‘Claims and Disclaimers: Knowledge,Reflexivity and Representation in Feminist Research’

Introduction

1.1 In this article I explore issues and dynamics of ‘intellectual craftswomanship’ and consider the tensions involved when aiming to produce ‘accountable knowledge’ (Stanley 1999). Following Stanley (1999) I believe that attention to these concerns, alongside the analytical and political engagement with substantive issues that affect the lives of women and men should be the project of critical feminism. I am not suggesting that critical reflexive ‘craftspersonship’ is only undertaken by feminists, indeed there is a long tradition in interpretive sociologies of writers making transparent the working practices involved in producing knowledge claims (e.g. from Gouldner 1970 onwards). But feminists’ work combines particular analytical, ethical and political dimensions and I am concerned with this interplay here.

1.2 As Harding (1987: 3) notes ‘it is not by looking at research methods that one will be able to identify the distinctive features of the best feminist research’: i.e. there is no such thing as a feminist method, rather as Millen notes what is distinctive about feminist research is the incorporation of two main aims:

1.3 Yet, as others have noted, although there are points of commonality between feminists there are ‘many varieties of feminism and many ways of being feminist’ (Temple 1997: 1.6 and also see Stanley 1990, Fonow and Cook 1991, Reinharz 1992, Maynard and Purvis 1994, Ribbens and Edwards 1998 and so on). In this article I consider some commonalities and varieties.

1.4 Using my own feminist sociological doctoral research as an example I add to feminist debate on epistemological authority and consider in particular the epistemological status of my work in relation to reflexivity and representation. Within this I outline what I think are some of the contradictions involved in ‘theorising about feminist epistemology’ and ‘doing feminist research’. In my research I was concerned to explore the social, emotional and medical experience (predominantly women's) of ‘infertility’ and ‘involuntary childlessness’. ^[1] My original motive for undertaking the research was a response to the fact that I felt that the experience of ‘infertility’ and ‘involuntary childlessness’ was misunderstood and misinterpreted by many people. My political interest was in part stimulated by my own experiences in that I fit the medical definition of ‘infertile’ and at the time of the fieldwork I was ‘involuntarily childlessness’ ^[2] . This auto/biographical element is obviously relevant to my discussion here and I consider throughout my positioning in the research process and product.

1.5 The main part of this article is divided into four sections. I begin by outlining my research both in relation to method and ‘findings’. I follow this by reflecting on my study and make some initial claims and disclaimers outlining what I think I did and did not do. Next I consider the relationship between my approach to research and my presentation of respondents’ experiences and highlight what I think are some of the contradictions and tensions in feminist theorising and feminist research. I extend this discussion through a consideration of the significance of the researcher's and the respondents’ place in the research . Finally, I conclude the article by briefly re/locating my argument within wider debates concerned with the value and status of epistemological debate.

Setting the Research Scene

2.1 My research was qualitative in its approach and in-depth semi-structured interviews (combining a life-history approach with some standard questions) and research by correspondence were the methods used to collect the data. A processual approach (i.e. series of interviews or letters with each person) was adopted to obtain access to different viewpoints on experience over time.

2.2 Like similar research in this area (e.g. Pfeffer and Woollett 1983, Monach 1993, Franklin 1997) my work has highlighted the fact that the discussion of the status and experience of ‘infertility’ and ‘involuntary childlessness’ is affected by several dominant discourses including the discourse of medical and scientific superiority (in relation to treatment); the discourse of the deserving and the undeserving (in relation to who should and should not parent) and the discourse of loss and damaged identity (in relation to non-parenthood). All of this in turn further affects the experience of ‘infertility’ and ‘involuntary childlessness’.

2.3 The data suggested that the ‘infertile’ and ‘involuntarily childless’ are aware that these dominant discourses and the associated ‘ideal’ identities affect the ways in which their status and experience is described and indeed often affects the way they describe it themselves. Thus, the ‘infertile’ and the ‘involuntarily childless’ are affected by the stereotypes used to describe the experience - stereotypes which often conflate the two statuses. Yet, as in other areas of life individuals in this position resist the stereotypes if they feel that they do not adequately describe their experience. This resistance sometimes includes reference to feelings of ambivalence in relation to their experience. Recognition of ambivalence leads to consideration of self-identity. Self-identity in this instance is again affected by ‘ideals’ and respondents recognised the value of ‘ideal’ themselves. Thus, my research suggests that the ‘infertile’ and ‘involuntarily childless’ often feel a sense of damaged identity as individuals, as a couple, as a family. Some reported a sense of despair. Yet many resisted and rejected desperation as the dominant expression of their feelings, whilst some try to make stereotypes of desperateness work for them in negotiating the medical setting. Due to the stereotypes surrounding the experience respondents spoke or wrote of ‘hiding’ aspects of their identity and experience in certain settings. Yet many also referred to the political importance of publicising their experience in order to ‘tell it like it is’, which in part explains the motivation to be involved with the research (for further discussion see Letherby 1997, 1999).

2.4 My research provides further evidence that women (and men) without children represent the ‘other’ in a society that values parenthood. This affects relationships with family, friends and partners. Therefore the ‘infertile’ and ‘involuntarily childless’ often feel excluded. Yet this is complex. The ‘infertile’ and ‘involuntarily childless’ feel excluded in different places, in different ways. Individuals who don't have children, and those who have achieved parenthood in ‘unusual’ circumstances, report feeling excluded in different situations. All of this results in the experience involving individuals in large amounts of emotional work and management. This suggests that contact with those who understand, i.e. other members of the ‘in-group’, is desirable. But my research suggests that the ‘infertile'/'involuntarily childless’ have complex experiences of, and feelings towards, this sort of support. Counseling is overwhelmingly supported, although little is available, whereas the topic of self-help and support groups brings differing responses. Some of my respondents had never heard of these, whilst others had decided that such support was not for them. There are various reasons for this, including the fact that respondents perceived these groups as focusing on a particular aspect of experience that was not relevant to them. My research suggests that the whole experience of ‘infertility’ and ‘involuntary childlessness’ needs to be related to issues of time and control. Time is relevant with respect to the ‘right time’ to have children, to the fact that treatment takes time, and to changes over time in terms of adaptation. Yet this adaptation is tentative: as the life course continues, new adjustments are necessary. With respect to control, respondents reported lack of control over their body and their life in general, again this sometimes changes over time (Letherby 1997, 1999).

2.5 Overall, my work highlighted the necessity of placing any consideration of the experience and status of ‘infertility’ and ‘involuntary childlessness’ within wider political ideologies and structures.

Initial Claims and Disclaimers

3.1 Whilst much of this is not new in itself I would argue that with the doctoral research ‘aim’ in mind my research does indeed make several ‘specific contributions to knowledge’. Thus:

3.2 Firstly, by working with a diverse group of individuals who have varying experiences in relation to ‘infertility’ and ‘involuntary childlessness’, rather than focusing on a distinct group - undergoing a particular type of treatment, planning adoption and so on - as others have done (Franklin 1997, Monach 1993, Meerabeau 1989), I have begun to develop a feminist sociological analysis of ‘infertility'/'involuntary childlessness’ that relates to the whole experience: social, emotional and medical, and places this experience within the broader context of individuals’ lives. This has never, to my knowledge, been done before.

3.3 Secondly and relatedly, I considered in detail the experience of non-motherhood and the experience of women who have achieved motherhood in an ‘unusual’ way i.e.: non-biological mothers and mothers of so called ‘miracle babies’ (babies born as the result of medically assisted conception) (Letherby 1999). The initial point of the research was to explore the experience of a group of individuals at a particular moment in time. However, because of the fact that the study group represented a wide/broad group in terms of age and experience (some who have had treatment some who have not, some experiencing ‘infertility'/'involuntary childlessness’ now, some in the past etc.), and because my research relationships with respondents often extended over several months or even years (as some kept in touch to, as they said, ‘keep me up to date’) my doctoral thesis (and subsequent writings) in fact provides an analysis of experience over time and of different types of experience in relation to ‘infertility’ and ‘involuntary childlessness’.

3.4 Further, and thirdly, my study has implications for all women's lives, not least, in that given expectations of appropriate ‘womanhood’ (e.g. Connell, 1987, Hey, 1989, Letherby 1994, 1999, Gillespie 2000), motherhood/non-motherhood is a significant difference in women's lives. Other implications for the lives of all women (and men) are the issues raised by my consideration of specific aspects of the experience in relation to, for example, family and kinship identity, relationships with partners, family and friends, and medical care and medicalisation.

3.5 As well as these substantive contributions I would argue that my research makes a fourth contribution, in that my attempt at a grounded use of an auto/biographical epistemology is distinctive, although not unique, in that my presence was central in the research and in my writing. Relatedly, the recognition of this relates strongly to my epistemological concerns, in that I was concerned with issues of difference, power and representation. My work was grounded, as I shall show later, in a developing feminist epistemology.

3.6 In terms of my grounds for these ‘claims to know’, as Morley (1997) notes, feminist research often takes a grounded theory approach but can not be grounded theory. She writes that at one time, ‘grounded theory was seen as highly compatible with feminism, firstly, it was concerned to locate theory in participants’ worlds and secondly, it aided the process of breaking out of the confines of andocentric theory’ (p140). Morley adds that many feminists now reject grounded theory on the grounds that ‘no feminist study can be politically neutral, completely inductive or solely based on grounded theory, as all work is theoretically grounded’ (ibid). With respect to my research, I did have prior ideas about what I would discover which were influenced by my own experience, and from reading previous research and comparing my respondents’ experience to the respondents I read about in other similar research projects. My academic, political and personal interest in issues of definition, identity, relationships and networks in relation to the experience of ‘infertility’ and ‘involuntary childlessness’ and how all these change over time had an influence on the issues I raised in interviews and letters (though I did start by asking respondents to ‘tell me your story since you first decided that you would like to have children’). For much of my research I ordered data under these broad areas/themes. Within these parameters the data itself led the discussion. Therefore, my research took a grounded theory approach even though I agree with Morley's argument that feminist theory generated from feminist research can not be grounded theory.

3.7 I am conscious that I ‘took away their words’ and then analysed the data from my own political, personal and intellectual perspective. As Fine (1994) argues, research involves ‘carving out pieces of narrative evidence that we select, edit and deploy to border our arguments'(p22). Thus, I am aware that my voice is the loudest. With this in mind I attempted to be sensitive to issues of power and control throughout the whole research process. When writing up my data I highlighted my role in the selection and interpretation of respondents’ narratives and in terms of presentation of ‘findings’. In interviews and letters, I asked respondents to reflect on issues that they had spoken or written of previously and incorporated (some) individual changes into my analysis and writing. Yet, as Holland and Ramazanoglu argue, there is ‘no technique of analysis or methodological logic that can neutralise the social nature of interpretation’ (cited by Morley 1997 p142). As Morley adds ‘The difference with feminist research is that it admits it!’ (ibid). My work represents an attempt to work within an auto/biographical approach and, as I will consider further, later, there are tensions between theory and practice.

3.8 Stanley (1991) argues that ‘people theorize their own experience … and so researchers of the social are faced with an already “first order” theorized material social reality'(p208). Thus, ‘people observe, categorize, analyse, reach conclusions’ (ibid). I agree with this. We must respect what our respondents tell us. Yet, few (if any) research reports, provide complete transcripts of all respondents’ narratives, leaving any further analysis to the reader. The presentation of our work involves further categorization and analysis and we reach conclusions based on our interpretation of the data and the academic and political theories and understandings that we have access to. I started from an epistemological position that rejects a simplistic foundationalist/standpoint position. I do believe that I am in a position to generate the ‘true story’ of the experience of ‘infertility'/ ‘involuntary childlessness’. But I do believe that ‘my story’ can stand in opposition to and as a criticism of ‘other stories’ (both feminist and non-feminist, academic and lay) which I, and many of my respondents, see as at worst inaccurate and at best partial. I do not claim to have the ‘answer’. However, by starting to ask different questions of a different group I believe that my research highlights complexities of differences of experience that have previously not been considered. What is different is the focus on an under- researched group. This highlights the need for different explanations of the experience of ‘infertility’ and ‘involuntary childlessness’ than those which have previously been thought sufficient. I agree with Millen who writing about her own experience of research, writes:

3.9 I do not claim to have uncovered the ‘absolute truth’ or indeed that an ‘absolute truth’ is possible but I do claim to have widened the debate surrounding these issues, and, as a result, I am challenging aspects of previous work in the area of ‘infertility’ and ‘involuntary childlessness’, motherhood and non-motherhood.

3.10 Clearly, if my research had been undertaken by a different researcher, or by me at a different point in time, the result would be a different. Equally, if I had talked or corresponded with the same individuals at a different time they would have talked about or focused on different aspects of their experience. Thus, the research and the claims of my work are not representative of all. It is not possible to ‘prove’ the accuracy of my claims in an absolute sense but comparison of respondents’ experience, reference to other work in the area, and indeed to some of my respondents’ continuing experience suggests that a repetition of the research would uncover flavours of difference rather than result in distinctively different claims. Having established what I believe my work has and has not done I will now explore further some of the tensions raised by these claims and disclaimers.

Epistemology and Representation: one

4.1 As noted above, my epistemological approach rejects the view that there is one ‘reality’ out there that I as feminist sociological researcher can discover. I agree with Di Stephano (1990) that gender is a ‘difference that makes a difference’ (p78) but that differences between women (and between men) are themselves theoretically and politically important as are commonalities between women and men . I accept that knowledge is a material product: ‘something which is specific to time and place and person, and so which is contextually, grounded and material, as well as being rooted in the ‘point of view’ of particular knowledge producers … ‘ (Stanley 1997 p204 drawing on Rich 1986 and Haraway 1988). I also believe that respondents as well as researchers are reflexive, theorising individuals. Yet, I do make strong knowledge claims and argue that my research is ‘broader’, ‘fuller’ than what has gone before. So, I am arguing that my research is in some ways ‘superior’, and stands as a successor to what has gone before. Feminist research has political aims: it aims to effect social change, and my respondents wanted me to tell it ‘how it is’. My work represents my interpretation of ‘how it is’ rather than a matter of simple reportage and I agree with Temple (1997) who argues that:

4.2 Wilkinson and Kitzinger (1996), point out that ‘our work should not be so much about the other as about the interplay between the researcher and the other'(p18). But as they themselves add, ‘many feminists want both to enable the voices of Others to be heard, and to create social and political change for or on behalf of those Others’ (p20). This creates a dilemma and involves us in a struggle between acknowledgement of the impossibility of full representation and the assertion that our work makes a difference. This leaves me then supporting an approach which may possibly involve a less than complete representation of the other, but I suggest that this is better than no representation at all.

4.3 ‘Doing feminist research’ highlights for me the problems in taking an epistemological position. I find myself arguing for an epistemological position somewhere between ‘epistemic privilege’ and ‘post-modernism'/relativism. I do not claim that my work is by definition superior to other knowledge claims and indeed it should be subject to critical enquiry (Stanley 1996, Stanley and Wise 1993). Stanley and Wise (1993) argue for a methodological and epistemological position which they call Feminist Fractured Founationalist Epistemology: a position that does not dispute the existence of truth and a material reality but acknowledges that judgements about them are always relative to the context in which such knowledge is produced. From this perspective researchers acknowledge that they are not intellectually superior to their respondents and they have the responsibility to providing accounts of their research process so that readers can have access to the procedures which underlie the way that knowledge is presented and constructed by the researcher. Yet, as Maynard (1994) and Abbott and Wallace (1997) suggest there is a problem with this because of the difficulty of defining all accounts as equal with no way of selecting between them. With this in mind, and from my own research experiences, whilst I agree with Stanley and Wise that as researchers we are not intellectually superior to our respondents I do think it is important that we acknowledge our intellectual privileges.

4.4 My research can be compared to qualitative interpretations in general, where the goal is to consider as many interpretations as possible and to give respondents the authority to define themselves and their position. Even if this happens within the fieldwork, the thesis or report represents the researchers’ and not the respondents’ definitions. In epistemological terms, this is particularly relevant to me, because, my respondents comprised a very knowledgeable group: many of them knew a great deal about the medical research on ‘infertility’ and about ‘causes’ and ‘cures’. Yet, perhaps ‘knowledgeable’ here is sometimes better defined as possessing information or ‘detail’ rather than ‘knowledge’. For example, an individual experiencing unexplained ‘infertility’ is likely to have lots of information but does not have knowledge (and neither does the clinician) regarding their status. Similarly, a couple turned down for adoption may not be given the reasons why, and thus have no knowledge as to why they were not considered to be appropriate adoptive parents (see for example Franklin 1992, Letherby 1997). As a researcher, I have access to many different professional, political and lay bodies of information and knowledge. Clearly, I refer to these when I adjudicate between respondents’ accounts and between respondents’ views and my own. I didn't always agree with their views and/or felt I would have done things differently if I were them. At many places in my writing I have presented several sides of an argument but accept that I evaluated these in terms of MY feminist and sociological standpoints. So I have the final say. I am not only claiming a privilege here but also a ‘superiority’: a right to be regarded as a knower in a way that respondents do not have. This may involve some misrepresentation of their words but a final decision is necessary if (feminist) research is to say anything at all, have any effect at all, and not be concerned solely with issues of representation rather than ‘reality’ itself (Kelly et al 1994).

4.5 Furthermore, respondents are sometimes aware of and even supportive of this ‘intellectual privilege’. As Wolf (1996) notes, participatory research ‘can entail very disparate levels of input from research subjects’, and respondents may not wish this type of involvement, and instead wish the researcher to ‘speak for them’ (p26). Indeed, Annie (one of my respondents) described me as her ‘little soapbox’ telling me that she could have her say through me without identifying herself publicly (see Scott 1998 for a similar argument). Thus, there may be a tension between the desire to give women a voice and the making of knowledge (e.g. Millen 1997, Maynard and Purvis 1994) not least because individuals may not necessarily possess the knowledge (or have the desire) to explain everything about their lives. Furthermore, whilst we need to recognise and account for researchers’ power we should not characterise respondents as uniformly passive or powerless (see e.g. McRobbie 1982, Skeggs 1994, Millen 1997, Ribbens and Edwards 1998).

Representation and Reflexivity

5.1 My aim then was to be reflexive in relation to representation. Reflexivity was something that my respondents, as well as I, engaged in, and it is useful to discuss here the differences between ‘descriptive reflexivity’ and ‘analytical reflexivity’. As reflexivity can be defined as reflecting back on something, descriptive reflexivity is clearly a description of one's reflection. Analysis means breaking something down into its constituent parts or elements and examining the relationship between them so analytical reflexivity involves comparison and evaluation. All individuals reflect on their lives and on the lives of others. This can be demonstrated by consideration of respondents’ narratives and accounts. All my respondents described aspects of their experience, their views on motherhood and parenthood, their views of the opinions of others, their relationships with others and so on. Many also were clearly being reflexive while narrating events and feelings e.g.: in terms of their choices and their understandings of things. This led some respondents to analytical reflexivity: i.e. it led them to examine and evaluate their views and experience and the perceptions of others. Thus, it is possible to argue that the research process led to increased reflexivity in some respondents. Indeed, many spoke or wrote of how being involved in the research had made them feel differently about aspects of their experience and/or helped them to make decisions about future ‘choices. Whilst I am not suggesting that ‘feminist research is feminist politics’ (which Gluksmann 1994 p150 warns us against claiming) some of my respondents said that involvement in the research had made things better for them. The research process also led me to increased analytical reflexivity in relation to my own experience of ‘infertility’ and ‘involuntary childlessness’. I began to consider in greater detail my own experience, emotions and choices just as many of my respondents did and change and adaptation were not only important themes in the research but in my own life at this time (see Letherby 2000 for further discussion).

5.2 In my work I aimed to be ‘analytically reflexive’ in that I drew on and aimed to represent and interpret the experience of my respondents and the theorising that respondents engaged in. The analysis is mine, and it is an analysis that is located in a body of feminist and sociological theorising. As a researcher I had access to my own experience, that of my respondents, and access to theoretical explanations that were not available to many of them. I was in a privileged position as a researcher not only in terms of access to accounts and availability of time and space, but also in terms of my discipline training. This was not least because of my familiarity with the ‘sociological imagination’ as a theoretical research tool which enable me to engage in ‘second order theorising’ or what Giddens (1984) calls the ‘double hermeneutic’. This clearly involves ‘interpretation’, not just ‘description’, of their, as well as my, analytical processes.

5.3 I would also argue that I have access to more narratives of experience and more interpretative tools than my respondents and I have also been ‘given’ more time to think and particularly to theorise about these issues than many of the people I spoke and wrote to. My presentation is filtered through my understandings, but at the same time I have made a self conscious attempt to understand my respondents’ understandings in their own terms. So, I accept that my interpretation is grounded in time, place and person, and in accepting the ‘job’ of researcher I believe I have a responsibility to acknowledge my privileged resources and thus stand by the claims that I make. I have tried to adopt what Stanley and Wise (1993) call ‘a morally responsible epistemology’ (p200) which recognises that ‘the ‘objects’ of research are also subjects in their own right’ (p200), and that that ‘my work is not a ‘representation of ‘reality'’ but my own construction (ibid).

5.4 With respect to making analytical procedures transparent (Stanley 1997) I appreciate now that I could have been clearer about which aspect of an individual's narrative I was drawing on: which particular interview or letter I was quoting from, what they spoke or wrote about just before and after. More explanation could have been given about why I chose, for example, to include Gloria talking about this experience rather than that one, and why I chose Gloria's narrative or account over Ida's. I could have included details on my step-by-step decisions and choices surrounding selection and interpretation. As it stands, the interpretation is based on my ‘feel’ for respondents understandings and my familiarity with the data which built up over time and is related to my developing relationship with respondents over the fieldwork period and the visiting and re-visiting of tapes and letters. My thought processes were affected by respondents’ voices as well as my own. There is a tension when working within word limits and when trying to write a piece that is both accountable and representative. To do both thoroughly requires a lot of space and requires and attention to detail which may possibly submerge the respondents’ voices and which may involved so much extraneous material that essential analysis or argument becomes submerged in detail.

5.5 Descriptive and analytical reflexivity are both essential parts of the research process. Researchers are themselves people, with their own ‘responses, values, beliefs, and prejudices’ (Morley 1997 p139) and research involves selection, explanation, interpretation and judgement. Thus, it is important that the ‘person’ is made explicit and the processes involved in research procedures are clearly outlined in order to uncover the differences that we as researchers make (Jones 1997).

5.6 Conscious of this at the start of my research, I was concerned to ‘make myself vulnerable’ (Stanley and Wise 1993) within the research process, in terms of situating myself personally, politically and intellectually both with respondents and in my writing. I was aware of the danger of positioning my experience as the norm, against which others would be judged, and agree with Temple (1997) that ‘It is by listening and learning from other people's experiences that the researcher can learn that ‘the truth’ is not the same for everyone'(5.2). Yet I would argue that the inclusion of my autobiography is valuable as it enables respondents and readers of my work to compare my motivations, experience and views with those of respondents and other researchers. This enables them also to make their own judgements about my approach and in some respects my findings as well:

5.7 Throughout my writing I have made it clear that what is presented is affected by me as biologically childless, me as feminist, me as sociologist, me as researcher, me as doctoral candidate, me as Gayle. My ‘self’ and my understandings are explicit. Increasingly it is becoming usual for researchers to include aspects of the self in their writing (Wilkinson and Kitzinger 1996, Wolf 1996, Stanley 1993, Okely 1992, Sociology 1993), yet there still appears to be a tendency to keep this outside of the main report of a study (McMahon 1998, Aldridge 1993).

5.8 Despite the increased support for auto/biography approaches, I think that many people still feel uncomfortable with this way of writing. This is probably both for personal reasons, in terms of issues of privacy, and for academic reasons, in that they may be criticised for self-indulgence and sloppy intellectual work. This fear has some basis in reality (see Mykhalovskiy 1996, Letherby 2000). Whilst I appreciate that there is a fine line between situating the self and egotistic self-absorption as I have argued with Pamela Cotterill elements of ourselves are always present:

5.9 With this in mind I would argue that it is better to acknowledge our involvement rather than pretend to objectivity when writing up whilst at the same time being aware that within auto/biography some voices (not least by virtue of power of editorial control) are more prominent. Furthermore, agreeing collectively on the format and style of interviews and letters, inclusion of the researcher's ‘self’ and the joint exploration of issues are all ways to break down the subject/object split during the fieldwork period of research. However, once the fieldwork is over, involving respondents in writing and re-writing and presentation is time consuming, may not be wanted by respondents and is impossible to fully achieve when the researcher has the final editorial control. Further to this, my analysis involved listening and reading, transcribing, highlighting and interpreting, and at all times during this process I was aware of the different voices of my respondents. One of the main problems in terms of representation was that there were so many more experiences, emotions and evaluations than I had the space to include. I carried out 99 interviews (lasting between one hour and three) and received 100 plus letters, and as a result I had a great deal of data, and any reader of my and similar work needs to be aware of the shadows and silences of the resultant text. Although ‘quantity’ is not central to qualitative research and analysis, having data which included such a broad range of experience was beneficial to my research, both theoretically and politically. Yet it also resulted in less available ‘space’ in which I could discuss any one individual's life and experience. In my data analysis, I was concerned with themes and issues and in doing this I selected extracts from narratives and accounts that for me exemplified groups of respondents’ views on and about certain issues, the issues I considered most salient. In carrying out this analysis I also emphasised commonality and difference among my respondents. Whilst trying to present as many viewpoints as possible, it is also fair to say that in my view some respondents had more to ‘say’ than others. Some had a lot to say about certain issues and little about others. In interviews and letters, some of them appeared more comfortable with the ‘life-history’ approach, whilst others preferred a question and answer approach. This resulted in some respondents appearing more often than others, or in some Chapters of my thesis and not in others. For some respondents it is easier to follow the life-history of the experience of ‘infertility'/'involuntary childlessness’ and/or of non-motherhood to motherhood for them than for others of my respondents. Thus, my approach resulted in the fragmented representation of many of my respondents’ lives. This was accentuated by the fact that I decided not to write the thesis following a lifecourse format - for example from the decision to try to become pregnant - through the discovery that this might be difficult - through investigation and treatment - through to resignation to childlessness. My intention was to challenge simplistic representations of resolution in relation to the issue of ‘infertility’ and ‘involuntary childlessness’ which was I think valuable substantively but I am aware that this approach led to further fragmentation of individual stories.

Epistemology and Representation: two

6.1 At the beginning of the research I had hoped that I could ‘represent us all’. However, not being a complete novice to research, I was aware that there were likely to be changes to my approach to the subject. Of course there were. At the end of the whole process, I felt that I was even more sensitive to issues of power, sameness and difference, and I had a greater appreciation of just how hard it is to reach a ‘definition’ of what experience really is. And yet, on some levels, I would argue that my research does represent much of the experience of the ‘involuntary childless’ and the ‘infertile’ in ways others would find familiar and inclusive of their own experiences. And whilst I am not suggesting that all research should be auto/biographical many of us do draw on our own experiences when doing research and like Ribbens (1993) I believe:

6.2 My work in the area of ‘infertility’ and ‘involuntary childlessness’ represents the experience of those involved at specific times in their lives and their lives, like all lives, are subject to adaptation and change. This leads to a consideration of the value of this type of research. As Attar (1987) notes, experiential material is valuable:

6.3 Further to this, research of this type involves both first- and second-order theorising of experience by respondents themselves as part of their own self reflection, and by the researcher. Thus, it is possible to argue that this research may have value in explanatory terms, if not for its typicality, and it might therefore be relevant to others who find themselves in a similar situation (Clyde Mitchell 1983). Just as data can be revisited in relation to the consideration of different issues and themes, so too can specific biographies/stories be reworked in relation to different types of analysis. Having read transcriptions and letters, and indeed the doctoral thesis as a whole, many times now, each time I do, I see new gaps and think of new considerations. In the same way that as the lives of the individuals who I corresponded with and interviewed are liable to shift and change, so, too, is any researchers’ analysis and theoretical reflection. The project is never complete.

Concluding Points

7.1 Like many others I believe that by employing the sociological concept of reflexivity to produce ‘first person’ accounts we not only challenge the traditional objective/subject split within traditional research practices we also produce ‘accountable knowledge’ in which the reader has access to details of the contextually located reasoning process which gives rise to our ‘findings’ (Stanley 1991 p209). With this in mind I do not agree with Kelly et al (1994) that feminist researchers have concentrated on issues of epistemology at the expense of the intellectual and the political. They argue that feminist researchers seem:

7.2 With reference to my own work, and as I have suggested here, it was my aim that my work would make a difference to understandings of the status and experience of ‘infertility’ and ‘involuntary childlessness’. Therefore, like many others my hope is that my work will have a political emancipatory element (e.g. Acker, Barry and Esseverd 1991, Kelly et al 1994, Oakley 1998). However, I believe that epistemological reflection is part of and not separate from this political aim within feminism and this piece is part of a tradition that demonstrates the problem and tensions in arguing for a ‘position’ (e.g. Harding 1993, Millen 1997, Stanley 1999).