Context and connections
- May 28th, 2009
- Eric Johnson
I’ve been doing a lot of thinking about ways to get primary source documents to “talk” to each other and to the cloud of secondary sources that surround them. For example, at Monticello we’re working on a digital version of Jefferson’s memorandum books (60 years’ worth of purchases made, places visited, people seen, etc.) and want to enrich it far beyond simply getting the text on the web. Can we make that incredible information come alive in a rich and user-friendly way? Put these and other primary sources into a broader context of people, events, ideas? Connect these seamlessly with secondary sources treating the same topics? Can we decentralize the process to pull information from non-Monticello assets? What visualization tools will help?
Or another version of the same “problem.” Thomas Jefferson wrote between eighteen and nineteen thousand letters in his lifetime and received several times that number from other writers. What are ways to illuminate the connections among those letters? What are ways to permit an easy understanding of the larger (political, social, material, geographic) contexts in which that correspondence took place? Are there good tools that will let people explore letters by theme? And beyond that, can the same solutions be applied to other correspondents at other times in other places (and, ultimately, turned into a giant killer spiderweb of correspondence)?
May 28th, 2009 at 12:15 pm
And of course Jefferson’s letters are scattered in archives around the world, some online, some not. We need tools to enable the letters in your collection to talk to those at the Massachusetts and Missouri Historical Societies, for example, and to printed letters in various Google Books.
May 28th, 2009 at 2:15 pm
I really like the directions in which you are thinking. I could say a lot but I will start with three questions: 1) have you tried to obtain the query logs for the Jefferson Papers search engine on the Library of Congress web site?, 2) have you used web spiders or crawlers to identify the range of sites that link to Jefferson papers sites?, and 3) have you extracted the terms that appear in the indices to the volumes of the official published Jefferson papers?
May 28th, 2009 at 3:29 pm
Luckily, much of Jefferson’s correspondence has been pulled together into a new online form available through UVA’s Rotunda press. Unluckily, it doesn’t yet cover everything, because the two scholarly editorial projects working on his papers still have much of his presidential term and much of his retirement-era correspondence to do (some of his retirement material will be available online soon). Luckily, it’s got basic XML markup (the retirement-era project will have more robust XML). Unluckily, it’s a subscription database.
[What is going online are digitized versions of Princeton University Press’s Papers of Thomas Jefferson, the definitive scholarly edition of all of Jefferson’s correspondence. We have a parallel project here at Monticello tackling his retirement-era correspondence. Right now, everything from his earliest correspondence as a young man to the first eight weeks of his first presidential term is available online.]
Some of the basic connections, such as hyperlinking between related letters (“I received yours of the 13th inst and was much obliged . . .”), should be do-able as a matter of course. I’m especially interested in trying to pick people’s brains on other, more dynamic and expansive ways in which documents might be connected to one another (in other words, what would you as a user be interested in seeing?) and in picking brains on technical ways it can be done. And my bigger yet more nebulous question: how can we provide easily-accessible context “around” these primary source documents?
May 28th, 2009 at 11:43 pm
Maybe one way to start thinking about this is to imagine something like Diigo’s social annotation system, except with easy links to other (socially annotated/annotatable) materials, both primary and secondary. I don’t think Diigo’s code is open-source, but if you’re not looking for point-specific annotation/linkage, it should be possible to create a metadata engine that displays a document and then allows people to annotate, link, etc., in a navbar-like strip somewhere on the margin.
May 29th, 2009 at 11:48 am
[…] of the blog data at Semantic UMW. It sounds like this is similar stuff to Eric’s interest and Sterling’s interest — making connections — but in the domain of students and […]
May 29th, 2009 at 1:30 pm
Eric: Did you see this?
chronicle.com/review/brainstorm/katz/government-and-scholarship
It sounds like the federal government is going to take over or coordinate the digitization of the papers of the founding fathers. I hope you get to be involved.
June 1st, 2009 at 11:41 pm
Sterling, I am glad to report that the Papers of Thomas Jefferson: Retirement Series at Monticello is being held up as a (the?) model of the kind of approach NHPRC is interested in pushing along–relatively quick production, scholarly integrity, digital-friendly. The NHPRC has funded part of the work of our project.
Katz’s column is presumably a continuation of the discussion that launched into the public sphere with this Washington Post article among others and resulted in this report from the National Archives (something I hadn’t seen until your link prompted me to go poking around!).
June 3rd, 2009 at 5:19 pm
Hi Eric,
Have you seen the Finnish Culture Sampo site? It may of limited use to non-Finnish speakers, but the English interface at least lets you browse around and see the possibilities.
Their Museums and the Web paper (and presentation) are available at: www.archimuse.com/mw2009/abstracts/prg_335001985.html
This goes beyond just texts by also connecting material culture objects together (e.g. a painting that depicts a scene from a famous poem, or in your case maybe we could also link to the image of the desk upon which Jefferson wrote).
Also, I thought I saw something at HASTAC that had plotted Jefferson’s travels and correspondence on a map.
June 5th, 2009 at 11:21 pm
Musebrarian: what you saw sounds like www.jeffersonstravels.org/ from the Virginia Center for Digital History in collaboration with the Thomas Jefferson Foundation at Monticello.
This topic interests me greatly, and Eric has set up a number of productive questions. I wonder about a tension between curation and standardization. On the one hand, it would seem to be more efficient to plan for generalized representation and visualization tools that could spin out into a “giant killer spiderweb of correspondence.” But on the other hand, why do these contexts and connections warrant the effort to digitally represent them? What specifically do we want to learn or teach about Jefferson, or about history through Jefferson, that we don’t already know yet? Having particular questions in mind will cause the tool-building to go in certain directions that it wouldn’t necessarily go otherwise. But then highly curated tools in the service of a hypothesis or argument or specific direction of inquiry may be less generalizable.
June 22nd, 2009 at 12:10 pm
We have similar challenges (on a smaller scale) with our Paracelsus collection (not yet digitized or even on the road there yet, but is ‘on the list’) – we’d love to eventually have a site where scholars could compare our copies, page by page, to those held by other institutions around the world, but short of writing a big joint grant proposal, getting all that information in one place (or at least using the same standards) seems a far-off goal. I’d be very interested to hear the ideas that come out of this.
June 23rd, 2009 at 10:25 pm
Richard, thanks so much for bringing the Finnish Culture Sampo site to my attention–there’s some incredible food for thought there. I’ll have to hit that site hard over the next couple of days.
Douglas, I suspect you’re right about the Jefferson’s Travels site being the thing Richard’s thinking of. I was excited to realize that Bill Ferster, who is largely responsible for the programming behind the Travels site will also be at THATCamp. I’ve certainly seen the project in action, but I came late into the thing–this will give us a chance to really sit down and talk it out.
You also raise a good point about asking specific questions. Off hand, the biggest driver guiding my interest in these areas is around trying to get a handle on the mass of data presented in Jefferson’s correspondence. In an ideal world, we would develop tools that would illuminate in a largely automatic way the thematic, temporal, and spatial connections among Jefferson’s correspondents (and correspondence). I’m interested in understanding large patterns and rhythms, and if that can be shown in ways that don’t require a manual examination of 40000-plus letters, I’m all for it. It may be a while before we can move off of a letter-by-letter examination, but the discussions of semantic tools give me hope for down the road (and I have no problem with iterative development, with manual effort at first and automated effort later).
It would be great if out of these conversations a suite of tools were made available that would work for ANY set of correspondence (or other written texts). That would certainly be the kind of shared direction I think you’re getting at, Lisa.
May 4th, 2010 at 2:52 am
[…] thatcamp.org/2009/illuminating-context-and-connections/ […]