More on effectiveness, formality, etc.

More interesting conversation on Roly Perera's blog, happening at a glacial pace over the course of some months. This is connected to Kiczales' OOPSLA keynote and the ensuing discussion. My comment is long, but I'll copy a bit here, because I think it captures my current thinking about this question:

But I would go further. To summarize my view, I would say that in any particular view or way of looking at a system, there is a tension between effectiveness and informality (effectiveness requires formalism, while informality is hard to make effective). There is also a relationship between the multiple views themselves, and this relationship can also be viewed in multiple ways, some more effective and some more informal. What I believe we need (I don't know whether this is Gregor's point, or whether he would agree) are better tools for navigating and articulating the spectrum from effective to informal, and ways to apply these tools to the connections between multiple perspectives (some more effective, some more informal) onto software systems. There is a two-level structure here, but it should not be too confusing for a programmer.

This is sort of confusing, and I'd like to expand and clarify it, at least so as to present a clear target. But it will have to wait until some other time.

As I look back, it turns out that I'm basically restating my previous comments in different words. Which is fine: it's nice to see that I'm consistent.