Blog: Exegesis Saves! (Part 2) Transpection with James Bach

Last evening, after a long session of collecting and organizing a large number of contributed responses to yesterday’s testing challenge, I was going over my own perspectives on the sentence

“In successful agile development teams, every team member takes responsibility for quality.”

James Bach appeared on Skype, and we began an impromptu transpection session. It went more or less like this:

James: I saw your original challenge and a couple of replies. Are you familiar with the “tragedy of the commons”? That’s what the sentence reminds me of. However, I bet there is something real behind that statement; something good. I have been on management teams where I felt we were all responsible for quality, but not for all quality equally. We each had our jobs to do.

Anyway, I think “tedious” is a good descriptor. The sentence is a bland truism, like “honesty is good”. You can re-render it like this: “On a successful agile project, you won’t find anyone irresponsible about quality.” Oh THANKS for that insight.

Michael: I’ve got a bunch of problems with it. Is it a descriptive definition or normative definition? What is being distinguished here? The successful vs. the unsuccessful? The agile vs. the non-agile? Every team member vs. most team members vs. no team member? The responsible vs. the irresponsible? Does quality here mean critical thinking, or writing acceptance tests, or the absence of bugs, or making the customers happy? What’s the scope for quality here? Could you ever get anyone to admit seriously that they’re not responsible for the quality of their own work? It seems to me that, with those things all in play, the sentence doesn’t do any useful work. “On a successful agile team, every team member is responsible for apple pie.”

James: You could contrast it with waterfall: On a successful waterfall project, only some people need to be responsible for quality. That makes waterfall seem better than agile, acutally; less risk.

Michael: For somebody!

James: Or you could contrast it with the default: Under normal circumstances, people don’t take responsibility for quality (really?), unlike successful agile projects, where everyone does. Or you could take it to logical conclusion: if two people on an agile project disagree about quality, they must work through the diagreement or else one of them must leave the project (really?). Which means: Successful agile projects consist only of people who are in complete agreement about quality.

What does “responsibility for quality” actually mean in that context? Does it mean “making the product better?” Or does it mean “accepting how the product is?” Maybe the intention is: developers have to fix any bug that anyone on the project wants them to fix.

Michael: Yes. All kinds of questions are being begged, it seems to me. Now, to be fair, I haven’t yet provided a link to the article in which this exact sentence appeared (I’ll talk about that in my own reply), but when I look up keywords from the sentence on Google, I find a ton of articles that beg the quality and responsibility questions. It seems to me that the intention is usually this: “On an Agile team, everyone is responsible for preventing bugs.” That’s a nice idea, but even that statement would have serious problems. Or maybe the intention is “On an Agile team, testers get to go to the meetings too.” To some people, it means “on an Agile team, the programmers actually test their own code.” In fact, I remember in the early years of the Agile movement, a few programmers would tell me that they’d never worked with testers who had delivered any real value to the project. When I considered some of the places I had visited and testers I had met, I was dismayed to find that claim credible. In such contexts, really conscientious programmers would tend to be extra diligent about testing their own work, and would tend to assert that they and they alone were responsible for its quality. That’s my understanding of part of the genesis of XP—programmers asserting that with respect to coding bugs at least, the buck stopped with them. That seems like it could be a positive subtext.

James: I think I see this subtext, too: “On an agile project, there is no social structure or order. We are a perfect Einstein-Bose condensate of spontaneous harmony. Everyone does everything. Communication is instant. Agreement is assured. All skills are held in common. All experiences are equivalent.”

Which in practice means: The Strong Rule The Weak.

Michael: Ouch.

James: The sub-sub text is then this: “Don’t ask about social order. We don’t like to talk about that.”

Michael: “…but we do have this sentence that you’re not really supposed to question. It’s a form of Word Magic, so please don’t mess with it.”

James: How about this, then: “On an agile project, one thing that is beyond question is that everyone is resposible for quality. Also, unicorns are pretty.”

Still more to come. Stay tuned.

Want to know more? Learn about Rapid Software Testing classes here.

3 responses to “Exegesis Saves! (Part 2) Transpection with James Bach”

  1. Excellent! Especially the analogy about the Einstein-Bose condensate. It reminds me of Lene Hau who was the frist physicist to slow down, even stop light in such an entity. That is a perfect analogy to the agile team which delivers perfect quality, only the product just never escapes out of the team! I’ve seen that 🙂

  2. Tweets that mention Exegesis Saves! (Part 2) Transpection with James Bach « Developsense Blog -- says:

    […] This post was mentioned on Twitter by Michael Bolton, Michael Bolton, Anders Dinsen, Albert Gareev, Martin Jansson and others. Martin Jansson said: RT @michaelbolton: Published: Exegesis Saves! (Part 2) Transpection with James Bach #testing #softwaretesting #qa … […]

  3. […] all sorts of interesting stuff on it (see James Bach’s post here, some Michael Bolton posts here and here, and Stephen J. Hill’s post here), so we asked Michael Bolton if he would be willing […]

Leave a Reply

Your email address will not be published. Required fields are marked *