This post is a bit of a hodgepodge of follow up from stuff I was working on last week – we’ll see if it weaves itself into one piece by the end.
First up is a follow up on the social network analysis. Ethnobot was good enough to point me at the work of Damon Centola and his content on weak ties. Centola talks about weak ties in more detail and looks at how they work in order to pass ideas and information. As with the Granovetter work the conclusion continues to be that weak ties are the most effective for passing information but Centola goes further to talk about the ways in which weak ties can effect what he calls complex contagion. Simple and complex contagion is a metaphor taken from epidemiology and refers to the source of infection. In social network terms do you need to hear an idea once or do you need to hear it multiple times in order to act upon it? If its the latter then its a complex contagion and this is much more common if you are talking about behaviour change but still, according to Centola, track and traceable.
There are a number of different terms being used for the connection points between networks and sub-networks. Castells, for example, talks about ‘Switches’ but Centola calls them ‘Bridges’ which I think its nicely illustrative. Complex contagion requires either wide or multiple bridges so that people are presented with an idea or behaviour in different ways or from different people. Contrast with the simple contagion model of Malcolm Gladwell who’s Connectors apparently fulfil this role entirely on their own – something which is challenged by Centola’s work.
[as a complete aside you may want to read Bellweather by Connie Willis – its fiction – and takes a rather more amusing view as to how trends are set – with a solid bit of theory lurking within it.]
Earlier in my PHD I got very seduced by social capital and Robert Putnam’s work in ‘Bowling Alone’. One of the interesting aspects of working within a multidisciplinary research centre (I’m studying at SPRU at the University of Sussex) is that you are able to roam between fields to some extent. Social capital looks at the same effect as the social network analysis from the point of view of an economist as opposed to a sociologist. I’m not sure where those urban geographers (Wellman, Massey and others) come into it but this is crossover territory for people who accept that there is no point at looking at the built environment without considering the people who live within it – and I think perhaps we need to find a new category for researchers who are looking at the online world through the filter of the lives that people live within it. Ultimately I decided that, for me, the social capital analysis is too passive and looks at measuring an asset which is to a great extent unmeasurable until its tested – and if its unmeasurable then you can’t really organise around it. Social networks are measurable to a far greater degree and we are not putting any value on that measurement – we are just noting and describing the connections rather that the inherent value they may have. It does however make it important that we are able to describe these relationships in some detail.
This is not so say that a social network has a flat topography – links between nodes have different directions and strength. This question of strong and weak ties and the effects of them on information and behaviour change is about power manifests through networks – and who can wield it.
And this brings me to the other bit of follow up I was doing last week in reading a few articles from the International Journal of Communication’s recent supplement on multidimensional networks – which is worth a read. Yochai Benkler – who is someone I should have added to my piece on network society thinkers – one of his research interests is the “Effects of the networked public sphere on democracy” and his recent article”Networks of Power, Degrees of Freedom” explores manifestations of power, counter power and freedom. He looks at the way that networks and individuals exert power and uses the example of wikileaks to show how the network society has subverted traditional power structures. This is a fairly standard observation from a network society thinker but as a lawyer Benkler has an in-depth analysis of the wikileaks narrative which is worth reading.
Counter power is in fact a Castells concept that refers to the activity of resisting the imposition of power onto and indivudal – as distinct from freedom that allows you to exert your own power. Castells contribution to the IJoC supplement is a discussion of the four types of power of the network society:
- Networking Power: the power of the actors and organizations included in the networks that constitute the core of the global network society over human collectives and individuals who are not included in these global networks.
- Network Power: the power resulting from the standards required to coordinate social interaction in the networks. In this case, power is exercised not by exclusion from the networks but by the imposition of the rules of inclusion.
- Networked Power: the power of social actors over other social actors in the network. The forms and processes of networked power are specific to each network.
- Network-making Power: the power to program specific networks according to the interests and values of the programmers, and the power to switch different networks following the strategic alliances between the dominant actors of various networks.
Within the article he develops the idea of programmers – who are able to form and direct networks – and switchers who are able to connect different networks together.
Now, this is the point where the network society theorists start to trip over the social network analysis folks.
Centola’ multiple sources of complex contagion and his strong bridges are not compatible with Castells construct of the switcher (or indeed Gladwell’s connectors) – the data seems to indicate that more than one person needs to be involved if we are talking about behaviour change. Castells (and I think Benkler though more reading needed on this) do address this issue but instead of talking about multiple people refer to multidimensional relationships – power needs a variety of relationships to manifest.
There is an issue here with data vs theory and another reason why you need to be cautious with social network analysis. Though it might appear that we can connect the Centola result that indicates the need for multiple relationships to create complex contagion and that we can say that the manifestation of power is an active instance of complex contagion we are not going to be able to create the data set that would prove or disprove this theory. The multidimensional nature of the networks that Castells and Benkler are talking about are not measurable except by heroic data collection efforts as they rely on actors being able to categorise this connections where Centola just needs us to admit to a connection. We are therefore left with the slightly unsatisfactory feeling that Centola (and others) have given is a measurable theory that speaks to the complexity of behaviour change while Castells (and others) has given is a compelling explanation as to how power works through networks.
The question I ask myself then is whether or not it would be possible to look at this question at the local or hyperlocal level? Can we document the nature of the relationships of the core civic creators in an area in such detail that we are then able to track behaviour change through that network, and its connected networks? And can we then examine how this civic network connects to the democratic one in order to see whether it is the elected representatives that form the necessary bridges or are those bridges formed (or not formed) elsewhere?
I think so. Detailed SNA of this core group before and after an intervention with a specific call to action should show us how ideas have passed and who has been influential. This is however an experiment that needs some designing. We need to have a network in the first place and then we need to have some kind of traceable ‘infection’ in terms of behaviour that we can track through the network. We would also need to measure the network in terms of influenced / influencer. Alternatively we could, more simply, look at the network before and after our intervention to see whether there had been a change in the number of connections to the formal actors – and then examine those new connections in order to trace a source of the new tie.
I’ve been struggling a lot with my data collection over the last couple of months (not time) but have also been unusually dithering about getting going with it properly. In retrospect this may be a reflection on the way my thinking about this has been moving – I have been thinking about Civic Spaces (which I still am) but I should have focused on the people that write those spaces into being. Unless you go down a track of techno-determinism online space is about a manifestation of people and relationships. The technology design can help shape those behaviours in the same way as physical architecture influences offline behaviour but the dynamic force in all of this is brought by the people, the actors, within the space. If I am interested in creating co-produced civic space online then I need to think about the people who will co-create that space in at least as much detail as I consider the technological affordances of those spaces.