Back in 2009 American scholar Cass Sunstein published an important book entitled Going to Extremes: How Like Minds Unite and Divide. Drawing on group dynamics research he asserts that “some of the best and worst developments in social life are a product of group dynamics, in which members of organizations, both small and large, move one another in new directions”. Research on group dynamics reported in the book (including a detailed Appendix outlining important studies and scientific experiments) has found that “when people find themselves in groups of like-minded types, they are especially likely to move to extremes”. This is termed group polarisation. Other phenomena that can fuel extremism are also explored such as “social cascades” in which people are relying on “what they think others think” (e.g. information cascades, reputation cascades, etc).
Group polarisation is defined as a social process whereby “members of a deliberating group usually end up at a more extreme position in the same general direction as their inclinations before deliberation began”. In other words, researchers have found that “like-minded people tend to move to a more extreme version of what they thought before they started to talk”.
A related phenomenon termed “crippled epistemology” is explored in the book. Sunstein argues this is often the cause of “unjustified extreme movements”. Sunstein cites Russell Hardin’s work in arguing that extremists often know less than they think they do and what they know is often biased (e.g. ignoring evidence that challenges their views, excluding dissenting voices). The development of such an epistemology can be self-reinforcing, promoting further extremism.
Sunstein suggests that group polarisation and related issues like a crippled epistemology help to explain a wide variety of phenomena such as terrorist cells, the deliberations of the American government under George W. Bush culminating in the Iraq war, through to corporate scandals such as the Enron disaster (in which he contends that members of the firm’s Board were inclined to take unusual risks and it didn’t contain a sufficient amount of diversity and/or dissent).
Two general accounts of group polarisation are explored in Chapter 2. Some argue that it brings out already existing views, which previously weren’t articulated or conscious – a self-discovery process. Another account points to entirely new views being generated by group interactions.
I was prompted to look further into these phenomena by a case that I’m analysing in my doctoral research – a scenario exercise called the ‘Future Fuels Forum’. In this scenario exercise many participants who were highly concerned about peak oil and related issues became more concerned (as a result of participating), and some participants who weren’t concerned became even less concerned. This outcome was initially surprising because the participants were exposed to the same information and participated in the same exercises and discussions during this process. An important group dynamics-related discovery was that during the process participants formed like-minded sub-groups.
A related aspect of the case is that the more ‘extreme’ scenarios were emphasised (e.g. in the main public report). Less extreme scenarios were discussed in a separate modelling report.
Sunstein explores some underlying causal mechanisms that can result in group polarisation. These include: (i) the information pool in the group (i.e. exposure to information and arguments that is skewed in a particular direction); (ii) corroboration leading to increased confidence and, thereby, also to increased extremism; (iii) and the process of social comparison and related reputational concerns, where “people want to be perceived favorably by other group members, and also to perceive themselves favorably” which can lead to adjustments in a person’s position towards the direction of the dominant group position. In addition, Sunstein argues “social pressures, both informational and reputational [noted above], are heightened if group members have a high degree of solidarity and affection”.
Social status and networks can be influential too. An example is overestimating the knowledge and/or performance of high-status group members and giving them an unwarranted degree of deference. Social network effects are also highlighted: “social networks can operate as polarization machines because they help to confirm and thus amplify people’s antecedent views”.
The degree of uncertainty (or confidence) is important as conveyed in this passage:
“Recall that people moderate their opinions if they are unsure whether they are right. And other things being equal, confident people have an advantage in social deliberations. It follows that if group members tend toward extremism, and if the group is dominated by confident people, it is exceedingly likely to shift [i.e. to a more extreme version of what group members thought before they started to talk]”
Sunstein is careful to point out that group polarisation can occur for “perfectly sensible reasons” and that “sometimes extremism is defensible or even right”. These are important qualifications. But he also points to “two major wrinkles”. First, shifts in views can be motivated more by a concern for one’s reputation than good information and good arguments. Second, he argues that “much of the time, people do not seem to have anything like an adequate sense of the partiality and skew of the groups in which they find themselves”. Judgements can be warped when “we act as if those groups reflect an impartial sum of information, even when there is a systematic bias”.
A related process is called biased assimilation. Imagine that a mixed group is convened with people that equally hold opposite views – e.g. half of the group strongly favor capital punishment, and half strongly reject it. The group is then confronted with balanced information which provides arguments supporting both views. Rather than observing moderation of views, and hence depolarisation, what can happen is selective assimilation of information supporting their position and ignoring powerful contrary evidence. Rather than finding a middle ground, each side often “digs in”.
This insight that deliberation can drive people to sharper extremes rather that bring them together is a troubling one. This is something I’ve observed in deliberative engagement processes.
Additional chapters explore social movement dynamics (Chapter 3), ways of preventing “unjustified extremism” (Chapter 4), and “good extremism” e.g. civil rights movement, the fall of apartheid, etc (Chapter 5). Chapter 5 also introduces interesting new concepts like “enclave deliberation”. Sunstein notes that there is danger in such enclaves – which, for example, “may end up in violence and put social stability at risk” as per terrorist cells, cults of various kinds, etc – but also argues they can be socially beneficial in promoting the development of new positions on social issues.
In the context of prospective practices there are many potential applications of such theory, such as:
- Explaining the outcomes of scenario exercises, such as the greater polarisation of viewpoints seen in the Future Fuels Forum or the strengthening of preexisting inclinations;
- Applying group dynamics theory in process design and facilitation; and
- Using associated theories and concepts, such as social cascades and “enclave deliberation”, to better understand the social dynamics of extremism in futures thinking.
I intend to explore this further and would welcome hearing others’ thoughts/suggestions.
Lastly, I didn’t feel that Sunstein adequately explored when “extreme movements” are or aren’t justified. People pushing to abolish slavery were once “extremists”. Can some climate activists and sustainability activists today be considered extremists (I think, in many respects, the answer is yes), and is this justified? A balanced reading of history reveals that extremism is not always a bad thing. Sunstein similarly states that “it is obvious that extremism is not always bad” and argues that successful reform movements can be aided by polarisation and extremism. However, it is also clear than extremism can impair reform movements. More work is needed to unpack and understand these complexities.