Following my connections on Twitter is always interesting during conferences. Today, Cat Swetel is at DevOps Days Atlanta and mentioned an old post from Donald Reinertsen on The Cult of the Root Cause that talks about the abuse of the concept behind the Five Whys. He's specifically concerned when people apply the tool without bringing their brain to the game. Is it valuable to find the root cause? Is it more expedient to fix a symptom or intermediate cause? Does it depend on the time available?
Even more interesting (at least to me) from his article is the nature of the system one is analyzing - and the way many people (mistakenly) apply the five whys process. Most systems contain networks of causality - there isn't necessarily just one path from the symptom backwards to a cause or forward from the presumed cause to the symptom. Unfortunately, if the five whys leads to a linear A then B then C then D then E scenario, this may not fit directly.
This is where the conversation mentioned last week comes into play. Can you reverse the logic and make the connection between the presumed cause the the effect? Are there other effects that you would expect to see if the root cause were in play? What if you don't find them? (What if you do?) What else must exist for the cause to create the effect? Is there an "and" in your logic that must be articulated? Are there other legitimate causes that could create the effect? Engage the brain. Explore. Think.
The Theory of Constraints Thinking Processes (TP) have some mechanisms to help with this - it's a structured way of articulating the intuition and ideas people have. And there are tools within the TP to help people check if the thinking makes sense - are the logical connections unclear? are they missing some steps that need to be articulated?