Our paper focuses on the socio-semantic configuration of a digital Twitter territory, in particular around the publication of the Inter-governmental Panel on Climate Change (IPCC) panel Working Group 2 and 3 reports in March-April 2014. We show that this portion of the public space is characterized by specific sets of structural positions which, in turn, correspond to specific semantic positions, in terms of discourses and alignments in the underlying debate on climate change.
We distinguish an “IPCC heart”, actively discussing IPCC, from an “IPCC shadow”, which more anecdotally mentions IPPC and is likely to correspond to the remainder of a public space minimally interested in IPCC-related reports. More precisely, from the 35k distinct authors of about 90k English-speaking tweets mentioning “IPCC” during this period, we manually identified the top 2% having more than 15 tweets in the corpus. This yields a “heart” made of 629 authors who produced about 27% of the total amount of tweets (retweets were excluded to focus on utterances and links originally made by individuals rather than duplicated from others). We manually identify these heart users, according to their type and, most importantly, alignment in the climate change debate. We assume that users may have a specific alignment, depending on whether they are generally critical or supportive of anthropic climate change, or unaligned. User types simply correspond to the kind of entity behind the account, sorted into seven categories: corporate, research, governmental, individual, media, NGO-related accounts, or “other” (for uncategorized accounts, such as individuals whose category was not possible to identify from their Twitter account).
The goal of our analysis is to assess which positions are being occupied within this specialized digital public space, and by whom. We note the existence of four typical positions: stars (both heavily cited and active), famous users (cited without being very active), curious users (being rather active without receiving much attention), and silent users. In turn, these meta-positions are occupied by users whose type, alignment, and discourse very significantly diverges from a random baseline. Interesting observations relate to the position of critical users (who, albeit the minority, are both most visible and cited in this arena) among ”star” accounts featuring under-represented and unaligned media and governmental agency accounts. Casual users, on the other hand, essentially remain supportive yet invisible, while relying on concepts – discussing mitigation, cost, adaptation issues – which are in phase with those evoked by more institutional accounts and in stark contrast with those of the critical stars, who focus on scientific concepts (hockey stick, AGW, data, models and predictions, even referring to scientists and science per se). For climate change debate, it is interesting to note that while critical users are in minority as compared to supportive and unaligned users, they typically occupy star positions, indicating a dominant (highly visible and highly cited) position in the Twitter debate. In a nutshell, our socio-semantic analysis sheds light on the intertwining of cognitive and structural position of a tightly delimited territory of the online/Twitter public space.