Bob Martin responds to the age-old argument skepticism against incremental development:
doesn't an evolving architecture just become a big mess?

The problem is, the way I see it, that Bob's example of making a "big architectural change", effectively changing the signature of a method throughout a wiki software (please excuse me for the simplification...), simply does not count in the eyes of the skeptic. There's too many people out there that don't have a clue about what is software architecture -- in my experience, most people think of software architecture as the combination of software products the project/product being developed is built on, or whatever they would describe system architecture to be. For most people, (again, in my experience) a "big architectural change" is something a whole lot "bigger" than adding a new parameter to a widely used method.

Unfortunately, it's very difficult to use any other than trivial examples to communicate ideas or "prove" that something works. The trivial examples won't convince everybody. There's always a large number of people who don't just stop at "show me". Those guys continue with "show me more" until you can't feasibly come up with an example that a) is complex enough and b) fits into your budget for getting the message through to that someone.

Having said that, I can't offer a remedy. I wish I could, but I can't.