You are here

Analysing the Visuals Shared by the Different Sides of a Polarised Conflict

The final speaker on this first day of the Indicators of Social Cohesion symposium is the great Luca Rossi, presenting some of the outcomes of the PolarVis project to map online debate around climate change from a visual perspective.

The project is interested in the visual content that these groups share online, and in how this content is used to support their narratives. Visual elements have been especially important in climate change debates, both because of the emotional impact of metonymic depictions of climate change and the use of scientific visualisations to describe and forecast climate change and its implications, and because of the contestation of such images by climate change denialists.

The project gathered visual content from the facebook, Instagram, and TikTok accounts of some 454 European and international actors on either side of climate change debates, and clustered their content based on the similarity of the visuals. There are both similarities and differences in the content typically used by the opposing sides. Similarities would point to some level of shared reality, while differences indicate a visual form of polarisation between them. In other words, how substantial is the perceptual distance between the opposing sides?

Some types of visuals are shared between the two sides: this includes photos of wind power plants and of demonstrations, for instance, which are operationalised in very different ways by different ideological groups, and which may generate very different emotional reactions from their target audiences, too. To better understand this, it was also necessary to determine the likely ideological leaning and emotional reaction of the people who shared such images; this was achieved through the use of word embedding for ideological leaning and the ratio between ‘angry’ and ‘love’ reactions on Facebook for emotional responses.

From this analysis, it emerged that visual images from actors on either side of the conflict are predominantly shared subsequently by actors with the same ideological stance. When content does move from one side to the other, there tends to be a decrease in emotional reactions, and an increase in neutral feelings. Meanwhile, images shared by both sides of the debate often generated very different comments – suggesting that people actually interpret this content in very different ways (in other words, there is interpretive polarisation).

This also means that polarisation is not necessarily a problem with a lack of contact between the different sides of a conflict – but rather, with the very different waz in which different groups read and respond to the same content.