Founding Edition • Volume I • Judgment
Executive Essay No. 006
Daniel Kahneman
When Judgment Must Question Itself
On Subjective Authorship, Retrospective Confidence,
and the Effort of Examining Our Own Conclusions
Founding Edition · Volume I · Judgment · 13 pages
Opening Observation
When Judgment Must Question Itself
Executive responsibility rests on human judgment. Daniel Kahneman asks us to examine how much that judgment understands about itself.
In Thinking, Fast and Slow, intuitive impressions and deliberate reasoning are presented as interacting influences on what we believe and choose. The experience of having considered a decision does not reveal every influence that helped to form it.
For leaders, this opens a difficult inquiry. How does a conclusion come to feel like our own? What happens to our memory of uncertainty after an outcome is known? Under what conditions can reflection test a conviction rather than merely defend it?
This essay approaches those questions through Kahneman and related research on introspection and hindsight. It considers their implications for executive responsibility when analysis can increasingly be supplied by Artificial Intelligence.
The inquiry begins with a distinction: being responsible for a decision does not give us complete knowledge of how our judgment arose.
Being responsible for a decision does not give us complete knowledge of how our judgment arose.
Opening Reflection
The Feeling of Having Decided
A decision can feel entirely our own before we understand how we reached it.
We recognize the conclusion. We can explain its merits. We are prepared to accept responsibility for what follows.
Yet none of these establishes that we understand how the judgment was formed.
An impression may have arrived before the argument. A familiar explanation may have made one possibility easier to accept. Reasons assembled afterward may give the conclusion a coherence it did not possess when it first appeared.
For an executive, this creates a difficult problem.
Responsibility requires us to own our decisions. But ownership does not give us complete knowledge of their origins.
Daniel Kahneman brings this difficulty into the study of judgment. His work asks us to reconsider the confidence with which we describe our own thinking — particularly when an answer feels obvious, a story feels complete, or an outcome makes the past appear inevitable.
Ownership of a decision does not establish knowledge of how it was formed.
The Origins of Conviction
When a Conclusion Becomes Our Own
A board is considering an acquisition. Before the analysis is complete, one director finds the opportunity compelling. The market is familiar. The management team speaks with assurance. The strategic story is easy to follow.
These impressions may contain useful information. They may also give the proposal an advantage that subsequent scrutiny fails to examine.
As the discussion develops, the director offers reasons for proceeding. Some may have shaped the initial judgment. Others may have emerged while explaining it. From the quality of the explanation alone, the board cannot tell which is which.
Here, subjective authorship names an experience: the conclusion feels like something we have deliberately produced. It does not establish a complete account of how that conclusion arose.
The distinction matters because reasons have two possible roles. They can help us test a judgment. They can also help us defend one to which we have already become attached.
An executive therefore needs more than an articulate answer to the question: Why do I believe this?
A second question must remain open: What would lead me to revise it?
The first invites explanation. The second tests whether explanation has left room for correction.
A conclusion becomes more trustworthy when its reasons permit revision.
Retrospective Confidence
When the Outcome Rewrites the Past
Imagine that the acquisition succeeds. Two years later, the board returns to its original decision. The strategic fit now appears clear. The warning signs seem manageable. What once required a difficult choice begins to resemble an opportunity that good judgment was bound to recognize.
Now imagine that the same acquisition fails. The integration risks appear obvious. The optimistic assumptions seem careless. Directors ask how the warning signs could have been missed.
These are two possible histories of the same uncertain moment. Once the outcome is known, preserving that moment in its original uncertainty becomes difficult.
Retrospective confidence can quietly change the lesson an organization draws. A successful result may be treated as proof of a reliable decision process. A disappointing result may be taken to show that the error should have been apparent from the beginning.
Neither inference follows from the outcome alone.
The executive task is to distinguish what became known afterward from what could reasonably have been known beforehand.
A record made before the decision can help: the alternatives considered, the assumptions accepted, the probabilities assigned, and the evidence that would have changed the choice.
Such a record preserves something memory may struggle to retain: how much remained unresolved when action became necessary.
To learn from a decision, preserve the uncertainty in which it was made.
Metacognitive Effort
The Work of Questioning Ourselves
A leader may spend hours reconsidering a proposal and emerge more certain of the original conclusion. The time was real. The effort was real. Whether the judgment improved remains a separate question.
Reflection can reopen a decision. It can also become an extended search for reasons to preserve it.
The difficulty is that both activities may feel like careful thinking from within. We experience the effort of reasoning more directly than we can assess its independence from what we already want to believe.
Metacognitive effort means examining how we are thinking: what has captured our attention, which assumptions we have accepted, and whether our standards change when evidence becomes inconvenient.
This examination needs something beyond its own intensity. It needs points of reference that our preferred conclusion does not control.
Before reviewing a recommendation, write down the criteria it must satisfy. Seek an independent assessment before sharing your own. Compare the proposal with similar cases, including those whose outcomes were disappointing.
These practices give reflection something to encounter beyond the reasons already available to the person reflecting.
Their value lies in whether the judgment has faced a meaningful test. An assumption may change, an estimate may be revised, or the original conclusion may survive on stronger grounds.
The discipline of reflection lies in allowing our conclusions to be tested.
Judgment in the Age of AI
When an Answer Becomes a Conviction
Consider an executive who asks an AI system to assess an investment. The response arrives with a recommendation, an orderly explanation, and a persuasive account of the risks.
The executive reads it, recognizes familiar concerns, and agrees.
What has happened to the judgment?
The recommendation may have improved the decision. It may also have supplied the first interpretation around which subsequent thinking was organized. Agreement alone cannot establish how thoroughly the executive examined it.
The important moment is the passage from receiving an answer to accepting its reasons as sufficient grounds for action.
At that point, the questions developed throughout this essay return. Which assumptions have I checked? What alternatives have received independent attention? What evidence would change my position? What will I later remember as my own assessment?
A useful response can still leave essential work unfinished. The executive must connect its claims to evidence, examine what the analysis leaves unresolved, and decide whether its reasoning fits the circumstances in which action will occur.
That examination may support the recommendation. It may require a different one. Either result is compatible with responsible judgment, provided the conclusion remains open to scrutiny.
An organization can delegate parts of its analysis. It must still determine who will examine the grounds on which that analysis becomes a decision.
Before an answer becomes a basis for action, its grounds deserve examination.
The Conditions for Correction
Making Room for Revision
A leader can sincerely invite disagreement while making it difficult for anyone to disagree.
The preferred answer may already be visible. The meeting may reward speed. A previous commitment may have become so closely associated with the leader that questioning it appears to question the person.
Under these conditions, asking individuals to examine their thinking leaves part of the problem untouched. The organization must also examine what happens when someone changes a position.
Can a director revise an estimate without being treated as unreliable? Can a team report evidence against its own proposal before the project fails? Can an executive acknowledge that an earlier explanation was incomplete?
The answers determine how much correction an institution can absorb while action is still possible.
Independent assessments, recorded before discussion, preserve differences that conversation might otherwise erase. Decision records preserve assumptions that later success might conceal. Review dates create occasions to reconsider a commitment before circumstances force a retreat.
These arrangements require judgment in their design. Too much scrutiny can consume attention, delay necessary action, and obscure who must decide. The depth of examination should reflect the stakes, the uncertainty, and the difficulty of reversing the choice.
The purpose is to make consequential convictions answerable to evidence while there is still time to act on what that evidence reveals.
An institution learns when revision can occur before failure requires it.
An Enduring Contribution
The Judgment We Can Answer For
Perhaps this is Kahneman’s most enduring contribution to executive judgment: he makes the confidence of the person deciding part of what must be examined.
The inquiry does not end when a leader can explain a choice. It extends to the conditions under which the explanation was formed, the evidence it has encountered, and the possibility that conviction has outpaced understanding.
This places a demanding obligation on those who exercise authority. They must act while accepting that their account of their own thinking may be incomplete.
A conclusion can feel self-authored before its origins are understood. A result can make the past seem more predictable than it was. Reflection can strengthen an argument without exposing its weaknesses.
Together, these difficulties change what it means to stand behind a decision.
The executive can preserve the assumptions made before action, invite assessments formed independently, and establish occasions for reconsideration. These practices make the grounds of a choice available for examination beyond the moment of conviction.
Responsibility then includes a willingness to explain what was believed, what remained uncertain, and what would justify a different course.
Artificial Intelligence brings new forms of assistance into this work. The obligation to examine the grounds of action remains with those entrusted to decide.
The aim is a confidence proportionate to its evidence, and an authority capable of revision when that evidence changes.
To stand behind a decision is also to remain answerable for revising it.
Closing Thought
Closing Thought
There is a moment in every consequential decision when examination must give way to action.
The evidence remains incomplete. The alternatives carry different risks. Further reflection may help, but delay has consequences of its own.
The leader must decide.
Nothing in this essay removes that obligation. Questioning our judgment has a purpose: to improve the grounds on which we act and preserve our ability to respond when those grounds change.
We may never fully reconstruct how a conviction began. We can still examine the evidence it rests upon.
We may struggle to remember how uncertain the past once felt. We can still preserve a record before the outcome is known.
We may find that reflection returns us to the same conclusion. We can still ask whether it encountered anything capable of changing our minds.
Such disciplines leave room for confidence. They give that confidence something more durable than the feeling of being right.
A decision must eventually be made. The willingness to examine it can remain.
We must decide before we can know everything. We need not stop questioning once we have decided.
Questions for Leaders
Executive Reflections
A few questions to guide reflection, discussion and decision in your organization.
How did this conclusion become our own? Which reasons shaped our initial judgment, and which emerged only when we began to explain it?
What did we believe before we knew the outcome? Have we preserved our original assumptions, alternatives and uncertainties, or are we reconstructing them in hindsight?
What could our reflection actually change? Have we sought evidence and independent assessments capable of altering the conclusion we prefer?
What have we examined before accepting the answer? When a recommendation comes from AI or an adviser, which claims have we verified and which uncertainties remain?
Can we revise a judgment without losing authority? Do our meetings, incentives and review practices allow people to change course before failure makes revision unavoidable?
The quality of reflection depends on whether our conclusions remain open to correction.
Sources and Editorial Attribution
Selected References
Primary Works
Kahneman, Daniel. Thinking, Fast and Slow. New York: Farrar, Straus and Giroux, 2011.
Kahneman, Daniel, and Gary Klein. “Conditions for Intuitive Expertise: A Failure to Disagree.” American Psychologist 64, no. 6 (2009): 515–526.
Related Foundational Research
Fischhoff, Baruch. “Hindsight ≠ Foresight: The Effect of Outcome Knowledge on Judgment Under Uncertainty.” Journal of Experimental Psychology: Human Perception and Performance 1, no. 3 (1975): 288–299.
Nisbett, Richard E., and Timothy D. Wilson. “Telling More Than We Can Know: Verbal Reports on Mental Processes.” Psychological Review 84, no. 3 (1977): 231–259.
Editorial Note
This essay offers an independent interpretation of Kahneman alongside related research on hindsight and introspection. “Subjective authorship” is an editorial formulation. Applications to AI and executive governance, practical recommendations, and closing maxims are editorial contributions, not a framework or quotations attributed to Kahneman. The retrospective and introspective arguments also draw on Fischhoff and on Nisbett and Wilson, respectively.
Executive Essay No. 006 · Daniel Kahneman
Colophon
- Title
- When Judgment Must Question Itself
- Subtitle
- On Subjective Authorship, Retrospective Confidence, and the Effort of Examining Our Own Conclusions
- Intellectual Subject
- Daniel Kahneman
- Authorship
- An Executive Essay by The Executive Library
- Series
- The Executive Essay™
- Number
- No. 006
- Edition
- Founding Edition – Volume I: Judgment
- Edition Year
- 2026
Confidence accompanies a decision.
Judgment remains willing to examine it.
Founding Edition · Volume I · Judgment · 13 pages
Timeless Principles for Modern Leadership