Conway's Law[1] tells us that organizations build computer systems that mirror their own structure and organization. No surprise there.

Most organizations, when you look closely at them, exhibit complexities in their information flows that are mind-boggling. Sometimes, nobody - not even top management - have a good grasp of the so called "business rules" that govern the information flows.

Complex information flows seem to sprout naturally wherever you have people talking to each other. These informal information flows are often the ones that make the entire enterprise tick. No surpise there either perhaps.

What is surprising, to me at any rate, is the extent to which accepted wisdom in Enterprise Integration has it that Conway's Law is a scourge. The idea that the role of middleware in enterprises is somehow to eradicate all the back-roads and corridors of un-controlled information flow that Conway's Law tells us exist. The idea that computing can liberate an organization from the messy, wet complexity of human, social interactions and replace it with shiny, dry, metallic laser beams of rationalized, straight line logic.

Maybe in can, but I doubt it, and I strongly doubt I would like to work in the organization that resulted from such a logical integration of all informal information flows. Trite truism notwithstanding, organizations truly are all about people.

If there is one thing we know about people it is that people have disputes all the time. No surprises there. Politics, the rule of law, society, all those words are techniques we have developed over the millennia to deal with human-to-human disagreements.

Recently, I have been masticating the idea that enterprise integration is, when you boil it down, a dispute resolution problem. Computer system A sees the world one way, computer system B sees the world another way, computer system C has a third way and thinks A and B are deeply wrong in how they see the world. etc. etc. The custodians of systems A, B and C need to work together for the good of the greater organization. However, they need to do so without sacrificing their autonomy and without necessarily trusting or even liking the other parties in the integrated system.

Take computers out of the equation for the moment. If A, B and C were people, how would you approach the problem? The list of possibilities you come up with will read like a table of contents of a political science/industrial relations handbook. Now make A,B and C computers rather than people. Has the set of options changed as a result? I don't think so. Computers in organizations are just proxies for people. Conway's Law rules.

So where am I going with this? Well, I am not sure. I am just thinking out loud at the moment. I'll get somewhere eventually. I'm thinking of phrases like client/server and mapping them into mediator/mediatee relationships in my head. A client and a server are two machines "in dispute" that successfully work together thanks to the efforts of a mediator - an API or a protocol. Does that analogy work? Yes, I think it does. Can the same analysis be applied to the Web? What is the Web in industrial relations terms?

Maybe the Web is a giant secretariat for mediators and mediatees to get together and build their relationships on ready-made components such as URLs, HTTP and XML. Maybe the term "interoperability" can be best understood as successful dispute resolution. If so, then the Web is the most accomplished dispute resolution device in the history of the planet.

What makes a good electronic dispute resolver? Now there is a good question. Maybe when we have an answer to that, we will truly know what the Web really is.

Conway's Law - definition

ITworld.com, Ebusiness in the Enterprise

• Email
• Print
• Share
  • Slashdot
  • Digg
  • Stumble
  • Reddit
  • Newsvine