TRAVIS: Trust and Visuality: Trust in Everyday Digital Images

Lead Research Organisation: University of Oxford
Department Name: Geography - SoGE

Abstract

The second decade of the 21st century has been characterized by a deep crisis of trust and legitimacy. We need to recuperate trust, and to do so we need to understand it better. TRAVIS starts from the premise that trust is crucial for functioning social relations within and outside of national spheres. Whereas existing work predominantly discusses trust as integral for human-to-human relationships, an adequate understanding of trust is significantly complicated by our increasing reliance on online communication, where the intentions, commitments and even existence of other autonomous humans has to be inferred from representations unfolding on screens. These representations are increasingly shown in sequences influenced by algorithms intended to maximise engagement, volume, reach and share of voice. These metrics do not necessarily yield trustworthy interactions, or a multifaceted public sphere, but tend to amplify material that increases uncertainty or simply cannot be trusted.

TRAVIS will study practices of visual digital trust in two interrelated fields that are central to everyday enactments of trust, near ubiquitously digital, heavily visual, and existentially relevant to all of us: health and news. Specifically, we look at health news and social media health communication. Their centrality to questions of trust and the social implications of that centrality has been highlighted in the last two years of the ongoing pandemic - data visualizations of covid-19 statistics, graphics of the infection rates and personal stories by people both working in intensive care and those witnessing the immense societal challenges have merged in our daily media routines. Of course, visual digital representations related to our individual and collective experiences of health were central to our life already before - we produced and shared proud screenshots of step-counters, sad Instagram-posts about a friend's mastectomy, scary pictures of the black lung in the meme about smoking. Visual representations in health contexts have the potential to affect and inform, to discipline and empower us, individually and collectively, in both public and intimate contexts. But if and how these potentials are enacted, is closely related to questions of trust.

We will how trust in images is enacted by studying visual health communication on social media and visualization of health topics within digital news in four different cultural contexts - Austria, Estonia, Finland and the UK, thus combining perspectives from Nordic, Eastern-European/Post-Soviet, Anglo-Saxon and Germanic cultures, each with their own traditions and norms of trust.

Our three-stage project will use a range of in-depth qualitative methods:
- In the first phase of our research, we focus on the site of the audience - the trustors, and their practices of trusting by focusing on how they consume and interact with health visuals. This allows insight into the role of audience perceptions of the content creator (e.g. institutions vs influencers), those sharing and circulating content, the characteristics of the content itself, the inferred impact of software and hardware.
- In the second phase, we shift our attention to the site of production - the trustees, and their practices of trustworthiness through the production and sharing of health visuals: again, our approach incorporates study of the performance, statuses, experiences and skills of human and institutional content producers.
- Finally, in the third phase, we will focus on the site of the visuals and their platforms, in order to understand how the platforms as built, corporately owned structures shape trust practices. We trace circulation patterns in all three phases.

We will work with our non-academic collaborators to refine our questions and to develop effective pathways to sharing our findings about how to engender different kinds of trust in different digital visual contexts.

Publications

10 25 50