Visualizing word-image relations in Instagram posts

For my DECRA project, I analysed Instagram posts that were made by members of the public in relation to the Australian federal election of July 2 2016. I collected a total of 6299 posts including a photograph and caption (verbal text and hashtags), over a 5-day period from 30 June to 4 July. All of these posts made use of the hashtag #ausvotes. My aim was to analyse how Instagrammers express both affiliation with a political party (e.g. photographing themselves wearing a Greens campaign t-shirt, or explicitly stating in the caption ‘I voted Green’) and how they distance themselves from a particular party or politician (e.g. in the use of the hashtag #fuckyouturnbull). Sometimes they do both within one post (e.g. photographing themselves handing out how to vote cards for the Labor Party while using the hashtag #putthelibslast in the caption). My analysis demonstrated that Instagrammers made use of multiple resources (both verbal and visual) to express (dis)affiliation with political parties and their representatives.

While analysing and then collating individual instances of these affiliative strategies is straightforward, it is difficult to get a sense of the dataset as a whole, where these strategies occur (across texts), how they are realised (across semiotic modes) and the extent to which they converge and diverge. I use Kaleidographic (Figure 1) to demonstrate all of these relations.

Kaleidographic is an experience of analysed data. In this case study, it is designed to allow audiences to experience the various ways in which Instagrammers align with or distance themselves from political parties, how these strategies combine (e.g. aligning with the Greens while distancing from the Liberal Party) and where these strategies are realised (in the image, in the verbal text or in the hashtags). To see the Kaleidographic in action, click here or click on the diagram below.

A more detailed explanation of how to manipulate the different elements of Kaleidographic and what they mean can be found here. If Kaleidographic doesn’t appear or start playing, I recommend you update or change your browser (e.g. to Chrome). If you are interested in learning how to create your own Kaleidographic, visit for more information.

You can learn more about the results of this analysis in my forthcoming publications listed in the reference list at the end of this page.

Figure 1



Caple, H. (in press) ‘Analysing the multimodal text’. In C. Taylor & A. Marchi (Eds.) Corpus Approaches to Discourse: A Critical Review. London: Routledge.

Caple, H. (in preparation) ‘Lucy says today she is a Labordoodle’: How the Dogs-of-Instagram reveal voter preferences.