THE HUMAN MIRROR · The Stories Between Us

The Blame Trap: Why Finding Fault May Not Resolve the Problem

Responsibility for an error and the conditions that made it possible are different inquiries.

The comfort of a single cause

After a missed deadline, organisations often want a name, a cause and a corrective action. Identifying a person can create the appearance of closure. Yet a delay may involve ambiguous priorities, untested assumptions, late approvals and a person who did not escalate risk. Naming the last person in the chain may leave the process unchanged.

When a project fails, people want to know who made the mistake. That question can be essential: decisions have owners, professional duties exist, and some conduct warrants sanction. But the search for one culpable individual often becomes a shortcut for understanding a complicated outcome. A missed deadline may involve an unrealistic commitment, late decisions, unclear hand-offs and a colleague who failed to raise a known risk. Each element changes what could reasonably have been done. If the review stops at the most visible lapse, the organisation may discipline one person while preserving the conditions that made the lapse probable. The alternative is not to dissolve responsibility into an anonymous “system”. It is to examine the sequence of contributions, including who had authority to intervene and when.

Consider a service failure in which a team member missed a mandatory check before releasing a change. The omission is real and may justify a formal response. Yet the review is incomplete if the check had become difficult to perform during a predictable rush, if escalation routes were unclear, or if managers repeatedly rewarded rapid delivery while discouraging delay. James Reason's work on human error distinguishes attention to individual acts from attention to the conditions that make errors more likely. The distinction does not absolve the individual. It asks whether punishing one person would leave the next person exposed to the same conditions. A meaningful response may include both an accountability decision and a redesign of the workflow.

Accountability and contribution answer different questions

Difficult-conversation scholarship distinguishes assigning blame from examining contribution. The distinction is valuable only when used carefully. Responsibility may be asymmetric. An employee who falsifies a report has a different kind of accountability from colleagues who failed to notice it. Exploring conditions must never become a technique for diluting wrongdoing or making those harmed share responsibility for harm.

Accountability asks who was obliged to do what and whether they met that obligation. Contribution asks how the situation developed and what else influenced it. The two inquiries can coexist. A supplier can be accountable for missing a contractual milestone even if the client contributed through repeated specification changes. A leader can be responsible for a decision while acknowledging that poor reporting limited the available evidence. Mutual contribution does not imply equal contribution, nor does it imply shared blame. In cases of deliberate deception, harassment or abuse, treating the situation as a symmetrical misunderstanding can be actively harmful. A mature review makes the distinctions explicit: individual obligations, causal contributions, preventable conditions and safeguards for the future. This preserves fairness while increasing the chance that a similar failure will be prevented.

The timing of a review matters. Immediately after a failure, participants may be anxious about blame, while senior leaders may feel compelled to offer a quick explanation. Waiting too long can lead to reconstructed memories and lost records. A pragmatic approach preserves decisions, messages and timelines early, then provides a structured chance to test explanations when people can engage seriously. Distinguish the review used for learning from any formal disciplinary or legal process where procedural fairness requires different safeguards. Confusing the two can compromise both. An organisation that demands candid reflection while implying that every admission will be used as a confession should not be surprised when the resulting account is incomplete. A good process is candid about its purpose and its limits. It can be rigorous about evidence and consequences without encouraging speculative certainty about what somebody must have been thinking.

The investigative sequence matters. Start by establishing the observable event and the applicable standard; preserve records before memories are adjusted by blame. Next identify what the people involved knew at the time, which options were available and which constraints shaped their choices. Only then decide where authority, negligence, capability or system design played a part. This sequence protects against hindsight, where a warning looks obvious once the outcome is known. It also helps distinguish a deliberate rule violation from a mistaken judgment made with incomplete information. These distinctions may not change the seriousness of the result, but they are central to fairness and prevention.

A review that learns rather than prosecutes

First: who had which obligation, authority and information, and what happened? Second: what made the failure more likely or harder to detect? The first question establishes accountability. The second supports prevention. Collapsing them produces either punishment without learning or analysis without consequences. Both are inadequate.

Begin with a timeline of observable decisions and signals, not with the names of people believed to have failed. Mark what each person could know at each point, what authority they had and what alternatives were realistically available. Invite explanations, then compare them with records. Ask what would have prevented the event even if the individual had behaved as expected; this exposes dependencies and weak controls. Separately ask whether any duty was neglected and whether a response is warranted. The order matters because early accusations can narrow the evidence people are willing to provide. The review should end with a small number of owners and changes that can be checked, not a dense list of lessons nobody can operationalise.

In a missed launch, the person responsible for final testing becomes the immediate focus because a defect escaped. The timeline later shows that testing was compressed by a late scope change, several warnings were discussed informally but not escalated, and the release criteria were never updated. The testing owner may still have failed to perform an essential check. But disciplining that individual without examining the release decision leaves the next launch exposed to the same pressure. A useful review would ask who changed the scope, who accepted the residual risk, whether the test owner had the authority to delay and what evidence of the decision survives. The resulting actions might include clearer release gates and clearer individual obligations. That is more demanding than assigning blame and more practical than declaring that the system, in some vague sense, was at fault.

Leaders can accidentally teach the wrong lesson after a failure. A punitive response to a genuine mistake may encourage future concealment; an excessively forgiving response to reckless behaviour may tell conscientious staff that standards are negotiable. The governing question is what people will infer about the organisation's standards and the safety of reporting problems. This is not an argument for a universal “no-blame culture”. The phrase is too blunt for workplaces where legal obligations, customer safety and public trust matter. A better ambition is to understand causal conditions accurately, assign responsibility proportionately and make it easier to detect the next problem before harm occurs.

Where systems language becomes an escape

In the next project retrospective, separate a factual timeline from interpretations. Identify a decision point where a different action was realistically possible. Then test one process change: a named escalation trigger, a clarified decision right or an earlier evidence review. The early signal is that similar risks become visible sooner. The method fails when “systems thinking” becomes a euphemism for avoiding a difficult performance conversation.

Leaders sometimes invoke complexity precisely when clarity is needed. “The process failed” may be true but unhelpful if a known decision-maker repeatedly ignored warnings. “We all contributed” may silence the person who carried most of the consequence. Organisations can also use post-incident learning as a public-relations exercise while avoiding difficult personnel or governance decisions. A sound approach resists both extremes: simplistic blame and responsibility-free analysis. Test the quality of a review by asking whether it can describe the causal pattern and name the relevant obligations without contradiction. A good review leaves people knowing both what happened and what must change. It does not require choosing between compassion and standards.

A credible organisation can hold people accountable while still asking why the same failure would be plausible for someone else in the same conditions.

References

James Reason (2000), Human Error: Models and Management

Chris Argyris (1991), Teaching Smart People How to Learn