It strikes me that there are two, somewhat overlapping, reasons that we do visualization. They are a method of communication and a method of discovery.
Visualization for Communication
Visualizations are methods of communication, ways of communicating something that we already understand. In this case, things like Tuftee’s work on presenting data and information is squarely about communicating known things. Similarly, most of what I see on flowing data strikes me as this communicative tradition. In the realm of historical thinking and scholarship David Staley’s ideas about Visual Secondary sources in Computers, Visualization, and History: how new technology will transform our understanding of the past forward this communicative notion of visualization.
To make this a bit more concrete, the image below, from information is beautiful, illustrates (and illustrates is a key term) the effectiveness of different approaches to fundraising for Wikipedia doesn’t really tell us something new. All of the data is up online and if we read through the data the relationship is evident. The graphic below just communicates that relationship more forcefully.
Visualization for Discovery
Visualizations are also tools for discovery. In this sense, visualization is a method for finding out new things. Even in the case of something really simple, like Wordle, we create something visual that we can then examine and explore for a potential new ways of seeing or understanding something. For example, in the Wordle below I feed the entirety of René Descartes Discourse on the Method of Rightly Conducting One’s Reason and of Seeking Truth in the Sciences from Project Gutenberg into Wordle and was then presented with the following representation of the book in a word cloud.
I did not know what I would get when I hit the button. The result is not particularly good in terms of communication, largely because I didn’t intend it to communicate anything. I just wanted to see what would happen. Now in this case, I find it interesting that things like “heart” and “blood” are as big as they are. If I were interested in taking this further I might go back to the text and try and suss out why this is the case. Now, if you don’t want to use the kids-table version of this sort of thing you can pick up something much more sophisticated and do things like Now Analyze That.
This line of thinking, of visualization as a method of discovery, is largely in line with Jessop’s ideas about Visualization as Scholarly Activity, and Drucker’s notion of Graphesis , wherein visualization is understood as “generative and iterative, capable of producing new knowledge through aesthetic provocation.” I think this is also very much what Moretti is talking about in Graphs, Maps, and Trees.
Is Public History Visualization Somewhere In Between?
For me this becomes a central question. What is the goal of visualization for an online exhibit, or a cultural heritage collection? Do we want to communicate something we already know as clearly as possible? Or, are we trying for the generative and iterative new knowledge producing capabilities of aesthetic provocation? In some cases, I think there is also the possibility of attempting to put something in the hands of the public that lets them engage in their own exploration and discovery in the context of a collection.
For example, contrast Digital Harlem and PhillaPlace. Both offer map based interfaces to cultural heritage data. Both let us explore in our own ways. With that said, I think Digital Harlem falls much more on the side of providing a messy-data-sense-making-discovery-place while PhillaPlace offers a structured visual communication space.
In the image above you can see the dense interface to Digital Harlem which invites us to poke around in the data they have gathered together and explore how the picture changes as we poke.
In contrast, the point of entry to PhillaPlace is a map that moves on it’s own. We see the cultural heritage points flip through. While PhillaPlace does offer a rich map interface, it is less about surfacing any patterns in the underlying data and more about giving you a way to browse via location.
What Approach to Visualization are You Most Interested in and Why?
I imagine that there are going to be different answers to this question in different situations. With that said, I think it is essential that anyone thinking about using a visualization have a really good answer to the root of this question. That is, why are you making a visualization?
I would be curious to hear from the group, and anyone else listening in. What exactly is it that you want to get out of visualizations? Are you trying to communicate something as clearly as possible, or are you trying to generate something messy that we can use instrumentally to develop new knowledge?
Pingback: URL
Pingback: Communication or Discovery: Which Approach for Public History? via DHNow | #HIST5702x
Great post. Another way of thinking about this would be to ask whether we want visualizations to behave like secondary sources–which are ideally transparent, controlled by the author, and which strongly allow only a single reading–or like primary sources, which require learned techniques of interpretation, can be misleading, and from which we need to be alienated. Most everything I’ve done falls in the latter camp, which I’m painfully aware presents substantial problems for public history; you’re both trying to create an object which is, essentially, a new historical text in itself and teach people how to read this new sort of text, which can be very tricky. (It’s hard enough to expect museum visitors, say, to deal with an old letter sensitively; very frequently, they won’t be at all good at knowing how to read heatmaps, or network visualizations, or log scales). On the other hand, I tend to think that getting people used to reading statistical graphics from a humanistic-aesthetic point of view should be part of what digital humanities aims to teach the broader public.
Pingback: » Content and Context: Visualizations for the Public? [bracket]
Pingback: Content and Context: Visualizations for the Public? | Visualizing the Past
Smart post, Trevor, and I like Ben’s consideration of these as related to primary and secondary sources. Perhaps one of the things that digital humanities can do is to break down that distinction. (This is a point that Tom Scheinfeldt makes here (http://j.mp/xtIt7C) – and it’s even more true for visualizations. VIsualizations can be static or dynamic. Both we and our publics can play with them in their dynamic form to make their own discoveries and share them with each other in their static form, as a way of making an argument. Perhaps we can no longer get away with making claims based on data without sharing the data and letting our readers reproduce our discoveries for themselves!
Trevor, your conclusion offers a damned interesting dichotomy. As an historian I like serendipitous discovery. As a museum director, I need to know the audience. Web 2.0 blows this apart. Letting people play with our data and feeding their play back to us can do a lot for public engagement. And we may well learn something new.
Wow, what a useful question to ask about purpose. How about both? When I started working on the Driving through Time digital Blue Ridge Parkway project, I think part of my aim was to convey visually thing I already knew based on my primary research. These things were often things that I had found very hard to convey via text. But I also wanted people to be able to assemble visualizations, especially via georeferenced maps, that were meaningful to them. That is why I wanted all of our maps to be available for download to Google Earth (achieving this proved challening, but we did it). I wanted people to be able to create their own map sets. Anyway, I am not sure I thought through this as systematically as your are suggesting here.
Our project is here: http://docsouth.unc.edu/blueridgeparkway/
Pingback: Artist Analyst « Ludic Analytics