The Second Look

Do you see the rabbit—or the duck?

Do you see a rabbit first? Or a duck? If after a few seconds you can suddenly see both: welcome to the little stumbling block of perception.

Verified: The rabbit-duck drawing appeared 1892 in the Fliegende Blätter and was popularized 1899 by the psychologist Joseph Jastrow. Ambiguous images do not show two realities, but several plausible interpretations of the same template. And what you recognize first reveals nothing reliable about your character or your personality—even though social media likes to craft horoscopes out of it. Source on the rabbit-duck classification

Reminder: Perhaps you know this moment: A brief message comes in. "Can we talk briefly later?" No smiley. No context. And already a movie starts playing in your head: I did something wrong. There's trouble. This is probably it for my career—hopefully there's good Wi-Fi under the bridge.

Verified: The message itself initially contains only a little information. "There's trouble" is an interpretation, not an observable fact.

Assumption: Perhaps the person merely wanted to ask a follow-up question, offer help, or clarify an appointment. Perhaps there really is a problem. Both can be true; the message alone does not decide it.

This is exactly where the small, important gap lies: between what happens and what we make of it. Feelings do not arise detached from the situation; evaluations also help shape them. An interpretation is therefore not "merely imagined." It can be very convincing, very quick—and still provisional.

The ambiguous image does not demand: "See the duck already, you pessimist." It says something kinder: Your first look is a look, not a verdict from the universe. Sometimes the rabbit remains a rabbit. Sometimes the duck appears. And sometimes the most helpful thought is simply: "I don't yet know what I'm seeing there."

One word, four things—and none of them means magic

Verified: "Mindshift" is used inconsistently. It is not a precisely defined psychological technical term, but an everyday label for very different changes: interpreting a situation differently, examining a belief, redirecting attention, or taking a new learning path.

When it needs to be more precise, three terms help:

When something changes	More precise term
A distressing belief is examined on the basis of evidence	Cognitive restructuring
The meaning of a situation is reinterpreted so that it feels emotionally different	Cognitive reappraisal
One regards one's own thought from a distance instead of sitting inside it	Self-distancing

Verified: The APA describes cognitive restructuring as a technique of cognitive behavioral therapy: self-damaging beliefs and cognitive distortions are examined and replaced with more helpful evaluations. That explicitly does not mean: painting a real danger in nice colors. APA Dictionary: Cognitive Restructuring

Verified: Cognitive reappraisal means changing the interpretation of an emotional event. That is something other than merely smiling it away: the appointment can remain unpleasant; but perhaps it is not automatically proof that everything will go wrong. Overview of reappraisal

One word, four things
Verified: MindShift is the name of an educational offering by KQED. Verified: MindShift CBT is an app by Anxiety Canada that provides methods of cognitive behavioral therapy for anxiety. Verified: At the University of Marburg, a project on AI-supported neurostimulation for depression and anxiety disorders bears this name. It is still developing and validating; no proof of efficacy follows from that. Verified: Barbara Oakley's book Mindshift is about learning, career reorientation, and assumed limits.
Four names, four projects, no shared evidence for any miracle technique.

Assumption: Precisely because the word sounds so nicely like an inner switch, it is readily made bigger than it is. The more modest question is more useful: What exactly am I shifting right now—an evaluation, a habit, my perspective, or actually my action?

That is less glamorous than mental magic. But considerably more useful.

Your island is not the whole map

[Reminder] Do you know the moment? Someone says in a meeting: "We need to discuss this again." And already the translation is fixed: My work was bad. I am not trusted. This will be unpleasant. The other person, however, may merely have a follow-up question. Or time pressure. Or a calendar that looks as if it personally has a quarrel with them.

[Verified] Vera F. Birkenbihl does not call that a character flaw, but an effect of perceptual filters: We do not take in reality as a seamless transcript. Experiences, current mood, expectations, and fears influence what we pay attention to and what meaning we give to what we see. In Jeden Tag weniger ärgern! she speaks of the "filter of perception" and invites us not to confuse our own view with the whole of reality (p. 12, 4th edition 2025).

Their island is real. Only: it is not the whole map.

Not feeling less. Getting annoyed more efficiently.

[Verified] Birkenbihl's goal is explicitly not to abolish anger. Her sentence is: "Ärgern Sie sich effizienter!" (p. 8). What is meant is: to get annoyed for as short a time as possible, as rarely as possible, and as little intensely as possible—not to commit oneself to perpetual friendliness. Anger can indeed contain information: a boundary was touched, an agreement was violated, or something important was overlooked.

[Verified] At the same time, she warns against fueling anger further through repeated complaining: speaking and repeating can intensify feelings (p. 20). That is not a prohibition on speaking out. It is an invitation to notice the difference between a conversation that clarifies and a thought loop that keeps turning new rounds but arrives nowhere.

The island question

[Verified] Birkenbihl describes misunderstandings as a consequence of differing expectations and perspectives. Her island metaphor asks, in essence: How much of what I currently take for granted is also visible on the other person's island? She connects this with a percentage question: What proportion of the information, motives and constraints do I actually know? (p. 24–26)

Let us take the colleague who does not answer your message for three days.

On your island	What could additionally be true on the other island
"He is ignoring me."	He is prioritizing a deadline, caring for a sick child, or has overlooked the message.
"My request is unimportant."	Your request is important, but not yet decidable.
"I have to apply pressure now."	A precise follow-up question could lead to a usable answer more quickly.

[Assumption] None of these alternatives has to be true. That is exactly the point: The first thought is often one explanation among several, not the official final result of reality.

A question creates new data: "What currently has priority?" Or: "Is there still something you need from me in order to decide?" Perhaps the answer remains unpleasant. But perhaps a constraint, a different meaning, or simply a misunderstanding becomes visible. Then you have not thought prettily, but looked more closely.

Premises: the invisible directors

[Verified] Birkenbihl's work assignment is: "Examine your premises" (p. 26–30). Premises are the silent presuppositions beneath a judgment: Whoever respects me answers immediately. Or: A good colleague recognizes by herself what I need. Such sentences can be reasonable, exaggerated, or merely inappropriate in this specific situation. They still have an effect—often without our noticing them.

[Verified] For this purpose, Birkenbihl separates facts and assumptions (p. 29):

Fact: "My message has been unanswered since Tuesday."
Assumption: "She does not care about me."
Open question: "What do I not yet know?"

This is not an invitation to self-gaslighting. If someone repeatedly disregards agreements, the pattern can be real. The island question does not demand: Paint the situation nicely for yourself. It demands: Which parts do I know, which do I supplement—and what would I have to ask in order to find out?

[Context] The most useful new view is not the rosiest. It is the one that allows more information than the old one. Sometimes it is: "I really was treated unfairly." Sometimes: "I do not yet know the reason." And sometimes: "I can understand the reason and still set a boundary."

Is that true? Four questions, no thought police

A thought can appear with the force of a court ruling: "She does not respect me." "I failed." "This will go wrong." Its volume, however, is not proof of truth.

Verified: Byron Katie's reflection format The Work begins with four questions:

Is that true?
Can you absolutely know that it is true?
How do you react, what happens, when you believe that thought?
Who would you be without the thought?

After that, a voluntary "turnaround" can follow: for example, from "He does not respect me" tentatively to "I do not respect myself" or "I do not respect him." The questions and turnarounds come from The Work.

This is not thought police. The first question does not demand: "Please think something nicer." It also does not demand: "Your impression is false." It merely invites you to bring a thought down from the judge's bench and look at it.

A thought may be important without automatically having the last word.

Let us take the sentence: "My colleague is deliberately ignoring me."

Question	What it opens
Is that true?	What have I actually observed?
Can I be absolutely sure?	Do I know her intention—or am I filling in a gap?
What happens when I believe it?	Do I become cold, accusatory, do I withdraw?
Who would I be without the thought?	What would remain true before I assume an intention?

Reminder: Perhaps anger remains. But perhaps the more sober story is: "She has not replied for three days. I do not know why. I can ask." That is not rosier. Only more precise.

Questions instead of counter-speech

Verified: In cognitive behavioral therapy, automatic thoughts are treated as testable appraisals. Typical Socratic questions look for observations, evidence and counter-evidence, other explanations, and whether a thought is helpful. A review paper on cognitive restructuring describes this approach.

Practical questions can be:

What is the observable event—without interpretation?
What evidence supports my thought?
What evidence speaks against it?
What other explanation also fits the facts?
Am I confusing a possibility with certainty?
Does this thought help me act appropriately?

Verified: For observed cognitive restructuring and therapy outcome, a meta-analysis of four process studies with 353 patients found an association of \(r = 0{,}35\) (95% CI 0,24–0,44). This shows an association, but not an isolated causal proof: It is therefore not proven that a particular question alone produces therapeutic success. Ezawa and Hollon, 2023

Verified: There is no solid evidence that one particular type of Socratic question is reliably superior to all others. Good questions are not magic words; they must fit the situation, the person and the moment. Analysis of Socratic dialogue

Turnaround: perspective test, not judgment

Verified: The turnaround belongs to Katie's method. It is most useful as a voluntary test: Are there also places where the reversed statement reveals something about my role, my needs or my way of dealing with things?

Assumption: Sometimes it opens a surprising door. Sometimes it opens none. Both are allowed.

It is never a judgment about the reality of other people. In cases of violence, mobbing, discrimination, manipulation or gaslighting, "Just turn it around" would not be wisdom, but a dangerous oversimplification. Documenting observations, setting boundaries, seeking support and organizing protection then remain more important than any elegant question.

Not everything has to be reinterpreted

Verified: Cognitive reappraisal changes the meaning of a situation in order to influence its emotional impact. ACT defusion pursues something different: it does not first check whether a thought is correct in content, but rather creates distance from it.

The difference sounds small, but in practice it is large:

Reappraising: "Maybe the terse email is not criticism, but time pressure."
Defusion: "I notice the thought: this email means trouble."

Reminder: For some thoughts, the second form is more honest. If the situation really is unpleasant, you do not have to sprinkle glitter on it. It can be enough not to marry every thought immediately.

And sometimes the best answer to "Is that true?" is not "No," but: "Partly. And what do I do now with the part I know?"

Glasses on an open book, each lens showing a different landscape — The same page, two glasses – and two different worlds.

What research can do – and what it does not promise

Verified: Thinking differently about a distressing situation can influence feelings. In a meta-analysis with 306 experimental comparisons, the average effect of cognitive change on emotional outcomes was d⁺ = 0,36. For reappraising a trigger, it was also 0,36; reappraising one's own reaction was 0,23. These are small to medium average effects – not a remote control for life. Webb, Miles and Sheeran (2012)

Context: An average is not a prediction for a single Tuesday at 16:40 h. It does not mean that a conflict, an illness, financial hardship or unfair treatment disappears. It only means: under study conditions, a different appraisal on average also changed emotional outcomes a little.

Verified: Under acute strain, stress reappraisal improved performance on average by d = 0,23. Interventions that combined it with additional components were at d = 0,45. This must not be read as though reappraisal alone produced the larger value: the combination contains more than one lever. Bosshard and Gomez (2024)

The sobering, helpful news: a video is not a personality renovation. Verified: Two randomized studies initially changed stress mindsets markedly; after one week, largely little of that remained (d = 0,15). A lasting, statistically significant gain in well-being did not appear. Stress mindset studies (2024)

Even grand promises about "growth mindset" shrink on closer inspection to a manageable format. Verified: Growth mindset interventions showed only very small effects on school performance on average: d = 0,08; in a stricter analysis, d = 0,04, with a confidence interval that also allows no effect. Sisk et al. (2018)

Distance can help, but it too is no magic trick. Verified: A meta-analysis of 25 experiments found a small average advantage for self-distanced rather than self-immersed reflection (g = 0,19). So whoever thinks about themselves for a moment as though about another person sometimes gets more overview – not necessarily a solution. Murdoch et al. (2022)

And positive thinking? Verified: In two experiments, people with low self-esteem sometimes felt worse after repeating a positive self-statement than comparison participants. A sentence does not have to become true just because it looks good on a coffee mug. Wood, Perunovic and Lee (2009)

Mnemonic: Research does not promise a rosier story. It shows cautiously: a more plausible, more fitting or somewhat more distanced view can change room for action and feelings. Whether it fits is decided not by optimism, but by the situation.

Old answers, new labeling

Some of what sounds today like a fresh change of perspective is very old – and thereby becomes neither false nor automatically scientific.

Verified: In the New Testament, metanoia means a turning around or a religious-ethical reorientation toward God, not merely the swapping of a single thought. In Acts 3,19, repentance is connected with turning toward; Mark 1,15 links it with faith in the gospel. That is religious language with a claim on conduct of life – not a psychological study in the garb of an old term.

Verified: Epictetus begins his Handbook with a distinction that helps surprisingly well against mental entanglement: some things are up to us, such as judgments and desires; others are not, such as body, property and reputation. This can be read in Enchiridion 1. That too is not a therapy instruction. It is philosophy – and an imposition: do not take everything personally that strikes us. But also do not act as though everything were in our hands.

Then there is the new labeling: thoughts as an order placed with the universe. Rhonda Byrne's The Secret appeared in 2006 first as a film and then as a book; the publisher describes in it a universal law with which a joyful life can be created "deliberately and effortlessly." That is the publisher's documented claim.

An idea may comfort, motivate or give a morning a friendlier color. Context: It does not follow from that that thoughts cause external events independently of action, power relations, money, discrimination, illness or chance. That would be a very strong causal claim – and something different from hope, prayer or personal spirituality.

The universe is not complaint management with a thought form.

Verified: "Positive thinking" too is not automatically helpful. In two experiments, people with low self-esteem felt worse after repeating a positive self-statement or after one-sided reflection on it than comparison participants; people with high self-esteem benefited only slightly. Wood, Perunovic and Lee (2009) did not examine spirituality, but positive self-statements. The fair lesson is therefore not: forbid optimism. Rather: a formula does not have to be correct just because it sounds kind.

The most important switch: can I change anything?

Some thoughts do not need a friendlier answer, but first a map.

"This can't go on like this" can mean: I need to have a conversation. But it can also mean: I am mourning something that cannot be undone. From the outside, both sound similar. For the next step, it is a world of difference.

[Verified] The study by Troy, Shallcross and Mauss suggests an important context rule: How helpful reappraisal is depends on whether a burden is experienced as controllable. For uncontrollable burdens, the ability to reinterpret was associated with fewer depressive symptoms; for burdens experienced as controllable, the opposite relationship appeared. The study does not prove a simple recipe formula – but it supports a very reasonable preliminary question: "Emotion Regulation and Psychopathology" (Troy, Shallcross & Mauss, 2013).

The switch is not: "How do I think about this more beautifully?" Rather: "Can I change something?"

When something is changeable: first work on reality

[Heuristic] If you actually have influence, before reinterpreting, check:

Can I clearly say no or set a boundary?
Is a conversation needed, a concrete request or a complaint?
Should I document incidents, messages or agreements?
Who can support me: colleague, works council, counseling, circle of friends, legal advice?
What would be a small step that gives me new information?

This is not a call for endless fighting. Influence can also mean getting help or leaving a situation. A thought like "I can't change anything about this" is sometimes a precise description of a state. Sometimes it is the old protective formula of a person who had to fight alone for too long. Both deserve examination, not rah-rah.

[Reminder] A change of perspective does not replace a boundary. But it can help to recognize the next boundary more clearly.

When something is not changeable: change the relationship to it

A diagnosis, a loss, the past, another person's decision: Not everything can be influenced by a well-conducted conversation, a list or three color-coded calendar blocks. Here, reinterpreting can have its more sensible place.

Not: "This isn't actually so bad."

Rather: "It is bad. And I do not additionally have to fight every minute against the fact that it happened."

[Assumption] Often it is not the especially positive thought that relieves, but the more precise one: I have no influence on the event, but perhaps on how I deal with it next. That can mean accepting support, grieving, taking breaks or giving the thought less sole rule.

Stop: Not everything is a thinking task

[Verified] Discrimination is not merely a question of perspective. A meta-analysis of 73 studies with 12.097 people found that experimentally triggered discrimination immediately worsened mental health; with repeated discrimination, the effect was stronger. The meta-analysis therefore does not describe a "negative attitude," but the consequences of real disadvantage.

[Verified] Gaslighting is described in a systematic review on gender-specific violence as manipulation of reality and perception as well as emotional abuse in connection with power and control. Reported consequences include self-doubt, isolation and emotional stress. The review is a good reason, in such situations, not to begin with "Maybe I'm just seeing it too negatively?"

Safety rule: In cases of violence, threats, stalking, gaslighting, discrimination or acute distress, reinterpreting is no substitute for protection, documentation, counseling and help.

In acute danger, the following applies in Germany and many European countries: 112. Confidential help is provided by Telefonseelsorge Deutschland at 0800 111 0 111 or 0800 111 0 222, the Hilfetelefon "Gewalt gegen Frauen" at 116 016, in Austria TelefonSeelsorge at 142 and in Switzerland Die Dargebotene Hand at 143.

The decisive question is therefore not: Am I strong enough to think differently? Rather: What does this situation demand of me – action, protection, support or a different way of dealing with the unchangeable?

That is less glamorous than thought magic. But it has one decisive advantage: It takes reality seriously.

The exercise: Dethroning a thought

A thought may be loud. It may be unpleasant. It may even briefly appear as if it had a master key to the truth.

It therefore does not have to rule.

This exercise is a heuristic for a concrete, stuck everyday situation – not therapy and not a promise of effectiveness. Do not take anything that overwhelms you or acutely endangers you. In cases of violence, gaslighting, discrimination or acute distress, the following applies: safety, support, documentation and help take precedence over any change of perspective.

1. Notice: Write down the raw wording

Choose a situation that repeatedly occupies you. Not your whole life at once; one sentence is enough.

"She isn't answering because I don't matter to her." "I failed." "This will go wrong again."

Write down the thought as it comes. Without making it prettier, more reasonable or more positive.

Reminder: A thought is first of all an event in the head – not automatically a report about the world.

2. Accept: Do not argue anything away

Name what is there:

"I am hurt." "I am afraid." "I am angry – and from my point of view that currently makes sense."

You do not have to like the feeling. But it rarely becomes more cooperative if you dismiss it with a calendar quote.

Heuristic: First acknowledge, then examine. Not: "I am not allowed to feel this." Rather: "There it is. What is it telling me?"

3. Question: Ask the thought to be examined

The four questions from Byron Katie's The Work are a possible reflection format, not an interrogation and not a thought police:

Is that true?
Can I be absolutely sure that it is true?
How do I react when I believe this thought?
Who would I be in this situation without this thought?

[Verified:] These four questions are published by The Work. Cognitive behavioral therapy also works with questions about evidence, counter-evidence, alternative explanations and the usefulness of automatic thoughts. Ezawa & Hollon, 2023

Add two sober control questions:

Fact or film? What did I actually observe—and what am I adding?
What other explanation would be at least as plausible?

Example:

Check step	Answer
Fact	The message has been unanswered for two days.
Film	"I'm unimportant to her."
Other explanation	She could be busy, ill, overwhelmed or unsure.
Honest residual uncertainty	I don't know the reason yet.

It is not about winning the rosiest story. It is about not mistaking the first story for a court ruling.

4. Distance: Giving yourself company while thinking

Change perspective for a moment. Speak inwardly about yourself in the third person:

"What does Anna need right now?" "What would I say to Jonas if he had this thought?" "What does Lea see when she takes a step back?"

[Verified:] A meta-analysis of 25 experiments found a small average advantage of g = 0,19 for self-distanced versus self-immersed reflection. That is not a thought elevator to the seventh floor, but a possible step out of the thought loop. Murdoch et al., 2022

5. Seeing anew: Formulating the most plausible interpretation

Write a sentence that neither glosses over nor dramatizes:

Not: "Everything will be wonderful." But rather: "The feedback may be critical. I don't know it yet and can prepare myself."

Or:

Not: "I am irreplaceable." But rather: "I made a mistake. It does not follow that I am a mistake as a person."

Assumption: Merely formulating a more plausible interpretation can create room to maneuver. Whether and how strongly it does depends on the situation, habit and options for action.

6. The switch: Can I change something?

Answer this question as concretely as possible:

Yes, at least in part. What can I influence: a conversation, a boundary, a request, an application, a break, documentation, support?
No, not at present. What can I accept, mourn, limit or let go of without glossing over the situation?

[Verified:] Research on reappraisal suggests that context matters; the distinction between stressors that can be influenced and those that cannot is especially important. Troy, Shallcross & Mauss, 2013

7. A small step that brings new data

Choose an action that can begin in at most ten minutes:

formulate a message factually,
arrange an appointment,
note a boundary,
ask someone for concrete feedback,
work on the task for five minutes,
check a piece of information,
contact support.

The step does not have to look brave. It only has to be real.

Reminder: Thoughts do not change only through better counterarguments. Behavior provides experiences that an old thought did not yet know.

To take away:

"I notice this thought. I examine it. I do not have to leave it on the throne. And I can choose the next small step."

Sources, scope and what remains open

This page deliberately separates three things: confirmed findings, assumptions with a plausible direction and reminders that still need a robust source. That is less glamorous than a wallpaper of links—but considerably more useful.

Status	Source	What it can support here
Confirmed	Birkenbihl, Jeden Tag weniger ärgern!, 4th ed. 2025	Birkenbihl's own positions: "getting annoyed more efficiently" (p. 8), perceptual filters (p. 12), complaining (p. 20), misunderstandings and expectations (p. 24–25), the island and percentage question (p. 25–26), examining premises (p. 26–30), facts and assumptions (p. 29).
Confirmed	Webb, Miles & Sheeran (2012)	In experimental comparisons, cognitive change had on average a small to medium effect on emotional outcomes: d⁺ = 0,36. That is not proof that external problems disappear.
Confirmed	Bosshard & Gomez (2024)	Stress reappraisal improved performance under acute stress by a small average amount: d = 0,23. Higher values for combined methods are not attributable to reappraisal alone.
Confirmed	Stress-mindset RCTs (2024)	Short-term attitude changes largely did not persist after one week; a sustained significant gain in well-being did not appear.
Confirmed	Murdoch et al. (2022)	Self-distancing had on average a small advantage over mental immersion: g = 0,19.
Confirmed	Ezawa & Hollon (2023)	Cognitive restructuring and therapy success were associated, r = 0,35; however, four studies with 353 people are a narrow basis and not causal proof.
Confirmed	Sisk et al. (2018)	Growth-mindset interventions showed on average very small effects for school performance.
Confirmed	Wood, Perunovic & Lee (2009)	A positive self-statement can make people with low self-esteem feel worse.
Confirmed	Epton et al. (2015)	Values-based self-affirmation showed small effects on acceptance, intention and later behavior; it is not the same as positive recitation.
Confirmed	Simons & Chabris (1999)	Attention is limited; it does not follow that every one of one's own interpretations is wrong.
Reminder	Wiseman & Watt (2025)	The rabbit-duck reaction reveals nothing about personality. The primary bibliographic source should still be completed in full before publication.

Open: The detailed figures for Troy, Shallcross and Mauss (2013) are not repeated here because they still need to be checked against the full text. The useful but preliminary assumption is: whether reinterpretation helps also depends on whether a situation can be changed.

Open as well is the robust body of studies on Byron Katie's The Work: the four questions are documented as a method, but superior effectiveness compared with other forms of reflection is not thereby proven. The exact classification of Watkins's concreteness training also still needs full-text review: less rumination is not automatically a stronger reduction in depressive symptoms.

Another thought is not a new reality—but sometimes the beginning of a new action.