Our first archive visit was to the National Co-operative Archive. This was a great place to start our research, in part because of the way the archivists have helped to make their resources accessible. As I'll explain in this post, the level of detail used in the archive descriptions enabled us easily to identify items which might be of interest and to order them from storage before we arrived. This won't be the case at other archives we visit, so I'm going to use this post to explore how archivists' work can support research, whilst recognizing that such efforts aren't always possible. Since the nature of our research is so specific, i.e. finding out how those experimenting with economic systems might be thinking about time, we will have to put in some additional effort to understanding both the domain and the archives we use in order to get the most out of these resources.
There are 6 ways of searching the Co-op archive online, including an image search, which we did not use. In line with most users' preferences, the default option is a simple free text search. We used this to identify the majority of the items we used, by using keywords including: time, hour/s, work, working, clock/s, future, nature, change, evolution, progress, history, environment, flood/s, earthquake/s, disaster/s. However, as our list of search terms suggests, the disadvantage of this method is that even changing the word from singular to plural brings a different list of results, because the search engine simply looks for exact matches anywhere in a description. This creates a number of problems when trying to research time in the archives. A good example was our search for 'clocks'. Anyone looking for clocks may also be interested in watches, but because the word has another, more common meaning this can confuse the results. In our case, all the item descriptions that included the word 'watch' were for images involving onlookers. This means that it is possible that the Co-op archives include information about pocket and wrist watches (these were often retirement gifts for long serving employees in many businesses, for example), but unless the archivist has included the word in the item description it will not be accessible via a keyword search.
One of the ways archivists try to deal with this and other problems associated with keyword searching is to 'index' archives - that is to say, to associate the description with a number of 'authority terms' which the cataloguer decides have particular relevance for the unit being described. At the Co-op archive there are indexes for names (of people and organizations), subjects and places. Authority terms are created as a separate exercise and can involve considerable research, to ensure that, for example (and, not, as far as I am aware, featuring in the Co-op Archive), Robert Smith, equestrian, son of Harvey Smith, is not confused with Robert Smith, musician, lead singer in The Cure. When used consistently, indexes can often be the most efficient and effective means of searching - but they are labour intensive to create.
Of the three types of index, subject indexes are perhaps the most problematic for archives. From the point of view of an archivist, archives are not in the first instance information resources because - unlike books - this is not the purpose for which most records were created. Most records are created as evidence - to provide a persistent representation of a time-delimited event. Minutes from meetings provide a good example of this. As the influential archival writer Sir Hilary Jenkinson (1882-1961) declared, they 'were not drawn up in the interest of, or for the information, of posterity'. The American archival theorist T.R. Schellenberg modified this perspective, by suggesting that archives have two sets of values: primary values for their creators and secondary values for other users, which come into play in particular when records, created for current purposes, cross the 'archival threshold' and then may be used by a wide variety of users for a wide variety of purposes. Archivists want to open the records in their custody to the widest variety of users and uses. Subject indexing is intended to assist with this but archivists cannot predict all the possible topics future researchers may want to investigate. Trying to represent the subject matter of a document from the perspective of its creator/s follows archival theory's traditional emphasis on provenance (discussed further below) and thus appears theoretically straightforward, although it inevitably involves difficult decisions in practice, as anyone who has ever tried to describe a photograph without knowing what the photographer was trying to capture will recognize. Extending this indexing to include possible research topics is fraught with difficulties and is thus rarely attempted: a more common approach is the 'Subject Guide', whereby an archivist gathers together potential resources for a researcher interested in a particular subject area. However, when embarking on this research, we were aware that the elusive nature of time and temporal awareness was one of the difficulties we would have to overcome (which will be explored in our 'Methods Festival' in June) - there are no useful archive guides on temporal research to assist us.
Archivists' descriptions try to represent the context of creation of the records - this is what is known as 'provenance'. Whilst archives can be used to support an almost infinite variety of research topics, which will inevitably change over time, it is generally accepted that our understanding of the people and processes that created the records in the first place are likely to remain relatively constant and that this knowledge is vital for interpreting the records' historical meaning. This means that archivists try to maintain the original order of the records, and list them according to their creators. Thus at the Co-operative Archive, all the records of the Crumpsall Biscuit Works are described as a single group. They are, of course, not all the records created - the vast majority have not survived, but enough remained for us to get a sense of the importance the Co-op movement accorded both to worker welfare and to production efficiency in this model factory. In an archive where descriptions are less full than at the Co-op, the only way we may be able to identify records relevant to our research will be by identifying the types of organizations and the historical circumstances which might produce records of potential interest. Again, by starting with an archive with a very detailed catalogue, we have started to build up a picture of what sorts of series of records are likely to be of particular significance to our research.
In a later post, I'll explore how 'Time' is represented in other forms of resource discovery, such as library catalogues.