BIOGRAPHIES IN TALK: A NARRATIVE-DISCURSIVE RESEARCH APPROACH
This paper demonstrates the contribution a synthetic narrative-discursive approach can make to understanding biographical work within a research interview. Our focus is on biographical work as part of the ongoing, interactive process through which identities are taken up. This is of particular interest for people who, for example, are entering a new career and can be seen as 'novices' in the sense that they are constructing and claiming a new identity. Following a discussion of the theoretical and methodological background in narrative, discourse analytic and discursive work in social psychology (e.g. Bruner, 1990; Edley, 2001; Potter and Wetherell, 1987; Wetherell, 1998), the paper presents an analysis of biographical talk from an interview study with postgraduate Art and Design students. Our interest is in their identity work, including biographical work, as novices in their fields. The analysis illustrates the approach and the key analytic concepts of, first, shared discursive resources, such as interpretative repertoires (e.g. Edley 2001) and canonical narratives (e.g. Bruner 1991), and, secondly, troubled identities (e.g. Wetherell and Edley, 1998; Taylor 2005a) . It shows how speakers' biographical accounts are shaped and constrained by the meanings which prevail within the larger society. For our participants, these include established understandings of the nature and origins of an artistic or creative identity, and the biographical trajectory associated with it. The particular focus of our approach is on how, in a speaker's reflexive work to construct a biographical narrative, the versions produced in previous tellings become a constraint and a source of continuity.
Keywords: narrative-discursive, discursive resources, identity trouble, creative identities, novice identities.
Our starting point for this paper is a concern with people's identities as complex composites of, on the one hand, who they create themselves as and present to the world, as a way of 'acting upon' it (Plummer, 2001: 4), and on the other, who that world makes them and constrains them to be. The focus of our research is biographical talk. Our assumption is that this is shaped by both the unique circumstances of people's lives and the meanings in play within the wider society and culture. These meanings include established categorisations of people and places (see e.g. Taylor, 2001), values attached to particular categories (Reynolds and Wetherell, 2003), and expected connections of sequence and consequence (Taylor, 2003). The approach we present for investigating them builds on a well-established body of work in social psychology including discourse analysis and discursive psychology (e.g. Edley, 2001; Potter and Wetherell, 1987; Wetherell and Potter,1992). Following narrative analysis in psychology (e.g. Bruner,1990; Gergen, 1994), our approach introduces an additional focus on a personal biography or narrative as a version of the speaker's ongoing identity work across different interactions. We understand this biography as a situated construction, produced for and constituted within each new occasion of talk but shaped by previously presented versions and also by understandings which prevail in the wider discursive environment, such as expectations about the appropriate trajectory of a life.
The particular contribution of the narrative-discursive approach we present is to show more exactly how this wider discursive environment is implicated in speakers' biographical talk. The materials analysed are recordings and transcripts from interviews. An interview-based research project is presumed to be a context which selects for and makes salient shared features of participants' lives. A research interview is also taken to be a situation which is recognisable to participants and one to which they can bring certain expectations. We propose that a narrative-discursive analysis of interview material can explore the commonalities in participants' biographical talk, for example of established meanings and life trajectories which are seen as typical; in addition, it can show the identity work through which these available meanings are taken up or resisted and (re-)negotiated thereby resourcing the construction of a personal identity which accommodates the particular biographical events and life situation of the speaker. The approach therefore offers a way of investigating the social nature of biographical talk.
The initial sections of the paper outline the theoretical background drawn from psychology, specifically discourse analysis, discursive psychology and narrative psychology. This includes the notions of discursive construction and rhetoric, and the speaker as active and reflexive. We present our narrative-discursive approach as a development of the synthetic discursive psychological approach outlined by Wetherell (1998). Drawing on examples from previously published studies, we then discuss the sense in which talk is understood to be social, as a preliminary to the methodological discussion in the following sections. We go on to illustrate the narrative-discursive approach using material from a new research project, Creative Journeys, conducted with postgraduate Art and Design students in 2005. Our interest in this project is in the ways that prevailing understandings shape and constrain the talk through which a speaker constructs an identity as a creative person, including a personal biography. Our discussion of the project includes our rationale for collecting data through research interviews, and a detailed description of the process through which the transcribed interview materials are analysed. We then present an analysis of the interview material to show participants' use of discursive resources, including interpretative repertoires (Edley, 2001; Wetherell, 1998) and 'trouble' (Taylor, 2005a, 2005c; Wetherell, 1998) in their work to construct a creative identity.
The THEORETICAL BACKGROUND: CONSTRUCTION, RHETORIC AND REFLEXIVITY
Our narrative-discursive approach to biographical talk assumes that talk is constitutive. This is a basic premise of discursive psychology, originating in ethnomethodology and Foucauldian theory (see Wetherell, Taylor and Yates, 2001; Potter and Wetherell, 1987 for discussions of this background). The assumption is that meanings are not the stable properties of objects in the world but are constructed, carried and modified in talk and interaction. Following from this, people's identities are also understood to be performative: constructed and enacted in their talk (Abell, Stokoe, and Billig, 2004).
A further assumption is that a speaker is active in this identity work which is an ongoing project that includes constructing a personal biography (Gergen, 1994; see also Mishler, 1999). However, identities are also social because they are resourced and constrained by larger understandings which prevail in the speaker's social and cultural context. Our interest as analysts of biographical talk is in how these larger understandings shape biographical work and their implications for how people construct accounts of both their previous experience and the possible future trajectories of their lives. In this and the next two sections, we discuss these points in turn, as a preliminary to more detailed methodological discussion and data analysis.
The constructed nature of a personal biography, as an account of the past which does identity work for a speaker is discussed by Reynolds and Taylor (2005). They suggest that single women telling their life stories structure their personal narratives to orient to an established sequence and narrative form, the 'dominant coupledom narrative'. This is the story of a life which progresses through the stages and events of coupledom, such as courting, getting married and becoming parents. This established narrative is implicitly (and sometimes explicitly) echoed and challenged by the 'women alone' in the way they tell their own stories, for example, to present an alternative progressive narrative of personal development and inner growth. The analysis considers the work which is done by the biographical talk.
This kind of 'talking against' established ideas is what Billig calls 'rhetorical' work (Billig, 1987). It suggests that talk is not just an interaction with the other person(s) present but takes place on several levels simultaneously as a speaker also responds to imagined or previously experienced audiences and criticisms. For example, an analysis of the talk of New Zealanders in Britain (Taylor and Wetherell, 1999) shows how in answering an interviewer's questions about their own lives, speakers were also taking up positions in ongoing political debates around land claims by the indigenous Maori people and contests to the established historical narratives of European colonisation and settlement.
Corresponding to this conception of talk as shaped by several (potential) interactions and contexts is the notion that identities are complex. Postmodern theorists challenged simplistic analyses which reduced people to single identity categories ('black', 'female') and ignored how fragmentary identities intersect (Rattansi, 1995). In addition, discourse analysts and discursive psychologists emphasise that identities are multiple and occasioned, with a particular focus on how people are positioned in talk (e.g. Antaki and Widdecombe, 1998) However, this kind of approach has been criticised for understating the continuity of identity (e.g. Crossley, 2000)i. Analysts of life stories (e.g. Linde, 1997; Schiffrin, 1997) consider how participants' biographical narratives are shaped in the telling, possibly over the course of an extended interaction rather than in successive turns of conversation. A narrative-discursive approach unites these concerns, looking at a biographical narrative as a situated version of previous tellings, which is constructed as part of a speaker's identity work.
A SYNTHETIC NARRATIVE-DISCURSIVE APPROACH
Our approach is 'synthetic' in the broad sense proposed by Wetherell (1998) in that a speaker is assumed to be positioned by others as having a certain identity but also actively to position her or himself; in other words, identities are both conferred and actively claimed and contestedii. However, our approach does not share the discursive psychological concern with the detail of turn-taking. Wetherell's interest is in how the subject positions are made available and taken up in the turn-by-turn of the immediate interaction. It considers a person-in-situation but not a personal biography in the sense of connections between the series of temporally-linked situations which constitute an individual's unique experience. Our argument is that an expanded, discursive and narrative focus is needed to explore the possibilities and constraints which speakers bring to an encounter from their previous identity work, or, in other words, how they are positioned by who they already are (Taylor, 2005b).
In our approach, we do not look for a narrative in a single extended stretch of talk which has perhaps been elicited by a single question (cf. Labov and Waletsky, 1967/1997; Schiffrin, 1997). Our focus is not on the kind of structure explored by Labov and Waletsky (although an analysis might also consider this: cf. Reynolds and Taylor, 2005) but in the sequential or consequential structuring of the version of a personal biography produced in a particular interaction, in this case, the research interview. This structuring appears in the details of talk, for example, in brief references to past and future (Taylor, 2005b), memories (Taylor, 2001) and the unfolding of a life story (Reynolds and Taylor, 2005). In addition, we explore the discursive resources which establish possibilities and constraints for a speaker's identity work. We share with Linde (1997) an interest in how personal narratives are in part shaped by collectively held narratives. However, our interest is not in the narratives of a bounded collective, such as those which Linde discusses in relation to institutional memory. We would also agree with her emphasis on re-telling, as in her statement that 'The life story is ...comprised ...of the most significant narratives of a speaker's life, which are told and retold, reinterpreted and reshaped for different situations' (p.283). We look at how a version of a life story functions for a speaker in a particular interaction, for example, to support a claim to an identity as a creative person, and how understandings prevailing within the wider society facilitate or constrain such identity work, for example, because of the 'trouble' a speaker may have in reconciling it with other identity claims or positionings given by her or his life circumstances.
This does not necessarily assume that speakers have met before. At the outset of any encounter a person is always already positioned. For example, her appearance and the circumstances of the meeting 'tell' something about who she is, as do any references to past life and experiences. In addition, since most people in most circumstances are not attempting to present wholly new identities or deny their pasts, the versions which have been presented in previous interactions and tellings to others become what Davies and Harrďż˝have called the 'cumulative fragments of a lived autobiography' (Davies and Harrďż˝ 1990: 49). Taylor (2005b) argues that 'these prior positionings are a constraint on a speaker's identity work because they trouble new positionings which can appear to be inconsistent with them. They establish limits to the range of identity work which can take place within any occasion of talk and thereby create continuity across occasions of talk and a likelihood that patterns will be repeated (p.48).' These patterns are explored in a narrative-discursive analysis.
Our narrative-discursive approach therefore expands discursive psychology's conception of the active speaker to include her reflections on her ongoing and previous identity work and her awareness of consistencies and contradictions. This is not to say that identity is, or can be, wholly coherent and integrated. There are inevitably differences and disjunctions, for example between who I am and who I have been, and between who I am in different roles and contexts. Some inconsistencies are tolerable. Some I am unaware of, although a new experience may draw my attention to them. However, certain inconsistencies may require explanation or repair: they create 'trouble' in identity work (Taylor, 2005a; Wetherell and Edley, 1998).
TALK AS SOCIAL
The foci of a narrative-discursive analysis are given by our assumption, following discourse analysis and discursive psychology, that talk is not as a purely individual product or expression but is social, in several senses. The first is that talk is situated. Following Billig (1987), our approach considers talk, and a speaker's identity work within that talk, as taking place in a more complex aggregate of contexts. These include the immediate interactive context, such as an interview, and the larger context invoked in rhetorical work, including what Mishler (1999) calls the 'social and cultural frameworks of interpretation' (p.18), that is, the prevailing meanings and assumptions given by the speaker's society and culture(s). This complexity of context and possible foci is the reason that a story, even an often repeated one that is part of a personal biography, will vary with the occasion of telling, as already discussed. Our analysis looks at how a biography is constructed by a speaker.
A second way in which talk is social is that speakers necessarily use a common language which includes accrued ideas and associations. These pre-exist any particular occasion of talk and can be understood as resources for it. For example, women talking about where they live refer to the established identities and associations of specific places (Taylor, 2003). They also invoke a common and recognisable narrative of a long-term family connection to a place of origin, the 'born-and-bred narrative'; this can also be referred to in a speaker's work to position herself differently, as a person who lacks such a connection to her place of residence and therefore does not belong. This kind of identity work can only make sense to speaker and hearer because of the established interconnections of meanings and associations. These discursive resources are discussed by some analysts as interpretative repertoires (Potter and Wetherell, 1987; Edley, 2001). They can also be discussed in terms of expected connections of sequence and consequence which create narrative structure and trajectories, like Bruner's 'canonical narratives' (Bruner, 1987). They are a second focus of our analysis.
Another social aspect of talk is the constraints which operate on it. There is an onus on speakers to be consistent, both with their own previous identity work, as already discussed, and also with what is more generally recognised and expected. Unexpected associations or connections become a source of 'trouble' which requires repair (Wetherell, 1998). Taylor (2005a) discusses the example of a woman who had emigrated as a child. Her identity in the new country was 'troubled', even after many years of residence, because her accent marked her as different.
METHODOLOGICAL BACKGROUND: THE CREATIVE JOURNEYS RESEARCH PROJECT
The previous sections presented the theoretical background to our narrative-discursive approach, from discourse analysis and discursive psychology. The focus of the next three sections is methodological. We will discuss data collection and analysis with reference to a current research project, Creative Journeys. The idea for this project came out of our previous research on identity and life narratives (e.g. Taylor, 2001, 2003) and the creative experience of music students (Wirtanen and Littleton, 2004). Our aim was to investigate the identity work of novices in creative fields, including biographical narratives constructed by speakers. Our participants were postgraduate Art and Design students at a prestigious London college. A postgraduate qualification is widely recognised as a professional entry pointiii to a career in this field. In choosing to undertake an expensive and demanding postgraduate course, the students could be seen to be confirming a commitment to a longer-term career in Art and Design and an associated identity as a professional in a creative field. They were therefore at a threshold point appropriate to our interest in exploring their identity work as novices in the sense that they are constructing and claiming a new identity. We suggest that it is likely that novices have special difficulty in weaving together established and personal meanings and their identity work may be particularly fragmented. Because of this, novice talk is an especially appropriate site for looking at how a creative identity is taken up (Taylor and Littleton, 2005).
Our first contact with our participants was through the college. Students were invited to an interview about creative work and creative identities, including 'the paths that people follow in art and design' and 'when, how and why people make key decisions along this journey'. Twenty nine volunteers, studying a wide range of postgraduate coursesiv, were interviewed individually by our researcher, a qualified female art therapist. The interviews were informal and followed a general list of questions about each student's work, background in art and design, influences, current life and expectations for the future. Each interview lasted about an hour. Participants were assured confidentiality and asked to sign a form giving consent for extracts from the transcripts to be used in academic papers and publications. The participants were paid a nominal sum for their time (£20).
RESEARCH INTERVIEWS AS DATA COLLECTION
Some discursive theorists (e.g. Potter and Hepburn, 2005) have criticised the use of material collected in research interviews as data. Most of the criticisms are based in the conversation analytic notion of 'naturally occurring data'v (e.g. Schegloff, 1997). This argument against interview material is broadly the same as the conventional criticism of 'leading questions', namely, that the participants would not have talked about the research topic, or talked about it in the way they did without the researcher's guidance, so the feature being studied has actually been produced by the method of data collection.
A contrary argument is that an interview is a form of interaction which is as natural as any other and is widespread in Western societies (Atkinson and Silverman, 1997). It can be argued that an interview becomes a conversation. Somewhat differently, Shakespeare (1998) suggests that the research interview has its own conventions which are familiar to participants. Further to this, we would argue that interviews are culturally rooted communication situations in which meanings are reinforced, challenged and negotiated between interlocutors in the ongoing interaction (Westcott and Littleton, 2004). Taylor (2001) suggests that a research interview not only makes certain topics salient, but participants may volunteer to be interviewed for a project because their life circumstances make a research topic (such as the importance of where you live) a current concern or special interest. This also suggests a new interpretation of a research interview, as a congenial performance context for first person narration which speakers find pleasurable (Redman, 2005; Taylor, 2005). Another defence of research interviews derives from the point made earlier about versions of talk. If it is accepted that situated talk may be a new version of what has been said before, different in detail from previous tellings rather than a wholly original, never-before-expressed innovation, then the talk produced in a research interview can be analysed as part of the ongoing project which is the speaker's identity work. This could also suggest that a research interview can be an appropriate context for speakers to rehearse new versions, making it a particularly attractive context for the novice identity work which is our interest (Taylor and Littleton, 2005).
THE ANALYTIC PROCESS: PATTERNS IN DATA
The interviews with our participants were audio-recorded and transcribed for analysis. A narrative-discursive analysis is a detailed, sometimes laborious approach. Like other discursive analytic approaches used in psychology, it is, in Wooffitt's definition, a 'bottom-up' approach in that the 'goal of analysis is to describe the organisation of actual language practices, unencumbered in the first instance by theoretically derived characterisations of their import or nature' (Wooffitt, 2005, p.154). The approach shares the classic ethnographic purpose of making the familiar strange, in this case through a process of sorting and sifting to uncover features of the data which were not necessarily apparent on an initial reading or listening. These may then be related to other writers' theorising and discussions; however, the analysis is initially data-driven and involves examining the details of talk rather than the overall story of, say, the speaker's biography.
The findings are an interpretation based not on a single interview or one participant's talk but on the analyst's immersion in the larger body of material as data, and the search for patterns across it. These patterns are discussed in terms of discursive resources. The data extracts presented in research texts will therefore be illustrative of these larger patterns rather than a presentation of all the data analysed (as is sometimes claimed for discursive psychological and conversation analytic studies). However, any pattern will also be considered within the unique context of a particular occasion of talk, for the work it does there.
There are two aspects or tasks to the analytic process. These are not 'stages' because, as in any qualitative analysis, the process is not straightforwardly sequential but inevitably iterative, although it is systematic in that it involves rigorous reading, re-reading and sorting to ensure that all the transcribed material is considered. One task for the analyst is to look for common elements which occur across different interviews and also at different points in the same interview. As already noted, these are discussed in terms of discursive resources which pre-existed an individual speaker's talk. The second task is to consider a resource within the context of a particular interview and the biographical details the speaker presents there, in order to analyse the identity work accomplished by the use of the resource and also the possible 'trouble' (Taylor, 2005a; Taylor, 2005c; Wetherell and Edley, 1998) the resource gives rise to. It can be relevant here to look at contrasting or 'deviant' cases (Silverman, 2000: 107). In a conversation analytic study, an example which does not conform to a pattern can sometimes demonstrate participants' own awareness or 'normative expectation' (Wooffitt, 2005, p.61) that such a pattern exists. In our approach, as the analysis we present will show, contrasting cases can indicate the use of an alternative resource, perhaps because it can be more easily reconciled with how the speaker is already positioned, for example, in the versions of a biography presented in previous tellings.
The narrative-discursive approach retains from psychology an emphasis on detailed analytic procedures which is perhaps less common in other social sciences. However, it is interpretive and does not claim the same complete or correct status for its findings as analyses in some other areas of psychology, including discursive psychological work which is based on conversation analysisvi. As already noted, our approach does not particularly consider the kind of extended biographical talk which is of interest to life history researchers and others (see Plummer, 2001 for an overview). Instead, it analyses the emergent biographical details and the ways that these are mobilised and harnessed to support speakers' broader ongoing identity projects.
A NARRATIVE-DISCURSIVE ANALYSIS OF BIOGRAPHICAL WORK
Following these detailed discussions of theory and methodology, we now present a narrative-discursive analysis of material from the interviews with Art and Design students. The analysis is illustrative, in that it shows how we use analytic concepts proposed by discursive and narrative psychologists: discursive resources (specifically, interpretative repertoires, following Edley, 2001: Wetherell, 1998, and canonical narratives, following Bruner, 1987) and 'trouble' in identity work (following Wetherell, 1998; Taylor, 2005a, 2005c). However, it also fulfils a larger aim, in that it shows how these speakers' biographical talk is shaped by established and recognisable ideas about a creative life and identity. The influence of the understandings which prevail in the larger social context is shown in the first analysis below. In the second, we 'correct' the common misunderstanding that discursive construction is infinitely flexible, by discussing the constraints on one speaker's talk which derive from conflicting understandings given both by the larger context and that of her family. Our analysis shows the 'trouble' in her work to construct a past and future biographical trajectory compatible with an identity of creative success.
CONSTRUCTING AN ARTIST'S BIOGRAPHY
The students were asked how they first became interested in Art and Design, and what their early experiences and influences had been. Their answers of course varied in the details but there were several detectable patterns, illustrated in the following selection of extractsvii:
I think it's probably the same old story the young age type of thing and I mean I was always making things
I always liked to draw and I was always very creative
my father was always travelling so I was a lot alone with my mum and we would sit there and draw and she would play a guitar
I come from quite a creative family although my parents don't do anything creative they you know encouraged creativity and my grandma did watercolour painting my granddad on my dad's side did cartoons he was good at drawing as well but none of them really did anything professionally with it
there are no artists or something in my family but I was always interested actually yeah
It is possible to identify a number of interpretative repertoires (Edley, 2001; Wetherell, 1998) in the interview material which these extracts illustrate. Edley (2001) defines an interpretative repertoire as 'a relatively coherent way ... of talking about objects and events in the world' (Edley, 2001: 198). The term can therefore refer to meanings which are grouped by commonsense association rather than, necessarily, logic or rational argument. The reference in Extract a to 'the same old story' indicates the speaker's awareness of the established and recognisable aspect that what is being said, and this recognisability is part of the nature of a resource. The particular resource illustrated in Extracts a, b and e is one we would label the 'prodigiousness' repertoire. It is a group of references to having an early interest and talent in art and design.
There were also many references to a family member who had creative talent and interests. Sometimes it was implied that this person had been an influence on the student taking up creative work, whether through direct encouragement or indirectly by their lived example. We refer to this as the 'creative milieu' repertoire and it can be seen here in Extracts c (in the reference to 'my mum') and d (the 'parents'). Alternatively, the family member could be cited as evidence of an inherited aptitude for art and design, as in Extract d (the grandparents): We refer to this as the 'creative inheritance' repertoire. The same repertoire can be seen in Extract e. In contrasting the claim to an early interest ('always') with the point that there were no artists in the family (in the use of 'but'), as if this is surprising, the speaker is invoking the same commonsense logic that a creative interest can be inherited.
All three repertoires (prodigiousness, creative milieu, creative inheritance) do identity work because they present, in a minimal way, the speaker's claim to be a creative person. In addition, the use of three repertoires, separately and together, is part of a different kind of pattern in the talk, the construction of a narrative or general storyline for the speakers' biographies, of development from an early starting point, possibly with an implied extension of this progression into the future. This is a story of an early start and enduring, long-term involvement, whether through interest or aptitude, which recognisably works in everyday talk to ratify a claim to be a certain kind of person, in this case someone creative or artistic. We would suggest that this is an example of an established, in Bruner's term 'canonical' narrative (1987). It does identity work for these speakers because it is part of a recognisable biography, one which might be referred to as a 'portrait of the artist as a young person'. Interestingly, a reference to an early starting point was mostly presented to support a claim to a general identity as an artist or creative person. The specific field in which the students were working (such as painting, sculpture, photography or animation) was often presented unproblematically as a relatively recent interest. Many students had changed field (for example, from sculpture to animation) or avoided positioning themselves within a single category of creative work ('I don't think I'm completely a painter I make objects as well'). This may have reflected the commitment of the art college to encourage students to work across discipline boundaries.
Why do we refer to this biographical narrative as a 'construction' rather than simply a telling of 'how it was'? We are not denying the referential quality of talk or questioning the 'truth' of these accounts in ordinary terms, as if they were presented insincerely or with an intention to deceive. However, we are interested in them as retrospective constructions, shaped in the talk for the particular purposes of the current interaction. This is a starting premise of the approach we are using, as discussed, and as such cannot be 'proven' within the data. However, it is interesting to note that some of the examples of the 'prodigiousness' repertoire are somewhat tenuous in that they are activities which would not have been special to these children but common to most of their peers, for instance, at the same schools, many of whom presumably did not go on to study Art and Design. An example can be seen in the following extract.
I guess it's something that I was always good at you know I was always told that I good at as well and yeah I drew a lot of cartoons when I was a kid .... I remember painting in play group and enjoying it and being praised
The narrative of 'artist as a young person' did not appear in all the interviews. Its absence prompted us to re-examine interview material in search of other patterns. This is an example of the use of contrasting cases to guide an analysis, as already discussed. There is more than one way to construct any identity, including as an artist and an alternative can be seen in the talk of two speakers who explicitly denied any early interest or aptitude. Here is an extract from the first of them:
I found it really difficult to paint and to draw I mean I could do like the school level ... and now I find it really difficult and somehow to paint and I didn't feel really like that I had much talent or stuff I had just worked a lot there and wanted to do some more to improve my stuff .... I was more interested in just in the idea that I could work independently it was not so much that art it was just um be by myself and work with my own things without any technique or just with my own material and somehow develop my own material
In this extract we can see a different repertoire being invoked, of art as a solitary activity ('just be by myself and work with my own things') and with the demanding project ('worked a lot') of developing something new and original ('my own material'). Again, these are recognisable associations which come together in a particular image of the artist, as someone inspired but independent and distinctive. A second speaker also invoked this image, describing a narrative of becoming an artist through being different to everyone around him. He described how he grew up in a village and from 'quite early' felt that he didn't want to stay there and become part of
a community of people that are having girlfriends and living or having a flat like in a ten kilometres area or something or quite close and I had this feeling that I didn't really fit into this
He goes on to say
I don't know if it's already creativity or but I already had this feeling that I didn't want to be like the others I think that's maybe one of the first type of creativity because I think that all my work now and the way I live as well is all led by or pushed by this idea of want I don't want to be like everybody else
These two speakers can therefore be seen to construct a claim to a different but equally recognisable identity as an artist, in this case as an independent person living an unconventional life.
Trouble in identity work
The analysis presented up to this point could be taken to suggest that talk is infinitely flexible and speakers can construct or claim any identity they want. This is a common misinterpretation of the notion of discursive construction. It misses an important point of this and related approaches, namely, the constraints on talk. Our analysis in this section shows why a speaker is not 'free' to claim any identity; her identity work is constrained. We discuss the nature of this constraint and how it appears in our analysis as 'trouble' (Taylor, 2005a, 2005c; Wetherell, 1998).
One set of constraints on the identity which a speaker can construct or claim derives from the social nature of resources. A theorist of written rather than spoken biography makes this point elegantly in a discussion of the limited range of narratives available for 'writing a woman's life' and how these act on the biographer and women themselves. Heilbrun (1988) suggests that 'women have been deprived of the narratives, or the texts, plots, or examples, by which they might assume power over their own lives' (p.17). In the terms we are using, there are no resources for certain kinds of life stories or the identities that might be associated with them. For example, Heilbrun suggests that the available narratives of women's lives emphasise 'safety and closure' rather than 'adventure, or experience, or life' (p.20), and marriage and family over work and public life as a source of fulfilment. In the terms of the narrative-discursive approach presented in this paper, the identities of either an ambitious woman or a professionally successful married woman are 'troubled' because the established associations of their different elements ('woman' and 'ambition', 'professional success' and 'married woman') are not easy to reconcile. This is not to say that these identities can never be constructed, but they are likely to need repair, such as repeated restatement and explanation to counter expectations of something different.
Heilbrun's example draws attention to the possible gendered nature of trouble (see also Taylor, 2001, 2005a, 2005c). However, this is not the focus of our current analysis (although it is a point for further investigation in our data). The point for attention in the next extract is how a speaker's identity work may be troubled because different resources cannot be reconciled. The extract is taken from an interview with a young woman who was just at the point of completing a postgraduate course. She had been successful in her career up to this point (for example, in being accepted to study at prestigious art colleges for both her undergraduate and postgraduate courses) and her discussion of her work conveyed her enthusiasm for art and her involvement. However, she also expressed uncertainty about her future commitment to Art and Design. In the interview she mentioned a member of her family who had wanted to be a painter but could not make a living and had given up art for a successful business career.
He tried to be a painter for about a year or two and he realised he couldn't make any money out of it and he just dropped it and went on to try and be a businessman
This story offered a narrative sequence in which art or creative work is attempted then put aside for success in a different kind of work. This story could seen as a local resource which is available as part of her family culture.
She also referred to a more general narrative of the established structure for a career and life narrative which shaped her family's expectations of an appropriate trajectory for her life. As she describes it
there's kind of this perception that you do your BA that's after A-levels you do your three year BA and then you get a job whereas the fact that I've done a Foundationviii and three years BA and two years MA that's already like 6 years so they were already thinking Well you'd better hurry up when are you going to get a job you're going to be 30 before you get a job
A number of theorists and researchers have challenged the assumption that a life is structured in this way, as a progression through discrete steps or stages (e.g. Mishler 1999), but the relevant point about this narrative structure is not its confirmed (in)accuracy as a description but its established and recognisable status. It is a canonical narrative (Bruner, 1987) which depicts a notion of how a life 'should' unfold, or at least one way for it to do so. Many of the art students discussed the financial and professional uncertainty of a career in Art and Design and the impossibility of achieving this kind of certain progression. (A larger point, which is outside the scope of this paper, is whether in contemporary Western society any career choice and qualification guarantees a job and a continuing progression through stages.) This speaker has reached the end of her course without a definite offer of future work, either salaried or commissioned. Her situation was not unusual among the students. However, there is conflict between the trajectory of her life so far and that expected by her parents, and this makes it difficult for her to envisage her future. In addition, the family narrative of the painter who failed weakens any claim she might make to be an artist because of its implication that this identity is fragile and inevitably illusory. The student herself presents a possible resolution of the conflict, and one favoured by her family, which is to train as a secondary school art teacher. However, she postpones this decision, at least on the occasion of the interview, saying she will probably take a year to think about it. The following extract suggests that this is an attempt to step outside the narrative her parents favour without making a definite choice of the alternative, or alternatively to reconcile the two:
I think if after a year I haven't it turns out I haven't actually done much I've just whiled away the time then I think I probably go back to the teacher training
The emphasis on going 'back' to teacher training evokes a fallback position but also the failed painter's narrative in that teaching is a link back to a career that would be sensible, as if her (successful) art work has been a diversion or time out from 'real life'. In the terms we have presented both the established cultural narrative and the local family narrative of the painter trouble her claim to an identity and a future as an artist.
There are further points for investigation in this materialix. As already noted, one of these is gender. It is noticeable how this student's situation fits with Heilbrun's analysis in that the trouble is linked to a lack of safety and closure. Another is class. The student's parents were relatively affluent and had subsidised her study of art, but this position of apparent privilege became a constraint because of both her financial dependence and their expectations about an appropriate and attainable level of financial security. These points will be explored more fully in future analyses. The particular interest here is in the conflict between the kind of open-ended trajectory which might accommodate the uncertainty of an Art and Design career and the narrative structure of her family's expectations, and the implications of this conflict for the speaker's identity work.
With its focus on the situated construction and performance of identity and personal, reflexive biographical work, the synthetic narrative-discursive approach presented in this paper offers a distinctive means of exploring both the commonalities in participants' biographical talk and the implications of prior tellings, established meanings and expected trajectories for the identity work of individual speakers. It therefore makes two distinctive contributions to the study of biographical talk over, say, more established discursive psychological approaches which have been taken up in sociology (e.g. Edley, 2001; Potter and Wetherell, 1987).
The first contribution is that the approach extends the analytic focus beyond the immediate interactive context. It acknowledges the continuity of a speaker's reflexive project to construct a biography and the links across the multiple occasioned and situated tellings of 'who I am'. In doing this, the approach addresses a common criticism of discursive approaches, that they overemphasise 'flux, variability and incoherence' and fail to recognise continuities in the narrating of a life (Crossley, 2000, p. 528).
A second contribution is that the approach extends or expands the notion of discursive resources. In addition to the broader cultural resources, such as canonical narratives and interpretative repertoires, it shows how more local resources are implicated in the construction of a personal biography. As well as the versions produced in previous tellings, these may include the resources in play within, say, the local culture of the family, as illustrated here in the family story of the failed painter which resources the young woman artist's biographical talk. Both points have implications for the nature and scope of future research, suggesting a need for work that is sensitive to the interplay of local and wider discursive resources as well as the flux and continuity in novices' biographical talk.
i See also Wetherell, 2003 for a discussion of this point.
Abell, Jackie, Elizabeth H. Stokoe and Michael Billig (2004) "Narrative and the discursive (re)construction of events" in The Uses of
Taylor, Stephanie and Karen Littleton (2006) "Biographies in talk: A narrative-discursive research approach". Qualitative Sociology
© Qualitative Sociology Review 2006