Since October 2012, Ralf Grötker from Debattenprofis has been conducting a media experiment involving argument maps and swarm intelligence. In the so-called Faktencheck (Factcheck) series, Grötker sets up and moderates online forums on controversial issues (e.g., “boycott of textiles — helpful or not?”). Debattenprofis use argument maps to aggregate the discussions. In a recent article Grötker sums up his experience so far.
This is how Grötker describes the overall design of the experiment:
“Faktencheck” brings together elements of reader participation and journalistic investigation on controversial issues with a scientific background. There’s a live investigation for a period of 3 days. Readers have the opportunity to join the investigation in a forum. A moderator transfers both comments in the online forum as well as other results from the investigation into a so-called argument map.
According to Grötker, the role of argument maps was ambivalent. On the negative side he notes:
The great hope, which we associated with using argument maps, was that it would be feasible to somehow order the discussion in the forum. In this regard, the experiment was rather disappointing. Only a few commentators made use of the opportunity to refer directly to the branches of the map. At closer look, however, this is hardly surprising: In many contributions, commentators looked for engaging in a dialog — mostly, by the way, in an astonishingly respectful way.
Let me jump in here. It’s clear from a brief look at the webpages which document the experiment that the argument maps didn’t help to structure the online discussions. So, is argument mapping in general unsuitable for such a task? I’m not prepared to draw this conclusion yet. As has been noted by participants of Faktencheck before, it seems a major problem that the argument map is totally disconnected from the online debate.
- The arguments in the map don’t link to the contributions they reconstruct. (Although some arguments contain roll-over quotes from the online discussion.)
- The contributions in the online forum, in turn, don’t link to arguments that represent the contribution, either.
- The argument map is, moreover, embedded on a different page than the online forum.
(To make things worse, the discussion seems to be conducted in several, separate online forums.)
That might explain why the map played virtually no role in structuring the discussion.
Another issue is the software Debattenprofis have used. Look at the startscreen of the map about textile boycotts:
That doesn’t look very inviting. Even more importantly,
- it’s really cumbersome to navigate the map, and difficult to get an overview;
- more specifically, it’s not straightforward to focus on one argument and its adjacent reasons only;
- the debate doesn’t exhibit a macrostructure;
- the hierarchical layout is very space consuming, which is a major disadvantage if the map is embedded.
Let me also note that the Faktencheck argument maps are not based on detailed reconstructions of the arguments advanced in the online forum, but merely “sketch” (as we use to say) the debate. As a consequence, the dialectic relations between the arguments are still provisional and seem to some extent arbitrary.
Ok, back to Grötker’s upshot, which also highlights a positive effect of argument mapping in the Faktencheck experiment:
The visualization fulfills an important function nonetheless. As the argument map strived to represent all pros and cons (concerning the corresponding claim), it goes along with a certain promise of neutrality. The opponents don’t have to agree on a common evaluation of the arguments in the map. But they have to agree that the controversial issues are comprehensively represented. This promise of neutrality was particularly helpful when we asked external experts to join the factcheck. The opinions of experts could enter the map without us being required to come to the same conclusion as the experts in our final summary — a fair set-up for all parties.
It’s an obvious virtue of argument mapping that it allows opponents at least to agree on what the arguments are and how they relate to each other. Also, to agree on an argument map doesn’t compel one to endorse a certain position. Faktencheck seems to have vindicated these argument mapping advantages.
To connect the dots, however, imagine an argument map browser that allows users to enter their assessment of the arguments (‘refute’ vs. ‘accept’) as they skim through the debate. This might not only enhance user experience but help to integrate the debate visualization and the online discussion.
Be that as it may, Ralf Grötker has conducted an exciting experiment and I’m looking forward to Debattenprofis’ future argument mapping projects!