Two presentations in one

In the middle of the year I presented two conference papers:

– ‘What we talk about when we talk about things’ at Digital Humanities Australasia in Hobart, Australia (20-23 June 2016)

– ‘Mind the gaps: missing connections in museum documentation’ at the 24th General Conference of the International Council of Museums (ICOM) in Milan, Italy (3-9 July 2016)

This post combines those two presentations to provide a summary of some aspects of my graduate research.

We are all familiar with the types of displays presented in museum exhibition spaces. The two examples shown below are from Museum Victoria’s Love and Sorrow exhibition. Many types of item are brought together to produce meaning, including three dimensional objects, photographs, documents and text. The things presented are not expected to work in isolation. They are enriched and contextualised through aggregation.

Historians and other writers work the same way when preparing journal articles and books. This page, from Philip Jones, refers to a whole range of historical and contemporary entities to construct a narrative, including places, people, events, and archival records.

These examples align with the illustration initially developed by @gapingvoid showing ‘information’ and ‘knowledge’, and extended by @bestqualitycrab (aka Deb Verhoeven) to include ‘data’ and ‘humanities knowledge’. The move from data to humanities knowledge involves adding additional context to data, connecting up data elements, and adding context and weight to these connections.

But when we look at museum collections online we are often presented with what seem to be quite disparate collections of discrete objects. We might see lots of sets of very similar things, or things from different collections and time periods collected by different people and worked with by different curators and collections managers. It can be difficult for users to work out how this material fits together; it’s often hard to connect things in a way which produces the types of meanings and narrative (’humanities knowledge’) which help us understand the meaning and significance of objects in exhibition spaces and in books like those by Philip Jones.

When thinking about museums, we can split the preceding diagram into parts. Online museum collections often contain data and information without providing the connections and context required to turn those collections into knowledge. That is left up to the user, and each subsequent user has to go through the same process.

This is all, of course, something of a generalisation. Looking at a specific example, currently on display in the Melbourne Museum is a rope with harpoons from the Donald Thomson Collection on long-term loan to Museum Victoria from the University of Melbourne.

(At the conference in Milan, the first question following my presentation was: ‘What’s a dugong?’ Also known as a ‘sea cow’ it’s a wonderful, blubbery animal found in, among other places, the waters around Cape York Peninsula where Thomson collected his rope and harpoons.)

The ‘Summary’ for this object gives the user some general detail about dugong and turtle harpooning. Interestingly, this text appears in the ‘Summary’ field for multiple items – a potential issue for maintaining collection documentation – and is provided without references or links to additional information.

Underneath these descriptive passages comes a section of detailed, fielded metadata. All this information is valuable, particularly for pinning down the specifics of the object itself. But there are limitations. Clicking on ‘Donald F. Thomson’, for example, will bring up all the items collected by Thomson but not provide information on who he was or the context of his work. The other links perform a similar function, triggering a search of other items with that same fielded data. But if a user wants to locate richer material, a fuller description of dugong hunting, photographs, or other related context they need to conduct their own research.

This is not intended as a specific critique of Museum Victoria and its Collections Online site, which is an impressive interface, often featuring rich metadata and beautiful images with generous usage rights. It is more an example of the way museum collections are often presented more broadly – as discrete items with metadata attached, linked to various options for filtering and browsing other discrete items but rarely connected to other items, or to the sorts of context found in either exhibition spaces, catalogues, books or articles.

When this material is aggregated into library-based catalogues the ability to navigate to related content becomes even more limited.

Some museums attempt to use search indexing and facets to provide access to a broader range of content. The Smithsonian’s National Museum of American History has a ‘Collections Search’ which (as seen in the ‘data source’ section of the shaded column on the right hand side) includes results from the Archives Center and the Smithsonian Libraries as well as the object collection.

But when a user clicks on an archival item, they find themselves in a separate interface for the Archives, Manuscripts, Photographs Catalog…

… and likewise, if the user clicks on an object they find themselves in a Collections catalogue with no reference to archives, published material, or other context.

As noted earlier, museum collections online (and often in underlying catalogues) appear as discrete collection records with common fields, sometimes incorporating ‘authoritative’ subject terms, collector names or other data. Though the user can view these records as lists of search results, or move between individual records, there are few defined pathways available.

One of the primary reasons for this is the prevalence of the search box. With it comes a belief that, with the right text and keywords, users can bring up the content they need. While there is immense value in textual content (particularly structured, fielded content) when it is indexed and made available as part of search, it is often the only option.

A user can bring up the rope used for dugong hunting collected by Thomson using search. They can also find other content by Thomson, or other material from the same place. But if there are archival records, publications, photographs or other collection material directly related to the dugong rope – material that was created at the same time, in the same place, or on the same expedition – users are left to uncover the majority of these explicit connections for themselves.

Before the advent of search (and when sufficient resources were available), different techniques were employed. The page shown below comes from the Thomson microfiche catalogue* which in turn was based on over a decade of work documenting, describing and organising the collection.

The middle item – DT3324 – is the rope with harpoons shown earlier. The last line of the entry is ‘Refs: TPH2813-28; TPUB 5’. The TPH references are for a series of photographs Thomson took of Cape York people manufacturing hibiscus fibre rope. TPUB 5 is an article Donald Thomson wrote on dugong hunting on the cape** which includes among other things a lengthy and evocative description of a dugong hunt and analysis of various harpoon heads and ropes. While these references have been transferred into a text field in the in-house collections management database used by Museum Victoria, they are not available to the general user.

The items above and below include additional references – TFN 192; DT BOT 381A – to specific Thomson field books and botanical specimens related to these harpoon heads. Though not accessible via the microfiche guide, a reader would at least be aware that contemporary textual, photographic and botanical material is available related to these items. Researchers have been given pathways to other content they can follow to find out more.

Diagrammatically, then, our discrete collection records (the large nodes) can be joined by creating explicit cross-references or relationships (as seen in the microfiche catalogue), providing a pathway by which users can find known related content. That is, users are given the information they need to find content which previous users (curators, researchers and similar) have already identified as being related. They can follow in their footsteps.

Sometimes defining relationships can also help users infer connections. For example, if the red node above is a field book, and two items – the white nodes – are explicitly referenced by that field book, the white nodes have common documentation and may have come from a shared context.

Some institutions have started to document more explicit connections between items, documents (archival material), and other entities such as biographies; and art museums such as the Tate and the Carnegie Museum of Art are developing sites which presents this type of content online..

The American Museum of Natural History’s Division of Anthropology has also made progress in this area. Using a database developed in-house, they provide links by which online users can view related photographs, expedition field notes (digitised and transcribed), museum register pages, guides to archival records, and narrative descriptions of expeditions. However, there are limitations here too. While users can go from an object to related expedition field notes, for example, they cannot then go from these field notes to other objects collected at the same time, on the same expedition. Similarly, the museum register cannot be used as a node by which the user can navigate to other collection objects registered during the same period.

Finally, in all the cases examined links between objects and objects, or objects and related contextual material (archives, publications, biographies and similar) function as simple connectors or hyperlinks. While valuable, this overlooks the fact that the meaning of many things lies in the way they connect to other things. An object may have been collected on an expedition, on a certain date, by a certain person, and at a certain place. Field books contain explanations of objects, written by particular people at a particular time. The Thomson photographs of Cape York men making rope may in some cases show a similar rope being prepared to show the manufacturing process, or may be an image of the particular rope in question being manufactured. Similarly, the connection to field books may be because they describe when the object was collected, or they may contain contextual information like descriptions of dugong hunting.

Simple ’A connects to B’ relationships do not explain this type of detail, continuing to leave the task of working out why two things are connected to the user. Therefore, if many attributes of items are about the relationships between things, not just the things themselves, this means connections between things are ‘entities’ worthy of description and detailed explication in their own right.

Ian Hodder suggests that it is in connections, in the way things flow into other things, that the ‘thingness’ of things resides. If we accept this proposition (and I propose we do) we need to move away from a reliance on search. We also need to go beyond simple hyperlinks or ‘thin’ relationships in our collection documentation – beyond shared keywords and tags – and start documenting how things interconnect and relate. We need to capture and document the nature of relationships, and explicitly describe the ways in which items and their context intersect. To do so will help capture more of what we know about collections, and support new, different and deeper forms of interaction and use.

As is no doubt apparent, these presentations are more about defining the nature of a problem than providing solutions. That comes next year (I hope). For those interested in finding out more about my work, see my article for Museums and the Web Asia 2015:

Jones, Michael, ‘Artefacts and archives: considering cross-collection contextual information networks in museums,’ MWA2015: Museums and the Web Asia 2015, 5-8 October 2015, Melbourne, Australia. Printed in Museums & the Web: Selected Papers and Proceedings from Two International Conferences, Nancy Proctor and Rich Cherry (eds.), Museums and the Web, Silver Spring, MD, 2016, pp. 123-135. http://mwa2015.museumsandtheweb.com/paper/artefacts-and-archives-considering-cross-collection-knowledge-networks-in-museums/

I also have a book chapter coming out soon on archives, museum documentation and the Donald Thomson Collection which expands on some of the content in these presentations. I will post the details here when they are available.

Finally, this is an active project and I still have a way to go, so comments, feedback, or critiques are welcome!

* Ramsay, E. G. Aboriginal Artefacts in the Donald Thomson Collection a Microfiche Catalogue. Melbourne: University of Melbourne and the Museum of Victoria, 1987.

** Thomson, Donald F. “The Dugong Hunters of Cape York.” The Journal of the Royal Anthropological Institute of Great Britain and Ireland 64 (1934): 237–64. doi:10.2307/2843809