Today I want to write about the claim we, qualitative researchers, often make about our research. We say that we give voice to those who speak to us. But do we?
The post was triggered by this tweet:
This is why I rarely take part in qual res. Researchers often don’t honour participant voices, maybe don’t hear them above their own voices?
The question of what it means that qualitative research can give voice to the participant seems easy to answer. After all, we focus on individual narratives, we open spaces for sharing experience, the individual has the opportunity to make their own relevancies as we celebrate them. Qualitative syrup is flowing, as we, the qualitatives, reassured in our superiority, get on with empowering the participants. Obviously!
But I think that when we stop winking, close the door and be honest with each other and ourselves, and actually do stop to think about what it actually means, things get a bit hazy. So, I would like to offer brief account of what it might mean to give voice to those who participate in our research and, needless to say, what it means to me.
I think there are two aspects of the claim of giving voice. First, it’s the issue of data collection. Although not easy at all to implement, at least the matter is fairly straightforward to think about. Qualitative research ‘offers the space’ for the story; in contrast to the quantitative methods, it allows the participants to speak any way they want. That’s the official spiel at least. I also use it. I also have significant long-term experience in research interviewing and I think my interviews were good only in one project. In fact I wrote a post on interviewing misery. I think the interviews were very good because I was emotionally engaged with my interviewees, but it also had consequences which I had not been prepared for.
When I had started interviewing years before, reading transcripts of my interviews was just painful. I interrupted, cut the informant off, interpreted, knew better. Bloody hell, it was awful. Over time, I learnt to concentrate and not ask the question that had just sprung into my mind, wait for the informant to end. And it took years to achieve. So, do I open the space and all the rest of it? At the beginning I certainly didn’t. Now? I think I begin to. Mostly, because over the years, I’ve learnt to shut up. How many of my fellow qualitatives reflect on their interview practices? I don’t know and I prefer not to guess.
And then there are our questions. No, I’m not going to have another rant about the vacuous insistence on asking open questions, I prefer to ask how conducive my questions are to opening the narrative space. How often we ask such questions of our interviews, I don’t know. I do know that we very rarely talk about such things as we present our research at various conferences. Time pressure is an obvious reason, but is it always? I wonder how many of us actually reflect on what we do in the interview. And I want to be very clear – this question I direct also at myself.
There is much more here that can be said, of course. Ethnography, for example, is a minefield for such considerations. But let me finish with a point which is also relevant outside research activities. Just because we have wonderful rules and accounts of how to carry out interviews (research or otherwise) doesn’t mean that ‘we’ actually follow those rules. In other words, written guides into interviewing do not offer any insight at all into actual interviewing practices. I thought I’d just make this point.
Now, the second aspect of ‘giving voice’ is analysis. We come to our data with a conceptual frame which impose on them. Preserving ‘the voice’ within such a frame is not obvious or straightforward. I see it mostly in terms of the status of researchers’ interpretation of what participants said. There are two kinds of methods here, I think. On the one hand, there are those which primarily focus on formal aspects on what our informants say, on the other, there are those which tend to offer interpret what our informants say and, in some cases, mean. Let’s consider them.
In the first group are assorted versions on (linguistic) discourse analysis. Whether narrative analysis focuses on the structure of informants’ stories, Critical Discourse Studies on the use of agentive structures, or Conversation Analysts on the structure of interaction (I do realise that for some putting CA into linguistic DA is a slur – I beg forgiveness), all these approaches focus on what is objectively (again, I do realise the issues here) available. In other words, when I focus on passive voice, it is more than likely ‘we all’ can identify what passive voice is and where to find it in the data. And so, none of such approaches make claims what an informant’s story means, let alone what the informant meant by it.
That said, also here our interpretations, especially insofar as our suppositions as to the source of informants’ narratives, their social and cultural contextualisations are prone to imposing perspectives with which our informants might not agree. What is crucial, however, is that when we offer interpretations, we see our data as part of larger discourses and do not make any claims as to the individual story. If we do blur ‘the voice’, we blur it at a social rather than individual level (excuse the distinction – it’s only a blog).
As I said, the other strand of qualitative analyses focuses more on interpreting what a story, utterance, or other stretch of discourse means. As researchers in Interpretive Phenomenological Analysis and (versions of) thematic analysis offer insights into what is behind the narrative, such interpretations may impose themselves on a story not only blurring the voice of the individual, but also, potentially, replacing it with a different ‘voice’.
Indeed, I wrote before that whenever I review a paper based on thematic analysis or IPA, I wonder about how the interpretations are arrived at. I can add that I also wonder about who is talking to me. Is it the informants or is it the researchers who decided to offer a particular interpretation of what their informants said? And when I read the extracts quoted in the papers, I always come up with different themes and interpretations. Where is the voice of the informant? I don’t know.
As can be seen, I think that ‘giving voice’ can be difficult in either of the approaches. And given that I am more sympathetic to the former, I am not going to extol its virtues and say that when we, discourse analysts, don’t honour our participants’ voice, it’s better, because it’s at a different level. Is it better? I don’t know and it’s not for me to say. To be completely honest, I have been told before that this whole focus on ‘grammar’ is nonsense anyway and it misses the point of what the informant said. Go figure.
So, what about the voice? Are there solutions? Yes, we can ask the participants (and informant participation in the analysis is an issue for a different text), but does it solve the problem? Am I the only expert in what I say or write? I’m not so sure, actually. And so, we are left with a critical look at the assumptions we take to our data. But we can also make them explicit and make an account of how we arrive at an interpretation. Why do we claim that the informant said A and not B. In such a way at least make our reasoning transparent. For me, I can say that I often wonder about what authors of the suicide notes I have been writing about lately would make of my analyses and to what extent I ‘give voice’ to them and do justice to their sensitivities. I will never have answers, but it is important to me.
But I think there is one other way in which we honour the narrative. We quote from it, in the form we got it. No sanitisation, no embellishing, no removing of the ‘bad’ words. We preserve our informants’ voice by offering it both in our publications and….And here we come to the very difficult issue of sharing qualitative (and especially interview) data. This is probably the best way in which we can ‘give voice’ to those who haven’t got it. How exactly to do it is not clear to me (yet); the ethical issues are complex, but they can be solved. I think ‘we’ should give more and more thought (and action) to sharing our data. This is the best and most certain way of ‘giving voice’.