Starfish? Spiders? More Like Birds.

In the circles I move in, there’s been a lot of discussion lately about Starfish and Spiders; reference to the 2006 book by Ori Brafman and Rod Beckstrom.

The idea behind the book, which I have not read (why should I have to fund these guys just to understand their point), is that top-down command and control style organizations resemble a spider, and that if the head of the spider is removed, the organization dies.

A starfish organization, in contrast, can survive damage, and in fact after one of its arms is severed can not only repair the arm, but the severed arm can re-grow a new body. Nice enough analogy, and good for getting the point across to thick-skulled CEO’s still mourning the apparent loss of their cheese.

However, I find the analogy a bit weak; the “starfish” concept doesn’t actually explain a lot of the behavioral properties that underlie “starfish” organizations. Folks in the coworking community rightly believe that it is a starfish-style movement: leaderless and self-healing.

Flocking behavior (as seen with birds and insects) is a more instructive analogy to me. On first glance, many naïvely assume that flocks follow a leader. Not true. Individual members of a flock obey just three simple rules, and this is all that’s required to produce complex flocking behavior:

  • Separation: Steer to avoid crowding local flockmates
  • Alignment: Steer towards the average heading of local flockmates
  • Cohesion: Steer to move towards the average position of local flockmates

Quoting from Wikipedia (so it must be true), “In flocking simulations, there is no central control; each bird behaves autonomously. In other words, each bird has to decide for itself which flocks to consider as its environment. Usually environment is defined as a circle (2D) or sphere (3D) with a certain radius (representing reach). A basic implementation of a flocking algorithm has complexity O(n2) – each bird searches through all other birds to find those who falls into his environment.

The implementation of coworking is flock-like. The spread of coworking is starfishy.

The reason why so many people have trouble defining coworking is because it defies centralized control, or the notion of a flock leader.  The reason why people say, “the only way to understand coworking is to do it,” is because it is fundamentally a flocking behavior which relies on individual execution of the flocking algorithm rules.

Flocking also explains why so many coworking environments end up selecting for the right people, with no defined rules or central control; each bird chooses whether the environment is right for her. The flock self selects.

So, if you’re having trouble explaining why your local coworking group has anything to do with starfish, maybe it’s time to start talking about your flock instead.


#1 Zvi Band on 11.10.08 at 11:18 am

Interesting perspective.

However, there is some central control – a catalyst to initiate a meetup. Otherwise the flocking algorithm doesn’t apply – we’re all flying by ourselves.

I’m working on organizing a co-working group in DC this week, we’ll see how that goes!

#2 davetroy on 11.10.08 at 11:43 am

Zvi – thanks for the input. I’d argue however that catalyst!=control. Once started, that initial catalyst is no longer required. And any one individual could have been that catalyst, no one “leader” with unique skills was really required.

Anyway, we’re splitting hairs… good luck on your efforts in DC. As you know I’m trying to be that catalyst in Baltimore too!

#3 Todd Sundsted on 11.10.08 at 1:23 pm

Great model! It describes both what I experience and what I see practiced in coworking, especially as the number of spaces grows and coworkers have options.

This ties in to another model I’ve been thinking about recently (thanks to Tony)… the tipping point. One person can’t flock, two people can’t flock, but clearly, at some point, a group with enough identity and cohesion develops to demonstrate this kind of behavior.

My feeling is that there are many kinds of catalysts. Sometimes it is a person, sometimes it is a cause, sometimes it emerges from a space, virtual or real. It’s necessary until the flock forms. At the tipping point, the flock behavior takes over.

#4 davetroy on 11.10.08 at 1:38 pm

Todd – I agree. The flocking algorithms don’t work until the number of peers surrounding each individual is at some critical number, perhaps 2-3 at a minimum.

Once that critical flock mass is met, then adding new members is done at a low marginal cost, and the original catalysation need not be repeated.

This explains why the best possible catalyst is to *do* a Jelly, or something like it. Simply doing the act creates the prerequisite critical mass for the flocking behavior, in all its elegance and simplicity, to take hold for the long term.

#5 Jack Hadley on 11.10.08 at 2:08 pm

Thanks, Dave, for the insightful post. And Todd, I agree with you that there are many catalysts. We are just about to open a coworking space here in Utah. Our impetus has been multi faceted. In part a person, in part a cause. We have been following the wiki and dialog for many months, but have been too busy trying to get our space open to comment. Would love your input and feedback as we get closer to opening. Here’s to flocking!

#6 Matthew Wettergreen on 11.10.08 at 2:56 pm

Great post, Dave. The presence of a catalyst is important for initiation but that catalyst does not necessarily need to be an individual. For example environmental considerations or economic situations which drive a flock to take flight and perform according to the algorithm. Similar to the economic and social conditions that exist now and are perfect for the formation of coworking worldwide.