Colin Camerer’s fascinating talk Cognitive Neuroscience and Regulatory Paternalism, given at the RSA last Friday to publicly launch the Social Brain project, is now available as a podcast.

Behavioural economics tends to go hand in hand with libertarian paternalism, which is the idea that we guide behaviour to certain ends without forbidding anything.


But what about going further and using it to create self-directing normative behaviour (to coin an ugly phrase)? Think of the Dutch idea of removing or reducing traffic furniture and signals.


This puts the responsibility back with the communities involved – drivers slowing down not because of sanctions, but because of concern for pedestrians and wariness of askance looks from passers by.


People going round a roundabout


We know that certain sorts of sanction and rule-enforcement make people think of decisions as bearing economic rather than human cost. And we know that how we perceive others to perceive us is a big influence on how we behave. So why not harness these insights to put power back with communities where we can?


Last Friday I attended a fascinating seminar on neuroeconomics at the Wellcome Trust Centre for Neuroimaging. Over the next few days I’ll be blogging about questions and issues that arose from these events.


I’d just like to say something quickly about Tim Harford’s contribution. Tim’s personable and open demeanour belie a keen, subtle and formidable economic brain. His contributions to the day were many. But in general, he had three worries:


1)     that behavioural economics (BE) and neuroeconomics (NE) have a few interesting things to say about economic behaviour and how markets function, but there is no need to rethink the rational choice theories of neo-classical economics;

2)     that proponents of BE and NE (usually popularisers rather than actual scientists) tend to extrapolate grand conclusions about paradigm changes in how we think about economic decision-making, and decision-making more widely, from a few meagre (often experimental) results;

3)     that policies based on BE and NE still need to do the hard economic number-crunching work of thinking about long-term effects of policy, and politicians wowed by the ‘cool factor’ of BE and NE have forgotten to do this.


I think Tim is absolutely right to urge caution. He is an analytical thorn in the side of those who hail a new era of post-individualistic thinking about economics. In his exchange with Pete Lunn in Prospect he both brought doubt on the validity of the claims of BE and argued that good old-fashioned rational-choice theory can explain anything BE can. So given the cognitive virtue of conservatism about theories (if it ain’t broke don’t fix it), why all this talk of a paradigm shift?


I think there is great value in Tim’s questioning of the eager neophytes. But I think there is an ideological bias in his position. For example, on Friday he argued that one could make the case for saying that markets had not failed at all in the credit crunch. Rather, they had told us in no uncertain terms that complex derivatives were stupid things to invest too heavily in (which is the conclusion George Soros apparently drew).


But surely, Tim’s ideological commitment to the individualism of neo-classical economics is on display here? Crashes, crunches and depressions have happened before many times – investors creating bubbles is no new thing. The point is that we seem incapable of learning some kinds of lessons. So perhaps we should think about policies that reflect our shortcomings rather than vainly hoping we will rid ourselves of them? And BE and NE can do two things in this regard:


1)     give us a precise idea of where our cognitive shortcomings lie;

2)     help us to design systems and institutions etc in light of this knowledge.


But seeing things this way is a kind of paradigm-shift is it not? It’s accepting that we model economic behaviour, at least in part, in terms of the cognitive idiosyncracies of human beings, rather than in terms of an idealised rational agent.

There is obviously a psychological dimension to the Isreali/Palestinian conflict. One aspect of it is the trauma experienced by both sets of people (if speaking of sets of people isn’t already part of the problem – see Chris Dillow’s post on the inherent dangers involved in such broad-brush thinking). The Israeli trauma is rooted in the Jewish experience of the Holocaust. The Palestinian trauma in the Nakba of 1948 when many people either fled or were forced to flee their villages.


Discussion of this subject is a minefield. Here is a very small contribution to it. Jonah Lehrer recently wrote about the brain’s hard-wired capacity for ‘meta-cognition’. This is where we think about thinking (see my post on the evolution of the social brain in terms of thinking about thinking in other people’s minds). This capacity is incredibly useful, allowing us to reflect on our cognitive techniques and refine them.


But there is an unfortunate side-effect to this amazingly useful capacity for meta-cognition. The brain checks on itself to see if a thought process has started, keeping it in mind so we can alter it if it ‘jars’ with other information we are receiving. This is why if we try not to think of something we often do so anyway to our chagrin.


When a thought is particularly powerful or vivid (for example painful) this meta-cognitive process becomes pervasive and takes centre stage. And this is basically what happens in trauma (there may be other factors at work in the experience of trauma, but this meta-cognitive element seems to be central).


One thing that gets discussed a lot with regard to the Isreali/Palestinian conflict is whose trauma is more important, legitimate or painful. That seems like a wrongheaded discussion to me. Denying, cheapening, even simply quantifying the trauma of ‘the other side’ contributes to the dehumanising dynamic that makes this conflict so intractable.


I don’t want to suggest that the actors in this conflict are being driven by their brains and are blameless. But perhaps one of the reasons for the inability of some on ‘both sides’ to see the trauma and suffering of the ‘other side’ is the pervasiveness of their own trauma. In terms of mediating the conflict, remembering this fact might be useful.



Something I have come across recently is signalling theory. Part of the theory (as I understand it) is that humans communicate almost entirely through conventional signals where there is no inherent cost if communication is dishonest or not (compare a ‘natural’ signal such as one antelope communicating to its cohorts that there is a pack of hunting lions nearby – the costs to the ‘honesty’ or otherwise of this signal are immediate and extremely high). Signalling theory says that the costs of deception must be high enough to keep signals within the expected range of honesty of signallers and receivers.


Studies of how social networks work by the sociologist Granovetter have suggested that lots of weak ties in a large network are best for the free-flow of information, but that clusters of strong ties may be better for getting jobs done where intense effort is required.


Following on from my post yesterday about trust and social networks, perhaps signalling theory and Granovetter’s work can tell us something about why the financial system recently broke down? For example, perhaps the costs of deceptive signalling are too cheap in large networks of financial workers? Perhaps the very fact of trading in derivatives that nobody really understands encourages a culture where deception is accepted as the norm (with the ‘irrational’ discounting of future consequences beloved of behavioural economists compounding matters)? Perhaps the networks that structure international finance are geared to the fast-flow of information rather than close ties that ensure intense effort and scrutiny?


These are open questions but it is clear that how companies and institutions organise themselves in terms of network size and composition is something that is very complex and demands a lot of thought and planning. And it seems to me that there is a distinct lack of awareness of this fact.

As Matthew Taylor has reported in his blog, we hosted on Tuesday an RSA/NIACE event on neuroscience and life-long learning. I want to pick up on two points that were made by speakers. First, Sarah-Jayne Blakemore talked about the ‘seductive allure’ of neuroscientific explanations – that even when neuroscience adds nothing to an explanation, the mere use of neuroscience language, or images of the brain, adds gravitas. Second, Paul Howard-Jones mentioned briefly that when teenage learners were taught about the links between brain function, learning and environment, they were more motivated to engage with learning and change their behaviour.


I wonder if these points are connected? Perhaps it is just our empiricist culture, but there is something convincing about literally seeing images of the brain. Images of neglected children’s brains really ram the message home that early years deprivation is a terrible thing. But we know it is anyway, so why is it that brain images make such a difference? I would conjecture that we associate the brain very closely with our sense of agency – our sense of self as a source of actions and decisions. And because this agency is so closely bound up with our identities as individuals, brain images both draw on our empathy and make us think very seriously about possible harmful affects upon ourselves.


I wonder whether it is viable to use neuroimaging to inform certain cohorts of young people about how their behaviour/environment affect their brains? Perhaps this would add, in explanatory terms, nothing more than ‘seductive allure’. But if it made such people think seriously about their agency and empowered them with the motivation to change their behaviour (where they could), perhaps this doesn’t matter?

One aspect of the financial crisis that hasn’t been examined is the role of complex social networks and relationships of trust. There are some reasons to think that networks beyond a certain size and with certain sorts of internal structure might lead to a breakdown in trust.

As I mentioned in my post, last week I attended a fascinating conference on the social brain and social networks. Two of the talks at the conference dealt with the issue of social network size and composition in terms of the evolved parametrics of the human brain (an organ designed to compute complex social relationships). The basic point was that there are limits beyond which networks become unmanageable and a dramatic drop-off in emotional closeness occurs. The upper limit is thought to be about 150. But there are also limits on the internal structure of the network. As far as I understand it, although ratios vary according to the propensity to communicate and extroversion of an individual (etc.), it appears that the best functioning networks are ones which scale up at a ratio of 3 from a core ‘inner’ network of close friends and/or family.

A common pattern (again, as far as I understand things) is an inner network of, say, 5, connected to a network of about 15, connected to a network of 45 and an outer network of over 100 (the scaling is not perfectly manifested in the network composition). As we travel from inner to outer networks several relationship characteristics reduce in intensity and capacity:

1) emotional closeness, including relationships like trust;

2) ‘theory of mind’ (what we understand about how someone thinks, their beliefs etc., what we think they are likely to do in given situations);

3) time spent with individuals.

For primates, emotional closeness and theory of mind (if that is not a misnomer) require immense amounts of time to be built up through grooming rituals – some primates only need to spend 2% of daylight hours on grooming for hygiene reasons but in fact spend 20%. It is estimated that humans would need to spend 40% of daylight time on grooming to sustain the complexity of the social networks they move within. But humans have language and vastly enhanced memories thus reducing the contact time necessary for emotional closeness and adequate theory of mind to develop and be sustained (think of the emotional closeness a well-chosen word can generate and how long that closeness can last).

The jury is out on what governs the size and scaling ratios of human social networks, but the influencing factors are likely to be a combination of the three relationship characteristics listed above.

My thought is this. Perhaps the technological, global and corporate world of finance allows individuals to hold together networks of such size that little is known about what anyone in the network is likely to do, believe or think. Moreover, perhaps the scaling ratios are skewed, so that emotional closeness is lost (the inner networks are too big perhaps, or the overall network size too large). And perhaps because of all this too little time is spent on developing trusting relationships.

These factors together could contribute to a kind of trust developing which lacks the normal interpersonal knowledge, closeness and time commitment. So perhaps the corporate world should consider the evolved parametrics of the social brain when organising its institutional structures?

The Social Brain project tries to wed together three elements:


1) big ‘paradigm ideas’ or philosophical ‘meta-norms’;


2) new knowledge in decision-making theory (largely neuroscience, the behavioural sciences, the study of forms of social organisation);

3) practical application of 1) and 2) to social challenges.

This is to some extent an ambitious remit. But if we don’t include the big paradigm ideas we miss something out. This is because thinkers, practitioners and policymakers can become gripped by a single view of things that is so commonplace it goes unseen. The behavioural economist Shiller predicted the credit crunch on the grounds of a kind of mass groupthink. We might say that those running our western societies (and perhaps some non-western societies too) have been gripped by a form of groupthink for the past thirty years or so. What has held them captive has been a conception of human decision-making processes as articulated by the simplifying assumptions of ‘economic man’. These assumptions were that consumers and traders alike acted on purely rational and self-interested grounds. This made the job of Government the facilitation of markets that did not obstruct the smooth functioning of said rationality. But of course we are now living with the aftermath of policy driven by these simplifying assumptions – the credit crunch and an impending global depression.

So it’s important to consider the big ideas because they become internalised and act as ‘meta-norms’ that govern our thinking without our really being reflectively aware of this fact.

On the other hand, the point of introducing into the debate new knowledge in decision-making theory is twofold: it can be used to put to bed the defunct paradigm in the face of stalwarts who won’t give it up; and it can be used to help construct a successor paradigm that is better at predicting behaviour and facilitating the kind of socio-economic world we want.

And of course it is important to bring in practical application. For there is little point in connecting the big ideas with new knowledge in order to construct a new paradigm if this does not gain traction on social reality. Moreover, there should be bottom-up feedback that informs and fine-tunes the policies that the paradigm yields, lest groupthink set in again.

On Tuesday I went to a conference that was the product of a project that tries to weave together all these three elements (a project after the heart of the Social Brain project and it’s namesake too!).

One of the project’s big ideas is that the brain has evolved into the rational information-processing hub it is because it has evolved to cope with considerable social complexity. A large element in this complexity is representing to oneself other people’s thinking in order to build alliances and coalitions (e.g. friendships of different orders of closeness). This is something we do without thinking (as it were). But in fact, if we reflect on it, this kind of mental representation is incredibly complex. For example, we can understand with relative ease a sentence such as:

(a) ‘Jack thought that Jill was thinking what he was thinking and that both desired what their parents wanted.’

But this sentence stands for an incredibly complex thought. A regular old first-order thought such as ‘the apple is red’ is about an object, an apple. But the thought ‘Jill is thinking about Roger’ is about what Jill is thinking about. This means it represents both Jill and what she thinks – it has two layers of ‘aboutness’ or what philosophers call ‘intentionality’. The sentence (a) quoted above has five layers of intentionality! The social brain theory in evolutionary psychology holds that it is representing this kind of complexity that primarily caused our brains to evolve to the size they did.

This challenges the paradigm idea of individualism – that our brains evolved solely to represent the world truthfully in order that we might be able to better predict its behaviour and thus satisfy our self-interested desires (so that, for example, we be able to obtain food better by understanding the causal relation between a pack of hyenas following a scent and there being a dead animal somewhere in the vicinity). Of course our brains did evolve to be as big and clever as they are to do this. But social brain theory simply challenges the idea that this is solely what they evolved to do.

Individualists with this view of human nature (like Hobbes or Locke) then proposed a theory of social cooperation where humans come together in order to better satisfy their individual desires. But if the social brain theory is right, such theory and its presumed individualism is bunk.

This is a great example of how new knowledge is challenging a paradigm idea and giving us a clue about what to replace it with – a thoroughly social conception of human decision-making. And now that ‘economic man’ has fallen, that is just the kind of conception we need to better predict behaviour and create the kind of world we want. But there’s more. The project I visited yesterday applies knowledge from evolutionary psychology to policy – it also attempts practical application. For example, there seems to be an optimal maximum number (150 – dubbed ‘Dunbar’s number‘) beyond which social organisation becomes difficult (emotional closeness dissipates, the cognitive task of keeping in mind the members of the group’s attitudes and alliances becomes too hard, peer pressure as social sanction for the sake of social cohesion begins to fail). If you organise – say – institutional departments in excess of this number they will begin to function badly, for they are fighting against the evolved parametrics of the brain.

So there you go. A very good example of the kind of integrated three-level approach to understanding human decision-making to which the Social Brain project is committed. We are not alone.

I shall be blogging on other ideas that came up at the ‘Social Brains and Social Networks’ conference over the next few days.

I read in the Times yesterday that the gap between public pessimism and private optimism has reached new limits with regard to the economic outlook for the UK in the coming year or so. Most people when asked about their own prospects are bullish, but regard the prospects for the economy as a whole as dire. This appears to be a form of irrationality or cognitive bias – if a vast majority of individuals are going to do okay, then how will the economy as a whole nose-dive?


Perhaps we just love to moan. But beyond that explanation, the gap is one instance of the common psychological phenomenon of over-estimating our individual prospects, attributes and capabilities.


What I want to briefly suggest is that this phenomenon is connected to a conception of ourselves that goes quite deep (and is for that reason largely unseen and unquestioned). We think of ourselves as almost completely in control of events, as little centres of autonomy, outwards from which flows the coherence and consistency of our individual lives. This is the individualistic paradigm I questioned in my last post on Self-help, individualism and the social brain.


But much recent research puts this idea of an almost wholly autonomous individual in doubt. Apart from the now fairly well-known phenomenon of unconscious or semi-conscious cognition (popularised by Malcolm Gladwell in his book Blink), our agency is in large part defined  by relations to other people in social networks. As I have commented before, a capability central to autonomous agency such as self-control is in fact dependent on what Avner Offer has called ‘commitment devices’ – relationships with social structures that enable a person to keep herself on track. (Think of the advice of a good friend, parent or sibling. Or perhaps even something as ephemeral as thinking what a character in a book, play or TV show would do in a certain situation.) In fact, the very production of the neurotransmitter that enables self-control (Serotonin) is dependent on empathic relations with others.


A recent study of happiness suggests an epidemiological understanding of its spread and sustenance. A person who is happy is a node within a network of people that she has meaningful and fulfilling relationships with. Happiness then spreads along the network like an epidemic. When someone in the network is sad, this spreads too. These are of course just the networks of concern that structure our emotional lives. But note, the decisions a person makes regarding her happiness – perhaps the biggest decisions she makes – are in large part dependent on the other nodes in the network (her friends and family).


So in the language of Mark Earls’ Herd we are much more pull than push – much more of the most important decisions we make are dependent on thoroughly social relations that to some extent exceed our control (yet we still maintain some control, we are not all pull rather than push). If we internalised this fact – if we really learnt it – perhaps this would go some way to closing the gap between private and public pessimism. We would realise that we are not the all-singing-all-dancing super agents that we take ourselves to be. And perhaps then we would start to see the public realm as less alienating (as more of the realm in which we ourselves as individual agents have our efficacy).


How to bring about this change? We could start with Education. The RSA’s education charter puts learning as part of engaged project work at the centre of its ethos. It also holds that children should learn the skills (including social and emotional skills) needed to enable them to realise their potential. Part of the sea-change we are trying to bring about is getting kids to think of themselves as active social agents rather than islands of autonomy divorced from the public realm.

The self-help movement was a real late twentieth century phenomenon. It was (is) quite different from the self-improvement societies of (say) Victorian Britain, or the (often somewhat cultish) self-improvement societies of the twenties and thirties (the nadir of these societies being Hitler’s National Socialism, which had its roots in cult-of-the-body groups from the twenties). All these movements were fundamentally social. But self-help since the liberal individualism of the sixties has been thoroughly individualistic (think of the Tom Cruise character in Paul Thomas-Anderson’s flawed masterpiece Magnolia as the nadir of this kind of movement).


The following quote typifies a common view; that the self-help phenomenon is an achievement of Enlightenment culture (a culture that is optimistic about human potential and committed to social progress):


I would say that, over the last twenty-five years, the biggest triumph of the Enlightenment view is that people have grasped the concept of their own happiness as a real goal in life. You can see this in the emergence of the self-help industry. Here is a huge industry that hardly existed a generation ago. People today go to seminars, and take classes, and buy books, for no purpose except to be happier in their personal lives. Of course, a lot of it is garbage. There’s a lot of self-indulgence, irrationality, and subjectivism involved. But the very fact that the self-help movement exists is a triumph for the individualist, Enlightenment outlook. (“The State of the Culture, 1997,Navigator, Vol. 1, No. 1, p. 11)


But this view also equates Enlightenment culture with individualism. But does this equation hold? Here at the RSA one thing we are concerned about is what is often referred to as ‘the retreat to the private sphere’ – the withering of concern for others and collective social responses to problems that is driven by a private individualism and a public sphere that is highly bureaucratic and doesn’t engage citizens (so the way the ‘public’ sphere is organised reinforces private individualism). As an organisation we recognise that this public/private sphere combination is an impediment to social progress. We all moan about the retreat to the private sphere, but we also live highly individualistic lives. The RSA thinks this has to change if we are to live in the more pro-social world we would like, and deal with the problems we face (such as climate change, rampant and entrenched inequality). For these problems require collective action and thus an engaged citizenry.


So the self-help phenomenon, for all it might do for individuals, looks like part of the problem not the solution. And if it is the highest achievement of Enlightenment culture recently, then perhaps that culture is in trouble?


I don’t think so. It’s true that the Enlightenment from the start was always individualistic. Kant’s dictum ‘think for yourself’ is expressed in the singular after all. But it is also about taking note of scientific research – being committed to a continual non-dogmatic re-appraisal of our beliefs in light of what science discovers about the world, and indeed, for science itself to be an ongoing journey taken with an open mind.


If we look at Enlightenment culture that way, then self-help culture need not be its recent zenith. In fact, what science now tells us about the brain seems to refute individualism. I am attending a very exciting conference tomorrow hosted by my project’s namesake Lucy to Language: The Archaeology of the Social Brain’.  The conference title is ‘Social Brains and Social Networks. The basic theme, as far as I understand it, is that our brains have evolved to function within social networks. For example, the adequate production of the neurotransmitter/modulator Serotonin is designed to occur within empathic relationships. Take the relationships away, and problems ensue. Or, another example: feelings transmit information about help we need to others, even if we haven’t realised we need the help ourselves. And Daniel Goleman has popularised the idea that the production of certain mirror neurons makes possible altruistic behaviour beneficial to social cohesion.


Self-help culture then, seems out of kilter with what we know about the social brain. If a person wants to be happier she is better off working on her friendships and relationships. If a person wants to increase her self-esteem then she should approach this through her relationships with others too. If a person wants to be more effective in the world, then she should work on her social and emotional intelligence. The self-contained individualism of self-help culture helps with none of this. Perhaps as that culture catches up with the science, it will become less individualistic. Then it might actually be able to contribute to the wider problems we face, such as the retreat to the private sphere. And it can still be an expression of Enlightenment culture, just not individualistically conceived. As Sartre said, ‘Hell is other people’. But Sartre also recognised that a person’s sense of  self is co-constituted by her ‘being-with-others’ – her perception of herself as an entity is mediated by states of mind such as shame which fundamentally related to other people.


So Sartre foreshadowed the idea of the social brain. He recognised that everything that’s important to us – happiness, self-esteem, self-efficacy – are socially constructed (everything that self-help culture places ‘inside us’, is actually both there and ‘outside us’ in others as well). But he didn’t like the idea of such dependency. But what’s not to like? For the individualistic self-reliance he was enamoured with has been perhaps the largest impediment to social progress (something he was also enamoured with) in the last thirty yeas or so.