Ask why repeatedly to move from symptom to cause; each answer becomes the next question.
Sakichi Toyoda's method. Its known weakness is that it produces a single chain by construction, so it can generate a plausible causal chain while missing the others. Complex failures commonly have branching or interacting causes and need evidence beyond repeated questioning.
In practice
- An outage traced through five whys lands on "the engineer forgot", which stops the inquiry exactly where the interesting part begins.
- Asked five times why a supplier missed a date, a buyer arrives at one chain and misses that three separate conditions had to coincide.
- Possible explanations
- One explanation for its usefulness is that it forces a team past the first plausible answer. Another is organisational: a named method makes it socially acceptable to keep questioning a senior person's account.
- In its favour
- It is cheap, needs no training or data, and in manufacturing it repeatedly finds a fixable process cause where blame would otherwise stop at the operator.
- Against it
- It produces a single chain by construction, and safety research on complex systems finds branching and interacting causes, so the method can end in a plausible story the evidence does not support.
- Standing
- Prone to the root cause fallacy: it produces a single chain by construction, and complex failures usually have branching or interacting causes.
- Further reading
- Wikipedia
- Subject
- Organizations & Bureaucracy
- Essay
- Rules for running organisations, and for pressing them