ProjekinAja. Clear Thinking Stage 2 · Applied Judgment Lesson 2.5
● Stage 2 · Lesson 2.5

Ethics & Moral Consistency

Moral reasoning requires pinning down the exact principle being claimed. Three different principles can each condemn different things — and applying them inconsistently is a problem, not a debating trick.

⏱ ~16 min read 🧩 Ethical reasoning 🎮 Principle picker

🎯 By the end you'll be able to

Identify which moral principle is actually being claimed in an ethical argument; explain why consequentialism, deontology, and virtue ethics yield different verdicts on the same case; apply the universalization test to any moral principle; and diagnose inconsistency in how principles are applied across cases.

FOUNDATIONSThree ethical frameworks

Most moral arguments, even informal ones, implicitly invoke one of three major frameworks. Knowing which one is being used — and which one would reach a different conclusion — is essential to evaluating the argument.

FrameworkCore questionWhat makes an action right?Famous version
Consequentialism What outcomes does this produce? An action is right if it produces the best overall outcomes — most welfare, least suffering — across all affected parties. Utilitarianism (Bentham, Mill, Singer)
Deontological ethics What duties and rights are at stake? An action is right if it conforms to moral duties and respects rights, regardless of consequences. Kantian ethics; rights-based theories (Kant, Nozick)
Virtue ethics What would a person of good character do? An action is right if it expresses and develops virtuous character — honesty, justice, practical wisdom, generosity. Aristotelian ethics (Aristotle, MacIntyre)

These frameworks don't always agree. An action that produces good outcomes might violate someone's rights (consequentialism says do it; deontology says don't). An action that follows a rule might produce bad outcomes (deontology says do it; consequentialism says don't). An action that is technically permitted might reflect poor character (virtue ethics objects where the others might not). Real moral reasoning often requires navigating these conflicts — not ignoring them.

🔑 The key move

When evaluating a moral argument, ask: "Which framework is this argument implicitly using? Would a different framework reach a different conclusion? And is the framework applied consistently?"

CONCEPTSKey moral concepts

Several concepts appear across frameworks and need to be distinguished before we can evaluate moral arguments precisely.

ConceptMeaningWhat it requires
Rights Entitlements that constrain how a person may be treated, regardless of aggregate benefit Identifying whose right is at stake and what it entails
Consent Voluntary agreement to an action or arrangement Was the agreement informed, free of coercion, and from someone capable of giving it?
Fairness Distribution of benefits and burdens according to relevant moral criteria Specifying which criteria are relevant — equal treatment? proportional contribution? need?
Exploitation Gaining advantage by taking unfair benefit from another's vulnerability or lack of alternatives Distinguishing mutually beneficial exchange from transactions that are beneficial but unfair
Harm Making someone worse off than a relevant baseline Specifying the baseline — worse off than before? than they would have been? than is fair?
Responsibility Being the appropriate target of praise, blame, or obligation Establishing causal contribution, foreseeability, and whether alternatives existed

Notice how much work the specification of each concept does. "This is harmful" is not a complete moral claim until you've specified: harmful to whom? compared to what baseline? caused directly or indirectly? foreseeably or not? These specifications are where most moral disagreements actually live.

CORE ANALYSISThe three possible principles behind "investing is wrong"

The argument in our running example ("it is wrong to profit when another person loses; investing is zero-sum; therefore investing is wrong") does not specify which moral principle it is invoking. There are at least three plausible candidates, and they are not equivalent.

📖 Principle A

"It is wrong whenever my financial gain produces another person's financial loss."

This is the broadest reading. It condemns any transaction — trade, employment, negotiation — where one side ends up with more money than before and the other ends up with less. On this reading, selling anything at a profit, negotiating a raise, or driving a hard bargain is wrong.

📖 Principle B

"It is wrong to deceive, manipulate, or coerce another person in order to profit from them."

This is a narrower, deontological reading. It condemns fraud, manipulation, and exploitation — not voluntary exchange. A seller who accurately represents their product and a buyer who accepts the price have not wronged each other, even if one benefits more than the other.

📖 Principle C

"It is wrong to participate in systems that distribute benefits and burdens unfairly across society."

This is a systemic, distributional reading. It condemns structural features of an economic system — not necessarily any particular voluntary transaction within it. It may condemn speculative activity that concentrates wealth without productive contribution, while permitting productive investment.

These three principles are genuinely different. Here is why it matters:

Someone who says "investing is wrong" must be invoking one of these (or another specific principle). Which one they mean determines what conclusions actually follow — and what the counterevidence and counterexamples are.

💡 The practical move

When you encounter a moral argument, ask: "Can you state the principle precisely enough that I could test it on three other cases?" If not, the principle is underspecified and cannot be properly evaluated. Precision is not pedantry — it's what makes moral arguments honest.

TESTUniversalization — Kant's acid test

One of the most powerful tools for testing a moral principle comes from Kant: universalization. Ask: "What if everyone acted on this principle?" If universal adoption produces a contradiction or an obviously unacceptable world, the principle has a serious problem.

Applying universalization to our three principles

The universalization test doesn't settle all moral questions, but it is a powerful filter: a principle that breaks down when universalized is doing something structurally wrong, even before we've evaluated its specific content.

CORE SKILLMoral consistency — the hardest requirement

Moral consistency means applying the same principle to relevantly similar cases. It is not the same as never changing your mind — it is the requirement that if a principle condemns Case A, it must also condemn Case B when Case B is morally relevantly similar to Case A.

Inconsistency is a genuine moral problem, not just a debating trick, for this reason: if you apply a principle selectively — invoking it when it condemns things you already dislike and ignoring it when it would condemn things you don't mind — then the principle is not actually doing the moral work. Your prior preferences are. The principle is just a post-hoc label.

The claw-machine case

This is where our running example's side fact becomes morally relevant. The person claims:

Claw machines are structurally more zero-sum than most investment: the operator consistently profits, the average player consistently loses, and no new value is created. Under Principle A (all profit-from-another's-loss is wrong), claw machine play is condemned more clearly than long-term equity ownership.

This is a genuine consistency problem — not a tu quoque ("you're inconsistent, therefore wrong"). Inconsistency doesn't make the conclusion false; it demands an explanation of the morally relevant difference that justifies treating the two cases differently. There might be one! Perhaps the person would argue that claw machines involve entertainment value that makes the loss acceptable, or that the scale of harm is different. Those responses require engaging with the principle's application, not dismissing the consistency challenge.

⚠ Consistency vs tu quoque

Pointing out inconsistency is not the same as "you're a hypocrite, therefore your argument is wrong." The inconsistency challenge says: "Your principle, applied consistently, implies X about this other case. Do you accept that? If not, explain the morally relevant difference." It is a demand for coherence, not an ad hominem attack.

INTERACTIVEPrinciple picker — which principle condemns what?

For each scenario below, select which of the three principles (A, B, or C) condemns it, permits it, or requires more information. Then reveal the model analysis.

⚖ Interactive · Principle Picker

Select a scenario and choose which principle(s) condemn it before revealing the analysis.

APPLYING THE LESSONBack to the running example

We now have the tools to be more precise about what the running argument actually requires.

The argument's P1 — "it is wrong to profit when another person loses" — most naturally reads as Principle A. But as we've seen:

This doesn't mean the person's concern is groundless. The strongest version of their position might be something like: "Speculative financial activity that extracts value without creating it, especially when it concentrates wealth and increases systemic risk, is unjust — and ordinary people who participate in it without understanding what they're doing are contributing to systems they would not endorse on reflection."

That is a defensible position. It is also not what P2 of the original argument says, and it requires far more careful specification than "whenever someone profits, another person loses." We'll build the strongest version of this argument — and then evaluate it properly — in the final lesson.

WEAKNESSCommon mistakes in moral reasoning

✗ Treating moral intuitions as self-evident principles.

A strong moral intuition ("this feels wrong") is evidence that a principle is at stake — but it isn't the principle itself. The work of moral reasoning is making the principle explicit, testing it, and checking whether it's the intuition's real source or just a post-hoc rationalization.

✗ Conflating "harmful" with "wrongful."

Not all harm is wrong, and not all wrong is harmful in the conventional sense. Declining to rescue someone (when you easily could) may be wrongful without causing direct harm you inflicted. A surgeon causing pain during surgery harms but doesn't wrong. Distinguishing these requires specifying the moral concept being applied.

✗ Applying one framework selectively.

Invoking consequences when they support your conclusion, rights when they do, and character when those do — without acknowledging that these frameworks can conflict — is not moral reasoning. It's cherry-picking justifications. Real moral reasoning requires acknowledging which framework you're using and what it implies in cases you don't prefer.

✗ Treating consistency challenges as personal attacks.

"But you do X, which your principle also condemns" is a philosophical challenge requiring a response about the morally relevant difference, not a personal attack to be deflected. Deflecting instead of engaging is a sign that the principle is underspecified.

PRACTICE🔨 Weekly exercise

This week's task

Take one of your own strong moral judgments — something you believe clearly and confidently — and do the following:

  1. Write down the exact moral principle behind it, precisely enough that a stranger could apply it to new cases.
  2. Apply that principle to three other cases — ideally ones where you don't have a strong pre-formed view.
  3. Check: does the principle give verdicts you find acceptable in all three cases? If not, either revise the principle or explain the morally relevant differences.
  4. Apply the universalization test: what happens if everyone acts on this principle?

This exercise is uncomfortable. The point of the discomfort is discovering where your stated principle and your actual values diverge — which is information you want.

GLOSSARYKey terms introduced

Consequentialism
The view that the rightness of an action is determined by its outcomes — specifically, whether it maximizes overall welfare.
Deontological ethics
The view that the rightness of an action is determined by whether it conforms to moral duties and respects rights, regardless of consequences.
Virtue ethics
The view that morality is fundamentally about character — acting rightly is what a person of good character would do.
Universalization
Kant's test: ask whether a principle could be consistently adopted by everyone. Principles that fail this test have structural moral problems.
Moral consistency
Applying the same moral principle to relevantly similar cases — a requirement for coherent moral reasoning, not just a debating tactic.
Exploitation
Gaining advantage by taking unfair benefit from another's vulnerability or lack of alternatives — distinguished from mutually beneficial voluntary exchange.
Harm
Making someone worse off than a relevant baseline — requires specifying the baseline to be a complete moral claim.
Tu quoque
The logical fallacy of dismissing an argument because the person making it behaves inconsistently — not the same as a legitimate consistency challenge.

ACTIVE RECALLCheck your understanding

Answer in your own words first, then reveal.

1What is the difference between Principles A, B, and C as applied to the running example? Why does it matter which one the argument is invoking?›

Principle A condemns all profit when another loses financially — extremely demanding, fails universalization, likely not what is actually believed. Principle B condemns deception, manipulation, and coercion — much narrower, widely defensible, does not automatically condemn voluntary investment. Principle C condemns participation in systems distributing benefits unfairly — requires specifying what counts as unfair, may condemn some financial activities but not others. It matters because the three principles yield different verdicts on ordinary investment, speculation, fraud, and claw-machine play — and the argument's conclusion only follows if the right principle is being invoked and correctly applied to the actual activities in question.

2What is universalization, and how does it expose problems with Principle A?›

Universalization asks: what if everyone acted on this principle? Principle A — "it is wrong whenever my profit involves another's financial loss" — universalized, would prohibit all voluntary market exchange (every sale involves one party with more money and one with less). Trade, employment, and commerce would be logically impossible. A principle whose universal adoption produces a self-defeating contradiction fails Kant's test. Principle A, despite expressing a genuine moral concern, is too broadly stated to be the actual principle anyone consistently holds.

3Why is pointing out that someone plays claw machines a legitimate moral challenge, not just tu quoque?›

Tu quoque is "you're inconsistent, therefore your argument is wrong" — that's a fallacy. The legitimate consistency challenge is: "Your principle, applied consistently, would also condemn claw machines — and claw machines are structurally more zero-sum than much long-term investment. Do you accept that verdict? If not, explain the morally relevant difference that justifies treating them differently." This is a demand for coherence, not an ad hominem. It may reveal that the principle is underspecified (distinguishing entertainment from finance?) or that the person doesn't actually endorse the strong reading of P1. Either way, it is philosophically productive.

4Why is "this feels wrong" not a complete moral argument? What is missing?›

A moral intuition is evidence that a principle is being violated — but it is not the principle itself. Without making the principle explicit, we cannot: (1) test whether it applies to this case correctly; (2) check whether it is applied consistently to relevantly similar cases; (3) apply the universalization test; (4) evaluate conflicts between principles. Moral reasoning requires extracting the principle, specifying it precisely, and then evaluating it. The intuition is the starting point, not the argument.

Progress saves in this browser. Next: argument mapping and the complete method →