The Beauty of Broken Systems

An Appeal for a Measure of Anarchy in the Church

About 6 or 8 months ago [spring/summer 2004] I was reading David Weinberger‘s Small Pieces Loosely Joined: A Unified Theory of the Web. Amid Weinberger’s discussion of the Web, I found something to which I’ve returned several times and re-read, because the first time through, a light went on. Weinberger’s style is entertaining and easy to read, so even if you’re not particularly interrested in a theory of the Internet, you can stick with it while I quote at length, because I’m going somewhere with this… and I’ll comment below after you understand why the Internet is inherently chaotic and broken.

Tim Berners-Lee, the Web’s inventor, is reported to have said, the “Web will always be a little bit broken.”

Now we come… to the Web. Links don’t work. Email messages are misspelled. Discussion boards devoted to medical matters of life and death contain claims so false that they’d be funny if they weren’t so dangerous. The site that downloaded in a second an hour ago now takes five minutes. The link you thought would take you to pictures of endangered species instead sinks you into a porno site that sproutes new windows like poison ivy. Where’s perfection when we need it?

The imperfection of the Web isn’t a temporary lapse; it’s a design decision. It flows directly from the fact that the Web is broken on purpose.

In an imperfect world, software can be an island of perfection. Software does exactly what the programmer tells it to do; following laws of logic embedded in the programming language and carefully explicated in the programming language’s manual. …Unlike the real world, the world of programs is perfectly knowable, perfectly predictable, and perfectly controllable.

But the design assumption of the Internet was that it’s an imperfect world. The Internet isn’t just a program. It’s a physical system that uses hardware and wires. For the Internet to be robust, its designers had to keep in mind that hardware sometimes fails and cables are sometimes cut by overzealous gardeners with their Rototillers set to eleven. The original designers of the Internet understood this well. In a perfect world, a central dispatcher (a piece of software) would know where every bit needs to go and would route each the most efficient way. But that router’s every hiccough would have the effect on the rest of the system of a cardiac arrest. So the Internet was designed to have many decentralized routers, each making decisions about where to send packets next. If one of the routers goes offline – for routine maintenance or because a volcano erupts under Topeka – the packets are simply sent to another router. The Internet routes around disruption.

This feels chaotic and messy. Rather than a centralized brain making decisions based on an overview of the entire system, smaller collections of neurons act by reflex with little concern for anything except moving packets one more leg on their journey. The difference is between on the one hand, having your automobile club lay out a map that shows you a direct route from New York to San Francisco and, on the other hand, navigating by asking gas station attendants along the way who give replies such as, “Gosh, I don’t know how to get you to San Francisco, but I think you’ll be closer if you drive northwest to the next Sunoco station and ask again.” The first way works fine so long as the recommended road hasn’t been shut down for repaving and there are never any traffic jams – not to mention every “driver” in this analogy would first have to drive to the automobile club’s headquarters to pick up the map. When it comes to packets in a highly dynamic highway system, the stop-and-ask technique turns out to be not only more robust but also more efficient.

This surprises us only because we have long assumed that centralized power and efficiency go hand in hand. When you look at the systems and institutions that have advanced our culture – education, government, business, religion – the same basic picture emerges: the bigger a system is, the more control it requires. Or, to be more accurate, the more complex a system is, the more control it requires, which means the amount of required control increases even more steeply. So, while it may take only one adult and a sketch on the back of an envelope to build a tree house, by the time you get to a project such as the Hoover Dam, you have not only hundreds of managers but also managers managing the managers; and then you bring in management consulting firms to help manage the way the managers are managing the managers.

Management is good. It brings efficiency and accountability. But we also know that we like building big management structures ofr other, less honorable reasons. The larger the fiefdom, the more powerful the monarch. If your project gets large enough, you can actually become a pharaoh and have your brains hooked out through your nose and achieve an imperviousness indistinguishable from death. Management is, in short, about power as much as about efficiency. As Edward Tufte has said, “Power corrupts. PowerPoint corrupts absolutely.”

Now consider how we would have gone about building the web had we deliberately set out to do so. Generating the billions of pages on the Web, all interlinked, would have required a mobilization on the order of a world war. Because complexity requires management, we would have planned it, budgeted it, managed it,… and we would have failed miserably. If everything had to be coordinated and controlled, we’d still be processing Requests to Join and Requests to Post. We’d have editors poring through those pages, authenticating them, vetting them for scandalous and pornographic material, classifying the, and obtaining sign-offs and permissions to avoid the inevitable law suits. Yet we – all of us – have built the global Web without a single person with a business card that says “Manager, WWW.”

Our biggest joint undertaking as a species is working out splendidly, but only because we forgot to apply the theory that has guided us ever since the pyramids were built. Whether we’ve thought about it explicitly or not, we all tacitly recognize – it’s part of the Web’s common sense – that what’s on the Web was put there without permission. We know that we can say what we want in an email or on a discussion board without permission. The sense of freedom on the Web is palpable. The Web is profoundly permission-free and management-free, and we all know it.

If we say that the Web is self-organizing, it’s crucial to recognize that “self-organizing” does not imply that the Web is very organized at all. It is not, for example, as consistent, predictable, or purposeful as a protozoan….

Yet the Web works. It grows without much maintenance. It invents at insane speeds. We can get done what we want, although usually only after clicking down some dead ends. Beyond any reasonable expectation, it works. But it works only because it has remained true to its founding decision: remove the controls and we’ll have to put up with a lot of broken links and awful information, but in return we’ll get a vibrant new world, accessible to everyone and constantly in the throes of self-invention. The Web works because it’s broken.

So there it is. The Internet would probably not be what it is if we tried to engineer it and manage it — and it would be broken, but in a bad way. As it is, it’s broken in a good way, one which actually makes it work. I do appreciate irony.

Now we come to the part which isn’t really about the Internet, it’s about the church. We have a natural tendency to over-manage and over-organize things, to keep controls, procedures, and protocols in place. But we don’t always know why, and we don’t always realize what those conventions cost us. And here’s where a little anarchy would do us a world of good: instead of vetting every idea until its time has passed before we try it, why not just run with it and see what happens? This is the difference between approaching the pastor or church leaders with some ministry idea and getting the response, “Sounds interresting, put together a plan for us to look at” and getting the response, “Great, go work it out and let us know if you need anything – and let us know how it goes!”

It’s the difference between releasing and repressing.

So I want to appeal for an ecclesiastical structure (you may read “wineskin”) that is inherently geared toward releasing. One that’s not afraid of mistakes, of brokenness, of imperfection. One that not only accepts but expects these things. We are but dust, after all. Thankfully God remembers this even though we tend not to; thinking too highly of ourselves causes us to think we’re capable of more than we really are. You know, like perfection. Or real wisdom. Or the idea that we can make mistakes too big for God to cover over.

Hmmm, there’s a good one – we know theologically that we can’t commit a sin to great for God to forgive, yet we act as though we could make a mistake too big for him to deal with. By this I mean that we sometimes can stand paralyzed for fear of making the wrong decision. We tend to think things like this matter far more than they really do. I used to tell people considering becoming involved in some church ministry or other, “Go ahead and try it on and see if it fits – if it doesn’t, take it off and try something else.” They seemed to appreciate it, and seemed to be helped by it. The people I was talking to generally tended to be afraid of making a mistake and getting involved in the “wrong” ministry.

So the wineskin for which I’m appealing shouldn’t take itself so seriously that it can’t achieve any action. Primary goals should include flexibility and the willingness to fail. And whatever else, it shouldn’t waste too much energy on the structure and the organization when it could be spending said energy on the function.

I’m appealing then for something that may lack apparent central management. It may feel “chaotic and messy”, to use Weinberger’s description… but in the context of the church, one must presume that whatever central management is necessary can be accomplished by the Holy Spirit. I wonder if this could be estabilished in some fashion that would look like the early church, meeting in homes and meeting together, and just being about the business of doing church. And just maybe we could get somewhere by asking for directions at each stop, rather than insisting on having the whole roadmap up front. It’s probably a better means of navigation if you’re trying to arrive at “the land that I will show you” …which is not the type of marking I’ve ever seen on any map.

I don’t wish to insult the reader with spoon-feeding, so I urge reflection on the Weinberger quote to see if you don’t begin to see some possibility for ecclesiastical “organization” in it.

Quotation from Weinberger, David: Small Pieces Loosely Joined, Perseus Publishing 2002. Pages 76-83. (Amazon Link)