The Two-Way Language Bridge:Co-Constructing Bilingual LanguageLearning OpportunitiesMELINDA MARTIN–BELTRANUniversity of MarylandCollege of Education2311 Benjamin BuildingCollege Park, MD 20742Email: memb@umd.edu

Using a sociocultural theoretical lens, this study examines the nature of student interactionsin a dual immersion school to analyze affordances for bilingual language learning, languageexchange, and co-construction of language expertise. This article focuses on data from audio-and video-recorded interactions of fifth-grade students engaged in joint writing activities (inSpanish and English). A qualitative analysis of discourse found that students seized opportu-nities to use two languages simultaneously, which multiplied opportunities for metalinguisticanalysis and bridged understanding across interlocutors. Findings suggest that language learn-ing affordances could be fostered in linguistically diverse classrooms by allowing interplaybetween languages and by creating activities that encourage learners to co-construct text. Thisstudy contributes to the expansion and reconceptualization of the field of language educationresearch by attending to bilingual language learners, or first language/second language users,whose reciprocal language learning experiences show how concepts from the fields of secondlanguage acquisition and bilingualism are necessarily linked. This study also contributes tolanguage learning research using a sociocultural perspective by revealing the ways that two lan-guages can simultaneously become mediational tools and objects of analyses within bilingualinteractional spaces.

DUAL IMMERSION EDUCATION AIMS TOincrease language learning affordances forboth minority-language1 students and majority-language students by offering a space where learn-ers are encouraged to draw on two or morelanguages as academic resources. Dual immersionprogram design is built on the assumption that in-teraction between learners is central to the learn-ing process, yet we know surprisingly little abouthow or whether the students interact meaning-fully (Genesee, 1985; Saunders & O’Brien, 2006;Valdes, 1997). Several studies have confirmedthe success of dual immersion programs with

The Modern Language Journal, 94, ii, (2010)0026-7902/10/254–277 $1.50/0C©2010 The Modern Language Journal

measures of high academic achievement (Alanis,2000; Christian, Howard, & Loeb, 2000; de Jong,2002; Lindholm-Leary, 2001, 2004; Thomas & Col-lier, 2002), yet there is less research about theeveryday interactions between language learnersand classroom contexts, which may lead to suc-cessful language learning and academic achieve-ment. As Hayes (2005) has shown, many stud-ies and reports have focused on program design(the “recipe approach,” p. 93) rather than thedialogic processes of language learning at workin the school. Angelova, Gunawardena, and Volk(2006) have begun to build this body of researchby describing strategies students used during peer-teaching. The findings from the present study cor-roborate findings from the Angelova et al. studyand go beyond to more closely analyze studentinteractions as collaborative dialogue in whichwe can observe “microgenesis” (Lantolf, 2001;

Melinda Martin-Beltran 255

Vygotsky, 1978) or processes of language learningunfolding during the interactions.

Recent studies (Fitts, 2006; Hayes, 2005; Lee,Bonnet-Hill, & Gillespie, 2008) have documentedthe challenges dual immersion programs faceand suggested that programs often foster “paral-lel monolingualism” instead of bilingualism. Thisseparation of language acquisition processes ininstructional design reflects the theoretical dividein second language acquisition (SLA) research,which has often conceptualized first language(L1) and second language (L2) acquisition as sep-arate and sequential rather than examining con-tinuous bilingual language development. Thesestudies call for further research to understandhow to create bilingual interactional spaces thatoffer a rich context for language learning. Thepresent study responds to this need for more re-search by focusing on moments when bilingual in-teractional spaces are successfully created, whichmay increase language learning affordances forstudents with a wide range of language experi-ences. The purpose of this study was to exam-ine the nature of student interactions using asociocultural lens to analyze potential opportu-nities for language learning, language exchange(across Spanish and English), and creative co-construction of language expertise.

THEORETICAL FRAMEWORK

Dual immersion program design draws fromtheoretical and empirical work in SLA, which hasdeveloped the input–interaction–output (IIO)model for language learning (see Block, 2003;Gass & Selinker, 2001; Long, 1996; Mackey, 2007).This model explains that interaction between L2learners and “native speakers”2 (as well as in-teraction between learners) promotes languagelearning through negotiation for meaning, modi-fied, comprehensible input, and opportunities forlearners to produce language and test new out-put hypotheses (Gass, 1997; Hatch, 1992; Long,1996; Long & Porter, 1985; Mackey, 2007; Mackey,Gass, & McDonough, 2000; Pica, 1994; Swain,1985). Several empirical studies (see review inMackey, 2007) have found a positive relation-ship between opportunities for learners to inter-act and language development (Keck, Iberri-Shea,Tracy-Ventura, & Wa-Mbaleka, 2006). Cognitive orcomputational models of SLA view input and in-teraction as activating individual, self-containedpsycholinguistic processes (Ellis, 2005), whereas asociocultural theory of the mind is primarily con-cerned with distributed cognition and conscious-

ness as mediated through social interaction andcultural artifacts.

This study employs a sociocultural conceptualframework, which shifts the focus from the indi-vidual learner to the social activity of learning, par-ticularly in dialogic interaction. This study buildson research (Anton & DiCamilla, 1998; Donato& Lantolf, 1990; Foster & Ohta, 2005; Lantolf,2000; Storch & Wigglesworth, 2003; Swain, 2000,2006; Swain, Brooks, & Tocalli-Beller, 2002; Swain& Lapkin, 1998; van Lier, 2000) employing a socio-cultural theoretical framework influenced by thework of Vygotsky (1978), which provides a lens tounderstand microprocesses of language learningin action that have often gone unexamined. Swainet al. have argued that peer–peer interaction fos-ters language learning and that a microgeneticanalysis (analysis of moment-to-moment commu-nication) will help us to better understand howlearners can create further language learning af-fordances (van Lier, 2000).

I draw from research examining collaborativedialogue (Swain, 2000; Swain & Lapkin, 2000) asan important window to understand how learn-ers work together to solve linguistic problems andco-construct knowledge about language. I foundthe language-related episode (LRE)—as a unit ofanalysis that describes a specific kind of collabo-rative dialogue—to be helpful both theoreticallyand methodologically. Swain and Lapkin (1998)defined an LRE as “any part of a dialogue wherethe students talk about the language they are pro-ducing, question their language use, or correctthemselves or others” (p. 326). My study builton this description of LREs informed by otherstudies (Foster & Ohta, 2005; Lapkin, Swain, &Smith, 2002; Mackey 2007; Mayo, 2002; Swain &Lapkin, 2002) to further investigate interactionsbetween young students and the practices that af-ford and/or constrain learning opportunities ina dual-language setting.

Recently, scholars have shown how the analy-sis of LREs can link cognitive and socioculturalaspects of language learning (Foster & Ohta,2005; Tocalli-Beller & Swain, 2007). They arguethat LREs capture more than moments of ne-gotiation for meaning (which are often aboutbreakdowns in communication) by shedding lighton sociocultural dimensions of communicationsuccesses, innovations, and co-construction oflanguage in interactions. Extending this work,Swain (2006) coined the term languaging to de-scribe the way that learners use speaking andwriting to mediate cognitively complex activ-ities. “Through languaging—a crucial mediat-ing psychological and cultural activity—learners

256 The Modern Language Journal 94 (2010)

articulate and transform their thinking into an ar-tifactual form, and in doing so, make it availableas a source of further reflection” (Swain & Deters,2007, p. 822). Data from the present study foundseveral examples of young students involved inlanguaging.

BROADENING THE CONCEPTUALFRAMEWORK BEYOND ONE–WAYLANGUAGE LEARNING

Much of the interaction research in SLA fromboth cognitive and sociocultural perspectives hasfocused on one-way language learning contextsrather than bilingual or two-way language learn-ing contexts (for a review, see Mackey, 2007 andSwain et al., 2002). Scholars such as Potowski(2004) and Valdes (1997) explain how the ra-tionale for dual immersion programs draws fromresearch in one-way contexts, which does not cap-ture the complexity of dual-language schools thatinclude heritage language speakers and a diversegroup of L2 learners/users. The present study re-sponds to Valdes’s (2005) call for a reconceptu-alization and expansion of the field of SLA byincluding bilingual, heritage language students(for whom the concepts of L1, L2, and nativespeaker are problematic) and by examining lan-guage learning in an educational context that in-volves several types of language development.

Although there is a growing body of empiri-cal studies that have been carried out in dual im-mersion programs (for reviews of the literature,see Howard & Sugarman, 2007; Howard, Sugar-man, & Christian, 2003), the theoretical frame-work underlying these studies and program de-sign is derived from SLA research that has focusedon the acquisition of one target language. Ratherthan recognize multilingual contributions to in-teractions, research analyses have been rooted inmonolingual assumptions prevalent in much SLAliterature that conceptualize L1 and L2 as sepa-rate and sequential rather than continuous (forcritiques of monolingual bias, see Block, 2003;Canagarajah, 2007; V. Cook, 1999, 2001; Levine,2003; Motha, 2006; Pavlenko, 2000). The concep-tual framework of this study considers continu-ous, simultaneous and reciprocal affordances forlanguage learning during interactions betweenL1/L2 users.3

The present study seeks to link research in SLAwith work done in bilingual settings in which allparticipants are developing multilingual reper-toires and may draw on these multiple linguis-tic resources within a shared interactional space.Within studies of bilingualism (Heller, 1999;

Romaine, 1996; Valdes, 2003; Wei, 2000; Zentella,1997), scholars have recognized the complexity ofmultilingual speakers who use a multitude of lan-guage varieties that are not separate and discreet;however, this is rarely recognized within SLA re-search (for exceptions, see Harris, 1999; Leung,Harris, & Rampton, 1997).

Although many studies recognize the impor-tance of students’ other languages and havepushed the field forward to consider the use ofthe L14 in L2 instruction (Anton & DiCamilla,1998; Brooks & Donato, 1994; Centeno-Cortes &Jimenez, 2004; Scott & de la Fuente, 2008; Storch& Wigglesworth, 2003; Swain & Lapkin, 1998),these studies have analyzed the functions of theL1 in service of learning the L2. In other words,L2 learning is the end goal (or object) and theL1 is considered separately as a mediational tool.This study broadens this research by consideringways that two languages can go back and forthsymbiotically as mediational tools and objects ofanalysis within the same interaction.

SCHOOL CONTEXT

This study took place in at the Escuela Unida,5

a dual immersion bilingual school located in anagricultural region in central California. As a pub-lic charter school whose enrollment was open tostudents beyond the local neighborhood, EscuelaUnida brought together students who might oth-erwise have little contact with each other, dueto housing segregation within the school district(that followed linguistic, ethnic, and socioeco-nomic lines). The student body was 90% Latino(including recent immigrants and U.S.-born chil-dren), 10% white and mixed heritage students,75% English language learners, and 87% receiv-ing reduced or free lunch. The school used a90/10 dual-language program whereby studentsbegan with 90% of their instruction in Spanishin kindergarten and reached a 50/50 balance inSpanish and English instruction by the fifth grade.

For this study, I chose to focus on one groupof 30 fifth-grade students who represented awide range of language experiences (from emer-gent to proficient bilinguals). In this class, therewere 3 newcomers from Mexico (who arrivedto the United States in the fourth and fifthgrade), 20 bilingual children (or heritage lan-guage speakers) who used mostly Spanish withparents and varying degrees of English at home,and 7 children who came from homes in whichthey spoke primarily in English. This class hadthree teachers—two Spanish-model teachers inthe morning and an English-model teacher in the

Melinda Martin-Beltran 257

afternoon. Each teacher had more than 7 years ofexperience and demonstrated great commitmentto the bilingual program and enthusiasm in theirteaching. They were cooperative partners open toreflecting on their teaching and discussing newideas for their classroom.

RESEARCH METHODS

Data Collection

The research methodologies that guided thedata collection for this project were modeled af-ter ethnographic and sociolinguistic studies thatinclude participant observation, interviews, andaudio recordings of classroom discourse (Free-man, 1998; Valdes, 2001; Zentella, 1997). Follow-ing guidelines for interpretive inquiry, ethnogra-phy, and participant observation (Creswell, 2007;Emerson, Fretz, & Shaw, 1995; Erickson, 1998;Hammersley & Atkinson, 1995; Miles & Huber-man, 1994), my fieldwork included intensive long-term participation in the school, careful record-ing and documentation (field notes, observationprotocols, analysis memos, questionnaires, stu-dent work, digital audio and video recordings),and an analytical reflection about the documen-tary record during the data collection process. Ivisited the school two to three times a week overthe academic year (August–June). In addition toclassroom, lunchtime, and recess observations, Iobserved school meetings, parent forums, parent–teacher conferences, after-school programs, anddistrict school board meetings in order to ex-plore the school conditions and social contextsthat supported or constrained opportunities fortwo-way language learning. To triangulate my ob-servations, I interviewed and surveyed students,teachers, parents, and the principal about lan-guage use in and out of the school, as well as theirperception of good language helpers and socialgroups. During class activities that required peerinteraction, I placed audio recorders on students’desks supplemented by a video camera in the cor-ner of the room while taking detailed field notes.In the subsequent description, I will focus on datafrom audio and video recordings during a jointwriting activity.

Description of Joint Activity

This classroom was not without its challenges(see discussion in Martin-Beltran, 2006); how-ever, I observed several occasions when studentswere successfully engaged in language exchange,collaborative dialogue, and languaging (Swain,

2006). From my classroom observations and pre-liminary analysis of the audio recordings ofinteractive activities, I identified activities that pro-moted the most LREs. I found the most LREs dur-ing activities that involved the creation and revi-sion of written text. I sought to build on the LRE-rich peer interactions during writing revisions tocreate activities that involved collaborative writingin both Spanish and English. My focus on writingactivities was guided by the theoretical claim thatwriting allows learners time to monitor (Krashen,1985) and reflect on their language productionand may be a tool for co-construction of knowl-edge (Haneda & Wells, 2000; Wells, 1999). Storch(2002, 2008) suggested that one way to encouragemetatalk about language is to require joint writtenoutput and reflection of the language produced.To increase interdependence and collaboration,Cohen (1994) recommended that groups workon a common written product; thus, the studentswere instructed to write a shared letter authoredcollaboratively. Drawing from suggestions inCohen’s review, the teacher and I modeled col-laborative writing, practiced metatalk about lan-guage with students, and provided self-evaluationand teacher-evaluation rubrics with specific col-laboration criteria so that the students understoodthat collaborative behaviors were an importantpart of this assignment. In my discussions with theteachers to create activities, we struggled with waysto design equitable, collaborative writing projectsthat might offer shared language learning affor-dances. Despite our efforts to guide and modelcollaboration, there were varying degrees of mu-tuality and equality (see Storch, 2002) among thepairs.

In her study that demonstrated the importanceof the collaborative nature of peer interactions,Storch (2002) identified patterns of interactionin terms of equality (control over the task) andmutuality (engagements with partner’s contribu-tions). I found Storch’s descriptive categories tobe helpful as I examined student interactions andconsidered ways that students engaged with eachother’s ideas, shared (or assumed) control overthe direction of the task, and made shared (orunilateral) decisions.

During the joint writing activity analyzed forthis article, students co-wrote two letters with theirpartner—one for a Spanish-speaking friend or rel-ative and the other for an English-speaking friendor relative (during Spanish instructional time andduring English instructional time, respectively).The assignment was designed to give the studentsan authentic purpose and audience for their writ-ing in both languages. In the letters, the students

258 The Modern Language Journal 94 (2010)

described themselves, focusing on the classmatewhom the letter recipient did not know. To brain-storm content for the letter, they followed an in-terview guide that asked the students to recounta memorable experience and to find out some-thing they had in common with each other. Theywere required to co-author a shared piece of writ-ing, in contrast to their usual writing assignmentsthat required independent authorship. Some ofthe students struggled with their shared, narra-tive voices, reflected in their usage of mixed first-person and third-person perspectives. However,this did not mean that one student did all of thewriting; in fact, it was common for a student towrite about herself in the third person. It is un-clear how much of this struggle with author voicewas due to a lack of familiarity with the formatof shared narrative and how much of this strug-gle was a true battle over writing territory. Dueto the limited scope of this article, a discussionof this point will be reserved for future research.To further encourage equitable collaboration, thestudents were required to switch roles as the writer(i.e., interpreted as being in control of the pen-cil) every few sentences and were encouraged to“write out loud” so their writing partner couldhear what they were writing during the composi-tion process. Although the students’ writing didnot often reflect a shared narrative voice, I foundthat the students co-created the texts in otherways throughout the process, which will be dis-cussed later. For example, their oral discussionswhile writing reflected a high degree of mutual-ity (Storch, 2002). After the letter writing activ-ity in Spanish and English, students completedself-evaluation rubrics about collaboration, whichguided them think about what collaboration lookslike in action. After they completed the activi-ties, I interviewed the students individually to askabout where they received or gave assistance withlanguage.

Data Analysis

I read through field notes and transcriptionsseveral times to seek out salient categories of talkthat emerged from the data (following qualitativeanalysis described in Emerson et al., 1995; Erick-son, 1998; Miles & Huberman, 1994). To focus thecoding of transcriptions and to capture the learn-ers’ attention to language, I identified LREs dur-ing the interactions. An LRE was identified whenstudents explicitly reflected on their language us-age, asked questions about language, or playedwith new language forms. I initially used LRE cod-ing categories from previous studies (Fortune &

Thorp, 2001; Foster & Ohta, 2005; Mackey et al.,2000; Swain & Lapkin, 1998); however, I foundthat much of the talk between the students was notcaptured by previously defined categories of LREs(morphosyntax, phonology, lexicon, and seman-tics), so I developed three additional coding cat-egories (crosslinguistic analysis, sociopragmatics,and written conventions) to describe other salientfeatures of student talk about language (for moredetailed examples of each category, see Martin-Beltran, 2006). An LRE could be short, involvingtwo turns of speech, or could occur over severalturns of speech with tangential remarks insertedin between their questions or doubts about lan-guage usage. In the excerpts presented here, thereader will notice that students may engage in sev-eral LREs that overlap and build upon each other.

Language-related episodes involving lexiconwere the most common kind of episode foundin the recordings of peer interactions during thefocal activity in both Spanish and English instruc-tional time. This was consistent with classroom ob-servations of other activities when students oftensolicited help from one another with quick confir-mation checks or “How do you say x?” During thiswriting activity, even apparently simple word-to-word translations often led to further discussion,other linguistic problems, or reformulation ofentire sentences. Lexical LREs often overlappedwith other categories, such as semantics and syn-tactic structure of sentences, which is exemplifiedin the excerpts. Please note that counting LREcategories was not the focus of my study. Instead,I focused on the nature of the LREs using a so-ciocultural lens to analyze moments at which lan-guage competence was explicitly discussed and si-multaneously constructed. The excerpts selectedrepresent the content and kind of speech fromthe larger sample of LREs found throughout stu-dent interactions (across 20 students involved inthe same activity).

Student Participants

I have chosen excerpts from 8 students (fourdyads) who represent a range of languageproficiencies and literacy levels (as seen inTable 1). Among the 20 students in the class wasa varying degree of mutuality and equality; how-ever, for the purposes of this discussion I delib-erately chose examples of interaction for whichstudents were involved in LREs, which dependedon a certain level of mutuality and collaborativebehavior (see Storch, 2002). Missed opportuni-ties and challenges are discussed further in otherworks (Martin-Beltran, 2007).

Melinda Martin-Beltran 259

TA

BL

E1

Foca

lStu

den

tCh

arac

teri

stic

s

Stud

ents

Shad

edbo

xes

Rea

din

gR

eadi

ng

Lan

guag

eL

earn

erH

ome

Lan

guag

eH

ome

Lan

guag

eA

DE

PTd

are

todi

ffer

entia

teL

evel

Lev

elSt

atus

Rep

orte

dU

seR

epor

ted

Beg

inn

ing/

En

dst

uden

tpai

rsSp

anis

ha

En

glis

h(a

sde

fin

edby

sch

oolb

)W

ith

Pare

nts

cU

seW

ith

Sibl

ings

Year

Scor

es

Hea

ther

Atg

rade

leve

lA

bove

En

glis

h-o

nly

/Spa

nis

hle

arn

erE

ngl

ish

En

glis

hN

/AIl

ian

aA

bove

Bel

owE

LL

Span

ish

Mos

tly

Span

ish

(som

eE

ngl

ish

)1/

2

Javi

erA

bove

Atg

rade

leve

lR

FEP

(red

esig

nat

edth

ird

grad

e)M

ostl

ySp

anis

hB

oth

N/A

Dan

iel

Atg

rade

leve

lA

bove

En

glis

h-o

nly

/Spa

nis

hle

arn

erM

ostl

yE

ngl

ish

En

glis

hN

/AIg

nac

ioB

elow

Bel

owE

LL

/RFE

P(r

edes

ign

ated

end

offi

fth

grad

e)M

ostl

ySp

anis

hB

oth

3/4

Rub

enA

tgra

dele

vel

Atg

rade

leve

lR

FEP

(red

esig

nat

edfo

urth

grad

e)Sp

anis

hB

oth

N/A

Mar

ioA

tgra

dele

vel

Atg

rade

leve

lE

LL

Span

ish

Mos

tly

Span

ish

3/3

Rya

nA

tgra

dele

vel

Atg

rade

leve

lE

ngl

ish

-on

ly/S

pan

ish

lear

ner

En

glis

hE

ngl

ish

N/A

a Bot

hSp

anis

han

dE

ngl

ish

read

ing

leve

lsw

ere

teac

her

asse

ssed

each

trim

este

rus

ing

sch

ool-w

ide

asse

ssm

entb

ench

mar

ks,w

hic

hw

ere

guid

edby

adop

ted

text

book

refe

ren

ces.

bE

LL

(En

glis

hla

ngu

age

lear

ner

)de

scri

bes

stud

ents

wh

osp

eak

ala

ngu

age

oth

erth

anE

ngl

ish

ath

ome

and

wh

oh

ave

not

yetm

etth

ecr

iter

iato

bere

desi

gnat

ed.

RFE

P(R

edes

ign

ated

Flue

ntE

ngl

ish

Profi

cien

t)re

fers

toa

stud

entw

ho

ente

red

sch

oola

san

EL

Lan

dbe

cam

epr

ofici

enti

nE

ngl

ish

.Th

isis

dete

rmin

edby

aco

mbi

nat

ion

ofst

anda

rdiz

edte

stsc

ores

,in

clud

ing

the

Cal

ifor

nia

En

glis

hL

angu

age

Dev

elop

men

tTes

t(C

EL

DT

),th

eE

ngl

ish

lan

guag

ear

tste

st,a

sw

ella

ste

ach

erre

com

men

dati

ons.

c Fro

ma

surv

eyth

atof

fere

dth

efo

llow

ing

choi

ces:

Span

ish

-on

ly,M

ostl

ySp

anis

h(s

ome

En

glis

h),

Bot

h(e

qual

ly),

Mos

tly

En

glis

h(s

ome

Span

ish

),E

ngl

ish

-on

ly.

dA

DE

PT(A

Dev

elop

men

talE

ngl

ish

Profi

cien

cyTe

st)

isa

lan

guag

eas

sess

men

tin

stru

men

t(al

ign

edw

ith

the

CE

LD

T)

used

wit

hst

uden

tsac

ross

grad

ele

vels

K–8

that

isle

vele

dfr

om1–

5(p

repr

oduc

tion

toin

term

edia

tefl

uen

cyII

).T

his

asse

ssm

entw

ason

lyus

edw

ith

EL

Ls;

no

data

wer

eav

aila

ble

for

non

-EL

Ls.

260 The Modern Language Journal 94 (2010)

Table 1 provides a summary of basic languageproficiency descriptors of all of the students high-lighted in the excerpts. I offer this chart to re-flect the kind of information that was availableto the teacher and used to inform instructionaldecisions; however, it is important to recognizethe limitations of such measures to capture thesituated nature of language competency. Seven ofthe 8 students (all except Heather) highlighted inthese excerpts could be considered heritage lan-guage speakers of Spanish because they have somefamily members who speak Spanish, althoughValdes (2005) and others have explained that thislabel refers to a wide range of language experi-ences. Valdes (2005) described heritage languagespeakers as “L1/L2 users” who “fluctuate in theirpreference or perceived strengths in each lan-guage . . . ” depending on contextual factors (p.414). To contextualize each student’s languageexperiences, additional background informationabout each participant will be discussed in eachexcerpt.

Findings

The findings of this study indicate that studentinteractions offer rich affordances for languagelearning when students are given the opportunityto draw on two or more languages simultaneouslyin dialogue with members of distinct linguisticcommunities as they participate in joint activities.Given time constraints on this classroom-basedstudy, it was not possible to develop individualizedassessments that would assess learning or changesin language development as a result of these in-teractions; instead, this analysis focused on LREsas opportunities or contexts for potential languagelearning . As I discuss the findings, I will drawon a sociocultural framework to qualitatively ex-amine how learning opportunities emerged dur-ing student interactions. The following excerptsdemonstrate the linguistic and metacognitive ac-complishments of the students who were simul-taneously expanding opportunities for languagelearning. I argue that using two languages astools for mediation further expanded opportu-nities for learning. The students’ dual languageuse not only deepened metalinguistic analysis butalso multiplied the language learning affordanceswithin the interactions.

Two Languages Bridge Gaps for One Student, Be-come Learning Opportunities for Another . The con-versation samples presented in Excerpt 1 takeplace between Heather and Iliana, who wereboth academically strong students, yet they bothexpressed frustrations and explicit awareness of

their weaknesses in their L2s. Heather was per-ceived as a strong English speaker and a weakSpanish speaker by her peers, although her liter-acy skills were considered at grade level in bothlanguages. Heather described herself as one of theonly blonde girls in school, which she recalled asa difference she felt when she entered EscuelaUnida in the first grade. Heather was confident ofher academic abilities, and when working in smallgroups, she often assumed a leadership role. Asa recent arrival from Mexico, Iliana came withstrong language and literacy skills in Spanish butvery little prior instruction in English. Iliana ar-rived with her family from Michoacan the summerprior to this study to join her cousins in California.Prior to enrolling at Escuela Unida, she attended5 years of school in Mexico, where she developedstrong literacy and study skills. Although she wasa new student, she demonstrated confidence andleadership when working in groups. Outside ofthe classroom, Iliana and Heather rarely spenttime together socially.

The girls described the complexity of theirbilingual competence, which included imbal-anced perceived proficiencies (Martin-Beltran,2007) when they wrote, “We are both bilingualand speak more one language than the other.”Their difference in proficiency levels suggested an“expert/novice” relationship in terms of Storch’s(2002) definition of low equality and high mutu-ality among participants. Although Iliana seemedto take control of the writing task at times, theirresolutions and revisions seemed acceptable toboth participants and there was evidence of re-ciprocal feedback between both girls. I observedeach student take on the role as novice andexpert at different points throughout the inter-actions, especially when comparing their letterwriting across Spanish and English. This schoolsetting, which recognized both Spanish and En-glish as important linguistic resources, offered animportant opportunity for an English languagelearner (ELL) to be placed in a position of pres-tige and power in contrast to mainstream ed-ucational settings in which ELLs are often as-signed low academic status (Christian & Bloome,2004; Yoon, 2007, 2008). Although Heather con-tributed her academic capital and experience asa veteran in this program, she also dependedon her newcomer partner, Iliana, to co-constructthe text and often take the lead in composing inSpanish.

In Excerpt 1, Heather was describing a fright-ening moment when she and her friend almostfell out of a tree. Iliana acted as an active listeneras she transposed this experience to written formfor the letter to her cousin.

Melinda Martin-Beltran 261

EXCERPT 1

Original Utterance6 English Gloss (when needed)[actions and comments in brackets]

1 Heather: Yo estaba como diez pies del . . . suelo I was like 10 feet from . . . the ground[She lifts her arm to show height]

2 “and” una cosa que se cayo en su lado “and” a thing fell by her side[pantomimes a falling branch]

3 y . . . yo estaba . . . “leaning?” and . . . I was . . . . “ leaning?”4 Iliana: ¿Que? What?5 Heather: Leaning?. . .6 Iliana: [look of confusion from Iliana]7 Heather: Como ası Like this“I’m leaning” en mi brazo [showing movement by leaning over her

chair]“I’m leaning” on my arm

8 Iliana: ¿inclinada? Inclined/ leaning9 Heather:: . . . estaba inclinada en una cosita

del arbol como ası. . .I was leaning on a little thing from the tree likethis[leaning over on chair]

10 y inmediatamente se quebro esta cosita y fuecomo ası en el arbol y ya mero me caı!

and immediately it broke, this little thing, and itwas like this on the tree and I almost fell![pretending to fall]

11 H& I: @@@ [both girls laugh]12 Heather: Estaba como ası en el arbol! I was like this in the tree!

[She shows herself hanging on, leaning overher chair]

13 Iliana: A ver. . . Let’s see. . . (pause 12 seconds) [She puts thepencil to her mouth as she is thinking, thenshe begins to write out loud]

14 ‘Ella estaba con su amiga en un arbol cuandouna’ . . . este . . . ¿una parte?

She was with her friend in a tree whena . . . um . . . a part? [with rising intonation]

15 Heather: Un pedazo del arbol se cayo y yamero se cayo :: al piso. . .@@. . .al suelo::

A piece of the tree fell and she almost fell to thefloor [laugh] . . . to the ground

16 Iliana: ::una parte del arbol se cayo:: . . .a part of the tree fell[she speaks softly out loud what she writes]

17 Heather::: It’s so confusing!::[she whispers under her breath, frustrated]18 Teacher: What’s so confusing?19 Heather: How like outside it’s different than

inside on the floor. . .“piso or suelo.” Likefloor and dirt. We don’t say that, we sayground and ground or floor and floor :: ordirt or no uh ::

20 Iliana: Ok, que les parece? Ok, how does this sound to you?[She announces and reads what she haswritten]

21 ‘Ella estaba con su amiga en un arbol cuandouna parte del arbol se quebro y estaban apunto de caerse’

She was with her friend in a tree when a part of thetree broke and they were just at the point of falling

22 Heather: Sı, esta bien Yes that’s good

262 The Modern Language Journal 94 (2010)

During the interaction transcribed in Excerpt 1,several kinds of LREs occurred. Heather and Il-iana discussed and repaired lexicon and syntaxand created simultaneous opportunities for lan-guage learning in both Spanish and in English(even though this activity took place during Span-ish instructional time). When Heather encoun-tered a lexical dilemma during her story telling(line 3), she paused and then inserted the wordin English (the nontarget language) in a risingintonation to signal questioning and a request forhelp from her partner. Using English (her dom-inant language) gave her a way to continue thestory; yet when Iliana did not immediately trans-late the English word, Heather was compelled touse other strategies (such as gesture and physicaldemonstration) to communicate her ideas andelucidate the meaning of the unfamiliar word toIliana. While demonstrating the action, Heathercontextualized the English word in a Spanish sen-tence (“I’m leaning en mi brazo”), codeswitchingcoherently and adapting a common teacher strat-egy to explain a new word by using that wordin a sentence. Different from teacher talk, whichpromoted a monolingual model in this instruc-tional space, the students created a bilingual in-teractional space (Lee et al., 2008) drawing ontwo linguistic codes simultaneously. The teachercould have extended this crosslinguistic compari-son by calling attention to the different ways thatSpanish and English words express the mannerin which motion events occur.7 In this instance,the teacher did not intervene but allowed the stu-dents to resolve this problem of word choice ontheir own, thus emphasizing the importance ofpeer collaboration and interdependence.

By repeating the new English word “leaning”three times and contextualizing this word withinIliana’s dominant language, Heather providedcomprehensible input and together they createdincreased language learning affordances in bothSpanish and English. Iliana demonstrated her un-derstanding of the new English word “leaning”when she offered the Spanish word inclinada,which Heather appropriated to continue her nextsentence (line 9). This evidence of uptake (or useand incorporation of the new vocabulary) is con-sidered an important step toward language devel-opment (Gass & Selinker, 2001; Mackey, Oliver, &Leeman, 2003; Oliver, 2000). Beyond simply re-ceiving and using the new language (which tradi-tional definitions of uptake suggest), I argue thatHeather and Iliana are appropriating new knowl-edge through this dialogic interaction. Appropri-ation is an ongoing process that describes the waysthat learners are transforming texts, learning pro-cesses, artifacts, and situations as they make them

their own (see Rogoff, 1995; Wells, 1999; Wertsch,1998). In the discussion section, I will further ex-plain how the sociocultural concept of appropri-ation illuminates the transformational processesand co-construction of language knowledge dur-ing these interactions.

Although a monolingual analysis of this interac-tion might focus on the gap in Heather’s Spanishlexicon, a bilingual, two-way analysis reveals theway this gap created additional opportunities forlearning both English and Spanish. In this case,Heather’s struggle with language created an un-expected, unplanned (in terms of the teacher’sobjectives), and often unacknowledged opportu-nity for Iliana to learn a wider range of English vo-cabulary. By using English in place of the missingSpanish word, Iliana was presented with new En-glish lexicon within a meaningful context. Theiruse of two languages to understand the meaningof “leaning” also afforded additional opportuni-ties for the students to compare (or at least be-come exposed to) different languages’ concep-tual frameworks for expressing motion events,which might be expanded in the future. A bilin-gual analysis of this interaction calls attention tobidirectional language learning; for example, inthis moment when both students were strugglingand negotiating for meaning, both students werealso potentially developing their respective targetlanguages.

Refining Understanding of Academic LanguageThrough Interaction. As Iliana began to read aloudher written work to clarify meaning and ask for ap-proval from her co-writer, she echoed Heather’searlier strategy to initiate an LRE—rising intona-tion to signal a question and a request for help—asshe searched for a more specific, academic wordto replace una cosita ‘a little thing’ (line 14). Il-iana’s questioning elicited other suggestions fromHeather, who offered a synonym un pedazo ‘apiece’ (line 15). Although Iliana did not take upHeather’s suggestion, she did incorporate the restof Heather’s sentence (beginning with se cayo inline 15)—demonstrating the ways they were co-constructing this text. Foster and Ohta (2005) de-scribed co-construction as “allowing learners toparticipate in forming utterances that they cannotcomplete individually, building language skills inthe process” (p. 420).

At the end of this excerpt, we see furtherevidence of linguistic work accomplished dur-ing the collaborative interaction, when Ilianareformulated Heather’s sentences by substitutinglexicon (una parte for cosita or Heather’s sugges-tion of pedazo) and repairing syntax (estaban apunto de caerse—‘they were at the point of falling’).

Melinda Martin-Beltran 263

Iliana also contributes her understanding or sen-sitivity to academic register when she transformsthe colloquial, informal phrase ya mero to a moreformal register, estaban a punto de caerse . The na-ture of the assignment required both participantsto pay attention to linguistic transformations, asthe oral telling took on a written form using aca-demic language and both students shared in thecomposing and reading of the new text.

Two Languages Promote Metalinguistic Aware-ness and Private Speech Becomes Collaborative . AsHeather was retelling her story (line 15), she ini-tiated another LRE as she struggled with wordchoice (piso/suelo) and considered different con-texts in which to use and distinguish these wordsacross her two languages. After Heather had amoment to reflect on her sentence, which sherepeated for Iliana to write, she verbalized herthinking in the form of private speech (Thorne &Lantolf, 2007; Vygotsky, 1978) as she expressedher confusion between lexical choices (“It’s soconfusing!” in line 17). The teacher played akey role in this interaction when she noticedHeather’s remark and she encouraged Heather’sverbal problem-solving process by asking, “What’sso confusing?” Heather’s private speech, or self-directed questioning to help direct one’s ownmental activity when confronting cognitive diffi-culties (DiCamilla & Anton, 2004; Donato, 2000;Thorne & Lantolf, 2007; Vygotsky, 1978), be-came socially mediated when she engaged in con-versation with the teacher. This is an exampleof how “cognitive processes derived dialogicallycan be observed directly in linguistic interactionsthat arise among speakers as they participate inproblem solving tasks” (Donato & Lantolf, 1990,p. 85). This dialogue with the teacher provideda window into Heather’s cognitive processes ofassessing the appropriateness of alternative lexi-con for “floor/ground” depending on the con-text. Although Heather’s linguistic confusion wasleft unresolved, this metacognitive moment af-forded Heather an opportunity to think metalin-guistically as she verbalized the usage rules for“floor/ground” by comparing Spanish and En-glish. Through this crosslinguistic comparison,she used language as a tool for mediation andas an object for analysis. This demonstrates fur-ther evidence for how two languages offered ad-ditional affordances to engage in metalinguisticanalysis, comparing semantics and lexical choicein different contexts across languages.

Heather’s switch to English, as private speech,exemplifies a case where a learner draws on herL1 in order to think metacognitively (see Anton

& Di Camilla, 1998); however, as social speech,this switch to English also excluded Iliana fromthe conversation (as Iliana was often positionedas a nonmember8 of the English community andit was more difficult for her to participate inEnglish-only conversations). While the teacherand Heather were having their conversation onthe side, Iliana was busily writing the text. Theteacher might have redirected this situation todraw Iliana into this questioning and crosslinguis-tic comparison, thus multiplying opportunities forincreased metalinguistic awareness for both stu-dents in both languages.

Two Languages Inspire Crosslinguistic Word Anal-ysis and Linguistic Creativity. In contrast to thefirst pair discussed, the students in Excerpt 2 weremore similar in their language strengths as her-itage language learners, although they also camefrom homes in which different languages were pri-marily spoken. Daniel and Javier were both recog-nized by their teachers as good students in bothSpanish and English. The boys told me they weregood friends, which is also apparent from the waythey played together during recess and lunch timewith a common group of friends.

Javier’s parents were born in Mexico and spokeonly Spanish at home, but Javier told me aboutan extended network of family members whowere bilingual and well established in California,where he was born. On the schoolyard and dur-ing small-group classroom activities, Javier movedcomfortably between Spanish and English withhis peers. He was often called on to explain con-cepts in both English and Spanish classes and hispeers recognized him as a good helper in bothlanguages.9

Daniel reported that his parents knew Span-ish, but they spoke mostly English at home withhim and his two older siblings. He described hismother as Filipina and European in origin andalso as a fluent Spanish speaker who majored inSpanish in college. Daniel described his father asa Mexican American who grew up in Texas in abilingual family. Daniel spoke both Spanish andEnglish with his friends but showed higher pref-erence for English in social situations.

Excerpt 2 was part of a conversation betweenDaniel and Javier who were discussing Javier’sscariest memory to share in their letter. Javier toldDaniel that he did not have nightmares after hisscary experience and Daniel wrote this in the let-ter. As Daniel wrote the word “nightmares” andread it out loud, he and Javier began to talk spon-taneously about the meaning and structure of theword.

264 The Modern Language Journal 94 (2010)

EXCERPT 2

{English translations in curly brackets when needed}

1 Javier: When he looked back . . . there was nobody there2 Daniel: There was nobody there [Daniel talks slowly as he is writing]3 Javier: He thought it was the grim reaper @@4 Daniel: @@ ‘He thought it . . . was the grim reaper.’ [writing what Javier said] Did you have

nightmares?5 Javier: No6 Daniel: ‘But he didn’t have nightmares’ [Daniel writes, reading out loud] . . . .Why do they

call them nightmares if they’re not - night mirrors?7 Teacher: That’s a good question!8 Javier: That’s a scientist ::[inaudible overlap]. . .mirror::9 Daniel: ::Yeah, I’ve been thinking since like 5 years old . . . mirrors10 Teacher: Do you think they’re like a mirror of what you’ve seen all day?11 Daniel: Yeah. . . Once I watched a movie and . . . there was this garden and this lady wanted

to save the kids . . . the movie was called ‘The Haunting.’ In the garden she saw somebodyhanging from the top, the kid . . . and she was like [drops his jaw, slaps his hands on cheeks]@@

12 Javier: @@13 Daniel: And then all the other people were like ahh! . . . And when I went to sleep I had a

nightmare. You know how flashing stuff like . . . scares you?14 Javier: Oh flashing like . . . flashing through your eyes?15 Daniel: Like shoom, shoom, that happened to me and I was like sweating!16 Javier: Ok, My turn my turn! One time, at night it was the middle of the woods or something17 Daniel: Oooooh [pretending to be scared]18 Javier: I was hiding. Here’s like the tree, I was hiding like that [with the pencil

demonstration and the eraser behind the tree] He was floating like that . . . I turned back,he said boo! I ran but I couldn’t run that fast because it was an escalator. Then some deadpeople rised up from the ground and I waked up . . . It was like one in the morning and Icould not sleep any more! It was a bad nightmare!

19 Teacher: That sounds scary. So it’s called nightmare in English. How do you say it inSpanish?

20 Daniel: Un neetmare [trying to use Spanish phonology]21 Javier: No no . . . . Sueno mal {bad dream} or something like that . . . or pesa . . . oh yeah,

::pesadilla:: [overlap with Daniel]22 Daniel: ::Pensamiento?:: {thought}23 Javier: No, pesadilla {nightmare}24 Daniel: Oh yeah, “pesos” . . . Oh I know pesa like your dıa was pesado {Oh yeah “weights” Oh I

know like “to weigh” like your “day” was “hard/heavy”}25 Javier: Dia like hard day, pesado dia like pesa . . . dıa {Day like hard day, heavy day like

nightmare}26 Daniel: Oh! . . . . I think I know what ‘mare’ means, it’s like a schmare . . . is like something

scary that doesn’t really happen, but you believe in it!27 Both: @@@ [they both laugh]28 Daniel: Something we both have in common? [reading the interview template]29 Javier: We both have scary dreams.30 Both: @@@ [they both laugh]

Melinda Martin-Beltran 265

Excerpt 2 exhibits several key qualities that wereimportant for creating a rich context for languagelearning in this interaction. First, the participantscreated a space to play with language and askquestions about language. To use Swain’s (2006)sociocultural term, they were involved in languag-ing and language play (G. Cook, 2000; Tocalli-Beller & Swain, 2007). This conversation betweenDaniel and Javier called attention to languagewhen Daniel asked a rhetorical question aboutlanguage (line 6). This practice of asking ques-tions about language was often observed in theclassrooms in which teachers and students wereencouraged to be curious about words and dif-ferences between languages. In this excerpt, theteacher advanced this curiosity by asking morequestions (lines 7 and 19). Javier affirmed thatDaniel was acting like a “scientist,” who asks goodquestions—a phrase often repeated by their teach-ers (line 8). The students were able to explore themeaning of “nightmare” further and share rich,descriptive language as they each told their sto-ries of scary nightmares (lines 11–18). Similar toExcerpt 1, this interaction in two languages ex-panded the students’ vocabulary and may haverefined academic vocabulary. For example, in-stead of limiting his vocabulary to “bad dream,”Javier was encouraged to use the more specificterm for “nightmare” and thus exposed Danielto more academic language in Spanish. Simi-lar to Excerpt 1, the new lexicon (pesadilla) wasembedded in a meaningful context, increasingthe comprehensibility of this language for bothlearners.

The teacher was an important participant inthis interaction as a co-constructor of the collab-orative dialogue. She encouraged their questionsabout language (line 7) and asked her own ques-tions, which inspired further dialogue (line 10).She became involved in the language play and en-couraged the students to expand and extend theirword analysis across language boundaries. With-out the teacher’s question comparing English andSpanish (see line 19), it is unlikely these studentswould have reached this level of crosslinguisticword analysis on their own. The teacher’s scaf-folding and inquiry about the comparison of thetwo languages opened further opportunities formetalinguistic awareness and co-construction oflanguage knowledge.

Building Off Each Other’s Expertise . The boysdemonstrated a high level of collaborative co-construction evidenced in their back-and-forthstorytelling and their animated dialogue. Theircollaborative dialogue about writing illustrated

many instances of co-constructions as defined byFoster and Ohta (2005). For example, Danieland Javier echoed and built on each other’sprevious statements as new ideas emerged, de-veloped, and changed. Their interaction dur-ing this activity was very friendly and would bedescribed as collaborative, in terms of Storch’s(2002) definitions of high equality and high mu-tuality. Both Javier and Daniel took directionsfrom each other, they contributed jointly to thecomposition, and they engaged with each other’sideas, showing high levels of cohesion (repeat-ing utterances and extending on each other’sutterances).

This excerpt also highlighted the importance ofthe social activity of learning during interactions.The playful nature of the dialogue (evidenced bylaughter, overlapping speech, ludic talk, and one-up-manship) seemed to encourage languaging.A sense of solidarity framed their interaction, asevidenced in the final lines of the transcript whenthey recognized they shared common experiencesand perspectives (“we both have scary dreams,”line 29).

Another feature of this dialogue that a sociocul-tural lens helps to highlight is the way they usedlanguage explicitly as a tool for mediation andan object for analysis. This example expands pre-vious empirical evidence that has demonstratedthe use of the L1 as a tool for mediation to an-alyze the L2 (Anton & DiCamilla, 1998). In thiscase, both languages became “objects for analysis”and “tools for mediation” during the interaction.This interaction demonstrates the complexity ofa two-way learning process that constantly movesback and forth between languages. There wereseveral instances when the use of two languagesoffered greater affordances for the students to an-alyze language. For example, in response to theteacher’s question about how to say “nightmare”in English, Daniel demonstrated his metalinguis-tic understanding as he tried to extend cognaterules (phonological transfer patterns) and triedto apply these rules (albeit incorrectly) to a newcontext to create new knowledge (see line 20). Hispartner, Javier, drew on his own linguistic fundsof knowledge (Moll, Amanti, Neff, & Gonzalez,1992) as a dominant Spanish speaker and con-tributed to this construction of knowledge byproviding authentic Spanish alternatives. Javierdemonstrated his expertise by correcting Danieland finally producing the precise word in lines21–25.

After Javier shares the word pesadilla, we ob-serve an interesting case of uptake (that went be-yond traditional definitions of uptake in terms

266 The Modern Language Journal 94 (2010)

of using new linguistic input) when the boys col-laboratively engaged in the transformation of theword. Instead of passively receiving this word orsimply incorporating this word into their next ut-terance, they deconstructed the new word andanalyzed each part of the word to further under-stand its meaning. The back-and-forth languageplay using two languages provided an importantscaffolding opportunity for David, for whom pe-sadilla was an unfamiliar word prior to this trans-formative dialogue. In lines 20–26, the boys usedboth Spanish and English as tools of mediationas they transformed the word parts pesado ‘heavy’and dıa ‘day’ into the phrase “hard day.” Danielverbalized his analysis and Javier agreed and ex-tended this analysis, which seemed to bring aboutfurther revelation for how word meanings andorigins may fit together. Finally, Daniel returnedto his original question about the English word“nightmare” and he hypothesized a new mean-ing for the word part “mare” (or “schmare”) as“something scary that doesn’t really happen, butyou believe in it” (line 26). Javier and Danielinvented their own etymology of this word thatwas not exactly accurate, but, more importantly,they went through a transformative analysis pro-cess that inspired further thinking about lan-guage and appropriation of new co-constructedknowledge.

This collaborative dialogue offered a uniquespace for linguistic creativity. In Javier andDaniel’s interaction, we observed how the stu-dents used “texts as thinking devices and re-sponded to them in such a way that new mean-ings were generated” (Wertsch, 1998, p. 115).For example, when they played with the words“nightmare” and pesadilla, not only did they gen-erate new meanings for word roots, but they alsogenerated a new definition for their inventedword “schmare.” The excerpt demonstrated theways that students drew on their resources inboth Spanish and English to solve linguisticproblems. Their dialogue could be described asthrowing a metaphorical boomerang across lan-guages, when they began their metalinguisticanalysis in language A, transferred this knowledgeto analyze language B, then returned to languageA with new insight and possibly a deeper levelof analysis that had transformed their learningprocesses.

Two Languages in Dialogue Allow for Analysis ofMultiple Meanings and Social Relationships. Ex-cerpt 3 occurred between Ignacio and Ruben,

who were of similar language backgrounds. BothIgnacio and Ruben spoke Spanish with their par-ents (who had emigrated from Mexico duringthe past decade), and they used both Englishand Spanish with their siblings. In an interview,Ruben’s mother had told me how her son of-ten acted as a translator for the family. She en-couraged her son to use well-formed Spanish,avoiding palabras acomodadas, or borrowed En-glish words incorporated into young people’sSpanish vocabulary. Ruben brought this languageawareness to school and also proved to be an ag-ile translator in the classroom. Ruben was classi-fied as a “redesignated fluent English proficient”student. He was reading at grade level in bothSpanish and English. Ignacio and Ruben bothused English and Spanish fluently during socialactivities.

In fifth grade, Ignacio had not yet tested as of-ficially English proficient, but his teachers sug-gested that his low English Language Develop-ment scores were an indicator of Ignacio’s lowliteracy skills. He was reading below grade level inboth Spanish and English. Ignacio’s mother toldme she was proud of her son’s proficiency in En-glish, but she was worried about her son’s progressin school because of behavior issues. In the subse-quent excerpt, Ignacio displayed his class-clown-like behavior when he chose to discuss a topicthat he knew may be considered inappropriatebut would appeal to a fifth-grade boy’s sense ofhumor.

Ignacio and Ruben’s interaction was collabo-rative, in terms of Storch’s (2002) definitions ofhigh equality and high mutuality. Both Ignacioand Ruben contributed jointly to the composi-tion, they engaged with each other’s ideas, andthey extended each other’s utterances. Their play-ful, animated dialogue also highlighted the highlysocial nature of their interaction where laughter,overlapping speech, and ludic talk seemed to cre-ate a context for languaging. Excerpt 3 comesfrom a conversation when Ignacio and Rubenwere writing about Ruben’s funniest moment in aletter to Ignacio’s grandfather.

As Ignacio was reading out loud what he hadwritten to Ruben about his funniest moment, theyengaged in an LRE about word appropriateness.In line 2, Ruben questioned Ignacio’s use of theword “gas,” which he had chosen as a more po-lite way to say “fart.” Ruben explained that theremight be confusion because the word “gas” hasmultiple meanings. Ruben gave evidence for an-other meaning of the word “gas” by placing it in

Melinda Martin-Beltran 267

EXCERPT 3

Original Utterance English Gloss (when needed)

1 Ignacio: ‘La cosa mas chistosa que paso fuecuando el hermano de mi amigo se echo ungas @@ en la cara de otro nino’ @@

‘The funniest thing that happened was when mybrother’s friend passed gas @@ in another boy’sface’ @@[whispering and giggling, he reads what he haswritten]

2 Ruben: ¿Un gas?? @@@ A gas?? @@@3 Ignacio: @@@ He did like this [fart noise]4 Ruben: It is gas @@ but he’s gonna think it’s

gas of CAR! Does he know the otherword . . . for gas?

5 Ignacio: @@@ Yes, pero, quedigo . . . [whispers] pedo?

Yes, but, what should I say. . . [whispers] fart?

6 Ruben: @@ ¿sabe esta palabra . . . tu abuelo? Does your grandpa know this word?7 Ignacio: Claro @@ pero . . . ¡pero es grosero! Of course@@ but . . . but it’s rude!8 Ruben: ¡Eres un nino grosero! You’re a rude boy!9 Both: @@@ [shared laughter]10 Ignacio: My grandpa knows @@ I make funny

noises @@

another context in English (line 4, “gas of car”).Then Ruben asked his partner about the letterrecipient’s familiarity with the “other” word (line4).

Ignacio demonstrated his understanding that“knowing” a word goes beyond knowing itsmeaning, but also knowing when it might beconsidered rude or inappropriate (line 7). He ex-plained that the dilemma was not simply whetherhis grandfather knew the word but whether theword would be the right word to use in this con-text because this word would be categorized asgrosero ‘rude.’ Ruben plays on Ignacio’s descrip-tive word grosero to come back to describe Ignacioas a rude user of the word (“¡Eres un nino grosero!”in line 8). By whispering and giggling, both boysdemonstrated their knowledge that pedo ‘fart’ wastaboo in their school setting and would be con-sidered impolite with some older adults, like theteacher or Ignacio’s grandfather. They discussedthe appropriateness and perception of their au-dience. In the end (line 10), Ignacio pointed outthat his grandfather was familiar with him andmay allow for a more intimate or informal kindof language that is different from school settings.As a result of this LRE, they erased the word “gas”and changed it to pedo. This change also reflectedtheir willingness to co-construct the text.

Like the other excerpts, this was another caseof dialogic interaction when the students drew ontheir two languages as resources and tools to ana-lyze language as an abstract object. These studentsdid not stick strictly to the language of instruc-tion (Spanish) but instead used both English andSpanish in their interaction. This is also an inter-esting case because—contrary to earlier studiesthat have shown how the L1 is often used as a toolto analyze the L2—these students used what mostwould call their “L2” to analyze and talk metacog-nitively about their “L1.” This is an example whereL1 and L2 are not appropriate labels for these stu-dents who have developed bilingual proficienciesin a highly bilingual context.

This interaction was one of the many exam-ples where students drew on their pragmaticand cultural knowledge learned in their multiplecommunities of practice. What is unique aboutthis excerpt and the next is the participants’ ex-plicit attention to the cultural meaning of lan-guage that went beyond their knowledge of lin-guistic systems or la langue (Agar, 1994; Saus-sure, 1959). The students in this dual immersioncontext were able to accomplish sophisticatedforms of crosslinguistic comparison not solely be-cause two languages were the topics of academicstudy. In addition to their academic and linguistic

268 The Modern Language Journal 94 (2010)

knowledge, students offered deeper understand-ings of social language, which reflected theirparticipation in very different communities ofpractice beyond the boundaries of the classroom(Lave & Wenger, 1991). These LREs went beyondwhat one might expect in a one-way context oran L2 classroom that is distanced from authentictarget language community members.

Interpreting Word Choice: Letter Greetings to MarkSocial Relationships. The dialogue in Excerpt 4demonstrates these students’ awareness that wordchoice signifies relationships between people (inthis case, between the letter recipients and the stu-dents writing the letters). The students involved inthis dialogue shared similar home languages andimmigration status—both students had been bornin the United States and spoke Spanish with theirparents, who had emigrated from Mexico 10–15

EXCERPT 4

Original Utterance English Gloss (when needed)[actions and comments in brackets]

1 Lorenzo: ‘Querido Miguel’ Dear Miguel[Reading the written letter aloud]

Estimado mejor ¿no? Better to say esteemed, no?2 Teacher: Depende de . . . ¿que crees? It depends . . . what do you think?3 Lorenzo: ¿Lo quieres mucho, mucho,

mucho?. . .Do you really, really, really love him? . . .

[asking Johnny]. . . o ‘estimado’ . . . Or [write] ‘esteemed’4 Johnny: Estimado Esteemed5 Teacher: ¿Por que? Tu crees que

estimado es mas como–Why do you think esteemed is more like . . .

6 Johnny: I don’t know.7 Lorenzo: Porque querido suena

como mas . . . Como . . .Because dear sounds more like . . . like

8 Teacher: ¿Con quien usas querido? With whom do you use ‘dear’?9 Lorenzo: Como querido . . . como con

::tu mama::Like dear . . . like with your mom

10 Johnny: :: Como con alguien que quieres?¿O amo? ::

Like with someone you like ?. . . or love?

11 Teacher: :: ¿Con tu mama?:: With your mother?12 Johnny: I like my cousin!13 Teacher: Entonces ¿Martın es un

primo querido para ti?So, Martin is a dear [close] cousin for you?

14 Lorenzo: Mejor . . . xxx Better.. xxx[decides to leave ‘querido’ on paper and goeson to reading the next sentence]

years ago. Lorenzo told me that his Spanish wasmuch better than his English and he preferredto use Spanish with his older siblings and friends.Johnny told me his English was better than hisSpanish, but he used both languages equally withfamily and friends. The nature of this interactionis more difficult to define using Storch’s (2002)four quadrants, in part because the teacher playsan intervening role. This interactive relationshipshowed qualities of low to moderate equalityand mutuality, falling between expert/novice anddominant/passive. Lorenzo seemed to assume au-thority over the task, and Johnny was ready to sim-ply accept Lorenzo’s revision without discussion(see line 4). However, Lorenzo showed highermutuality when the teacher intervened and askedthem to think about why they would choose aparticular word. Lorenzo paused (in line 3) toask Johnny for his opinion, thus encouraging his

Melinda Martin-Beltran 269

partner’s participation. However, there were alsoinstances when Lorenzo ignored Johnny and in-stead addressed the teacher. In Excerpt 4, thesestudents were rereading and revising their letterto Johnny’s cousin when they engaged in an LREto discuss the appropriateness of a letter saluta-tion.

As Lorenzo read what Johnny had written tobegin the letter, he suggested that they writeestimado rather than querido as a salutation be-cause querido10 may express feelings of tender-ness that are not appropriate for a formal letter tohis cousin. To further evaluate the word choice,Lorenzo asked Johnny how close he really felt tohis cousin (line 3), suggesting that querido in aletter would be the exception and estimado wouldbe more appropriate unless he “really liked” hiscousin. At first, Johnny quickly agreed that theyshould use estimado. When the teacher intervenedto ask why they wanted to use estimado, Lorenzoexplained that it sounded strange because queridois a word to use with one’s mother (lines 7 and 9).Johnny extended Lorenzo’s definition for queridoto include any person one cares about (line 10).On further consideration, Johnny draws on En-glish to demand attention (line 12: “I like mycousin!”), to defend his position, and possibly todraw attention to a distinction between the feel-ings of love and like that he was having troubleexplaining in Spanish. With his counterevidenceand with the teacher’s support, Johnny convincedLorenzo that his word choice of querido was ap-propriate for this context.

Although the text was not modified in this case,their disagreement over word choice presented anopportunity to analyze the significance of wordsand to consider the implications of their languagefor social relationships with their potential read-ers. The teacher acted as a key mediator in thissituation to redirect their thinking about register.Without the teacher’s intervention in this inter-action, this kind of metalinguistic conversationmay not have been possible between the boys. Al-though they brought different understanding ofregister (and range of word choice) to the sit-uation; together they created an opportunity tolearn about what register may signal and whatkind of language is appropriate in certain set-tings and relationships. Thinking about why theyselected certain words allowed them to draw onand demonstrate their cultural knowledge of lan-guage use. The discussions about word choice andappropriateness illustrate the inextricability oflanguage and the cultural beings who use the lan-guage. This dialogue demonstrates the situated

nature of languaculture knowledge (Agar, 1994;Lantolf, 2006) that shifts and only becomes mean-ingful in social relationship with others. Theseexamples suggest researchers need to give moreattention to languaculture, which recognizes theunity of language and cultural practices. Due tospace limitations of this article, the concept oflanguaculture and the importance of social dis-course will not be elaborated here but should beexplored in future research (see Martin-Beltran,2006).

In Excerpts 3 and 4 the students ex-tended their conversations to include relation-ships beyond the immediate class context tochoose their language appropriately. Studentdiscussions about sociopragmatics embodied acomplex form of metalinguistic talk, as theinterlocutors were compelled to consider sev-eral layers of linguistic and social understandingto make choices or evaluations for what soundsgood. I observed many such examples during theschool day when students engaged in discussionabout word appropriateness, register, or wordsthat were considered taboo in their community(for more examples, see Martin-Beltran, 2006).Even a brief discussion about the gravity of a“bad word” involved nuances and social signalsunderstood implicitly by participation in a com-munity with shared norms. Such knowledge of so-cial discourse or appropriateness requires interac-tion with other members of a speech communityin more intimate domains, not available in tradi-tional classrooms. This sharing of sociolinguisticcompetence and metalinguistic analysis was madepossible because students were able to draw ontheir prior experiences in which they used mul-tiple languages for authentic, social communi-cation in their communities outside of school.This dual immersion school context brought to-gether members of distinct ethnolinguistic com-munities who were given the opportunity tocontribute their unique understanding of socialdiscourse, thereby enriching opportunities forbilingual learning at school.

DISCUSSION

The findings from this study shed light onthe multiple (and sometimes unexpected) sitesfor metalinguistic analysis and potential languagelearning as students interact in dual immer-sion contexts. This study considered continu-ous, simultaneous, and reciprocal affordances forlanguage learning during interactions betweenbilingual students who used two languages to go

270 The Modern Language Journal 94 (2010)

back and forth symbiotically as mediational toolsand objects of analysis. Findings corroborate withprevious studies in which peer interaction scaf-folds language development (Mackey, 2007; Swainet al., 2002) and also reveal the importance ofteacher intervention in these interactions. As anexpert (or a more able member of society as de-scribed by Vygotsky, 1978), the teacher bringsmore advanced metalinguistic knowledge to theinteraction, and she is attuned to the learners’needs in ways that may make the interaction morefruitful for all participants.

The excerpts exemplified the kinds oflinguistic-knowledge-building dialogues that arepossible in dual immersion programs both withand without teacher intervention. Whereas Ex-cerpts 1 and 3 showed students engaging in LREswithout teacher intervention, the teacher playeda more pivotal role triggering and guiding themetalinguistic analysis in Excerpts 2 and 4. In Ex-cerpt 2, the teacher was an important participantas she advanced David and Javier’s curiosity aboutwords, and she encouraged the students to ex-pand and extend their word analysis across lan-guage boundaries. The teacher’s inquiry aboutthe comparison of the two languages opened fur-ther opportunities for metalinguistic awarenessand co-construction of language knowledge. InExcerpt 4, the teacher acted as a key mediator toencourage Lorenzo and Johnny’s thinking aboutword choice and register rather than allowing onestudent to dominate the decision making withoutdeliberation. The teacher also became a partic-ipant in Excerpt 1 when she noticed Heather’sconfusion and she encouraged Heather’s verbalproblem-solving process by asking, “What’s so con-fusing?” The excerpts also revealed examples ofmissed opportunities when teachers could haveintervened to deepen language analysis or en-sure that all students were equitably participat-ing in the interaction (e.g., the teacher missedthe opportunity to include Iliana in Heather’scrosslinguistic comparison of piso/suelo). An im-portant question for future research would be tolook more closely at the implications of teacherintervention in these dialogues. More researchis needed to examine how often students havethese kinds of LREs when the teacher does anddoes not intervene to ensure collaborative work isconstructive.

Common across all of the examples was theuse of two languages, which opened up new pos-sibilities for analyzing language. The studentsdrew on a range of bilingual resources to under-stand a new concept, which created greater accessto unfamiliar language. Using two or more lan-

guages created opportunities for multiple partic-ipants to be learning multiple languages in rele-vant moment-to-moment interactions, which havebeen neglected in previous research. For exam-ple, when Heather inserted an English word intoa Spanish phrase, this was simultaneously a bid forassistance and an opportunity for dual-languageexchange. In other words, both participants werenegotiating and appropriating new meaning intwo languages. Thus, whereas the teacher’s objec-tive was not to teach English to Iliana in that partic-ular instructional context, the students created anopportunity where Iliana would continue learningEnglish. Different from traditional, monolingualteacher talk, these interactions were concrete ex-amples of bilingual interactional spaces (Lee etal., 2008) where speakers used two languages tobuild off of each other.

By reaching outside of the “language of in-struction” into their other languages as resources,the students were able to bring back new insightinto the language of instruction, thus demonstrat-ing more profound metalinguistic awareness (i.e.,Daniel and Javier’s analysis of “nightmare”). Con-trary to earlier studies that have shown how the L1is often used as a tool to analyze the L2, this studyfound that students also used what most wouldcall their L2 to analyze and talk metacognitivelyabout their L1. The case of these heritage lan-guage learners also demonstrated that L1 and L2are not appropriate labels for learners who havedeveloped bilingual proficiencies in a highly bilin-gual context.

Using a sociocultural lens revealed ways thatyoung students and their teachers used two lan-guages concurrently as tools for mediation andanalysis during their joint activities. These splicesof interactions in the excerpts demonstrate the di-alogic, back-and-forth processes of appropriation(Rogoff, 1995). Rogoff (1995) and Wells (1999)described appropriation as a participatory processthat involves the transformation of the learners’thought processes, the transformation of the ar-tifact (in this case, the text, as it is assimilatedand reconstructed by the learners), and the trans-formation of the sociocultural context or situ-ation, which brings about change in the socialpractice. This study found that the ways that thestudents perceived the problem and participatedin problem-solving processes were transformedthroughout the collaborative dialogues. The ini-tial problem-solving process often began as an in-dividual telling of one’s story and evolved into ajoint telling as they negotiated meaning and wordchoice. Learners invited their partners into theproblem-solving process by signaling for help or

Melinda Martin-Beltran 271

asking questions. The problem of “telling a story”was further transformed into a joint problem of“choosing the appropriate words” and “express-ing ideas clearly and fluently.” In several excerpts,the problem-solving process was taken in a new di-rection as learners drew on their two languages tocompare metalinguistic qualities and appropriateusage.

The students were involved in collaborativedialogue, which Swain (2000) described as “so-cial and cognitive activity . . . linguistic problem-solving through social interaction” (p. 111) andis therefore an occasion for language learning.Findings from my study corroborate Swain’s claimthat it was not only the input or the output alonein collaborative dialogue that may promote lan-guage development but “the joint construction ofknowledge resulting from questions and replies”(2000, p. 112). These students’ questions werenot simply answered but involved complex, mul-tilayered language analysis as they drew on theirmultiple languaculture resources available in thisbilingual context. In Javier and Daniel’s interac-tion, we observed how the students used “texts asthinking devices and responded to them in such away that new meanings were generated” (Wertsch,1998, p. 115). The students in all of the excerptsengaged in talk about language and questionedtheir own language use across a wide range of lin-guistic and social categories.

The transformation of cultural artifacts (in thiscase, the written and spoken text) could be ob-served most directly in the way that the students’utterances were reconstructed by both partici-pants as they composed their final draft of theletters and as they drew on their prior knowledgeto construct the artifact in the moment. Throughthe process of questioning, writing, reading aloud,asking for opinions, and questioning again, thelanguage of the text was transformed. Their ver-balizations and writing became an object of analy-sis or “available for scrutiny” (Swain, 2000, p. 104).The students could observe the transformation ofthe text as they wrote and rewrote their spokenstories, discussing linguistic form and bringing inimplicit knowledge about social norms and com-munity practices.

The situation was transformed from an aca-demic assignment into a personal, social interac-tion, evidenced by the children’s shared laugh-ter, their playful banter, and their questions aboutrelationships with the letter recipients. For ex-ample, as Heather told her story with animatedpantomime and gesture, her participation in theinteraction went beyond the purpose of gettingthis narrative on the paper and became more

about involving her partner Iliana in the excite-ment of the moment. In the interactions betweenRuben and Ignacio and between Lorenzo andJohnny, they transformed the sociocultural con-text by extending their conversations to includerelationships beyond the immediate pair (i.e..Ignacio’s grandfather and Johnny’s cousin). Theirinteractions demonstrated that language is morethan linguistic form but also a way of mediatingoneself and relationships. One can only specu-late how these opportunities for extended per-sonal conversation between these students mayhave changed their future interactions across lan-guages and participation in new communities ofpractice.

An important part of recognizing two languagesas resources in a dual immersion setting is alsorecognizing peers as co-constructors of knowl-edge. Collaborative relationships in this schoolsetting were supported both in moment-to-moment interactions and in the larger institu-tional structures of the school (Gebhard, 1999),where different voices and different languageswere valued as important resources.

Students positioned themselves as both expertsand novices throughout the interactions as theyactively requested help from their partners andoffered advice. This study found that co-writingactivities offered several opportunities for co-construction (as defined by Foster & Ohta, 2005),which recognized the value of different speakers’contributions to the text and created affordancesfor each participant to engage their unique lan-guage expertise. For example, the LRE that be-gan with Daniel’s English expertise to analyzethe word “nightmare” developed further whenJavier contributed his Spanish expertise with theword pesadilla, and this expertise was shared asboth of them played with the two languages inter-changeably to generate new meaning (Wertsch,1998).

The students’ recognition of shared expertiseand understanding of “partner as resource” wasalso confirmed during the postactivity interviews.In the interviews, all of the students highlightedin these excerpts were able to refer back to spe-cific examples in the text of the letter to point outwhen they gave and received help thinking aboutlanguage. In their self-evaluations, the studentsrecognized collaboration and shared decisionmaking with their partner, which suggested highmutuality and equality (as described by Storch,2002). It is important to note that the relation-ships between students were built over time withina classroom culture that supports collaboration.The excerpts highlighted in this article were

272 The Modern Language Journal 94 (2010)

limited in that they could not capture the full pic-ture of these relationships and the ways they werefostered within the school context over time. Forfurther discussion of the conditions and contextsthat afford and constrain language exchange, seeMartin-Beltran (2006).

IMPLICATIONS FOR RESEARCHAND PRACTICE

This study fills a gap in educational researchby closely examining opportunities for bilinguallanguage exchange during student interactions indual immersion programs. The interactions ana-lyzed illustrate the unique opportunities for lin-guistic problem solving in a dual immersion con-text where two languages are framed (and utilizeddaily) as academic and social resources. Findingsfrom this study provide a window into the com-plexity of collaborative interactions and offer ed-ucators a new lens to think about the ways in whichstudents develop language simultaneously. Educa-tors in diverse classrooms (bilingual, heritage lan-guage, English as a second language, and main-stream) can learn from dual immersion programsthat explicitly recognize the diverse linguistic re-sources students bring to the classroom from theirdistinct communities of practice (Lave & Wenger,1991). Findings from this study suggest that teach-ers may promote language development by ac-tively creating supportive spaces where studentscan play with language, co-construct language ex-pertise, and recognize each other as resources.In this study, the teacher and researcher used ex-tensive guides for collaboration, including mod-eling and self-evaluation rubrics that placed greatimportance on collaboration within the activities.The teacher plays a key role in interactional spacesto call attention to opportunities for learning,to extend student thinking about language andto intervene or step back in order to encouragecollaborative interactions among the students. Asteachers plan instruction they need to considerways to provide guidance for co-construction oftext by explicitly valuing multiple voices and mod-eling ways that participants can draw on their dif-ferent strengths to learn together. More researchis needed to understand how teachers might opti-mize the opportunities for language learning thatcould occur in interactions.

The classroom teacher alone cannot providethe wide range of language and social experi-ences that are representative of bilingual com-munities of practice. The examples of studentdialogue involving discussion about social dis-course demonstrate why it is so important that stu-dents have meaningful interactions with other stu-

dents who are members of distinct target languagespeech communities beyond the school. The stu-dents enrich this learning context as they embody,construct, reflect, and re-create the social commu-nities from which language emerges.

The findings from this study showed how learn-ers created further opportunities for languagelearning as they broke language norms (violat-ing the language of instruction) to solve a com-municative problem. The students created theirown bilingual norms, which were more represen-tative of the multilingual communities outsideof the classroom (Levine, 2003). In contrast toearlier studies that found dual immersion pro-grams may promote “parallel monolingualism”(Fitts, 2006), this study sheds light on what learn-ers are able to accomplish when given opportuni-ties to participate in and co-construct bilingualinteractional spaces (Lee et al., 2008). I arguethat language learning affordances could be in-creased in such settings by allowing interplay be-tween languages, by expanding language bound-aries across instructional domains, and by cre-ating multiple opportunities for language learn-ers to hear their target languages embedded intheir home languages. It is also important to givemore attention to the learning opportunities thatarise when students struggle with language. Whatmay be one student’s weakness could become anopportunity to build strength in both languagessimultaneously.

In a setting where two languages were avail-able all of the time, this study demonstrated thatthe ongoing processes of acquisition in two lan-guages can also occur simultaneously. Whereasbilingual programs aim to develop bilingual com-petence, teachers within classrooms separated bylanguage tend to focus on students’ language de-velopment in only one language. Many educatorsseem to overlook that their students could be si-multaneously accomplishing their goals develop-ing two languages within student interactions intheir class.

In response to those who may critique bilin-gual programs for not providing enough instruc-tion or linguistic input in the majority language(English), this study provides evidence that En-glish learners have increased access to mean-ingful language in ways that are not availablein monolingual English settings. By participat-ing in activities that require meaningful inter-actions between learners with different languagestrengths, students create learning opportunitiesbased on their own authentic, communicativeneeds.

This study seeks to encourage future dialogueacross the fields of SLA and bilingualism. This

Melinda Martin-Beltran 273

study opens up new possibilities and contexts forthe study of SLA and the teaching of L2 learn-ers by shedding light on moments of reciprocallanguage learning. Unlike earlier studies thathave focused on LREs in one-way immersion orforeign language classrooms (Anton & DiCamil-lla, 1998; Foster & Ohta, 2005; Swain & Lapkin,1998; Swain et al., 2002), this research took placein a bilingual setting in which two languages wereutilized simultaneously as tools for mediation andobjects of analysis. The students’ dialogue alsoillustrated the complexity of bilingualism wherelanguages are not easily separated into L1 or L2(Bailey, 2000; Block, 2003; Heller, 1999; Zentella,1997). By demonstrating some (even limited) ex-pertise in each other’s languages, the studentswere able to create linguistic bridges, draw onmultiple resources to meaningfully communicate,and create new opportunities for language learn-ing. This intermingling of languages presentsanother challenge to SLA researchers who haveoften assumed the L2 as a discrete unit ofanalysis. In this unique context, students werefound to engage in crosslinguistic comparisonsand to combine strategies from two languages togenerate a creative approach to problem solving.This unique use of two languages as tools for me-diation raises further questions for research in thefield of L2 learning and urges future research tomore closely consider simultaneous and recipro-cal affordances for language learning during in-teractions.

The collaborative dialogue described in this ar-ticle does not happen automatically in dual im-mersion programs, but it is built up over timewithin a supportive classroom culture (see Martin-Beltran, 2006, for further discussion of classroomculture and school ecology). While I do not claimthat there was language learning happening ev-ery time these students were brought together tospeak with one another, I do argue that by sup-porting these encounters, teachers were creatingthe possibility that these students would come to-gether again and continue their pursuit for learn-ing and social understanding. These students andteachers were actively building the environmentin which this kind of social and language ex-change can continue to grow and transform theworld around them.

ACKNOWLEDGMENTS

I am grateful to the Spencer Foundation for fund-ing the research on which this article is based. I wish to

express my gratitude to the teachers and students whowelcomed me into their multilingual worlds. I also wishto thank Guadalupe Valdes, Ray McDermott, FranciscoRamirez, Phil Hubbard, and Stanford LEEP colleaguesfor their encouragement and support throughout myresearch. Finally, I would like to thank the journal edi-tors and the reviewers for their helpful comments. Thiswork does not necessarily represent the views of thosementioned here; I alone am responsible for any short-comings.

NOTES

1In the United States, language-minority students arealso known as English language learners previously la-beled Limited English Proficient. Majority-language stu-dents refer to those who speak English at home. Thedefinition of “language-minority student” and “nativespeaker” is a point of contention and ambiguity in botheducational research literature and SLA studies. I usethese terms here to make societal language and powerrelations explicit. In my research, I question these cat-egories and I seek definitions relevant to local context.I use the term “dual immersion,” although it should benoted that programs are also known as “Two-Way Immer-sion” or “Dual Language” (Howard, Sugarman, Chris-tian, Lindholm-Leary, & Rogers, 2007). I refer to a pro-gram that maintains the goals of bilingualism/biliteracy,academic achievement in two languages, and cross-cultural understanding between language-minority andlanguage-majority students.

2Although SLA research often uses categories of “na-tive speaker” and “nonnative speaker,” these terms areproblematic in several ways. These categories are basedon a monolingual bias, which does not adequately takeinto account bilingual communities in which peoplemay be “native speakers” of more than one languageor a variety of dialects. I will use this term when refer-ring to literature that refers to “native speakers,” but inmy own research I seek descriptors that more accuratelydescribe language experience.

3Valdes (2005) uses the term L1/L2 user , building onV. Cook’s (2002) work, to describe the complexity ofheritage students’ language experiences. I use this termto describe the variety of language backgrounds of thestudents in dual immersion programs.

4L1 refers to first language and L2 refers to secondlanguage; however, this may perpetuate the misconcep-tion that languages are always learned separately andsequentially rather than simultaneously.

5All identifying information and names have beenreplaced with pseudonyms to ensure confidentiality ofall participants.

6See the Appendix for transcription conventions.7In this example, Heather demonstrated her sensitiv-

ity for manner by choosing a more fine-grained motionverb, “to lean.” It is interesting that she applied gesture(showing with her body) to express manner similar tofindings by Negueruela, Lantolf, Jordan, and Gelabert

274 The Modern Language Journal 94 (2010)

(2004), who found that Spanish L1 speakers preferredto encode manner through gesture. This presents aninteresting question for further research to examinewhether this exposure to two “conceptual frameworksfor talking and thinking about motion events” (Lantolf,2006, p. 86) impacts bilingual students’ choice of wordsand gestures to encode manner.

8See Martin-Beltran (2006) for more discussion ofthe discursive construction of perceived proficienciesand their power of to afford or constrain participationin language learning.

9This was reported on a survey given to students about“good helpers” in Spanish and/or English.

10The literal English translation for estimado is “es-teemed” or “respected,” but this is a common salutationto begin a formal letter in Spanish and is equivalent tothe English “dear” in this context. Querido is used inmore personal and informal letters.

REFERENCES

Agar, M. (1994). Language shock: Understanding the cul-ture of conversation. New York: Quill Press.

Alanis, I. (2000). A Texas two-way bilingual program: Itseffects on linguistic and academic achievement.Bilingual Research Journal , 24, 225–248.

Angelova, M., Gunawardena, D., & Volk, D. (2006). Peerteaching and learning: Co-constructing languagein a dual language first grade. Language and Edu-cation, 20 , 173–190.

Anton, M., & DiCamilla, F. (1998). Socio-cognitive func-tions of L1 collaborative interaction in the L2 class-room. Canadian Modern Language Review, 54, 314–342.

Bailey, B. (2000). Language and negotiation of eth-nic/racial identity among Dominican Americans.Language in Society, 29 , 555–582.

Block, D. (2003). The social turn in second language ac-quisition. Washington, DC: Georgetown UniversityPress.

Brooks, F. B., & Donato, R. (1994). Vygotskyan ap-proaches to understanding foreign languagelearner discourse during communicative tasks.Hispania, 77 , 262–274.

Canagarajah, S. (2007). Lingua franca English, multilin-gual communities, and language acquisition. Mod-ern Language Journal , 91, 923–939.

Centeno-Cortes, B., & Jimenez, A. (2004). Problem solv-ing tasks in a foreign language: The importanceof the L1 in private verbal thinking. InternationalJournal of Applied Linguistics, 14, 7–35.

Christian, D., Howard, E. R., & Loeb, M. I. (2000). Bilin-gualism for all: Two-way immersion education inthe United States. Theory into Practice, 39 , 258–266.

Cohen, E. G. (1994). Restructuring the classroom: Con-ditions for productive small groups. Review of Ed-ucational Research, 64, 1–35.

Cook, G. (2000). Language play. Oxford: Oxford Uni-versity Press.

Cook, V. (1999). Going beyond the native speaker lan-guage teaching. TESOL Quarterly, 33, 185–209.

Cook, V. (2001). Using the first language in the class-room. Canadian Modern Language Review, 57 ,402–423.

Cook, V. (2002). Background of the L2 user. In V. Cook(Ed.), Portraits of the L2 user (pp. 1–28). Clevedon,UK: Multilingual Matters.

Creswell, J. W. (2007). Qualitative inquiry and researchdesign: Choosing among five approaches. ThousandOaks, CA: Sage.

de Jong, E. J. (2002). Effective bilingual education: Fromtheory to academic achievement in a two-way bilin-gual program. Bilingual Research Journal , 26 , 77–108.

DiCamilla, F. J., & Anton, M. (2004). Private speech: Astudy of language for thought in the collabora-tive interaction of language learners. InternationalJournal of Applied Linguistics, 14, 36–69.

Donato, R. (2000). Sociocultural contributions to un-derstanding the foreign and second languageclassroom. In J. Lantolf (Ed.), Sociocultural theoryand second language learning (pp. 27–50). Oxford:Oxford University Press.

Donato, R., & Lantolf, J. P. (1990). The dialogic originsof L2 monitoring. Pragmatics and Language Learn-ing , 1, 83–97.

Ellis, R. (2005). Instructed Second Language Acquisition:A Literature Review. New Zealand Ministry of Edu-cation, Research Division.

Emerson, R. M., Fretz, R. I., & Shaw, L. L. (1995). Writ-ing ethnographic fieldnotes. Chicago: University ofChicago Press.

Erickson, F. (1998). Qualitative research methods forscience education. In B. J. Fraser & K. G. Tobin(Eds.), International handbook of science education(pp. 1155–1173). Dordrecht, The Netherlands:Kluwer.

Fitts, S. (2006). Reconstructing the status quo: Linguisticinteraction in a dual-language school. BilingualResearch Journal , 30 , 337–365.

Fortune, A., & Thorp, D. (2001). Knotted and entan-gled: New light on the identification, classification,and value of language related episodes in collabo-rative output tasks. Language Awareness, 10 , 143–160.

Foster, P., & Ohta, A. S. (2005) Negotiation for mean-ing and peer assistance in second language class-rooms. Applied Linguistics, 26 , 402–430.

Freeman, R. (1998). Bilingual education and socialchange . Clevedon, UK: Multilingual Matters.

Gass, S. (1997). Input, interaction, and the second languagelearner , Mahwah, NJ: Erlbaum.

Gass, S., & Selinker, L. (2001). Second language acquisi-tion: An introductory course (2nd ed.). Mahwah, NJ:Erlbaum.

Gebhard, M. (1999) Debates in SLA Studies: RedefiningClassroom SLA as an Institutional Phenomenon.TESOL Quarterly, 33, 544–557.

Melinda Martin-Beltran 275

Genesee, F. (1985). Second language learning throughimmersion: A review of U.S. programs. Review ofEducational Research, 55, 541–561.

Hammersley, M., & Atkinson, P. (1995). Ethnography:Principles in practice (2nd ed.). London: Rout-ledge.

Haneda, M., & Wells, G. (2000). Writing in knowledge-building communities. Research in the Teaching ofEnglish, 34, 430–457.

Harris, R. (1999). Rethinking the bilingual learner. InA.Turo, C. Leung, & C. Cable (Eds.), RethinkingLanguage Education (pp. 67–81). London: CILT.

Hatch, E. (1992). Discourse and language education. Cam-bridge: Cambridge University Press.

Hayes, R. (2005). Conversation, negotiation, and theword as deed: Linguistic interaction in a duallanguage program. Linguistics and Education, 16 ,93–112.

Heller, M. (1999). Linguistic minorities and modernity: Asociolinguistic ethnography. Harlow, UK: Addison-Wesley Longman.

Howard, E. R., & Sugarman, J. (2007). Realizing the visionof two-way immersion: Fostering effective programs andclassrooms. Washington, DC/McHenry, IL: Centerfor Applied Linguistics/Delta Systems, Co.

Howard, E. R., Sugarman, J., & Christian, D. (2003).Trends in two-way immersion education: A review ofthe research (Rep. No. 63). Baltimore, MD: Centerfor Research on the Education of Students PlacedAt Risk.

Howard, E. R., Sugarman, J., Christian, D., Lindholm-Leary, K., & Rogers, D. (2007). Guiding principlesfor dual language education (2nd ed.). Washington,DC: Center for Applied Linguistics.

Keck, C. M., Iberri-Shea, G., Tracy-Ventura, N., & Wa-Mbaleka, S. (2006). Investigating the empiricallink between task-based interaction and acquisi-tion: A meta-analysis. In J. Norris & L. Ortega(Eds.), Synthesizing research on language learningand teaching (pp. 91–131). Philadelphia: Ben-jamins.

Krashen, S. D. (1985). The input hypothesis: Issues andimplications. New York: Longman.

Lantolf, J. P. (2000). Sociocultural theory and second lan-guage learning . Oxford: Oxford University Press.

Lantolf, J. P. (2006). Re(de)fining language proficiencyin light of the concept of “languaculture.” In H.Byrnes (Ed.), Advanced language learning: The con-tribution of Halliday and Vygotsky (pp. 72–91). Lon-don: Continuum.

Lapkin, S., Swain, M., & Smith, M. (2002). Reformula-tion and the learning of French pronominal verbsin a Canadian French immersion context. ModernLanguage Journal , 86 , 485–507.

Lave, J., & Wenger, E. (1991). Situated learning: Le-gitimate peripheral participation. Cambridge: Cam-bridge University Press.

Lee, J. S., Bonnet-Hill, L., & Gillespie, J. (2008). Learn-ing in two languages: Interactional space for be-coming bilingual speakers. International Journal ofBilingual Education and Bilingualism, 11, 75–94.

Leung, C., Harris, R., & Rampton, B. (1997). The ide-alized native speaker, reified ethnicities, and class-room realities. TESOL Quarterly, 31, 543–560.

Levine, G. S. (2003). Student and instructor beliefs andattitudes about target language use, first languageuse, and anxiety: Report of a questionnaire study.Modern Language Journal , 87 , 343–364.

Lindholm-Leary, K. J. (2001) Dual language education.Clevedon, UK: Multilingual Matters.

Lindholm-Leary, K. J. (2004). The rich promise of two-way immersion. Educational Leadership, 62, 56–59.

Long, M. (1996). The role of linguistic environment insecond language acquisition. In W. Ritchie & T.Bhatia (Eds.), Handbook of second language acquisi-tion (pp. 413–468). London: Academic Press.

Long, M., & Porter, P. (1985). Group work, inter-language talk, and second language acquisition.TESOL Quarterly, 19 , 207–228.

Mackey, A. (2007). Conversational interaction in secondlanguage acquisition: A series of empirical studies. Ox-ford: Oxford University Press.

Mackey, A., Gass, S., & McDonough, K. (2000). How dolearners perceive interactional feedback? Studiesin Second Language Acquisition, 22, 471–497.

Mackey, A., Oliver, R., & Leeman, J. (2003). Interac-tional input and the incorporation of feedback:An exploration of NS–NNS and NNS–NNS adultand child dyads. Language Learning , 53, 35–66.

Martin-Beltran, M. (2006) Opportunities for languageexchange between language-minority and language-majority students. Unpublished doctoral disserta-tion, Stanford University.

Martin-Beltran, M. (March, 2007) Co-constructing compe-tence and conditions for language exchange amonglanguage-minority and language-majority students.Paper prepared for the American Educational Re-search Association Conference, Chicago, IL.

Mayo, M. d. P. G. (2002). The effectiveness of two form-focused tasks in advanced EFL pedagogy. Inter-national Journal of Applied Linguistics, 12, 156–175.

Miles, M., & Huberman, A. M. (1994). Data manage-ment and analysis methods. In N. Denzin & Y. Lin-coln (Eds.), Handbook of qualitative research (pp.428–444). Thousand Oaks, CA: Sage.

Moll, L. C., Amanti, C., Neff, D., & Gonzalez, N. (1992).Funds of knowledge for teaching: Using a qualita-tive approach to connect homes and classrooms.Theory into Practice, 31, 132–141.

Motha, S. (2006). Decolonizing ESOL: Negotiating lin-guistic power in U.S. public school classrooms.Critical Inquiry in Language Studies, 3, 75–100.

Negueruela, E., Lantolf, J. P., Jordan, S. R., & Gelabert,J. (2004). The “private function” of gesture in sec-ond language speaking activity: A study of motionverbs and gesturing in English and Spanish. Inter-national Journal of Applied Linguistics, 14, 113–147.

Oliver, R. (2000). Age differences in negotiation andpairwork. Language Learning , 50 , 119–151.

Pavlenko, A. (2000). L2 influence on L1 in late bilin-gualism. Issues in Applied Linguistics, 11, 175–206.

276 The Modern Language Journal 94 (2010)

Pica, T. (1994). Research on negotiation: What does itreveal about second-language learning conditions,processes, and outcomes? Language Learning , 44,493–527.

Potowski, K. (2004). Student Spanish use and invest-ment in a dual immersion classroom: Implicationsfor second language acquisition and heritage lan-guage maintenance. Modern Language Journal , 88,75–101.

Rogoff, B. (1995). Observing sociocultural activity onthree planes: Participatory appropriation, guidedparticipation and apprenticeship. In J. Wertsch,P. Rio, & A. Alvarez (Eds.), Sociocultural studies ofthe mind (pp. 139–164). Cambridge: CambridgeUniversity Press.

Romaine, S. (1996). Bilingualism. In W. Ritchie & T.Bhatia (Eds.), Handbook of Language Acquisition(pp. 571–604). New York: Academic Press.

Saunders, W., & O’Brien, G. (2006). Oral language. InF. Genesse, K. Lindholm-Leary, W. Saunders, &D. Christian (Eds.), Educating English languagelearners: A synthesis of research evidence (pp. 14–63).New York: Cambridge University Press.

Saussure, F. (1959). Course in general linguistics (W.Baskin, Trans.). New York: The Philosophical Li-brary.

Scott, V., & de la Fuente, M. J. (2008). What’s theproblem? L2 learners’ use of the L1 duringconsciousness-raising, form-focused tasks. ModernLanguage Journal , 92, 100–113.

Storch, N. (2002) Patterns of interaction in ESL pairwork. Language Learning , 52, 119–158.

Storch, N. (2008). Metatalk in a pair work activity:Level of engagement and implications for lan-guage development. Language Awareness, 17 , 95–114.

Storch, N., & Wigglesworth, G. (2003). Is there a rolefor the use of the L1 in an L2 setting? TESOLQuarterly, 37 , 760–770.

Swain, M. (1985). Communicative competence: Someroles of comprehensible input and comprehen-sible output in its development. In S. Gass &C. Madden (Eds.), Input in second language ac-quisition (pp. 235–253). Rowley, MA: NewburyHouse.

Swain, M. (2000). The output hypothesis and beyond:Mediating acquisition through collaborative dia-logue. In J. Lantolf (Ed.), Sociocultural theory andsecond language learning (pp. 97–114). Oxford: Ox-ford University Press.

Swain, M. (2006). Languaging, agency and collabora-tion in advanced second language learning. InH. Byrnes (Ed.), Advanced language learning: Thecontribution of Halliday and Vygotsky (pp. 95–108).London: Continuum.

Swain, M., & Deters, P. (2007). “New” mainstream SLAtheory: Expanded and enriched. Modern LanguageJournal , 91, 820–836.

Swain, M., Brooks, L., & Tocalli-Beller, A. (2002). Peer–peer dialogue as a means of second languagelearning. Annual Review of Applied Linguistics, 22,171–185.

Swain, M., & Lapkin, S. (1998). Interaction and secondlanguage learning: Two adolescent French immer-sion students working together. Modern LanguageJournal , 82, 320–337.

Swain, M., & Lapkin, S. (2000). Task-based second lan-guage learning: The uses of the first language.Language Teaching Research, 4, 251–274.

Swain, M., & Lapkin, S. (2002). Talking it through: TwoFrench immersion learners’ response to reformu-lation. International Journal of Educational Research,37 , 285–304.

Thomas, W., & Collier, V. (2002). A national study ofschool effectiveness for language minority students’long-term academic achievement . Santa Cruz, CA:Center for Research on Education, Diversity & Ex-cellence.

Thorne, S. L., & Lantolf, J. P. (2007). A linguistics ofcommunicative activity. In S. Makoni & A. Pen-nycook (Eds.), Disinventing and reconstituting lan-guages (pp. 170–195). Clevedon, UK: MultilingualMatters.

Tocalli-Beller, A., & Swain, M. (2007). Riddles and punsin the ESL classroom: Adults talk to learn. In A.Mackey (Ed.), Conversational interaction in secondlanguage acquisition: Empirical studies (pp. 143–167). Oxford: Oxford University Press.

Valdes, G. (1997). Dual-language immersion programs:A cautionary note concerning the education oflanguage-minority students. Harvard EducationalReview, 67 , 391–429.

Valdes, G. (2001). Learning and note learning English:Latino students in American schools. New York:Teachers College Press.

Valdes, G. (2003). Expanding definitions of giftedness:Young interpreters of immigrant background . Mah-wah, NJ: Erlbaum.

Valdes, G. (2005). Bilingualism, heritage languagelearners, and SLA research: Opportunities lostor seized? Modern Language Journal , 89 , 410–434.

van Lier, L. (2000). From input to affordance: Social-interactive learning from an ecological perspec-tive. In J. Lantolf (Ed.), Sociocultural theory andsecond language learning (pp. 245–259). Oxford:Oxford University Press.

Vygotsky, L. (1978). Mind in society. Cambridge, MA:Harvard University Press.

Wei, L. (2000). The bilingualism reader . London: Rout-ledge.

Wells, G. (1999). Using L1 to master L2: A responseto Anton and DiCamilla’s “Socio-cognitive func-tions of L1 collaborative interaction in the L2 class-room.” Modern Language Journal , 83, 248–254.

Wertsch, J. (1998). Mind as action. Oxford: Oxford Uni-versity Press.

Yoon, B. (2008). Uninvited guests: The influence ofteachers’ roles and pedagogies on the position-ing of English language learners in the regularclassroom. American Educational Research Journal,45, 495–522.

Zentella, A. C. (1997). Growing up bilingual: Puerto Ricanchildren in New York. Malden, MA: Blackwell.

Melinda Martin-Beltran 277

APPENDIXTranscription Conventions

[square brackets] actions, comments and context@@ laughter, roughly use @ for each syllable:: double colon overlapping speech (included only where significant)dash- truncated word (sudden cutoff)“double quotes” indicates code switching (i.e., use “espanol” during English LOI or usar “English”

durante clase de espanol)‘single quote’ participants indicate written language? rising intonation (indicating question)! exclamatory intonationunderline word emphasized by speakeritalics translation/gloss of original textCAPS indicates shouting or raised volume of speakerxx unintelligible words(# seconds) pauseBold type highlighted for analytical purposes

Note. Excerpts are taken from larger transcriptions of student interactions. Line numbers are not fromoriginal transcripts. They are renumbered here.