Doing intersectionality in empirical research

Most of my Gender Studies students are well versed in the theory and politics of intersectionality. However, this often seems to fall by the wayside when it comes to designing their research projects. Intersectionality is easy to discuss, but takes work to apply; this is work of designing and redesigning, questioning and (in Crenshaw’s words) ‘asking the other question’. In her famous article ‘Mapping the Margins’, Crenshaw defines three levels of intersectionality:

  1. Structural: how the social locations of black women make their lived experiences qualitatively different from those of white women
  2. Political: how feminist and antiracist politics have both marginalised the concerns of women of colour
  3. Representational: how the cultural construction of women of colour is produced by ideas about gender and race

When students attempt to apply intersectionality, the representational level often feels easier and more natural. However, without attention to the political and structural, this tends to lend itself to a superficial approach focused on ‘adding’ particular groups rather than exploring how systems and identities are co-constructed (what Patricia Hill Collins calls the ‘matrix of domination‘). In what follows I will attempt to sketch out some suggested protocols for ‘doing’ intersectionality: central to these is the fact that intersectionality is not an additive principle but an inherent one which requires us to interrogate the very foundations of our work. In other words, we need to apply it right from our ontologies, through our research questions and sampling, to the knowledge claims we make.

Ontology

Research always proceeds from ontology, whether this is a well-developed theoretical perspective or a simpler set of ideas about life. It is how you think the world works. If you are not intersectional in your ideas about the world, it will come through in your research. This is not just about acknowledging the existence of different types of people: crucially, you also need to think about how you define and locate structures such as the family, religion and the state. Our structural interpretations are often constructed from the perspective of a particular group, usually the dominant one. For example, since the 19th century black feminists have pointed out that state institutions such as law enforcement can be understood/experienced radically differently according to race. Privileged white women tend to look to the police for protection: for black women law enforcement is more often an agency of state violence against themselves and their families (usually perpetrated in the name of protecting whites). Despite this, the ‘neutral’ account of law enforcement is that they are here for everyone’s security: if you conduct research on an issue such as the under-reporting of sexual violence based on this ontology, your project will be exclusionary.

Developing an intersectional ontology also means interrogating key concepts such as gender, power and violence. This demands that we understand power relations both between genders and within them, mediated by categories such as race, class, sexual orientation, (dis)ability and age. It also means accounting for geopolitical flows of power between groups, nations and states in different regions of the world. Understanding a concept such as violence intersectionally asks us to broaden it from physical and sexual forms to include state, political, cultural and symbolic ones, which affect some communities more intensely and implicate others as perpetrators. Colonialism is the paradigm example. Within this framework, a term such as ‘violence against women’ becomes one-dimensional and inadequate. We must constantly challenge and complexify our ideas as we map the ontological foundations of our research.

Research questions

The ontologies underpinning our work should define the questions we choose to focus on. However, sometimes even with an intersectional worldview it is easy to slip back into two dimensions when we think about practical questions for an empirical project. To make your research questions more intersectional, check that you are allowing for difference and ‘asking the other question’, where necessary, about your topic. For instance, in a project on the under-reporting of sexual violence, make sure your questions incorporate and acknowledge different understandings and experiences of law enforcement. If you are researching gender equality in parliamentary politics, understanding ‘women’ as a homogeneous group risks allowing the success of some white middle class women to conceal the continued struggles of those who do not fit this category. Make sure your questions are designed to avoid this pitfall: this might involve asking specifically about particular groups of women in the political system. It may also be necessary to interrogate your ontology of ‘progress’: if this is defined as any woman holding political office, you may not be ‘asking the other question’ about how politics and policies led by privileged women might affect others who are more marginalised. A more intersectional appreciation of ‘progress’ might be necessary, and you should frame your questions accordingly.

Sampling

We should usually aim for diverse samples in our empirical work. However, intersectional research can be done using a limited and very specific sample, as long as you are honest about it. In fact, specificity can be a strength. Your desired sample will sometimes be dictated by your topic and what you aim to explore: if your research brief is to understand gendered street harassment in broad terms, for example, you will need as diverse a sample as possible. Women are sexualised in varying ways depending on intersecting categories such as class, race, disability and age, and gender-nonconforming people are also subjected to street harassment which has different dynamics. Often in qualitative research, samples are convenience-led and we must work with what we are given. Imagine you are asked to study a local women’s yoga group and you find that it is exclusively white and middle class. If approached in an intersectional way, the specificity of this sample could add depth to your research, allowing you to investigate how whiteness and class privilege are articulated in, and police the boundaries of, the space.

Knowledge claims

Many research projects in Gender Studies are grounded in the epistemology of experience, as a challenge to more masculinised, positivistic approaches. This is often situated within a testimonial politics focused on allowing more marginalised people to speak for themselves. If you are familiar with intersectionality theory you will already have a critical appreciation of terms such as ‘women’s experience’, knowing that this is not unitary or static and that to ground your research in such a principle may implicitly privilege the narratives and concerns of the dominant. In an intersectional research project you need to engage critically with different epistemologies in deciding where to locate yourself, realising that all knowledge claims are partial.

With this in mind, when you derive conclusions from your data make sure they are not over-generalised and that they are appropriate to your sample. In the project on the women’s yoga group, for example, you should not be making claims about ‘women’s experience of yoga’ but much more precise points about this particular white, middle class community of practice. This does not preclude raising broader questions or linking your work to more general themes: for example, the relationships between whiteness, privilege and the appropriation of Eastern physical-spiritual traditions in the West, and the historical and geopolitical contexts in which these are formed. However, you must be clear on what your particular dataset confirms, what has the status of interpretation and what needs to be left unanswered for now.

You should also ensure that you are not just generalising about your sample when there is differentiation within it. Imagine you are researching with a small group of sex workers, many of whom have extremely negative experiences of outreach and support services. You could derive legitimate conclusions here about sex worker stigma and judgment in the statutory and third sectors. However, an intersectional approach would require you to think about additional factors which might be at play. It might become apparent that the sex workers reporting the worst experiences are women over 45, linking to themes around how age, gender and sexuality are co-produced. You might begin to develop an analysis around perceptions of sex work as ‘sex’ rather than ‘work’, and how this interacts with the desexualisation of older women.

***

After doing all the above, you may end up feeling completely confused and as though you are unable to say anything at all. Congratulations! You have started to do intersectional research. The challenge for all of us is how to hold on to the complexities of social life with its multiple dynamics of privilege and marginality, while constructing narratives through our research which are engaging and intelligible. You will never, ever see the finished picture: but if you are lucky, you will get to be part of the process of finding a piece.

Advertisements