The bridge failure below is fictional. It is not a description of a real accident and does not imply that every disaster has the same causal structure.
A bridge collapses hours after a routine inspection. The report contains no mention of a crack, and attention immediately turns to the inspector who signed it.
The inspector may have failed and may bear professional, moral or legal responsibility. That alone does not show that the error fully explains the collapse. Why was the defect not found earlier? How did the second-line check work? What did management know about deferred maintenance? Which warning signals were recorded but never converted into action?
1. Accountability and causation are different questions
Accountability asks who breached a duty, what they could reasonably know and which consequences they should bear. Causal analysis examines the chain of conditions, decisions and failed defences without which the event would not have happened or would have caused less harm.
Both conclusions may be true: a person made a serious error, and the organisation created conditions that made such an error more likely and less detectable. A systems approach does not automatically excuse the individual. It asks a different question.
| Layer | Question | What it does not decide alone |
|---|---|---|
| Finding | What happened and what can be documented? | Motive or degree of culpability. |
| Human action | Which decision or omission contributed? | Whether it was the only or decisive cause. |
| System conditions | How did procedures, resources, oversight, culture and time pressure operate? | Whether the individual can be accountable. |
| Normative conclusion | What follows under law, professional rules or morality? | Which technical measure will prevent recurrence. |
2. What the evidence actually supports
James Reason’s review distinguished a person approach from a system approach to error. The first focuses on unsafe acts by people at the sharp end. The second examines working conditions and defences intended to catch ordinary human fallibility or limit its consequences. This is a safety framework, not an experiment proving that an individual is never the main cause.
Roese and Vohs review hindsight bias: once an outcome is known, the earlier event can appear more predictable than it did beforehand. The work spans laboratory and applied research and describes how outcome knowledge may narrow attention to one causal story. The effect is not identical in every person or investigation.
Webster’s three experiments found that experimentally induced or measured need for closure could strengthen overattribution to personal or situational causes, depending on the task. The study supports a motivational influence on attribution; it does not establish a universal public need to blame someone after every disaster.
3. Other explanations for rapid blame
Focusing quickly on one person need not be a psychological bias. Their failure may be well documented, an institution may be withholding information, or legal procedure may properly begin with an individual act. A name is also easier to communicate than a dispersed chain of organisational decisions.
It is therefore unsafe to diagnose “fundamental attribution error” from a headline or to claim that blame always hides the truth. The better question is whether the available evidence supports additional causal layers and whether investigators actually tested them.
4. Investigating without a false choice
- Reconstruct the timeline: separate documented events from later interpretation.
- Identify failed defences: why did one error pass through controls and reach the outcome?
- Compare no-harm cases: did the same deviation occur before without a catastrophic result?
- Separate authority from outcome: what could the person actually control and know at the time?
- Do not end with a sanction: discipline alone does not repair weak oversight, unclear procedures or a missing technical safeguard.
5. What this article claims — and what it does not
Outcome knowledge can affect perceived predictability, and systems analysis can reveal conditions that personal blame alone leaves unresolved.
Not every accountable person is a scapegoat, a system is not always more culpable, and psychological studies do not decide legal responsibility.
6. What would weaken the main claim?
It would be weakened by rigorous analyses across different accident types in which systematic investigation of organisational conditions repeatedly produced no additional preventable findings beyond a precisely identified individual error. Even then, that conclusion would need evidence rather than assumption.
Accountability determines who should bear consequences. Causal analysis asks what must change to prevent recurrence.
— Jiný Kontext