AN INTERACTIVE EXHIBITION BY SAHIL TAYAL
The
Identity
Laboratory.
You know what matters to you.
So why doesn't that always become what you do?
Walk through six small experiments about intention, friction, environment and the stories we tell about ourselves. There are no personality types to unlock. No score for being a better human. Just a more interesting question.
A failed action is not automatically a failed person.
Someone sincerely values reading and finishes no books this month. A colleague believes in speaking openly and says nothing in a meeting. A friend intends to call and doesn't. The usual explanation arrives quickly: discipline. This exhibition asks us to delay that verdict. Which action, in which situation, amid which demands, with what information and opportunity?
The Identity–Action Gap is Sahil Tayal's proposed framing for a familiar puzzle: the distance between self-understanding and observable behaviour. The broader intention–behaviour gap has a substantial research literature. The particular formulation used here is not a validated clinical scale or predictive theory. Read the original inquiry ↗
IDENTITY × BEHAVIOUR
The label
is not the act.
“I am a creative person.” That may be true, meaningful and sincerely held. Yet the sentence does not specify what will happen on a crowded Tuesday at 9:40 pm.
Identity helps organise attention: some futures feel like mine; some choices feel worth defending. But an identity is a description at a different level of analysis from the action it may inspire. “I care about my health” is a broad orientation. “I will walk ten minutes after lunch today” is an observable act with a time, place and competing demands.
That distinction matters because retrospective stories are seductive. When an action succeeds, we may say, “That is who I am.” When it fails, we may say, “I suppose I never really cared.” Both conclusions can be too quick. A change in conduct might instead reflect fatigue, uncertainty, a broken process, conflicting duties—or a thoughtful revision of priorities.
Identity is not irrelevant. People can be motivated by self-congruent goals. The caution is against treating self-description as an event-producing mechanism all by itself.
Think of identity as the compass, not the vehicle. A compass may point north while the path is flooded, the bridge is closed and the traveller has good reasons to remain where they are.
Zoom in on a statement.
Pick an identity, then an action. Watch how the second introduces real-world conditions.
Now ask: is there a seat, enough light, and a book within reach?
A TRANSPARENT TOY MODEL
The slope
of the path.
A two-minute task can feel enormous when it contains an uncertain first move, fear of embarrassment, a hard deadline, or no safe opportunity to act.
“Friction” is a useful working label for what stands between contemplating an action and carrying it out. Sometimes it is physical—distance, inaccessible transport, money. Sometimes cognitive—not knowing which document to open. Sometimes social—a manager who invites dissent and punishes it. These should not be collapsed into a moral verdict.
Below is an intentionally invented numerical demonstration, not a validated formulation of the Identity–Action Gap. Move four sliders to see how a mathematical metaphor responds. Notice how a tiny change in friction can interact with opportunity and cue clarity. The bars show a toy index, never your probability of taking action.
Why use any numbers at all? Because transparent assumptions invite criticism. Hidden assumptions dressed up as a test do not. This exhibit makes the arithmetic visible so that you can ask: Should these factors be weighted equally? Is “energy” the right construct? What is missing?
Real behaviour is not reducible to four sliders. Identity, environment, habit, changing priorities, social power and unforeseeable events cannot be safely inferred from a decorative index.
Change the conditions.
All inputs are hypothetical, scaled 0–10. They describe a situation—not a person.
In this illustration, the conditions are mixed; reducing a barrier may be more useful than demanding greater motivation.
SHOW THE EQUATION & ASSUMPTIONS +
Start with a constant −0.6. Add 0.24 × opportunity, 0.19 × cue clarity, 0.17 × capacity; subtract 0.28 × friction. Pass the result through 100 / (1 + e^(−z)), then round. If an absolute barrier is ticked, set the display to zero. These weights and the logistic shape are invented for learning, not fitted to data.
Current z = −0.6 + 0.24(5) + 0.19(3) + 0.17(5) − 0.28(6) = 0.34.
ENVIRONMENT × POWER
Same person.
Different room.
We praise courage in one environment and penalise it in another. Opportunity isn't only the presence of a door. It's whether someone can open it without paying an unreasonable price.
Consider an employee who wants to raise a safety concern. In one workplace, concerns are recorded, supervisors respond and retaliation is forbidden and actually investigated. In another, dissent is formally welcomed but informally punished. A silent employee does not prove apathy: that conclusion ignores the very incentives their organisation created.
Psychological capacity and individual motivation matter. So do authority, childcare, access, incentives and social norms. The COM-B framework (capability, opportunity, motivation → behaviour) is a useful established alternative to explanations that live entirely inside an individual's head. Its constructs do not mean every influence is easily measurable.
A second trap lurks here: blaming the system for every outcome can hide choice, while blaming the person for every outcome can hide constraint. Good analysis locates the boundary rather than pretending it never moves.
Making conduct easier is not automatically good. Friction sometimes protects privacy, consent and reflection. Ask whose behaviour is being changed, for whose benefit, and with whose permission.
Switch the organisation, not the worker.
Alex has the same concern in both scenes. Select a workplace to reveal why the silence might mean something different.
Speaking is relatively safer.
Alex still needs to decide what to say and when. But a functioning process lowers the expected interpersonal cost. Silence might reflect uncertainty, timing or competing priorities—not necessarily fear of punishment.
Try the other room. Does your judgment of Alex change even though Alex doesn't?
SALIENCE × IMPLEMENTATION INTENTIONS
Give tomorrow
a handle.
“I'll do it later” contains a destination but no departure time. A specific cue can help an intended action meet the actual day.
Psychologist Peter Gollwitzer's work on implementation intentions studies plans of the form if situation X occurs, then I will do Y. The strategy makes a relevant opportunity easier to recognise and can help bridge some intention–action gaps. It does not guarantee execution, solve material barriers or make every goal worth pursuing.
A cue needs to be recognisable. “Whenever I become my best self” is grand but hard to spot. “After I put the kettle on” is mundane and specific. The response should be within reach. If the action requires a quiet room you don't have, a perfect sentence won't manufacture one.
The Fogg Behavior Model similarly emphasises the convergence of motivation, ability and a prompt at the moment of behaviour. These are distinct research traditions, not proof that the proposed Identity–Action Gap has been experimentally confirmed. Their shared usefulness here is practical: the moment matters.
Planning can itself become sophisticated procrastination. A plan is a cue to attempt something, not a substitute for doing it. And sometimes an unexpected interruption makes the original plan inappropriate.
Make a plan the calendar can recognise.
If I finish breakfast, then I will read two pages of a book.
Works best when the cue occurs and the action is genuinely possible.REPETITION × STABILITY
The room
starts remembering.
Repetition in a stable context may gradually make a response easier to retrieve. That is different from being motivated every single morning.
Habit research distinguishes deliberate intentions from patterns triggered by recurring contexts. Wendy Wood and colleagues describe how repeated responses in familiar situations can become less dependent on moment-to-moment deliberation. The ordinary environment begins to carry part of the behavioural load.
But the same mechanism can preserve unwanted patterns. Returning home exhausted and opening an app without deciding to do so is not evidence that one truly wanted the app more than a meaningful conversation. The cue was familiar and the response rehearsed. To change habits, it can help to change the cue or friction around the response—not just deliver a louder pep talk.
Below, four fictional mornings show why consistency of context matters. This is not a habit-formation clock. Four repetitions do not establish automaticity; no universal “days to form a habit” promise is made. The illustrated memory strands merely track how many times our fictional gardener encountered the same cue and chose the small act.
Do not confuse rehearsing an action in a stable setting with creating an automatic habit. The latter needs longitudinal evidence, and individuals and behaviours vary.
Choose what happens each day.
Fictional gardener Mina wants to sketch one leaf after morning tea. For each day, choose the situation and whether she sketches.
Mina repeated the action three times in the same setting. That is a promising practice pattern in this story, not evidence that a habit has formed.
OBSERVATION × INTERPRETATION
Do not mistake
the verdict
for the data.
When intention and action diverge, we love explaining why. But explanation requires evidence. Otherwise a story about character can harden around one missed act.
Imagine someone who promised to send a draft on Friday but did not. The observation is narrow: the draft was not sent by the agreed time. “They do not care” is one interpretation. “They were overloaded,” “They misunderstood the deadline,” and “They intentionally deprioritised it” are others. Those hypotheses imply different responses—and each needs more information.
Feedback can help us recalibrate, but only when we understand what it measures. A missed attempt may show that the cue was weak, the cost higher than expected, or the goal no longer deserved attention. It may also show avoidable neglect. Ethical inquiry leaves room for accountability without pretending to read minds.
The final exercise does not pronounce anyone innocent or guilty. It asks you to practise separating the known, the possible and the next reasonable question. The goal is not to excuse every discrepancy, but to replace overconfident storytelling with testable curiosity.
In situations involving repeated harm or broken commitments, the observable pattern and its impact matter regardless of unverified motives. Curiosity about causes must not erase boundaries or responsibility.
What do we actually know?
A teammate, Noor, did not send the agreed draft by Friday evening. Sort each statement by its epistemic status.
“The agreed draft did not arrive by Friday evening.”
Perhaps the
question was
never
“Why am I like this?”
Perhaps a better question is: What does this action demand, in this situation, from this person, and what could responsibly change?
Aspirations are not the same as observable behaviour.
Cost and uncertainty are not moral attributes.
Resources, power and permission change what's possible.
A moment to act can be more useful than a grand intention.
Repetition can make an environment a behavioural ally—or trap.
Separate what happened from your theory of why.
One question to take outside.
What is one action whose conditions you would like to understand better? This note stays on your device until you copy it; it is not submitted, recorded or saved here.
THE LAB HAS WINDOWS
Where the
ideas come from.
The sources below support particular background ideas. They do not validate Sahil Tayal's proposed Identity–Action Gap model or the site's invented toy equation. All source titles link to the originating academic publication or research centre.
The Intention–Behavior Gap Sheeran & Webb, 2016
A review of why intentions do not always lead to behaviour, and approaches for understanding that discrepancy.
Fogg Behavior Model Stanford Behavior Design Lab
Frames action in terms of motivation, ability and a prompt converging in a given moment. The lab's equation is not Fogg's model.
Implementation Intentions Gollwitzer, 1999
Foundational work on if–then plans for translating goal intentions into situation-linked responses.
The Behaviour Change Wheel Michie, van Stralen & West, 2011
Introduces the COM-B framework: capability, opportunity and motivation as conditions of behaviour.
Psychology of Habit Wood & Rünger, 2016
Reviews how repeated action in stable contexts can establish cue–response tendencies that interact with goals.
Habits, Goals, and Effective Behavior Change Wood, 2024
Discusses interventions addressing cues, rewards and friction rather than relying on good intentions alone.
Editorial note: Scenarios involving Alex, Mina and Noor are fictional teaching devices. Interface indices, coefficients and labels were made for this exhibition; they are not empirical findings, estimates, diagnostics or advice. Accessible links were verified when this edition was prepared; academic access may vary.
OTHER DOORS IN THE SAHILVERSE
The question
doesn't end here.
The Identity–Action Gap
The proposed framing behind this exhibition, with its distinctions and limitations.
READ THE INQUIRY ↗COMPANION / 02Agency Lab
Choice, power, responsibility and the actual conditions in which decisions occur.
EXPLORE AGENCY ↗COMPANION / 03Still Becoming
On revising beliefs without treating every change as automatic growth.
FOLLOW THE THREAD ↗