JOURNAL OF SECOND LANGUAGE WRITING, 5 (I), 51-75 (1996)

Peer Revision in the L2 Classroom:

Social-Cognitive Activities,
Mediating Strategies, and Aspects
of Social Behavior
OLGA S. VILLAMIL
MARIA C.M. DE GUERRERO
Inter American University of Puerto Rico

Little is known about what actually happens when two L2 students are
involved in peer revision of written texts. This article reports the results of a
study conducted among Spanish-speaking students in Puerto Rico which
sought to investigate (a) the kind of revision activities students engage in
while working in pairs, (b) the strategies peers employ in order to facilitate
the revision process, ond (c) significant aspects of social behavior in dyadic
peer revision. The participants were 54 intermediate ESL college students
enrolled in a writing course. Interactions between pairs of students during
two revision sessions were recorded and transcribed. Analysis of the tran-
scripts yielded seven types of social-cognitive activities the students engaged
in (reading, assessing, dealing with troublesources, composing, writing com-
ments, copying, and discussing task procedures), five different mediating
strategies used to facilitate the revision process (employing symbols and
external resources, using the Ll, providing scaffolding, resorting to interlan-
guage knowledge, and vocalizing private speech), and four significant
aspects of social behavior (management of authorial control, affectivity, col-
laboration, and adopting reader/writer roles). Results reveal an extremely
complex interactive process as well as highlight the importance of activating
and enhancing cognitive processes through social interaction in the L2 writ-
ing classroom.

The last two decades have seen a change in the approach to the teaching of
writing. Traditionally, composition teachers focused on the finished product,
that is, on a final text produced by the students on their own. Evaluation of the
finished product was the teacher’s main task and prerogative (Rothschild &
Klingenberg, 1990). The focus has shifted now to the process of writing, with
instruction critical in helping students go through the different stages of com-

We gratefully acknowledge the aid of the Faculty Research Program of Inter American Uni-
versity of Puerto Rico, which made conducting this research possible. We also want to thank the
editors and two anonymous reviewers of the Journal of SecondLanguage Writing as well as our
colleague Janice Heber for their helpful revision comments,
Correspondence and requests for reprints should be sent to Olga S. Villamil, Inter American
University of Puerto Rico, Metropolitan Campus, Division of Humanistic Studies, PO. Box
191293, San Juan, Puerto Rico 00919-1293.

51
52 VILLAMIL AND DE GUERRERO

posing-prewriting, writing, and revision-and in stressing student strategies

conducive to better written outputs (Emig, 1971; Flower, 1985; Flower &
Hayes, 1981; Hillocks, 1986; Zamel, 1987). Revision, or the transformation of
text through multiple drafts, has been identified as a crucial factor in achieving
quality in the final product (Bereiter & Scardamalia, 1987; Elbow, 1981; Som-
mers, 1980; Zamel, 1983). To help students with this process, a frequently used
technique which allows for the intervention of other students as audience and
collaborators is peer revision.
Although peer revision has been widely adopted by writing teachers in Ll
(Bruffee, 1984; Elbow, 1973, 1981; Gere, 1987; Nystrand, 1986; Spear, 1988)
and L2 classrooms (Bell, 1991; Hafemik, 1983; Hvitfeldt, 1986), there remain
some doubts as to the efficacy of the practice (Berkenkotter, 1984; Carson &
Nelson, 1994; Nelson & Murphy, 1992a, 1992b; Newkirk 1984). An area in
question is the capacity of novice writers for helping each other when working
in groups or dyads (Nelson & Murphy, 1992b). In actuality, this type of con-
cern cannot be adequately addressed unless more is known about what happens
when students gather to critique and revise their writing.
Most of the published studies on peer revision have focused on showing how
to implement the technique in the classroom and on appraising the effects of
peer revision on final drafts, from an Ll (Brakel, 1990; Broome, 1988; Droge,
1991; Horgan & Bamett, 1991; Matsuhashi & Gordon, 1985; Sitko, 1989) an
L2 (Chaudron, 1986; Rothschild & Klingenberg, 1990), and a foreign language
(FL) perspective (Hedgcock & Lefkowitz, 1992). Yet, as Di Pardo and Freed-
man (1988) state in their thorough review of peer response groups: “More stud-
ies are needed of the actual functioning of peer talk within writing classrooms”
(p. 143). Knowing what actually occurs when students get together to talk about
a writing piece may provide important clues to why peer revision is or is not
successful and ultimately may help practitioners make informed decisions as to
its classroom use. This issue is especially meaningful when it involves L2
learners, for whom the language of written and oral communication in the class-
room is simultaneously the language they are trying to learn. In this article, we
attempt to answer some questions concerning the actual dynamics of peer revi-
sion among ESL learners with a homogeneous language background.
Relatively few studies have investigated peer talk during revision. From an
Ll perspective, Gere and Abbot (1985) examined the language of writing
groups at three grade levels--5th, 8th, and 11-12th. An analysis of idea units
showed that writers focused primarily on content and remained mostly on task.
In general, students demonstrated that they were reflective about what they did
while writing; they recognized problems, considered possibilities, and assessed
the worth of their product, and thus exercised their metacognitive capacities.
Freedman (1992) also investigated peer talk in two 9th-grade Ll English class-
rooms and found a remarkable variety of behaviors and types of interaction
within response groups. Sixty percent of students’ talk showed how they
worked on response sheets and 40% how they informally and spontaneously
 PEER REVISION 53

discussed the content of their writing. Even though some resistance was found
to giving negative evaluation, two kinds of productive talk permeated the
response groups: talk about content and self-response as writers corrected them-
selves while reading aloud.
From an L2 perspective, Stanley (1992) studied the interactions of 15 col-
lege students who had received training in peer responding and compared them
to the interactions of students who had only participated in a short demonstra-
tion session. Findings showed that elaborate student preparation resulted in
more productive conversations about writing. Mangelsdorf and Schlumberger
(1992) examined the stance of 60 ESL freshman composition students when
responding in writing to essays written the previous semester by other ESL stu-
dents. Results demonstrated that most students assumed a prescriptive stance,
paying more attention to correct form than to communicating meaning. Also,
Nelson and Murphy (1992a) explored the task and social dimensions of a four-
membered writing group at the college level over six different collaborative ses-
sions. The high percentage of on-task comments indicated that students were
successful in the task dimension; however, when the dynamics of the group
were examined, the social dimension was less successful.
Other studies in L2 have investigated how social interaction shapes revision
activities. Connor and Asenavage (1994) conducted a case study of two four-
membered response groups at the college level in order to determine the impact
of peer responses on revisions and to compare influences from peer comments,
teacher comments, and other readers such as tutors in the writing center. Find-
ings revealed that 5% of text-based revisions resulted from peer comments, 35%
from teacher comments, and 60% from self and tutors’ comments. Mendonca
and Johnson (1994), in a study of peer review among 12 advanced ESL students,
found that students’ negotiations in interaction generated questions, explanations,
suggestions, and grammar correction. Students used peer comments to revise,
but used suggestions selectively, deciding ultimately what to accept. The stu-
dents’ perceptions of peer revision showed the practice to be an effective one.
In an effort to explore the social-cognitive dimensions of peer interaction
during revision, we conducted a study (Guerrero & Villamil, 1994) which
adopted a Vygotskyan perspective. In that study, we examined how 54 Spanish-
speaking ESL students working in pairs regulated themselves and others during
peer review and what types of social relationships were created between the
learners. Results showed individual fluctuation through object-, other-, or self-
regulation during the review sessions and a predominance of asymmetrical
social relationships between a self-regulated member and an other-regulated
one. Although the study we are reporting here is based on the same data
(recorded interactions), its focus, approach to the data, leading questions, and
method of analysis are entirely different.
As in the previous study, we draw theoretical support from Vygotsky’s work
(1978, 1934/1986). According to Di Pardo and Freedman (1988) “Vygotsky’s
theories suggest a close relationship between talk and writing and the impor-
54 VlLl.AMlL AND DE GUERRERO

tance of a research framework that leads to understanding how social interac-

tions, in this case in the form of peer talk, can contribute to writing develop-
ment” (p. 122). Vygotsky (1978) asserts that an individual’s mastery of higher
mental functions is derived from social interaction which has been mediated by
communicative language. His concept of “zone of proximal development,“*
which recognizes the importance of peer assistance in the solution of tasks and,
consequently, in learning, seems particularly applicable to the kind of collabo-
rative instructional activity that occurs during peer revision. In the belief that
“instruction usually precedes development,” Vygotsky (1934/1986, p. 184)
offers writing as “a good illustration” (p. 180) of the role that instruction has
in fostering development. Writing is a highly abstract, voluntary, and conscious
activity, even more so than oral or inner speech, and thus, in Vygotsky’s view,
its development depends upon the “contributions of instruction” (p. 184).
Instruction-based on assistance, imitation, and cooperation-has a critical role
in providing the individual with the psychological tools, usually semiotic in
nature, that are necessary for higher mental functioning (Vygotsky, 1978,
1934/1986). These and other notions rooted in Vygotsky’s theories (Bruner,
1985; Lantolf & Appel, 1994; Nystrand, 1986; Wertsch, 1985) provide many of
the underpinnings of our analysis of peer interactions.
Our review of the literature suggests that various aspects of the L2 peer revi-
sion process remain unexplored, in particular the question of what actually
occurs when two L2 learners engage in joint revision and how they manage to
assist each other. In order to narrow the focus of our investigation, we designed
the following research questions:

1. What kind of revision activities do students engage in while working in

pairs?
2. What strategies do students employ in order to facilitate the peer revision
process?
3. What are some significant aspects of social behavior that characterize
dyadic peer revision?

METHOD

Participants
The participants were 54 Inter American University of Puerto Rico students drawn
from three intact sections of an ESL course which emphasized the development
of writing skills.2 Although the overall number of students enrolled in these sec-
tions was 75, only students who met the following sampling criteria qualified as
participants in the study: (a) a score of 500 to 599 on the College Board ESLAI,3
the English as a Second Language Achievement Test used for admission purpos-
es; (b) Spanish as the native language; (c) no more than a year of residence in an
English-speaking environment; and (d) two previous intermediate-level ESL
courses at the university. To screen the participants for these qualities, a bio-data
 PEER REVISION 55

questionnaire was administered at the beginning of the course, and pertinent uni-
versity records were verified. All students had a chance to participate in the train-
ing, composing, and recording sessions designed for the study; however, only 54
students who met the sampling criteria provided the data for analysis.

Procedures
Two rhetorical modes were taught in this writing course: narration and persua-
sion. All three sections followed the same schedule of activities and were taught
by the same professor (one of the researchers). A training period of about 4
weeks was devoted to the teaching of each mode (narration preceding persua-
sion). During this period, the students were also introduced to the use of peer
revision, and materials, such as writing prompts (see the appendix) and sample
essays, were distributed. At the end of this period, a composing session took
place in class. The purpose of this session was to produce a first draft which
would be jointly revised with a peer during the next session.4
All students produced a first draft in this way. However, for the revision ses-
sion, care was taken to pair students who qualified as participants together. In
order to establish dyads, pairs of first drafts were randomly formed. To deter-
mine whose composition would be revised, two outside raters were instructed
to “choose the composition that needed more revision” from each pair. A third
rater was used in case of discrepancies. Interrater agreement was 83% for the
narrative mode and 82% for the persuasive mode. The students were thus paired
for each revision session. In each pair, there was a “writer,” whose composition
would be revised, and a “reader,” whose task was to help the author revise
his/her paper. These writer/reader labels, however, were not explicitly men-
tioned to the students. A tape recorder and a Revision Sheet with the following
written instructions were provided to each pair:

Today you will revise with your partner one of the compositions written in class.
You will record all your comments during the revision process whether in English
or Spanish. You may also use this sheet to take notes or make additional com-
ments. Try to focus first on content and organization and then on language use
and mechanics.

Writers were also orally instructed to read the composition aloud before joint
revision began. In addition to the Revision Sheet, some students made use of
the writing prompts distributed during the training. After each peer revision ses-
sion, the students were asked to work at home on a final version of their com-
position and submit it a week later. Qualifying participants’ interactions record-
ed during the revision sessions were then preliminarily screened for their
appropriateness as data and their technical quality. Four recordings were dis-
carded in this process because large pieces of the interaction had been erased
or were inaudible. The 40 recordings which remained were transcribed by grad-
uate students, and the transcripts were verified by the researchers.
56 VILLAMIL AND DE GUERRERO

Data Analysis
The analysis was based primarily on the transcripts made of 40 recorded dyadic
interactions, 17 from the narrative mode session and 23 from the persuasive
one.5 Additional data for analysis were the students’ first drafts, which were the
focus of revision during interaction, any revised version produced during the
task, and comments written on the Revision Sheets.
To analyze the various activities, strategies, and significant aspects of social
behavior that characterized the participants’ interactions, a predominantly qual-
itative approach was followed, with emphasis on discovering and classifying
revision behaviors that could be quantitatively analyzed in future studies. Due
to the extended and complex nature of the processes on which we wanted to
focus, whole transcripts of interactions between two peers constituted the units
of analysis. However, because the focus of our research questions was on “revi-
sion” processes, we concentrated our attention on those segments of the inter-
actions in which the students were on-task, that is, dealing with some aspect of
the revision task assigned to them. Off-task conversation, that is, talk unrelated
in topic or purpose to the contents of the composition or the task itself, did
occur but to a relatively small degree.6
An iterative method of analysis was followed during which we recursively
examined and reexamined transcripts. First, we defined the three areas that con-
stituted the foci of our research as follows:

Social-Cognitive Activities: activities displayed during peer interaction which

were thought to be the basis for cognitive processes related to revision.
Mediating Strategies: semiotically encoded actions which facilitated the achieve-
ment of task goals, that is, revising the text.
Significant Aspects of Social Behavior: salient behavioral issues that indicated
how peers handled their mutual interaction regarding the text.

Then, we drew a preliminary taxonomy of activities, strategies, and aspects of

social behavior. We subsequently analyzed transcripts to verify the appropriate-
ness of such taxonomy. Categories were thus added, deleted, or modified. A
final analysis was then conducted to determine the frequency, saliency, or pat-
tern in which categories occurred. In some cases, percentages of occurrence
were obtained by calculating in how many of the total number of transcripts the
observed behaviors took place.

RESULTS

Research Question 1
What kind of revision activities do students engage in while working in pairs?
As can be observed in Figure 1, we identified seven major social-cognitive
activities. The first major social-cognitive activity we detected was reading.
 PEER REVISION 57

1. Reading
2. Assessing
3. Dealing with troublesources
4. Composing
5. Writing comments
6. Copying
7. Discussing task procedures

Figure 1. Social-Cognitive Activities

This activity took place in two main instances: initially, as the writer read aloud
the text, and during peer interaction. Only in 1 out of 40 recorded sessions did
reading of the text occur after task completion, that is, after peer interaction had
ceased. Initial reading aloud, which occurred in 38 out of 40 interactions (95%),
was characterized by self-response revisions made by the writer.7 Self-response
revisions were comments or asides made by the writer as he/she read the text
aloud to the reader. Reading during peer interaction took various shapes: silent
reading; reading in chunks with alternating revisions; reading aloud in the
beginning, middle, or end of the session; rereading aloud text with corrections;
and reading instructions, guidelines on prompt sheets, or comments written on
the Revision Sheet. No particular person was in charge of the reading; in other
words, either partner read, sometimes taking turns in doing so, or both read in
unison. Only in two of the recorded sessions did the students not engage in
reading during interaction. In one of these cases, the students went off-task
immediately after the initial reading aloud phase, and in the other, the pair
restricted themselves to making general assessment comments.
Assessing-in the form of evaluative comments on the quality, absence, or
presence of textual elements or aspects of the writing process-varied among
students: Some started the task with a preliminary assessment, in general or
about a specific point; other students assessed in the middle or at the end. Only
four pairs performed no assessing at all. Interestingly, one kind of preliminary
assessment was “There’s nothing to revise,” although the students did find
sources of revision later. Some used the prompts provided during training and
went through them point by point or just focused on a particular item. Some
assessed if parts of the composition existed (“You need to write a conclusion”),8
whereas others judged the piece in general (“This was a good story”). Still oth-
ers assessed the task itself (“It was easy”) or made evaluative comments about
the effectiveness of peer or self as writer or reader (“Wow, you’re really good
at catching errors”). Frequently, the students wrote their assessment comments
on the Revision Sheet. Two things were absolutely clear about assessing: (a)
that it was a responsibility assumed by both writers and readers and (b) that
assessing did not imply taking any specific action in terms of revision. For this
58 VILLAMIL AND DE GUERRERO

reason, we are distinguishing between assessing and dealing with trouble-

sources-the next activity to be discussed.
Dealing with troublesources constituted the students’ main concern during
the revision session. Thirty-eight out of the 40 dyads (95%) engaged in this
activity. According to Nystrand (1986), troublesources are “choice points”
(p. 18), that is, parts of the text which learners choose to discuss because they
are “unclear or troublesome in some way” (p. 211). In our study, we defined
troublesources as perceived or potential problems, errors, or deficiencies in the
text, and they could range from grammar and mechanics to high-level concerns
such as organization and content. The following examples show how students
dealt with troublesources:

1. And here, oh, OK, I got it. . You don’t say “I didn’t have not even. . .”
I’m using the negative two times. You’re supposed to say “I didn’t have
even . .“.
2. With this sentence you can start another paragraph because you see . . . this
is something else you are going to talk about.
3. I think you need to include more details, more action.

As has been mentioned, recognizing and repairing errors or textual deficien-

cies started in the form of self-response as early as during initial reading aloud.
Like Freedman (1992), we found the purpose of these self-response revisions
to be either giving explanations to the listener or identifying a problem of clar-
ity, form, or content. Here are some examples of how writers started dealing
with troublesources while reading aloud:

1. Text version: There’s educated programs

Oral version: There’s ah . education programs
2. Text version: I told my lawyer and the judge [word judge is illegible]
Oral version: I told my lawyer and the juice? oh, my God, what does it
say here?

There was great variation in how the students dealt with troublesources.
Some pairs revised in a methodical way, starting from the top or the bottom of
the composition or working recursively in either direction. The students’ self-
imposed procedures involved reading and revising sentence by sentence, para-
graph by paragraph, or following the prompt sheet. Other pairs displayed an
erratic form, such as troubleshooting discrete points within the composition.
Discrete troubleshooting was triggered by various stimuli: specific questions
posed by the interlocutor, problems spotted while reading, or whatever caught
the eye. Dealing with troublesources was highly recursive. Some students
undertook the top-down revision pattern two or three times, each time spotting
new troublesources or going over old ones. Some students solved the problem
right away; others procrastinated and considered the aspect later on; and others,
 PEER REVISION 59

unable to solve the problem, either abandoned it or asked for outside help, as
in the case of consulting with a neighboring student or the teacher. In a few
cases, the students deliberately postponed revising a troublesource:

Reader: I don’t know how to say ‘exagerar.’ Do we leave that pending?

Writer: Yes.

Reader: Okay, then (continues with another troublesource).

Finally, it should be noted that troublesources were not dealt with at all in only
two of the transcripts, with students limiting themselves to assessing and/or
talking about task procedures or unrelated matters.
Assessing and dealing with troublesources sometimes resulted in composing
new sentences or missing parts, such as a conclusion or supporting paragraph. This
activity was undertaken in 26 out of 40 interactions (65%) by either writer or read-
er on his/her own or together by reader and writer. Quite interestingly, some writ-
ers provided content in the Ll while the reader translated to the L2 and wrote.
Writing comments during revision was quite common, occurring in 33 of the
40 recorded sessions (82.5%). Both reader and writer wrote comments through-
out the whole revision session but most frequently towards the end, as a con-
sequence of final assessment. Students wrote comments on the text or on the
Revision Sheet. These comments varied in purpose, from evaluative judgments
or opinions to reminders of what to do (add, delete, or modify) for the final ver-
sion. Some comments were very specific. (“I misspelled words like may-
my”), whereas others were more general (“And most important that everything
you write no matter what has to be complete and clear”).
Copying was observed in 40% of the transcripts (16 out of 40). Either read-
er or writer copied the corrected version on the Revision Sheet, usually one stu-
dent dictating to the other. Sometimes copying involved additional revisions as
students became aware of new troublesources or went over already revised ones.
Discussing task procedures was quite frequent, taking place in 31 of the
recorded sessions (77.5%). In discussing task procedures, the students talked
“about” the task rather than reading, assessing, dealing with troublesources, or
composing. This type of activity had a variety of purposes: clarifying instruc-
tions (“Do we have to use red ink to correct?‘), establishing turns (“You ask
me questions and I answer them”), assigning responsibilities (“I dictate it while
you write it”), wondering about what to do (“Do I have to write it here or on
the other paper?‘), establishing sequence of procedures (“First we have to write
the comments”), announcing next step (“Now we are going to copy it”), check-
ing on task development (“What are you doing now?‘), voicing complaints
about peer’s lack of participation (“You can’t even be heard. Say something so
that at least they know you were here”), and generally negotiating and making
decisions about how to carry on with the task (“Let’s write the comments in
English”). Both readers and writers participated in this type of discussion.
60 VILLAMIL AND DE GUERRERO

Sometimes, concern with task procedures led students to call upon the teacher
for directives and clarification. Among the students’ most common concerns as
they consulted with the teacher were whether or not to copy the new revised
version on the Revision Sheet, whether or not to leave the tape recorder on at
all times, and whether or not to use the writing prompts. Nine pairs did not dis-
play any discussion of task procedures. We can only speculate on two reasons
for this: Either those segments of talk were not recorded or the students had
internalized instructions and were absolutely sure about the goal of the task and
how to achieve it.

Research Question 2
What strategies do students employ in order to facilitate the peer revision
process? In Vygotskyan psycholinguistics (see Lantolf & Appel, 1994), it is
assumed that goal-oriented human social interaction is semiotically “mediated,”
that is, aided by psychological tools involving signs and language. In our study,
peers made use of five different mediating strategies in order to cope with task
demands (see Figure 2).
Employing symbols and external resources (dictionaries, prompt sheets,
teacher, classmates) was a quick and efficient way of organizing revision and
solving pressing textual concerns. This strategy had two purposes: to complete
the task at hand and to facilitate the writing of the final version. To cope with
their task, the students used such aids as dictionaries and prompt sheets. They
also turned to the teacher or neighboring students for help, much as if these
were readily available “human” dictionaries or textbooks. To facilitate the writ-
ing of the final version, the students used numbers or asterisks on drafts and
Revision Sheets, circled portions of the text, as well as utilized symbols such
as parentheses, brackets, and arrows. In 30% of the interactions (12 out of 40)
no symbols or external resources were employed.
Using the Ll was an important strategy to gain control of the task. In an over-
whelming majority of the cases (38 out of 40 interactions), Spanish provided a
verbal matrix for the interaction, with English used mainly to refer to specific
parts of the text or during reading, copying, and composing. In the remaining
two interactions, English predominated throughout. For most of the students, the
Ll was an essential tool for making meaning of text, retrieving language from
memory, exploring and expanding content, guiding their action through the task,
and maintaining dialogue. To make meaning of text, for example, students point-
ed to a word in English and translated it to Spanish to verify the meaning intend-
ed (“Downward . . ‘bajando,’ is that what you wanted to say?“). In some cases,
whole explanations of the meaning of the text were produced in Spanish either
by the writer for the reader to understand or by the reader to check his/her
understanding of the text. Sometimes, it was necessary to think about a word in
Spanish first to secure meaning and then retrieve the English equivalent from
memory, the dictionary, or teacher (“Aduana, aduana, [customs, customs] how
do you say that?‘). Frequently, the students explored aspects of the text contents
 PEER REVISION 61

1. Employing symbols and external resources

2. Using the Ll
3. Providing scaffolding
4. Resorting to interlanguage knowledge
5. Vocalizing private speech

Figure 2. Mediating Strategies

in Spanish either to have a broader grasp of it or to make additions. One student

confessed she had transferred the contents of a persuasive text she had written
for a Spanish class to English in order to comply with the assignment. In gen-
eral, Spanish was essential for the students to direct themselves through the var-
ious stages of the task, for example, when making comments such as “Vamos a
pasarlo y despues vamos a verificarlo [Let’s copy it and then we’ll check it],”
and for them to achieve the task goals through conversation.
Providing scaffolding was a general strategy whose main function was for the
members of the dyad to assist each other verbally in achieving task goals. Both
readers and writers provided scaffolding to one another by resorting to various
substrategies shown in Figure 3. 9 Among these scaffolding substrategies, the
most salient were advising and responding to advice, eliciting and responding to
elicitation, reacting, and requesting clarification. Some pairs adopted specific
patterns in the use of these substrategies; for example, writer elicited and read-
er responded to elicitation; reader reacted and advised and writer responded to
advice; reader requested clarification and writer either clarified or justified. In
addition, some students created strict rules for their own discourse, such as the
pair in which the writer told the reader to ask her questions so that she could
then answer. An interesting scaffolding substrategy, which clearly shows the
more “expert” peer assuming a didactic attitude, was instructing:

Reader to writer: “Remember this I always do it having in mind that the mean-
ing of are and is is plural and singular. That’s what I always keep in mind.”

Resorting to interlanguage knowledge was a frequent strategy when dealing

with troublesources. Mostly, students relied on “ear,” that is, they corrected the text
according to their familiarity with the word or grammatical structure in question.
“This doesn’t sound right” was a familiar remark in almost all transcripts. Simi-
larly, students turned to their own rules when revising or composing. False repairs
based on incorrect or incomplete knowledge of the target language occurred but
were not common. There were a few instances of overgeneralization (“I spell with
a ‘t’ all those words that end in Jan]“) and of hypercorrection, that is, correcting
when it was not necessary. Of course, the students also “corrected correctly,” that
is, they made use of standard rules of English which were part of their interlan-
guage repertoire.
 62 VILLAML AND DE GUERRERO

SOBSTNATNGT DEFINITION KXAUPLES

requesting asking for suggestions "IS the title OK?"

.d”iC4 "Should I expand it more?"
"Tell me "hot it needs."

suggesting revision or recommending that "You can p"t this in another

cbangas be made; going beyond advising by paragra.ph."
composing nev sentences for the writer or "You should say that it helped you
offerillg rpmzific solutions grow as a hum.n being."

.ccep+ing changes or solutions; questioning "OK ."

the rationale or validity of the advice; "Yer,. that’s tnJ.3:.
rejecting the suggestions made "Why [should I gut a colon]?"
"I don't think SO._

drawing out opinion or reaction, additional "You, as the reader, what do you
information or content, background knowledge, need?"
or understanding of text from peer "who took you to the airport?"
"What's your point of view?"
-Keep telling me, keep telling me
Other negative factors."

responding to giving opinion of reaction, additional "Yaah, it’s

all right."
elicitation information or content, or background "ub, footbPl1 is one sport where
knowledge; giving response about meaning as you have full contact with not one
-e&ad by peer but maybe fifteen persons."

reacting making evaluative comments about specific or "That was a nice erpe,rie"ce."
genera1 aspect* Of the text “It is we11 writtell:.
"This is an a-ror."

requesting asking interlocutor to clarify illegible *Is this what you wanted to say?"
clarification handwriting or intended meaning "What da you mean here?"
"Unuanted pregnancy?"

clarifying offering clarification of handwriting or "What I wanted to say there

meaning "BE..._

c
interpreting interlocutor's responseor
paraphrasing
caning

informing about
or about missing
text on

the
the basis

contents
parts
of understood

of a paragraph
"Do you mean 11
"So what you are
it's a debatable

"Here's the first 1easCxl:~

"The conclusion is missing."
students?"
saying
opinion."
is that

"This is another paragraph."

justifying explaining and defending choices or decisions "I didn't give many details
made about the text because I‘m not a wad observer."
“I’m goingto stop-this because
I'm not going to explain a11 the
political system of...."

instructing giving “mini” lessons on grammar.

voc.bu1arY. “[text says: 3 months] Don't use
stylistic conventians, or other aspects Of figures to exprelm numbers;
Vi-iting figures are used in math. You'd
etter write them in letters.
That's what my teachers have told
0."
:
giving rdering peer to take action (read, write, '*Let's go now to the
directives continue "itb the task, etc.) intmd”ct.io” ..
“You re.d now, so I can start
grasping the ideas."

co-nicrting specific id8.s by me.“= of

content-free spaceholders of feedback ["Yeah. yeah...."

Figure 3. Substrategies for providing scaffolding. (All examples have been given in English;
however, many of them were soid in Spanish.)

Vocalizing private speech was most noticeable when composing and reading
during interaction. In contrast to self-response during reading aloud in which
comments were made by the writer about his/her own text for the benefit of
 PEER REVISION 63

the peer who was listening, private speech utterances were comments to self
made with no discernible intent to have the peer hear what was said. Usually,
these private speech comments were mumblings or asides captured by the tape
recorder for which there was no response or reaction from the peer. As Lantolf
and McCafferty (1994) explain, private speech provides “externalized symbol-
ic expression to aid in the execution of a task” (p. 3). In this study, private
speech had two main purposes: to guide behavior or action and to release affec-
tive load. Private speech forms such as fillers, repetitions, questions to self, and
self-reminders were frequently used by students to guide themselves through
the task as they read portions of the text and composed new sentences. Inter-
jections, local idioms, and self-chastising comments, such as “iQuC bruta!
]How dumb of me!],” were verbal displays of the emotions generated by the
task.

Research Question 3
What are some significant aspects of social behavior that characterize dyadic peer
revision? As shown in Figure 4, four main issues of social behavior during inter-
action were identified. Within the first group, management of authorial control,
there was a variety of behaviors. One was relinquishing authorship, which
occurred when writers surrendered their rights as authors and gave up the lead
during revision. In some instances, the writer relinquished his/her author status
from the very beginning with comments such as “You can do whatever you
want”; in other instances, the reader gradually took the lead in revision and the
writer mostly responded in terms of procedures. In both cases, the writer com-
plied to the peer’s demands at the expense of losing his/her own voice. Appro-
priating was the counterpart of relinquishing. The writer or reader took posses-
sion of the task and obliterated the presence of the peer. Appropriation sometimes
occurred gradually until complete usurpation took place. For example, the read-
er’s use of “I,” as in “I think this ought to be eliminated, but Z will fix it for you,”
showed the intention of appropriation. Other examples of complete takeover were
when the reader asked the writer to read so “I can correct the text,” when the
reader mumbled while composing on his/her own a new version of the last para-
graph, or when he/she dictated new sentences in a commanding tone. In all these
cases, the writer became a backseat passenger who assented, mumbled, or
yawned because of displacement and lack of involvement.
Respect for authorship was shown when readers acknowledged the writer’s
rights over the text and treated it as private property. There was a delicate bal-
ance on the reader’s part between contributing ideas and avoiding invasion of
the writer’s territory. The reader’s use of the second person in comments such
as: “I am making this up, but you let me know if you don’t like it,” or “No, I
can’t change that; this is your composition,” or even the use of the third per-
son when referring to the text as something impersonal, “There is an error
here,” showed deference and tactfulness toward the author. Lack of respect for
the writer’s authorship was seldom observed. This behavior occurred, for exam-
ple, when the reader treated the peer as a child, as in “Callate, nene [Shut up,
 1. Management of authorial control
relinquishing
appropriating
respect for authorship
lack of respect for authorship
struggle for authorial control
maintaining authorial control
2. Collaboration
3. Affectivity
4. Adopting reader/writer roles

Figure 4. Significant Aspects of Social Behavior

you boy],” or laughed at the writer’s ideas and the way they were expressed.
This behavior seemed to be upsetting and intimidating to the author.
There was tension in the struggle for authorial control regarding who con-
trolled the interaction and led revision. This was observed when the writer gave
up authorial control to the peer and later on assumed authorship again. In very
few extreme cases, this tension was also characterized by disagreements or by
clashes. Participants manifested their opposition to changes, as in, “No, no, no,
I don’t like that; that is not what I want to say,” or imposed their opinion, as
in, “Well, you write it if you want to, if not . . . well, that’s my opinion.” Clash-
es occurred because both participants had strong personalities and did not want
to yield to the other. In either case, this extreme struggle for control was not
conducive to major revisions. In general, maintaining authorial control reflect-
ed the writer’s responsibility and ability to keep control of authorship. One way
of maintaining this control was through decision making, as when the reader
suggested a change and the writer replied, “No, no, leave it like that,” or by
clarifying doubts: “This is what I want to say.” Another observed way of main-
taining authorial control was persuading the partner to change a particular posi-
tion or opinion. For example, in the argumentative revision session, a student
used the text to pursue his political views.
The second main aspect of social behavior was collaboration. There were
two types of peer interventions: collaborative and noncollaborative. In collabo-
rative interventions, both individuals recognized and respected the other’s pri-
vate world. Both felt free to present their views for the creation of meaningful
text. Likewise, collaborative interventions occurred when one of the partici-
pants tried to see the text through the eyes of the author in order to help him/her
achieve the task goals. The peer became a strategic assistant who provided scaf-
folding to the other member and promoted development of text. Representative
comments of these interventions were: “What do you mean when you say
. . . 7”
.) “I don’t know if this is correct, what do you think?,” “OK, let’s go on
 PEER REVISION 65

working,” and “You see, we can work better this way . . . when I work alone,
sometimes I don’t know what to do, but two heads think better than one.” These
comments imply a bilateral view of the problems at hand and, therefore, an
understanding that through negotiation and compromise goals could be
achieved. This compromise could also be observed in the frequent use of the
first person plural, which in turn reflected joint authorship and responsibility,
and in the interchangeability of roles+ither one could read, compose, or jot
down comments, or either one could initiate and lead revision. Collaboration
was seen throughout dealing with troublesources, but it was mostly observed in
particular moments such as during composing, in which both reader and writer
had to use all their available resources.
Noncollaborative interventions, on the other hand, were characterized by
either an authoritative attitude or resistance to collaboration. In the former, one
of the participants controlled the task, imposing his/her views on the other. This
led to cases of appropriating, relinquishing, and struggling for authorial control
as previously mentioned. Resistance to collaboration occurred when one or both
reader and writer demonstrated passive behaviors, in other words, indifference
or unwillingness to participate in the revision of text. The passive attitude of a
writer moved his partner to say exasperatedly, “An opinion, Peter,10 please,
give your opinion.” Since the author did not become an active participant, the
reader’s reaction was to conclude the interaction. A different type of resistance
occurred when there was a perceived breakdown in communication. That is,
participants may have recognized a troublesource and may have displayed a
collaborative attitude, but because of limitations-linguistic constraints or lack
of knowledge of concepts-they could not work in unison; therefore, goals
could not be achieved. This was the case when one of the partners admitted his
limitations and justified himself saying, “I don’t know, details, you mean? To
tell you the truth, all of this writing in English and Spanish . . . (starts singing)
. . . I always do so bad.” In both cases of noncollaborative interventions, the
spirit of teamwork was lost and revision efforts declined.
The third aspect, ufSectivity, was crucial in collaborative interactions. It
encompassed camaraderie, empathy, and concern for not hurting each other’s
feelings. Oftentimes, readers and writers congratulated each other on a job well
done. Comments that could be interpreted as a pat on the shoulder (“This is
quite good.” “Wow, you know how to revise well.” “Your composition does not
have as many errors as you think.“) were enough stimulus to open up the social
relationship and make both participants less apprehensive and willing to col-
laborate. In some cases, empathy led peers to long conversations after task com-
pletion, consolidating the bond created during social interaction.
The fourth main aspect, adopting reader/writer roles, was characterized by
two salient behaviors: awareness of role and role switching. The first one,
awareness of role, meant that both readers and writers recognized and accept-
ed their roles-which had been tacitly assigned-and were cognizant of the
functions each role entailed. In general, there was no overt awareness; only a
66 Vll.lAMll. AND DE GUERRERO

few students mentioned the fact that roles were changed, for example, “Today
we have another role. . . . Last time you were the writer, remember?’ Each
member, however, kept to his/her function during interaction. In general, an
active reader was a critical eye, provider of suggestions, and simplifier of task.
An active writer, in turn, was a reviser aware of audience needs and receptive
to feedback, a composer, and final decision maker. But this reader/writer pro-
file did not inhibit members from assigning each other duties such as copying
the final version or jotting down comments.
Changes in attitude were observed between sessions when reader became
writer and vice versa, as well as when they exchanged partners. For example,
as the reader, Rosa was such an active collaborator and worked so closely with
the writer that many times she assumed authorship. In contrast, as the writer,
when she worked with Julio, she relinquished authorship and let him appropri-
ate the text. She picked up an active role at the end of the session when copy-
ing the corrected version. Joe’s case was quite intriguing, too. In both sessions,
he was the writer. In the narrative session, his peer was a more competent revis-
er, but in the persuasive one, his peer was not apparently perceived as such.
Joe’s change of attitude towards his peer was noteworthy. In the first task, Joe
was not willing to say much even though the reader tried to elicit information
for development of composition. He justified himself saying, “I am not a good
observer and do not like to write much.” However, in the second interaction, it
was Joe who constantly elicited from the reader and got upset when the reader
became distracted and turned to silly behavior.

DISCUSSION

The discovery of a wide variety of revision activities undertaken by peers con-

firms previous ch~acte~zations (Freedman, 1992; Nelson & Murphy, 1992a) of
peer revision as a comptex process involving a myriad of recursive behaviors.
Indeed, as demonstrated by the activities the students engaged in, peer revision
emerged as a total communicative experience in which students not only wrote
and read but also spoke and listened. The collaborative dyadic situation in
which peers found themselves fostered the occurrence of seven distinct social-
cognitive revision activities: reading, assessing, dealing with troublesources,
composing, writing comments, copying, and discussing task procedures.
Each social-cognitive activity had a critical role in the overall peer revision
process. Initial reading aloud was extremely important because it initiated revi-
sion as the writer stopped to make repairs, while providing the writer with a
sense of audience. Silent reading, on the other hand, seemed to indicate that
readers needed more than listening to the text; it seemed vital for them to estab-
lish visual contact, to interact with the text before proceeding to adopt the “read-
er and critic” role. Assessing was an activity which, we believe, evidenced the
effects of training. Students made comments related to the content of the com-
position using the terminology or following the sequence of items presented in
 PEER REVISION 67

the prompts. Dealing with troublesources manifested not only the recursive
nature of writing but also how individuals dealt differently and creatively with
the problem-solving situation. Composing revealed that in some cases the stu-
dents were able to produce new text as a result of peer intervention. Writing
comments and copying showed the collaborative spirit of teamwork, as either
reader or writer made comments or copied text on the Revision Sheet. Dis-
cussing task procedures denoted the importance of speaking as a cognitive activ-
ity during social interaction. Speaking “about” the task is, according to Brooks
and Donato (1994), essential in verbal interactions because it enables learners to
define the situation, set goals, and share orientation throughout the task.
The importance of these jointly undertaken activities resides in their role as
precursors of the conscious, volitional processes that characterize individual
writing activity. For Vygotsky (1934/1986), writing is a highly abstract and
deliberate action which presupposes the existence of inner speech, a mental
function whose evolution is in turn tied to external social speech (p. 182). Fur-
thermore, it is collaborative work and imitation, in Vygotsky’s view, that form
the basis of conscious mental activity: “What the child can do in cooperation
today he can do alone tomorrow” (p. 188). On these premises, we believe the
observed peer activities constitute the social basis for the development of cog-
nitive processes that are essential for revision. In short, the findings related to
our first research question are in agreement with Nystrand’s (1986) observation:
“The composing processes and discourse strategies that writers take from their
groups largely emerge on the basis of reflective and regulative processes that
can be observed to occur first in the social interaction of peer review” (p. 18).
Five different strategies were identified as facilitators of the peer revision
process. These strategies had the characteristic of being mediated in all cases by
semiotic or linguistic tools. By resorting to such forms of “semiotic mediation”
(Wertsch, 1985, p. 12) as symbols and dictionaries, the native language, verbal
scaffolding, interlanguage knowledge, and private speech, the L2 learners of our
study were able to gain control of the revision task. Our finding regarding the
use of the native language highlights the idiosyncratic nature of peer discourse
in homogeneous L2 contexts. For the Spanish-speaking students of our study, the
Ll appeared as a natural crutch for conducting interactions and solving revision
problems. This, of course, is not a novel finding. Research in the area of L2 writ-
ing shows L2 learners make use of the Ll to retrieve information from memo-
ry, make meaning of text, generate ideas, and generally improve the quality of
writing (Cumming, 1990; Friedlander, 1990; Lay, 1982). Our study not only con-
firms these uses of the native language but also reveals its power as a tool of
task control whenever partners share a common Ll background.
Providing mutual scaffolding in various forms was another important way
for students to achieve task goals. The term “scaffolding,” which has its origins
in Vygotsky’s (1978, 1934/1986) concept of zone of proximal development, has
been traditionally used in the literature (Bruner, 1978; Lehr, 1985; Searle, 1984;
Wood, Bruner, & Ross, 1976) to refer to the kind of unidirectional help that a
68 VILLAMIL AND DE GUERRERO

more expert partner provides to a less skillful one in collaborative problem

solving. We believe, however, that the notion of “mutual” scaffolding (Donato,
1994), by which two novices-rather than an expert and a novice-are capable
of providing guided support to each other through dialogic interaction, reflects
more accurately what we observed in our data. Although pairs had been formed
with the intention that the reader would assist the writer in improving his/her
composition (which had been selected on the basis of being the “one that need-
ed more revision”), our data indicated that assistance moved to and from read-
ers and writers.
How the peers managed authorial control, collaborated with each other,
showed affectivity, and adopted reader/writer roles were significant aspects of
social behavior observed in the dyads’ interactions. The students’ management
of authorial control revealed that this issue is as complex among L2 learners as
it is among Ll students (Berkenkotter, 1984). Definitely, the most obvious char-
acteristic during interaction was collaboration, in contrast to the findings of a
study conducted by Mangelsdorf and Schlumberger (1992), which presents a
prevalence of a “prescriptive” stance-rather than a “collaborative” or an
“interpretive” one-among L2 students responding to a text written a previous
semester by another student. Perhaps the lack of immediate response from the
author may have accounted for those results. In our study, most dyads estab-
lished joint responsibility for accomplishing the task.
The issue of collaboration seems to be tightly related to the social context
of the instructional situation. The students of our study, in general, tried to
establish a working atmosphere of camaraderie and compromise. Our hypothe-
sis is that this happened because joint work and collaboration are culturally per-
ceived as pleasurable and beneficial in Puerto Rico (Nine-Curt, 1993). Students
typically welcome team work in the classroom. In addition, interacting for
Puerto Ricans does not normally pose the communicative problems that other
heterogeneous L2 groups may have, inasmuch as Spanish is for them the com-
mon vernacular. Obviously, this cultural scenario is not repeated in all L2 teach-
ing contexts (see Carson & Nelson, 1994).
In general, we need to point out that social relationships were not static but
rather fluctuated as they developed during interaction and as roles changed. Many
interactions that started as collaborative continued as such until the end, but some
showed how individuals struggled for control as authorship shifted hands. Fluc-
tuation could also be seen when some readers, though respectful of authorship at
first, became so involved that they usurped the writer’s role or when writers, who
had relinquished control at the very beginning, gained it back during the course
of the interaction.

IMPLICATIONS AND RECOMMENDATIONS

The subtly woven tapestry of activities, strategies, and aspects of social behav-
ior that emerged from our study not only corroborates the notion that peer revi-
 PEER REVISION 69

sion is indeed a very complex process but also enlarges the picture of what hap-
pens during interaction and highlights some of the benefits of collaborative
writing in the L2 classroom.
It may be of comfort for teachers to learn, for example, that-when condi-
tions similar to those of our study are created-dyadic peer revision offers an
opportunity for bilateral, rather than unilateral, participation and learning; in
other words, both peers may give and receive help, both peers may “teach” and
learn how to revise. It may be encouraging for teachers likewise to know that
peer revision fosters a myriad of communicative behaviors, from reading and
composing in the L2 to making meaning of the L2 in the native language, and
to realize that students have at their disposal a wide range of strategies to
achieve task goals. Teachers may also find it worthwhile to see how students
profit from mutual collaboration. Sometimes when a teacher reads a student
paper without the benefit of a conference, he/she cannot fully grasp the stu-
dent’s goals and intended meanings. Peer revision, on the other hand, gives stu-
dents a chance to explain, defend, and clarify their points of view. In addition,
peer revision has the potential for bringing out into the open the students’ lim-
itations and creating awareness, without which remedial action would never be
successfully undertaken.
As in every human activity, social interaction during peer revision was not
always flawless. Certainly, there were cases of clashes and disagreements, atti-
tudes of passivity and authoritativeness, usurpation of text, and other unpro-
ductive behaviors among the students of our study. Teachers need to be fully
aware of the possibility of these behaviors. They may even want to forewarn
students against these behaviors during training and discourage these as much
as possible during revision sessions. At any rate, we view peer interaction, in
all its behavioral complexity, as a useful training ground for the students’ future
social and professional relationships. Peer collaboration in the writing class-
room may provide the stage for students to learn to cope with different types
of people, with different abilities and points of view, as well as to regulate their
own behavior accordingly.
The purpose of this study was to further the investigation of dyadic peer
revision in the L2 classroom. In doing so, we uncovered a wealth of activities,
strategies, and behaviors, some of which we could only touch on lightly. We
feel some topics we raised deserve greater attention than we did grant them,
and we admit this as a limitation of our study. Yet, we preferred to give the
more global, all-encompassing view of the terrain we discovered rather than a
narrower one for two reasons. First of all, we want to impress on our readers a
sense of the intricacy and totality of the whole peer revision process as we
observed it, and second, we believe the issues we raised may serve as points of
departure for future, more in-depth, studies.
Most of the issues which have been approached from a qualitative point of
view in this article can be quantitatively investigated in future research: for
example, the different types of troublesources students are concerned with, the
70 VILLAML AND DE GUERRERO

use and frequency of scaffolding strategies among the students, and a compar-
ison of collaborative and noncollaborative interventions. An important line of
research to be pursued (which we did not attempt at all because of the process
approach to the analysis of revision that we adopted) is the impact that joint
revision has on the final version produced by a student. In other words, to what
extent do writers incorporate revisions done in a peer review session in the final
version of their essays? Are there differences in the way collaborative and non-
collaborative interventions during peer revision affect the final writing of an
essay? Are final versions produced after a peer revision session significantly
better than draft versions?
In conclusion, the results of this study lead us to believe that peer revision
constitutes a unique opportunity for L2 students to discuss and formulate ideas
about the content of their writing as well as to assist each other in the devel-
opment of writing skills and discourse strategies. It is in the exchange of ideas
during interaction, where both peers extend and receive help, that they are able
to advance their knowledge. The wide variety of social-cognitive activities,
mediating strategies, and aspects of social behavior the students displayed sug-
gest that peer revision is not only the spark that activates collaboration but also
the spark that ignites cognitive processes. We need to take these findings into
account if we want to have a clearer and more realistic picture of peer revision
in an L2 context.

Notes
1. Vygotsky (1978) defines the zone of proximal development as “the distance
between the actual developmental level as determined by independent problem solving
and the level of potential development as determined through problem solving under
adult guidance or in collaboration with more capable peers” (p. 86).
2. The participants were 2nd- and 3rd-year students ranging in age from 18 to 25
years (M = 19.75).
3. Students taking the ESLAT test would score roughly about 100 points less on the
TOEFL (Alderman, 1981).
4. The topics for this first draft were, for the narrative task, “Narrate an experience
that taught you something about yourself or an experience that made you reflect on life”
and, for the persuasive task, “Convince/persuade your audience about that controversial
issue that you have been working on for the last few weeks.”
5. It should be noted that 26 students provided data from both sessions and 28 from
only one of these, thus accounting for 54 as the number of participants whose data were
analyzed.
6. In only 17 of the 40 transcripts did off-task conversation occur. Of course, this
result is based solely on “recorded” interactions; we admit the great likelihood that addi-
tional off-task behavior must have indeed occurred but was cut off from the recordings
as students self-manipulated the tape recorders.
7. Even though initial reading aloud was an activity the students were instructed to
do, we have included it among the types of reading performed because it was an inte-
gral part of the whole interactive revision process and because of the important behav-
iors that took place during its occurrence (e.g., self-response revisions).
8. Although most of the students’ comments were originally said in Spanish, they
have been translated to English in the text to facilitate reading.
 PEER REVISION 71

9. Many of the categories offered in Figure 3 have appeared in the literature of peer
response under various headings: “peer evaluator” and “writer” responses (Stanley,
1992), “mechanisms bridging interaction and achievement” (Webb, 1982), “language
functions” (Gere & Abbott, 1985), “tutor’s evaluative strategies” (Matsuhashi, Gillam,
Conley, & Moss, 1989), and “categories of student behavior and small-group interac-
tion” (Peterson, Wilkinson, Spinelli, & Swing, 1984).
10. Student names have been changed to preserve confidentiality.

REFERENCES

Alderman, D.L. (1981). Language proficiency as a moderator variable in testing academic aptitude
(TOEFL Research Report No. 81.41). Princeton, NJ: Educational Testing Service.
Bell, J.H. (1991). Using peer response groups in ESL writing classes. TESL Canada Journal, 8,
65-7 1.
Bereiter, C., & Scardamalia, M. (1987). The psychology of written composition. Hillsdale, NJ: Erl-
baum.
Berkenkotter, C. (1984). Student writers and their sense of authority over texts. College Composi-
tion and Communication, 35, 312-3 19.
Brakel, V.L. (1990). The revising process of sixth-grade writers with and without feedback. Jour-
nal of Educational Research, 84, 22-29.
Brooks, F.B., & Donato, R. (1994). Vygotskyan approaches to understanding foreign language
learner discourse during communicative tasks. Hispania, 77, 262-274.
Broome, M.C. (1988). A study of revision practices in the composition classroom. Dissertation
Abstracts International, 50, 657A-658A. (University Microfilms No. 89-10, 233)
Bruffee, K.A. (1984). Collaborative learning and the “conversation of mankind.” College English,
46, 635-652.
Bruner, J. (1978). The role of dialogue in language acquisition. In S.R.J. Jarvella & W.J.M. Levelt
(Eds.), The child’s conception of language (pp. 241-256). New York: Max-Plank-Institut for
Pshycholinguistik.
Bruner, J. (1985). Vygotsky: A historical and conceptual perspective. In J.V. Wertsch (Ed.), Culture,
communication and cognition: bgotskian perspectives (pp. 21-34). Cambridge: Cambridge
University Press.
Carson, J.G., & Nelson, G.L. (1994). Writing groups: Cross-cultural issues. Journal of Second Lan-
guage Writing, 3, 17-30.
Chaudron, C. (1986). The effects of feedback on students’ composition revisions. RELC Journal,
25, 1-16.
Connor, U., & Asenavage, K. (1994). Peer response groups in ESL writing classes: How much
impact on revision? Journal of Second Language Writing, 3, 257-276.
Cumming, A. (1990). Metalinguistic and ideational thinking in second language composing. Writ-
ten Communication, 7, 482-5 11,
Di Pardo, A., & Freedman, S.W. (1988). Peer response groups in the writing classroom: Theoretic
foundations and new directions. Review of Educational Research, 58, 119-149.
Donato, R. (1994). Collective scaffolding in second language learning. In J.P. Lantolf & G. Appel
(Eds.), Vygotskian approaches to second language research (pp. 33-56). Norwood, NJ:
Ablex.
Droge, E.F. (1991). How college students revise in response to written peer review exclusive of
oral components. Dissertation Abstracts International, 52, 2002A. (University Microfilms
No. 91-32, 399)
Elbow, P. (1973). Writing without teachers. Oxford: Oxford University Press.
Elbow, P. (1981). Writing with power. New York: Oxford University Press.
Emig, I. (197 1). The composing process of twelfth graders. Urbana, IL: National Council of Teach-
ers of English.
72 VlLl.AMlL AND DE GUERRERO

Flower, L. (1985). Problem-solving strategies for wrifing (2nd ed.). New York: Harcourt Brace
Jovanovich.
Flower, L., & Hayes, J.R. (1981). A cognitive process theory of writing. College Composition and
Communication, 32, 363-387.
Freedman, SW. (1992). Outside-in and inside-out: Peer response groups in two ninth-grade class-
es. Research in the Teaching of English, 26, 71-106.
Friedlander, A. (1990). Composing in English: Effects of a first language on writing in English as
a second language. In B. Kroll (Ed.), Second language writing: Research insights for the
classroom (pp. 109-125). New York: Cambridge University Press.
Gere, A.R. (1987). Writing groups: History, theory and implications. Carbondale: Southern Illinois
University Press.
Gere, A.R., & Abbott, R.D. (1985). Talking about writing: The language of writing groups.
Research in the Teaching of English, 19, 362-385.
Guerrero, M.C.M. de, & Villamil, O.S. (1994). Social-cognitive dimensions of interaction in L2
peer revision. The Modern Language Journal, 78, 484-496.
Hafemik, J.J. (1983, April). The how and why of peer editing in the ESL writing class. Paper pre-
sented at the State Meeting of the California Association of TESOL, Los Angeles. (ERIC
Document Reproduction Service No. ED 253 064)
Hedgcock, J., & Lefkowitz, N. (1992). Collaborative oral/aural revision in foreign language writ-
ing instruction. Journal of Second Language Writing, I, 255-276.
Hillocks, Cl., Jr. (1986). Research on written composition: New directions for teaching. Urbana, IL:
ERIC Clearinghouse on Reading and Communication Skills.
Horgan, D.D., & Barnett, L. (1991, April). Peer review: It works. Paper presented at the annual
meeting of the American Educational Research Association, Chicago. (ERIC Document
Reproduction Service No. ED 334 203)
Hvitfeldt, C. (1986, November). Guided peer critique in ESL writing at the college level. Paper pre-
sented at the annual meeting of the Japan Association of Language Teachers, International
Conference on Language Teaching and Learning, Hamamatsu, Japan. (ERIC Document
Reproduction Service No. ED 282 438)
Lantolf, J.P., & Appel, G. (Eds.). (1994). bgotskyan approaches to second language research. Nor-
wood, NJ: Ablex.
Lantolf, JI?, & McCafferty, S. (1994, March). Eliciting and analyzing private speech. Workshop
conducted at the annual Conference of the American Association for Applied Linguistics,
Baltimore, MD.
Lay, N. (1982). Composing processes of adult ESL learners: A case study. TESOL Quarterly, 16,
406.
Lehr, E (1985). ERIC/RCS report: Instructional scaffolding. Language Arts, 62, 667-672.
Mangelsdorf, K., & Schlumberger, A. (1992). ESL student response stances in a peer-review task.
Journal of Second Language Writing, 1, 235-254.
Matsuhashi, A., Gillam, A., Conley, R., & Moss, B. (1989). A theoretical framework for studying
peer tutoring as response. In C. Anson (Ed.), Writing and response: Theo?, practice and
research (pp. 293-316). Urbana, IL: NCTE.
Matsuhashi, A., & Gordon, E. (1985). Revision, addition, and the power of the unseen text. In S.W.
Freedman (Ed.), The acquisition of written language: Response and revision (pp. 226-249).
Norwood, NJ: Ablex.
Mendonca C.O., & Johnson, K.E. (1994). Peer review negotiations: Revision activities in ESL writ-
ing instruction. TESOL Quarterly, 28, 745-769.
Nelson, G., & Murphy, J.M. (1992a). An L2 writing group: Task and social dimensions. Journal of
Second Language Writing, I, 17 l-l 93.
Nelson, G., & Murphy, J.M. (1992b). Writing groups and the less proficient ESL student. TESOL
Journal, Z(2), 23-26.
Newkirk, T. (1984). Directions and misdirections in peer response. College Composition and Com-
munication, 35, 301-326.
 PEER REVISION 73

Nine-Curt, C.J. (1993). Intercultural interaction in the Hispanic-Anglo classroom from a nonverbal
perspective. In J. Fayer (Ed.), Puerto Rican communicarion studies (pp. 99-109). San Juan,
Puerto Rico: Fundaci6n ArqueGgica, Antropol6gica. e Hist6rica.
Nystrand, M. (1986). Learning to write by talking about writing: A summary of research on inten-
sive peer review in expository writing instruction at the University of Wisconsin-Madison.
In M. Nystrand (Ed.), The structure of written communication (pp. 179-211). Orlando, FL:
Academic.
Peterson, P.L., Wilkinson, L.C., Spinelli, E, & Swing, S.R. (1984). Merging the process-product
and the sociolinguistic paradigms: Research on small-group processes. In P.L. Peterson, L.C.
Wilkinson, & M. Hallinan (Eds.), The social context of instruction: Group organization and
group processes (pp. 125-152). New York: Academic.
Rothschild, D., & Klingenberg, F. (1990). Self and peer evaluation of writing in the interactive ESL
classroom: An exploratory study. TESL Canada Journal, 8, 5245.
Searle, D. (1984). Scaffolding: Who’s building whose building. Language Arts, 61, 48%483.
Sitko, B.M. (1989, March). How readers use “real audience” feedback: Problem solving in revi-
sion. Paper presented at the annual meeting of the American Educational Research Associ-
ation, San Francisco. (ERIC Reproduction Services No. ED 308 515)
Spear, K. (1988). Sharing writing. Peer response groups in English classes. Portsmouth, NH:
Heinemann.
Sommers, N. (1980). Revision strategies of student writers and experienced adult writers. College
Composirion and Communication, 31, 378-388.
Stanley, J. (1992). Coaching student writers to be more effective peer evaluators. Journal of Sec-
ond Language Writing, I, 217-233.
Vygotsky, L.S. (1978). Mind in society: The development of higher psychological processes. Cam-
bridge, MA: Harvard University Press.
Vygotsky, L.S. (1986). Thought and language (A. Kozulin, Trans.). Cambridge, MA: MIT Press.
(Original work published 1934)
Webb, N.M. (1982). Student interaction and learning in small groups. Review of Educational
Research, 52, 421445.
Wertsch, J.V. (1985). Introduction. In J.V. Wertsch (Ed.), Culture, communicarion, and cog&ion:
l+gorskian perspectives (pp. l-l 8). Cambridge, United Kingdom: Cambridge University
Press.
Wood, D., Bruner, J.S., & Ross, G. (1976). The role of tutoring in problem solving. Journal of
Child Psychology and Psychiatv, 17, 89-100.
Zamel, V. (1983). The composing processes of advanced ESL students: Six case studies. TESOL
Quarterly, 17, 165- 187.
Zamel, V. (1987). Recent research in writing pedagogy. TESOL Quarterly, 21, 697-7 15.
74 VILLAMIL AND DE GUERRERO

APPENDIX

Writing Prompts for Peer Revision of Narrative

and Persuasive Essays
All of the following passages are first drafts written by intermediate students in
previous courses. As you read them, you will notice they seem to be incom-
plete and/or underdeveloped and therefore need a lot of revision and editing.
First, we will focus on revising content and organization. Working in pairs, you
will read each composition and indicate which elements are present and which
are missing and thus need to be included to make it better. For example, you
will decide:

(Narrative Essays) (Persuasive Essays)

-if there is a setting, a beginning, a -if the topic of the essay is a

development of action, and an debatable opinion about a serious
ending issue
-if ideas relate to one main event or -if the opinion on the topic is clearly
action stated
-if you can follow a logical sequence -if the opinion is stated in the
of events; if words such as first, introductory paragraph
then, afterward, two hours later, -if the opinion is supported by at least
signal sequence of time and help two reasons
you follow the action -if each reason is supported by some
-if you can answer questions such as: kind of evidence such as facts,
who, what, when, where statistics, examples, or quotations
-if ideas are clear and developed and -if appropriate background
there is a sense of completeness information for the audience’s
-if there are enough details so the needs is provided
reader can see and feel what the -if some points that the audience may
writer sees and feels argue are included and debated
-if there is interesting dialogue -if the tone is consistent throughout
allowing the characters to speak to the essay
the reader -if ideas are connected logically and
-if the title is appealing and related to clearly
the ideas -if the concluding paragraph restates
-if the introduction invites the reader the opinion and summarizes the
to continue argument
-if there is a message-stated or
implied
-if the conclusion helps the reader
understand the writer’s feelings

Now, let’s work on editing. You will focus this time on language use and
mechanics. Select one passage and circle all language errors. Next to it, write
 PEER REVISION 75

what you think the error is. For example, you may notice that most frequent
errors are related to:

vocabulary-wrong use or invented word

connectors-more transition words needed to organize ideas or guide the
reader word order
verbs-wrong tense (past, present, etc.)
prepositions-in, on, at, from, of, etc.
word form-if the word is used as noun, verb, or adjective
subject and verb agreement-for example, he has/they have
sentences-no fragments or run-ons
spelling, punctuation, and capitalization
paragraphing+ne paragraph for each main idea and dialogues