Doing Conversation, Discourse and Document Analysis. Tim Rapley
public).
Assorted handwritten and typed notes to myself, memos and quotes.
For me, all these materials were ‘my data’. Or rather, as I prefer it, all these materials made up my archive. This archive, combined with conversations with the research team, friends, strangers, watching television, listening to the radio, reading novels, et cetera, and most importantly, bolts from the blue (often over a large cup of coffee), enable the production of specific research findings and papers.
I could easily have called this whole section ‘Generating data’. And I would have probably focused on roughly the same things. However, the scare quotes I have placed around the word ‘data’ may suggest I have a problem with the term. I am never quite sure when I am conducting my own research what actually is my ‘data’ and what is not my ‘data’. Is a quote I take from a recording of a doctor–patient consultation data? Whereas the quote I take from a social science article discussing doctor–patient consultations is not data? I use both to develop my argument, so for me both are data. Equally, are the topics covered in the interview data? Whereas the interview schedule itself and the reading and discussions that led to its development are somehow not data? All these areas of activity are central to producing my arguments.
Rather than just thinking about ‘generating data’, in any narrow sense, you need to think about generating or producing an archive – a diverse collection of materials that enable you to engage with and think about the specific research problem or questions. On a practical level, this means collecting and managing an array of different materials. Obviously, what materials make up your archive is directed by both your specific research question and your theoretical trajectory. In order to briefly review the possible sources of researchable materials, I am going to divide the following discussion into two areas: document-based sources and audio- and visual-based sources. I should note that this is a wholly arbitrary division. For example, when an audio file is transcribed it is, in one sense, translated into a document. However, this arbitrary or working division is just that, and is based on my need to offer you an accessible and manageable story.
Document-based sources
On the face of it, this would appear the easier archive to generate. And to be honest, it can be. As a lot of the material you could be working on will already be in the ‘public domain’ – either published on paper or on the web – you do not have to go through the process of getting consent to use the material or recruiting and recording often very busy people. Your major considerations are often related to how to initially discover, then source, and then make some form of recording of the documents. In this section, I am just going to offer a few examples of documents you could use, and talk about some of the general pitfalls and problems. In Chapter 9, I shall offer some more detailed cases.
The most ubiquitous and accessible source of documents is newspaper and magazine articles. Both web- and paper-based content are a massive potential resource for most academic projects. You only have to think about the diversity of weekly and daily local, national and international newspapers, as well as the ever-growing numbers of general and specialist magazines we are confronted with on a day-to-day basis, to realize how much material is easily available for analysis. You can learn a lot about the trajectory of culture and institutional practice through engaging with these materials.
Let us take, for example, a recent phenomenon – articles and magazines specifically targeted at men and their relationship with their bodies. When you look at the print cover of a monthly magazine like Men’s Health you will be witnessing specific discourses of masculinity. Each cover seems to have headlines like ‘how to tone your stomach in one month’ or ‘six exercises to develop a six pack’, alongside pictures of men with washboard stomachs. This raises various questions, including: What (new) versions of masculinity are being promoted? What (new) connections are men meant to make between their bodies and self-identity? Are men (now) objects of a female gaze? I hope it is easy to see how an analysis of such headlines, and the articles that they refer to, might raise some very interesting questions about new formations of masculinity.
Research has also focused solely on newspaper headlines; for example, Lee (1984) offers a rich analysis of the headline ‘Girl Guide Aged 14 Raped at Hell’s Angels Convention’. Among many things, he highlights how this headline works to attract our attention and persuade us to engage with that story. The headline raises a puzzle: how is it that these two categories of people – ‘Girl Guides’ and ‘Hell’s Angels’ – not routinely connected, are together? How is it that they were in the same space? And this is not any space, this is a specific type of space with specific un-Girl-Guide-like associations and activities, this is a ‘Hell’s Angels Convention’. In part, we may engage with this story to find a solution to this puzzle. That we engage with this as a potential puzzle in search of potential solutions, is intimately dependent on some shared understanding of culture and cultural categories. Research has also focused on the text (and images) in such things as advertisements, magazine front covers, dating or ‘would like to meet’ adverts, as well as magazine articles and newspaper reports.
So work has ranged from focusing on a single headline, a single article or a single publication to a very large number of national and international newspapers. With studies that work with a larger archive, you obviously somehow have to make your archive manageable. For example, Seale (2002) studied how cancer is portrayed in regional and national newspapers across different cultures. His archive was based on one week’s newspaper publications from the English-language press all over the world that contained the words ‘cancer’ and ‘leukaemia’ or ‘leukemia’. His choice to concentrate on just one week’s press was driven by various practical concerns, including:
Generating a manageable number of articles. His initial search generated 2,419 English-language articles.
The cost of identifying and gathering this amount of material. He went through a specialized company that collates the material and emails it to you. These days, you can use various news-based search engines and databases, often for free.
He only worked on articles written in English due to the cost of and potential problems with translation.
You need to be aware that simply finding, collecting, sifting through and then physically and/or electronically working with a large number of articles can take large amounts of time and money.
Another massive potential source of documents for research are academic publications. These can range from publications in the areas of the sciences and medicine, to the arts and humanities and the social sciences and can be either more historical or more contemporary documents. Obviously, the term ‘historical’ is contingent, and can refer to anything from documents produced ten years ago to, say, 300 years ago (and more). A lot of contemporary work is focused on analyzing the contemporary or historical academic articles and books of other researchers. Much of this work rarely offers a novice reader any hint as to how the materials included were discovered or selected (or even analyzed).
As with much research work, analyzing academic publications involves a lot of common-sense practices and some related detective work. Emerging from some research I was doing on how social researchers conduct qualitative interviews, drawing on audio files of actual interviews, I wanted to discover what researchers are told interviewing should look like, what prescriptions of ideal practice were being promoted. I felt this was important as qualitative interviewing had become the method of choice in the social sciences, especially in sociology, and so a vast array of literature on ‘how to be a good interviewer’ existed. I already knew of some of the key texts to focus on. In part this was because these were the articles and books that ‘everybody’ seemed to reference when justifying their choice of using interviews. I went to these articles and books and looked at their reference lists and followed the trail back, a practice known as ‘reference chaining’. I then found these articles and repeated the process. It is a hit-and-miss affair; some were highly relevant whilst others were redundant.