Copyright Sociological Research Online, 2000


Katherine Reed (2000) 'Dealing with Difference: Researching Health Beliefs and Behaviours of British Asian Mothers'
Sociological Research Online, vol. 4, no. 4, <>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 2/8/1999      Accepted: 15/2/2000      Published: 29/2/2000


Contemporary sociology is faced by a central problem of conceptualising and rendering empirically operable the concept of difference without dissolution into perpetual plurality on the one hand, and recourse to fixed hierarchical relations on the other. Drawing on attempts to operationalise research categories within a research project on the health beliefs and behaviours of South Asian mothers, the paper explores the difficulties of operating concepts of difference at epistemological analytical and methodological levels. For example, within the research there are difficulties in operationalising concepts of local/global difference and differences between western and non-western medical systems without fixing one in a privileged position relative to the other or without seeing them as necessarily always equal. The research also raises questions of how to sample across multiple difference and develop interview and writing strategies which do not fix relations between researcher/researched in either equal or hierarchical relations. The paper draws on attempts to cope with these problems. It engages with post-modern approaches to difference but stops short of complete deconstruction, developing these approaches instead within a dialectical framework. A dialectical approach attempts to contextualise difference, recognising the interrelationship and contradiction between research categories of difference, temporally locating hierarchies between them. Methodologically, it also strives to develop an approach which steers a course in between a position of researcher as 'expert' and a position where our knowledge of others is treated as inconceivable.

Categorisation; Dialectic; Difference; Hierarchies; Post-modernism


This paper explores the problems of conceptualising difference in the context of my research into the health beliefs and behaviours of British South Asian mothers. The research adopts a theoretical framework which highlights both global syncrecy and the importance of material and geographical context. In the course of this research I have been faced with two related sets of problems. On the one hand there is a need to avoid the danger of over deconstruction which can lead to a potentially limitless plurality of categories, which creates insurmountable problems of operationalisation and analysis in the context of empirical research. Conversely, there is also a need to avoid the creation of categorisations of difference which are fixed in unequal relations and deny transformation and fluidity. These difficulties lie at epistemological, analytical and methodological levels of the research. The problem of diversity has always lain at the heart of sociological accounts, however this problem has become more pressing in the light of post-modern attempts to reconfigure diversity in terms of a celebration of difference.

Within this paper I use my attempt to operationalise my research categories to explore and work through some of these issues, focusing on problems which arose in the research project. The first problem is that of configuring difference between the local and the global whilst adequately recognising the interconnections between the two. The substance of the women's accounts quickly revealed that it was important to avoid positioning either the local or the global as automatically more important, but rather it was necessary to see them as specific to time and context. Secondly, it became apparent that there is a difficulty in operationalising differences between western and non-western medical systems. Again, it was necessary to avoid placing difference within rigid hierarchies where one (in particular western biomedicine) always takes precedence over the other. Thirdly, as a white researcher there are also issues which arise out of researching across structural difference. How can sensitive research across structural difference be practically achieved? Finally, there is the problem of operationalising categories of research, in the light of such difference. How far is it possible to take the deconstruction of categories in empirical research? What are the implications for sampling, interviewing and 'writing difference'?

In the course of my research, I found that while it was necessary to engage with post-modern approaches to difference it was also helpful to develop and situate these approaches to difference within a framework of dialectics. In the accounts of the women I researched I found widely varying responses to categories of difference. At times certain differences (for example the local or the global) would be privileged hierarchically, at other times there were significant commonalities across apparent categories of difference. There were also often contradictions in the women's expressions of difference. Analytically it is therefore helpful to see such differences as polyvocal, allowing for plurality within difference and within the women's accounts. The categories of difference employed by the women were not necessarily mutually exclusive, there was communication across categories and they were often held in dialogue. Through such communication differences at times informed and connected but at other times they clashed and contradicted one another.

Seeing such accounts through a dialectical framework means that differences can be seen as being held syncretically in tension with one another, at different times and in different contexts. This need not lead to the creation of a new more advanced 'synthesis' of differences which merge together, however (there is no self-completing truth or history). And whilst the dialectic appears to have different stages, these should not be seen as necessarily sequential. The dialectical approach is an analytical not a chronological one and differing phases of the dialectic should be located at different times according to historical, local and personal context (Reed 1998, Shrijvers 1993). By adopting this dynamic and processual dialectical framework in my research it is possible to explore more fully the multiple, shifting and sometimes hierarchical responses to difference that the women in the research expressed.

Researching Health Beliefs and Behaviours

The research project, which this paper will draw on, explores the health beliefs and behaviours of [1]British South Asian women in contrasting areas of Leicester. The project explores the idea that British Asian women's health beliefs and behaviours reflect trends towards global syncrecy. The concepts of syncrecy and syncretism have been used in research on health, and ethnicity and identity. The concepts are used to explore and denote crossovers and tensions between categories of difference (Fitzpatrick 1984, Parker 1995). The benefit of these concepts here lies in their resistance to completely collapsing categories of difference. Whilst recognising crossovers, syncrecy and syncretism also allow the potential for tensions between different categories. Within this study the concept of syncrecy is used to look at the way in which west and non-west are seen as mixed, both locally and globally, in a reciprocal context. The aim is to use this to look at the tension for British Asian women between west and non-west. In addition, the research explores the role of material and contextual circumstances on women's beliefs and behaviours, to ask how these affect their use of syncretic discourses.

The project is based on thirty in-depth interviews with British South Asian women from Leicester. Within the conceptual framework a distinction between health beliefs and health behaviours was recognised along the lines of McAllister and Farquhar's (1992) distinction: beliefs as feelings about lifestyles and health behaviours as health related activity. However, the distinction was not seen as fixed and the aim was to see whether respondents themselves made the distinction. Within the interviews no clear-cut distinction was made. The respondents of the study adhere to various religions, they come from a range of social classes and their ages range from early twenties to early forties. The women all have at least one child under the age of ten. Blaxter and Paterson (1982) highlight the important role of women as mediators of family health. The decision to focus on mothers arose out of this for the purpose of finding out not just about the women's beliefs and behaviours relating to themselves but also about their beliefs and behaviours regarding their families' health.

Respondents were drawn from playgroups and women's centres. Initially an informal pilot discussion group was conducted in order to refine the method for the main body of interviewing. This involved four women from a community playgroup, all of whom had small children. This was followed by informal interviews with two other women. The main interviewing process within the main body of research was staggered over a seven-month period. I took a multi-layered approach to analysis. My approach involved developing initial theories, which were then explored within the field. Having collected the data, themes were identified through the continual process of transcription. After themes were identified, the next stage of analysis took the form of accounting and membership categorisation; the final phase involved a reflection back to the initial theory. On writing up the interviews have remained confidential and names have been changed.

The findings of the research suggest that women's beliefs and behaviours do reflect trends to global syncrecy. These moves to global syncrecy however are not fixed. They are temporally specific and are significantly influenced (and dependent on) women's contextual and material circumstances. Women's position as 'British Asian' plays a significant role in influencing moves to syncrecy. In addition their moves to syncrecy must also be seen as part of a complex interweaving of their contextual and material circumstances and moves to broader global process. In particular, women's beliefs about, and use of syncretic discourses relates to their family, to particular types of illnesses, to their ethnic community and religion, to their location in Leicester and to their access to members of a broader global Asian diasporic network. The main body of this paper is concerned with the process of this research rather than with the findings in particular. At each stage of research outlined above I was faced with the difficulty of conceptualising and empirically operating difference. It is to each of these difficulties I will now turn, drawing on wider sociological attempts to conceptualise difference and developing a dialectical approach in order to transcend them.

Local/global Difference: Introducing a Dialectical Approach

Local and global processes are central to the research project. The current powerful phase of globalisation as argued by many authors, (e.g. Massey 1994) is marked by increased time/space compression, the opening of global boundaries and transcultural flows of goods. Such global movements can then be juxtaposed with the resurgence of place-bound traditions, languages and ways of life emphasising a renewed sense of localism. With this in mind, I wanted to look within the project at the importance of global process within the women's accounts of health whilst at the same time recognising the importance of the local and contextual factors. This focuses on how the situation of women's context in Leicester has wider global interconnections with India. I had difficulties in operating these concepts as interconnected but different, and in not wanting to prioritise one over another. My problem was in trying to avoid conceptualising categories as fixed in a particular hierarchical relationship where one takes precedence over the other.

In trying to conceptualise and operate these local/global categories of difference I turned to some of the key contemporary writers on diversity. Derrida has been highly influential here in terms of reconfiguring difference along non-hierarchical terms (Hobson 1998). In Derrida's (1978) terms, difference is a move away from hierarchical categorisation. It allows for complex sets of distinctions without opposition. It has been used and developed most fervently in thinking about gender, and difference in sexuality. This approach is useful in conceptualising local/global categorisation within the research. It points to the need within my research to bring categories of diversity such as local/global to the centre of the research. Empirically however, can moves to non-hierarchical forms of categorisation help operationalise concepts such as the local/global? Women within the research do not automatically see the local/global as different but equal. At times women do privilege the local over the global and vice versa. The women in the study often privileged global resources at particular times and in the context of particular illness and the local at others. A good example of this comes from Sakeena, a Muslim respondent. Within the research she talked about faith in local/global resources. It was quite clear that she privileged local non-western and western health resources over global ones:

Sakeena: Because I reckon here in Leicester, in the UK we've got more and better facilities. Although things go wrong here as well I'm not saying they're 100% but if you compare it to India or Pakistan (local facilities are better)

In writing about difference however, Derrida does acknowledge that escaping such hierarchies of knowledge is a difficult process. He does recognise that such metaphysical thought is embedded in our history. We can't merely rid ourselves of the urge to forge first principles and hierarchies. Derrida argues that we can however continually move away from such hierarchy through the process of deconstruction (Eagleton 1996). As soon as hierarchies are created they can be obliterated. With regard to the above quote, if we take a deconstructive approach the privileging of local over global resources can be deconstructed to make room for a reconfiguration of the local/global dynamic. This is a useful way of locating the contextual nature of hierarchy and difference. Such continual deconstruction however has the potential to lead to limitless plurality of categories creating difficulty with operationalisation and analysis in empirical work. In order to make deconstruction operationally more feasible I want to situate it within a framework of dialectics.

As argued earlier, I take a dialectic approach as one, which is processual. It is an analytic rather than necessarily sequential framework. As such it can move from one state of difference to another, privileging each at different times. Differences, which are multi and polyvocal, can conflict and contradict, but can also hold dialogue and communicate with one another. The outcome of such communication is not necessarily some self-completing 'synthesis' where differences merge (although that can be one outcome). Rather it is the case that differences often become held in tension in syncrecy and are located within historical, local and personal context. This dialectical framework can be used in the context of this research as an analytic and methodological framework. The proof of the value of this analytical and methodological framework is that when we look at women's accounts we can make more sense because we don't fix categories of difference, nor do we end up with a sea of endless difference.

We can see such processes in operation in the context of the local and the global; within women's accounts they focus on one or the other at particular times. For instance, many women felt that they had access to a whole variety of western and non-western health products and services, within Leicester. This access to a whole plethora of resources meant that at times there was a privileging of these resources. In talking about the availability of non-western medical products in Leicester Samina, a Muslim woman originally from Blackburn, pointed out:

Samina: In Leicester I've seen it (laughs), you can get everything in Leicester. Where I come from in Blackburn, we always had to get them sent down from India, like Tiger balm, don't forget the tiger balm!

This signifies a privileging of local resources. While local resources were privileged at some points women often moved onto global resources when local ones were insufficient, where certain services and products were available only in India. A good example of this can be seen in an account from Gurinder, one of the Sikh respondents. Her parents had been long term eczema and diabetes sufferers. Within Leicester her father had been on steroids for eczema and her mother was an insulin-dependent diabetic. Both were tired of long-term biomedical drug dependency and so they went to India to obtain specific non-western health products. The quote demonstrates her mother's quests:

Gurinder: My mum has diabetes and she's trying these herbal tablets from India to help with it. The treatment for diabetes is free and you have to take it for 40 days and you have to cut out a lot of things as well. They are herbal tablets; you can only get them from India.

This demonstrates a shift in many women's accounts from local to global. While women's location in Leicester was important in terms of access to western and non-western discourses, resources in India were important at other times. This movement from one type of difference to another is a two-way one, women also moved from global to local resources. It is also worth noting here that what women define as local or global depends on what context they are in. When in Leicester women identify India as being global in the sense that it is part of the global South Asian diaspora. The converse may be true when women are in India; India being local to Leicester's global.

The concept of the dialectic then makes room for such movement from the local to the global and vice versa. It also recognises that the two rather than being seen as mutually exclusive can be held in a dialogue and communicate with one another. This can be seen if we look at the way women talk about the transcultural flow of goods between India and Leicester. Women took health products such as painkillers and vitamins over to India to circulate them among local networks there. As the account of Sita, a Hindu woman demonstrates:

Sita: I always take multi-vitamin tablets out there when I go to India. My family over there they really like them from here. If anybody's going, we take them over.

The transcultural flow of goods is two-way as women also brought a variety of goods back to the UK. While being in dialogue with one another, the local and the global at times also contradict. The account provided by Shahnaz, a Muslim respondent originally from Blackburn, demonstrates this point. She tells the story of her brother in-law who has quite recently died as a result of stomach cancer. Health care services within Leicester failed to diagnose this cancer initially and he wasn't diagnosed with cancer until he went to India to use biomedical healthcare there. He was operated on in India and he came back to the UK to finish his chemotherapy. There was a conflict over the recommendation of the prescribed dose of chemotherapy between health care in India and that in Leicester. Shahnaz blames this conflict and in particular Leicester health services for his deterioration and death:

Shahnaz: He got the tumours removed in India. He was prescribed a dose of chemotherapy there, had some of the treatment there. He came home to Leicester and they wouldn't give him the dose prescribed in India. Within 3 months his symptoms came back.

The account from Mumtaz highlights the communication and conflict within a local and global dialectical approach. Data collected from the interview with Shahnaz also show how these contradictions between the local/global often move on to be held syncretically in tension with one another. Again this supports a dialectical approach:

Shahnaz: I mean, it doesn't matter whether it's a GP here (in Leicester) or a GP there (India), or a religious or herbal healer here or there. Whatever you keep them in conjunction and use them at different times.

The local/global dialectic is historically, locally and personally situated. This acknowledges the possibility that one or the other will be privileged, contradict and held syncretically in different women's accounts at differing times and contexts. This enables us to contextualise difference in the local and the global allowing for differing experiences.

Between West and Non-West: Difference in Medical Systems

The second problem in understanding issues of difference in my research was with conceptualising and empirically operating different health systems without suggesting one took automatic precedence over another. Brady (1997) demonstrates how medical systems tend to be looked at separately and hierarchically. Non-western medicine is often set up in opposition to Cartesian dualist models of western biomedicine. Biomedicine occupies a hegemonic position and is seen as a yardstick to which non-western medicine must measure. I wanted to avoid setting up a fixed hierarchical relationship between Asian and western medical systems. I also wanted to recognise a medical pluralism without suggesting a necessarily unchanging hierarchical relationship to this pluralism on one hand and without recourse to synthesis on the other. Finally, I wanted to be able to recognise the potential for diversity within medical systems themselves. For instance, as Worsley (1997) argues Ayurveda, the official medical system of Hinduism is far from monolithic but varies depending on origins in Sri Lanka, or India. The project here was focusing on women's use of a number of medical belief systems for example Western, Unani, Folk, Ayurvedic and Spiritual.

Brady et al (1997) appear to suggest a deconstruction of difference between medical systems. They argue that western and non-western medical systems should not be looked at separately. They propose that systems merge and are not in fact polar opposites. If we took a Derridian (Hobson 1998) approach to difference and medical system categorisation, these two systems would be deconstructed and recognised as different but equal. Both approaches are useful in deconstructing fixed hierarchical relations between systems. However, I would argue that to see systems themselves as blurred, as only ever a mix or match of difference denies that systems have characteristics, which are specific to them. To see them as different but equal also fails to take into account the long history of inequality between biomedicine and non-western systems. In the eyes of the state, biomedicine is still viewed as superior to any alternatives, which have to strive for official recognition (Cant and Sharma 1999). This can be recognised within the women's accounts as systems are definitely privileged over one another at times. If we look at Rambha's account of her niece's diabetes we can see the privileging of biomedicine:

Rambha: My brother was diagnosed as diabetic recently then his daughter was as well. She's going to the doctors all the time because she's insulin- dependent. She has to be careful what she eats, you feel sorry for her but yeah it has to be treated by the doctors.

In dealing with these difficulties again a dialectical approach has proved more useful in the context of this research. It enables us to contextualise the importance of temporality and fluidity of differing medical systems, in individual women's accounts at any given time. As with the local/global, women may move from favouring one medical system over the other within differing contexts and over time. This can be seen in women's accounts as they move throughout the lifecourse. Gurinder reflected on some of the changes in her attitudes to health and medical discourses:

Gurinder: A few years ago I would have thought you need a hospital, a doctor, whatever. Over the years I don't think the same thing. I think I would seek alternative medicine. I would pass that onto my children too. I wouldn't mind trying some of my mum's herbal remedies and stuff you know, what my mum did.

Again through a dialectical approach we can also see how medical systems communicate and enter into dialogue with one another and can actually at times cross over. Women's use of remedies for general illness highlights the crossovers between medical systems. Some remedies were prevalent in both western and non western medical discourses. This can be seen in a Muslim respondent's account at a basic level, with the use of honey. The healing properties of honey can be claimed by many medical discourses but she talks about it in the context of the Koran:

Sakeena: If it's a cough we use a little bit of turmeric powder in milk before we're going to bed and drink that. I do that regularly. Honey is also very good and it's in our religious book as well, there's a cure in honey, if we have bruises.

These systems also conflict and contradict one another. Sakeena's account of the treatment of her daughter's asthma and eczema supports this. Her teenage daughter suffered from asthma and was treated by the GP and used inhalers. Although the asthma was under control, Sakeena had felt that nothing her GP was giving her daughter for her eczema was helping. As a result Sakeena took her daughter to a [2]Hakim but found that the treatment by the Hakim for eczema conflicted with the GP's treatment of the asthma.

Sakeena: I took my daughter to a Hakim for her eczema. But what happened was they stopped everyday other things right, the doctor's medication for asthma. If she stops them an asthma attack comes on and she has to be admitted to the doctors, hospital, I couldn't do that.

Sakeena consequently ended treatments with the Hakim and stayed with treatments from the GP even though they appear ineffective. Taking the dialectic then we have seen how women may move their preference for one system to the other over time and in different contexts. We can also see through a dialectical approach how these systems communicate with one another and also at times conflict. Taking the dialectic further we can also see that they can be held syncretically or simultaneously. For instance we may look at women's use of medical systems for specific illness. We can see that women move in and out of preference for one medical system over different illnesses, from a predominance of biomedicine for diabetes to predominantly spiritual health systems for mental health problems. If we focus on general ailments such as common colds we can see that they draw on medical systems syncretically. As Kishwar, a Muslim woman within the study, observed:

Kishwar: I mean with coughs and colds we'd use Lemsip. That helps a lot and also for sore throats we use honey and lemon and we use other remedies from the kitchen, you know various herbal powders and Asian balms such as Tiger Balm that you can put on your chest.

Again the dialectical approach is a useful way of conceptualising and operating medical systems. It allows us to see how women move in and out of preferences over different systems. These systems are at times in dialogue and at others conflict and often are held in syncrecy. They are fixed neither in hierarchical relations nor in relations of equality. Again it must be recognised that the dialectical framework is an analytic not a sequential framework and therefore different parts of the dialectic in relation to medical systems will be relevant for different women at differing times and contexts, intersecting the local, personal and historical.

Knowing Difference: Epistemologies of Process

Within the research I was faced with the actual practical problem of researching across difference, across race and class. In epistemological terms, could I situate myself as a 'knowing' researcher? The implicit presumption of knowledge across difference and the operationalisation of this through the interview situation is seen by some writers to re-invoke a fixed relationship within which the researcher and researched are placed within particular unequal relations (Schenrich 1987). The researcher holds a dominant position over the researched. The possibility that one can research across structural difference has been questioned. As Ram (1996) argues, to do such research is difficult; conversely to hold a similar structural position to respondents is particularly beneficial when gaining information. How can one know what it is like to be in a different structural position? This is most keenly felt within this research regarding race. Rhodes (1994) and others have raised the issue of the feasibility of white researchers undertaking research with non-white respondents. As a white researcher could I understand and conduct research with South Asian women with variable class positions, which in some cases were far removed from my own? How could I do this without in some sense 'othering' and re-invoking a fixed hierarchical relationship between my respondents and myself?

With a complete deconstruction of difference we are left with a sea of endless difference where there can no longer be 'insiders' or 'outsiders' within research. Researchers must therefore either attempt to build bridges across difference or give up on social research altogether. Authors who argue for the feasibility of research across structural difference focus on building bridges over that difference. In discussing research across race, Anderson (1993) argues for the need to be culturally sensitive, suggesting white scholars can develop and utilise tensions in their own cultural identities to enable them to see different aspects of minority group experiences and beliefs. Feminist writers on epistemology and methodology have argued for similar bridge building endeavours (Skeggs 1997). One of the routes taken with reference to differences between women lies in the adoption of a 'partial perspective' (Haraway 1988). This is a perspective, which recognises that there are differences in experiences, positioning and power relations between women but that there are also commonalities. We can 'partially identify' with other women through our various experiences. While a partial perspective can be useful in attempting to research across difference there are limits. It can be critiqued for neglect of difference. Similarities are focused on and differences ignored. I would argue differences within research should not be ignored or explained away because differences matter, as do similarities as both affect research outcomes. Women within the study sometimes made a point about differences between us. For instance if we look at the account from Usha a Hindu respondent. In talking about family influence on beliefs and behaviours she was quick point to the differences between us.

Usha: You're different being young and not married or a mother, I suppose you don't have the same family constraints as me.

This affected the account this respondent gave to me about family and health. Had I been married myself, I think she would have been more likely to be more forthcoming about family effects. I would argue there is a need to incorporate difference into a framework, which also recognises commonality. Again this is why a dialectical approach is useful. Taking a dialectical approach we can recognise that we move in and out of similarity and differences. At times there are opportunities to develop commonalities between researchers and respondents. For example whilst not overcoming structural difference, my originating from Leicester fostered commonality across difference. My experience of spending time in India also built on commonalities as we shared tales of visits and compared areas. One respondent was so taken with my love of India it prompted her to offer quite seriously on several occasions:

Musarat: We are going to India at Christmas, would you like to come? We could find you a husband there.

Within the research situation there was often movement from similarity to difference between researcher and respondent and vice versa. A partial perspective focuses on fostering commonalties across difference. A dialectical approach recognises that these differences and commonalities are also often contingent, contradictory and acknowledges the plurality within the category of difference itself. The contradiction and plurality within difference can be used to develop a syncretic position between researcher and respondents, one, which holds similarities and differences in tension. For example, in some cases within the research, particularly regarding the interviews taking place in community centres, stark structural differences between the "community" and myself were used to actually build bridges between my respondents and myself. I interviewed Inder, a manager at a Sikh women's centre within her private office. During the course of the interview we were interrupted on several occasions by people wanting to take a look at me and find out why I was there.

Inder: I mean here now. They'll not say, "oh she's got someone with her" (means us in Inders office). They will still come in wanting to know a) why you are here, b) what you are saying to me (both laugh). But they want to know all the bits and bobs, what's going on. They're nosy, let's give them something to talk about!

Such explicit conflict between the community centre, and myself rather than inhibit the research situation with Inder and myself actually enhanced rapport, as we became complicit against community centre members.

The dialectic enables us to recognise movements between commonalties and differences. It also enables us to see that at times these conflict and contradict, and that actually recognition in difference itself enables us to develop rapport with respondents across difference. While recognising that there are limits to attempts to building bridges across difference, the dialectic with its focus on a politics of location then grounds commonality and difference historically, locally and personally. Taking such an approach puts issues raised by the partial perspective into context. It experientially anchors the commonalties between my respondents and myself then grounds these in tension with our differences.

'Doing' Difference: Methodological Operations

If I could deal with the above issues effectively, how could I then go about identifying my sample and carry out the interviews and write up the research? How do I operate diversity on an empirical level and also without recourse to hierarchies between categories, which become fixed? Within the research for instance regarding sampling, I focus on the health beliefs and behaviours of British Asian women. This could assume that South Asian women are a coherent group. In order to avoid this, the aim was to focus on just one religion, Hindu women. However this carries with it the danger of neglecting recognition of diversity among Hindu women. Hindu women are not a static categorisation; they are an incredibly diverse group not just in terms of geographical origins (e.g. East-Africa, Gujarat etc.), but also in terms of caste and class. In the face of this, was the research feasible and who would my respondents be? This is not something that I recognised fully until I went out into the field where seemingly endless diversity seemed to leave my research methodologically inoperable.

I drew on respondents from a range of religions: Muslims, Hindus and Sikhs; these respondents came from various socio-economic classes. A fully deconstructionist approach to such difference would involve an obliteration of categories leading to limitless difference and complexity in trying to deal with sampling. Taking a dialectical approach I proceeded from recognition that there were both similarities and differences within and between each of these groupings. Women move from one type of commonality with some women to other types of commonality with others. Taking this approach meant that I must recognise that any one part of women's identity, such as class or religion may be more important at any one time, and that these interrelate across the women's accounts fostering other commonalities and difference.

For example, in some women's accounts at particular times, such as during pregnancy and childbirth, religion seemed very important in influencing their health beliefs and behaviours. For instance the consumption of 'hot' and 'cold' foods at particular points of the pregnancy seemed to be particularly important for Muslim women in the research. Women ate particular types of food throughout the pregnancy. Hot foods were consumed at the end of the pregnancy, cooler foods at the beginning. Women talked about hot and cold foods not in the context of being physically hot or cold. Foods are hot and cold in the way that they react to your body, for instance milk or bananas are cool and help bring down a temperature. These ideas were central to many of the Muslim respondents' accounts as Shahnaz and Sakeena demonstrate:

Shahnaz: If you have food that is especially hot it will make you bleed more. That's why you're not supposed to have it at an early stage, as you don't want to bleed you might miscarry.
Sakeena: When you're nine months, you should start eating foods, which are hot and slippery, like ghee. It helps the baby come quickly.
The Hindu and Sikh women within the study were less focused on diet in pregnancy and did not talk

about food in terms of 'hot' and 'cold' properties. Women move from one type of commonality to focus on another. At other times for example, other factors appeared more prevalent than religion. For example if we look at East African women within the study who were either Hindu or Muslim, their sense of space and place can at times be seen to have more in common then those with the same religion from different places. On the whole these women are less likely to use non-western health care than those from India are. These commonalities are in dialogue with one another. They also conflict and contradict one another as differing contextual influences fight for recognition. This is where through a dialectical approach we can see that commonalties and differences can be held syncretically. For instance if we look at accounts taken from the interview with Sakeena we can see the way differences can be held in tension with one another. Sakeena moves from different types of influences on her health beliefs and behaviours from friendship commonalities cross differing religions, to strictly Muslim influence;

Sakeena: I have built up quite a social group of women who are Hindu and Sikh. I learn so much from them. They tell me about all sorts of health remedies specific to their religions and I try them. Some I use regularly. My own religion is also very influential on my beliefs and behaviours. There are so many remedies I have learnt from my mum, and yes they are specific to the Muslim faith.

Taking a dialectical approach here draws on arguments made within the previous section about building bridges across difference, but recognising tensions among this difference. Women move through commonality, difference, conflict and syncrecy in differing contexts determined historically, locally and personally. In recognising the processual nature of the dialectic, these categories of women were not seen as fixed but as continually shifting and being reconfigured, depending on what they were talking about, who they were talking to, and their temporal location.

Moving on from issues of sampling to the actual process of interviewing and writing difference, we can see how the differences between respondents and myself within the interviews were continually shifting and were rarely equal. By taking a dialectical approach we can recognise that the existing research hierarchy is always in process, moving from the researcher being dominant at particular times and respondents being dominant at others. At times it would seem that I dominated the research situation, after all I was asking the questions from an interview schedule, which I had designed. Within some of the interviews I was also perceived to be 'medical expert' and women asked advise. Harpreet was a Sikh respondent who had suffered many health problems relating to gallstones. She was particularly interested in my interpretation of her health problems, especially those relating to diet and weight loss:

Harpreet: I do try and diet and lose a bit of weight. I go to the gym but my weight just stays where it is, do you know anything that might help?

While I was in some cases perceived to be an 'expert', or in control of the research situation, the dynamic of researched/researcher and lay/expert dyad was often subverted as we moved from a position of my dominance to that of my respondents. Again this reinforces through the dialectical process the contextual location of knowledge and the production of knowledge through dialogue which makes room for a plurality of voices. Respondents led me to new spaces in the sense of giving me advice on alternative, non-western medicine and remedies. What was interesting was that women would often check out my health related issues and offered advice. For instance in talking about beliefs and behaviours around general health Sakeena was quick to evaluate my skin complexion and suggest remedies to improve it:

Sakeena: Ok, so your skin is combination/oily, use a little yoghurt, gram flower, mix it and make a paste and put it on your face. Leave it to dry and upwards, it improves circulation on the face and gets rid of excess oil.

At times there was also conflict and contradiction of hierarchical relations between my respondents and myself as often we talked over one another, metaphorically fighting for recognition. This usually led to more syncretic hierarchies between us, our positions being held in tension with one another. Difficulties associated with fixing difference between the researcher and the researched in a hierarchical relationship can also be translated into the writing up phase of research. One approach, which tries to deal with that, is a post-modern ethnographic approach. This type of approach attempts to obliterate the hierarchical and dualist nature of the researcher-researched relationship in a textual sense. It aims to bring the subject to the centre of research and argues for polyvocality, reflexivity and deconstruction of the authorial voice (Clifford and Marcus 1986). With polyvocality, voices previously silenced are able to speak. The research is no longer located within the discourse of the author. This approach is useful in bringing our attention to unequal relations between the researcher and researched. However, while it might appear feasible at a conceptual level, a complete deconstruction of a research situation is a somewhat na´ve objective. It does not acknowledge the potential for hierarchical relations between researcher and researched. As Parker (1995) argues, no matter how much shared experience there has been, the sharing stops when the writing begins.

Again a dialectical approach provides a useful alternative here. As Schrijvers (1994) argues, one of the challenges of a dialectical conception of knowledge is to find ways of constructing the established dichotomy between the personal and professional in the process of academic writing which is no easy task. How to integrate the ethnographic account itself with the interaction between researcher and wider situation within which research takes place, overturning the gaze. This is where I think a dialectical approach has a particular edge over the post-modern approach outlined by Clifford and Marcus (1986). It is unfeasible to aim for complete polyvocality of the written text, seeing the text as a joint endeavour. The dialectical approach takes us beyond this by recognising the hierarchy within the writing process, but recognising that this does not remain unchallenged. Within the research while my voice might be at times the loudest, it is not the only one. Through the utilisation of 'thick descriptions' (Geertz 1993) my voice was often challenged by the respondents' voices. My respondents' voices and mine became held syncretically with one another.

Conclusion: Dealing with Difference Dialectically

In this paper I have looked at some of the central dilemmas faced by conceptualising parameters of difference which were raised in my research on the health beliefs and behaviours of British Asian mothers. Part of the difficulty lies in conceptualising and operationalising categories such as the local/global and western and non-western medical systems in my empirical research. Other difficulties emerged out of the epistemological and methodological practicalities of researching across difference. These dilemmas can be applied to wider difficulties in contemporary sociology in rendering empirically operable concepts of difference. Sociologists it would seem are now faced by a world of incommensurable differences, and with discriminating in a complex reality, which does not slot into fixed categories.

There has been a proliferation of 'new sociologies' working along the axes of diversity which attempt to reconfigure difference through post-modern conceptualisations of difference. These approaches are useful in their obliteration of hierarchical categories and in their movements towards plurality. As Lemert (1995) argues, we can no longer scurry away in the face of difference. We must face and reconceptualise difference, which is not based on hierarchies of knowledge that become fixed in time. However, in the context of research such obliteration of categories is limiting and does not tell us how to proceed either methodologically or analytically.

In this paper I have explored an alternative approach which enabled me to work through my own research dilemmas, by using the framework of dialectics. Whilst this recognises difference as shifting and relational, and sees difference as always in process, it also recognises that hierarchy occurs. In the accounts of the women I researched difference and diversity was sometimes experienced and expressed as hierarchical or contradictory and at other times as complimentary or in dialogue. An adequate approach would need to recognise these variations in the experience of difference and in the shifts between them. This entails the recognition that difference is temporally located and is continually being reconfigured. The concept of dialectics attempts to account for this. It allows for the plurality of voices within difference itself and ultimately contextualises difference within a politics of location, situating difference in personal and historical context. Proceeding from this approach I was able to explore the tensions and interconnections within women's accounts between the local and global, and to move away from a hierarchical view of medical systems in which western biomedicine always takes precedence. The dialectic offers instead a more contextually located view of any hierarchical relationship between them.

In terms of researching across difference at an epistemological level the dialectic fosters attempts to build interpretative understanding across that difference. The dialectic recognises limits to such understanding through locating this difference in a politics of location. In attempting to operationalise categories of difference, similarity and differences in women's identities are held in tension to move away from over-complexity in sampling. Finally within interviewing and writing up, the processual nature of the dialectic moves us beyond the fixed hierarchical relationship between the researcher and researched. This relationship is, rather, continually being reconfigured. A dialectical approach, then, though not the only approach to difference, appears a useful one in dealing with difference in the context of empirical research.


1 The term British South Asian woman here is used to refer to women who have either been born in Britain or have moved to Britain under the age of five.

2Hakims are practitioners of the Muslim medical system Unani; Vedas are the practitioners of the Hindu medical system, Ayurveda.


I would like to thank the editor Liz Stanley and anonymous referees for their helpful comments and suggestions. I would especially like to thank Wendy Bottero and Graham Crow for all their patience, support and comments on the many drafts of this paper.


ANDERSON, M.L (1993) 'Studying Across Difference: Race, Class and Gender in Qualitative Research' J.H Stanfield and R.M Dennis (Editors) Race and Ethnicity in Research Methods London: Sage.

BLAXTER, M and PATERSON, E (1982) Mothers and Daughters: A Three Generational Study of Health Attitudes and Behaviours London: Heinemann.

BRADY, M (1997) 'Australian Aborigines conceptualisations of health and the world health organisation' in M Worboys and L Marks (Editors). Migrants, Minorities and Health: Historical and Contemporary Studies. London: Routledge.

CANT, S and SHARMA, U (1999) A New Medical Pluralism: Alternative Medicine, Doctors, Patients and the State London: UCL Press.

CLIFFORD, J and MARCUS G (1986) Writing Culture: The Poetics and the Politics of Ethnography. California: UCLA Press.

DERRIDA, J (1978) Writing and Difference. London: Routledge and Kegan Paul.

EAGLETON, T (1996) Literary Theory: An Introduction Oxford: Blackwell Publishers Ltd.

FITZPATRICK, R (1984) 'Lay concepts of illness' in R Fitzpatrick, J Hinton, S Newman, G Scrambler and J Thompson (Editors) The Experience of Illness London: Tavistock.

GEERTZ, C (1993) The Interpretation of Cultures: Selected Essays. London: Fontana.

HARAWAY, D (1988) 'Situated Knowledge's: The Science Question in Feminism and the privilege of the partial perspective' Feminist Studies Vol. 14, Number 3, pp. 575-600.

HOBSON, M (1998) Jacques Derrida: Opening Lines London: Routledge.

LEMERT, C (1995) Sociology After the Crisis. Colorado, USA: Westview Press.

LEMERT, C (1997) Postmodernism is Not What You Think. Oxford: Blackwell Publishers Ltd. MASSEY, D (1994) Space, Place and Gender. Cambridge: Polity Press.

McALLISTER, G and FARQUHAR, M (1992) 'Health Beliefs: A Cultural Division?' Journal of Advanced Nursing Vol 17, Number 12, pp.1447-1454.

PARKER, D (1995) Through Different Eyes: The Cultural Identities of Chinese People in Britain. Aldershot: Avebury.

RAM, M (1996) 'Ethnography, Ethnicity and Work: Unpacking the West Midlands Clothing Industry' in E.S Lyon, and J Busfield (Editors). Methodological Imaginations. London: Macmillan

REED, K (1998) 'Contextualising Comparative Research: Health Beliefs and Behaviours of American and British Born Asian Women' Current Research Proceedings New York State Sociological Association, New York: FIT, State University of New York.

RHODES, P.J. (1994) 'Race of Interviewer Affects: A Brief Comment' Sociology Vol. 28, Number 2, pp 547-58.

SCHENRICH, J (1987) Research Methods in the Postmodern London: Falmer Press.

SCHRIJVERS, J (1993) 'Motherhood Experienced and Conceptualised: Changing Images in Sri Lanka and the Netherlands' in D. Bell, P. Caplan and W.J Karim (Editors), Gendered Fields: Women, Men and Ethnography. London: Routledge.

SKEGGS, B (1997) Formations of Class and Gender (Theory, culture and society series) London: Sage.

WORSLEY, P (1997) Knowledges: What Different People Make of the World London: Profile Books.

Copyright Sociological Research Online, 2000