Metaethical topics Andrew is invoking (and why)

Semantics of moral language
- Cognitivism vs. non-cognitivism: Are moral claims truth-apt (state facts) or primarily express attitudes/commands?
- Meaning of “wrong/ought”: Do these pick out stance-independent reasons, or are they instrumental (“given a goal, you ought…”)?
Normativity & reasons
- Categorical vs. hypothetical “ought”: Categorical binds regardless of goals; hypothetical binds relative to aims.
- Pro tanto vs. all-things-considered: A pro tanto reason counts against an act but can be overridden; “all-things-considered wrong” means the net of reasons forbids it.
- Internalism vs. externalism about reasons: Must reasons connect to the agent’s motivations (internal), or can they be objective (external)?
Metaphysics of morality
- Realism vs. anti-realism: Are there stance-independent moral facts (realism), or are truths constructed/attitudinal/relational (constructivism, expressivism, error theory)?
- Supervenience: Moral facts depend on non-moral facts—no moral difference without some non-moral difference.
Epistemology & method
- Axioms / starting points: To avoid infinite regress, we begin with defeasible basic commitments and adjust via reflective equilibrium (mutual adjustment of principles and judgments).
- Burden of proof & dialectic: In applied ethics, both sides owe at least one defeasible reason; metaethics clarifies terms, but it’s not a shield against offering reasons.
“Rights” talk
- What a right is: A claim, power, or immunity that others have defeasible duties to respect.
- Where rights “exist”: Competing accounts—natural rights (personhood/rational agency), Kantian/Rawlsian constructivism (what free and equal agents would agree to), contractualism (Scanlon-style “what we could not reasonably reject”), or practice-dependent/legal.

One-line, debate-tight definitions you can deploy

Entailment: “X entails Y” means that in every admissible case where X is true, Y must be true (semantic entailment: X ⊨ Y).
Normative “ought/should”: A claim about reasons that count in favor of or against actions independent of mere desire satisfaction.
Instrumental “should”: “Given goal G, you should do A as an effective means to G.”
Pro tanto vs. all-things-considered: Pro tanto = defeasible reason; all-things-considered = the net verdict after weighing all reasons.
Axioms (in ethics): Minimal starting commitments we provisionally accept to stop regress and revise by reflective equilibrium.
Right: A claim-right that others have defeasible duties to respect (non-violation is normally required).
Wrong (my usage): “φ is wrong” = “There is decisive, all-things-considered reason not to φ (normally because φ violates justified rights or imposes unjustified harm).”

“Answer fast, then pivot” replies to Andrew’s moves

Use the same rhythm each time:
(i) Give the crisp answer in one sentence. (ii) Tie it to the topic. (iii) Pivot back.

1) “What do you mean by entailment?”

Answer: “By entailment I mean: in every case where X holds, Y must hold—formally, X ⊨ Y.”
Tie: “So when I say your premise entails your conclusion, I mean it’s not just suggestive; it’s necessary given your own commitments.”
Pivot: “We have that definition now—so please give your first substantive reason for calling promiscuity wrong.”

2) “What does ‘wrong’ mean here?”

Answer: “I’m using ‘wrong’ to mean ‘there is decisive, all-things-considered reason not to do it—typically because it violates justified rights or imposes unjustified harm.’”
Tie: “That makes our dispute empirical-normative: does promiscuity, as described, violate rights or impose unjustified harm?”
Pivot: “With ‘wrong’ fixed, what’s your first concrete reason that promiscuity meets that threshold?”

3) “Why should we use axioms?”

Answer: “Because without provisional starting points, justification regresses infinitely; axioms are defeasible anchors tested by reflective equilibrium.”
Tie: “A minimal axiom we can share is: ‘Avoid imposing serious, non-consensual harm; respect valid consent.’”
Pivot: “Granting that minimal axiom, explain how consenting adult promiscuity violates it.”

4) “Where do rights exist?”

Option A (Realist): “Rights are stance-independent normative facts about persons as agents; they supervene on features like autonomy and vulnerability.”
Option B (Constructivist): “Rights are what free and equal agents could not reasonably reject under fair conditions; they’re constructed but robust.”
Either way—Tie: “On both views, a rights-violation requires more than disapproval or distaste.”
Pivot: “Which right do you claim promiscuity violates—bodily, property, associational—and by what mechanism?”

5) “Are your ‘oughts’ categorical or hypothetical?”

Answer: “The moral ‘ought’ I’m using is categorical (not goal-conditioned), though instrumental oughts are fine for means-ends talk.”
Tie: “So if you say ‘promiscuity is wrong,’ you’re making a categorical claim—one that needs a categorical reason.”
Pivot: “Name that categorical reason—what makes the act wrong regardless of anyone’s goals?”

6) “Isn’t metaethics necessary before we proceed?”

Answer: “It’s necessary to the extent of fixing terms; we’ve done that in under a minute.”
Tie: “Further delay doesn’t add clarity; it avoids substance.”
Pivot: “Back to content: offer your strongest reason, then we can test it against the agreed terms.”

If he flips burden of proof or “answers a question with a question”

Use the Closed-Loop Challenge:

Restate the task: “You claimed promiscuity is wrong and shaming is good.”
Confirm shared terms: “We fixed ‘wrong,’ ‘entailment,’ and the minimal axiom.”
Demand a reason: “Provide one reason showing a rights-violation or unjustified harm.”
Time-box: “Do it in a single sentence; we’ll unpack after.”
If he still pivots:
“Noted: multiple pivots, zero reasons. I’ll proceed on record that no reason has been provided.”

Pre-emptive answers to his likely follow-ups (keep them pocket-sized)

“Feelings vs. reasons?”
“Appeal to feeling isn’t a reason; at best it’s evidence of an attitude. We’re asking for a reason that survives intersubjective scrutiny.”
“Utilitarian angle—social harm?”
“Then specify the non-speculative harm and its mechanism (coercion, deception, disease externalities), show it’s not preventable by regulation/education, and show the harm outweighs the liberty interests of consenting adults.”
“Virtue ethics angle—character corruption?”
“Call it pro tanto at most; show stable vice formation beyond cases of deception/exploitation. Otherwise it’s not an all-things-considered wrong.”
“Community norms / disgust?”
“Disgust is not a defeater; you need either a rights argument or a demonstrated, non-trivial harm.”

A compact script you can deliver live

You: “Quick definitions so we’re efficient: ‘entailment’ = X ⊨ Y (if X, necessarily Y). ‘Wrong’ = decisive, all-things-considered reason not to do the act, typically because it violates justified rights or imposes unjustified harm. We’ll use a minimal axiom—avoid non-consensual serious harm and respect valid consent—and reflective equilibrium for revision. With those fixed, what’s your first concrete reason that consensual adult promiscuity violates a right or imposes unjustified harm?”
(If he asks “Where do rights exist?”)
You: “Pick realism or constructivism; on either, a rights-violation requires more than disapproval. Name the right being violated and the mechanism.”
(If he says “Why axioms?”)
You: “To stop regress; they’re provisional and revisable. We’ve adopted the thinnest one we both accept. Now your reason, please.”
(If he keeps pivoting)
You: “We’ve settled semantics. You’ve offered no reason. Either provide one sentence identifying the violated right or specific unjustified harm—or concede you have none.”

Why this silences the meta-stall (without ad hominem)

You give precise, orthodox definitions in seconds—removing his pretext for delay.
You pin the dialectical burden: he made a categorical moral claim; he owes at least one categorical reason.
You force specificity (which right? what harm? what mechanism?), where vague moralizing usually collapses.
You document pivots and keep returning to the claim, which makes evasion obvious to the audience.

Below is a concise but thorough map you can use mid-debate. It separates (A) what moral sentences mean (semantics) from (B) what moral facts (if any) are like (metaphysics) and (C) how we know them (epistemology). Your questions center on (A).

Semantics of moral language

1) Cognitivism vs. non-cognitivism

Cognitivism (moral sentences are truth-apt claims)

Core thesis: When someone says “Lying is wrong,” they assert a proposition that can be true or false.
Why people adopt it (semantic evidence):

Embedding/logic: Moral claims appear in complex constructions that presuppose truth values: conditionals, negations, quantifiers.
- Example: “If lying is wrong, then teaching children to lie is wrong.” This looks like ordinary truth-functional reasoning.
Disagreement looks factual: “Lying is wrong” vs “Lying isn’t wrong” seems like a genuine contradiction, not just clashing attitudes.
Assertoric force & inference: We argue from and to moral claims the way we do with non-moral claims.
Main variations:
Non-naturalist realism (Moore, Parfit): Moral terms refer to irreducible normative properties; “goodness” is not identical to any natural property (Open Question Argument).
Naturalist realism (Cornell realists): Moral terms refer to natural/response-dependent properties (e.g., flourishing, well-being, cooperation).
Constructivism (Korsgaard/Rawls): Truth of moral claims depends on what would be justified by a certain procedure (e.g., what free and equal agents could not reasonably reject). Still propositional/truth-apt.
Subjectivism/relativism: Truth of moral claims is indexed to speakers, cultures, or standards (still truth-apt, but relative).
Pressure points:
Queerness worries (metaphysical, but often raised): what kind of property is “wrongness”?
Motivational internalism challenge: If moral judgments are beliefs, why do they (often) motivate? Cognitivists answer via desire-belief models or “besire”/hybrid accounts.

Non-cognitivism (moral sentences primarily express non-truth-apt attitudes/commands)

Core thesis: “Lying is wrong” functions like expressing disapproval (“Boo lying!”) or issuing a prescription (“Don’t lie!”). It’s not (primarily) stating a fact.
Why people adopt it (practical-language evidence):

Action-guiding character: Moral talk seems to nudge, guide, and commit to action.
Motivation link: Sincere moral judgment often comes with motivation; attitudes/imperatives naturally explain that.
Disagreement as clash of attitudes: Disputes look like conflicts in non-cognitive states (attitudes, plans).
Main variations:
Emotivism (Ayer/Stevenson): Moral utterances express approval/disapproval and attempt to influence others.
Prescriptivism (R. M. Hare): Moral language issues universalizable prescriptions (“Don’t lie” for all relevantly similar cases).
Expressivism / Quasi-realism (Blackburn/Gibbard): Moral claims express plan-like or normative attitudes, but we can project a realist-looking discourse (truth talk, inference) without committing to moral facts.
Classic challenge and replies:
Frege–Geach problem: If moral sentences lack truth conditions, how can they behave in embedded contexts (e.g., “If lying is wrong, …”) and preserve logical validity?
- Modern expressivist reply: Give compositional semantics for attitudes (e.g., sets of plans/norms) so embedding works; adopt minimalist/deflationary truth to recover logical form.

Hybrid views

Hybrid expressivism/cognitivism (Ridge and others): Moral sentences simultaneously express a belief (truth-apt) and a non-cognitive attitude. This aims to capture logical behavior and close motivational links.

2) The meaning of “wrong/ought”

A. Stance-independent (categorical) vs. instrumental (hypothetical) “ought”

Categorical/moral ‘ought’: Not conditional on any contingent goal. “You ought not torture” binds regardless of what you want.
- Semantics: Purports to state a reason that does not derive from the agent’s current aims.
- Goes naturally with: Moral realism (naturalist or non-naturalist) or constructivism that yields categorical reasons.
Instrumental/hypothetical ‘ought’: Conditional on an end. “If you want to pass the exam, you ought to study.”
- Semantics: Means-end normativity; truth depends on facts about effectiveness for the goal.
  Two useful refinements:
Narrow- vs. wide-scope ought:
- Narrow: “If you intend G, you ought to do A.”
- Wide: “You ought: (if you intend G, then do A)” (so revising the intention can also satisfy the ought).
Objective vs. subjective/evidential ought:
- Objective/fact-relative: Based on actual outcomes/facts.
- Subjective/evidence-relative: Based on what your evidence supports.

B. “Wrong”

On a cognitivist reading, “φ is wrong” asserts a proposition with truth conditions. Competing analyses:

Deontic/rights-violation analysis: φ is wrong iff φ violates justified claims (rights/duties).
Harm-based analysis: φ is wrong iff it imposes unjustified harm or risk.
Constructivist analysis: φ is wrong iff φ could not be justified to others under fair conditions (e.g., fails a reasonable-rejectability test).
Relativist analysis: φ is wrong relative to a standard S.
On a non-cognitivist/expressivist reading:
“φ is wrong” expresses a state like disapproval under a system of norms, with logical roles captured by sophisticated attitude semantics. “Truth” enters via deflationary devices that let us keep ordinary logical talk while the core psychological function remains expressive/prescriptive.

C. Thin vs. thick terms; supervenience

Thin terms: right, wrong, ought, good, bad—highly abstract.
Thick terms: cruel, cowardly, generous—mix descriptive content with evaluative force.
Supervenience (semantic corollary): If two acts are non-morally identical in all respects, they cannot differ in moral status. This is a widely accepted constraint on any adequate semantics.

3) Quick diagnostics you can use live

Test 1: Embedding

If your opponent’s semantics can’t handle “If lying is wrong, encouraging lying is wrong,” their account is in trouble (Frege–Geach).
Cognitivists: straightforward. Expressivists: must show a compositional attitude semantics (quasi-realists do).
Test 2: Disagreement
If “Lying is wrong” vs “Lying isn’t wrong” is mere attitude conflict, how do we explain the apparent contradiction and standard argument patterns?
Cognitivists: contradiction is literal. Expressivists: disagreement in plans/norms can mimic contradiction via advanced semantics.
Test 3: Motivation
If moral judgment is purely belief, why the tight link to motivation?
Non-cognitivists: easy. Cognitivists: add desire-like states, besires, or internalist theses about rational agency.

4) How to answer Andrew’s two recurring questions crisply

Q1. “Are moral claims truth-apt (facts) or just attitudes/commands?”

Debate-ready answer options:

Cognitivist (clean and fast):
“I’m a cognitivist: ‘wrong’ and ‘ought’ are truth-apt. The embedding behavior, contradiction, and our inferential practices support this. If you prefer expressivism, use a version that solves Frege–Geach. Either way, our shared logical inferences still go through—so let’s apply them to promiscuity.”
Quasi-realist/expressivist (if you want motivational oomph):
“I’m an expressivist of the quasi-realist kind: moral claims express plan-like attitudes but we recover truth-talk and valid inference via a compositional semantics. That preserves motivation and logic. With that secured, what non-speculative reason makes promiscuity wrong?”
Either answer neutralizes the “we must do metaethics first” stall, because both routes preserve valid inference and disagreement.

Q2. “Do ‘wrong/ought’ pick out stance-independent reasons, or are they merely instrumental?”

Crisp categorical stance (recommended for moral debates):
“By ‘wrong’ and the moral ‘ought’ I mean categorical reasons—stance-independent in the sense that they don’t depend on anyone’s contingent aims. Instrumental ‘oughts’ are fine for means–ends talk, but your claim that promiscuity is wrong is a categorical claim, so it needs a categorical reason (e.g., a rights violation or unjustified harm). Please specify which and by what mechanism.”
If pressed on categories quickly:

Objective vs. evidential ought: “I’m using the evidential should in deliberation and the objective should in assessment; either way, the burden is to show a decisive moral reason.”
Wide- vs. narrow-scope: “I’m using wide-scope to avoid bootstrapping; you can satisfy the ought by revising either the action or the aim. For moral wrongs, the relevant ought is categorical and wide-scope.”

5) Short examples that illustrate the distinctions

Instrumental ought: “If you want cardiovascular health (goal), you ought to run (means).” False if running won’t help, true if it will—wholly goal-relative.
Categorical moral ought: “You ought not assault people.” Purports to bind regardless of your goals; we evaluate by rights/harm/constructivist tests.
Cognitivist reading: “Assault is wrong” is true if assault violates justified rights; false if (counter-intuitively) it didn’t.
Expressivist reading: “Assault is wrong” expresses a conclusive plan against assault; sophisticated semantics explains why “If assault is wrong, hiring assailants is wrong” is a valid inference.

6) A compact, debate-ready script

Set the semantics in one breath:
“We can proceed either way. Cognitivist: moral claims are truth-apt; logic and disagreement work normally. Quasi-realist: moral claims express plan-like attitudes but we have a compositional semantics that preserves validity and truth-talk. In both cases, ‘wrong’ in moral contexts is categorical, not merely instrumental.”
Force the application:
“Given that, your claim ‘promiscuity is wrong’ must identify a categorical reason—typically a rights violationor unjustified harm. Which is it, and via what mechanism? If it’s only instrumental (‘bad for certain goals’), that’s a prudential ought, not a moral wrong.”
This gives you the semantic scaffolding and the on-ramp back to substance—without getting trapped in an endless pre-game over terms.

A clear, big-picture map of moral philosophy

Below is a plain-English overview that organizes the major pieces—so the terms stop blurring together and you can see how they fit. We’ll move from the “what are we even doing?” level down to concrete moral theories, and finish with a short checklist you can use in any debate.

1) The three layers of moral philosophy

Metaethics (Foundations)
- Asks: What do moral words mean? Are there moral facts? How could we know them? Why do they have authority over us?
- Keywords: cognitivism, non-cognitivism, realism, anti-realism, constructivism, internalism, externalism, categorical vs. hypothetical oughts.
Normative ethics (Frameworks)
- Asks: What general principles determine right and wrong?
- Major families: consequentialism, deontology, virtue ethics, contractualism, care ethics, rights theories, pluralism.
Applied ethics (Cases)
- Asks: Given a framework (or multiple), what should we do about concrete issues—e.g., lying, abortion, war, punishment, sexuality?
  Think: metaethics = the grammar, normative ethics = the playbook, applied ethics = calling plays on the field.

2) Key metaethical terms (with crisp explanations)

A. Semantics of moral language

Cognitivism: Moral sentences state propositions that can be true or false.
- “Lying is wrong” aims to say something true or false about lying.
Non-cognitivism: Moral sentences mainly express attitudes/commands.
- “Lying is wrong” functions like “Boo lying!” or “Don’t lie!”
Expressivism / Quasi-realism: A refined non-cognitivism—moral talk expresses attitudes but is engineered to behave as if truth-apt so we can still argue, negate, and embed moral claims in conditionals.
Frege–Geach problem (why it matters): Moral sentences appear inside “if…then…”, “not…”, etc. Any view must explain how moral logic works there.

B. Metaphysics of morality (what moral facts would be like)

Realism: There are stance-independent moral facts (naturalist or non-naturalist).
Anti-realism:
- Constructivism: Moral truths are what would be justified by an ideal procedure (e.g., what free and equal agents could not reasonably reject).
- Relativism: Truth varies by person/culture/standard.
- Error theory: All positive moral claims are systematically false (there are no moral facts of the kind we talk as if there were).

C. Normativity & reasons (why “should” bites)

Categorical ought: Binds regardless of your goals (“You ought not torture”).
Hypothetical (instrumental) ought: Goal-conditioned (“If you want health, you ought to exercise”).
Internalism vs. externalism about reasons:
- Internalism: A genuine reason must connect to an agent’s motivations (or idealized motivations).
- Externalism: Reasons can apply to you even if you’re not currently moved by them.
Pro tanto vs. all-things-considered:
- Pro tanto: A real but defeasible reason.
- All-things-considered: The final verdict after weighing reasons.

D. Epistemology (how we’d know moral truths)

Reflective equilibrium: We adjust principles and particular judgments until they cohere.
Methodological pluralism: Use arguments from cases, principles, consequences, virtues, rights—see what survives scrutiny.

E. Other useful semantic tools

Thin vs. thick terms: Right, wrong, ought (thin) vs. cruel, generous, deceitful (thick = descriptive + evaluative).
Supervenience: No moral difference without some non-moral difference. (Two acts exactly alike in all descriptive facts can’t differ morally.)

3) The main normative frameworks (how people argue “right/wrong”)

Consequentialism (esp. utilitarianism)
- Right action = the one with the best outcomes (often, maximum well-being).
- Strength: Clear decision rule; sensitive to harm/benefit.
- Pressure point: Can permit troubling acts if they increase net utility.
Deontology (Kantian and rule-based)
- Right action = respects duties/constraints (e.g., don’t use people merely as means; universalize your maxims).
- Strength: Protects rights/constraints even against utility trade-offs.
- Pressure point: What to do when duties conflict?
Virtue ethics (Aristotelian lineage)
- Right action = what a virtuous person would characteristically do; aim at flourishing (eudaimonia).
- Strength: Integrates character, context, and human development.
- Pressure point: Less algorithmic; depends on contested ideals of virtue.
Contractualism / Contractarianism
- Right action = what no one could reasonably reject (Scanlon) or what rational agents would agree to (Rawls), or what self-interested parties would bargain for (Gauthier).
- Strength: Centers justification to each person.
- Pressure point: Disagreement over “reasonable rejection” and bargaining baselines.
Rights theories
- Right action = respects claim-rights (bodily integrity, property, association, etc.).
- Strength: Protects individuals against majoritarian harms.
- Pressure point: Which rights, why those, how to weigh conflicts?
Care ethics
- Right action = sustains responsive, non-dominating relationships; attends to dependency and context.
- Strength: Captures moral salience of care, vulnerability, and partiality.
- Pressure point: Generalizing care norms to large, impersonal institutions.
Pluralism (e.g., W. D. Ross)
- Several basic duties (fidelity, non-maleficence, beneficence, justice, etc.); no single master principle.
- Strength: Realistic about moral complexity.
- Pressure point: How to adjudicate when prima facie duties conflict.

4) What “wrong” and “ought” usually mean in debate

When people say “X is wrong”, they typically mean a categorical, all-things-considered verdict: there are decisive reasons not to do X, often because it violates a justified right or imposes unjustified harm.
When people say “you ought to do Y” in everyday advice, they often mean a hypothetical/instrumental claim: given your goal, Y is a good means.
Many arguments go off the rails when someone makes a categorical moral claim but can only supply instrumental reasons (e.g., “It’s wrong because it won’t help you achieve goal G”). That supports a prudential recommendation, not a moral prohibition.

5) How these pieces interact in real debates

Semantics gates the game: If your opponent insists moral talk is “just feelings,” ask how they account for logical embedding and contradiction. If they’re expressivist, they need the sophisticated version that preserves logic.
Burden of proof: A categorical claim needs a categorical reason. Typically, that means one of:
1. Rights-violation (identify the right and the mechanism of violation), or
2. Unjustified harm (specify the harm, why it’s not consented to, preventable, or outweighed).
Cross-framework robustness: The best arguments give reasons that would look compelling from multiple frameworks (e.g., a practice causes coercion and predictable social harms and erodes key virtues).

6) A simple, reusable checklist for evaluating any moral claim

Clarify the claim type
- Is it categorical (“wrong, period”) or instrumental (“bad for your goals”)?
Fix meanings quickly
- “Wrong” = decisive all-things-considered reason not to do it.
- “Ought” = categorical (moral) or hypothetical (means-ends)—which is being used?
Ask for a reason
- Is it: rights-based, harm-based, duty-based, virtue-based, contractualist, or consequentialist?
- What is the mechanism (how, concretely, does the harm/violation occur)?
Test scope and evidence
- Is it pro tanto (usually a reason) or all-things-considered (decisive)?
- Are there countervailing reasons (consent, prevention, regulation, context)?
Check generalizability and consistency
- Would the principle still hold under relevantly similar cases?
- Does it contradict other principles they accept?
Look for framework convergence
- Do multiple theories point in the same direction? That’s a robustness sign.

7) A quick “translation” guide for common confusions

“It’s wrong because I don’t like it.” → Attitude, not a reason. Ask for a rights/harm/virtue/contractualist basis.
“It’s wrong because it won’t help you achieve your goals.” → That’s an instrumental “ought,” not a moral one.
“It violates rights.” → Which right, precisely? Where does that right come from? How does this action violate it?
“It causes harm.” → What harm, to whom, by what mechanism, with what probability, and are there less-restrictive ways to prevent it?

8) If you want a one-paragraph summary to keep handy

Moral philosophy has three layers: metaethics (what moral talk means and whether there are moral facts), normative ethics (general principles of right action), and applied ethics (particular cases). Semantically, cognitivists treat moral sentences as truth-apt; non-cognitivists/expressivists treat them as expressing attitudes but must still explain valid moral inference. Metaphysically, realists think there are stance-independent moral facts; constructivists/relativists/error theorists deny or reframe that. Normatively, major frameworks—consequentialism, deontology, virtue ethics, contractualism, rights, care, pluralism—offer different lenses. In debate, pin down whether the claim is categorical or instrumental; demand a specific rights or harm mechanism (or a clear duty/virtue/contractualist principle); weigh pro tanto vs. all-things-considered reasons; and look for cross-framework robustness.

Quick starter pack (read in this order)

Alexander Miller — Contemporary Metaethics: An Introduction
A clear, even-handed survey of all the major views (realism, expressivism, error theory, constructivism) and the classic problems (Frege–Geach, internalism/externalism). Ideal on-ramp.
James Dreier — Ethics and Metaethics: An Introduction (or his earlier Metaethics: A Contemporary Introduction)
Complements Miller with crisp distinctions and argument maps; very readable.
e (eds. Tristram McPherson & David Plunkett)
State-of-the-art overviews by top scholars; use it as a reference to deepen on any subtopic.
Russ Shafer-Landau — Moral Realism: A Defence
Canonical case for robust moral facts; sharp on objections to realism and the “queerness” worry.
Simon Blackburn — Ruling Passions (or Essays in Quasi-Realism)
The flagship expressivist/quasi-realist view that preserves logic and disagreement while treating moral talk as attitude-expressing.
J. L. Mackie — Ethics: Inventing Right and Wrong
Classic, punchy case for error theory (“queerness” and relativity arguments). Even if you disagree, it clarifies what realism must answer.

Focused books by issue/position

Semantics of moral language (truth-apt claims vs. attitudes; Frege–Geach; meaning of “ought”)

Allan Gibbard — Thinking How to Live (and Wise Choices, Apt Feelings)
Norm-expressivism with a rigorous compositional semantics; direct engagement with Frege–Geach.
Mark Schroeder — Noncognitivism in Ethics
The best single-author map of modern expressivism and its semantic machinery.
Stephen Finlay — Confusion of Tongues: A Theory of Normative Language
A linguistic-pragmatics account of “ought,” “good,” and reasons as end-relational; superb on instrumental vs. categorical readings.
Michael Smith — The Moral Problem
Anchors the debate on judgment internalism/externalism, motivation, and how moral belief relates to desire.

Moral realism and reasons (robust, stance-independent “ought”)

David Brink — Moral Realism and the Foundations of Ethics
Classic statement of naturalist realism with sophisticated replies to relativism and internalism.
Derek Parfit — On What Matters, Vols. 1–2 (start with the Introductions and “Reasons and Rationality”)
A grand unifying attempt across deontology, consequentialism, and contractualism; deep on reasons and normativity.
Ralph Wedgwood — The Nature of Normativity
Metaphysics and semantics of normative reasons; dense but rewarding.
David Enoch — Taking Morality Seriously: A Defense of Robust Realism
Accessible, forceful case for stance-independent moral truths and why we should believe in them.
Terence Cuneo — The Normative Web
If you accept epistemic norms, Cuneo argues, moral norms come along for the ride.

Non-cognitivism / expressivism (attitudes, prescriptions, quasi-realism)

Allan Gibbard — Thinking How to Live (listed above)
Simon Blackburn — Ruling Passions (listed above)
Michael Ridge — Impassioned Belief
A hybrid expressivist view that combines belief and desire-like states to explain motivation and logical behavior.

Error theory / fictionalism

J. L. Mackie — Ethics: Inventing Right and Wrong (listed above)
Richard Joyce — The Myth of Morality
Contemporary, sophisticated defense of error theory; also explores moral fictionalism as a practical stance.

Constructivism / contractualism (justification by procedure rather than stance-independent facts)

Christine Korsgaard — The Sources of Normativity
Why reasons bind us: a Kantian constructivist account of authority and obligation.
T. M. Scanlon — What We Owe to Each Other
Contractualism in its canonical form: wrong acts are those not justifiable to others on grounds they could not reasonably reject.
John Rawls — Political Liberalism (Lectures on Moral Philosophy also helpful)
Not purely metaethics, but foundational for constructivist method and justification to free and equal persons.

Reasons, “ought,” and decision theory cross-overs

Mark Schroeder — Slaves of the Passions
A Humean theory of reasons grounded in desires that still does serious normative work.
Joseph Raz — Practical Reason and Norms
Reasons, authority, and how norms structure deliberation.
Niko Kolodny — papers on rational requirements (if you dip into articles later): wide- vs. narrow-scope “ought,” rationality, and coherence.

Pluralism, particularism, thick concepts

W. D. Ross — The Right and the Good
Prima facie duties (fidelity, non-maleficence, beneficence, justice, etc.); a pluralist alternative to single-principle theories.
Jonathan Dancy — Moral Reasons
Particularism: moral relevance is highly context-sensitive; resists fixed hierarchies of reasons.
Philippa Foot — Natural Goodness
Bridges virtue and naturalistic accounts of evaluation; helpful for thick terms (cruel, courageous, etc.).

Anthologies and handbooks (for targeted deep dives)

The Oxford Handbook of Metaethics (McPherson & Plunkett, eds.)
The Routledge Handbook of Metaethics (Guy Fletcher, ed.)
Oxford Studies in Metaethics (annual series, Russ Shafer-Landau, ed.) — advanced but gold-standard snapshots of current debates.

A sensible reading path (8–10 weeks)

Phase 1 (foundations):
Miller (intro) → Dreier (intro) → Mackie (error theory)
Phase 2 (two big camps):
Shafer-Landau (realism) ↔ Blackburn or Gibbard (expressivism)
Phase 3 (reasons and ought):
Smith (The Moral Problem) → Finlay (Confusion of Tongues) → Schroeder (Slaves of the Passions)
Phase 4 (constructivist alternative):
Korsgaard (Sources of Normativity) → Scanlon (What We Owe to Each Other)
Phase 5 (pluralism/particularism, optional):
Ross (The Right and the Good) → Dancy (Moral Reasons)
Use the handbooks as a reference alongside any phase when you hit a knotty topic (Frege–Geach, supervenience, internalism vs. externalism, categorical vs. instrumental “ought,” etc.).

How to read them efficiently (quick checklist)

Pin the claim type: Are they defending truth-apt moral facts, expressive attitudes, or a constructivist procedure?
Track “ought”: Categorical or hypothetical? Wide- or narrow-scope? Fact-relative or evidence-relative?
Note their semantics: How do they handle embedding (if…then, not, believe-that)?
Watch motivation: Do moral judgments necessarily motivate (internalism) or only sometimes (externalism)?
Log the burdens: What must their opponents show to defeat them (e.g., queerness, debunking explanations, disagreement)?
Extract test cases: How would their view handle lying to save a life, promise-keeping, sexual ethics, punishment, distributive justice?
If you want, I can turn this into a printable reading plan with checkboxes and one-page chapter guides for the first three books.

Analysis of Andrew Wilson Debate (Moral Language)