Analysing human societies as complex systems can provide an insight into historical processes and the strengths and weaknesses of capitalism.

Complex systems theory is a scientific theory coming out of a tradition of catastrophe theory, chaos theory, control theory, and especially cybernetics. Complex systems theory arises as a rejection of the traditional programme of reductionist science. It rejects reductionism as being both unnecessary and intractable.

Traditional physics has attempted to find fundamental laws at the smallest granularity possible. However, it is now known that because of the nature of interactions at very detailed granularity, it can be computationally intractable to predict behaviour of aggregate systems. Water, for instance, is best described (under most conditions) with fluid flow equations which describe aggregate behaviour with relatively simple (non-linear) equations. Attempting to describe it as an aggregate wave function of 1026 interacting quantum particles is not computationally feasible. Instead complex systems theory attempts to describe systems at a natural granularity that allows for tractable prediction of behaviour.

One of the fundamental notions in complex systems is that of emergent behaviour. That is, from a system with a large number of actors with simple rules, can emerge complex behaviour. This is an echo of the notion of a metasystem transition which was expressed by the cyberneticists1 . Some of the earliest descriptions of emergent behaviour actually come out of the Austrian school of economics and can be seen in the writings of Ludwig von Mises, where he describes the aggregate behaviour of capitalism as a type of optimisation which arises as a result of the self-interest of the actors2 .

There are a number of rules-of-thumb that systems theory gives us as tools of analysis for understanding actors and the emergence of aggregate actors3 . If you have a simple system interacting with a complex environment, it will have a low probability of maintaining its identity, that is, its internal structural integrity. Complex systems can interact with a complex environment in such a way that they have a higher probability of survival. We see that simple organisms in nature often have a strategy of massive reproduction because of the low likelihood of survival, which is in contrast to the method used by humans.

Complex systems theory and society

In order to understand how societies can be modelled by systems theory it is instructive to look at some simple examples. In feudal Europe the organisation of society was exceptionally hierarchical. This is modeled in systems theory by a sort of control graph, which is a tree, with the lord at the top and his immediate vassals below him. In this structure it was possible to approximate, in many circumstances, control over a group of people with control over the leader of the hierarchy. This has a large number of consequences. If the behaviour of the system can be modeled by behaviour of the lord, then the system can not act in ways more complex than the lord. Because of this, the system remains simple. It also means that the system can easily act coherently. It is capable of leading armies, and interacting with other feudal states in simple ways.

In reality no perfect control hierarchies exist. There will always be lateral control links, various types of conspiratorial actions etc. However, for feudalism this model often remains a good approximation.As we move through history to early capitalism we start seeing a move towards more “hybrid” models of control, where many more lateral links exist and the system takes on the possibility of evolving more decentralised, more complex behaviours. In addition, it becomes less brittle. One might conjecture that feudalism was in some sense doomed when capitalism arose because the environment of interaction became too complex.

The modern world has moved to a highly interconnected network-model capitalism. This is almost the antithesis of feudalism within the framework of the connectivity of the model.It is important to note a few things about the network model. Networks can have vary different internal structure. A large amount of interconnectedness does not rule out particular internal patterns, in fact we know that many complex systems, including social networks, don’t have “random” graph structures. This internal structure can have big effects on emergent behaviour. All networks are not the same.

In addition, the emergent behaviour of the system is *strongly* dependent on the interaction paradigm of the actors. The current economic system is a result of the paradigm of capitalist social relations. There is nothing “naturalistic” about the emergence of capitalism from these social relationships. It’s a bi-product of person to person social institutions.The atomisation of actors is arbitrary. It is actually often the case that systems can be re- atomised into a different notion of actor or communication. A good example of this is class politics. The analytical framework of the state, the bourgeoisie and the working class reifies entities and their interactions in ways that are easier to analyse then the mass interactions. This gives a mechanism for feasible reasoning about economics. This is in stark contrast to the obliteration of class dynamics that occurs in the intellectual framework of neoclassical economics.

Structure and behaviour

The aggregate behaviour of systems in terms of their control behaviour is something which can be very instructive to anarchists when thinking about how anarchism relates to the rest of the social environment. Idealised hierarchies can be modeled by their controlling entity. These aggregates are capable of what is known as “coherent” activity. Armies provide a good example of these types of systems. They can move in orchestrated blocks. However, the notional *objectives* which can be achieved with coherent systems *must* admit simple descriptions. Genocide, for instance, can be simply described and would be a description that one can expect an army to carry out. “Set up democracy in Iraq” however is something which an army has no capacity to do.

Democratic states and large corporations often fit more closely into the realm of the hybrid model hierarchy. These systems are starting to show system behaviour more complex then that of an individual actor. The behaviour that they engage in is becoming less strongly “coherent” and more “correlated”. You can’t expect things to move in lock step, but the system will move with a general correlated direction. You can also expect that some hybrids will be able to cope with a more complex environment than even a single actor might be able to cope with.

Finally in a networked system, where there is little or no notion of hierarchy, there is a possibility for truly complex emergent behaviour. Some types of systems which exhibit this are the human organism and social networks.

Current political structures

The state, being a fairly hierarchical creature, would like to make things function coherently. However its greatest weakness is its incapacity to find solutions to problem descriptions of large complexity. As an example militaries and states are finding it increasingly impossible to deal with the emergence of guerrilla warfare and terrorism. The models of organisation used in these social structures are highly decentralised and highly non-hierarchical. In the end, the state has little chance of eradicating such movements. The state must find simple descriptions of objectives and is at a fundamental disadvantage because of the more limited capacity to deal with complexity.Capitalism and corporate globalisation however are fearsome beasts. The internal model is highly networked. These creatures move across state boundaries with great agility. They have emergent behaviour which is ridiculously complex. Nobody even understands how the stock market functions (or dysfunctions). Capitalism is also incredibly robust, being able to adapt to circumstances in ways that Marx would have never thought possible. The only chance of combating an enemy of this type is with one of at least equal complexity.

Revolution and structure

The global behaviour and properties of complex systems are dependent on the interaction regime of the individuals. With human interactions there is a possibility of a feedback cycle that actually drives fundamental changes at the level of actor communication. This means that extreme changes in global behaviour are possible. Revolution is the radical modification of the emergent properties through changes in the interaction dynamics4 .

A lesson to take away from complex systems is that there is no one-single-correct model of societal interaction. Even if we knew in detail the interaction paradigm, it was a fixed parameter (social interactions of the atomic actors did not change) and the world was in fact strictly deterministic, this would not give us a social theory where we can predict outcomes. There is no positivist or naturalist method with which to proclaim the inevitable “march of history”. It is both computationally and methodologically irrelevant.

So what then can we take from complex systems theory in terms of application to our thinking on human society and revolutionary change? We know from the areas of empirical research in the natural science, and from historical information that radical paradigmatic changes are actually a very natural behaviour even though they are largely unpredictable. Revolutions in social order have occurred repeatedly throughout history.

There are several critical factors involved in the manifestation of genuinely new social orders. In terms of the generative events that create these changes, they happen by a process of increasing disorganisation or decay of the old order. This usually involves the injection of large amounts of energy into the old system. When these energetic events dissipate we have a solidification into a new order.

Symmetry-breaking is a common systems behaviour which is particularly interesting in the context of paradigmatic changes. Symmetry-breaking occurs when a system falls into one of two choices of lower probability based on small local deviations. The choice of which emergent behaviours are chosen canappear arbitrary to an external observer. In fact it starts from small internal fluctuations in the behaviour of the system.

Anarchist communism, when expressed in relation to complex systems, is a description of both the emergent and the individual behaviour. It requires behaviour at the global level with communism and mutual-aid among communities, as well as cooperative and collective decisions and solutions on the scale of the problems that face us. At the same time it asks for the removal of hierarchy and coercive power relations down to the level of individual actors.In the framework of complex systems, anarchist communism actually looks like it has very good chances for survival. It posits a non-hierarchical network model as a starting point for human organisation. This means it has a theoretical capacity to display complex emergent behaviour.

Additionally the role of the revolutionary organisation can be seen to be critical to the preservation of the libertarian quality of the revolution. Small fluctuations in a revolutionary situation can have disproportionately large impacts. The kernel of the new society will exist in the tendency of the organisations whose characters lend themselves to the movement. It is therefore critical that the organisationaltendency have the structural integrity and replicable knowledge of interaction dynamics required to crystallise the broader movement into one with a libertarian communist character.
Tactically, the use of complex systems thinking for analysis leaves as many questions as it answers. From this perspective many sorts of emergent behaviour will not be calculable a priori, but must be decided from empirical study or the weaker method of attempting to find appropriate historical analogies. We are left with complicated problems. We must find solutions in which tactical methods best enable escalation of class struggle. Additionally we must find the internal structures that are most scalable and replicable such that they can quickly be effected during the heightened period of sensitivity that occurs in the revolutionary moment.

Glossary
Intractable - Problems that can be solved in theory, but not fast enough for the solution to be usable on human time scales.
Reductionism - An approach to understanding the nature of complex things by reducing them to the interactions of their parts, or to simpler, more fundamental things
Neo-Classical economics - A general approach to economics focusing on the determination of prices, outputs, and income distributions in markets through supply and demand.

• 1The Phenomenon of Science. Valentin Turchin
• 2The Ultimate Foundation of Economic Sciences. Ludwig Von Mises
• 3Complexity Rising: From Human Beings to Human Civilization, a Complexity Profile. Yaneer Bar-Yam
• 4Phase Shifts or Flip-flops in Complex Systems. Henry A Regier and James J Kay

Over the last few years, the Workers Solidarity Movement, the anarchist organisation that publishes this magazine, has grown considerably. We went from being an organisation with only a dozen members or so, to an organisation six times that size. As part of that growth we have had to reassess our internal workings and devise a range of new processes and structures for communicating, coordinating and democratic decision making. This article describes this process of change. It is hoped that it may serve as a useful case-study for other groups facing similar problems and as a small demonstration of the how anarchist organisational principles can be applied in practice.

Up until about 2002, the WSM never really consisted of more than a single branch in Dublin of about a dozen people and a couple of members in Cork, with a few sympathisers scattered around the country. As an anarchist organisation, even when we were only a handful, we were careful to ensure that we had structures and processes in place which would allow our members to have a full democratic input into our policies and activities. Our constitution specifies that our twice yearly conference, open to all members, was our supreme decision making forum. These conferences provide an opportunity for any member to propose a change to any of our policies. Being an organisation which strives to put our money where our mouth is, our actions are guided by our policies. Thus, our conferences serve as democratic decision making forums which enable our members to direct the organisation’s activity.

Even when we numbered less than a dozen, our conferences were highly formal affairs. In order to create new policies, or change existing ones, members had to submit motions in writing, weeks in advance. These motions, often accompanied by articles arguing their merits, were collated into an internal bulletin which was circulated to all members by post. Members then had a chance to submit written amendments to motions. The conference itself was devoted to debate and voting on motions. Tremendous care was taken to make sure that all points of view were heard, and the strictest democratic principles were followed, including providing private balloting and proxy voting.

When we numbered a dozen, the elaborate care that we devoted to internal democracy made our conferences occasionally tortuous. Even when less than ten members attended they could take two entire days, with attendees increasingly irritable as time passed and procedural debates came to the fore. Yet, despite the procedural frustrations, the conferences proved productive. The WSM developed a set of detailed policy documents, continually debated and amended over the years, which distilled the organisation’s wisdom and experience into guides for future action.

The coherence that these policies gave the organisation proved invaluable when the anti-globalisation and anti-war movements emerged, bringing with it a relatively large number of people who were sympathetic or at least open to anarchist ideas. Through a lot of hard work within broad and relatively informal anti-authoritarian groups such as Reclaim The Streets and the Grassroots Networks, we gained respect among activists, a higher profile for the organisation, and a wealth of practical experience of the problems associated with organising among larger groups with dispersed membership. Eventually, many of the anarchist activists who had worked alongside us inthese groups came to join the organisation. Many of those who worked within looser, less formal groups and campaigns, repeatedly ran into organisational problems concerning communications and decision making. The experience of working alongside the WSM gave them an appreciation of how useful formal structures and a capacity for coherent action can be – and that viable anarchist models can be built.

Thus, in the last five or six years, our organisation has experienced a steady influx of new members. Although our current membership of roughly 70 is hardly going to send the ruling class scuttling to their bunkers in fear of revolution, over the last few years our rate of growth has been such that we have doubled in size every two years or so. This steady growth has required us to continually re- examine our processes for communication and decision making across the organisation.

Happily, the formal processes which underpin our conferences have coped admirably with our expansion. What used to seem somewhat constricting and excessively formal, now seems to be an eminently sensible and valuable system which allows all of our members to have a genuine opportunity to change the organisation’s policy. Indeed, as we have grown, our conferences have actually become more efficient and inclusive decision making bodies. For example, at our conference in November 2007, some 30 motions, proposing some 80 amendments to our policy documents were proposed, debated and mostly passed, including the replacement of our entire position paper on the environment. Most of these changes were put forward by members who had been in the organisation for less than 2 years. To a considerable extent our conferences work better know than at any time in the past – the formal structures that we put in place have come into their own now that they are obviously needed. We have not had to change our conference structures at all in order to cope with larger numbers, simply formalise some of our standing orders. The success of our conference structures is largely due to the fact that they were originally borrowed from the democratic structures developed by anarcho- syndicalist trade unions, which are retained by modern trade-unions, albeit as a poor shadow of their former selves.Our conferences, as they stand, provide an excellent example of direct democracy in practice and they help to show new members exactly what our politics mean in practice. However, if we continue our recent rate of growth, within the next few years we will face new challenges as it becomes more difficult to fit everybody into a room together and the time available for debate and voting diminishes. Happily, however, we should be well prepared for this eventuality. By the start of the twentieth century, anarcho-syndicalist trade unions had already developed conference structures based on tightly-mandated delegates, which allowed vast numbers of members to have a direct say in policy changes. We therefore have a wealth of models available to us which should enable us to continue making directly democratic decisions at our conferences for the foreseeable future. However, delegate-based conferences are much less useful forums for debate – if the delegates are already mandated to vote in a certain way, they can’t be persuaded by argument, it is only if they have an open mandate that there is any point in hearing arguments. We already use the Internet to circulate debate pieces and arguments for and against proposed amendments in advance of conferences, and as our conferences become more delegate based, these debates and arguments will need to take on a greater role, and may need to be given more formal structures.

While our conferences have proved well suited to coping with a larger organisation, the rest of our structures were much less well prepared. Conferences basically define the organisation’s core policies. There is, however, never a formulaic way to translate general policies into concrete actions. That is to say, that no matter how thoroughly our policies describe our political positions, we constantly face decisions, sometimes requiring subtle judgement calls, as to how we apply them in practice. As a group of political activists, we face important tactical decisions all the time – which campaigns to join, which political groups to cooperate with, which demonstrations to go to, what leaflets to print and so on. We also face a whole host of day to day operational decisions – who will lay out our publications, write articles and leaflets, attend meetings, carry our banner on marches and so on. As we grew, we discovered that our structures were poorly suited to these types of decisions.

Once a group grows beyond a couple of dozen members, particularly when those members live in different regions, towns and neighbourhoods, it needs to change in character. In small groups with relatively steady membership over years, everybody can know everybody else reasonably well and much day to day decision making and the sharing out of responsibilities can rely upon informal communication channels. When you have over 50 members, this becomes impossible and informal arrangements become barriers to new members integrating into the organisation and can lead to the emergence of an informal leadership based upon knowledge of the various informal mechanisms that keep the organisation ticking over. Therefore, as we grew, we had to develop new, formal structures to coordinate our day to day decision making and communication and to ensure that the organisation worked in a transparent way where everybody had an equal chance to have an input into tactical and operational decisions.

The problem was not that we had not thought about such problems in advance – just that the measures for coordinating day to day activity that we had defined in our constitution weren’t capable of addressing all of our requirements. The constitution defined a National Committee, which had to meet at least once between each conference. However, due to our small size, the national committee was open to all members and due to the fact that we only had a single branch, it effectively amounted to our Dublin branch with an extra member from Cork in attendance. In these circumstances there was little need for the national committee to meet since it was much easier for the Dublin group to make most of the tactical decisions, occasionally telephoning our comrades in Cork for their opinions when needed. The constitution also defined Commissions, bodies designed to coordinate the organisation’s work in particular areas. However, once again, these commissions never really got off the ground. We had a single branch which met regularly and commissions would simply have amounted to a subset of these people having an extra meeting – extra work without any real advantage.Thus, when we started to experience sustained growth in the period since 2003, our structures for day to day coordination were rather theoretical and untested in nature. One of the first real practical problems that we encountered was how to divide up the organisation. This became pressing in late 2004, when our Dublin branch became too large to fit in our office’s meeting room. In addition to the cramped nature of meetings, with up to 20 people attending, we found debates and discussions becoming increasingly lengthy, cranky and frustrating – there were too many people present to give everybody a chance to express their opinions and anarchists are, if anything, full of opinions!

Therefore, we had to find a way to split our Dublin branch up. We had no formal mechanism for doing so, so we simply gathered all of our Dublin members together in a general meeting and, based on the arguments and proposals that were raised, we eventually divided ourselves up into groups based on two factors:
1. Where we lived.
2. When we could attend meetings.

It was generally envisaged that continued and sustained growth would require our branches to be increasingly tied to local community activism. Thus, the location of each member’s residence was considered to be important in allocating them to branches. However, the problem is that in modern cities, with their wide socio-economic spread, community identities tend towards the hyper- local. In practice, the various local issues that interest people can vary widely from estate to estate and neighbourhood to neighbourhood. Therefore, in order to have a real impact on a community, an activist group needs to have a lot of members living in the same neighbourhood. Furthermore, community activism is, by its nature, slow-burning, requiring steady work over years to have a real impact. As many of our members are relatively young and transient, moving around the city frequently and living in rented accommodation, they were not in a position to really implement a community- based strategy.
A year after our Dublin branch’s initial sub- division, we yet again found that our branches had grown too big. Again, we got together and re-divided ourselves, this time into three branches based on where each person lived and when they could make meetings. Once, again, despite our efforts, this did not work as well as we would have hoped and this is still a live issue in the organisation. Of our 3 Dublin branches, only one is mostly formed of people who are long term residents in the same area.By early 2005, we had 3 branches, 2 in Dublin, while our Cork membership had also grown to the point where it formed a stable branch. A year later this had grown to four, with the creation of a third Dublin branch. In 2008, a new Belfast branch was launched . As the number of branches grew, the problem of day to day coordination between them increased. As an organisation which prioritises theoretical and tactical unity, we had to come up with structures which allowed us to take decisions which would guide the whole organisation, even in cases where there were significant disagreements about the best course of action.

The way that most organisations deal with this problem, even those that are based around a democratic, policy-setting national conference, is to appoint an executive committee, responsible for implementing the organisation’s policy. Such arrangements, however, invariably degenerate over time. Even when everybody acts with the best of intentions, the fact that democratic decisions take longer than the decisions of an executive officer means that the officers tend to become more powerful over time. Alongside this tendency, the relatively privileged position that executive officers occupy, allows them great influence over the formation of policy. Before too long the conference has become a talking shop with policy being effectively controlled by a small leadership. One only needs to look at the conferences of the labour party, or many of the trade unions to see stark examples of this process in action. Furthermore, by limiting day to day decision making power to a tiny number of people, the organisation vastly under-utilises its collective intelligence. For all these reasons and more, an executive was not the solution that we wanted.We instead established a Delegate Council which would be responsible for coordination and decision making across the organisation. This body was established at our conference in Autumn 2006. Its structure has been refined twice in the period since then, after some inadequacies in its initial specification became clear. It meets once a month, with a different branch hosting each meeting – meaning that its location is rotated. Any member or branch can submit a motion to the council – asking the organisation to commit itself to a certain course of action. All the motions are circulated through our website and branches, sufficiently in advance so that each branch has a chance to discuss them at a meeting. Each branch then selects a number of delegates – one for every five members and these delegates attend the council.

Initially, our Delegate Council meetings were not particularly successful. Each motion would have been discussed and voted upon at eachbranch and the delegates would merely report how many votes each motion had received, for and against, from their branch. Although it allowed us to make decisions, the motions had already been voted upon and the delegates had no mandates to do anything but report those decisions. The meetings were really limited to a bit of informal discussion and the tallying of votes – something that could have been done by email or over the phone. The real benefit of face-to-face meetings is that they make it much easier for groups to make compromises. Written motions are often formulated in such a way that they do not include the proposer’s arguments and these may be then misunderstood by others who only see the wording of the formal motion. This can, for example, lead to situations where a motion is voted down, yet no solution to the problem that it was addressing is put in its place.

A face-to-face meeting gives each delegate the chance to explain the thinking behind their motions and any concerns their branches may have about other motions and to try to see if compromises can be reached which accommodate everybody’s desires. Thus, as these problems became apparent, although motions were still voted on in advance, delegates were often provided with mandates to seek compromises or amendments to motions. Since each delegate is only representing five members and has to report back to their branch, there is very limited space for these looser mandates to result in abuses. At our Autumn conference in 2008, we formalised this in a vote at our national conference.Although the Delegate Council has proved to be reasonably effective in coming up with tactical decisions, since it meets only once a month, there are occasionally situation where we have to make decisions to a tighter deadline. To enable this our constitution allows any branch, or group of members, to call an emergency Delegate Council meeting. This has been used on a couple of occasions when we faced important decisions and time was felt to be of the essence.

However, there are also frequently operational questions which need to be made at short notice – should we send a speaker to aparticular meeting, should we bring a banner to a demonstration called at short notice? In such cases, it is not realistic to convene a national meeting to make the decision. Therefore, at our most recent conference in November 2007, we instituted an Interim Decisions Committee, a body made up of 3 members, including the national secretary, which has the authority to make operational decisions at short notice. Its power is subordinate to Delegate Council, which is itself subordinate to national conference.

In it’s first year and a bit of operation, the IDC has been called on to make only a handful of decisions. Initially these were confined to routine matters - endorsing a demonstration that was in line with organisational policy and precedent, and turning down a request that was deemed outside of its mandate. However, in recent months, the IDC has been called on to make decisions on a couple of occasions over matters of some controversy. These decisions were disagreed with by some, but they were implemented without any problems. Nevertheless, the debates surrounding these decisions have revealed that there remains some unease throughout the organisation about this committee, both within the committee itself and across the rest of the organisation. The IDC remains very much a work in progress and it remains possible that it will be replaced with another mechanism which allows broader input.In summary, over the last few years the WSM has put in place a range of structures in order to allow the organisation to take common decisions. These structures are intended to maximise democracy, while still allowing snap decisions to be made in emergencies. The WSM’s steady growth over the last few years prompted these changes, and the shape of these structures is in constant evolution as new problems are encountered and overcome. Although the WSM remains a very small organisation, this example does show the viability of anarchist direct democratic decision making principles when applied to organisations that are geographically dispersed and have enough members so that not everybody knows everybody else.