The discussions in this forum have raised some big issues, ranging from the implications of two types of social glue for the evolution of groups (e.g. Waring; Smith) to the practical and ethical challenges of seeking public policy interventions based on our scientific theories and findings (e.g. Lanman; Waring). I agree with most of the comments that have been posted and as everybody points out we need more evidence before much more can be said. But there are two issues I’d like to pick up. The first is a very basic question about whether there really are two kinds of social glue (Kavanagh; Buhrmester) or just one with varying degrees of ‘stickiness’ (Ingram and Prochownik; Reeve and Johnson). The second is about whether social glue is really the most important issue in addressing my three wishes for the world or if other sources of altruistic behaviour should receive equal or greater priority (e.g. Swann; Smith). Altruism has many sources but in my view social glue plays an especially important role in solving collective actions problems that carry high individual costs.
One or two kinds of social glue?
As an anthropologist who studies religion, I am always an outsider looking in – albeit sympathetically. And that is often how I feel when trying to untangle the intricacies of the social identity perspective on group psychology, which appears (again, from the outside) to be forcibly reminiscent of a religious organization. Although the social identity perspective has two branches – Social Identity Theory (SIT) and Self-Categorization Theory (SCT) – all followers subscribe to a mainstream orthodoxy with varying degrees of piety and commitment. Core beliefs include that everything in group psychology, including the phenomenon of identity fusion discussed in my target article, falls within the ambit of the social identity perspective. As such, fusion appears to be just one more sect under the infinitely extendable umbrella of the mother church and her encompassing doctrines. Of course, outsider impressions can be misleading and I cannot claim to have conducted long-term ethnographic research among social identity theorists. So my comments on this topic will be as tentative as they are curious and well intentioned – and no doubt shaped to a degree by my having been trained to recognize the trappings of religious fervour no matter how subtly it is expressed.
Field notebook in hand, I have spent the last few days tracking down informants. Let’s begin by giving some space to their voices (it’s standard practice in ethnographic research to disguise sources): “A while back, advocates of SIT and SCT (self-categorization theory) realized that many of their own workers didn’t believe in the main tenets of one or more cardinal doctrines, e.g. functional antagonism (that activation of personal self reduces salience of social self and vice-versa) and depersonalization (that groups members are categorically interchangeable when the group is salient).” Another informant described what happened next: “Instead of providing a formal revision to the theory, some researchers simply endorsed arguments that were in direct opposition to the original tenets. The result was that the informally revised ‘theory’ now embraced both the original theoretical ideas (e.g. functional antagonism and collective ties) and their opposites (e.g. identity synergy and relational ties). The resulting ‘social identity perspective’ was immune to falsification.”
Unfalsifiable? This sounds like a set of beliefs that can’t be resolved on empirical grounds. Like a religious system perhaps? Let’s explore this in more detail…
According to Ingram and Prochownik identity fusion is probably just an “extreme form of social identification.” Reeve and Johnson agreed with this take on things, arguing that the notion of identity fusion merely “expands the SIT paradigm.” But as Kavanagh and Buhrmester carefully argued in a series of posts, the empirical evidence points to systematic differences between fusion and identification. Kavanagh cited a body of empirical evidence showing that existing measures of social identification simply cannot subsume identity fusion. If, for example, personal and social identities are hydraulically related, what is one to make of evidence that activating a personal identity enhances the tendency for fused individuals to enact pro-group behaviour? Building on this, Buhrmester pointed out that fusion theory focuses on the causal role of relational ties to other group members as well as collective ties to the group whereas the social identity perspective is only concerned with the latter. The two kinds of social glue predict different psychological and behavioural outcomes.
What motivates efforts to make identity fusion part of the social identity perspective rather than, as the evidence suggests, an alternative theory of how groups are glued together? The motivation doesn’t appear to stem from either empirical or logical considerations but from a desire to maintain the sovereignty of a tradition. It is easy to underestimate the extent to which academics (including good scientists) can form distinctive cultural traditions that, just like any other traditions, can glue adherents to each other and to a set of values and beliefs. And this makes the study of social glue all the more complicated.
If we think of the social identity perspective as a kind of church it is a relatively new one. There is of course a much bigger and older church, called ‘social science’. As Michael E. Smith reminded us, in an informative post entitled somewhat indignantly “You Folks Should Pay Attention To Social Science”, there is a grand tradition out there, one that has a more illustrious history than the social identity perspective and a greater plethora of special terms for things, ways of talking about them, and other special customs and beliefs. When I started to read Smith’s commentary, I thought at first he was going to say that we’d made a basic error that could have been avoided if only we’d known about some previous research on the topic. But as I read on it became clear that his main point was something quite different – that there is another academic tradition over the hills that has lots of doctrines about social glue, including what to call it and how to think about it.
To my mind, however, the most thought-provoking response of all, from a big-picture perspective, was the one written by Lanman. He reminded us that when we’re asking questions about human psychology we should ask about both mechanism and function. To put this in the language of the evolutionary sciences, we should address both proximate and ultimate causation. The social identity perspective has taught us much of importance about the proximate level but when we broaden the focus of groups research to consider issues of ultimate causation we begin to understand social glue rather differently.
To appreciate why fusion and identification may be different it could help to unpack their evolutionary histories. Lanman and I hypothesize that the categorical ties studied by social identity theorists evolved to bind together tribes and ethnic groups whereas identity fusion emerged to hold kin groups together: two functions, two psychological mechanisms, and two kinds of social glue. Kin psychology (on this view) regulates behavior among genetically related individuals, facilitating exceptionally high levels of altruism towards the group, rooted in the fusion of personal and social selves. Ethnic psychology, by contrast, solves collective action problems using categorical ties based on identification with groups. As Lanman succinctly put it: “Whitehouse’s account of identity fusion as ‘psychological kinship’… which lies at the heart of the imagistic mode of cohesion… utilizes the findings of both evolutionary psychology and socio-cultural anthropology and can serve as a reminder of the insights we can reach when we move past the more exclusionist rhetoric sometimes used by scholars in these fields.”
How important is social glue?
Several contributors to this discussion pointed out that a range of factors contributes to altruistic behavior and that ‘social glue’ is only one of them. Bill Swann, for example, mentioned shared interests, top down incentives and deterrents, and ideologies that appeal to our capacities for empathy or right and wrong. Arguably, however, social glue has a special role in motivating altruism. You and I might share similar concerns about social problems, discover common vested interests, and even agree on the best courses of collective action, but I doubt whether any of this would motivate most of us to make big sacrifices for the greater good.
Not everyone shares my hunch. Some prominent activists insist, for example, that morality rather than social glue is the best way to address the major collective action problems faced by our species. One of my three wishes for the world was to eliminate extreme poverty. The moral philosopher Peter Singer famously gives away a third of his income to the charity OXFAM and he urges everyone else to do the same. I vividly remember a lecture given by Singer in Oxford, at which he pointed out the wrongness of allowing extreme poverty to persist in the world and the fact that we could eliminate the problem overnight if we all set our minds to it.
But that’s the problem. We won’t all do it. And since we know that other people won’t all do it, we typically decide not to do it ourselves. After the lecture, I sat next to Singer at lunch and I put it to him that the moral argument wasn’t going to change things. His response? It simply had to – there was no other way. But as Swann points out there are indeed other ways.
Aside from moralizing we’ve seen many efforts to solve world problems using a diversity of strategies. For example, over several decades concerted effort has been made via high level international initiatives to redistribute wealth from the world’s wealthiest countries to help the poorest and there has long been broad agreement that 0.7% of GNP is a realistic target for provision of aid. But apart from some outliers like Scandinavia, we have fallen woefully short in achieving these kinds of targets. Swann mentions various mechanisms of regulation in modern states that can be used to solve collective action problems without relying directly on social glue – we can incentivize. legislate, tax, subsidize, and do other things in a top-down fashion to tackle poverty. But I would argue that none of these approaches works very well without the right kinds and quantities of social glue.
Social glue plays a vital role in solving collective action problems in a sustainable fashion. Without it, other mechanisms deteriorate and fail. Examples are legion but to take just one from my own country: social glue was essential to setting up the welfare state in the UK but it is now eroding because of a culture of sponging and entitlement symptomatic of a progressive weakening of national cohesion. Whereas the social sciences have traditionally provided quite sophisticated ways of understanding systems of regulation – in economics, law, governance, politics, and so on – a more basic aspect of coordination in society – social glue – has not been understood so well, and figures less prominently in the thinking of policy makers and advisors. I think it’s time to rectify the neglect.
After much debate on points of detail, Gordon Ingram concluded an informative thread in this debate by taking issue with my suggestion that “when we fight back against injustice it’s because we believe that its victims share our suffering. The victims are, in an important sense, one with us.” Ingram retorted: “This, I think, is quite wrong: I predict that it is not necessary to feel fused with someone in order to feel a duty of care towards them… it comes down to the need for more empirical research: these are two testable predictions and I hope that someone will test them soon. Until then, Harvey is not really justified in making this sort of claim.” Ingram’s call for more data is well taken. And I should have been careful to emphasize repeatedly that my fusion-based explanation for altruism (including parochial altruism) is no more than a working hypothesis rather than an established fact. But by the same token Gordon is offering a counter-prediction rather than demonstrating the wrongness of mine. Although we do need more evidence, the idea that empathizing and moral reasoning are sufficient to motivate extreme sacrifice for the group warrants skepticism. By contrast, there is already quite compelling evidence that when compassion and morality are bolstered by a visceral belief that the group is me, self-sacrificial commitment markedly increases.