Philosophy for Absolute Beginners: Thought Experiments and Testing Intuitions Carefully

What Thought Experiments Are For

A thought experiment is a carefully designed “what if?” used to learn something about your values, principles, or concepts. You are not trying to predict what would really happen in the world. You are trying to see what your judgments commit you to.

Good thought experiments typically do three jobs:

• Isolate a factor: Hold most things steady and change one feature so you can see what that feature does to your judgment.
• Reveal implications: If you accept a claim in the case, what else follows? Would you also have to accept something you dislike?
• Test consistency: If two cases are relevantly similar, do you judge them similarly? If not, what difference explains the change?

Think of a thought experiment like adjusting one slider at a time on a control panel. If you move many sliders at once, you cannot tell which one caused the result.

Three common targets

• Values: What matters most here—fairness, privacy, loyalty, safety, honesty, autonomy?
• Principles: Rules you implicitly rely on (e.g., “people deserve credit proportional to contribution”).
• Concept boundaries: Where you draw lines (e.g., what counts as “consent,” “coercion,” “privacy invasion,” “keeping a promise”).

Everyday Thought Experiments (Original Scenarios)

Scenario A: The bonus split rule

You manage a small team of four. The company gives your team a $4,000 bonus to split. You must choose a rule and apply it.

• Alex handled the hardest client and worked late often.
• Bea did steady work and helped others troubleshoot.
• Chen joined halfway through the year but made one key improvement that saved time.
• Dina did less visible work (documentation) that prevented mistakes.

You consider three possible rules:

Continue in our app.

• Listen to the audio with the screen off.
• Earn a certificate upon completion.
• Over 5000 courses for you to explore!

Or continue reading below...

Download the app

• Equal split: Everyone gets $1,000.
• Contribution-based: Split proportional to measurable output (tickets closed, features shipped).
• Impact-based: Split based on overall benefit to the team, including invisible work.

This is a thought experiment because it forces you to clarify what you mean by “fair.” Is fairness equality, desert, impact, need, or something else?

Scenario B: The smart doorbell privacy tradeoff

You install a smart doorbell camera. It records video when it detects motion and stores clips in the cloud. It also captures parts of the sidewalk and sometimes your neighbor’s front steps.

One night, a package theft happens on your street. A neighbor asks for your footage. You can share it, but the clip also shows:

• a delivery worker’s face clearly,
• a passerby who appears to be visiting a nearby clinic,
• your neighbor’s child playing outside.

You have options:

• Share the full clip with the neighbor.
• Share only with police if requested.
• Share a redacted clip (blur faces / crop the frame).
• Refuse to share unless there is a warrant.

This case isolates tensions between safety, community cooperation, and privacy. It also tests what you think “reasonable privacy in public” means when recording is continuous and searchable.

Scenario C: A promise made under pressure

You are interviewing for a job. The interviewer says, “We need someone who will stay at least two years. Can you promise you won’t leave before then?” You feel pressure: you need the job, and you suspect that refusing will cost you the offer.

Consider three versions:

• Version 1: You say, “I promise,” but you privately think you might leave if a better offer appears.
• Version 2: You say, “I intend to stay two years, but I can’t promise.”
• Version 3: You say, “I promise,” because you believe the promise is not binding when demanded in an interview.

This case helps you test what makes a promise valid: sincerity, freedom from pressure, mutual understanding, or something else.

Rules for Responsible Use (So the Case Teaches You Something)

1) Keep the case clear and minimal

Write the scenario in plain language. Include only details that might matter. If you add too much, you will not know what drove your judgment.

• Too messy: “Alex is a single parent, Bea is rich, Chen is disliked, Dina is your best friend…”
• Clearer: “Here are the roles and contributions; choose a rule for splitting.”

2) Track what changes (and what stays fixed)

When you compare two versions, list the single feature you changed. If more than one feature changed, you cannot interpret your own reaction.

3) Don’t smuggle in emotional language

Loaded wording can force a reaction without you noticing. Replace charged labels with neutral descriptions.

• Loaded: “Your creepy surveillance doorbell catches an innocent person.”
• Neutral: “Your doorbell camera records motion and stores clips.”
• Loaded: “Your greedy coworker demands more bonus.”
• Neutral: “A coworker argues their contribution was higher.”

Emotions can be relevant, but they should enter as explicit factors (e.g., “the person feels unsafe”) rather than as hidden cues in the wording.

4) Compare multiple variations, not just one dramatic case

A single scenario can mislead you because it may be unusual. Use a small “family” of cases to see whether your judgment is stable.

Example: For the doorbell case, compare:

• Variation: The clip shows only the thief, no bystanders.
• Variation: The clip includes sensitive bystander information.
• Variation: The request comes from police with a case number.
• Variation: The request comes from a neighbor who is likely to post it online.

5) Separate your immediate intuition from your considered judgment

First reactions are data, not verdicts. Write your initial answer, then ask what it depends on. If your answer changes after reflection, that is not failure; it is information about which reasons you find stronger.

Practice: Answer, Then Adjust One Feature at a Time

Step 1: Commit to an initial answer (no hedging)

Choose one practice scenario below and answer it in one sentence. Then add one sentence explaining why.

Practice scenario (choose one)

• Bonus: “Which rule should be used to split the $4,000, and why?”
• Doorbell: “Should you share the clip with your neighbor, and why?”
• Promise: “Is it wrong to promise two years in Version 1, and why?”

Step 2: Identify the key factor you think is doing the work

Circle the factor your explanation relies on most. Examples:

• Bonus: equality, measurable contribution, invisible labor, team morale, incentives, need.
• Doorbell: consent, public space, risk reduction, control over data, likelihood of misuse.
• Promise: sincerity, coercion/pressure, mutual expectations, harm from breaking, meaning of “promise.”

Step 3: Create three “one-change” variations

Make three new versions by changing only one feature each time. Use this template:

Base case: (your chosen scenario)  My judgment: ____  Because: ____  Key factor: ____  Variation 1 (change one feature): ____  New judgment: ____  What changed in my reasons: ____  Variation 2 (change one feature): ____  New judgment: ____  What changed in my reasons: ____  Variation 3 (change one feature): ____  New judgment: ____  What changed in my reasons: ____

Here are ready-made variations you can use.

Bonus variations (change one feature each)

• Variation A1 (need): Dina is facing a medical bill and asks for a larger share. Everything else stays the same.
• Variation A2 (measurement): Output metrics are known to be unreliable and miss documentation and mentoring.
• Variation A3 (choice): Before the year began, the team agreed in writing to an equal split, regardless of contribution.

Doorbell variations (change one feature each)

• Variation B1 (consent): Your neighbor previously asked you not to record their steps, and you agreed to “try.”
• Variation B2 (risk of misuse): The neighbor has a habit of posting videos of strangers online to “warn the community.”
• Variation B3 (authority): Police request the clip with a formal case number and explain it will be used only for identification.

Promise variations (change one feature each)

• Variation C1 (pressure): The interviewer says, “If you can’t promise, we won’t hire you.”
• Variation C2 (clarity): The interviewer says, “A promise means you will not accept another offer, even if your situation changes.”
• Variation C3 (stakes): Leaving early would cause your team to lose a major contract and likely lead to layoffs.

Step 4: Record which judgments changed and diagnose why

Make a simple change-log. The goal is not to defend your first answer; it is to learn what your answer depends on.

If your judgment flips, ask: did you discover a hidden condition in your original view (e.g., “sharing is fine only if it won’t be broadcast”)? If your judgment stays the same, ask: are you being consistent, or are you ignoring a morally relevant change?

From Intuition to a Candidate Principle (Then Stress-Test It)

1) Translate your judgment into a principle

Take your most stable judgment across variations and express it as a general rule. Keep it specific enough to guide action, but general enough to apply beyond the original story.

Examples of candidate principles (you may agree or disagree):

• Bonus principle: “A distribution is fair when it reflects overall contribution, including invisible work, unless the group agreed in advance to a different rule.”
• Doorbell principle: “Sharing surveillance footage is permissible only when the recipient will use it for a limited purpose and bystanders’ privacy is reasonably protected (e.g., redaction).”
• Promise principle: “A promise is binding only if it is made with sincere intent and without unreasonable pressure, and both sides share an understanding of what is being promised.”

2) Check the principle against a new case

Now test your principle on a different everyday scenario to see whether it gives a plausible result. If it gives a result you reject, revise the principle (not just the verdict) and test again.

New test cases

• Bonus test case: A team member contributed less this year because they were training a new hire at your request. Should that count as contribution?
• Doorbell test case: Your camera captures a car accident. A journalist asks for the clip. Sharing could improve public safety but also expose victims.
• Promise test case: You promised a friend to help them move on Saturday. On Friday, your parent has a medical emergency. Does the promise still bind, and what is owed?

As you test, keep the method disciplined: state the principle, apply it, note any discomfort, and revise by changing as little as needed. The aim is to make your principles match your best-considered judgments across a range of clear cases—not just to win a single scenario.