Raising the profile of women in Discovery

Caroline Catchpole – Digital Development Officer, November 2023

Caroline Catchpole explores a recent data enhancement project she undertook to research and enhance authority files of women in The National Archives’ online catalogue, Discovery.

Caroline works in The National Archives’ Archive Sector Leadership team, focusing on providing support to archives who use Discovery, and projects to increase digital capacity and confidence in the archives sector.

The importance of authority files

Discovery is a treasure trove, and not just for the wealth of catalogue descriptions. There are also over 273,000 authority files of Record Creators, giving users an authoritative source of information on specific individuals, organisations, businesses, families or manors, as well as an understanding of the location(s) of collections created by or connected to them.

An authority file is a form of authority control archivists have increasingly sought to apply to cataloguing practice. Authority control is defined (by the Society of American Archivists’ Glossary) as:

‘the process of establishing standardized names and index terms for use in archival or bibliographic description and ensuring their consistent application’

At The National Archives, we use the term ‘Record Creator’ and authority file interchangeably.

Building a complete picture

The creation and management of the authority files are in part a function of our work under the Historical Manuscripts Commission (HMC) Royal Warrant.

In 1869, the HMC was formed to document the location of records and papers in private hands. The work of the Commission raised awareness of the importance of privately owned material for a more complete picture of the nation’s archival heritage and history.

In 1945, following recommendations of the Master of the Rolls’ archives committee, the National Register of Archives (NRA) was set up under the Commission to carry out a more comprehensive survey of all privately owned collections and papers. In 1995, the NRA indexes were made available online and list where archival collections relating to the Record Creator can be found.

The data challenge

The NRA has been curated over the last 70 years and today there are 53,451 people authority files available on Discovery. However, there is a seismic gender split, with less than 10% of the authority files being of women.

In the UK, the NCA rules for the construction of personal, place and corporate names (1997) sets out how to construct an authority file. The International Standard Archival Authority Record for Corporate Bodies, Persons and Families (ISAAR(CPF)), (1996), provide a general framework in which the NCA Rules may be applied for consistency in archival descriptive practices.

It is common practice to include an epithet in the authority, which describes what the person was known for in their life. It is commonly their profession or for nobility or military, their rank and/or title.

However, amongst female authority files it was common practice to define a woman by her relationship to a man, most notable as ‘wife of’ but also including sister, daughter, granddaughter, goddaughter, niece, mother, grandmother, aunt and mistress. This practice derives from NCA rule 2.5D.8 ‘Relationship to another person’ which states:

‘The relationship of one person to another, better-known individual or to someone whose authorised name has already been established, should be used for identification only where no other epithet would serve.’

Of the 5,509 female authority files, 758 had epithets where they were either exclusively described by their relationship to a man or with their profession and their relationship status. This represented 11% of the overall female authority file dataset.

Some examples included:

  • Cheverton, Vera Brooke, (1898-1974), wife of Frank Cheverton, engineer, Indian Army
  • Hawarden, Eleanor, (fl 1941-1966), writer, wife of Julius Lewin

Defining project requirements and methodology

I recently worked on a data project to address this inherent bias and to give a more representative picture of the professional and personal achievements of these women who have historically only been described in terms of their relationship to a man.

When the authority files were created, the construction made sense as a workflow in following NCA rule 2.5D.8. However, I saw an opportunity to enhance this dataset to centre personal and professional achievements. The aim was to give equity to these women whilst still preserving the relationship data, as this remains an important reference point for researchers.

I created a straightforward set of requirements to enhance the dataset:

  • Keep the relationship data – just move it elsewhere
  • Accept feminising titles as an epithet – e.g. ‘Duchess’ instead of ‘wife of’ Duke
  • Make other enhancements where possible, eg add maiden name, birth/death dates adjusted if required or dates confirmed (where c. or fl/ dates were recorded) to improve the overall accuracy of our dataset.

Where the man has his own authority file on Discovery, I linked them with an explanation of the relationship, so users can navigate between linked relationships, enhancing the discoverability of each entity.

Where the man did not have a Record Creator page on Discovery, I was able to use a field in our database, ‘Alternative Name Form’, that has the option ‘Married To’. This meant I could preserve the data in the event I wasn’t able to link two entities together.

I undertook desk research and used sites such as the Oxford Dictionary of National Biography and Wikipedia to learn more about the careers and achievements of these women, which enabled enhancements to create a more representative epithet.

Outcomes and opportunities

One woman included in the project was:

The original construction of her authority file was:

  • Gordon, Ishbel Maria, (1857-1939), wife of 7th Earl of Aberdeen

You can see that not only was I able to enhance the data to correct her surname but also to convey the incredible professional career she had with many achievements. The relationship data was preserved by recording it in the ‘Alternative name form’ field.

I was able to enhance data on 529 women. The outstanding batch were not easily researched with straightforward internet research, so an opportunity remains to conduct more in-depth research to see if this data can also be enhanced.