Content and Context: Visualizations for the Public?

[Cross-posted from the Visualizing the Past blog for the NCPH2012 working group “Graphs, Maps, and Trees: Imagining the Future of Public Interfaces to Cultural Heritage Collections”]

In the very useful survey of the “history web” in their 2005 book Digital History: A Guide to Gathering, Preserving, and Presenting the Past on the Web , Dan Cohen and Roy Rosenzweig identify the range of genres that encompassed the historical content on the web: archival sites, exhibits and scholarly essays, teaching and learning sites, and discussion forums and organizational sites. Even though Cohen and Rosenzweig failed to account for the way that blogs, YouTube, and social media would eventually permeate the history web, I like their categories because they continue to the give us a way to think about what we do when we create public history online. We tend to provide access to collections, to offer interpretation, to offer instruction, and to offer a forum for conversation, both general and professional. So, as I began to think about the critical issuing in effectively using data visualizations in public history, I wanted to consider them in relationship to the activities above. Since Sheila has already written a great post on collections and enhancing access with visualizations, I’d like to focus both on their interpretative and instructive use, building on Trevor’s thoughts from his last post on discovery and communication.

For public historians, the mode of online outreach that has the longest history is that of interpretative exhibits, whether as companions to a physical exhibit or as independent works of scholarship. Despite the liberating possibilities for disjunction, many of exhibits hue very closely to the linear narrative structure of traditional narrative history. In doing so, they have demonstrated varying degrees of success in offering the public a glimpse of the richness of the past. Two sites from the National Museum of American History demonstrate the wide range of approaches. Both “The Price of Freedom” and “A More Perfect Union” are beautiful sites, but one presents a linear and reductive narrative of military history and the other presents the difficult topic of Japanese internment during World War II with a range of voices and perspectives that highlights historical complexity. The difference here is in the effort to bring together evidence in a user interface that allows for the consideration of many perspectives and multiple causality, as opposed to offering a single perspective that simplifies the past.

Successful or unsuccessful, most exhibit sites have the benefit of offering visitors a range of contextual information in both the text of the basic narrative and in the descriptions that accompany individual artifacts, images, or documents. This contextual information is essential for a public who may not have a deep background to bring to their encounter with primary historical materials. Data visualizations can short circuit the tendency to present simplistic narratives about our collections. Unfortunately, however, data visualizations that concentrate the user interface into a single interactive screen can also significantly reduce our ability to offer the public necessary historical context if we’re not careful.

Take, for example, the great interactive correspondence visualization created by the historians and computer scientists at Stanford for the Mapping the Republic of Letters project. This complex interface really only makes sense to individuals who are content experts, and sometime then only after they’ve read the accompanying pdf explaining the different facets of the tool. For a content novice, the tool is little more than a colorful toy because it links to primary sources in a subscription database, and because it lacks the biographical data on the correspondents that might make their network connections intelligible. If the visualization included access to a larger context of the enlightenment and background on the individual correspondents, it could be a powerful and concentrated way for the public to learn about the development of this period in Euro-American intellectual history. To some extent, we can excuse the Stanford project because it is explicitly a research venture targeting scholars of the enlightenment, rather than members of the general public

The Digital Vaults site from the National Archives and Records Administration, on the other hand, was created precisely to engage the public. Unfortunately, the project is completely hampered by its abstraction. The Flash version of the site gives users access to a seven randomly selected sources from a database of over 1,200. Clicking on a document, the user enters a web of connections to other documents based on shared tags. (The HTML version simply offers an alphabetical list of tags.) The sources have minimal accompanying metadata–usually title, date, and a brief description. While this environment is attractive and fun to play with, it fails to offer users enough context to make any historical sense out of the materials they encounter. Rather than offering and entree into NARA’s rich collections, the site leave users at sea with only their pre-existing historical knowledge to support them.

Unlike Digital Vaults, visualizations that make use of geospatial and temporal cues offer users more necessary orientation. One successful example is Minnesota Historical Society’s True North project, which offers users the ability to layer information in space in time to glean some understanding of the state’s history. While the interface does not link out to individual primary sources, it manages to offer enough cohesiveness that users can start to construct their own narratives of change. The National Museum of Australia’s History Wall is even more successful. Built on the backbone of a flexible timeline, the interface allows users to explore the lives of Irish in Australia between 1770 and the present, drawing on, among other things, the Australian Dictionary of Biography, and amazing Trove database that aggregates over 280 million sources from National Library of Australia. Together these elements let a user to explore within a much more deeply layered context that can push them to use heuristics that are important to historians as they make sense of the evidence from the past.

Considering this range of examples, I hope that we can begin to have a conversation about how to create and frame data visualizations that provide the public with new ways to access our content but also offer them enough context to help them begin to make sense of those materials in meaningful ways.

Comments 4

  • Thank you for all of the wonderful examples and analysis, this will be very helpful in moving my own project forward.

  • It’s my pleasure, Alex. I’ve been following the exciting work at the CHI Initiative, and I’m looking forward to seeing how the projects develop.

  • My name is Kelsey, and I am a student in a digital humanities seminar called “Hamlet in the Humanities Lab” at the University of Calgary: .

    I found this page very inspirational and, I would like to base my argument for my final paper on your blog post. You can read my paper after April 25th on the course blog:

    Thank you for your unique, and very helpful insight!

  • […] What would Shakespeare have done if he were to find there was a technology to break down his entire works into categories, word count, or frequency? If there were something to link all his meanings together? He would probably rewrite his plays to make them that much more cryptic! The wondrous thing about his plays is that they were even complicated for the time they were created. Nowadays there are scholars who devote their lives to discovering the meanings of Shakespeare and the voices of Shakespeare. The article I chose is speculative on the future of museums and archives whether it will be possible to provide on-site enough information to let the average viewer read a work of art or historical artefact like an expert. She imagines a world where information is immediately available to those who seek it. It sounds like the future for those who would take the time to pursue it. I believe this is the future of the Digital Humanities. It would not be that while reading Hamlet notes appear at the side of the text to divine meanings. There are already books suited especially for that. Instead, this would have the power of the internet behind it. All of the searches we can do in the five tools are to deliver what you are searching for in their location, location, location. Sometimes you can even figure out who delivers the line, how often and if similar words were delivered. It is up to us to understand it. The difference, and what I believe is the future, would be to deliver context to the seeker. Not just the immediate ability to see the context and meanings of Shakespeare in notes by previous scholars, but also maps of discovery. What this would mean is that a person would find Hamlet in a digital tool, and not just find a word. The word would come with the initial and evolved definition since Shakespeare’s time, any idiomatic references it may contain of the 1600s, and the option to dive deeper in to what other scholars think about it. From this our understanding would not only be our own limited experience. Sharon Leon wrote: “The difference here is in the effort to bring together evidence in a user interface that allows for the consideration of many perspectives and multiple causality, as opposed to offering a single perspective that simplifies the past” ( […]

Leave a Reply

Your email address will not be published. Required fields are marked *