Controversy Mapping. Tommaso Venturini

to classified sources, there is little chance that you could study them seriously. If a debate does not disclose new information, then its investigation risks degenerating into a conspiracy theory in which collective dynamics are explained through the deployment of some surreptitious plan rather than as the outcome of many diverging and mutually interfering agendas. As Karl Popper beautifully puts it:

It is one of the striking things about social life that nothing ever comes off as intended. Things always turn out a little bit differently. We hardly ever produce in social life precisely the effect that we wish to produce, and we usually get things that we do not want into the bargain … the people who approach the social sciences with a readymade conspiracy theory thereby deny themselves the possibility of ever understanding what the task of the social sciences is, for they assume that we can explain practically everything in society by asking who wanted it, whereas the real task of the social sciences is to explain those things which nobody wants. (Popper, 2002 [1963], pp. 165–8)

      On the contrary, considering controversies that are accessible because close at hand is always a smart move. Controversy mapping is a sort of “secondary analysis” of the scientific debates that other researchers are engaged in. Hence, controversy mappers who work in academia can find great subjects just by interviewing their teachers and colleagues. Likewise, local controversies (e.g., connected to architectural projects; Yaneva, 2011) are often attractive choices, especially if they take place nearby.

      Finally, and for reasons directly connected to the point we just made, controversies that develop online are cartographically convenient because they leave digital traces readily at hand for everyone who has a computer. As in the advice to privilege scientific controversies, digital should here be taken in the widest possible sense, and certainly not only limited to dominant platforms. In the last few years, these platforms have grown especially popular among social scientists (and even more among commercial and political marketers) for the way in which their APIs (application programming interfaces) redistribute some of the records they collect. Yet, platforms are by no means the only digital sources that controversy mappers can and should exploit and, depending on the object of research, other datasets may be more interesting.

      We are not saying that social media records are useless for social investigation. Quite the contrary! Google, for example, can be suitable for studying the cycles of online attention (Choi & Varian, 2012); Facebook can offer a decent proxy of Web sociability (Rieder, 2013); YouTube can be interesting for comparing mainstream and marginal media discourses (Arthurs et al., 2018; Rieder et al., 2018); Wikipedia is great for studying knowledge debates (Niederer & Van Dijck, 2010; Borra et al., 2014, 2015). Still, digital mapping can and should extend further than social media. It is one of our duties as critical scholars to interrogate and oppose the project of Silicon Valley giants to claim online sociality as their fiefdom. One way we can do this is by refusing to consider the data collected by these corporations as the gold standard of digital traceability and engage in other forms of digital fieldwork (Venturini & Rogers, 2019; Perriam et al., 2020), another is to critically examine the consequences of repurposing their APIs for online controversy mapping (Munk & Olesen, 2020).

On the risk of being overwhelmed

      It should be clear from the discussion above that the most common pitfall in controversy mapping is the mismatch between the richness of the terrain and the resources available for its investigation. As a rule of thumb, controversies are always larger and richer than they seem at first glance and, while few mapping projects shipwreck for lack of ambition, many do for lack of modesty.

      Sometimes it is possible to carve a feasibly sized topic out of an unmanageably large debate by focusing on a specific sub-controversy, but this carving does not always make sense. In studying nanotechnology, for example, there is no point in trying to single out a debate about a particular nano-molecule. Despite the diversity of such molecules, the advocacy around them tends to address them globally rather than one by one, which the cartographer must respect. The point is not to artificially tear large controversies into smaller pieces but to identify debates which are commensurate with available time and resources.

      In lively controversies, actors go to every available agora to plead their cases. They will make documents available, offer interpretations, and even explain the arguments of their opponents (only to contradict them, of course). If you choose a good controversy, its actors should provide most of the inquiry, leaving you with “just” the task of reporting on their work and to make sure that it sheds light on all the important aspects of the situation. Indeed, most controversies will typically open sociotechnical black boxes selectively. Consider, for instance, the Fukushima Daiichi disaster, where the 2011 Tōhoku earthquake and tsunami broke open not only a Japanese nuclear reactor, but also several of the black boxes underpinning nuclear power production – such as the safety of generation-2 reactors and the resilience of plants in seismic zones. In the aftermath of the disaster, this destabilization had consequences for energy policy across the globe, encouraging a shift away from nuclear power, in many cases towards fossil fuels. Of course, fossil alternatives have significant disaster risks in their own right, but these risks were not made available for inquiry through the Fukushima Daiichi catastrophe (though they were later brought into the debate by environmental activists). Controversies apply no “fairness” in the way they distribute scrutiny across claims and positions. They shed light on some collective phenomena but cast shadows on others.

Still, the problem of “not observing enough” tends to be secondary in controversy mapping. Imagine being a gold miner patiently sifting through the gravel of a mountain stream. At first you are not finding much, but at least you can set your own pace and control your workflow. This is the situation at the outset of most controversy mapping projects. Now imagine that every pebble in the river decides that it is going to help by jumping into your sieve. Buried under the avalanche, the mountain scenery loses most of its appeal. This is a caricature but an accurate one: controversies can be exasperating for the sheer quantity of arguments, documents, evidence, and questions that they impose on the cartographer. All controversy mapping projects therefore tend to go through their own form of “hype cycle” (see figure 10) (Fenn & Raskino, 2008). Cartographers are usually optimistic at the outset: their subject looks interesting; the simplest Web search provides tons of documents, interviews and position statements. But as the investigation proceeds, their optimism tends to fade: the more they inquire, the less they seem to make sense of the controversy. Things turn out to be far more intricate than expected. The most visible actors are not necessarily the most influential; new and unexpected characters sprout everywhere; the arguments are more subtle and charged with countless ifs and buts; simple ways of grasping the debate prove inapt, but sticking to the raw complexity produces very confusing maps. A long plateau of disillusionment must be crossed before things start to make sense again (and from there, it is still a long way to reach a satisfactory mapping).

      Figure 10 The hype cycle of controversy mapping (created by the authors; released by the authors under CC BY-SA 4.0).

      Acknowledging the overabundant nature of controversies leads directly to the question that haunts all controversy mappers: where to stop? This question is crucial and less trivial than it may appear (Strathern, 1996). Most controversies (and most social phenomena) do not have clear-cut boundaries. If you follow the web of collective actions, you will find no obvious points to stop. The social fabric is continuous and extends indefinitely from interference to interference. Knowing where to stop is always, at least in part, a subjective decision, as observed by John Law and Evelyn Ruppert:

Where we set the boundaries is an analytical and political matter … [it] depends on our own questions, and our own agendas. It also tells us that whatever the stories that we tell … these will only ever be partial. There can be no exhaustive analysis of devices. There is always more to be told, more and different, about the work that they do, and about what and