Making anarchist organisations work - Dunbar’s number, administration and care

There is a strong tendency, almost a rule, that anarchist groups tend to fall apart once they have more than 20-30 members in any city or 50 to 60 overall. Or at the very least an organisation that once felt like it worked very well becomes one that feels sluggish and starts requiring too much effort to achieve limited results in the longer term. There are exceptions which mean this is not inevitable but why does this happen and more importantly how can we avoid it in our organising?

The cause may be simply a limitation of our brains and in particular the number of complex inter relationships between people we can track. Or, more correctly, a failure to acknowledge that this limit means that informality will fail and formal administration is more and more necessary as group size rises.  A lesson that is not just relevent to anarchist but to all attempts at horizontal organisation. 


(If you arrived here from a search for Dunbar's number
and know little or nothing about anarchist organisation
you might want to read
Are Anarchists in Favour of Organisation

For a group to work well at an informal level everyone needs to get on and to understand how they relate to everyone else in the group and how those people relate to each other.  This is easy enough for 2 or 3 people but impossibly difficult for 200.  The complexity rises much faster than the number of people in the group as I illustrate below (or skip this section but watch the video at the end of it).

If there are two people A and B then each only needs to track one relationship, the one between them.

When a third person C arrives then A not only has a relationship with B to understand and a relationship with C to understand but also needs to understands the relationship between C and B and how BC as a group rather than two individuals relates to them. As anyone who has shared a house will know a relationship with a couple is going to have complexities beyond the sum of the relationships with both individuals.

Aidan, one of the pre publication readers of this piece, told me this section reminded him of some TV show I saw years ago about psychology that talked about Shakespeare’s Othello, and how at a certain point in writing Shakespeare needed to understand the audience's perception of Iago's perception of Othello's perception of Iago’'s perception of Desdemonda's perception of Othello.  That’s a good illustration of how complex thing can become even when you are only dealing with the audience member and 3 characters and of course the tragic results of getting it wrong.

The arrival of a 4th person D adds to the complexity as we now have
and also
A with BC
A with BD
B with CD
A with BCD

and also
A with BC
A with BD
A with BE
B with CD
B with DE
C with DE
A with BCD
A with BCE
B with CDE
A with BCDE

And I may have missed a couple with that last one, but tracking relationships between 5 people has already become hard enough to simply list the possibilities. The number of relationships to track increases much much faster than the number of people involved. Pretty quickly you get to very big numbers indeed. It’s an exponential relationship where according to Reed’s Law if the number of participants are N the number of subgroups is 2^N − N − 1. For a group of 20 that is already 1,048,555 relationships to track.

There is a theory

What’s above is a crude illustrative sketch but it turns out there has been scientific study of this complexity. In particular a British anthropologist and evolutionary psychologist who is a specialist in primate behaviour called Robin Dunbar has written a good deal about this. As often happens with science the number he suggests is the maximum size for a group has come to be referred to as Dunbar’s number.

A lot of recent interest in Dunbar’s number comes about because of social networking. While I’ve written in the past about the benefits for organising of the Network effect where as the network becomes larger the benefit of network becomes larger still, Dunbar’s number represents the cost side of that benefit. The ‘transaction cost’ of communication rises with the number of people.

The odds are the number of Facebook friends you have is some multiple of Dunbar’s number unless you are one of those people that do an annual cull. In which case it might well come close to defining the size of your Dunbar group. Otherwise perhaps the whole reason we spend so much time staring at Facebook is in part down to the demands placed by trying to maintain such large Dunbar groups. But that is for another discussion.

Dunbar's number is defined as the suggested maximum cognitive limit to the number of people we can maintain stable social relationships with. That is relationships in which we know not only who each person is but also how each person in a group relates to every other person in that group. The idea that there is a maximum limit is an important one for anarchist organising because it marks the line where informal organisation will certainly fail and so where we will need formal organisations or the much dreaded ‘bureaucracy’.

I’m generally wary of using concepts from evolutionary psychology because they tend towards biological determinism and from there to the acceptance of soft racism, sexism etc as part of some natural order. But here we are on safe enough ground as Robin Dunbar calculated that the maximum group size for humans would be 148. That some of us live in cities of 10 million plus and nations over 300 million shows that humans have long ago found ways to overcome the limitations of our brain size in building stable social relationships. If we hadn’t you wouldn’t be reading this on a technology that required the cooperation of tens of millions of people to function.

Dunbar actually didn’t take group-to-one relationships into account so his complexity grows considerable more slowly that the sketch I opened with. But I’m not so much interested in the precise details as the idea that the number of people with whom we can have strong relationships is limited and it’s very very much harder to maintain strong relationships as group size increases. If that limit is 50 or 150 or 300 isn’t the important thing, just that there is a limit and in terms of the lives we live it’s actually quite small.

Dunbar calculated that limit at 148 (normally rounded to 150 for discussion purposes) but argued that this is very much an upper limit. Primates maintain relationships through ‘social grooming’, in human society this often revolves around eating together, drinking together and other social rituals. All that takes up a lot of time. Dunbar calculated that to actually maintain a group size of 150 would require a lot of time devoted to such grooming, in fact about 42% of group time. So numbers that large are only found when there was intense survival pressure for people to stick together in large groups.

Dunbar suggested that we work best in small groups of 5 and that for each time you multiply that group size by 3 you reduce the strength of the relationships between people.

We come into contact with a lot of people

At the level of the individual we can see that in most circumstances in modern society we are in contact with very much more than 150 people. When you catch a plane at an airport you are seeing far more people in that space of time than any human would have seen in their entire lifetime for almost all the time humans have been on the planet. It’s only in the last 15,000 years that we have started to congregate in larger numbers although before that date there may have been occasional super gatherings of multiple very much smaller groups - a bit like a music festival.

I’m going to use the concept of a ‘Dunbar group’ in most of the rest of this piece to suggest the group of people we construct with whom we seek to maintain stable social relationships. That is the group of people where we maintain “relationships in which we know who each person is but also how each person in a group relates to every other person.” Of course we probably don’t think of it that way, rather we have those lists of people we might ask to a party we were holding or that we go for a pint after work with or, importantly for this essay, whom we try to do intense ongoing political activity with. We need to know not only how they get on with us but also how they get on with each other.

Most of us don’t consciously add and subtract people from that group but if we look back over a few years its easy to identify people we have in fact added and others we have mostly lost contact with. Dunbar’s number was based on the relations in a single group where everyone knew each other, that’s an assumption that seldom exists in modern life where instead each of us has relationships with multiple groups but those groups may have little or a lot of contact with each other. The point again being not to get hung up on the exact number but to understand that limits exist.

In physical space city dwellers are continually not putting people into their individual Dunbar group. I stop off on my way to work at a Lidl every Monday morning, there is only ever one checkout in operation and 90% of the time over the last two years the same woman is working it.

We say “Hi” in recognition but I know nothing about her at all, I’ve certainly no idea how she relates to the hundreds or thousands of other customers who must come in regularly and I’ve no relationship with any of them. If management told her to build a personal relationship with her customers that would be impossible. That we know nothing about each other is not because either of us is a bad, uncaring person but because it’s impossible to communicate at that level with everyone most people meet in their day to day lives in a city.

Politicians employ people whose sole job it is to track who they are going to be meeting and remind them of their name and some basic facts about them. I’ve talked to union organisers who try to retain one key fact about each member they are trying to organise, they try to identify and remember what that person's primary issue is. Try, in other words, to create a false sense of having that person in your Dunbar group because they are then more likely to trust you.

Finally this tracking ability is a talent that some people are probably going to be better at than others. It’s not unusual for the sort of people who are attracted to small political groups to be quite introspective. In terms of organisation that can shape the way the organisation appears to outsiders and how easy it is to engage with.

The state

There are movements related to anarchism that long for simpler, small scale societies where everyone knows everyone else. Primitivists for instance. They want to live in groups smaller than 150 because they want the simplified social relations that come with small groups.

The idea of spending my entire life in a small community horrifies me. I’ve visited enough small islands and isolated villages in my time to be aware of the unpleasant side of living in a society where everyone knows every aspect of everyone else’s business. Very often group cohesion is maintained by rigorously policing the group’s defined boundaries against outside influences.

On the other hand there is mainstream liberalism and its understanding of the state. This insists that because of something like Dunbar’s number civilisation needs strong leaders and rigid rules at its centre lest it fall apart in chaos. It’s likely that the state as we know it came into being at the point where agricultural surplus and storage meant that in terms of production it became possible to exceed Dunbar’s number on a permanent basis.

At first that was done through the application of force to make most of the population’s wishes irrelevant to those who ruled. The only strong relationship that was needed to maintain coherence in society was the terror the ruled felt for the ruler. A top down structure of power where the King had a ‘Dunbar group’ of trusted generals and princes and each of those in turn a Dunbar group of captains and lords. All the way down to the village strongman who ruled over the serfs. That is basically the structure of rule of a lot of old empires and, as many commentators have pointed out, Dunbar sized groups are often as the size of military formations.

For a fair chunk of human history a ruling elite built a mass society to serve them through brute force along those lines. The awful methods of executions used by the Romans to maintain slavery or the German Princes to crush the Anabaptists were a way of imposing coherence on those mass societies. There is nothing like witnessing a crucifixion or the physical dismemberment of a living conscious person with hot tongs, muscle by muscle, over hours to instruct the masses on the danger of stepping out of line.

But again we escaped biological determinism or the insistence there once was on an iron law that says mass societies must just be based on brutality. Over the last 300 years the naked brute force societies of slavery, feudalism and early capitalism have been replaced by representative democracies where the appliance of naked force is a secondary resort. In some cases, notably the USA, the application of brute force, including execution, is still fairly routine in order to keep the population in check. This tends to be a requirement of societies where there is a huge division between the rich and the poor.

In others like Norway and even Ireland the application of state force against the population is exceptional, the building of an apparent consensus is far more fundamental to maintaining stability. The club comes out when it is needed but most of the time its shadow is sufficient.

There have been large scale experiments in systems of direct democracy that have exceeded Dunbar’s number by several magnitudes, the Zapatista rebellion in Chiapas which is now over 20 years visible being a well known example. Rural ommunities there typically number 4-700. So we have already proved that you can exceed the Dunbar limitation in a mass society without recourse to brutality, that is of course essential for the anarchist project. If it wasn’t possible than anarchism would be a utopian dream.

The anarchist organisation

When I first became an anarchist the first organisations I was involved in were tiny. The Anarcho-Communist Group was five students in the same college engaged in a militant period of students struggle. Four of the five of us also travelled to London together to squat and organise around the Poll Tax. We were from Ireland and we also took part in various anti-imperialist solidarity demonstrations demanding the withdrawal of British troops from Northern Ireland. We didn’t have much in the way of formal ways of maintaining group stability but then we were living in constant close proximity and facing hostile external enemies so cohesion was easy to maintain.  We were also at the group size that Dunbar suggests would have the strongest relationships.

I next became a member of the Workers Solidarity Movement. But again this was a very small group, 5 in Dublin and one in Cork that initally grew to 10-15 after we joined.  Which is more or less the size of the second strongest releationship group Dunbar suggests.. We spent a lot of time socialising together, in Ireland ‘social grooming’ operates around pubs and parties. For the first time I was part of an anarchist group that did have formal collective methods like minute taking and chairing but we were a close knit bunch so these were add ons rather than requirements.

Over the following 20 years the WSM grew, peaking at perhaps 65 formal members so entering the weakest relationship group Dunbar suggests. It hit problems that at the time were hard to understand. For instance members would have fallings out that seemed to be based about misunderstandings. Or some members would develop very strong grudges against others, often based around the idea that the other person was not pulling their weight. Sometimes I knew there was little accuracy to those feelings, I remember in particular being at the receiving end of a heated drunken rant about the supposed laziness of one member in circumstances where I knew the actual problem was the ranter was simply unaware of the activity the person they were ranting about was involved with. But that sort of stuff bred demoralisation on the one hand and undermined commitment on the other, over time it proved very corrosive indeed.

Retrospectively I think the problem was that we had reached a size where people were exceeding a Dunbar’s number limitation. Being in an active revolutionary organisation subject to times of considerable pressure requires a lot of collective feeling and trust and the social grooming of pints after a meeting in the pub is only sufficient for quite small numbers.

People have families, flatmates, friends and colleagues who must absorb a lot of their Dunbar’s number. By the time we reached 30-50 members we had crossed that line and trust between everyone in the group was probably impossible to build informally. Dunbar has pointed out that 50 is “the typical overnight camp size among traditional hunter-gatherers like the Australian Aboriginals or the San Bushmen of southern Africa” which isn’t far off the level I think I started to notice problems.

A diversion on Care

The other aspect is how Dunbar’s number relates to care. We expect people around us to care for us, to be aware of our lives, to spot problems and intervene without being asked. Actually it’s not really people we expect this of, it’s mostly women. In our society this expectation tends to be very gendered, that is there is an expectation that women will do the work of care. Likewise administration is gendered.

I often heard people express at meetings a frustration with what I would now call the lack of care they were receiving. This was actually expressed in many ways, but most often as a frustration with some failing where the solution should really have been collective and thus the complainer should have been creating the solution rather than bemoaning the lack of it. In other words a sense that someone should look after them but without a willingness to collectively construct that someone.

I was national secretary a couple of times in this period and in recently comparing experiences with a woman who filled that role at other times we realised that while she had experiences of members who expressed feeling let down with her because she had failed to care for them I didn’t have such experiences. And again because these expectations are gendered the reality is she was probably quite good at this whereas I would mostly not have even conceived of an expectation that care was part of my role.

These problems were part of the reason for organisational crisis that saw us lose many members, those who departed and stated reasons for doing so very often expressed a sense of being let down by the organisation. But as members the same people had generally not been people who prioritised work around creating organisational solutions to problems of communication and care. Instead communications solutions tended to be rejected as ‘organisational solutions to political problems’. Worse still care was rejected because people who found conflict emotionally exhausting were seen as not really suitable for political activism as things would be much tougher in our external work.

Now as an organisation goes into a crisis of demoralisation the need for ‘social grooming’ tends to increase while the willingness to put time into it decreases. Or as I experienced things people became far more likely to misunderstand each other, fight in an unproductive way as a result of these misunderstandings, and maintain grudges more intensely and for longer periods afterwards. On the way up the WSM had felt like a warm, close knit community with a lot of informal mutual support, on the way down very much less so. A surprising amount of ex members identified ‘the problem’ as being with particular people, the more self aware ones at least made that a particular type of person rather than named individuals.

If we accept that Dunbar’s number tops out at 150 and we think that on average someone might have 120 people in their social network than as WSM approached 30 lots of individual members were either having to drop relationships with people outside of WSM or not able to put enough effort into maintaining connections within WSM to avoid such misunderstanding and overcome friction.

As I recall what happened is that people started to put their effort into maintaining relationships with those most like them culturally or politically. Informal sub networks emerged of for example those very involved in Rossport who would hang out together after conferences, and ring each other for a chat quite separately from WSM business. And as with all group formation there was a tendency for those spontaneous groups to also start to define themselves against the members who didn’t fit into them. At least at times this also meant that members who didn’t easily fit into any group felt excluded from the whole, and of course this most affected those coming from already marginalised groups such as migrants.

We were aware of the weakening internal social bonds although we didn’t consider that perhaps the real issue was that we were too many to rely on tight bonds as an organisational solution until it was too late. On one level we had some grasp of the problem as we did do things like have a joint meeting of all three Dublin branches once a month, try to organise social nights and, later on, once a month have a meal after a branch meeting to which we would also ask contacts. None of these really worked well as they tended to only attract people who were already in the core group that did or had socialised together. Retrospectively that was really what we should have expected because in effect we were demanding that members fit additional WSMers into their Dunbar group which as a solution would have required excluding family, friends etc to make room.

Escaping Dunbar

For anarchists to win we need to overcome the limitations of Dunbar's number. The importance of the concept is understanding that this cannot be a matter of becoming better, more caring individuals who spend more time looking after each other in genuine ways. If you go ‘full cult’ and expect members to cut connections with everyone else you might get to 150 that way. But we need to organise tens of millions as a transition to six plus billion people becoming self organised. Being better people won’t do that.

How is this done? That’s not so hard to answer in a general sense as we already live in a mass society so we can see how this happens.

1. Ideology

Mass societies develop collective understandings a of why they are and where they are going which can be summed up under the term ideology. Generally the most successful form will be seen as ‘common sense’ or ‘the way things are’ but it’s ideology all the same. They are developed from the top down, in the interests of those who have wealth and power. But there will be other competing ideas of how things could be, anarchism being one of those.

In the USA today for instance the ‘American dream,’ now expressed through the ideology of neo-liberalism, is a common point of departure for many. If you work hard you can get ahead is something strongly believed by many people who have worked hard and are stuck where they are.

Religion is essentially ideology and for long periods of European history the ideological cement in societies was that ‘we’ are Protestant while ‘they’ over the river are Catholic and vice versa. Ideology was what held the mass communist parties of the 20th century together as Russia sent tanks into Hungary in 1956 to put down a workers rebellion. It’s a powerful glue for holding mass organisations and indeed whole societies together.

So those who argue that organisational problems must have political solutions are not entirely wrong. The construction of a better ideological glue can overcome communication problems, at least for a while.

But using ideology as your primary glue to hold people together comes with costs, chief amongst those is the promotion of an unchanging conservatism, since to enter into free discussion softens the glue. And while political solutions might alleviate the symptoms they may not in fact address the cause if the cause is indeed insufficient administration and communications.

But, even or perhaps especially for anarchists ideology is important. After all it’s our common understanding of what is wrong with the world and what needs to be done to put it right that separates us out from the mass of society and caused us to seek each other out in the first place.

Ideology also has the major advantage of being easy to reproduce faithfully and transmit over both time and space. It scales very easily and doesn’t necessarily require a large group to produce. Indeed in the networked age of social communication one person can create and transmit ideology as never before.

2. Common transparent membership expectations

The huge advantage a Dunbar group has is a very deep organic common trust based on a deep understanding of where each other is at. In the smallest groups on the Dunbar scale of 3-5 people the work everyone does is inevitably very visible.  Loss of that makes organising very much harder indeed. But at groups above 50 or so it’s impossible to have that trust organically generated and maintained. To deal with the problems this causes it’s important that expectations be formally stated and visible rather than assumed as they are in informal small groups.

There should be clear expectation of what members are expected to do. How much of their income should they be donating to financing the organisation? How much work are they expected to put into its work? If these things are clearly defined and visibly stuck to the resentment over relative workloads shouldn’t be a problem providing members who are not filling those collectively agreed expectations are removed from membership on a regular basis.

There is a ‘yes but’ here which is what if there is a good reason why someone isn’t pulling their weight. The only workable answer is that membership is defined by meeting the collectively agreed expectations and if someone over time isn’t doing that they are no longer a member.

Another answer would be to ignore the situations of members not doing what they have committed to. The problem with that it is quickly corrosive in a couple of ways. Firstly it creates hostility, distrust and suspicion, in particular where differences of political perspective exist. Why should someone have a vote on what sort of work should be done for instance when you never seem them actually doing any work for the organisation. Why should a meeting have to spend time going back over an issue for someone who seldom bothers to come to meetings. And secondly resentment will mean other members in turn will be much less motivated to carry through what they have committed to.

The third answer, the popular but I believe unworkable one is that a judgement is made whether it’s ok that someone isn’t meeting the collective membership commitments. That seems wonderfully fair until you asks the question of who is to make this judgement and on what grounds will they make it?

The most useful approach is to treat membership not as badge of honour but a combination of both political agreement and carrying out the requirements of membership. When someone stops carrying out those requirements then they are no longer a member. This isn’t new. In the past organisations like the IWW have made that process transparent by requiring members to have a card on which stamps were placed when they filled membership requirements and the display of the card was a requirement to vote.

Aileen  commenting on a draft of this article noted “I like to take a slightly different tack and say that the organisation has to have agreement on what constitutes "work". This agreement is often absent. Another anarchist women used the phrase "social reproduction of the organisation" to describe the vital admin work that she does for an anarchist group elsewhere. It's a useful turn of phrase because it mirrors the "reproduction of labour" that (most often) women do under capitalism. Mirrors it because like domestic labour admin work is often invisible and not valued as "work", in a gendered way. I think many revolutionaries I have met do something similar in terms of their understanding of what "counts" as revolutionary work - going to a picket, attending a meeting, giving a talk, writing an article - all these are undoubtedly seen as real work (they are also in the main collective and public). Spending a weekend doing layout, graphic design, ringing people to arrange a meeting, fixing a database or the website - these types of work are invisible and don't count. So there needs to be an understanding on what constitutes "work", and understanding that values all tasks equally and doesn't have a "hierarchy of importance".

3. Administration

Many anarchists tend to avoid administrative work. Not surprising as it is routine, often boring and there is little glory attached to it. No one composes songs about the anarchists who must have ran the administrative systems that got food and ammunition to the front during the Spanish revolution. As with care, administration is often very gendered, the burden falling mostly on women members, even when they are in a small minority. The boys preferring to write the articles, give the talks and be the visible external organisers.

Administration is often an area anarchists refuse to seriously discuss, any serious discussion tending to be dismissed as ‘an organisational solution to a political problem’ even in circumstances where it’s rather obvious that the problem is organisational. Instead of ‘bureaucratic’ fixes we are told organisational matters can be dealt with by boosting moral (see care), political agreement (see ideology) etc.

Those fixes work in small groups (i.e. where everyone has everyone else in the group inside their Dunbar limit). But the don’t work in large ones. Which is why an entire branch of anarchism has developed that attempts to make that bug into a feature, most insurrectionists reject formal (large scale) organisation and restrict themselves to small affinity groups.

If you are an insurrectionist you probably have some theory that explains how networks of affinity groups can create a revolutionary situation. So most of this discussion is probably not that relevant to you. Except of course that while you might imagine a revolution being made by a multitude of small groups with no strong links between them its impossible to imagine society being self-managed in such a fashion. That manufacture of a laptop or the running of a hospital requires the coordination of millions when you take into account all the materials that must be extracted, refined, reacted, shaped, placed, trained for and operated.

If you see a need for organisations of more than 50 or so to exist then Dunbar’s number requires that you embrace administration.

Administration is essentially processes of collecting information and then storing and presenting it in ways that are accessible to those who need to know. Generally, in an anarchist organisation, the membership. Administration is also the workload of following the methods by which decisions can be made from such information. For instance the planning for and physical work of collating the texts of motions, preparing multiple copies of an Internal Bulletin and physically getting it to every member by the required deadline.

In larger groups that decision making process needs to be standardised, so results can be tabulated across an organisation by having every branch discuss and vote on the same text for instance. Administration can also be standardising the way work is done with contacts so people’s impressions of the organisation are not primarily formed by whoever they happened to talk to.

When well done, administration as distinct from decision making is often almost invisible because it simply happens. Invisible because the entire purpose is to remove sources of friction that can prevent organising occurring or which causes organisational efforts to fail. If the Internal Bulletin arrives in your hands in plenty of time you are unlikely to give much thought to the work that made than happen because you will be interested in the content within. Unlike success, failure is very visible, no Internal Bulletin, no way of discussing motions. If in terms of more public failure if, you have printed 50,000 leaflets for a mass demonstration and all the members arrive to discover that no one thought of how to get the leaflets to the meetup point, then that probably turns the day into a total failure. If 20 boxes of leaflets are just there then no one gives it further thought.

That 50,000 leaflet example isn’t being used accidentally but because its quite a good example of a situation where the Dunbar limit is important. If you want to distribute that number of leaflets within the short space of time a demonstration is forming up for you need to be coordinating with 50 - 100 other people in a relatively complex multistep process.

Let’s break down the steps to use as an example;

- a process of discussion / decision making through which the decision is made that this act (leaflet distribution) will be the major priority. This in itself might be quite complex, for instance if it requires bringing everyone together in the same space at the same time for a discussion at which such a decision can be made.
- the delegation of key tasks to individuals and subgroups including
drafting the text
- agreeing the draft text
- laying out the text in a presentable form
- printing 50,000 copies of the text
- paying for that printing, the sum is around 1000 euro so there may be a whole set of sub processes around getting that money together, getting approval for expenditure and communicating with the printer so that they are confident enough about being paid to do the work
- taking delivery of the leaflets during working hours in a location where they can be at least briefly stored. This is far from trivial as if it’s an A4 leaflet on glossy paper that’s probably 70-100 boxes, each of a weight that makes it hard to carry more than one for any distance.
- Setting a meet up point for everyone taking part in the distribution at the right time and place from the distribution, communicating this to people and ensuring they agree / are going to turn up
- Transporting 70 plus heavy boxes to the meet up point
- Dividing up the leaflets among distributors and ensuring that they are dispersed through the gathering crowd in a way that maximises distribution opportunities

The complexity of the task should be clear from that example. But now consider if you only want to distribute 1000 leaflets. Although all the same stages are involved that is suddenly a very much easier task because it can be carried out by 2 or 3 people. A chat over coffee or a pint, a couple of last minutes phone calls and everything will probably come together.

If you are an insurrectionist then the answer is perhaps only to distribute small number of leaflets so that large scale coordination is never required. That’s the reason why informal scenes are dominated by zines and one person / small group blogs. It may also be why informal scenes are often quite dominated by argument and serious fallings out. That after all is a mechanism by which a group is kept inside the Dunbar limit, growth is dealt with through fragmentation and if need be expulsion.

If you are a Leninist or some other variety of authoritarian socialist, liberal or green the distribution process is simplier. One or two leaders make the key decisions and then order the members to provide the funding, do the collection & carrying, turn up for the distribution and go to their pre-arranged points. Generally such leaders are at least in part selected on their ability to get people to happily follow such orders - this side of the revolution.

If you are a business then you simply pay people to carry out all the different tasks described, at differential rates according to how much skill each tasks requires. Probably the person managing the process overall gets quite well paid, the writer & designer pretty well paid, the person organising the distributors gets an alright wage and the distributors get as little as is legally required and as much as makes doing the work attractive. Most cities have free newspaper distributed during rush hour by precarious and often migrant workers in such a manner.

By now the nature of the challenge for organisational anarchists (those of us who see a need for groups larger than 25-30) should be coming into focus. I’d bet that very few organisational anarchists reading this piece have ever taken part in anarchist literature distribution at this scale & timetable. You have probably only done it as part of broader coalitions using either the command or pay methods described above. And that’s because the tasks described from fundraising to physical carrying break the Dunbar limit in a complex enough way that defeats your organisation’s existing coordinating ability.

Leafleting is also a useful example as in most circumstances it’s quite a boring piece of political work that people will only put limited effort into because the stakes are quite low. That is there is neither much risk in it nor all that much to be gained out of doing it successfully. If this wasn’t the case we’d probably start using a lot more of our brains on overcoming the obstacles. That is the reason why the level of complex cooperation in a mass direct action or indeed a riot often pulls in much larger numbers of people coordinating than low stakes tasks. It is much easier to maintain a sense of group cohesion in situations where group activity represents a real risk than for more mundane tasks.

4. Rules & Standard Operating Procedures

So how do we escape or at least minimise the effect of Dunbar’s limit on our organising? The answer is administration or to use a less popular term bureaucracy. Above I described how capitalist organisations or those of authoritarian political parties use administration to overcome the limit. Unless we choose to believe we will never need organisation above 30-50 we need to find anarchist administration methods. Avoiding administration isn’t a possibility unless we stay in small groups that have no meaningful links between them.

What are the requirements of anarchist administration?

The rules under which members operate have to transparent and formally stated. What gets advertised on an organisation's web page for instance or who can put out a press release in the name of the organisation. What sort of training are new members required to go through. How much agreement with an organisation's policies is needed to join and where is that defined. How are those policies changed? What sort of spending can an officer sign off on without consulting the organisation ( presuming we don’t want to require a federation wide vote every time a paper clip needs to be bought). And so on.

In small groups a lot of this is often simply done through trust. It’s as organisations grow that major problems can occur as the Dunbar limit is exceeded and so officers no longer have the sort of relationships with the members where an informal trust really exists. Often it takes a while for a crisis to arise because officers will tend to drawn from people who have been members for longer and this means a combination of respect for them and a relative lack of confidence of newcomers will mean problems will take a while to be raised. But a crisis point is often where a relatively new member replaces an old timer as an officer and they immediately hit difficulties because they lack the accumulated respect that gave their predecessor the required social status to fulfil the role. The result can be paralysis as they both lack the social weight to compel others to fulfil tasks easily and lack the trust required to simply get on with it, make edge case decisions and risk the consequences. On many of the occasions that I’ve witnessed such a replacement it has gone badly and over time in WSM it meant that officerships kept defaulting back to the same small group of long term members.

A fix for this is training and also the development of Standard Operating Procedures so that new people in roles will have a well defined sense of what is expected of them and what they can demand of others. You literally write down what the steps of a complex task are and how they should be followed. This article actually arose after I saw an outraged online reaction to my use of the term Standard Operating Procedure and reflected on the lack of awareness behind that reaction.

Alongside this it would be a good idea to have new officers required to come from a secretariat that had been carrying out the particular type of work around the previous person holding the role. This would also have the advantage of clarifying is someone has the skill and aptitude for a particular task before taking in on in a major way. In an anarchist organisation there shouldn’t be significant decision making powers with a role so that sort of rotation system could be preferable to an electoral one, although of course people could be elected (or delegated) to the secretariat.

Such roles should simply be about administration of decisions that are made by collective bodies that all members have inputs to. That protection is far more meaningful than trying to avoid having such roles or trying to rotate them constantly or even never having such a role as a paid position. Paid administrative positions are going to be needed in effective organisations with memberships in the hundreds because at that point administration becomes a full time job. There are huge potential pitfalls with bad implementation, just look at your local trotskyist group for examples but that’s a discussion to be developed at another point.

Modern anarchists have often built their organisational approaches around a negative reaction to the manipulative and authoritarian structures of the rest of the left. That is wherever the left put a plus we have tended to put a minus. This hasn’t worked for us, rather than building a negative mirror of authoritarian left organisation we need to build an anti authoritarian model of organisation that fills our needs.

This isn’t just for us, in the sense of just for anarchist organisations. The challenges of good administration on a mass basis is the challenge of building a society without leaders and led, bosses and workers. It’s a challenge that constantly rises in mass struggles with the danger that otherwise people default to either a tyranny of the best connected / loudest voices or copy the command systems they see in work or mainstream politics. Very often it is already the case that the major contribution anarchists can bring into such organisations is a knowledge and experience of ways of organising that offer a genuine alternative.

Anarchists have both built mass organisations that minimise such problems, such as the anarcho-syndicalist unions. We have spent a lot of time understanding the work of others on the left that have developed their own similar skills, for instance the Zapatistas in Chiapas. Delegation rather than representation and Federalism rather than Centralism being two of the key tools of such approaches.

It’s the challenge the left has always failed on, mostly because it turns administration into command and as tasks become complex the command becomes more distant and brutal. Again rule through money or brutality being the requirement of command administration systems.

Our task is to break from that and break from its cousin, administration through the invisible hand of the market. Here we can stop and consider that we have in fact returned in a manner to one of the original anarchist slogans, the one on the rear of the WSM banner.

“Freedom without Socialism is privilege and injustice;
Socialism without freedom is slavery and brutality”

WORDS: Andrew Flood (Follow Andrew on Twitter - or if you liked this piece tweet me and let me know)

With thanks for disagreements, error corrections and suggested modifications to Aileen, Aidan, Mark, Paul, Cormac and Martin. 


Like what you are reading?  Get a notification whenever we post a new article to

Anarchist Writers via Facebook or Twitter

where you can also like and comment on our articles