Exploring the Dynamics of Subjectivity and Power Between Researcher and Researched

Abstract: Three studies are described and examined in terms of the power dynamics created through the subjective positions made salient for both researchers and the participants by the research process. The reflexive accounts of these studies are informed by the poststructuralist critique of reflexivity as both a truthful representation of the research process and one that can be produced by stable and unitary authors. In this paper subjectivity and power are explored through the use of different narrative styles that work to highlight the contradictory and fragmented nature of reflexivity as a new construction of (a past) reality.

In the first investigation a female researcher exploring women's experiences of anger describes the process of taking analysis back to her participants to enhance the researcher's understanding of her data. Taking the approach to reflexivity as one of introspection and collaboration a single narrator tells the tale of conflict and resolution between her subjective positions of feminist-researcher, feminist and researcher. In the second study, a female researcher who interviewed men working in professional employment creates a dialogical inquiry through polyvocality to produce an account of reflexivity as social critique. In particular, she explores the subjective positions created through identities attached to her gender and her role as a researcher. The third study approaches reflexivity as discursive deconstruction and employs non-dialogical polyvocality to explore the multiple and contradictory nature of reflexive understandings created through subjective positions derived from the research experience, nationality and motherhood.

In examining the participant-researcher relationships that were enabled or dis-enabled when the researchers inhabited the subjective position of "researcher", the use of three different approaches to reflexivity with correspondingly different narrative styles, produced new understandings of subjectivity and power.

Key words: reflexivity, power, subjectivity, feminism, post structuralism, social constructionism, gender, anger, childcare, masculinities

In all relationships dominance is inescapably at stake. In every instance of talk power is being negotiated, and the researcher-researched relationship is no different. Taking a Foucauldian position that power is understood as the ability to have particular forms of knowledge or meaning ratified, we explore the possibilities of existence that were created by the subjective positions taken up in our PhD research experiences. [1]

In this paper we, Sharon, Wendy and Sarah, describe our PhD research experiences, from the benefit of at least a couple of years of hindsight. From this temporal distance we produce new reflexive accounts that examine the subjective position of "researcher" that we inhabited. We explore the researcher-researched relationships that were enabled through the interaction of our sense of selves as researchers and other subjectivities made salient in the research process. [2]

In focusing on the relationships between participants and researchers in the production of knowledge, we define reflexivity as the conscious attempt to identify how and what social understandings have been produced in the process of research. The understandings that we use when representing ourselves, our research and our participants, can be mapped onto wider social relations. Reflexivity therefore, requires the researcher to consider how the whole process of research is structured around issues of dominance, gender, sexuality, class, age, and race (BURMAN, 1990). Such a critical gaze is understood as enhancing egalitarian relations between researchers and participants (MARECEK, 1989) because reflexivity enables a process of self-awareness that can make power dynamics visible (FINLAY, 2002a). [3]

The practice of reflexivity is, however, not unproblematic. Leaving aside the need to work within an academe that constructs the validity of research through notions of objectivity which can absent the researcher's role in the research (GILL, 1998; GREED, 1990), there are also problems within the circle of those who practice reflexivity. Discussions have focused on excessive self-analysis at the expense of attending to the research participants (FINLAY, 2002a); the use of reflexivity as a tool to reinforce rather than question the author's voice (GILL, 1995); and the coercion of researchers into a confessional subject (BURMAN, 1997). [4]

However, what we consider to be a more challenging critique, and a major consideration in this paper, is the poststructuralist rejection of reflexivity a truthful representation of the research process and the "authoring" of texts by stable and unitary authors (ALVESSON & SKÖLDBERG, 2000). In contrast, post structuralism argues that our descriptions are not neutral accounts of a real world, but are constructions that can be "opened, deconstructed and analysed in ways that bring into focus the accommodation of contradiction" (PHOENIX, 1999, p.11). [5]

One approach to engaging with the tensions between reflexivity and post structuralism is to understand a reflexive account as one (of infinitely many) versions of the processes and experiences that occurred during the research. From this perspective the reflexive account is a new construction of (a past) reality, rather than some kind of reflection of the past. As a version rather than a reflection of reality, a reflexive account can be interrogated as a text and analysed for the constructions that make it meaningful. This meaning making is understood as embedded within power relations that occur in the research process. [6]

An analysis of the constructions that produce a reflexive account need to be "capable of recognising its own textuality" (ASHMORE, 1989, p.18). Self-recognising texts often work by destabilising the single authors voice and notions of consensus through the application of different narrative styles and a focus on contested and conflicting understandings. In this paper we attempt to destabilise the text through the use of three approaches to reflexivity presented in three different narrative styles. [7]

We present stories from our PhD research experiences. Looking back after these events we examine the subjective position of "researcher" and its interaction with other subjectivities made salient in the research process. To examine the utility of the poststructuralist understanding of reflexivity as a version of an account, we present our stories using three approaches to reflexivity. First, Sharon examines the conflicts created through her subjective position of feminist-researcher by using reflexivity as introspection and mutual collaboration, as she explores the impact of her own experience and personal meanings in her interactions with her participants. Second, Sarah uses reflexivity as social critique, examining the power imbalances produced through different social positions held by researchers and participants, in this case the conflicts that arose from gendered and research-participant subjectivities. Third, Wendy works with reflexivity as discursive deconstruction to analyse the positions produced by nationality and motherhood in relation to researcher-participant positions. Discursive deconstruction aims to make salient the multiple interpretations researchers have to explore their experiences, often working towards producing an account in which no single interpretation is readily available (see FINLAY, 2002b, for further details of this taxonomy). [8]

We present our reflexive stories using different narrative styles appropriate to the reflexive approaches of introspection, critique, and deconstruction that are taken. Sharon provides introspection through a first person narrative style. Sarah breaks coherence through the use of multiple voices that present different perspectives and who then enter into dialogue. Wendy also uses polyvocality, but through an analysis that avoids consensus in order to produce contested meanings, version of events that are so different that they cannot speak to each other. In presenting accounts that appropriate different narrative styles we aim to explore subjectivity and power by making salient the tensions, inconsistencies, and narratives that are produced in the telling of a past history, highlighting the "power aspects, fictions, contradictions and cracks which invariably emerge in any discourse" (ALVESSON & SKÖLDBERG, 2000, p.148). [9]

My PhD was concerned with exploring women's discourses of anger. I took a feminist and constructionist standpoint and used Q-methodology to elicit the participants' subjectivities. This methodology asks the participant to sort a number of statements along a continuum from "I agree" to "I disagree", allowing concepts to be ranked according to how important they are to the participant. Inverted factor analysis is then performed on the data, which groups participants with people who sorted the statements in a similar fashion (model Q-sorts). The resulting data is then treated as multiple subjective positions and not as taxonomies. GALLIVAN notes that, "Q-methodology is not just another rating scale, people are asked to model their own point of view (....) what emerges is the real complexity and contradictions characteristic of people's subjective views" (1996, p.2). [10]

I chose Q-methodology to examine women's experiences of anger for two reasons. First, the ability of Q-methodology to deal with subjective accounts that are diverse and contradictory made this an appropriate method to examine women's experiences of anger, given that previous research had suggested such experiences were diverse and conflictual. For example, a woman may be able to express her anger at work but not at home (FISHER, 1993; LUPTON, 1998). Second, the quantitative analysis derived from the factor analysis is not the end point, most researchers take the model Q-sorts back to participants whose own Q-sort exemplifies the model. This interviewing enables the development of the analysis, as the participants work with the researcher to explore what meanings can be drawn on to interpret the model Q-sort. [11]

With most of the participants, exploring the Q-sort was a gratifying experience, and at the time I interpreted this positive emotional experience as being derived from two sources. First, from the pleasurable validating experience that can occur when people discuss a topic that interests them and with which they can negotiate a shared understanding. Despite only two of the sixty participants identifying themselves as feminist, my own interpretations were not in conflict with their experiences and understandings. From their comments, I concluded that the participants had enjoyed this experience as much as I did. Secondly, our ability to focus on the Q-sort had co-produced an analysis that was coherent, plausible, supported by evidence, and which produced a new story, allowing me to fulfil the validation criteria of a good social constructionist analysis. [12]

The pleasure my participants and I experience through the rapport developed in these interviews validated my sense of self as a feminist-researcher. Throughout my PhD I had emotionally invested into the subjective position of feminist-researcher, which I interpreted as a researcher who studies women; empowers her participants (e.g. through the process of participating in interpretation as well as data production); and whose analysis identifies or acts to resist barriers to women's emancipation (WILKINSON, 1996). I perceived the two parts of my feminist-researcher identity to be symbiotic, mutually supportive and producing a standpoint that was conscious, ethical, and supportive of my participants. [13]

In the cases I have discussed above, I did not have great difficulty producing a coherent account of the Q-sorts exemplified by the participants. The role of the interviews had therefore been one of confirming my interpretation, adding to it and then developing this analysis in more detail and conceptual depth. For example, in one interview we discussed the social boundaries that society imposes on women when they express anger, this had been my "stand-alone" interpretation but it was developed further with the participant. Together we explored an understanding of why and how women cope with this imposition when anger is part of most women's everyday experience. In two cases, however, I had difficulty in producing a coherent concept of the model Q-sorts and the interpretations that I used to try and make sense of these Q-sorts were contrary to my feminist standpoint. This conflict produced a shift I had not anticipated, my feminist-researcher identity was being challenged. As a feminist I did not want to participate in the production of knowledge that engaged in what I understood to be patriarchal discourses that were linked to material practices that disempowered women and thus "service(ed) an oppressive system" (BORDO, 1993, p.31). However, as a researcher (aiming to get a PhD), my goal was to get as close as I could to articulating my participants' understanding of my research topic. [14]

This splitting of feminist and researcher subjective positions was an emotional and painful one. I did not know how I was going to communicate with these participants in order to come to an understanding of these model Q-sorts. I was aware that I could not force my political feminist subjective position into the analysis but yet did not want to forgo it completely. My focus had been clarity with my participants' understandings of the Q-sorts and not with attempting to understand my now potentially competing subjective positions of feminist and researcher. [15]

I give the following case, "Agnes", as an example of how my analysis lead to my feminist identity coming in conflict with my sense of self as researcher, and how I worked to engage with this tension. [16]

Agnes lived in Glasgow, was married with children, and identified herself as a mother and a community worker living in an inner city housing "scheme" (what in the US would be called a "project"). She was my contact for one group of women living in this part of Glasgow and participated in all three phases of the research: focus groups that were part of preliminary work used to develop the Q-sort statements, the Q-sort and the post Q-sort interview. [17]

The model Q-sort that Agnes exemplified positioned gender differences as salient, with men as having influence on women's experiences, while women's experiences of anger were not valued. A structural explanation for this inequality was rejected. Instead my interpretation of the analysis led me to believe that this understanding was produced through a biological discourse, given that there was agreement with the following two statements:

Men think that women are pre-menstrual when they were angry
[strongly agreed with +5]

Alone, I explored the possibility that the participants whose experiences made up this model Q-sort used their menstrual cycle as a way of diffusing or excusing their anger. My feminist standpoint positioned me in such a way that all I could see was the political leverage of biological arguments to disempower women and I did not want to uphold that particular story. Yet it seemed to be the story of these women. How was I then to have dialogue with Agnes? As a feminist I felt uncomfortable with the idea of presenting this story but as a (feminist-) researcher it was essential that I engaged in a productive dialogue with Agnes, in order for me to represent these women's views. In considering this conflict, I realised that while I may have been explicit with my participants on my feminist standpoint, I had not been explicit enough with myself. I had not explored the difficulties of taking a standpoint to which I had a deep emotional commitment, but one that was not always going to be shared by my participants. [19]

Drawing on the poststructuralist understanding of subjective positions as accounts that enable or dis-enable certain forms of knowledge and their associated social actions, I considered how my identity as feminist-researcher, had become two conflicting positions. A feminist position that did not enable dialogue (for this particular story) and a researcher position that did. I decided to experiment with situating myself solely as a researcher. [20]

The social actions associated with the subjective position of researcher were derived from a goal of increasing understanding. Since Agnes and I seemed so different, then as a researcher I could use these differences to explore what Agnes meant. I could approach this dialogue from the researcher as "professional stranger" one that needed things explained (BANISTER, 1994, p.30). I attempted to perceive the differences between Agnes and me as welcome research opportunities. With my feminist subjective position no longer blocking a dialogue with Agnes I was able to enquire in what felt like an open manner, hearing her perspective from an interested and uncritical standpoint and resisting an acceptance of some understandings as if they were natural. Instead, I used our differences to explore the forms of knowledge that I had previously taken for granted. For example, that biological discourses can be appropriated to empower women. This allowed our discussion to shift the power dynamics between Agnes and me as we moved between the positions of knower and to be known and we came to understandings that allowed us to produce a more coherent interpretation of the Q-sort Agnes exemplified. [21]

Agnes talked about how the symptoms of being pre-menstrual, such as a raised temperature, affected her ability to deal with an anger-provoking situation. Through this discussion I came to understand that Agnes used her biology, not as an excuse, but as a means of explaining why at times it is more difficult to control one's anger than at others. For Agnes, there was a difference between someone else using her menstrual cycle as an excuse not to listen to her point of view and her own understanding of how it affected her ability to cope with the events of everyday life. Thus, while both men and women may be understood as attributing women's anger to biology the understandings behind this attribution differed greatly. Agnes could incorporate her menstrual cycle into her understanding of her anger without reducing it to just her biological make-up, a position that was created when her partner said "och hey you're just pre-menstrual". [22]

By celebrating difference a consensus had been reached. Paradoxically my experimental and temporary "putting aside" of my feminist identity had enabled me to engage in feminist research. The analysis was now not just "about" Agnes, but with Agnes. Furthermore, in exploring my construction of this past event, I was able to develop my analysis by understanding the contradictory nature of women's experience of anger in terms of my participants' incorporation of the perceived power for others to legitimate their experiences. My participants' understandings of anger therefore incorporated both their own positions, and the perceived positions of significant others, positions that were not always harmonious. This analysis provided a novel understanding for the apparently contradictory and plural meanings identified in previous research on women's experience of anger. [23]

In taking a social constructionist approach to reflexivity I engaged with my research question to produce knowledge that both my participants and I felt represented their position and which further developed the psychological understanding of women's experience of anger. When my dual feminist-researcher identity was split into two opposing positions I moved to a unitary position and took up the dialogue from there. Attempting to "step out" of a subjective position with so much emotion attached to it was not easy. In hindsight, it was made easier by the fact that my researcher identity was also highly salient for me, so that (to continue the metaphor) I was not having to stand on completely new turf. The conscious choice to shift positions enabled a dialogical inquiry that showed that one of the elements in women's contradictory experiences of anger is through the use of several standpoints to produce their understandings. Being able to give up an identity (feminist) that was important to me reduced my power to define the participants' understandings, but empowered me to engage with another identity (researcher) that allowed a more consensual production of knowledge. [24]

By writing about my experience I alert the reader to the possibilities and difficulties of working collaboratively. As Wendy and Sarah discuss below, sometimes power does not lie with the obvious person i.e. the researcher. In my own experience the power dynamic moved between myself and the participant and was dependent on what was being discussed, who agreed or disagreed on a particular issue, and what joint understanding we came to. Being reflexive thus enabled me as a researcher to both explicitly explore the power dynamics in the research process between myself and the participants and provide a way of recognising myself within that process. I would like, however, to end my section with a note of caution. The narrative I provide above could be read as a successful struggle towards the goal of egalitarian research, and as such it fails to highlight that this account is focused on my understanding, my reflexivity, and my co-operation. I support FINLAY (2002a) when she argues that despite intense reflexivity, if the researcher is the sole author of the text, then as the author, the researcher should recognise that s/he continues to occupy a position of authority. [25]

In my account below I approach reflexivity as social critique and explore conflicting understandings of my PhD research experience by presenting a dialogical inquiry between three "voices". The first voice is called "PhD Sarah", she considers the interpersonal processes that were enabled by the subjective positions inhabited by her and her participants. This account seeks to uncover some of the power dynamics that occurred in both the interviewing, analysis and dissemination aspects of the research. The second voice, "Reviewer B" responds to "PhD Sarah" to produce alternative accounts of her experiences and behaviour, through which a dialogical inquiry develops, producing a new conclusion. This is followed by new challenges from the third voice, which was inspired by Ken and Mary GERGEN's argument that social constructionism should be used to produce empowering and relationship oriented understandings (e.g. GERGEN 1994; GERGEN & DAVIS, 2002). In taking such a positive standpoint the GERGENs' work resists the cynical and destructive approach that is sometimes characteristic of poststructuralist analyses (e.g. FINLAY, 2002b). To highlight this positive orientation of the GERGENs' approach in a playful way, the third entity is entitled "the fluffy Gergens". [26]

My PhD involved interviewing nearly 50 professional men about contemporary gender relations. The text books told me that if I wanted "rich" data (and who doesn't) I needed an interview technique that was person centred and that made my participants feel comfortable. To this end I opted for an informal and non adversarial interviewing style congruent with much work in my field (e.g. WETHERELL & POTTER, 1992; although see SPEER & POTTER, 2000 for an alternative interviewing technique). I used a lot of "back channel" conversational behaviour such as "um" and "yes" responses while also nodding and smiling. I tried to avoid interrupting, confrontation, or appearing judgemental. If I did suggest alternative arguments I was careful to frame them in an unchallenging way, for example by distancing myself from my statements ("some participants have argued that"). With many participants this approach helped me develop what felt like a genuine and mutual rapport, with some I had a different experience. It is these few difficult experiences I focus on below. [27]

A framework often used when theorising power in the research process is one in which the researcher is more powerful than the researched. This has been challenged by some feminist writers (e.g. WILLOTT, 1998; TAYLOR, 1996), who found the power dynamics of their research were not mapped out along an axis of researcher (powerful)—researched (unpowerful) but on gendered dimensions of male (powerful)—female (unpowerful). On occasion, I also shared this experience. Reflecting on my interviews I felt that some of these men enjoyed the power of having a young fairly attractive woman listening to their opinions and nodding, smiling and laughing at their jokes. I needed these men's co-operation for my study, but in my desire to "please" these men, I entered a set of power relations that I experienced as extremely uncomfortable. This discomfort was created through an interaction of several subjective positions, as well as "researcher" these included being social, sexual, vulnerable and manipulative. [28]

During some interviews I listened to opinions I felt strongly against without arguing against my participants. Two subjective positions worked to maintain this silence. My sense of self as researcher, who had the aim of recording these men's opinions regardless of whether I agreed with them or not; and myself as a person who enjoys harmonious social relationships, and who would not therefore be so rude as to do otherwise than listen. However, the act of silencing myself produced a sense of powerlessness, which was amplified by my only semiconscious recognition that I was also situated in a sexualised subject-position that I did not like, in that it balanced female youth and attractiveness to male knowledge and intelligence. In appearing not to disagree with ideas I disagreed with, I also became situated as a vulnerable person, I endured discomfort because my need of these participants was greater than their need of me. Finally, this act of silencing myself, created a further subjective position that stood in contradiction to "researcher" and "harmonious". In not arguing back I was producing a false self. I can on occasion be charming and courteous, it is much less in my self definition to suffer in silence. In being false, I was now also manipulative. [29]

When analysing my transcripts I drew on discourse analysis literature, feminist texts and the "crisis of masculinity" literature. Such works are not always mutually exclusive, since many take a standpoint that challenges hegemonic understandings of gender. It is not surprising, then, that while my analysis identified some of the rocks and hard places professional men experience, it's core theme was highlighting accounts that naturalised androcentrism, enabling even "liberal" men to construct a world that normalised male privilege. [30]

I wrote a summary of my findings for my participants. I wanted a fair dialogue and wrote as clearly as I could, while highlighting the less critical findings, for example, that women's opportunities to engage in the public sphere were often celebrated. I also wrote of how the changes in women's opportunities were often constructed as negative for men, and how the meaning making that supported these accounts enabled an androcentrism that worked to reduce women's opportunities without appearing discriminatory. I was once again faced with problems in communication; I was concerned that my participants would read my report as criticism of them. I attempted to deal with this problem by emphasising that my analysis focused on the construction of texts, not the identification of internal thoughts or opinions. This avenue though was also problematic; I was now "boggling" them (and myself) with post-structuralism. The problem was compounded through the academic voice I unquestioningly used to explain an academic study. Subjectivities in play thus included harmonious (but now somewhat weakened in salience by the social distance in written, as opposed to verbal, communication), academic, professional, and person struggling to be academic and professional. In not paying attention to enabling a forum for open communication my act of "feedback" thus (re-)claimed power and silenced its recipients, although invited, I received no comments. [31]

My experiences show the interplay of the many identities available to us in research and how each enables or dis-enables different forms of interaction. I shared the experience with other female researchers of being inadequately prepared for being positioned within a sexualised subjectivity and in attempting to reclaim power through alternative positions that accrue greater social power. WILLOTT (1998), for example, drew on traditional gender relations to position herself as wife to a husband, while TAYLOR (1996) and I drew on our academic identities to silence participants through control of the tape recorder or academic writing, respectively. Our accounts share two features. First, being addressed according to our sexual subjectivities "wrong footed" us and second, when we experienced a vulnerable gendered identity we did not work to re-create these gendered positions as more empowering for ourselves, but drew on alternative subjectivities that were embedded within dominant power dynamics. The female researcher becomes situated as a man's wife or within the academy. This suggests a greater focus is needed in highlighting the experience of vulnerability researchers can have through their social identities and theorising creative ways of re-working subjectivities that allow us to celebrate these positions and avoid reclaiming a position of power over the participants through other hegemonic power matrices. [32]

Well apart from an hour of your delightful company and the opportunity to participate in the kind of abusive research that I've spent more than the last decade resisting, what did your participants gain from this study? Why shouldn't you pay for your data? Seems to me that you acted like you felt like acting and now you're using vulnerability as a weapon to justify it. [33]

Seems to me that you're invalidating everything I've just said, women can and do feel vulnerable when gendered subjectivities becomes salient during research. And I have owned up to resisting this position through situating myself in other subjectivities that bestow on me the power to define. I wanted to elaborate on these processes, since they happen even in good research. After all haven't you used your powerful status as important and anonymous reviewer to construct me as manipulative and deserving of pain? [34]

Why does my "version" bother you so much, after all, from your position its only one of many narratives? Stop trying to reclaim power by taking my voice as your own and think of the applications for my interpretation. [35]

You're saying that I could understand my research as an arena in which I created my own battle of the sexes, where as researcher I always had power—to start it, stop it, set the questions, be polite because I knew how to "work" professional men (I am my daddy's girl after all), that I took their freely given words, and that I resisted their situating of me within a framework of older wise man to young pretty girl and did so by drawing on all my institutional power to produce the "real" meanings of what they had said masked as a form of intellectual textual analysis? And then had the nerves to construct the situation as a narrative of vulnerability, oppression and struggle, in which I was the victim (and hero)? [36]

Well considering this (rather uncomfortable) position in relation to subjectivity and power, what comes to the fore are the contradictions and relational nature of power. In terms of contradictions, I can understand my account as both an example of when female researchers are vulnerable and when this gendered experience is used to shore up power, in that I could temporarily accept my position as passive female because at a later date I would be able to create my participants anew through my critical analysis of their talk. In terms of the relational nature of power I can see that I had originally been constructing power as shifting between the entities of participant and researcher (they were more powerful at the interview, I at analysis). Instead I can reconceptualise power, through a more Foucauldian perspective, as working within the relationship that is made between people. In this case the participants were only able to make salient my "traditional" passive female role because I also engaged with it in a way that made it meaningful. [37]

So far you've been talking about the subjectivities used to make sense of your experience, but you also need to explore their sociohistoric context. For starters where does the understanding come from that the researcher is powerful in comparison to the researched? We could argue that it comes as a resistance to an understanding of previous research methods that constructed the participant as an object from which the researcher-scientist needed to extract information. The notion of researcher as powerful resists this understanding, and puts relationships, subjectivity and ethics as a salient concepts within the research process. As such we could see it as part of a liberal project, which created a new understanding of the "good researcher" as one who engages with their power over participants with view to creating a more egalitarian interaction. However, this ideology of the good researcher can also act to dis-enable as well as enable, in this case producing a sense of disempowerment for female researchers, both at the time of "data gathering" and in reporting their experiences, given that disclosing problems with one's participants leaves one vulnerable to accusations of being unprofessional (TAYLOR, 1996). Lets shift our thinking and move away from what looks like becoming a critique of a critique. Instead let us present you with a question. What options do you have after having an argument with someone? The first might be to try and reflect with the other person and work out what happened to cause the argument, but this might lead to further argument, over who said what and in what way; and the relationship fails to move forward positively. Instead, what if the two protagonists sat down and told each other what they liked about each other? (GERGEN & GERGEN, 2002) [38]

So far I've explored the relationships between power and subjectivity in my research by looking at the problems that occurred. Instead I could think about what worked and explore those processes? [39]

My focus on trying to explore what went wrong meant I didn't identify when differences worked. For example, one participant told me how he had been accused of verbal sexual harassment, he had had to justify himself to his boss, and apologise both to the women he had originally spoken to and to the female colleague, who had overheard his remark and made the complaint against him. Given the context in which he made these remarks I was able, through my difference to him in terms of my subjective positions of feminist and female, validate his sense that he hadn't been inappropriate, reflecting back to him a positive identity. [41]

When else did these kinds of conversations work? Don't you feel that you have to stay in the research context. [42]

Well about that time I used to occasionally go with my boyfriend to the pub with his colleagues on a Friday night and I used to get into in-depth arguments on gender relations with his boss. We often disagreed, but enjoyed our debates, partly I think because we were quite different from each other in terms of age, sex, life experiences and ideas. In this context debate thrived and I didn't mind those gender roles I critiqued above. If I had considered my conversations with "Jim", enabling a research space that allowed, if not, celebrated disagreements would have seemed possible. [43]

In using polyvocality and dialogue I have written a text that focused on some of the different narratives that could be used to understand my PhD research experience. This dialogue enabled me to produce new understandings of both the research process and my research topic. Examining different perspectives on the interactions between subjectivity and power in my research, I have understood how I used my researcher identity in favour of my more vulnerable sexualised position. In terms of understanding gender issues, I have seen how both I and my participants "did gender" and I have been able to examine the contexts in which doing these forms of gender can be experienced as empowering and disempowering. In my analysis above I have used polyvocality to create interactions between different positions. In Wendy's account below, she takes the notion of contradictory accounts a step further, and explores what we can learn from versions of events that differ so greatly that there is no discursive space for dialogue. [44]

Below I describe two studies, one conducted in Greece and the other in England. Both of the studies focused on childcare, but in different ways. The Greek research aimed to use interviews to produce quantitative data on attitudes towards childcare and is narrated through the format of two research diaries, purportedly of the same experience, two tales of one city. In the English study, interviews were used to create data that would be qualitatively analysed from a discursive approach. The research experience is narrated through four subjective positions that held salience for myself and my participants. In employing narrative styles that produce reflexive accounts that do not enter into dialogue with each other, I explore how subjectivities can simultaneously produce multiple and contradictory positions and power dynamics. [45]

Below two research diaries are presented, in which Wendy reflects on her subjective experience as a "professional" researcher who is familiar with her research field (Diary 1), and as a foreign researcher in an unfamiliar research setting (Diary 2). [46]

This job is so exciting, how great is it to be paid to travel, meet new people and study something I am really interested in? I'm so glad I did all that research before I came over. I thought it would really help to know as much about the Greek childcare system as I could before I went—but phew, it never occurred to me that the research centre I'm seconded with would have such out of date material. I know more than they do about their own country! I'm starting to make real sense too of my research methods training, all this study that seemed so academic at the time has really helped me produce a useful interview schedule. I have clarified my research questions, and now have a set of questions that can be presented in a clear and consistent manner (FRASER & BURCHELL, 2001, p.394) and will allow me to produce the data we need to develop previous work on the Greek childcare system. It's a comfortable position to be in, feeling like a competent researcher, if I didn't feel like this, all the unfamiliarity of new country might be a lot more daunting, instead I feel like a detective. SQUIRE (1990) notes that the detective narrative analogy draws attention to the researcher's understanding of him/herself as an investigator who, working "undercover" if necessary, seeks to discover the truth. And I am a good detective (although, gladly not an undercover one!), when the people at the research center were having trouble recruiting for me, I was able to get in touch with a professor at the University here. She shared a major piece of recent not yet published work with me, and between us I can write a report answering my research questions. [47]

I am so glad I met those Spanish researchers, I was starting to think that it was just me and I was doing something terrible. Hardly anyone will speak to me! People seem really irritated with me and the idea that I've come over to find out about their childcare systems. Those who talk to me tell me to study my own country, or say that while I might come from a rich country, they are the people who created western civilisation and who, today, put children before materialism by caring for them in the family. Most people just wave their hands in my face and walk off. That happened even when I asked where the coffee machine was at the research centre. Not having much Greek, I asked in English, but I've seen those people at a conference before, I know their English is as good as mine! Even my designated mentor won't help, on my second day, I showed her my research plan and after spending ten minutes asking about my personal life and why my husband had given me permission to got to Greece alone, she said something in Greek to her colleagues, they smiled and she walked away! I am really concerned about the research, I've done everything I can think of at the centre. Am trying something else now. A Dutch professor I know gave me the name of a colleague of hers in Athens, and she should be able to help. It wasn't what I had aimed for, but I can do an analysis on secondary data from a recent study and answer my questions that way. [48]

A narrators voice to set the scene: Wendy, a PhD student studying mothers' constructions of childcare, goes to interview Caroline, mother of two children aged 14 months and 3 years, at Caroline's house. Having recently given birth herself, and unable to find childcare, Wendy arrives with her researcher equipment—tape recorder, interview schedule, consent forms—and her baby—Celeste, 4 weeks old, 9 lbs. After the initial greetings, the interview starts, and so does Celeste. Caroline picks up the baby, settles her in her arms and the interview continues. But what if the tape recorder picked up Caroline and Wendy's thoughts rather than their words? If given a voice, what would the "young mother", the "experienced mother", the "researcher" and the "participant" have to say? [49]

The researcher (a position inhabited by Wendy): Back again and feeling good. Done my homework, and know a lot about the politics and theorising of childcare, you could even say I'm an expert on mothers and childcare systems in the EU! I have a good schedule, and I have participants! I've learnt from my Greek experience. I feel good about myself in terms of my professional identity—I am capable of sound theoretical and methodological research through my training and hard work; this is a position that makes me feel secure. [50]

The inexperienced mother (a position also inhabited by Wendy): This identity is new to me and I am still working through this. My little girl is one month old and I am wondering "am I a good mother?" I don't want my performance of mothering to be witnessed just yet, I'm just learning this trade! Can she tell how insecure I feel? Celly doesn't even feel like an independent being yet, she's still an extension of my body to the extent that I feel really uncomfortable when people hold her for too long. [51]

The participant (a position inhabited by Caroline): Wendy is from the University, bound to be very clever. She's Dutch, but her English is perfect. I wish I were bi-lingual. She knows a lot about childcare too. I wonder if psychologists can't help but analyse people they talk to, I wonder if she's analysing me now? I did a big tidy of the house before she came round, but now I wish I'd left some toys out, I hope she doesn't think my kids don't have enough toys to play with. I hope I don't say something silly, what if she thinks I'm a bad mum? [52]

The experienced mother (a position also inhabited by Caroline): It's true, being a mum is hard work, but it's wonderful too. Even better when they sleep through the night! I remember my first one, and how anxious I was to do the right thing, and how uncertain I was at first. It was so much easier with my second, I could really relax into the role then, having a good idea what was needed and when. Her little one's just started complaining, good, it will give me an excuse to pick her up, I love the feeling of having a baby settle in my arms. [53]

The aim of this paper was to explore the relationships between subjectivity and power through a social constructionist approach to reflexivity. A social constructionist approach to reflexivity argues that we should interrogate our reflexive accounts as texts that create versions of past events in order for us to explore the multiplicity of our experiences, celebrating contradiction as much as coherence. [54]

We have explored multiple and conflicting accounts of our research experiences in the application of differing approaches to reflexivity (introspection, social critique and deconstruction) presented in different narrative styles (single narrator, polyvocality in dialogue, polyvocality in contestation). We have used this re-telling of our research experiences to examine the power dynamics enabled through our subjective positions as researchers. DAY (2002) has explored polyvocality already in this series, using it to develop the presentation of her narrative, enriching a coherent story of her research. In our paper we have taken a post structuralist approach and used different narrative styles to develop a series of alternative textualities. These multiple texts were employed to highlight the fragmented nature of our reflexive accounts and to produce a playful dialogical inquiry on the meanings that enable these accounts. [55]

By positioning reflexive accounts as created narratives we do not seek to find a "deeper" truth, rather our validity criteria becomes the production of new ways of understanding our research, both in terms of the power dynamics enacted during the research experience and in terms of developing our understandings around research topic under question (ASHMORE, 1989; ALVESSON & SKÖLDBERG, 2000; WETHERELL & POTTER, 1992). [56]

In deliberately exploring contested versions of our research experiences we have shown how subjective positions create multiple relations of power. A researcher is not necessarily powerful, and other identities, such as gender and nationality, should be attended to. Wendy's oppositional narratives highlight the complexity of the interactions within our research. This is not to say that researchers should not consider themselves powerful and act in ways that aim to produce more egalitarian relationships, but that this aim should not reduce our theorising of a more complex power dynamic. Our analysis has for example explored how our identity as researchers was used to avoid situating ourselves within subjective positions that were understood as more vulnerable, rather than challenging the vulnerability of our sense of self as female, or new mother. In examining reflexivity as a textual account we have also explored the linguistic and cultural resources that help us create our reflexive experience, highlighting what we had previously taken for granted. For example, that a researcher inhabits a more powerful position than the researched or that biological arguments disempower women. In focusing on contested as well as consensual accounts of our experiences we have also highlighted the paradoxical nature of research, for example one might need to stop "being a feminist" to do research consistent with a feminist identity, or in Sarah's case, an understanding of how her need to do "nice" also lead to her doing "manipulative". [57]

MARSHALL (1986) suggests using the ambiguities of the research experience to develop the research process and the relationships between participants and researchers. Our accounts support this, describing the positive role of feeling uncomfortable within the research process. Sharon, for example, chooses to take what is experienced as an uncomfortable position when she rejects her feminist values in favour of communication with her participants, and Wendy's unsettling experience of having someone "take" her baby highlights the sense of power she had from positioning herself as researcher to her participants. Again, we are not saying that it is good to suffer for your research, but that in context such experiences can be used to develop thinking. [58]

In celebrating multiplicity and contradiction we have produced a set of reflexive narratives that were not spoken through a unitary or stable author. In destabilising the author and seeking contradiction we were able to produce different narratives about our research experience, creating novel reflexive accounts. Two aspects of the constructionist approach that we took to produce multiple accounts created these novel interpretations. First, the act of attempting to take on more than one standpoint leads logically to more than one understanding. Second, the framework of positioning these understandings as one interpretation amongst many acted to reduce the "face threatening" (DRYDEN, 1999) nature of accounts that were more troublesome to hear, creating a freedom to explore such accounts. As GERGEN (2001) contends, in the Western tradition argumentation is understood within a war metaphor, for example, one's position is defended and attacked. This metaphor leads to one's standpoint becoming almost indistinguishable from oneself, so that "(t)o criticize another's views is not, then, a mere linguistic exercise ... it is to invalidate the originary essence of the self" (p.49). We used polyvocality as a tool to separate the self from discourse, so that we could engage with contradictory understandings about our experiences, and learn from them. [59]

For example, Sharon came to understand that her feminist identity could reduce her ability to celebrate women's experiences through, paradoxically, its emphasis on female emancipation within gender relations. She could do this by keeping herself open to an understanding of feminism as non-emancipatory, even though she had great emotional and intellectual attachment to an understanding of a feminist identity as empowering. Allowing herself to be open to such alternatives was made possible because examining the validity of such alternatives was understood in terms of its utility in exploring her research process, not in whether it was the "true" explanation for her experiences. Sarah gained a new understanding of how her subjectivity derived from being female acted to position her as powerful, as well as vulnerable. Taking a multiple perspective enabled her to feel that her experience of being dis-empowered through gender relations in her research experience could be validated, while simultaneously understanding her active role in participating in such gendered dynamics. Similarly, Wendy was able to explore the power dynamics within her research studies, for example in exploring the attraction of being "a good researcher" in comparison to the discursive space available to both her and her participants as "mothers". [60]

The social constructionist approach we took enabled us not only to explore novel reflexive understandings of our research experiences, but also new understandings of our research topics themselves. Sharon heard that anger is a complex and sometimes paradoxical emotion that ironically mirrored her own contradictory experience in the research encounter. Sarah, researching gender, understood the experience of speaking through (or of being spoken through) her gendered subjective position; and Wendy, studying mothers, gained an insight into the difficulties in negotiating a safe discursive space from which to understand oneself as a mother. [61]

FINLAY (2002b) critiques poststructuralist reflexive accounts in terms of their potential to be so critical as to become nihilistic and lose meaning in relation to the research context. In this paper we aim to have answered this critique by producing new meanings and understandings of subjectivity and power through a poststructuralist approach that worked to explore plurality and contradiction in reflexive narratives. [62]

Alvesson, Mats & Sköldberg, Kay (2000). Reflexive Methodology: New Vistas for Qualitative Research. London: Sage Publications.

Ashmore, Malcolm (1989). The Reflexive Thesis: Wrighting Sociology of Scientific Knowledge. London: The University of Chicago Press.

Banister, Peter (1994). Observation. In Peter Banister, Erica Burman, Ian Parker, Mary Taylor, & Carol Tindall, Qualitative Methods in Psychology (pp.17-33). Buckingham: Open University Press.

Bordo, Susan. (1993). Unbearable Weight: Feminism, Western Culture and the Body. Berkely, CA: University of California Press.

Burman, Erica (1990). Differing with deconstruction: A feminist critique. In Ian Parker & John Shotter (Eds.), Deconstructing social psychology (pp.208-221). London: Routledge.

Burman, Erica (1997). Power, Research and Problems with Discursive Approaches. Paper presented at Discursive Practice: Qualitative Developments in Psychological Research Workshop, Bolton Institute of Higher Education, 16-18 April.

Day, Eileen (2002, September). Me, My*self and I: Personal and Professional Re-Constructions in Ethnographic Research [59 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research [On-line Journal], 3(3). Available at: http://www.qualitative-research.net/fqs-texte/3-02/3-02day-e.htm [Date of Access: 02, 24, 2003].

Dryden, Caroline (1999). Being Married, Doing Gender: A Critical Analysis of Gender Relationships in Marriage. London: Routledge.

Finlay, Linda (2002a). "Outing" the researcher: The provenance, process, and practice of reflexivity. Qualitative Health Researcher, 12(4), 531-545.

Finlay, Linda (2002b). Negotiating the swamp: The opportunity and challenge of reflexivity in research practice. Qualitative Research, 2(2), 209-230.

Fraser, Colin & Burchell, Brendan (2001). Introducing Social Psychology. Cambridge: Polity Press.

Gallivan, J. (1996). Gender and humour: What makes a difference? Paper presented at Canadian Women's Studies Association, St. Catherine's, Ontario, May 1996.

Gergen, Kenneth J. (1994). Realities and Relationships: Soundings in Social Constructionism. Cambridge, M.A.: Harvard University Press.

Gergen, Kenneth J. & Gergen, Mary (2002). Dialogues between action research & social constructionist perspectives. Paper presented at the Emerging Approaches to Inquiry 9, Stroud, UK, 18-21st September.

Gergen, Mary & Davis, Sara M., (1997). Toward a New Psychology of Gender. London: Routledge.

Gill, Rosalind (1995). Relativism, reflexivity and politics: Interrogating discourse analysis from a feminist perspective. In Sue Wilkinson & Celia Kitzinger (Eds.), Feminism and Discourse: Psychological Perspectives (pp.165-187). London: Sage.

Gill, Rosalind (1998). Dialogues and differences: Writing, reflexivity and the crisis of representation. In Karen Henwood, Chris Griffin & Anne Phoenix (Eds.), Standpoints and Differences: Essays in the Practice of Feminist Psychology (pp.18-43). London: Sage.

Greed, Clara (1990). The Professional and the Personal. A study of women quantity surveyors. In Liz Stanley (Ed.), Feminist Praxis. Research Theory and Epistemology in Feminist Sociology (pp.145-159). London: Routledge.

Marecek, Jeanne (1989). Gender, politics and psychology's ways of knowing, American Psychologist, 50, 162-3.

Marshall, Judy (1986). Exploring the experiences of women managers: Towards qualitative rigor. In Sue Wilkinson (Ed.), Feminist Social Psychology: Developing theory and practice (pp.193-211). Milton Keynes: Open University Press.

Speer, Susan & Potter, Jonathan (2000). The management of heterosexist talk: conversational resources and prejudiced claims. Discourse & Society, 11(4), 543-572.

Squire, Corinne (1990). Crisis what crisis? Discourses and narratives of the "social" in social psychology. In Ian Parker & John Shotter (Eds.), Deconstructing Social Psychology (pp.33-46). London: Routledge.

Taylor, Karen (1996). Keeping mum: The paradoxes of gendered power relations in interviewing. In Erica Burman, Pam Alldred, Catherine Bewley, Brenda Goldberg, Colleen Heenan, Deborah Marks, Jane Marshall, Karen Taylor, Robina Ullah & Sam Warner (Eds.), Challenging Women. Psychology's Exclusions, feminist possibilities (pp.106-125). Buckingham: Open University Press.

Wetherell, Margaret & Potter, Jonathan (1992). Mapping the Language of Racism. London: Harvester Wheatsheaf.

Wilkinson, Sue (1996) Feminist Social Psychologies: International Perspectives. Buckingham: Open University Press.

Willott, Sara (1998). An outsider within: A feminist doing research with men. In Karen Henwood, Chris Griffin & Anne Phoenix (Eds.), Standpoints and Differences: Essays in the Practice of Feminist Psychology (pp.174-190). London: Sage.

Sarah RILEY is a lecturer at the University of Bath, with research interests in discourse analysis, gender, embodiment and recreational drug use.

Wendy SCHOUTEN is a PhD student, also at the University of Bath who is investigating ideological dilemmas around childcare with Dutch and English mothers of children under two years.

Sharon CAHILL is a lecturer at the University of East London, her research interests include emotion, feminism, gender, and embodiment.

Riley, Sarah; Schouten, Wendy & Cahill, Sharon (2003). Exploring the Dynamics of Subjectivity and Power Between Researcher and Researched [62 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 4(2), Art. 40, http://nbn-resolving.de/urn:nbn:de:0114-fqs0302400.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License