Agency in the Making: Adult Immigrants’ Accounts of Language Learning and | Work

MILLER University of North Carolina at Charlotte Charlotte, North Carolina, United States This article considers language learner agency from a poststructuralistperspective, focusing on

Agency in the Making: Adult Immigrants’ Accounts of Language Learning and

Work

ELIZABETH R MILLER

University of North Carolina at Charlotte

Charlotte, North Carolina, United States

This article considers language learner agency from a poststructuralistperspective, focusing on how agency is discursively constituted asindividuals position themselves and are positioned as (potential) agentswithin ideologically defined spaces As such, I regard agency asinherently unstable and as a discursively mobilized capacity to act.Drawing on a corpus of 18 interviews with adult immigrant smallbusiness owners in the United States, this study uses both quantitativeand qualitative discourse analytic approaches in considering (a)recurrent linguistic constructs used across interviews to positioninterviewees as (in)agentive characters in the story worlds of theirautobiographical accounts; (b) how these constructs are mobilized inthe co-constructed positioning work of interviewer and interviewees;and (c) so-called common sense ideological discourses by which theinterviewees are constituted as agents who rationally and responsiblymake self-generated choices and act on them This multilayeredpositioning work constrained interviewees to speaking from positions

of language learner or immigrant or small business owner, but at the sametime such positioning mobilized recognizable subjectivities for them,enabling them to act in interpretable and meaningful ways

doi: 10.5054/tq.2010.226854

T he interrelationship of language, identity, and learning has beeninvestigated extensively by second language acquisition (SLA)researchers in the past decade and a half, resulting in what Block(2007) labeled as a ‘‘boom in publications linking identity and SLA’’(p 864) And, as Block notes, much of this research has adopted apoststructuralist perspective to identity construction Rather thantreating identity as a manifestation of one’s essential self, this researchemphasizes the dynamism, fragmentation, and contested nature ofidentities Relatedly, learner agency has increasingly come to beregarded as a necessary construct in understanding language learning

as well (Lantolf & Thorne, 2006; Pavlenko & Lantolf, 2000; Swain &

Deters, 2007; van Lier, 2008), and much of the poststructurally informedresearch on identity has viewed learners as able to exercise their agency

in making identity choices and in positioning themselves within and inresponse to local and larger social constraints (DaSilva Iddings & Katz,2007; McKay & Wong, 1996; Norton, 2000; Ros i Soli, 2007; Vitanova,2005) Understanding these mediating enablements and constraintsbecomes particularly consequential when considering adult immigrantlanguage learners, individuals who frequently are marginalized in thedominant society (Norton, 2000; Vitanova, 2005)

In this study, I adopt Ahearn’s (2001) ‘‘provisional definition’’ ofagency, which is the ‘‘socioculturally mediated capacity to act’’ (p 112),and I draw from van Lier’s (2008) work in second language (L2)research van Lier (2008) notes that language learners’ agentive capacity

to act runs contrary to any notion of linguistic competence as somethingone can possess; agency is instead ‘‘action potential, mediated by social,interactional, cultural, institutional and other contextual factors’’(p 171) He also cautions against treating apparent action, such asactive participation in a language classroom, as an indication of learneragency at work; one can, after all, express one’s agency by deliberatelynot acting Relatedly, Lantolf and Pavlenko (2001) contends that agencyencompasses ‘‘more than performance, or doing’’ (p 145, cited inLantolf & Thorne, 2006, p 349) It is also linked to how individuals

‘‘assign relevance and significance to things and events’’ (Lantolf &Thorne, 2006, p 143)

Aligning with a number of studies that have drawn on interviewresearch in exploring the construction of identity and agency (de Fina,2003; McKendy, 2006; O’Connor, 1994; Ros i Soli, 2007; Vitanova, 2005),

I examine here the autobiographical accounts coconstructed in views with 18 adult immigrants to the United States, as they positionedthemselves and were positioned as variously agentive I advance thenotion that the socioculturally mediated capacity to act and ability toassign relevance and significance to such acts emerge as individuals arepositioned as (potential) agents within ideologically defined spaces Thispositioning occurs, in part, as individuals ‘‘speak themselves into being’’through discursive ‘‘regularities’’ (Poynton & Lee, 2000, p 5), theinteractional and linguistically recurring ways of speaking that allow us

inter-to make sense of our selves, inter-to (re)enact those selves, and inter-to respond inter-tobeing positioned by our interlocutors Furthermore, I consider how suchpositioning can reconstitute or resist ideological views of how individualsoperate in the social world (Bamberg, 2004) Here I use the termideology not as the obfuscation of unequal power relations but as thediscursive construction of so-called common sense discourses or ways ofunderstanding and reconstituting social reality (Miller, 2009;Pennycook, 2001) Ultimately, I propose the need for us to rethink

how we understand and research agency among language learners, whenadopting a poststructuralist approach Learner agency, I argue, must beunderstood as inherently unstable and as inevitably enabled andconstrained in the ongoing co-constitution of identity and social reality.And it is this understanding that I explore as I consider how theparticipants in my study come to be constituted as agents of languagelearning and work-related actions.

FEMINIST POSTSTRUCTURALIST APPROACHES TO AGENCY

Poststructural research focusing on identity and language learning, andincluding considerations of agency, can be traced to Norton’s (1995, 2000)formative work among adult immigrant women learning English inCanada More recent scholarship has continued to advance the notion oflearner agency as constructed in the interrelationships of individuals andsocial discourses Ros i Soli (2007), for example, sees L2 users as ‘‘dynamicagents who take the initiative and take charge of their own learning’’ whilealso noting that agency is ‘‘co-constructed, both by the socioculturalenvironment and by those around the L2 user’’ (p 205) Vitanova (2005)regards the adult immigrant language learners in her study as ‘‘active,responsible and languaged sel[ves]’’ (p 153), who, through ‘‘everyday acts

of creativity’’ (p 166) are able to reestablish voices for themselves in thetarget language and culture These researchers and others work to addressthe challenge posed by Pennycook (2001) to

find a way to theorize human agency within structures of power and totheorize ways in which we think, act and behave that on the one handacknowledge our locations within social, cultural, economic, ideological,discursive frameworks but on the other hand allow us at least some possibility

of freedom of action and change (p 120)

More compellingly, Pennycook adds that agency never works ‘‘outsidesome domain of power’’ and neither is it ‘‘merely a dialecticalrelation between macro structure and micro agency’’ but it is rather ‘‘aconstant recycling of different forms of power through our everydaywords and actions’’ (p 120) Adopting such a poststructuralist approach

to agency presents an interesting theoretical and analytic challenge, for,

as Price (1996) argues, researchers often implicitly grant a priori agency

to individuals and groups, portraying learners’ agency as exercised inrelationship between ‘‘pre-given’’ subject-agents and prevailing dis-courses (p 332) Price further argues that such an approach treatsidentity positioning as the outcome of ‘‘individual capacities’’ (p 332)

rather than as co-constituted in local and larger social discourses, asadvocated in poststructural theory.

A number of feminist poststructuralists have argued that subjectivity (oridentity) and agency do not exist prior to their production in linguistic anddiscursive practices (Butler, 1990, 1993, 1997; Davies, 1991; Weedon, 1999).Importantly, they contend that one cannot achieve agency withoutsubjectivity, that is, one cannot act in ways that are deemed relevant orsignificant, unless one has a recognized identity position from which to act

On this, Butler (1993) contends that the ‘‘paradox of subjectivation’’ (p 15)

is that the constraints that are imposed in constructing a particular kind ofsubject at a given moment in time simultaneously enable that individual toact meaningfully in that interactional space At the core of Butler’s (1997)understanding of agency is her contention that anyone who ‘‘acts actsprecisely to the extent that he or she is constituted as an actor and henceoperating within a linguistic field of enabling constraints from the outset’’(p 16) Thus discursively constituted and ideologically recognizable subjectpositions such as ‘‘adult immigrant’’ or ‘‘language learner’’ or ‘‘smallbusiness owner’’ can enable individuals to act meaningfully and also resistand transform such positioning

Butler (1997) acknowledges that people operate with a ‘‘commonsense’’ perception that we and others act and make choices indepen-dently, that we can at times resist norms and alter dominant discoursesthrough our own power Butler (1990) has famously proposed thatseemingly essential identity categories such as ‘‘woman’’ or ‘‘man’’ need

to be understood as a doing, ‘‘a set of repeated acts within a highly rigidregulatory frame that congeal over time to produce the appearance ofsubstance, of a ‘natural’ kind of being’’ (p 32) Importantly, thissedimentation of acts, repeated over time, can give the appearance of abeing who is acting independently in the world because he or she is aparticular kind of individual, rather than someone who is positioned assuch in and through a web of ideological discourses regardingpersonhood and responsibility (Davies, 1991)

USING AUTOBIOGRAPHICAL ACCOUNTS IN INTERVIEW RESEARCH

An important consideration for L2 researchers is how to investigatesuch positioning work among language learners van Lier (2008) speaks

to the difficulty of ‘‘locating agency’’ (p 164), noting that relevantmediating factors of our sociocultural–historical contexts are not

‘‘immediately visible’’ in naturally occurring talk, or are, at best, only

‘‘[ambiguously] locatable’’ (pp 175–176) My aim is to investigate howthe adult immigrants in my study come to be constituted as agents, how

they seem to perceive their own capacity to act, rather than to locatecases of agency ‘‘in action.’’ I do so by examining the patterned ways theresearch participants talk about themselves and their past experiences,using both quantitative and qualitative discourse analytic approaches.Though poststructural research frequently views discourse as ‘‘post’’ anykind of structuralist understanding of language, Poynton (1993)contends that poststructuralist research cannot overlook the linguisticmeans by which subjects come to be constituted (see also Pennycook,2001) I cite Poynton (1993) at some length here, because she arguesthat poststructural analytic practice must include

critical aspects of representation, concerned particularly with questions ofthe agency of grammatical participants and the relative focus (foreground-ing/backgrounding) on those participants [which] involve highly specificgrammatical features at the level of the individual clause Such ‘‘choices’’ arenot on the whole under conscious control, so do not imply conscious volition

on the part of the individual Their habitual use within the ways of speakingcharacteristic of a particular culture carry significant meanings, however,concerning the shape of habitual and hence proper relationships within thatculture (pp 6–7)

In examining some of the ‘‘habitual ways of speaking’’ in myparticipants’ autobiographical accounts,1 I have focused on the

‘‘material phenomena’’ of language (Poynton, 1993, p 6) as it emerged

in the interview interactions I reuse familiar metalinguistic labels andconsider language units developed in structuralist approaches tolanguage However, I adopt Pennycook’s (2007) view of systematicity

in language as emerging from iterative social and linguistic activities,that is, as ‘‘the product of ritualized social performatives that becomesedimented into temporary subsystems’’ (p 110; see Makoni &Pennycook, 2007, and Pennycook 2007, for in-depth discussion oflanguage as performatively constituted) Further, I understand thesesedimented ways of positioning self and others in interview talk to beinfluenced, in part, by common sense ideologies of language learners asparticular kinds of subjects, as individuals who are believed to alreadyhave agency and thus have responsibility to act

Relatedly, I treat interviews as occasions for assembling meanings andconstructing selves, not merely eliciting reports, and understand that theknowledge generated in interviews is created from the actions taken toobtain it (Gubrium & Holstein, 2003) I thus view seemingly mundaneand familiar ways of constructing versions of self and experience to be

1 Even though the participants in this study may not have developed all of the habitual ways

of using English as found among native speakers, they appear to have appropriated some

of the normative ways to speak about self in relation to past and ongoing experiences.

co-constructed by the interviewer and interviewee, interlocutors whosimultaneously reconstitute (or resist) normative understandings of thesocial world that render these versions ‘‘sensible’’ (Miller, 2009) Inanalyzing these accounts, I want to foreground Quigley’s (2001) cautionthat ‘‘there is no need to talk about causality’’ (p 152), that is,positioning oneself as an agentive being in autobiographical talk doesnot cause one to be agentive And yet, as Quigley (2000) notes, such talkprovides the primary site for us to ‘‘organize all our agentive encounters

in the world’’ (p 189), as we make use of conventional ways of speaking

in constituting recognizable subject positions and story worlds

THE STUDY

This study draws on a corpus of 18 interviews with individuals who (a)immigrated to the United States after childhood, (b) learned English asadults or after early childhood, and (c) opened their own businesses.The interview study was initiated to investigate how adult immigrantslearn and use English outside of classroom settings and in workplacecontexts These individuals’ considerable achievement in establishingtheir own businesses in English-dominant communities motivated mydesire to focus on their linguistic strengths and the negotiationstrategies they have developed in order to survive and thrive as theyinteract with English-speaking suppliers or clients

I recruited the research participants through several liaisons,colleagues, and community members acquainted with me and theinterviewees I conducted all of the semistructured interviews atinterviewees’ places of business in order to better understand thecontexts in which they work One exception was Lou, whom I met at acoffee shop, given that he no longer owned his import–export business

at the time of the interview As shown in Table 1, there were 10 femalesand eight males, hailing from nine different countries, who owned avariety of restaurants, small shops, and other businesses The interviewslasted 24 minutes, on average, with the shortest lasting only 12 minutes,and the longest 36 minutes The interviews were fully transcribed, andthe analysis entailed working with sound files and the transcribed textssimultaneously

DATA ANALYSIS: POSITIONING FOR AGENCY

The main analytic concept used in analyzing the discourse data is that

of subject positioning (Bamberg, 2004) within and through ‘‘linguisticfield[s] of enabling constraints’’ (Butler, 1997, p 16) In analyzing mycorpus of interview data (approximately 280 pages of transcription text),

I used a combination of quantitative and qualitative analytic approaches.

As van de Mieroop (2005) demonstrated in her exploration ofinstitutional identity, integrating both approaches enables ‘‘both an in-depth view and an overview of the corpus’’ (p 108) I use quantitativeanalysis to examine the recurrent (or sedimented) linguistic constructsused by interviewees in positioning themselves as participants in theirstory worlds, the talked-about content of their autobiographical accounts(see also de Fina, 2003) I then look more closely at the turn-by-turndevelopment of co-constructed interaction in representative excerpts ofinterviewee-interviewer talk And finally, I consider how these inter-locutors orient to common sense ideologies regarding language learners

as responsible agents of their learning success

Positioning Participants in the Story World

In analyzing several of the recurrent linguistic constructions used byinterviewees, I examine how interviewees came to be constituted asagents of the actions expressed in clausal predicates (Scheibman, 2002)

in which they are named participants (typically through using thepersonal pronouns I, me, or we) I selected only those utterances in whichinterviewees directly addressed topics regarding their experiences intheir (1) early learning of English, (2) starting and running their ownbusinesses, and (3) using and continuing to learn English at work Asshown in Figure 1, the largest percentage of interviewees’ utterances,across all three topics, used constructions in which the interviewee is

TABLE 1

Interview Participants

Pseudonym Gender Country of origin Type of business owned

Lan Female Vietnam Hail salon

Kay Female Vietnam Nail salon

Jin Female Vietnam Facial shop

Tony Male Vietnam Vietnamese sandwich shop

Don Male Vietnam Restaurant supply store

Hannah Female China Chinese restaurant

Joe Male China Chinese restaurant

Keith Male China Chinese pastry shop

Lois Female Korea Nail shop

Soo Female Korea Dry cleaning pickup store

Hee Female Korea Korean restaurant

Donna Female Laos Lao dress-making shop

Dorothy Female Laos Lao music and variety store Paul Male Burma Sushi supply business

Jenny Female Brazil House cleaning business

George Male Greece Barbecue restaurants

Ivan Male France Bakery

Lou Male Italy Import/export business

positioned as the agent of action predicates Though influenced byDuranti’s (1994, 2006) explication of semantic role types,2I use the label

‘‘agent’’ here to identify occasions when interviewees constructthemselves as the subjects of predicates that position them as havingthe ‘‘capacity to act’’ (Ahearn, 2001) in some way and that position them

as having some ‘‘degree of control over their own behavior,’’ suggestingthat they ‘‘could have acted otherwise’’ (Duranti, 2006, pp 453–454).Clearly, these individuals come from varied cultural and nationalbackgrounds, with different economic and educational histories, andhave diverse goals for themselves in the United States, all of whichimpact how they perceive themselves as capable of acting in their bestinterests However, my intention here is not to show how such factorscan explain individual choices but to understand how individuals come

to be constituted as recognizable agents with incumbent responsibilities

in local and ideological discourses

As shown in Figure 1, 52.3% of the 130 subject–predicateutterances related to interviewees’ early learning of English constituteinterviewees as agents who have some control over their learning of

2 Duranti distinguished between agent and actor roles for subjects Agents affect another entity in the sentence, as in ‘‘The boy chased the dog,’’ versus actors, who act but not on an object entity, as in ‘‘The boy went to America’’ (Duranti, 1994, p 122).

FIGURE 1 Subject-predicate utterances

English (agent of action predicates) These utterance types wereproduced by all 18 interviewees when they spoke of themselves asindividuals who ‘‘go/went’’ to class, ‘‘took’’ classes, ‘‘study’’ English,

‘‘pick’’ tutors, and so on, in 68 utterances Table 2 offers illustrativeutterances of the subject–predicate linguistic constructions according

to topical category

Another frequently used construction in their early learning of Englishaccounts positioned interviewees as subjects of the verb learn (23.8%).These utterances are more ambiguous with respect to interviewees’agentive activity, tending rather to diminish the agency of the subject as

in, ‘‘so the basic I learn there’’ (Hee) or ‘‘the only I learn is uh speakingspeaking a bit more fluent’’ (Tony) Though they position themselves asundergoing change, they do not construct themselves as actively pursuingthe study of English, and thus I interpret these constructs as mitigating thenarrated character’s agency to some degree In another 5.4% of these earlylearning of English utterances, interviewees positioned themselves asobjects of others’ actions, and thus as nonagents

Considering the utterance types constructed in interviewees’ accounts,one finds variation, but it seems the preferred pattern was for interviewees

to position themselves as agentive characters who actively pursued thelearning of English early in their residence in the United States Threeindividuals, who indicated that they did not have opportunity to takeEnglish classes, still positioned themselves as agents of learning Hee, forexample, reported learning English on her own by studying the Bible with

a Korean–English dictionary She became a Jehovah’s Witness convert andtook part in weekly Bible studies In order to understand the teachings of

Early learning of English

Starting and running a business

Using and tinuing to learn English at work Participant as agent

con-of action predicates

‘‘I go school night time like two hours Tuesday and Thursday.’’

‘‘Then I open the ness.’’

busi-‘‘I use mostly English with American.’’ Participant as subject

of ‘‘learn’’ predicates

‘‘First few year um I learn speak in uh broken English with a lot of Americans.’’

‘‘I learned a little bit how you have to organize.’’

‘‘While you talk with your client you you learn a lot.’’

‘‘If I interested in like, oh

I need more of these, something like that, I have

to go.’’

‘‘I need to write oh I’m sorry I broke something.’’

the church, she describes herself this way: ‘‘I dig dig in dictionary andalways I reading aloud So I do and slowly I understand it.’’

As already alluded to, the agent of action predicate constructs werealso adopted most frequently in these individuals’ accounts of startingand running their own businesses (66.4%, see Figure 1) They spoke ofthemselves as agents who ‘‘open a shop’’ (Dorothy) or ‘‘build anothershop’’ (Jin) or ‘‘left and took barbecue business’’ (George) and whosedaily activities involve ordering, cooking, sanitizing, and so on Similarly,this construction is used in 60.6% of interviewees’ accounts regardingusing and continuing to learn English at work, such as in ‘‘we write it [inEnglish] but most of the time we order by phone’’ (Kay) To a lesserdegree interviewees also spoke of themselves as ‘‘learners’’ or as theobject or recipient of others’ efforts in their utterances relating to thesetwo topic categories

One way to account for the predominance of participants as agents

in these accounts is the nature of the interaction itself Scheibman(2002) noted that in narrative accounts, or reporting activity,interlocutors are most likely to use constructions involving a subjectwith human animacy and active predicates And yet, these intervieweesdid not always position themselves as agentive subjects, and thus theactivity of autobiographical account giving does not absolutelyconstrain interlocutors in whether or not they position themselves asagents In order to gain a more contextualized understanding of thisgrammaticized positioning, one needs to move beyond an overview ofinterviewees’ abstracted utterances and consider how the local co-constructed interaction contributes to mobilizing particular languagechoices and identity positions Due to space constraints, I examine onlyseveral excerpts in this article; however, I find they illustrate theprocesses of positioning self and being positioned that were true of allinterview accounts

Co-constructed Positioning

In Excerpt 1 below, Hannah, an owner of a Chinese restaurant whohad been in the United States nearly 20 years at the time of theinterview, positions herself as an agentive character who took action to

go to adult school in order to learn English after her arrival in theUnited States As was typical, early in the conversation I ask Hannah totalk about when she first learned English (lines 1–2).3

3 Transcription conventions used: square bracket 5 overlapping talk, question mark 5 final rising intonation, period 5 final falling intonation, comma 5 falling but continuing intonation.

Excerpt 1a

1 INT: All right My first question I ask everyone is um when did you first

2 start to learn English

3 HAN: Um when is we we learning English in China [for the students

4 INT: [Did you?

5 HAN: Yeah

6 INT: Yeah?

7 HAN: But not not talk a lot

8 INT: Yeah

9 HAN: Yeah then and I then forgot

Following my question about when she first started to ‘‘learn’’ English(1), Hannah reissues my verb choice in her response, ‘‘we we learningEnglish in China for the students’’ (3) In the next few moments of talk(not included here), she describes her experience giving birth to herfirst child soon after arriving in the United States and her inability tounderstand the hospital staff An interpreter was supposed to help her,but, as she indicates below, she ‘‘wait for the whole day, nobody willcome’’ (25) She links this experience to her recognition that sheneeded to learn English (‘‘that’s why’’ [28])

Excerpt 1b

25 HAN: And then we wait for the whole day nobody will come [you

26 know,

27 INT: [Oh no

28 HAN: ((laughter)) That’s why and then and then I I got to learn

29 some ti- learn English right?

30 INT: Yeah yeah Did you ever go to an English class?

31 HAN: Yeah And I going to adult school?

38INT: Good for you

39 HAN: Uh huh And and about ho- how many years And then I pick