 Network Dynamics is the study of how networks change over time as new connections and structures are formed or removed. With the study of political network dynamics, we're asking the question of how do socio-political networks form, mature and eventually disintegrate? In this analysis, we're interested in such questions as how do things spread across the network as a function of the structure of the network of connections and various questions of network resilience and robustness to outside influences? The study of network dynamics is used as a tool to study the formation of many different kinds of socio-political networks such as criminal networks, terrorist networks, political movements, innovation networks or changes in culture etc. This approach is particularly relevant when dealing with informal political processes and organisations, where the organisation may have no formal structure but consist of purely functional connections, as would be the case in the early formation of a political protest, a new belief system or the workings of various criminal organisations. A primary factor to note is that networks are a very organic form of organisation. They typically emerge or grow out of the local interactions made by actors without global structures being imposed. Connections are typically made by individuals on a functional basis according to some cost-benefit analysis of the individual. People exist within some environments and they wish to do certain things such as move around, communicate, form acquaintances, exchange goods etc. All of these activities create connections but which connections are made and which are not made are a function of both the individual's propensity to make those connections and the limiting constraints within that environment. Most connections both have some advantage for the individual but also cost the individual something to make those connections. Thus there is both an outward force to create connections that is driven by the individual's motives to access whatever it is that is on the network but there is also some limiting factor requiring them to expend some resources to make that connection. For example we can think of a criminal network what we call a dark network. Dark networks are illegal and covert networks such as drug cartels, hacker networks, illegal flows of arms and money, terrorist groups etc. Dark networks are particularly interesting to study with respect to socio-political networks because their central defining characteristic is that they're trying to achieve something without being detected and this makes them different for most formal organizations in society which have both a manifest formal structure and an informal network structure. These dark networks in country are purely functional networks. For example if you want someone to traffic drugs across a border for you you cannot just call up some formal business to do this for you. You have to reach out to your informal network and find someone there. Individuals in such networks often do not trust each other and thus each link in the network can be seen to cost the individual and the more links an individual has the more exposed detection they are. Thus we can see here the two forces that are relevant in most social networks. The outward incentives driving the individuals to make connections but also the counter force that places the cost on every connection they make working to reduce the overall number of connections. For a network to form then the benefits to the individual must in some fashion outweigh the costs being imposed on them for forming part of that network. As an example we could think of the socio-political network forming a movement to try and change an oppressive political regime. Individuals are motivated to form the connections out of a desire to change the political order and the need to band together with others to do this. But also they face a potential cost that the regime imposes on the network by threatening them with arrest and persecution. In such a situation we can identify a number of key parameters to the development of the network. Firstly the individuals internal desire for change. This motivation can be seen to be driven by the emotions of anger and hope which are emotions that drive people into action but also we can identify the level of oppression that the individuals experience as another important factor. Oppressive regimes are maintained through fear which is an oppressor of behavior. Most social structures that regulate political systems are based on fear which works to dampen down change and maintain the system within its existing configuration. In order to overcome fear people's anger has to be strong enough and their vision of an alternative future has to be strong enough but added to this they typically have to be connected with others. Revolutions happen when people are no longer able to control their oppression and outrage any longer. One example of this being the self-immolation of Muhammad Barzizi a street vendor in Tunisia who set himself on fire on the street in response to the confiscation of his wares the harassment and humiliation that he said was inflicted on him by a municipal officer in her aides. An event that is seen to be a catalyst that ultimately led to the Arab Spring Revolutions. Most control is maintained by controlling the means of communication within a society so that those who are resistant to existing structures cannot connect and communicate with each other at large. Throughout history the control over information and communications has been the key element in regulation because they are the means through which people connect and can form counter movements. Ultimately humans are governed by the neural networks in their brains and the sociopolitical structures of a society and its operations are determined by the neural networks of the individuals and how those individuals are interconnected within the overall system. If the neural networks of the individuals conform to that of the authority and if they do not have the means to connect with each other through unregulated channels then change will not happen. Change happens when people's neural networks diverge from those of the authority creating a desire for change from within the individual and when they can connect through unregulated peer-to-peer networks. This is why the internet is potentially such a powerful political force in the world today. Not only does it provide alternative narratives that can diverge from the dominant ideology but also it gives people the potential to connect within unregulated environments. Historically in virtually all processes of political change people come together and collect in public spaces and if they can do that then they may experience a sense of solidarity that strengthens their resolve to overcome their fear of the regime. With the rise of information technology increasingly this public space where people can connect is on the internet. What we see with many movements today is that they are born on the internet and then move into public spaces. Networks are non-linear systems that means they have non-linear dynamics and typically grow or decay at an exponential rate with tipping point dynamics. Social networks like Facebook and WeChat have grown at such a rapid pace because of the network effect. Like dark networks protest movements are also informal functional networks occurring spontaneously without formal organization which can grow or decay rapidly. Many revolutions of the modern era have happened rapidly and unexpectedly including the French Revolution, the Russian Revolution of 1917 and the Iranian Revolution of 1978. These revolutions all took the people by surprise. In the example of the Iranian Revolution none of the major intelligence organizations like the CIA or the KGB expected the regime to collapse. Right up to the revolution they expected the incumbent leaders to weather out the movement. The Shah's fall came as a surprise even to the Ayatollah Khomeini, the cleric who from abroad in exile coordinated the revolutionary process that was to install him as the future leader of Iran. In a recent paper on the subject the authors illustrate the non-linear dynamics behind this process of change. Their model hinges on the observation that people who come to dislike their government are apt to hide their desire for change as long as the opposition seems weak. Because of this hidden preference a government that appears unshakable may see its support crumble following a slight surge in the opposition's apparent size caused by events insignificant in and of themselves. Unlikely though the revolution may have appeared in foresight, it will in hindsight appear inevitable because its occurrence exposes a panoply of previously hidden conflicts. Networks have thresholds, tipping points and feedback dynamics that make them subject to rapid and unpredictable change. However all political movements are temporary in nature, they only last a relatively brief period of time, either they die out or they become incorporated into the formal structures of the political system, in which case they will become more structured as non-linear dynamics give way to linear change. A network's resilience and functionality are closely related to the flow on that network, without any flow we do not have a network and no capacity for the system to operate as a whole and achieve certain ends. Our bodies function because there is a flow of blood and other resources around the whole system, likewise the community can form a functional unit when there is a flow of social capital through the interconnections within the group. As we integrate the system creating more connections and pathways for resources to circulate and potentially for the system to operate as a whole its capacity to realise work and resist change may go up, the inverse is also true, removing links reduces the flow and robustness of the system. Connectivity is one of the basic metrics of graph theory, it asks for the minimum number of elements, nodes or edges that need to be removed to disconnect the remaining nodes from each other. The connectivity of a graph is an important measure of its resilience as a network. A graph is connected when there is a path between every pair of vertices. In a connected graph there are no unreachable nodes. Connectivity reduces the path length on the network and makes people feel closer to each other and thus perceive that they form part of some common organisation. Resilience is one of the characteristics that is often associated with networks. We see time and again how small terrorist networks can persist, adapt and innovate in the face of an opposing security force that may be thousands of times their size. For example, the provisional IRA, a small terrorist network in the Republic of Ireland, managed to survive for 25 years while under continuous pressure from the British security services. There was a continuous interplay between them as the dark network of the IRA would conduct an attack. The British security services would reverse engineer them with the network then adapting. There would be many attacks conducted using a particular set of methods and technology before the security services figured it out and the British adapted to resist them before the IRA would adapt new strategies, again attacking with a new set of methods. Not all networks are resilient though, particularly centralised networks suffer the same vulnerabilities that hierarchies do. If the central hub is removed then this can drastically reduce the overall level of connectivity in the system. Centralised hubs are effective at rapidly connecting systems but the opposite is also true, they are single points of failure for rapidly disconnecting the system. When we talk about a brittle dictatorship like that of pre-war Iraq we can understand this as a network with a dominant centralised hub holding it all together with limited distributed networks connecting the whole of civil society. When the centralised hub is removed and without matured distributed networks of civil society the system can fall into the kind of chaos we saw in Iraq. This dynamic relates to how the network was formed and the distinction between connections made under rules of preferential attachment or the fitness model. With preferential attachment people connect to those who are already well connected thus irrespective of their merit those who are central will become more central making them over time too big to fail while at the same time those major nodes can be inefficient leading to a critical state in the system. A fitness model to the network development would though be room for others of merit to rise where the system may still be centralised and vulnerable but instead is dependent upon hubs that are more efficient with the rule structure that is better able to select for new more efficient nodes to rise. A key question for many is in how one intervenes to foster the development or disintegration of a social network. Policy makers may want to foster the development of innovation clusters by looking at the network structure or securities may want to disintegrate a criminal network. How one intervenes and tries to affect the development of the system changes in fundamental ways as organisations go from linear and well bounded to non-linear networked organisations whereas with traditional centralised hierarchical organisations we have one or few large nodes in the centre at the top who push out instructions to the rest of the organisation and it is seen that we just have to affect the central location to affect the mass of the organisation in the same way that we traditionally see the potential for change as lying with the president of a country or equally the way that mass marketing has tried to influence the mainstream of members within society by affecting the middle of the distribution. They do not look at how things are interconnected, they simply look at the characteristics of the individuals and aim for those characteristics that are most prevalent in the population thus trying to speak to the mass of people and in so doing hoping to affect the whole system. Whereas when dealing with linear systems of organisation you can only really affect the whole by affecting those large components near the centre of the mass. This changes though when dealing with networks where one can have much higher leverage by being strategic in one's intervention by analysing the network according to different metrics to look not only at what elements are most connected but also at other metrics such as between the centrality and how irreplaceable any given node in the network is and then affecting the network in a specific place. For example in attempting to break up a cannabis network the law enforcement agencies in Holland use data mining and network analysis to map out the dark network of the cannabis group and traders consisting of data on 30,000 members even though the real network was actually much larger than this. Traditionally the police had an approach of focusing on the kingpins in the network but they were finding that this approach did not break the whole network in fact it appeared to make it stronger. They combined hard data from criminal reports with soft data to map out the different functional roles within the cannabis network and various other attributes the members and how they were interconnected. It became apparent from this network model why the network did not fall apart when given the arrest of so many key actors each year. The police had thought that these networks were organised in a top-down fashion like the mafia but that turned out not to be the case. If it had been the best way to destroy it would have been through removing these centralised nodes however when the data was combined it was shown that the interactions were not top-down but had a different structure that was more functional and dynamic. The primary network had over time spawned other criminal and social networks. When a node was removed it was able to adapt and find replacements within those tangent organisations. The police decided to adopt a different approach looking at the specific roles that people played within the network to try and find those with functional roles that might be most difficult to replace. Over time they came to identify those doing the electrical work on the cannabis growing rooms to be scarce and difficult to replace. This model showed that every time an electrical worker was taken out the network's efficiency greatly dropped. The bosses were only coordinators and managers they were not so difficult to replace. The key was in looking at the specific roles and identifying those that were hard to replace. Those who could insulate rooms or install irrigation systems along with the electrical workers were the hardest roles to replace. When one of these were removed they were replaced by another freelancer but if the police stayed targeting those specialists the supply of such skills would run out and the network would fall apart over time. This is a good example of how in non-linear systems and complex systems the details matter. If you remove them through statistical aggregations then you remove precisely what it is that you need to succeed in your modeling or interventions and likewise this is why you need to use computational methods on real data. But we are of course not always trying to destroy sociopolitical networks just as often people wish to find methods to foster their development. Distributed networks cannot be controlled in a traditional sense through centralized control systems but they can be developed and modulated altering the costs of making connections in order to influence it to develop in a particular direction. This is the essence of managing complex systems. You cannot control all the parts but you can influence it based upon its evolutionary potential because power becomes diffused out within distributed networks. The whole system becomes greater than any of the parts. No one's in control or have the capacity to affect all the parts directly. The only option for affecting the whole system is to understand the current potential of the system and then create platforms or connections that facilitate the system to evolve in a particular direction that is possible given its past evolutionary history. Complex systems are path dependent meaning the set of possibilities available to the system now are contingent on the historical trajectory that brought it to this point and which conditions the current set of possibilities going forward. If you want to influence the system in the most effective manner you have to understand its history which tells you its current potential and then develop connections that facilitate to evolve within one of the current possibilities.