Monday, 17 November 2008

Networks, nodes and hubs

I've been reading a rather obscure libertarian anti-capitalist publication recently called Turbulence. I found some articles more interesting and useful than others, but one that stuck out was Network Organisation for the 21st Century by Harry Halpin and Kay Summer. In it they suggest an alternative to both the 'horizontal' and 'vertical' models of political organisation current on the anti-capitalist left, something I've been trying to work out in a very untheoretical way for a number of years.

Their argument goes something like this:
  • Questions of political organisation have traditionally revolved around two poles of attraction, the centralised structure with clear leadership and common ideology, and the loose decentralised network with no coherent agreement on politics. Many organisations mix aspects of these two, but the debate is viewed through this lens.
  • Both structures have benefits and problems – the ability to act quickly can be contrasted with the benefits of wider participation for instance. Organisations using both models were involved in the post-Seattle alter-globalisation movement, and neither managed to sustain that movement with the vibrancy it once had.
  • We can start to move beyond this dichotomy of political models by looking at what principles define a well-functioning network, whether this be political networks, ecosystems or the internet.
  • A network consists of connections between otherwise disparate elements which are called nodes. In the context of a social movement these could be people, groups, a social centre, a website, social forums and so on.
  • In a successful network, some nodes have more connections, and are connected with more distant nodes than others. These supernodes can be called 'hubs'. This model can be contrasted with both a centralised network and a completely decentralised one. For example, in a postal network, centralised system would see all post being routed through a single hub, which would be vulnerable to overloading. A decentralised network would see long-distance mail routed through a series of local connections since it has no hubs. By contrast, a system with multiple hubs and many nodes is more efficient, even if some of the hubs replicate each others' functions.
  • Advocates of 'horizontality' have sometimes been suspicious of any hubs as signs of centralisation. Those in control of a hub may well want to sabotage other emerging hubs as competitors – a tendency visible in the anti-war movement in the UK for instance.
  • The emergence of hubs appears to be a sign of maturity in long-lasting networks. However, they must never be allowed to become static, and must remain partially redundant so that the movement as a whole doesn't depend on one hub.
  • Existing hubs should also encourage the development of new hubs and dense local connections between nodes. This often involves re-inventing the wheel with people learning new skills, knowledge and information, but this is necessary and useful for making the network resilient.
Of course they make other points too, and the article itself has far more illustrative examples than I can give in a summary, so it worth reading the original if you're interested in this sort of thing. What's interesting to me is that some of the ideas here have also emerged through the common sense of activism I've been involved in, trying to organise in an efficient but non-authoritarian way - but what the article does is thinks the implications through, puts them in a slightly more theoretical framework and feeds them back to people like me to help with organisational problems.

No doubt this is not a definitive answer, and the authors don't make any such claims, but they do offer a potentially useful approach for social movements at the moment when we need to renew the radical challenge to an unjust and unsustainable system.

No comments: