Thoughts on context and bias

The importance of context is always emphasised when thinking about how to present archives to researchers. At a recent seminar series I attended in the beautiful town of Lewes, East Sussex (pictured), Mike Savage of the University of Manchester talked about a well-known social survey by Elizabeth Bott, carried out in the 1950s, where 32 couples were interviewed about their relationships. Much of the contextual material was left out of the resultant book, so it was effectively stripped away from the findings. But closer analysis of the survey shows that the selection of the couples themselves was significant – the notes (unpublished) reveal why people volunteered for the study. There was quite a long process of application and most people who ended up taking part had interest in the research as a social activity. This is an important piece of the whole picture and would have had an effect on the findings. The research process itself is an important part of the whole picture.

Social scientists need to find methods to extract key findings from diverse archive sources, often covering long periods. Mike referred to the need to avoid the ‘juicy quotes syndrome’ and talked in detail about sampling methods, all of which have their pros and cons. He referred, for example, to ‘trend analysis’, which strips out the contextual detail (e.g. economic indicators, studies of changing attitudes). Processes and methods get forgotten about.

Archived qualitative data does not allow this abstraction from context and hence cannot deploy representative or aggregate findings. In this sense, qualitative data may have something to teach the social scientist in terms of the importance of context.

Archivists need to think carefully about the whole picture: what they are presenting to users and what they are leaving out. The whole question of subjectivity is a complex one. The social scientist must build the biases of inquiry into their analysis of qualitative data, and this distinguishes it from quantitative data. There is a need to develop clear analytical strategies to allow rigorous yet partial examination of such data – it is important not to give a false sense of the completeness of the data.

At the seminar, there was a great deal of discussion about methodology, the bias of the archive and the life of the archive itself. A particularly interesting talk from Carolyn Hamilton of the University of Cape Town referred to ways of using archival sources to study pre-colonial South Africa. The colonial archive is itself an expression of the power and dominance of the ruling elite – so what can it meaningfully say about the indigenous population? It is profoundly contaminated as evidence, and yet by the very act of proclaiming their dominance, the rulers shed light on those they claim the right to rule. In fact, the colonial archive brims with material germane to the pre-colonial past, but it is important to think about how to approach it and analyse it. Historians tend to study the archive ‘against the grain’ in order to mine it against its basic bias.

A similar situation of bias, although in a very different context, occurs with a community ‘archive’ website such as MyBrightonAndHove: Jack Latimer of QueenSpark Books talked about how this Website has become a very successful community website where people post images, stories and comments about their local community and history. It is very active, with around 1,300 visits per day and around 10-20 comments put up per day. But of course, this is also a skewed history – maybe a history that is born out of nostalgia, and obviously a self-selecting group of people.

John Hay, of the University of Wolverhampton, gave us a very engaging presentation about archives relating to deaf people and deaf culture. One thing that struck me was his wish to have an archive that represents the achievements of deaf people within society – here we come to another sort of bias. This does, of course, sound like a very worthwhile idea, especially, as John explained, when you consider how the deaf have been treated in the past, pretty much as second class citizens and victims of an affliction. But it does raise the question of whether an archive should have a goal of celebration or creating a certain image. Should it actually seek to gather any and all materials and artefacts that reflect the history of deaf people in the UK? Or is it perfectly valid to want to create something that is intended to be positive and affirming?

Archives may be a result of discourses and may in turn mould discourses, which in turn may give shape to practices that shape the archive. This, as Ann Cvetkovich of the University of Texas postulated, could be thought of as the public life of archive. If we accept that the archive has public life, then maybe it requires methodologically its own biography. The Archive acquires a provenance, is a part of the history of institution housing it. The Archive itself could be seen as a biographical subject.

Use of archives by social scientists

I have just attended two seminars as part of a project on Archiving and Reusing Qualitative Data: Theory, Methods and Ethics Across Disciplines. They provided a great deal of food for thought, as seminars like this so often do. These seminars were particularly valuable because they drew together academics, particularly social scientists and archivists. Many of the participants were oral historians, and the challenges of oral history ran through many of the talks.

When archivists think about archival theory and description, they are generally thinking about archives as materials ‘created by an individual or organisation in the course of their life or work and considered worthy of permanent preservation’ (my quotes, to indicate that this is a classic definition of archives). But if we think about archives as any records considered worthy of preservation and with value for future researchers, then we can expand the definition to include records that social scientists refer to as archives. For them, archives are often data sets, created by researchers in the course of their research and then, possibly, reused.

Social scientists do not necessarily think in terms of business records or personal letters, or archives as a reflection of personal or organisational activity. They think in terms of longitudinal studies and oral histories; quantitative and qualitative data. These are archives that generally are created for the purposes of research, and so the perspective is rather different to those created in the course of individual or organisational activity. We have the UK Data Archive which has ‘the largest collection of digital data in the social sciences and humanities in the UK’, and this houses the History Data Service which ‘promotes the use of digital resources, which result from or support historical research, learning and teaching’, but I don’t think that there is a general sense amongst archivists that these are part of the archive community, in the sense that trainee archivists don’t really think about working for a data archive, and arhcival theory doesn’t appear to really encompass this type of archive. Certainly social scientists clearly see archives as both data archives (data sets) and traditional archives (archives as reflections of past activity), and the fact that the two were not explicitly distinguished during the seminars was striking in itself.

It may be that data archives require different ways of thinking to ‘historical archives’, in terms of how they are organised and managed, but now that archives are increasingly digital, and as all archives are a valuable source for research, surely there is sense in the two communities moving closer together?