Two experiments focusing on de-escalation oriented coverage of post-war conflicts

War coverage has a strong bias towards the promotion of conflict escalation and - though less pronounced - this bias often survives in post-war coverage as well. Even after the end of war, only a minority of journalists frame conflict in a firmly de-escalation oriented way. Do they have a chance to reach the public? Will their reports be respected by the audience as more balanced and unbiased? Will they have an impact on the audience's mental models of the conflict? Or will the Milo ević audience continue to cling to its prejudices and reject news articles which do not affirm the enemy images that emerged during wartime? The present paper investigates these questions by means of two experimental studies. In the first experiment, news articles on three events in former Yugoslavia after the fall of Milo ević were presented to a total of n=128 subjects, representative of the readership of the German quality press: (1) violent conflicts in Southern Serbia (December 2000), (2) the extradition of Milo ević to The Hague (June, 2001) and (3) the treaty between Serbia and Montenegro (March 2003). For each of the events, four different types of articles were used: moderately escalation oriented articles from prestigious German newspapers (Die Welt, Frankfurter Allgemeine Zeitung, Süddeutsche Zeitung) and three variants of these articles, (a) with increased escalation-oriented framing, (b) with moderate de-escalation oriented framing and (c) with more strongly de-escalation oriented framing of the events. Each subject was asked to read one article on each of the three events in chronological order and after each article (a) to narrate the reported events in their own words and (b) to fill out a questionnaire designed to measure the acceptance of the articles as unbiased, well-bgalanced, interesting, etc. The subjects' mental models of the reported events were inferred from their narratives by means of quantitative content analysis. The second experiment measured the acceptance of the articles only, but not their impact on the readers' mental models. Besides that, it employed the same design and used the same instruments, but with some modifications: The original articles stemmed from an Austrian regional paper (Voralberger Nachrichten), and the sample of subjects (n=126) was recruites from its readership. The reports about the treaty between Serbia and Montenegro were replaced by reports about Kostunica's reaction to Rugova's victory in the Kosovo elections (November, 2000), and the more strongly de-escalation oriented text versions were replaced by escalation-orientes ones with reversed partiality (pro Serbia). The results of the studies speak in favor of the peace journalism project. De-escalation oriented news articles were never accepted to a lesser degree than the other textversions. With text material from the quality press anhd its readership, they were even accepted to a higher degree and resulted in less polarized mental models of the events. With text material from the provicial press and its readership, no difference could be found with respect to the acceptance of the various texts versions. The results indicate, moreover, that the provincial audience was less interested in post-Yugoslavian affairs and more strongly influenced by traditional news factors like personalization and negativism