Why Asking an Expert “Why?” Is Not Enough

A consequential decision often becomes easier to explain after it has worked. The uncertainty drains out of the story. Details that once competed for attention are rearranged around the eventual result, and the expert can offer a principle that sounds as though it had been present, fully formed, from the beginning.

That account may be honest. It may also be incomplete in precisely the places a successor needs most: the first weak signal, the interpretation that was briefly plausible, the evidence that changed the call, or the threshold beyond which waiting became irresponsible. Memory is not a transcript, and expertise does not make it one.

The buyer’s question is therefore narrower than “Can this leader explain the role?” It is whether the material being preserved has enough evidentiary support to guide another person when the next case is similar but not identical.

Nisbett and Wilson’s classic review argued that people may have limited direct access to the higher-order mental processes that produce some judgments. When asked why, a person can report a plausible theory based on what seems likely rather than a complete account of what actually caused the response. The paper does not show that every verbal report is useless. It does show why confidence and coherence cannot be the only tests.

This matters in succession because senior people are practiced narrators of their work. They have board language, operating principles, and years of hindsight. The same fluency that makes an interview compelling can smooth over the moment when the situation stopped fitting the ordinary rule.

Skagway’s practitioner conclusion is modest: self-explanation should carry an evidence label. Recollection, interpretation, and inference each deserve a place in the record, but they should not be silently promoted into objective causal fact.

Klein, Calderwood, and MacGregor developed the Critical Decision Method for naturalistic work marked by pressure, dense information, and changing conditions. The method extends critical-incident interviewing with probes designed to elicit perceptual and conceptual discriminations, typicality judgments, and critical cues. Its practical power begins with a specific event rather than a request for a general philosophy.

A useful interview first builds the timeline. What was happening before the decision became visible? When did the incumbent’s reading change? What did they expect to happen next? Which information did they seek, ignore, or distrust? Only after the event has regained some of its original texture does the inquiry turn toward interpretation.

This does not solve the introspection problem. It gives the account constraints. Dates, artifacts, people, and changing expectations make it harder for a polished generality to stand in for the actual episode.

“Why?” invites a reason. A cognitive probe can ask something more discriminating: What did you notice that a less-experienced person might have missed? What would you have expected to see if your first interpretation were correct? Which single change in the facts would have reversed your decision?

Counterfactual and contrast questions are especially useful because they put pressure on the edges of a rule. If a leader says, “I always protect the customer relationship,” a contrasting incident may reveal that protection sometimes means conceding, sometimes refusing, and sometimes speaking privately before any formal action. The transferable material is found in the conditions separating those responses.

Applied Cognitive Task Analysis offers related tools for identifying difficult cognitive demands, potential errors, cues, and strategies, and for translating findings into training or design products. Neither CDM nor ACTA was validated as an end-to-end method for executive succession. Their value here is intellectual and practical lineage, not borrowed certification.

The incumbent’s account should be compared, where access and confidentiality permit, with contemporaneous records, decision artifacts, and other people who saw the event from a different position. The goal is not to stage a trial around an old decision. It is to see where independent traces support the account, where they add missing context, and where the organization should retain uncertainty.

Contradiction can be productive. A finance leader may remember that the decisive issue was covenant risk while an operating record shows customer concentration became visible first. Both can be true at different points in the timeline. Forcing a single clean reason would make the case easier to summarize and less useful to learn from.

A disciplined record can distinguish recollection, contemporaneous evidence, inference, corroboration, and tested prediction. Those labels tell a successor how much weight to place on the material and which claims still need to be challenged in practice.

A historical case can be richly documented and still fail to generalize. The next test is whether its distinctions survive variation. Change the stakeholder, shorten the time available, remove a familiar information source, or combine two conditions that did not occur together in the original event. Then ask the incumbent and successor to reason through the altered case separately.

Agreement is not the only useful outcome. A successor may reach a different decision for sound reasons because strategy, authority, or market conditions have changed. What matters is whether they recognize the consequential cues, make the tradeoff visible, and know where uncertainty warrants escalation.

Neither a transcript nor a scenario result proves that judgment has been transferred. Together they create better evidence than explanation alone: an inspectable account of what the organization believes, what supports it, and where dependency remains.

Some executive judgment will resist clean representation. Timing may depend on embodied experience; relationship history may change what can be said; part of the apparent expertise may reside in colleagues, routines, and artifacts around the incumbent. A method gains credibility by naming those limits.

The Map is useful when a real critical-person dependency justifies structured elicitation and triangulation before a leader leaves. It can surface consequential incidents, cues, exceptions, relationships, and thresholds, then organize them for successor practice or institutional use. It does not make an interview causally infallible or turn a human being into a complete model.

For boards, that restraint is practical. A partial map whose evidence and uncertainty are visible is safer than a confident document that cannot tell the reader how its conclusions were reached.