Frances Madden (orcid.org/0000-0002-5432-6116)

Posted 2 December 2020

The Victoria and Albert Museum (V&A) is a museum of art and design and houses over 2.3 million items including fashion, photography and sculpture. It is also home to the National Art Library and has an extensive archive collection.

The museum uses several different types of identifiers across the entire collection including the museum, archive and library collections. Museum or object numbers are assigned to collection items, but part numbers can be assigned to removable pieces of collection items, for example, a chest of drawers will have a museum number but every drawer will have a part number. Archive collections are catalogued to series level, which can include thousands of individual documents and pages, so many items within a series would not have an individual identifier assigned to them. However temporary identifiers are assigned to archival items if they are to be displayed in an exhibition or loaned. The library collection has two potential types of identifier: barcodes, which do not integrate with the library management system provided by OCLC, and shelf numbers. All of these identifiers address an internal use case but none are used to provide an online identifier or have any governance to ensure their persistence. Items in the collection can move between different areas resulting in their museum number being changed, leading to items having multiple numbers.

Collections management at the V&A, similar to many other museums, has a number of human processes associated with it, including manual accessioning into the collections. Practices also vary across museums about how museum numbers and part numbers are assigned. In addition, when items are accessioned into the collection they can go through a bulk inventory process where a number of objects, such as all the items in a box, are accessioned together.

For the museum’s digital collections, the online presence of the collections, the collections website at the V&A contains information about more than 1.2 million objects, over 600,000 of which have images. The collections site uses a standard syntax to construct URLs. This system utilises the information from the collection management system and pulls it through to web pages dedicated to each museum object. These URLs have been in place since 2009 and utilise the database identifier from the collection management system and either the title of the item, the object type in addition to the artist or maker if it is known.

For example:

These URLs are actively maintained and this will continue when the digital collections site is migrated in the near future. There was some debate about changing the URLs to improve the search engine optimisation of the site but it was agreed to keep them in an attempt at persistence. While there is no official governance or policy around the persistence of these URLs, there is a recognition by the staff of the museum that it would be best practice to maintain them as they are a long established method of access to the digital collection. This decision to maintain the collection site URLs demonstrates the importance of institutional memory and awareness of an external context in maintaining these identifiers when there is no additional governance in place which globally resolvable identifiers can provide.

The V&A's use of identifiers highlights some interesting points around the requirements for PIDs in a museum setting, particularly one with a large and heterogeneous collection. Identifiers would need to be assigned to a large number of different types of item and to be able to accommodate hierarchies and relationships for complex objects but also to express different representations of them, e.g. digitised versions versus the physical objects themselves. The capability for cross-walking these hierarchies in different ways is also an issue if these identifiers are to operate across institutions. The fact that the collections’ online presence is managed entirely separately to the internal identification system demonstrates the conceptual difference between physical and digitised collection items. This also means the link between the physical and digitised object lacks persistence as the database identifier used to connect the digitised and physical object could change if the collection management system were updated and the historical identifier was not maintained in some way.

This conceptual difference is particularly evident with photographic collections where a photograph within an album and which has been digitised can be catalogued in several different systems. Whether something is catalogued in the digital asset management system or in the main collection management system will affect the metadata captured about itand whether it is treated as a digital or physical item, and in turn the metadata made available both to internal and external users of the item. In addition the V&A has a significant cataloguing backlog where batch processes were used to inventory items, which means that there are a large number of items which cannot be assigned identifiers automatically. Even within catalogued items, across museums the vast majority of museum objects have no digital presence, or only minimal metadata description and this is an issue which can stymie the adoption of PIDs for citation of objects and in internal processes.

The appetite for making collections available digitally has not always been high in many museums, where visitor numbers are a key metric and the research and digital use of collections is not given the same value. While there is some interest in tracking citation of museum items, the question of what should be included in the citation remains. Should it be the museum number or a fully resolvable PID which refers to the item? Should the citation even include the museum number? The V&A does not have any explicit guidance on this and follows the individual publication’s recommendations or requirements.

The V&A is active in the Linked Art project which aims to provide linked open data approaches to collection data using CIDOC-CRM as a basis. Linked open data provides one of the key methods in which identifiers can be leveraged and viewed to provide value to a museum collection beyond a research citation value. The connection of different collections together is an application of the national collection and demonstrates the vital need of identifiers within collections.