The final day at the 2026 International Communication Association conference in Cape Town starts for me with a paper by Laura Liebig, exploring the German discourse around arms deliveries to Ukraine. She begins by outlining the issues around media criticism: such criticism is warranted at times, but can also be weaponised to elevate critics’ own positions and undermine trust in the media.
In Germany, critics have repeatedly accused mainstream media of one-sided coverage of arms support for Ukraine since Russia’s illegal full-scale invasion in 2022, and requested more viewpoint diversity. Such diversity is valuable in principle, and one of the normative ideals in journalism, but what exactly constitutes such diversity in practice is highly debatable.
There are liberal-aggregative, liberal-individualist, deliberative, and adversarial ideals which each describe a different form of viewpoint diversity; these are especially relevant in war reporting, which is often more oriented towards government elites. The arms debate is therefore a useful case study here.
The project examined some 55,000 articles from 14 German-language newspapers between 2022 and 2023, selected articles about weapon deliveries, and classified relevant sentences within these articles using LLMs. Viewpoint diversity was classed as for, against, or ambivalent; overall stance was balanced, but slightly more in favour towards arms support, except for the right-wing outlet Der Freitag. Such patterns were relatively stable over time too.
Various measures (Gini, Herfindahl-Hirschman, Shannon, etc.) show that such viewpoints were relatively balanced too; there is a fairly high degree of polarisation, with for and against stances represented in the articles.
Key co-occurring frames in the coverage are strategic, normative, security, international, escalation, and defence, while negotiations and pacifism are comparatively less common. Viewpoint diversity of frames is very concentrated on a small number of frames.
So, there is a balanced distribution of for and against stances; there is also high polarisation between articles, however. This means there is internal as well as external diversity. Stances are not static over time, but argumentative diversity is low – this means that the discussion is largely guided by a liberal-aggregative ideal.
This contradicts accusations of an overly uniform media reporting in this case, though highly oppositional viewpoints are not usually present. Formal diversity does not necessarily translate to argumentative diversity – and the question of what kind of diversity we should aim for remains open, of course.
