Timeline design for visualising cultural heritage data: Olivia Vane

The core of this research project is exploring new ways of interacting with cultural data. By ‘cultural data’ meaning the result of digitising cultural institution holdings. Cultural institutions include museums, galleries, archives and libraries. Their holdings are diverse, consisting of different types of objects and media ranging across artworks, artefacts, specimens, books, documents. Cultural institutions are increasingly digitising their holdings (Nauta & van den Heuvel, 2015:p.4). While the primary reason institutions digitise their holdings is to broaden and enhance access to the materials, there are also assumptions that once digital, these materials can be read, viewed, experienced etc. in new and creative ways (Hughes, 2004:pp.8-17). Essentially, there are things you can do with cultural collections as digital data that you cannot do with the physical items; one such possibility is data visualisation. But before discussing cultural data visualisation, it is helpful to characterise digitised collections.

Cultural data

Typically, when a collection is digitised, the associated institutional catalogue or index for the items is converted to digital data and it is now common for institutions to enter cataloguing information directly into a database. Catalogues perform a number of roles. They are finding aids for physically locating items, provide historical descriptive information necessary for understanding the material, flag related material in the collection, and record provenance (Riley, 2017:p.5). The catalogue can also serve an administrative function; institutions use these systems to track their acquisitions, exhibits and loans. The completeness and detail of catalogue records very much varies. The institutional catalogue both helps interpret a collection and describes the relationships and connections of items to each other.

The items themselves may be digitally reproduced: the objects and artworks may be digitally photographed, the texts may be transcribed to digital text files. Cataloguing information can be dynamic as new acquisitions are made and records are fortified or revised. Many institutions have an ongoing digitisation programme, describing their digital collection as a work in progress, constantly being improved and added to.

Cataloguing & classifying

Pertinently, the kind of data found in digitised collections, while it can be treated quantitatively, is not inherently so. It is largely nominal (as in, assigning names). As defined by Meirelles (2013:p.187), nominal data is distinguished on the basis of quality; examples include objects, names and concepts. Categorisation plays a major role in manipulating nominal data, and shared characteristics allow grouping.

How data is structured and what is recorded informs what can be done with it. Visualising data in a particular way is dependent on there being suitable attributes in the data, and the data being in a suitable form. What do cultural datasets allow or afford? Factors include: how are the objects to be catalogued divided into records (is a sketchbook assigned a single object, or every sketch within it, or both)? What is the nature of the classification system? What sorts of categories have been attended to? How is the metadata produced—manually? automated? How flexible is the dataset to change/ augmentation? To be clear, cataloguing is the general process of creating metadata representing cultural items. Classification meanwhile, which may be part of the cataloguing process, involves assigning that item to a set class in a classification system.

There are two important points here: that cultural datasets are not necessarily a definitive representation, and they are not set in stone. They can be changed, reworked, enriched—and the desire to use a particular visualisation template can be the impetus for changing cataloguing approaches. (Data can also deteriorate. For example, moving electronic records from one institution to another can introduce errors because of incompatible data systems).

Classification across cultural institutions

Different types of cultural institutions have traditionally had different attitudes towards, and practices of, classification and cataloguing, leading to datasets with particular characteristics. Robinson (2014) identifies fundamental differences between libraries, archives and museums.

Generally, libraries aim to provide broad access to their collection using a standardised categorisation scheme with rigid hierarchical subject definitions. These subject definitions form a “sort of epistemic cartography - mapping knowledge”(Olson, 2001:p.652). Dewey Decimal Classification is an example of such a scheme. (Although discrete digitised collections from libraries, rather than a broad view of their holdings, are often available as standalone datasets with flat data structure).

In contrast, for archives, the most important concern is “retaining the relationship between the documents and the institutional functions and activities that gave rise to them” (Robinson, 2014:p.418). The documents must be kept in the order in which they were received: the principle of provenance. This has an implication for data structure, in that archives often group documents in series (groups of similar items), rather than just indexing and describing items at item level. The 17 digital records for the Babbage Papers archive (relating to computing pioneer Charles Babbage), for example, follow this pattern; the data structure is a hierarchy of grouped documents.

Highly curated narrative

Museums have long employed timelines in analogue form for exhibition storytelling, for instance the Eames’ history wall for the 1971 exhibition ‘A Computer Perspective’ (Eames Office, 2014). The ultimate ‘curated’ digital timeline involves ordering a handcrafted, limited selection of digital media along a timeline—through a single, linear sequence— with additional custom text explaining the significance of items and the connections from one to the next in a clear, explicit narrative (eg. Mucha Foundation, 2012; Anne Frank House, 2010; McCloskey & Wei, 2014; Darwin Correspondence Project, 2017a—see Figure 18 and Figure 19). These designs are handcrafted, and tailored to display and storytelling rather than free exploration of collection data.

The Metropolitan Museum of Art’s (2000) ‘Heilbrunn Timeline of Art History’ organises content by time period and geography with the aim of “telling the story of art and global culture through the Museum’s collection”; it combines essays, chronologies, groups of collection items and timelines. Timelines here are not used to map items, but to contextualise them by visualising relevant historical periods (both political and from art history)—see Figure 20. This is a highly authored, encyclopedia-like platform rather than a search/browse interface, where the timelines provide context to curated groups of items.

Vane, O. (2020). Timeline design for visualising cultural heritage data. [online] Researchonline.rca.ac.uk. Available at: http://researchonline.rca.ac.uk/4325/1/TimelineDesignForVisualisingCulturalHeritageData_OliviaVane_redacted.pdf [Accessed 27 Feb. 2020].

http://researchonline.rca.ac.uk/4325/1/TimelineDesignForVisualisingCulturalHeritageData_OliviaVane_redacted.pdf