Swarm Orders: Time to Build on Hayek’s Taxonomy?

APRIL 16, 2013

by MAX BORDERS

If you’re a fan of Hayek, the distinction between cosmos and taxis is one you probably learned soon after you first uttered the words “spontaneous order.”
It’s not an easy distinction to grasp. But now that you get it, you realize that if more people got it, they’d see why the price system works and the planned economy does not. They’d see why it’s easier for someone to design a company than a country.
That’s fine as far as it goes. But is something missing?
As I understand it, cosmos is an unplanned or “emergent” order. Taxis is a planned order. According to Hayek, what distinguishes these two forms of order is that cosmos relies on impartial rules and taxis relies heavily on purposes, plans and commands.
Emergent orders (cosmos) have no purposes (goals or aims) per se. Only the people and organizations who make up the extended order have a purpose, and each purpose is very likely to be distinct from one person to the next and from one organization to the next.
On the other hand planned orders (taxis) do have purposes: Sell clothing at a profit might be one such purpose. Find a cure for breast cancer might be another. So while a planned order might have such a purpose, an emergent order cannot. Those who have certain purposes normally design an organization to carry out their purposes. Often they also carry out the planning and commanding, as a founder and CEO may very well plan and command with the purpose of selling shoes. But as Hayek rightly observed, mere humans run into trouble when they try to assign society some purpose, then proceed to plan as if the right sorts of people can implement it.
The extended order emerges; it cannot be planned.
The rules that provide a basis for cosmos and taxis, respectively, differ too. Hayek might have anticipated where we’re going with all this when in Law, Legislation and Liberty (vol. 1) he wrote:

Every organization in which the members are not mere tools of the organizer will determine by commands only the function to be performed by each member, the purposes to be achieved, and certain general aspects of the methods to be employed, and will leave the detail to be decided by the individuals on the basis of their respective knowledge and skills.

Hayek might have simply left it there. But he didn’t.

Does the master go too far with the following?

Rules of organizations are thus subsidiary to commands, filling in the gaps left by the commands. Such rules will be different for the different members of the organization according to the different rules which have been assigned to them, and they will have to be interpreted in light of the purposes determined by the commands. Without the assignment of a function and the determination of the ends to be pursued by particular commands, the bare abstract rule would not be sufficient to tell each individual what he must do.

This description is true for many organizations. It turns out, however, there are rules of organization that are not subsidiary to commands, and yet which suffice to make an organization thrive.

Yes, there’s cosmos. Yes, there is taxis. But is there something in between?

Some forms of organization have distinct purposes—that is, they have goals or aims—but they don't use central plans, commands, or directives at all. At companies like Valve or Morning Star, purpose and “command” are one in the same. These new organizational forms are structured like neural networks or profitable hives.
Why hives? Despite unfortunate terms like “queen” and “worker,” hives are distributed, non-hierarchical systems. As Steven Johnson writes:

The great bulk of ant information-processing relies on the chemical compounds of pheromones, also known as semiochemicals for the way they create a functional sign system among the ants. Ants secrete a finite number of chemicals from their rectal and sternal glands—and occasionally regurgitate recently digested food—as a means of communicating with other ants. Those chemical signals [rules] turn out to be the key to understanding swarm logic.

As with ants and bees, there are no managers, no directors, no assignments from above. Planning, as such, is carried out in highly localized fashion by ad hoc teams operating according to their commitment to some “mission” (read: purpose).

Recall that in this Freeman interview, Paul Green, Jr., says of his company:

At Morning Star, there are no titles; there’s no structural hierarchy. Each colleague comes into the enterprise with the same set of rights as any other colleague. Each colleague commits to a Personal Commercial Mission when they come aboard and, further, commits to what we call “Total Responsibility”—essentially, they agree that they are totally responsible for the success of the entire enterprise.

When pressed further about Morning Star operating in some sort of organizational “anarchy,” Green replies:

I guess it is anarchy in the sense that there’s no structural chain of command or hierarchy—no “government” of sorts. But it would be a mistake to assume that it’s disordered or without structure. On the contrary, it’s very ordered, and there is structure. The difference is in how we arrive at order and structure. It’s not through some sort of centralized command-and-control hierarchy; it’s a group of individuals developing order through a social network of sorts—perhaps you could call it “spontaneous order”—based on circumstances and needs.

But according to Hayek, planned orders can happen within spontaneous order, but not the other way around.

Yanis Varoufakis has recently enjoyed some viral love by going somewhat against Hayekian orthodoxy when it comes to cosmos and taxis. In a great post about his company Valve, he writes:

Hayek’s argument that markets protect us from serfdom (i.e. from authoritarian hierarchies) is weakened substantially by the fact that he has precious little to say about corporate serfdom; about the hierarchies that millions must submit to (when working for Wal-Mart or Microsoft for instance) in order to make a living or to get a chance to unfold their talents.

“Swarm intelligence,” “crowdsourcing,” “peer-to-peer companies,” and other tech-enabled trends seem to land squarely in between cosmos and taxis. These organizational forms have distinct goals but don’t prescribe any means of reaching said goals. A mission or purpose may be something more than a “bare abstract rule,” but it is certainly something less than a concrete plan implemented through command and control.
In any case, these swarm forms can’t be considered cosmic orders because they rely on purposive behavior to form. (For example: “Create high quality tomato-based products for customers” or “Collaborate to solve problem X.”) But they can’t be planned orders, either, because there is no formal management hierarchy and no one issuing commands within these organizations. And yet these orders have emergent properties, even though the agents that make them up share a purpose. Indeed, no executive or planner can determine how a problem will get solved through crowdsourcing, nor can any one person at a bossless organization replace peer collaboration with directives.
These new organizations throw a wrench in Frederick Taylor’s ideal factory firm with its master managers and workers as cogs. They also challenge—or at least stretch—Coase’s theory of the firm. (Coase thought more or less that firms would arrange hierarchically because it’s too costly to organize them like a market. But now those costs are coming down.)
What makes things somewhat more perplexing is that these different forms of order can be nested within—and sometimes dependent upon—the extended order (cosmos). For example, if we crowdsource a solution to a biochemistry problem or collaborate in a bossless organization, it may be that neither activity is likely to crop up in North Korea, though it might certainly happen in North Carolina and Northern California.
So if they’re neither cosmos nor taxis, what should we call these forms of organization?
I think we can safely call them something like smínos, following Hayek’s use of the Greek terms. Smínos (Σμ?νος) means “swarm” or “flock.”
This term may not be perfect. It may be that we need new words for yet other additions. Or it may be that “swarm” doesn’t satisfy our sense that humans are not operating on such strict algorithms as bees, ants, and birds, and that our agency makes our form of swarm intelligence something more peculiar. Fair enough. But I would like tentatively to propose a new category of order.
What if we start seeing more and more of these swarm orders forming in the private sector? If we start looking for ways to let orders emerge within our organizations, people may see what’s possible without formal authorities and central planners—and they may like what they see. As the venerable economist Lynne Kiesling writes: “The desire for fully-specified, legalistic, control-oriented regulation leads to this type of primitive, planned order.” Such a primitive form of organization does not just take its toll on the potential value achievable through open collaboration. But in strictly planned orders, “minds wither and atrophy, increasing the primitive and simplistic nature of the resulting order.”
On the other hand:

Rules that allow for the challenge and application of individual creativity and personal knowledge make for more robust institutions, and enable minds to thrive in an order that is complex beyond their understanding, without their having to understand its entirety. Thus relinquishing the base desire to control and manage is crucial to wellbeing, growth, and living together in civil society.

I think most of us can agree we want a condition of well-being, growth, and civil society. So now the question becomes: How do we move closer to it?

Once executives begin to see the power of decentralization inside their own organizations, the lessons of the extended order may follow much more easily. In other words, the best place to start making change is by implementing good rules and being less autocratic in our own businesses. We may find such change profitable.