Why Do We Often Find Reasons Only After We Have Chosen?

How a Person Builds a World · Season Three: Attention, Desire, and Self-Understanding · Article 5

Ask someone why they chose this garment, this candidate or this career, and they can usually provide reasons. The colour was more harmonious, the policy more credible, the work more consistent with long-term aims. The reasons may be complete and sincerely expressed, even describing a process of trade-off. But does the ability to state reasons show that those reasons, in the same form, drove the choice before it occurred?

In their classic 1977 paper “Telling More Than We Can Know”, Richard Nisbett and Timothy Wilson argued that people lack direct introspective access to some higher-order cognitive processes, yet explain their responses by drawing on plausible theories about what normally causes what. The paper initiated a long debate and should not be reduced to the slogan that all reasons are fabricated. It makes a more careful distinction: the capacity to explain oneself is not identical to the capacity to observe the process that actually generated a decision.

Choice blindness makes the reliability of reasons visible

In a 2005 choice-blindness experiment, Petter Johansson and colleagues asked participants to choose which of two photographed faces they found more attractive. Through a sleight-of-hand procedure, the researchers sometimes handed over the photograph that had not been selected and asked the participant to explain the choice. Some substitutions were not detected, and participants nevertheless produced specific explanations for the photograph now in front of them.

The task was narrow: it concerned rapid facial preference, an experimental arrangement and particular samples. It does not establish that people are equally blind in consequential decisions, and participants did detect some swaps. The experiment matters because it created an unusual condition in which the researchers knew that the outcome before the participant was not the original choice, while the participant’s language could still build a coherent rationale around it. The capacity to generate reasons could rapidly serve the fact that had already been presented.

There is no experimental magician switching cards in ordinary life, but outcomes often precede complete explanations. We accept an invitation intuitively and later say it suited our development. We reject an opportunity while exhausted and later describe the refusal as a principled choice. One path happens to become available; after walking it for a while, we write it into a long-standing ambition. The explanation need not be a lie. A speaker can sincerely mistake an order now visible for the original cause.

Reasons perform three different jobs

Confusion arises because we use reason for at least three things.

The first is causal explanation: which processes actually contributed to the choice? These may include the sequence of attention, bodily state, habit, situational cues, social influence and preferences that were never clearly articulated. The second is an agent’s reason: which considerations were understood and carried weight at the time? The third is a justificatory reason: regardless of how the choice arose, can it now be defended by standards that other people can examine?

The three may coincide or separate. Someone selects a political candidate through familiarity but sincerely believes policy was decisive; causal explanation and the agent’s stated reason diverge. A decision begins in accident and later produces a relationship worth maintaining; its origin supplies no justification, but a subsequent commitment may. Conversely, a person’s ability to state noble reasons does not establish that those reasons outweigh the harm actually caused.

Discovering post-hoc explanation therefore does not automatically destroy a choice’s significance. It removes one excess confidence: because I can now tell a coherent story, I must already know everything that caused me to act then.

Self-narrative protects a life that has already happened

People need to organise continuous action into an intelligible life. If every choice appeared only as an accidental impulse, identity and responsibility would be difficult to sustain. Retrospective explanation performs constructive work. It places events within a longer duration and says which choices should be continued, which errors repaired, and which chance encounters have become commitments.

The problem begins when narrative permits only one direction. An established choice changes the evidence available. Once we take a job, we become more familiar with its advantages. Once a relationship continues, shared memories accumulate. Once resources have been invested in a path, exit becomes more expensive. Genuine value formed later mixes with the original causes, and “I now choose to continue” is written backwards as “I knew clearly from the beginning”.

That backward writing may protect self-respect, sunk investment, or simply the coherence through which memory reconstructs the past. It can lead us to disparage roads not taken, describe uncertainty as conviction, and retell constrained circumstances as unrestricted autonomy. An excessively tidy life story is rarely evidence that life itself was tidy. It shows that the story is doing work to maintain consistency.

Reflection needs traces from before the choice

When an important decision is reviewed only after its outcome, it becomes difficult to distinguish the earlier judgement from the later explanation. One modest practice is to leave a limited record before choosing: which options do I presently see; which three conditions matter most; what do I fear; which facts remain uncertain; what new information would change the decision?

The purpose is neither perfect auditing nor converting life into a spreadsheet. A prior record allows the past self to exist without being represented entirely by the present one. Months later, it may show that earlier reasons remain sound, or that a decisive condition was never considered. Either discovery contributes more to knowledge than “I have always thought this”.

Other people can help, but the form of the question matters. “Why did you do it?” invites a complete explanation. “What did you notice first?”, “What alternatives were present then?”, or “If the outcome had been different, how might you describe the same choice?” may reveal more about formation. This is especially important for organisations. Writing formal reasons before a decision and preserving objections and uncertainties can prevent outcomes from retrospectively cleansing the decision record.

We can also observe growth in certainty within our own language. Before choosing, “I am probably leaning this way” becomes, after success, “I always knew”. Luck acknowledged beforehand disappears from the successful account, leaving judgement alone. Such movement is not proof of wrongdoing, but it is a reason to re-examine the narrative.

Honesty does not eliminate retrospective reasons; it dates them

No person can know themselves fully in advance of action. Many reasons reveal their weight only in consequences. Someone becomes a carer and only then understands why care matters; begins a kind of work and discovers its connection to capability and value; loses a way of life and finally sees what it supported. A reason formed later is not invalid merely because it is later.

What matters is whether time layers remain honest. A person can say, “Familiarity and opportunity moved me then. Only later did I recognise the public value of the work, and that is why I choose to continue now.” This is more accurate than “I did it for public value from the start”, and it shows more clearly how a person changed.

A testable self-explanation accommodates four possibilities: I may not know every cause; my present reasons may be affected by the outcome; value acquired afterwards may be genuine; and new evidence may require withdrawal of my earlier defence. None of this removes responsibility. Responsibility no longer depends on fictional self-transparency. It depends on the ability to revise explanation and subsequent action.

We often search for reasons after choosing because action can be generated faster than language can explain it, and because a person needs to organise what has happened into a world they can continue to inhabit. The danger is not that an explanation came later. It is that clarity acquired later is passed off as knowledge possessed from the beginning.

Mature self-understanding does not demand a faultless internal recording of every choice. It asks us to distinguish what may have caused the action, what we actually considered then, why we choose to continue now, and which part of the story must remain vulnerable to facts that could overturn it.

Primary sources and further reading

• Richard E. Nisbett and Timothy D. Wilson, “Telling More Than We Can Know: Verbal Reports on Mental Processes”, Psychological Review, 1977.
• Petter Johansson et al., “Failure to Detect Mismatches Between Intention and Outcome in a Simple Decision Task”, Science, 2005.
• Lars Hall, Petter Johansson and Thomas Strandberg, research and review on choice blindness and attitudes, 2017.

Continue reading: Explore the How a Person Builds a World series.