I’m investigating the overall theme of why people do not deliberate in the discursive activities that make up the public sphere. In general, designing for a ’deliberatory public sphere’ has been tried numerous times and the interfaces vary depending on the diagnose of why people do not behave ”as we want them to”. The ’error analysis’ was very helpful to imagine the designs and the potential errors of specific implementations.

For example, if the problem of our societal discussions is fake news, namely that wrong things are being published as if they were real news, then the easy way out is a to ban news that are fake. But this is a minefield, because journalists may be wrong without any bad intentions, and we do not want to ban journalists who are in good faith. Another example is the problem of informational polarization, i.e. the problem that people do not follow other news sources than the ones they agree with (and thus polarize into factions), and the easy way out is to force people to watch other types of news. But who are to decide on ’good news’, and how do we ’force’ people to watch it? The error analysis made such an issue harder and harder because the easy and bright solution is far from obvious.

As a last example, our social media esp. Twitter is flooded with bots that may be instrumental in causing fake waves of enthusiasm and trigger herd behavior involving real people. The fix to this problem that interrupts the political conversation is that bots are forbidden. But bots are scripts, and scripts are protected as freedom of speech. So, when prohibiting is not the solution, then what is?

I think that the error analysis helped such a complex problem with targeting all the ’obvious’ solutions. When these solutions come to the fore, they reveal how they may compromise values that we otherwise find are not ones to compromise.