If, let’s assume for a while, greed is a primary reason for us to function – this will affect both centralized and distributed organizations, although in a different way. I am not sure if we can name such patterns easily, but Girard’s mimetic rivalry is certainly something interesting at least to consider.

Re point #1: I wouldn’t agree that aggregation can be a solution to rivalry or other (possibly undesirable) behavioral patterns. I’d like to avoid dialectical thinking but if some human primary behavioral patterns are basically part of our intrinsic nature – it is impossible to avoid them just by aggregating more inputs. That’s what recent financial crisis demonstrated – just because everybody is doing the same mistake it does not make the whole thing better.

The relative strength of distributed communities (before or after the hypothetical peak) may be a purely evolutionary thing – distributed communities might be relative better off because of their variety. Just like elected democracy has one advantage over aristocracy – it’s easier to change people in power when things go bad. This means also that communities that fail they have to fail quickly. That said, however, does not preclude that short-term performance (is this “strength”?) cannot be more effectively (and faster) achieved in a centralized, hierarchical structure, like in a traditional business environment. A distributed may take longer, may be more durable, and possibly will fail quicker if the idea behind it is wrong. Everything depends if you want to “get rich quick” or not.

I think I’ll start with the question no 3 about Peak Hierarchies. I think I rushed a bit with positioning my idea as an argument against it in absolute terms – it is more just a hole in the reasoning that needs to be filled up. It is a question about the efficiency of organisations in one particular aspect – which is containing the mimetic rivalry. I have not seen anything proposed to answer that question. I don’t say that there are no other means to contain mimetism, or that hierarchy is the best way to do it. I only maintain that we need to propose something and then compare it’s efficiency in that role to the efficiency of hierarchies, because it is a sine qua non of any social order.

I guess this also answer the question no 2.

Ad question no 1 – The mechanism of mimetic rivalry is the same on-line and off-line. My point is not that much about the on-line – as it is about any new social organisation. Only because we do build new organisation forms online this is so important in that area.

But what is most interesting for me is not the theory – but the practice. I started to think about those things because I could not agree with the inefficiencies of current most predominant on-line modes of cooperation. This is still only my intuition of an amateur anthropolog – but my diagnosis for that is that mimetic rivarly is the key here. One example of that is the often analyzed here wikipedia. Another one – one that I know much better because of my insider position – is the Perl language and more precisely it’s Open Source libraries. It is widely acknowledged by the community that the libraries are unnecessarily fragmented with huge amount of ‘reinventing the wheel’. The problem with that is not only the inefficiency because of the duplicated efforts and infrastructure overhead – but, what is more important, it also destroys the overall value of CPAN (the main online library repository) by presenting too much choice and requiring too much work to evaluate each direction.

It is also interesting that while you herald the Peak Hierarchy – still the predominant model of Open Source projects is strictly hierarchic. I have no data to support that thesis – but this is my experience. And it is also a very interesting and complex matter – because preferential attachment amplifies everything – to the effect that it is safer to join a project lead by a very strong ‘macho’ leader (often this equals ‘narcistic’).

Thanks for your interesting contribution. There are certainly problems that adhere to online communities, and that may be related to what you call mimetic rivalry. However, it is not clear to me why this militates against the Peak Hierarchy meme.

So a few questions:

1) why are the problems of online communities different that those of others in that respect? (my sense is that human organisation is difficult, period, but that actually, the more free aggregation that online communities offer are a specific solution to avoid mimetic rivalries, i.e. a conflict avoidance scheme that limits conflicts to where they are really necessary …)

2) why would hierarchies be better suited to avoid it than decentralized or distributed organisational forms? To my knowledge, hierarchy driven social organisation where pretty much driven by conflicts and wars. My sense is that decentralized democracies and distributed open design communities would go less to war than hierarchical social structures.

3) Apart from my questions: peak hierarchy claims to be a factual statement (or rather, a hypothesis about reality, that needs to be tested and confirmed), i.e. that more distributed forms of organisation are more efficient than hierarchical ones. This is to my point of view, independent of mimetic rivalry, that operates in all social systems, not just in distributed ones.

Would I be correct in concluding that your point of view means that hierarchical systems are the best to control and mitigate mimetic rivalry and that therefore, hierarchy is poised to remain a dominant social form?

Michel