
4. NS: Perdón, el pico es la punta. El zapato termina en pico ( > ) y el
borde del sombrero y los zapatos también terminan en pico ( > ).Me
comprendes? [Elaboration]
(Sorry, the pico is a tip.The shoe ends in a point ( > ) and the edge of
hats and the shoes also end in a point ( > ). Do you understand me?) 
5. NNS: Si, ahora comprendo. En la punta de su zapato.
(Yes, Now I understand.At the tip of his shoe.) 
The illustration above shows that the NNS struggled to understand the
unknown word pico (tip in line 2). The NS twice elaborated on the meaning of
pico, and also produced a keyboard marker, to enhance comprehension (line 4). It
is logical to assume that NSs usually would not use Spanish to clarify input know-
ing their interlocutors were at the intermediate level of proficiency. In addition,
the use of Spanish language elaboration demands both learners’ linguistic and
cognitive skills to process modified input. In some cases, the negotiation occurred
over several turns in an attempt to achieve mutual comprehension. 
With regard to the use of English, the findings show that use of English was
the second preferred response move used by the NSs when Spanish elaboration as
the first attempt to try to solve a lexical and grammatical problem was unsuccess-
ful. Of the ten uses of English, five of them were used for grammar explanations. It
is possible that learners’ linguistic limitations forced the NSs to use English to
clarify unclear messages. For instance, as shown in example 9, after several nego-
tiated turns (line, 3, 4, 5, 6 and 7), the NS finally gave up and used L1 (line 8) to
explain the use of the aspectual problem, which the NNS understood and accepted: 
(9)
1. NNS: Antes muchas personas no sabieron usar la computadora.
(Before many people did not know how to use the computer.)
2. NS: Si, como mis abuelos no sabían nada de la computadora.
(Yes, like my grandparents did not know anything about the
computer.)
NS AND NNS FEEDBACK AND RESPONSES 167
3. NNS: Por que es “sabian” no “sabieron?” (Why is it “sabian” not
“sabieron”?)
4. NS: Es la descripción del pasado. (It is the description of the past.)
5. NNS: Pero ya termino la accion. Sabieron si? 
(But the action is complete. Sabieron yes?)
6. NS: “Supieron”quiere decir “enterarse de algo”
(“Supieron”means “enterarse de algo”)
7. NNS: Que es enterarse? No se la palabra.
(What is enterarse? I don’t know the word.)
8. NS: to find out or to discover.So in this case,you wanted to say “they
did not know” not “they did not find out” Me comprendes? (Do you
understand me?)
9. NNS: Si, no sabian. (Yes, they did not know.)
One possible explanation for resorting to English may be because it is fairly
difficult to explain an advanced grammatical concept in the target language. In
order to avoid confusion, English seems to be a good option after several turns of
negotiation. In certain circumstances, English may be necessary to support the
comprehension of Spanish. Until learners gain more knowledge and understand-
ing of Spanish, English is a good choice to solve linguistic or communication
problems and to keep the flow of the conversation going. 
Regarding the NNSs’ uptake, Table 14 displays the distribution of error types
across uptake types used by the NNSs. 43.5% of uptakes were repairs, whereas
21.7% involved the use of L1 and 17.4% were repetitions.
Table 14
Distribution of Error Types across Types of Uptake Moves Used by NNSs
Type of Uptake Moves  Lexical Syntactic Total 
Repair 14 (46.7%) 6 (37.5%) 20 (43.5%)
Elaboration 4 (13.3%) 2 (12.5%) 6 (13.1%)
Use of L1 6 (20%) 4 (25%) 10 (21.7%)
Repetition  5 (16.7%) 3 (18.8%) 8 (17.4%)
Use of keyboard symbol 1 (3.3%) 1 (6.2%) 2 (4.3%)
Total  30 (100%) 16 (100%) 46 (100%)
Of the fourty-six uptake moves, thirty (65%) were responses to lexical prob-
lems, whereas sixteen (35%) were responses to syntactic errors. These findings are
similar to those found in Morris’s (2002) and Pica’s (1994) studies confirming that
168 LEE
lexical items were negotiated and repaired more frequently than syntactic ele-
ments. Interestingly, L1 use (21.7%) as uptake was used similarly for lexical prob-
lems (20%) and syntactic errors (25%). In response to the implicit feedback from
the NS, the NNS tended to use their L1 to confirm the meaning of the unknown
lexical item (see example 2). One puzzling result is that L1 was used to respond to
the clarification request received from the NSs as shown in example 10: 
(10)
1. NNS: Nunca ha tomado una clase de Blackboard.
(He/She has never taken a Blackboard class.)
2. NS: ¿Quién? (Who?) [Clarification request]
3. NNS: I have never taken Blackboard. [Use of L1]
4. NS: Bien.“he”para yo y “ha”para él o ella. (O.K.“he” for I and “ha” for
he or she.)
5. NNS: Gracias! No he tomado. (Thanks! I have not taken.)
Negotiation occurred when the NNS misused the verb “to have” haber in line 1.
Instead of making the attempt to correct the morphological error from ha “he or she
has” to he “I have,” the learner used L1 to express the meaning (line 3). It is possible
that syntactic problems are more difficult to process than lexical errors as Morris
(2002) suggested in his study of negotiation moves, recasts, and learner repair in the
foreign language classroom. Another explanation could be that the occurrence of
the use of L1 might be related to learners’ language proficiency and personal learn-
ing style.
In this study, the corrective feedback provided by the NSs had a positive effect
on drawing learners’ attention to form that led to repairs. The results revealed that
46.7% of lexical errors were repaired after receiving either clarification or confir-
mation checks. In most cases, uptakes involved the confusion between two lexical
forms, such as cuento “story” vs. cuenta “bill” or brazo “arm” vs. abrazo “hug.”
Example 11 illustrates how the learner repaired (line 4) the lexical error from
derecha “right hand side” to derecho “right” immediately after receiving the con-
firmation check (line 2) provided by the NS: 
(11)
1. NNS: La gente no tiene derecha para leer los e-mails de otros.
(People don’t have right hand side to read other people’ e-mails.) 
2. NS: Derecha? (Right hand side?) [Confirmation check]
3. NNS: No tiene derecho, si? (Do not have right, yes?) [Repair]
4. NS: Si, estoy de acuerdo. Debemos respetar la privacidad de otras
personas.
(Yes, I agree.We should respect other people’s privacy.)
NS AND NNS FEEDBACK AND RESPONSES 169
Repair as uptake involving negotiation of meaning facilitated the NNS’s under-
standing of L2 vocabulary as the learner noticed the ill-formed lexical item. In
addition to lexical repairs, the NNSs self-corrected 37.5% of syntactic errors in the
use of temporal and aspectual morphosyntax, such as the use of fui “I went” versus
fue “he went” and trabajo “I work” versus trabajé “I worked.” These findings did
not corroborate those found in the studies of CMC conducted by Blake and Zyzik
(2003) and Tudini (2003). In their studies, the NNSs tended to ignore the syntacti-
cal errors in order to maintain the conversational flow.
Although the rate of the use of repetitions was low (17.4%), the rate at which
repetitions were used in the negotiation of lexical problems (16.7%) was very sim-
ilar to that for syntactic errors (18.8%) shown in Table 14. Anecdotal evidence
based on observations made of online exchanges suggests that the use of repetition
was often followed by a recast initiated by the NSs, especially with advanced gram-
matical structures such as is shown in example 12: 
(12)
1. NNS: Cuando me graduo, voy a hacer mucho dinero usando la
computadora.
(When I graduate, I’m going to make a lot of money using the
computer.)
2. NS: Cuando te gradues, vas a ganar mucho dinero? [Recast]
(When you graduate, are you going to make a lot of money?)
3. NNS: Si, cuando me gradue, gano mucho dinero. [Repetition]
(Yes, when I graduate, I make a lot of money.)
4. NS: Que bien.Tambien quiero ganar mucho.
(Great! I also want to make a lot.)
In this excerpt, the verb me graduo “I graduate” in the sentence Cuando me
graduo, voy a hacer mucho dinero (line 1) was used incorrectly. In this particular
case, the NNS did not know that certain adverbial conjunctions, such as cuando
“when” require the subjunctive when the event “to graduate” has not yet occurred.
The subjunctive me gradúe should be used instead of the indicative me graduo.
This type of grammatical concept is too advanced for learners at the intermediate
level to figure out on their own without the assistance of an expert. Not only the
target form of the verb graduarse in the present subjunctive but also the correct
lexical item ganar “to make money” instead of hacer “to do something” were
repeated correctly in the immediate follow-up response. It is possible that written
discourse, as represented on the screen via CMC, allows learners to notice the dif-
ference between target and non-target like forms (Pellettieri 2000; Warschauer
1996). Recasts via CMC seem to have a more positive effect on drawing the
learner’s attention to linguistic form than do the recasts occurring in the face-to-
face interaction. This current study makes no claim on the short-term or long-term
170 LEE
effect of immediate uptake on learners’ interlanguage systems. However, further
studies should address the long-term effect of incorporated recasts on learners’
linguistic improvement in NCI environments. 
Conclusion
Conversational interaction has become central to communicative language teach-
ing and learning. NCI allows L2 learners to extend communication beyond class-
room limits. More importantly, it affords unique opportunities for learners to use
the target language to develop their communicative language competence. Like
face-to-face interaction, NCI offers a powerful forum for learners to receive input,
provide feedback and produce output through interactive negotiation. The current
study focused on the examination of types of feedback and responses used by both
NSs and NNSs working in dyads to complete two task-based activities. The efficacy
of NCI made it possible to capture the commonly used feedback features—recasts
by the NSs and clarification checks by the NSs. Feedback, in particular as a source
of negative evidence, is essential to bring learners’ attention to particular linguis-
tic forms to make output adjustments. The evidence presented in this study
showed that the NSs played a teacher-like role that fostered negotiation of mean-
ing and focus-on-form interaction. Additionally, the NSs provided the NNSs with
authentic language discourse, new lexical items, and correct grammatical struc-
tures through recasts and modification devices. The findings of the current study
suggests that a reactive approach to form-focused instruction may induce learners
to pay attention to linguistic form, which in turn may restructure their inter-
language and increase language accuracy. 
Learner uptake in response to corrective feedback, including successful
repairs, is crucial for SLA. The findings demonstrate that the NNSs produced indi-
cations of uptake in 50% of their follow-up responses. A high rate of linguistic
error was repaired (43.5%) following confirmation and clarification checks. Given
that NCI is based on written communication, the availability of written visual dis-
play on the screen might have contributed to these high rates of learner uptake.
However, lexical errors (30 instances) were repaired at a higher rate than syntactic
errors (16 instances). This suggests that L2 vocabulary knowledge plays a crucial
role in NCI as it elicits learners’ feedback to unfamiliar linguistic input as well as
promotes corrective feedback on their developing interlanguage systems. Unlike
previous studies which have focused on NNSs’ uptake, this study examined NSs’
efforts to resolve negotiations of meaning. The results showed that L2 elaboration
for unknown lexical items and the use of L1 for syntactic problems had a positive
impact on learners’ comprehension. Until students at intermediate levels of profi-
ciency gain more control of advanced structures, the L1 is a good option for nego-
tiation of complex grammar. 
Although the findings support Long’s Interaction Hypothesis (1996), the cur-
rent study did not address whether responses to implicit feedback led to L2 devel-
opment but simply identified feedback features used by both NSs and NNSs to
negotiate meaning and form in the immediacy of ongoing dialogue. More
NS AND NNS FEEDBACK AND RESPONSES 171
empirical studies focusing on the relationship between feedback types and imme-
diate repairs are needed to determine the effect of uptake moves in NCI environ-
ments. Future studies could also address other factors that may affect the
negotiation process, such as task types, language proficiency levels, and gender. In
this study, two types of tasks were used; an open-ended question and a goal-
oriented activity, but the effect of each type of task was not compared. Additional
studies could address this issue to determine the relationship between task type
and corrective feedback. NCI supports the goal of empowering learners to become
active and effective language users, fosters a wide array of interaction, and pro-
motes negotiation of meaning. In conclusion, this research suggests that NCI has
the potential to extend and enrich students’ language learning experience beyond
the classroom setting. 
Notes
1. Both linguistic errors (e.g., lexical or syntactic items) and unclear or incomprehen-
sible written discourse (e.g., the entire message is problematic) are considered in
this study. Corrective feedback is provided by the NSs to indicate NNSs’ linguistic
errors while NNSs negotiate meaning or form when they encounter difficulty
understanding NSs’ output.
2. The term “negotiation moves” is used wherever previous SLA research is referenced
in this paper. Lee (2001, 2002b) uses “modification devices” in the analysis of the
data of the current study.
3. The Spanish Oral Proficiency Tests—SOPT (based on the ACTFL Proficiency
Guidelines) was administered to the NNSs before the study (for details, see Lee 2000).
4. Task type affects the extent to which the negotiation of meaning takes place. Long
(1996) examined the notion of a one-way versus a two-way task.A one-way task only
allows one participant to give information. In contrast, a two-way task involves the
exchange of information between two interlocutors and focuses on meaning, there-
fore, more modified interaction may occur in the process of the two-way exchange.
5. Blackboard is a software program that allows teachers to bring courses online
using multi-channel web tools (for more information about Blackboard, visit
http://www.blackboard.com).
6. The author is aware of Smith’s (2003) modification to the Interactionist coding
scheme in application to CMC. However, the current study is building on the
author’s previous research (see Lee 2001; Lee 2002b) drawing on a similar data
set.Thus, the same coding scheme created by Varonis and Gass (1985) is used for
this study.
7. All examples presented in this paper were taken from the current study and with-
out any correction.While accents, tildes and umlauts are supported in Blackboard,
students often do not use them.
8. As the current study is building on the author’s previous research (see Lee 2001;Lee
2002b), the term “clarification check” is used instead of “clarification request” to
indicate unknown lexical items and instances when a NNS asks for an explanation
from his or her interlocutor.
172 LEE
9. The category “non-understood message”often implies multiple linguistic errors that
indicate the entire discourse is problematic. In this study, requests for help refer to
instances when learners asked their interlocutors to elaborate on an earlier utterance.
10. In this study, both the NS and NNS immediate responses to feedback were con-
sidered to help identify distinct negotiation strategies used to reach mutual
understanding.
11. Repair is defined as the correction made by NNSs immediately after receiving cor-
rective feedback from NSs.
12. Words in square brackets with an asterisk represent linguistic errors.
13. It should be noted that in this study, morphological errors that caused lexical prob-
lems were coded as lexical errors.
References
Abrams, Zsuzsanna. 2003. The Effect of Synchronous and Asynchronous CMC on Oral
Performance in German.Modern Language Journal 87:157–167.
Ayoun, Dalila. 2001.The Role of Negative and Positive Feedback in the Second Language
Acquisition of the Passé Composé and the Imparfait. Modern Language Journal 85:
226–243.
Blake, Robert. 2000. Computer Mediated Communication: A Window on L2 Spanish
Interlanguage. Language Learning & Technology 4: 120–136.
Blake, Robert, and Eve Zyzik. 2003. Who’s Helping Whom?: Learner/Heritage-Speakers’
Networked Discussions in Spanish. Applied Linguistics 24: 519–544.
Chun, Dorothy. 1994. Using Computer Networking to Facilitate the Acquisition of
Interactive Competence. System 22: 17–31.
Darhower, Mark. 2002. Instructional Features of Synchronous Computer-Mediated
Communication in the L2 Class: A Sociocultural Case Study. CALICO Journal
19: 249–277.
Doughty, Catherine. 1993. Fine-timing of Feedback by Competent Speakers to Language
Learners. Georgetown University Roundtable on Languages and Linguistics:
96–108.
Doughty, Catherine, and Jessica Williams. 1998. Focus on Form in Classroom Second
Language Acquisition. Cambridge, UK: Cambridge University Press.
Duff, Patricia. 1999.Tasks and Interlanguage Performance:An SLA Research Perspective.
In Tasks and Language Learning: Integrating Theory and Practice, edited by
Graham Crookes and Susan Gass, 57–95. Bristol, PA: Multilingual Matters Ltd.
Ellis, Rod. 1999. Learning a Second Language through Interaction. Philadelphia, PA:
John Benjamins Publishing Company.
. 2003. Task-based Language Learning and Teaching. Oxford, England: Oxford
University Press.
Ellis, Rod, Helen Basturkmen, and Shawn Loewen. 2001. Learner Uptake in
Communicative ESL Lessons. Language Learning 51: 281–318.
Fernández-García,Marisol,and Asunción Martínez-Arbelaiz.2003.Learners’ Interactions:A
Comparison of Oral and Computer-Assisted Written Conversations. ReCALL 15:
113–136.
Gass, Susan. 1997. Input, Interaction and Second Language Learners. Hillsdale, NJ:
Lawrence Erlhaum.
Gass, Susan,Alison Mackey, and Teresa Pica. 1998.The Role of Input and Interaction in
Second Language Acquisition: An Introduction. The Modern Language Journal
82: 299–307.
NS AND NNS FEEDBACK AND RESPONSES 173
Gass, Susan, and Evangeline Varonis. 1994. Input, Interaction and Second Language
Production. Studies in Second Language Acquisition 16: 283–302.
González-Lloret, Marta. 2003. Designing Task-Based CALL to Promote Interaction en
Busca de Esmeraldas. Language Learning & Technology 7: 86–104.
Hatch, Evelyn, ed. 1978. Second Language Acquisition. Rowley, MA: Newbury House.
Iwashita, Noriko. 2003. Negative Feedback and Positive Evidence in Task-Based
Interaction. Studies in Second Language Acquisition 25: 1–30.
Izumi, Shinichi. 1998. Negative Feedback in Adult NS/NNS Task-Based Conversation.
Paper presented at the annual meeting of the American Association for Applied
Linguistics, Seattle,WA.
Kern, Richard. 1995. Restructuring Classroom Interaction with Networked Computers:
Effects on Quantity and Quality of Language Production. Modern Language Journal
79: 457–476.
Krashen, Stephen. 1985. The Input Hypothesis: Issues and Implications. London:
Longman.
Lee, Lina. 2000. Evaluating Intermediate Spanish Students’ Speaking Skills through a
Taped Test:A Pilot Study. Hispania 83: 127–138.
. 2001. Online Interaction: Negotiation of Meaning and Strategies Used among
Learners of Spanish. ReCALL 13: 232–244.
.2002a.Enhancing Learners’Communication Skills through Synchronous Electronic
Interaction and Task-Based Instruction.Foreign Language Annals 35:16–23.
. 2002b. Synchronous Online Exchanges: A Study of Modification Devices on
Nonnative Discourse Interaction. System 30: 275–288.
.2004.Learners’Perspectives on Networked Collaborative Interaction with Native
Speakers in the US. Learning Language & Technology 8: 83–100.
Leeman, Jennifer. 2003. Reasts and L2 Development: Beyond Negative Evidence. Studies
in Second Language Acquisition 25: 37–63.
Long, Michael. 1985. Input and Second Language Acquisition Theory. In Input in Second
Language Acquisition, edited by Susan Gass and Carolyn Madden, 377–393. Rowley,
MA: Newbury House.
. 1996. The Role of Linguistic Environment in Second Language Acquisition. In
Second Language Acquisition, edited by William Ritchie and Tej Bhatia, 413–478.
Handbook of Research on Language Acquisition:Vol. 2. San Diego:Academic Press.
Long, Michael, Shunji Inagaki, and Lourdes Ortega. 1998.The Role of Implicit Negative
Feedback in SLA:Models and Recasts in Japanese and Spanish.The Modern Language
Journal 82: 357–371.
Long, Michael, and Peter Robinson. 1998. Focus on Form:Theory, Research and Process.
In Focus on Form in Classroom Second Language Acquisition, edited by Catherine
Doughty and Jessica Williams, 15–41. New York: Cambridge University Press.
Lyster, Roy. 1994. Négociation de la Form: Stratégie Analytique en Classe D’immersion.
Canadian Modern Language Review 50: 447–465.
. 1998a. Recasts, Repetition, and Ambiguity in L2 Classroom Discourse. Studies in
Second Language Acquisition 20: 51–81.
. 1998b. Negotiation of Form, Recasts, and Explicit Correction in Relation to Error
Types and Learner Repair in Immersion Classrooms. Language Learning
48: 183–218.
. 2004. Differential Effects of Prompts and Recasts in Form-Focused Instruction.
Studies in Second Language Acquisition 26: 399–432.
Lyster, Roy, and Leila Ranta. 1997. Corrective Feedback and Learner Uptake. Studies in
Second Language Acquisition 19: 37–66.
Mackey, Alison. 1999. Input, Interaction and Second Language Development: An
Empirical Study of Question Formation in ESL. Studies in Second Language
Acquisition 21: 557–587.
174 LEE
Mackey, Alison. 2000. Interactional Feedback on the L2 Morpho-Syntax: Learners’
Perceptions and Developmental Outcomes. Paper presented at the American
Association of Applied Linguistics. Vancouver, British Columbia, Canada.
Mackey,Alison, Susan Gass, and Kim McDonough. 2000. Do Learners Recognize Implicit
Negative Feedback as Feedback? Studies in Second Language Acquisition 22:
471–497.
Mackey, Alison, Rhonda Oliver, and Jennifer Leeman. 2003. International Input
Incorporation of Feedback:An Exploration of NS-NNS and NNS-NNS Adult and Child
Dyads. Language Learning 53: 35–66.
Mackey,Alison, and Jenefer Philp.1998.Conversational Interaction and Second Language
Development. Recasts, Responses, and Red Herrings? The Modern Language Journal
82: 338–356.
Morris, Frank. 2002. Negotiation Moves and Recasts in Relation to Error Types and
Learner Repair in the Foreign Language Classroom. Foreign Language Annals
35: 395–404.
. 2005. Child-to-Child Interaction and Corrective Feedback in a Computer
Mediated L2 Class. Language Learning & Technology 9: 29–45.
Oliver, Rhonda. 1995. Negative Feedback in Child NS/NNS Conversation. Studies in
Second Language Acquisition 17: 459–483.
. 1998. Negotiation of Meaning in Child Interactions. The Modern Language
Journal 82: 372–386.
. 2000. Negotiation of Meaning in Child Interactions. The Relationship between
Conversational Interaction and Second Language Acquisition. The Modern Language
Journal 84: 119–151.
. 2002.The Patterns of Negotiation for Meaning in Child Interactions. The Modern
Language Journal 86: 97–111.
Payne, J. Scott, and Paul Whitney. 2003. Developing L2 Oral Proficiency through
Synchronous CMC: Output, Working Memory and Interlanguage Development.
CALICO Journal 20: 7–32.
Pellettieri, Jill.2000.Negotiation in Cyberspace:The Role of Chatting in the Development
of Grammatical Competence. In Network-Based Language Teaching: Concepts and
Practice, edited by Mark Warschauer and Richard Kern, 59–86. Cambridge, UK:
Cambridge University Press.
Pica,Teresa. 1994. Research on Negotiation:What Does it Reveal about Second Language
Learning Conditions, Processes and Outcomes? Language Learning 44: 493–527.
Schmidt, Richard. 1995. Consciousness and Foreign Language Learning:A Tutorial on the
Role Attention and Awareness in Learning. In Attention and Awareness in Foreign
Language Learning, edited by Richard Schmidt, 1–63. Technical Report No. 9.
Honolulu: University of Hawaii, Second Language Teaching and Curriculum Center.
Skehan, Peter. 1998. A Cognitive Approach to Language Learning. Oxford: Oxford
University Press.
Smith, D. Bryan. 2003. Computer-Mediated Negotiated Interaction:An Expanded Model.
The Modern Language Journal 87: 38–57.
Swain,Merrill.1995.Three Functions of Output in Second Language Learning.In Principle
and Practice in Applied Linguistics: Studies in Honour of H.G.Widdowson, edited by
Guy Cook, 125–144. Oxford: Oxford University Press.
Swain, Merrill, and Sharon Lapkin. 1998. Interaction and Second Language Learning:Two
Adolescent French Immersion Students Working Together. The Modern Language
Journal 82: 320–337.
Toyoda, Etsuko, and Richard Harrison. 2002. Categorization of Text Chat Communication
Between Learners and Native Speakers of Japanese.Language Learning & Technology
6(1): 82–99.
Tudini, Enza. 2003. Using Native Speakers in Chat. Language Learning & Technology
7(3): 141–159.
NS AND NNS FEEDBACK AND RESPONSES 175
VanPatten, Bill. 1990. Attending to Form and Content in the Input: An Experiment in
Consciousness. Studies in Second Language Acquisition 12: 287–301.
. 1996. Input Processing and Grammar Instruction in Second Language
Acquisition.Westport, CT:Ablex.
Varonis, Evangeline, and Susan Gass. 1985. Non-Native/Non-Native Conversation:
A Model for Negotiation. Applied Linguistics 6: 71–90.
Warschauer, Mark. 1997. Computer Mediated Collaborative Learning: Theory and
Practice. The Modern Language Journal 81: 470–481.
White, Lydia. 1990. Second Language Acquisition and Universal Grammar. Studies in
Second Language Acquisition 12: 121–133.
Williams, Jessica. 1999. Leaner-Generated Attention to Form. Language Learning 49:
583–625.
176 LEE