“Why is the world a failure?” asks Hilmer, speaking in a phone interview from Sydney, Australia, where he teaches at the Australian Graduate School of Management. “Because organizations are hierarchical machines, not using the thinking power of their employees?” The problem with that argument, he says, is that it doesn’t square with what you see if you take the trouble to observe what’s actually happening. “There are plenty of companies that seem to be working just fine. There are plenty of companies that have figured out ways to tap employee creativity within existing structures and hierarchies.”
But such arguments are all but drowned out these days by the crescendoing drumroll that heralds the unveiling of the reinvented, 21st century business enterprise. Just what will this radically new, post-modern, post-reengineering entity look like? Will it be the giant corporate nation-state (British Telecom/MCI, Boeing/McDonnell Douglas)? Will it be the loosely connected, virtual organization – the “adhocracy” of alliances and far-flung outsourcing contracts?
The answer is not yet clear. But whatever big new idea next emerges to capture the fancy of business thinkers, it’s hard to imagine a notion that will be more challenging than the one Wheatley and others are proposing: the biological organization; a complex, self-adaptive system; chaos theory as the next management pardigm.
“The world seeks organization,” writes Wheatley in her book. “It does not need us humans to organize it. . . . Organization wants to happen.” She instructs us to look to nature for examples of the world’s self-organizing handiwork. With no master flight plan to guide them, birds fly in flocks. Termites in Australia and Africa build towers soaring 30 feet into the air. These engineering marvels, laced with intricate tunnels and graced with arches, are the largest structures on earth in proportion to the size of their builders.
And yet, all this happens not from a detailed blueprint but as the improbable result of a curious work process, observes Wheatley. “With antennae waving, [termites] bump up against one another, notice what’s going on, and respond. Acting locally to accomplish what seems to be next, they build a complex structure that can last for centuries. Without engineers, their arches meet in the middle.”
Not to take anything away from the marvel of termite mounds, but convincing business leaders that we should run our airlines and petrochemical plants based on the termite model is likely to prove a tough sell.
When questioned by TRAINING about that potential difficulty, Wheatley characterized her book as not so much a call to action but rather “a meditative call to awareness. Awareness that a very different world view is available to us.” She acknowledges that for many people the shift to this new world view will not come easily.
As a metaphor, Wheatley’s hymn to self-organizing is not without appeal. There is, after all, evidence to show that workers are capable of self-organizing, and that the few companies that grant workers the power to make decisions affecting their jobs often prosper as a result.
But while Wheatley invokes self-organization as a metaphor, others are starting to speak of it as a new model – the next idea after reengineering. And that’s where the new world view becomes challenging. As Wheatley herself observes, “Life seeks order in a disorderly way…mess upon mess until something workable emerges.”
It is tempting to brand the self-organizing system as the next misguided management fad – the idea of the flattened hierarchy carried to the point of absurdity. (You want to order 40 boxes of photocopier paper? Please call back in 25 years. By then our new sales organization should have emerged.)
The problem with dismissing the ideas bubbling up around self-organization, however, is that, unlike some of the more faddish notions to come down the management pipeline in recent years, these ideas do have some scientific underpinnings.
Explorations into the world of chaos theory, and its spin-off field of complexity theory, are challenging some of science’s long-held assumptions about how the world works, just as quantum physics and Heisenberg’s uncertainty principle knocked Newtonian physics into a cocked hat earlier in this century. The question is, Will complexity theory provide us with useful models and metaphors for understanding the world, including the world of work? Or will it lead us down the path of strangeness to an intellectual cul-de-sac, where we can only shake our heads at the unpredictability of the world?
For decades most scientists have believed that the forces shaping the world and the universe are random, that there is no invisible, guiding hand organizing the show. But if that is so, then why, some scientists ask, do galaxies form pinwheels and certain marine creatures turn into chambered nautiluses?
The work of molecular biologist Stuart Kauffman and others at the Santa Fe Institute (SFI) has recently shed light on nature’s counterintuitive and previously invisible tendency to organize itself. Several years ago, Kauffman constructed a network of 200 lightbulbs in which each bulb was linked to two others using Boolean logic (e.g., bulb number 17 might be instructed to go on if bulb number 23 went off, and to turn itself off if bulb number 64 went on). The number of on-off configurations in such an arrangement is an astronomical [10.sup.30]. Given those numbers, chaos should reign; there should be no discernible pattern to the lighting arrangements. But in fact, after about 14 iterations, the lightbulb network settled into a pattern of just half a dozen on-off combinations.
“We have always known that simple physical systems exhibit spontaneous order: an oil droplet in water forms a sphere; snowflakes exhibit sixfold symmetry,” explains Kauffman in his book At Home in the Universe (Oxford University Press, 1995). “What is new is that the range of spontaneous order is enormously greater than we have supposed. Profound order is being discovered in large, complex and apparently random systems. I believe that this emergent order underlies not only the origin of life itself, but much of the order seen in organisms today.”
A nonliving arrangement of lightbulbs may seem far removed from a social system of human workers. But another early simulation provides a clue to complexity theory’s potential relevance. In his book Complexity (Simon & Schuster, 1993), M. Mitchell Waldrop describes how Craig Reynolds, a researcher at the nuclear physics lab in Los Alamos, NM, caused a stir in 1987 by simulating bird-flocking behavior on a computer screen. The birdlike objects, called “boids,” flew together in a flock and swerved as a unit to avoid obstacles. When forced to break apart to avoid an obstacle, they soon regrouped again into a new formation.
Yet nothing about their programming told the objects to display this collective behavior. There was no master flight plan that guided the motions of the flock. Each object was programmed individually with just three rules: Fly in the direction of the other objects; try to match velocity with neighboring boids; and avoid bumping into things. The essence of complexity theory is that simple agents obeying simple rules can interact to create elaborate and unexpected behaviors.
Taking their cue from these laboratory experiments and simulations, a few companies have found practical use for complexity theory, devising local, rule-based solutions for problems that in the past would have been addressed by a solution imposed from above.
General Motors Corp.’s truck plant in Fort Wayne, IN, formerly used a master scheduling program to determine which of 10 different paint booths painted which truck bodies as the trucks rolled off the assembly line. As long as all parts of the system moved along in sync, things worked well; but if any one piece of the system slowed down, things fell apart in a hurry.
In the early ’90s, GM switched to a complexity-based computer system in which a computer at each paint booth acts as an independent agent, “bidding” on each new paint job based on its ability to take on additional work. The calculations include the cost of each new job; for instance, can this job be done without a color changeover? This self-organizing system quickly evolved a pattern for painting trucks that reduced color changeovers at the paint booths by 50 percent and now saves GM more than 81 million a year.
Using “genetic algorithm” software that employs complexity concepts, Deere & Co., the farm-equipment maker, has developed an optimum scheduling program for the manufacture of customized seed planters, which can be assembled in a mind-boggling array of 1.6 million different configurations.
Successes such as these have drawn companies to the Santa Fe Institute by the dozens in recent years, each looking for ways to profitably apply complexity theory to its business. And these pilgrims are not confining their quests to operational applications. Some are hoping complexity theory will provide a new understanding of the development of social structures, including business organizations. Coopers & Lybrand, McKinsey & Co., and Ernst & Young have all sent people to SFI to learn more about its research, and, they hope, to find ways to roll complex adaptive systems theory into their consulting practices.
The idea that vastly complex organisms – be they cells, galaxies, economies or business organizations – can arise from a few agents interacting according to simple rules certainly does hold the promise of what Wheatley terms “a simpler way of being in this world.” Who needs 10 million lines of computer code outlining to the nth detail every step of a manufacturing or distribution chain when three or four simple rules suffice? Who needs a complicated plan for reorganizing the company when the company can organize itself? Who needs a vice president of strategic planning? And for that matter, do we really need a CEO?
But does complexity theory really lend itself to organizing human agents in a company? Michael McMaster, who works with a U.K.-based consulting firm called Knowledge Based Development Ltd., believes the answer is yes. McMaster has used complexity principles to design the work of a cross-functional team of pipefitters and welders building an off-shore oil platform – though design is probably not the right word. McMaster distilled project tasks down to just four basic rules and then set the workers, or agents, free to create their own work processes.
But while this demonstration of self-organizing theory applied to human agents may seem an important evolutionary step for complexity theorists, it doesn’t exactly break new ground in terms of describing how work gets done. The fact that groups of workers will, if given the chance, find ways to accomplish a task is hardly a revelation. This has been the driving force behind the creation of self-directed teams for years.
Even the term serf-organization is not unprecedented. Six years ago, researchers into workplace learning began to explore the ways in which workers organize themselves into “communities of practice” to accomplish jobs, and how these communities self-organize in ways that are often invisible to supervisors and managers (see “Communities of Practice,” TRAINING, February).
In a new field such as complexity theory, however, every modest success at applying the idea to the workplace quickly inflames the imagination with the hope of bigger things to come. Echoing the youthful prognostication of young Werner Heisenberg 75 years before them, the disciples of self-organization proclaim they are on the path of scientific discoveries that will shed light on the whole range of human intelligence. “Approaches based on the principles of complex adaptive systems theory will completely change the way we organize, compete, think of industries and do business,” declares McMaster.
But precisely how complexity theory will shape the work world is not clear. If you set the fight number of agents in motion, each agent following the fight set of three or four simple rules, complexity theory predicts that these agents will eventually, like boids, organize into something large, complex and unexpected. But what, exactly? The high-performance, global business organization of the 21st century? Or a wonderfully complex mechanism for achieving bankruptcy and ruin?
And in what sort of environment will these agents be set loose to do their self-organizing? Must they be unencumbered by any trace of hierarchy, or management structure, including a CEO and board of directors? Or will there still need to be someone, somewhere, calling the shots – at least some of the time?
These are legitimate questions for which budding complexity theorists have no answers yet, admits Christopher Meyer, who heads up Ernst & Young’s Center for Business Innovation in Boston. But Meyer believes an application of complexity theory to organizational design will one day be found. Last year he mailed 15,000 copies of At Home in the Universe to Ernst & Young clients and drew dozens of curious companies to a three-day symposium called “Embracing Complexity.”
Meyer acknowledges that a lot of ground remains to be covered before complexity theory will be palatable to the traditionalist in the business world. “There’s no solid theory yet to explain how aggregations of agents in something like a business organization will interact to create the emergent properties of a new organization,” he says. Computer simulations such as boids, while tantalizing, do not prompt one to rush straight out and demolish one’s reporting structure. “Today’s simulations don’t have what you’d call a hierarchy of objectives,” he says. “Teams operating under sets of three or four rules may do their tasks well, but what do you need to make those tasks come together in a larger sense?”
Meyer sees hope, however, and perhaps even a three-point mad map for bringing complexity theory to fruition as a management tool. The first point, which has already been reached, he says, is to apply complexity theory to operational problems. “Complexity-based models at General Motors and John Deere have proven that they solve operational problems better than linear techniques,” says Meyer.
The next phase, which researchers are starting to close in on, is to develop simulations that have what Meyer calls a “feel of real life to them.” The final step will be reached when these real-life simulations are so realistic that they can be used to solve organizational problems.
That third step, Meyer acknowledges, is still years in the future. But the second step, the real-life simulation, is being taken today, he believes. One such simulation can be found in a computer-generated world called Sugarscape.
In their book Growing Artificial Societies (MIT Press, 1996), researchers Joshua Epstein and Robert Axtell describe how they use agent-based computer modeling to create a landscape – essentially a large, two-dimensional grid – called Sugarscape. Each square on the grid is assigned a certain amount of sugar and the ability to replenish its sugar at varying rates. Some squares replenish quickly, others more slowly.
Set in motion in this Sugarscape are the computerized actors in the drama – agents that like to eat sugar. These agents are “born” into the landscape with the ability to see, a metabolism, and a set of genetic attributes. They are set in motion by a simple set of rules: “Look around as far as your vision permits, find the spot with the most sugar, go there and eat the sugar.” The agents metabolize sugar as they move from place to place. If their movements don’t turn up enough sugar to sustain their metabolic rate, they die. Sugarscape is a cruel world.
Now, if this simulation doesn’t exactly convey a “real-life” feel for what occurs in your workplace, it may be because the description here has been monstrously simplified. Or then again, it may be because Sugarscape is a crude first step at using complexity theory to model human behavior. Nonetheless, the authors argue that the actions of their Sugarscape agents reveal uncanny parallels to such human activities as “trade, migration, group formation, combat, interaction with an environment, transmission of culture, propagation of disease, and population dynamics.”
Interestingly enough, the folks at SFI who have pioneered the hard-science research into complexity theory make no promises as to the relevance their new ideas will find in the corporate world. “We understand the attractiveness of using complexity theory as an organizational model,” says Bruce Abel, SFI’s vice president of research. “Complex adaptive systems teach us that there is no stability, things are constantly changing. To businesses being buffeted by market changes and technological changes, it seems natural that there should be some applicability.”
But it’s not an easy connection to make, confesses Abel, especially once you start looking at large organizations and the myriad sets of relationships within them. “It’s a lot harder to study a city than it is to study an anthill,” he says.
The extent to which business leaders embrace complexity theory as a self-organizing model may ultimately depend on whether they come to perceive it as a true model, or simply an interesting way to think about the world.
To embrace complexity as a model means that business leaders also must be prepared to embrace the scenario described in A Simpler Way: employees turned loose to heap mess upon mess until a new workable system emerges. Moreover, according to the full-cloth version of self-organization, there’s absolutely no way to predict in any way, shape or form what will ultimately emerge, or how long the emergence will take. Remember, complexity theory derives from chaos theory, which gives us the famously cliched analogy to describe the unknownability of the world’s infinite interdependencies: A butterfly flaps its wing in Brazil; a stock market in Tokyo crumbles.
Barring the development of some sort of model that packages the theoretical power of chaos theory into a palatable form, we are left with self-organization as what? A metaphor? One more expression of a basic idea that has been recycling through the management vocabulary for years under various names: Theory Y, quality circles, self-managing teams, empowerment?
But if self-organization ends up as nothing more than a metaphor, how is it different from teams or empowerment? What does it bring that’s new?
“It’s a deeper way of thinking about these things,” says consultant Peter Block of West Mystic, CT. “We’ve been operating on a metaphor of engineering and control, a metaphor of economic scarcity. Here’s a new idea that proposes an abundance of possibilities instead of scarcity, that asks people to imagine all these new possibilities if they just let go of their controlling behavior. I think that’s good.”
One final speculative question about self-organizing systems: Suppose self-organizing doesn’t turn out to be just the latest craze on the seminar circuit? Suppose we do find a way to leverage these new ideas into something transformational? What, then, would the emergent self-organizing entities look like?
“There are two good examples, one obvious, one maybe less so,” says Thomas Malone, a professor at MIT’s Sloan School of Management and co-director of a project there called the Initiative for Inventing the Organizations of the 21st Century.
The obvious analogy, says Malone, is the Internet, a highly decentralized set of agreements on ways to communicate with no overarching control. “It’s interesting to contemplate whether the Internet would have grown as fast or as large as it has if it had been run in a more centralized way by someone like AT&T,” says Malone. “I don’t think so.”
The Internet is often cited as an example of what a self-organizing system might look like. But in the eyes of enthusiasts, ifs more than a model; it’s also a means of getting there. In order to reap the advantages of decentralized decision-making that self-organization promises, says Malone, you have to give everyone access to all information. “Information technology is what makes self-organizing possible.”
But this analogy ignores the fact that most companies are looking for something a little less chaotic than the Internet when they set up their own corporate intranets. “Most companies find ways to overlay a set of rules, some sort of road map for how information gets shared,” observes Carla O’Dell, president of the American Productivity and Quality Center in Houston, which is preparing to release a report on the role that information technology plays in corporate knowledge management strategies. “You can’t share all the information,” says O’Dell. “Without some organizing filter, e-mail systems are brought to their knees; workers are paralyzed by information overload.”
Not that O’Dell means to dismiss the idea of self-organization. “I think it’s the most powerful and interesting idea to come along in the last 20 years,” she says. “But it’s a difficult idea to grasp right now. How do you balance the freedom of local self-organizing with the need for global perspective and some degree of control?”