Opening up Election Debate Mapping
Here’s a follow-up for those of you who’ve been following the real time mapping of the first TV election debate, and are interested to pitch in, or for those of you interested in the technological aspects of these emerging tools for participatory sensemaking.
The original maps were mapped in real time, as the debate was broadcast, using Compendium, a desktop tool optimised for very fast mapping (it is often used to provide visual scaffolding for a team meeting either f2f or online – eg. hostage recovery, science teams, researchers). In the forthcoming release of Compendium v2.0, which I was using, this can be done over the internet by multiple analysts working on the same set of maps, rather like a visual wiki.
However, we can also upload the maps as XML, from Compendium, to our social media platform called Cohere [article]. This brings Web 2.0 capabilities to argument mapping. View the Cohere UK Election debate group which is housing all the election debates we map. Now every contribution to the debate has a unique URL, and can be embedded, tweeted, or linked to in a meaningful way (e.g. is inconsistent with, refutes, is evidence for, was predicted by… etc — whatever move you want to make).
Below, for instance, is one contribution to the election debate by Gordon Brown, embedded (like a YouTube movie) in this blog post — click on it to jump into Cohere, in order to see how it is connected to other contributions from Cameron and Clegg:
We can also embed a specific connection, or whole map, in a website, to enable others to view and contribute to it from many different sites. Below is a static screenshot of the network visualization applet we are using to provide interactive map browsing and editing (I haven’t embedded the live one in this page since you may not be running Java. Here’s the live applet if you want to try it).
Here’s the bird’s eye view of the whole of the first TV debate…
…and once you zoom in on a fragment of interest it becomes legible for navigation and editing:
But if views like this don’t work for you either cognitively or technically (and we’re about to replace this with the Adobe Flash version of flare.prefuse.com) then you can also work with the Connection Neighbourhood view.
Below we walk through the web browser interface via which anyone (who’s signed in) can make a new connection.
Here’s a view focused on the focal contribution to the debate, shown in the centre. Mousing over the menu icon brings up the menu of actions you can perform, which includes “Connect”…
In the pop-up dialogue, I’m asserting the Gordon Brown’s assertion reiterates Nick Clegg’s:
The new connection is confirmed, having been added to the map:
If I Get the Snippet for this specific connection, as shown, I can then embed it in this blog, as an interactive fragment of the map that I want to embed in a new commentary, thus:
Obviously, once there are many people adding new ideas and connections to a web debate of this sort, then there might be three different analysts shown in a given connection, with one analyst connecting contributions from two others (example: climate change analysis).
Another advantage of moving from real time private mapping to participatory, asynchronous mapping, is that each node in the network can now have an arbitrary number of websites added to it as backing evidence for the claim it makes. Obvious examples would be fragments from manifestos, or video clips of previous speeches, etc.
Just edit the node, or get very smart, and load the extension we provide for your Firefox web browser. Now when you’re viewing ANY website, and you want to connect something of importance on it to a node in the debate, just highlight the relevant text, and tie it to the idea in your margin sidebar. In the example below, I’m providing more detail from the NI Campaign on the Conservative party website, to an assertion made by Cameron in the debate… you’ll see my annotation on that page as shown, once you’ve loaded the Firefox extension:
Finally, you may have spotted that while this is a hypertext tool, we’re building on years of hypertext research which ha demonstrated how we can make meaningful connections between nodes in the network, not just the “go to” hyperlinks that dominate the Web at present. By using a connection language to model arguments, we can filter the network to show just specific kinds of links. In contrast to Compendium in which you hand-craft every map (great for precise presentations), in Cohere every view is automatically generated and laid out, since an unknown number of participants may be contributing to a given view. Semantic filters on these potentially very large, complex visualizations now start to become rather useful.
There’s a basic semantic filter into broadly positive, neutral or negative connections:
…which in our experimental Java applet, can be refined by any user-defined mix of connection types, or by using preset searches such (example below) Contrast which searches the database from a focal Idea on a specific subset of connection types of a contrasting nature: challenges, has counterexample, is inconsistent with, refutes:
OK, end of mini-tutorial. Hopefully that excites you about the potential of tools such as this to raise the level of public discourse around complex societal issues, the present election being just one timely example. I have a hunch that we also need to generate a more conventional reports which will be familiar to more people, and while harder to show complex interconnections, they can compress more information in a page/screen. Since it’s just another view onto the database, there would be no need to have to choose between them: just flip as required.
If this gets you going, I should also point you to the great work that my colleagues at Debategraph are doing, as they use their tool to map and open up for debate the political interviews conducted by CNN’s Christiane Amanpour, plus many other pressing political issues.
Note that this pilot work has since developed into a new project to map the anticipated 2015 Election Debates: http://edv-project.net
Simon