Skepticism leads to Understanding
For over four years the enthusiasm and energy of the XP/Agile community has been nearly overwhelming. The rest of the industry is still reeling with the onslaught of agile-this, and agile-that. However, that enthusiasm now needs to be tempered with a healthy dose of skepticism.
I'm not talking about skepticism that's negative and cold. I'm talking about healthy skepticism that doubts anything that is not well understood and demonstrated with empirical data. A skeptic is a person who looks closely and dispassionately at things. When a skeptic finally decides to use something like XP, it's because he understands it, not because he believes in it.
For years I was an OO enthusiast. It was not until I became an OO skeptic that I started understanding what OO was really all about. Oddly enough that understanding led me to be able to teach others what OO was about, and to sell a great number of skeptics that OO would actually be good for them. I anticipate the same kind of conversion with the XP/Agile disciplines. As we look carefully and dispationately at why and when these methods work, and when they don't we will understand them much better, and be able to teach that understanding to many others.
Append your comments below:
- Aug 21, 2004, Kelley Harris: Uncle Bob - You’re in a good position to help lead that scrutiny of agile methods, method-by-method. Consider having a blog & wiki page for it, and inviting the agile community to add insights/data and to continually refine/refactor them. While I greatly enjoyed Craig Larman’s Keynote at XP Agile Universe 2004, the data presented was more about the failures of waterfall methods than the successes of agile methods. It would help for the community to have a master list of project successes to point to.
- Sep 12, 2004, Jason Yip: Hear hear. Something I’m interested in now, is literature on the lines of “XP in excruciating detail” mainly to force the increase of awareness of why we do things and identify the areas where we don’t know why.
- Oct 2, 2004, James Bullock: Well, Uncle Bob, you’re right, and your commenters are wrong, in part. Yes, some skepticism is in order, because among other things most of the practices, and certainly the “values” sometimes collected under “agile” have been around for a long time. And sometimes didn’t take. And sometimes didn’t work. And it’s wonderful when they take. And making them take & work in any given organization at any given time is hard.
You don’t need another list of “waterfall failures” as Kelly Harris correctly points out. You don’t need a list of “agile successes” either, certainly not one collected and populated by “the community.” Data produced by advocates for, or advocated against is suspect. You need a place to talk about successes & failures together, to ask: “Why?” and “What’s different here?” and “How can we make this work?”
Many methodologies have been the grand solution, taken up by standard bearers, and so on. The much-maligned “waterfall” is simply an idealized, explanatory model hijacked by folks in pursuit of predictability. Perfectly understandable search for entirely desirable predictability. It simply wasn’t real, and Royce’s insight becomes tarnished through the mis-use.
Let’s not forget the SEI / CMM, the various big-document life-cycles (2167 / 2167a, 12207, SDLC and others.) And of course RUP. All of these were embraced without skepticism (by the less-than-skeptical) as was the “pc programming revolution” which was not even a little bit like building that other kind of software, just ask them, and god-help-us the “web / internet revolution” which was similarly a new, new thing. Not.
A little skepticism applied to any of these would have spared us a whole lot of grief then, and a fair amount of “agile” rhetoric now. These methods have successes and failures. “Agile” methods have successes and failures I suspect. Much is reported of success and failure, with the word “agile” attached, or with any one of the named methodologies, so I’m inclined to believe that there are some of both.
So, please, have a reasoned discussion of methods, and the challenges involved in embracing those methods. Have an experience-based discussion of how development works. Please inch us another little bit further along in this fascinating field. If it’s a dicussion, I’ll be right there with you. If I have to sign a manifesto to get in and play advocacy-du-trend well, sorry. Seen too much of that before. It’s not helpful.
- Jan 1, 2004, Anonymous: I've been skeptical of several things in the XP community for a while. I've long since opted out of forum discussions because it seems to be the same small choir preaching to itself, and vehemently putting down anyone who suggests they might be a bit off-key. As far as I'm concerned, the XP community has turned into one of the most dogmatic groups possible. It doesn't help when the key proponent rewrites the "manifesto" and adds to it a considerable amount of noise and political agenda.
Add Child Page to SkepticismLeadsToUnderstanding