Bridging the computational and visual turn: Re-tooling visual studies with image recognition and network analysis to study online climate images

In this article, we argue that to capture the liveliness of how visual public debates like the climate controversy unfold online, we must replace snapshot and single-platform approaches with a method that can capture their temporal and cross-platform dynamics. We suggest that such a methodology could be assembled by combining image recognition, visual network analysis, and a quali-quantitative approach within a digital methods framework. We demonstrate the potential application of the methodology in a two-fold case study of 1) how the human–nature relation is visually depicted on Instagram and Twitter, and 2) how visual genres in the climate debate on Twitter change from 2015 to 2017. Through these experiments, we analyse more than a quarter million social media images to produce novel insights about the climate debate, while showcasing how the computational and visual capabilities of social science can be bridged to open up opportunities for mapping complex visual debates across platforms and time.