RESEARCH-INFORMEDPROVISIONAL WHERE LABELLEDPUBLICLY REVISABLE
FIG. 01 / A QUESTION IN MOTION● SIGNAL
01QUESTIONWhat happened?
02FRAMINGWhat are we asking?
03EVIDENCEWhat do we know?
04ALTERNATIVESWhat else fits?
05REVISIONWhat now holds?
NO FINAL ANSWER REQUIRED.REVISED, NOT ERASED ↶
Welcome to the notebook.
This is an interactive field guide, not a test of intelligence. You can read every section and open the inquiry cards without JavaScript. The thought lab’s example is written out below as a fallback. Research is marked separately from proposed models and creative experiments.
✳CHOOSE YOUR READING DEPTH
Both versions share the same evidence.
THE CENTRAL DISTINCTION
A question is not an identity badge.
I move between organisational behaviour, human development, psychology, small retail, fragrance and creative technology. These interests are not evidence that I understand them equally well. They are different places where the same habit becomes useful: distinguish the thing I noticed from the explanation I want to believe.
Formal research, personal essays, early conceptual models and playful browser experiments should not be mistaken for one another. A proposal is allowed to be unfinished. An artistic simulation can illuminate a trade-off without proving that real people behave like its characters. A published paper deserves to be read for what it actually tested—not what its title makes convenient to claim.
I write this page as a record of method rather than an account of intellectual superiority. In a shop, one missing item might reflect demand, an ordering constraint or a supplier delay. At work, one silent meeting may reflect disengagement, caution or a poorly designed invitation to participate. In a relationship, one unanswered message is an event, not a complete biography of someone else.
None of this means postponing judgment forever. Decisions still need to be made. Better thinking is the ability to reach a proportionate conclusion while keeping the door open for the world to surprise us.
01Established researchPublished, bounded findings
02Working proposalsInteresting; not yet validated
03Public essaysArguments open to disagreement
04Creative experimentsArt and simulation, not data
01 / THE METHOD
Five moves of a question.
Not a proprietary scientific framework. A portable editorial checklist for turning a first impression into a more defensible view.
01 ↗
Notice
Begin with a trace, not a verdict.
A colleague goes quiet in a meeting. That silence is observable; “they do not care” is a story about it.
Write a sentence that a camera or transcript could confirm. Then write what you felt, separately. Keeping both matters: emotion is information about your experience, not an automatic report on somebody else’s mind.
IN A SMALL SHOP
A shopkeeper sees customers buying single units rather than larger packs. That is a transaction pattern, not proof that buyers lack planning or money. Stock availability, household needs, freshness, storage, and cash flow are rival possibilities.
02 ↗
Define
Turn a large word into a question you could investigate.
“Motivation is low” is vague. “How often did teammates receive a clear next action and time to do it?” is answerable.
Ask what counts as an instance, over what period, and for whom. If you cannot describe what would change your mind, you may be defending a mood rather than testing a claim. The point is not to strip life of nuance; it is to avoid smuggling a conclusion into an undefined term.
IN A SMALL SHOP
“The shop is doing well” could refer to revenue, cash collected, gross margin, net income, repeat custom or human sustainability. These are not interchangeable.
03 ↗
Separate
Observation, interpretation and value judgment are different sentences.
“Three people missed the optional session” is observation. “They are disengaged” is an inference. “They should be disciplined” is a value judgment.
Name the kind of sentence before acting on it. Moral evaluation is sometimes necessary, but it should not arrive disguised as a measurement. You can care deeply about an issue and still keep your evidentiary ledger tidy.
IN A SMALL SHOP
“A customer requested credit” is observable. “They are unreliable” requires further evidence. “Never extend credit” is a policy choice that also has consequences.
04 ↗
Compete
Give a rival explanation enough strength to embarrass your favourite one.
A training programme fails. Perhaps its content was weak; perhaps staff lacked time, permission, safety, or an opportunity to use it.
The test is not “Can I imagine another cause?” but “What observations would differ if the other cause were true?” Seek awkward cases, negative cases, and the pattern you hoped not to find. A rival that explains every possible result is not much of a rival.
IN A SMALL SHOP
Sales fell after a display change. Price, season, availability and neighbourhood traffic may have changed too. One shop and one week rarely isolate causation.
05 ↗
Revise
Replace certainty with a better-sized claim, then act proportionately.
Instead of “People resist change,” try “In this team, adoption stalled when the new workflow increased effort without changing incentives.”
Revision is not indecision. The aim is a conclusion strong enough to guide a reasonable next move, modest enough to survive new facts, and explicit about where it stops. Sometimes evidence supports a firm judgment. Sometimes the ethical action is to ask one more question.
IN A SMALL SHOP
“Our sample suggests…” is more honest than “Customers always…” A local observation can inspire a wider study; it cannot stand in for one.
THE TINY RULE
Before you say “this proves it,” ask: what else would make the same thing happen?
Eight puzzles. Each has a tempting story, a plausible competitor and a test worth imagining. Nothing here is a universal personality verdict.
SELECT A CARD TO OPEN THE EVIDENCE LEDGER ↓
01 / BEHAVIOUR / PROVISIONAL MODELWhy does an identity fail to become an action?Someone sincerely values learning, yet does not study. Is that hypocrisy—or is the question badly framed?Open the inquiry ↗
Competing explanations
The aspiration is weak or borrowed; the action is too costly; the prompt never arrives; the environment punishes attempts; the person has reconsidered the goal.
Evidence that would help
Follow specific intentions and actual opportunities across days. Record constraints before each opportunity, not only after failure. Compare a proposed “identity–action friction” account against established intention and behaviour models.
What remains uncertain
The Identity–Action Gap is Sahil’s proposed framing, not a validated psychological law. Its distinctive explanatory value remains to be demonstrated.
02 / WORK / INSTITUTIONSWhy might an organisation underdevelop its people?A company celebrates “talent” while giving employees little time to learn. Is the contradiction cultural, economic or structural?Open the inquiry ↗
Competing explanations
Short-term incentives reward throughput; managers cannot release staff; training is poorly designed; learning creates mobility risks; employees may want different kinds of development.
Evidence that would help
Examine budgets, actual protected learning time, manager incentives, usage of acquired skills and employee choice. Compare units with different institutional arrangements while controlling for selection effects.
What remains uncertain
The value of development is not in doubt as a human concern, but the best intervention depends on sector, resources, and who bears the costs.
03 / EDUCATION / MEASUREMENTWhat do school rankings leave unmeasured?An exam rank tells us something. When do we mistakenly ask it to tell us everything?Open the inquiry ↗
Competing explanations
Rankings coordinate selection under scarcity; they may track a narrow kind of mastery; they also reflect support, opportunity and test conditions; some valuable learning resists easy scoring.
Evidence that would help
Specify the intended outcome—content mastery, transfer, curiosity, wellbeing, collaboration. Compare multiple measures over time, accounting for prior opportunity and selection.
What remains uncertain
The problem is not that every ranking is meaningless. It is that a useful number can be promoted into an unwarranted verdict on a person.
04 / RELATIONSHIPS / INTERPRETATIONWhen does reciprocity become a story we invent?A message goes unanswered. Is there a pattern, a constraint, a boundary—or not enough evidence?Open the inquiry ↗
Competing explanations
The person is busy; the priority differs; the channel failed; the relationship is asymmetric; the silence expresses a boundary. Multiple explanations can coexist.
Evidence that would help
Look for observable behaviour over time and explicit communication where welcome. Examine your own choices and expectations; respect that other people need not disclose every motive.
What remains uncertain
You can judge whether a pattern suits your needs without claiming privileged access to another person’s mind.
05 / SCENT / EXPERIENCECan a fragrance carry an emotion?A smell feels nostalgic. Did the molecules contain nostalgia—or did a person and a memory meet them halfway?Open the inquiry ↗
Competing explanations
Personal association, cultural learning, context, perceptual qualities, expectancy and repeated exposure may each affect a response.
Evidence that would help
Ask participants about their prior associations, compare blinded and labelled samples, and separate immediate ratings from later recollection. Do not infer a universal emotion from one reaction.
What remains uncertain
Fragrance can be a compelling artistic medium without becoming a reliable technology for inducing identical moods in everyone.
06 / TECHNOLOGY / AUTHORSHIPWhat remains human when AI helps make the work?A prototype appears in hours rather than weeks. Did the tool do the thinking—or change where thinking happens?Open the inquiry ↗
Competing explanations
AI lowers execution costs; it can accelerate weak ideas too; creative direction, selection, correction, accountability and taste may still be substantial human contributions.
Evidence that would help
Document prompts, revisions, discarded alternatives, external testing and failures. Evaluate the finished artefact independently of how dramatic the production story sounds.
What remains uncertain
Speed is not quality, and volume is not originality. Authorship and responsibility need more than a count of outputs.
07 / FIELDWORK / SMALL RETAILWhat can a neighbourhood shop teach us?A tiny commercial decision changes cash, trust and inventory at once. Which consequences are visible—and which are not?Open the inquiry ↗
Competing explanations
A choice may make sense for liquidity but reduce goodwill, or favour a regular customer while making stock planning harder. No single lens captures it all.
Evidence that would help
Keep financial events and qualitative notes separate. Compare repeated, documented encounters; distinguish actual field observations from fictional model assumptions.
What remains uncertain
A single shop can generate hypotheses, not population-level conclusions. The linked observatory is an educational simulation, not real transaction data.
08 / AGENCY / STRUCTUREWhere does individual responsibility actually begin?Two people receive the same advice to “take initiative.” Do they have the same meaningful choices?Open the inquiry ↗
Competing explanations
Skill and belief matter; so do authority, resources, personal duties, institutional incentives and consequences of dissent.
Evidence that would help
Compare what options were realistically available, which supports were provided and what happened after similar actions in different environments.
What remains uncertain
Recognising constraints need not erase responsibility. Claiming responsibility need not erase constraints. Neither extreme resolves the case.
✳
These cards are an editorial map, not eight completed scientific studies. The links lead to distinct, publicly available works; their evidence status is stated in each card.
03 / INTERACTIVE FIELD LAB
Change my mind.
Choose an everyday claim. Move through the evidence slowly. The point is not to win an argument—it is to find a sentence the available evidence can actually support.
THE THINKING ROOM / EXERCISE 001NO SCORE · NO ACCOUNT · NO PROFILE
01 OBSERVE02 INFER03 COMPETE04 TEST05 REVISE
01 / OBSERVATION
What could a witness actually confirm?
In four weekly meetings, two employees proposed new tasks. The others did not volunteer proposals during those meetings.
1 of 5
A complete worked example: “Nobody on this team takes initiative.”
Observe: In four weekly meetings, two employees offered new tasks; others did not volunteer.
Infer: The manager believes the silent employees lack initiative.
Compete: They may have proposed ideas elsewhere, lacked permission, or learned that speaking up is penalised.
Test: Review proposal channels, follow-through and rewards. Ask privately what made participation easy or risky.
Revise: “Participation in these meetings was concentrated among two employees; we have not established why.”
An educational reasoning exercise, not a psychometric instrument, clinical assessment or a prediction about you. Selections stay in this page; nothing is submitted.
INTELLECTUAL HUMILITY, WITHOUT THE THEATRE
An unfinished idea can still be useful.
These are illustrative shifts in reasoning, not claims that I previously believed each first sentence. They show how to weaken an overconfident claim without draining it of meaning. If a view cannot be corrected, it is no longer functioning as an inquiry.
THE TEMPTATION
“People just need more motivation.”
THE MORE CAREFUL VERSION
An intention can be genuine while the action remains infeasible, badly cued or socially costly. Ask what the action demanded at the actual moment.
Counter-test: find cases where motivation stays similar but access, prompts or time change.
THE TEMPTATION
“If a metric exists, it must capture what matters.”
THE MORE CAREFUL VERSION
Metrics can guide decisions and still miss relationships, hidden labour, unobserved opportunities and long-term development. Keep the purpose of measurement explicit.
Counter-test: compare a score to independent outcomes it is supposed to represent.
THE TEMPTATION
“A consistent story proves a cause.”
THE MORE CAREFUL VERSION
Several mechanisms can predict the same observation. A polished narrative is not a natural experiment. Identify a discriminating test before becoming attached.
Counter-test: propose an inconvenient explanation and write its predicted evidence.
THE TEMPTATION
“A creative tool either replaces or liberates the thinker.”
THE MORE CAREFUL VERSION
Both narratives are too total. Tools can change the distribution of effort, enable experiments and introduce new errors. Judge a specific workflow and result.
Counter-test: inspect what a tool made easier, what it concealed and what the creator corrected.
When should we actually change our minds?
Not every disagreement deserves an instant revision. Some objections are poorly informed; some new facts are noisy. Revision should follow the weight and relevance of evidence, not the volume or confidence of whoever offers it. Yet the opposite habit—treating every criticism as a personal threat—turns ideas into possessions.
For a real research claim, I would want operational definitions, prospective predictions, comparison with established models, repeated observations and attention to failure cases. For a personal essay, I would ask whether the argument is fair to its strongest opposition. For an interactive artwork, I would ask whether the experience illuminates the intended question without pretending to have measured the world.
05 / A SMALL MAP OF INFLUENCES
Borrow ideas. Keep receipts.
These are lenses, not celebrity endorsements. Each source contributes a specific distinction—and none certifies the entire method on this page.
Related work in organisational behaviour, behaviour design and philosophy is used as intellectual context—not as a claim that Sahil Tayal's working proposals have received empirical validation.
06 / IDEAS LEAVE THE ROOM
Follow the questions outward.
Six real doors into essays, a provisional model and designed experiments. Pick the medium that interests you—not a grand tour of every page.
The research below supports particular concepts used in the notebook. The broader five-move sequence is a practical editorial synthesis, not a published or validated intervention. Work linked in the cabinet is the author's own reflection, proposal, or creative experiment—identified on its destination.