SOA Stupidity

In my previous post on SOA Chaos 2, I discussed Jeff Schneider's view that Stupid People Shouldn't Do SOA. 

I've now been looking at Stephen Swoyer's article SOA Fatigue: It's the People and Processes, Stupid (ADTmag September 2005). He talks about several forms of stupidity affecting people, processes and organizations, which interfere with successful SOA.

• bureaucratic infighting, turf wars, petty fiefdoms
• incentive incompatibility
• inertia

In such a context, it is the collective intelligence/stupidity of the organization that matters. Inside stupid organizations, most clever people either try to protect themselves from the stupidity, or find smarter ways to exploit it to their own advantage. I think the Conmergence blog is wrong to try to blame SOA failure on stupid people. 

But I don't think it's helpful simply to blame SOA failure on organizational stupidity either. I think it makes more sense to regard the problems discussed by Swoyer as problems of interoperability. And I think one of the implications of Swoyer's discussion is that interoperability is not just a technical question but a sociotechnical question (involving people, processes and organizations). Achieving (and motivating and funding) requisite levels of interoperability with SOA is a matter for SOA governance. 

To those that have ... Organizational intelligence is an enabler for effective SOA, and might even to some extent be a result of succesful SOA. 

Swoyer reports that Kareem Yusuf (director of SOA product management at IBM) advocates a strong top-down push for service-enablement. (Well he would, wouldn't he?) This approach might make sense for improving interoperability within one organization (endo-interoperability), but isn't going to help much with interoperability between separate organizations (exo-interoperability).

Related posts: SOA Chaos, SOA Chaos 2