Framing under Uncertainty: German and US Reactions to Nord Stream Sabotage

• In Event: Media, Social Media and Conspiracy Theories

Abstract

This paper traces the spread of news frames in the German and US information networks on the September 26, 2022 sabotage of the Nord Stream gas pipelines. We select these cases because of the nations’ dominance in NATO, their military contributions to the war in Ukraine, and because public support for the war is similarly divided (65 percent overall approval in the US, 56 percent in Germany), though there is more partisan division in the US than Germany.

Given the as yet unresolved mystery as to the identity of the responsible culprit, a comparison helps explain the contribution of the nations’ respective party and media systems as well as the degree of affective polarization--increasing in the US, stable in Germany (Garzia et al. 2023)--to the propagation and spread of relevant news frames to networked claim communities. We define networked claim communities as groups that have an interest in attributing a specific responsibility and motive for the sabotage with respect to its implications for support of Ukraine. To put this within a framing model perspective, such a community holds an interest in promoting a specific problem definition, assessment of blame, moral judgment, and policy solution (Entman, 1993).

Because news frames rarely present such elements in a comprehensive way expected in, say, a legal brief, we rely on a frame signature matrix an analytical tool used to define and identify one or more elements of a specific news frame, as well as the textual and visual features--”condensing symbols”--that indicate its presence (Gamson & Lasch, 1983).

Using Natural Language Processing algorithms, namely Large Language Models and text embedding, we automatically extract frames from text to trace the paths taken by news frames through the information networks over a period of a year following the explosions. The paths link a variety of internet sites (nodes) that include news as well as social media. For news media we use LexisNexis, a comprehensive and searchable repository of newspapers, news wires, blogs, and television transcripts (Germany, N=21,325; US, N=30,200). For social media we select Reddit, a platform that permits users to submit content to user communities (“subreddits”) who share similar interests and are permitted to register their approval or disapproval for the posted content and thereby either boost or hinder the spread of a frame. The goal of the research is to identify those paths that lead to the greatest likelihood of information cascades that advance a favored frame beyond the networked claim communities that propagated or endorsed it (Germany, N=1527; US=865).

We begin by deriving a network graph of media sources--nodes--that mention the sabotage where the size of each node are based on the number of sources that share the same claim and or frame. We do the same for subreddits and create a separate network of subreddits that mention each other. The idea for both is to gauge the issue frames they co-reference and the interrelationships between subreddits (e.g., the degree to which they refer to each other). In this way it will be possible to determine the origins of specific frames and the pattern of their spread through the networks over time.

We hypothesize that the multi-party system in Germany, its stable links to mainstream news sources, and relatively stable levels of affective polarization inhibit frame cascades that spread beyond their ideologically sympathetic claim communities. We also hypothesize confinement of the most extreme news frames to social media and less traffic from these nodes to mainstream news sites.

By contrast, we hypothesize that a highly polarized party and media environment in the US leads to the ready migration of news frames from social networks to news media, though enhanced on the right by the asymmetrical polarization of the US information system. Where left-leaning news sites remain committed to journalistic balanced norms, right-leaning sites are more partisan.

Preliminary analysis of German news-site data reveal four frames that blame (1) Russia by way of a false flag attempt to implicate Ukraine, (2) implicit US blame based on its interest in profiting from sales of replacement liquid gas for Europe, (3) explicit US blame, and (4) explicit blame of Ukraine. Direct US culpability is not advanced by any of the mainstream media or major political parties. The frame is confined to the extreme right (AfD) as well as the extreme left, most recently in the form of a new party created by Sahra Wagenknecht, former co-leader of Die Linke. The American blame frame does not gain currency in the mainstream media. It is predominantly advanced on social media.

In the US mainstream left-leaning media echo their mainstream German counterparts, highlighting Russian culpability. On the right, the US culpability frame is advanced, largely as a means of undermining the credibility and authority of the Biden administration (in conspiratorial terms, the "Deep State").

Authors

• Andrew Rojecki, University of Illinois, Chicago

• Alisa Baechner, Freie University Berlin

• Jianing Shen, Freie Universität Berlin

• Keith Burghardt

• Kristina Lerman, USC Information Sciences Institute