Meta-ethics

The meta-ethical map#

FoundationAQA 7172 3.2 Moral philosophy — Meta-ethicsDfE GCE Philosophy subject content

The meta-ethical taxonomy

Fig. 1Cognitivism divides into realism (naturalism, non-naturalism) and error theory; non-cognitivism into emotivism and prescriptivism.

Meta-ethics is second-order: it does not ask 'is euthanasia wrong?' but 'what are we DOING when we say euthanasia is wrong — stating a fact, or expressing an attitude — and could such a claim be true?'. Two cross-cutting distinctions organise the field. The first is between cognitivism and non-cognitivism, a thesis about moral LANGUAGE and psychology: cognitivism holds that moral judgements express beliefs and are truth-apt (capable of being true or false, aiming to describe how things morally are); non-cognitivism holds that moral judgements do NOT express truth-apt beliefs but some other mental state — an emotion, an attitude, a prescription — so they are not, strictly, true or false.

The second distinction is between moral realism and moral anti-realism, a thesis about moral FACTS and PROPERTIES: realism holds that there are mind-independent moral facts or properties that make moral judgements true; anti-realism denies this — either because there are no moral facts at all, or because moral 'facts' depend on our attitudes. These two distinctions generate the main positions. Moral realism is cognitivist AND holds some moral judgements are true in virtue of real moral facts: it divides into naturalism (moral properties are natural properties) and non-naturalism (moral properties are real but sui generis, non-natural — Moore's intuitionism).

Anti-realism divides according to whether it is cognitivist or non-cognitivist. Error theory (Mackie) is cognitivist but anti-realist: moral judgements DO make truth-apt claims (they assert the existence of objective moral facts), but there are no such facts, so all positive moral judgements are systematically FALSE — moral discourse rests on a massive error, like discourse about witches. Non-cognitivism is anti-realist by being non-cognitivist: since moral judgements do not even attempt to state facts, the question of mind-independent moral facts does not arise; the two main forms are emotivism (moral judgements express and arouse emotion — Ayer) and prescriptivism (they express universalisable prescriptions — Hare).

Holding the map steady is the discipline that keeps a meta-ethics essay precise. A useful test: does the theory say moral judgements are truth-apt (cognitivist) or not (non-cognitivist)? And does it say there are real moral facts (realist) or not (anti-realist)? Naturalism = cognitivist + realist (facts are natural); non-naturalism = cognitivist + realist (facts are non-natural); error theory = cognitivist + anti-realist (no facts, so all false); emotivism/prescriptivism = non-cognitivist + anti-realist (not truth-apt). Every argument in the topic is best read as attacking or defending one cell of this grid, and the master questions — is moral language truth-apt, and can moral judgements be true? — are the axes of the whole debate.

Worked example

Classifying a meta-ethical claim

'When I say stealing is wrong, I am not stating a fact but expressing my disapproval.' Classify this on both distinctions.

Truth-apt?
The claim denies that the judgement states a fact; it expresses disapproval — so it is non-cognitivist.
Which non-cognitivism?
Expressing disapproval/emotion is emotivism ('boo to stealing'); if it prescribed ('do not steal, everyone'), it would be prescriptivism.
Realist?
Non-cognitivism does not posit moral facts, so it is anti-realist.
Locate on the grid
Non-cognitivist + anti-realist = emotivism.

Result: The claim is emotivism: non-cognitivist (not truth-apt) and anti-realist (no moral facts) — a clear illustration of using the two distinctions to place a position.

Typical mistakes

Conflating non-cognitivism with anti-realism or with relativism; non-cognitivism is specifically the denial that moral judgements are truth-apt beliefs.
Classing error theory as non-cognitivist; it is cognitivist (moral claims are truth-apt assertions) but anti-realist (there are no moral facts, so the assertions are false).

Active revision

Explain the difference between cognitivism and non-cognitivism, and between moral realism and anti-realism, with an example of a theory in each category. (5-mark AO1 response)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Ethical naturalism and Moore's open-question argument#

StandardAQA 7172 3.2 Moral philosophy — Meta-ethicsDfE GCE Philosophy subject content

The open-question argument

Fig. 2For any natural definition of good, 'is the N good?' stays open, so good is not defined by N.

Ethical naturalism is the realist, cognitivist view that moral properties ARE natural properties — properties of the kind investigated by science or discoverable by ordinary observation — so that moral facts are a species of natural fact, and moral judgements can be true or false and known a posteriori. Utilitarianism is the paradigm: it identifies goodness with a natural property, the maximisation of pleasure or happiness, so that 'X is good' just means 'X maximises happiness', a claim about the natural world. Naturalism's attractions are considerable: it makes moral knowledge continuous with empirical knowledge, explains how moral claims can be true and reasoned about, and avoids positing any mysterious non-natural facts. If goodness is just a natural property, ethics is, in principle, as tractable as any other inquiry into how the world is.

G. E. Moore, in Principia Ethica (1903), argued that all such definitions commit the 'naturalistic fallacy': the mistake of trying to DEFINE the non-natural property goodness in terms of any natural property. His central argument is the open-question argument. Take any proposed naturalistic definition — 'good = pleasant', or 'good = what we desire to desire', or 'good = conducive to evolutionary fitness'. Now ask: 'this is pleasant, but is it good?'. Moore observes that this question is always OPEN — intelligible, substantive, not settled by the meanings of the words — whereas 'this is a triangle, but does it have three sides?' is CLOSED (a trivial, closed question). If 'good' really MEANT 'pleasant', then 'is the pleasant good?' would be as trivial as 'is a triangle three-sided?'; since it is not, 'good' does not mean 'pleasant', and by parity does not mean any natural property. Goodness is therefore indefinable and non-natural.

The open-question argument is meant to refute naturalism quite generally, not just one version: for ANY natural property N, 'X is N, but is X good?' remains an open question, so 'good' is never synonymous with 'N'. Moore concludes that goodness is a simple, unanalysable, non-natural property — simple in the way 'yellow' is simple (it cannot be broken into parts) but non-natural in that it is not the object of the senses. This clears the ground for his own non-naturalist intuitionism (next section): if good cannot be defined, it must be known some other way — by a kind of rational intuition.

The open-question argument has been influential but is now widely thought to prove less than Moore claimed. First, it seems to assume that if two terms are not SYNONYMOUS then they cannot refer to the same PROPERTY — but 'water' and 'H2O' are not synonymous (one can competently ask 'this is water, but is it H2O?'), yet water just IS H2O; likewise goodness might be a natural property (pleasure, say) even though 'good' and 'pleasant' are not synonymous, a synthetic a posteriori identity. This is the reply of 'Cornell realists' and other sophisticated naturalists, who identify moral properties with natural ones without claiming a definitional analysis. Second, the argument arguably begs the question: whether 'is the pleasant good?' is really open is exactly what is at issue, and a convinced naturalist may find it closed. A balanced verdict is that the open-question argument refutes ANALYTIC (definitional) naturalism — the claim that moral terms MEAN natural terms — but not synthetic naturalism, which identifies moral properties with natural properties without a meaning-analysis; so naturalism survives in its sophisticated form, and the debate moves to whether such property-identities can be established.

Worked example

The water/H2O reply to Moore

Explain how the water/H2O analogy blunts the open-question argument.

Moore's assumption
If 'good' does not MEAN 'pleasant', good is not the property pleasure — non-synonymy implies non-identity.
The counterexample
'Water' does not mean 'H2O' ('is water H2O?' was a genuine discovery), yet water IS H2O — a synthetic a posteriori identity.
Apply it
So 'good' could name the same property as 'pleasure' even though the terms are not synonymous; the open question shows only non-synonymy, not non-identity.
The residue
The argument refutes ANALYTIC naturalism (good MEANS N) but not SYNTHETIC naturalism (good IS N), which the sophisticated naturalist holds.

Result: The water/H2O reply shows the open-question argument establishes only that moral terms are not DEFINABLE by natural terms, leaving open that moral PROPERTIES are natural properties — so naturalism survives in its synthetic form.

Typical mistakes

Confusing the naturalistic fallacy (defining good in natural terms) with the is-ought gap (inferring 'ought' from 'is'); they are related but distinct — Hume's gap is about inference, Moore's fallacy about definition.
Treating the open-question argument as decisive; note the water/H2O reply, which lets a naturalist accept the argument yet identify goodness with a natural property non-definitionally.

Active revision

'Moore's open-question argument shows that goodness cannot be a natural property.' Assess this claim. (25-mark essay)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Non-naturalism and intuitionism#

StandardAQA 7172 3.2 Moral philosophy — Meta-ethicsDfE GCE Philosophy subject content

Ross's prima facie duties

Fig. 3A plurality of self-evident duties, weighed by judgement into the 'duty proper' when they conflict.

Non-naturalism agrees with naturalism that there are objective moral facts and that moral judgements are truth-apt (it is realist and cognitivist), but holds that moral properties are NON-natural — sui generis, irreducible to the natural properties of science or ordinary observation. Moore is the classic non-naturalist: goodness is a simple, unanalysable, non-natural property that supervenes on natural properties (two acts alike in all natural respects cannot differ in goodness) but is not identical to any of them. If moral properties are non-natural, they cannot be known by sense perception or empirical science, so non-naturalists standardly hold that basic moral truths are known by a special faculty of rational intuition — hence ethical intuitionism.

Intuitionism is the epistemology that typically accompanies non-naturalism: some moral truths are self-evident and known directly by intuition, not inferred from anything more basic. Moore held that the intrinsic goodness of certain things (personal affection, aesthetic enjoyment) is intuited. H. A. Prichard argued that our sense of obligation is immediate and underived — the question 'why should I do my duty?' is misconceived, because we simply apprehend, on reflection, that certain acts are obligatory. W. D. Ross developed the most sophisticated version with his theory of prima facie duties: there is a plurality of self-evident duties (fidelity, reparation, gratitude, justice, beneficence, self-improvement, non-maleficence), each of which we intuit as binding other things being equal; when they conflict, we must judge which is most stringent in the situation (the 'duty proper'), a judgement that is not itself a matter of intuiting a further rule but of weighing. Ross's view captures the pluralism and defeasibility of everyday moral thought better than monistic theories like utilitarianism or Kantianism.

Non-naturalism and intuitionism have real strengths. They take moral objectivity and truth at face value, honouring the phenomenology of moral experience (we FEEL that cruelty is really wrong, not merely disapproved of); they explain moral knowledge without reducing morality to natural facts (avoiding Moore's fallacy); and Ross's pluralism fits the texture of moral life, where duties genuinely conflict and no single principle settles every case. Intuitionism also provides a foundation — self-evident basic truths — that stops the regress of justification, in the way foundationalism does in epistemology.

The objections are serious. First, the metaphysical worry: non-natural moral properties look queer — unlike anything else in the world — and it is unclear how they fit into a naturalistic universe (this is one prong of Mackie's argument, next section). Second, the epistemological worry: 'intuition' looks like a name for a problem rather than a solution — what is this faculty, how does it work, and how do we tell a genuine intuition from a mere feeling, prejudice, or cultural conditioning? Third, the argument from disagreement: if moral truths were self-evident to intuition, we should expect convergence, yet there is deep and persistent moral disagreement, both between individuals and across cultures, which intuitionism struggles to explain (a defender replies that intuitions can be clouded by bias, self-interest and poor upbringing, as perception can be by bad conditions, and that there is more convergence on basics than the sceptic allows). A balanced evaluation grants that non-naturalism honours moral objectivity and that Ross's pluralism is attractive, but presses whether 'intuition' can bear the epistemic weight placed on it, and whether the queerness of non-natural properties is a price worth paying — questions that lead directly to Mackie's error theory and to non-cognitivism, which try to keep morality's importance without its dubious metaphysics.

Worked example

The argument from disagreement against intuitionism

Assess whether moral disagreement refutes intuitionism.

The objection
If basic moral truths were self-evident to intuition, we should expect wide convergence; but moral disagreement is deep and persistent, so either there are no such truths or intuition is unreliable.
The intuitionist reply
Intuitions can be clouded by self-interest, bias, upbringing and factual error, as perception is distorted by bad conditions; disagreement reflects these distortions, not the absence of moral truth.
Note the convergence
There is in fact wide agreement on basics (cruelty for its own sake is bad, gratuitous betrayal is wrong); much disagreement is over non-moral facts or application, not fundamental principle.
Weigh
The reply blunts but does not remove the worry: it is hard to specify 'ideal conditions' for intuition non-circularly, so the reliability of the faculty remains contestable.

Result: Moral disagreement does not refute intuitionism, since distortions can be invoked as with perception, but it does pressure the intuitionist to explain the intuitive faculty and its conditions of reliability — the topic's central weakness for the view.

Typical mistakes

Treating 'intuition' as mere gut feeling; intuitionists mean a rational apprehension of self-evident truths, and must distinguish it from prejudice — which is exactly the objection to answer.
Confusing non-naturalism (moral properties are real but non-natural) with non-cognitivism (moral judgements are not truth-apt); non-naturalism is a realist, cognitivist view.

Active revision

'Intuitionism cannot explain how we know moral truths.' Discuss. (25-mark essay)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Mackie's error theory#

AdvancedAQA 7172 3.2 Moral philosophy — Meta-ethicsDfE GCE Philosophy subject content

Mackie's argument from queerness

Fig. 4Objective values would be metaphysically and epistemologically queer, so we should reject them.

J. L. Mackie's error theory (Ethics: Inventing Right and Wrong, 1977) is the most striking anti-realist position because it is cognitivist yet holds that all positive moral judgements are FALSE. Mackie agrees with the moral realist about the MEANING of moral discourse: when we say 'cruelty is wrong', we are making a genuine assertion that purports to describe an objective, mind-independent moral fact — moral language is cognitive and truth-apt. But he denies that there are any such facts: the world contains no objective values. So ordinary moral discourse rests on a systematic error, asserting the existence of objective moral properties that do not exist, in the way discourse about witches asserted the existence of witches. Every positive moral claim, being an assertion of a non-existent fact, is false (or untrue).

Mackie supports this with two main arguments. The argument from relativity (or disagreement) observes the wide variation in moral codes across cultures and history and argues that this variation is better explained by supposing that moral judgements REFLECT ways of life and social arrangements — that people approve of the conduct their form of life requires — than by supposing they are (mostly failed) attempts to perceive a single objective moral truth. If there were objective moral facts, we might expect the kind of convergence we find in the sciences; the actual diversity fits the hypothesis that morality is a human institution, not a domain of fact. (The realist replies that scientific disagreement does not disprove scientific facts, and that much moral disagreement is over non-moral facts or is less than it appears.)

The argument from queerness is Mackie's central and more powerful argument, and it has two prongs. Metaphysically, objective values would be 'entities or qualities or relations of a very strange sort, utterly different from anything else in the universe' — non-natural properties that are somehow built into the fabric of the world. Epistemologically, to know such values we would need 'some special faculty of moral perception or intuition, utterly different from our ordinary ways of knowing everything else' — exactly the mysterious intuition non-naturalism appeals to. Moreover, objective values would have to be intrinsically motivating and action-guiding ('to-be-pursuedness' built into them), which no natural fact is: a fact that necessarily motivated anyone who apprehended it would be metaphysically queer. Since we should not believe in such queer entities and faculties without extraordinary evidence, Mackie concludes there are no objective values, and moral discourse, which presupposes them, is in error.

Error theory forces a choice about what to do next, and Mackie is a 'moral abolitionist' only in theory: he proposes we CONTINUE to use moral discourse, reinterpreted or reformed, because morality is a useful human invention for regulating conduct and cooperation, even though its literal claims are false ('fictionalism' about morality). The theory's strengths are its honesty about the metaphysical and epistemological costs of realism (it takes the queerness worry seriously) and its explanatory power regarding disagreement. Its difficulties are several: it is deeply revisionary, requiring us to regard all our confident moral convictions ('the Holocaust was wrong') as literally false, which strikes many as a reductio; the queerness argument arguably proves too much (numbers, logical necessity, and other abstracta are also 'queer' by its standards, yet we accept them, so queerness is not obviously fatal); and non-cognitivists object that Mackie mis-describes moral language as assertoric in the first place — if moral judgements EXPRESS attitudes rather than assert facts, there is no error, because they never claimed to state objective facts. This last objection is the pivot to non-cognitivism: perhaps the way to keep morality without queer facts is not to say it is all false, but to deny it was ever in the fact-stating business.

Worked example

The 'proves too much' reply to queerness

Assess the objection that Mackie's queerness argument proves too much.

The argument
Objective values are metaphysically and epistemologically queer (unlike anything else; needing a special faculty), so we should not believe in them.
The reply
Numbers, sets, logical and modal facts are also abstract, causally inert, and known by no ordinary sense — 'queer' by Mackie's standard — yet we rationally accept them.
The upshot
If queerness does not disqualify mathematics or logic, it should not automatically disqualify morality; queerness is a cost, not a refutation.
Mackie's rejoinder
Moral facts are queerer still, since they must be intrinsically motivating; but the reply shows the argument needs this extra premise, which is itself contestable.

Result: The 'proves too much' reply weakens the queerness argument by showing that abstractness and non-empirical knowledge are widely accepted; Mackie's case then rests on the special claim that moral facts must be intrinsically action-guiding, which is where the real dispute lies.

Typical mistakes

Saying error theory holds moral judgements are neither true nor false; it holds they are truth-apt and (the positive ones) uniformly FALSE — that is what makes it cognitivist.
Reducing queerness to mere strangeness; state both prongs (metaphysical queerness of the properties, epistemological queerness of the faculty, plus intrinsic motivation) for full credit.

Active revision

'There are no objective moral values.' Is this claim true? Defend your answer. (25-mark essay)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Non-cognitivism: emotivism and prescriptivism#

AdvancedAQA 7172 3.2 Moral philosophy — Meta-ethicsDfE GCE Philosophy subject content

Emotivism and prescriptivism compared

Fig. 5Both are non-cognitivist, but emotivism expresses emotion while prescriptivism issues universalisable prescriptions.

Non-cognitivism denies that moral judgements are truth-apt assertions at all: they do not state facts (true or false) but perform some other function — expressing feelings or issuing prescriptions. This dissolves the error theorist's problem (there is no error, because no false factual claim is made) and the realist's problem (no queer facts are needed), while explaining a striking feature of moral judgement that cognitivism struggles with: its intrinsic connection to MOTIVATION. To judge that something is wrong is (typically) to be against it, to be moved to avoid it — whereas merely believing a fact (that grass is green) motivates no one. Non-cognitivism explains this by making the judgement itself the expression of a con-attitude, not a belief about a fact.

Emotivism, associated with A. J. Ayer (Language, Truth and Logic, 1936) and developed by C. L. Stevenson, holds that moral judgements EXPRESS and AROUSE emotion. Ayer arrives at it via the verification principle: a statement is factually meaningful only if analytic or empirically verifiable; moral judgements are neither (no observation verifies 'stealing is wrong'), so they are not factual statements. But they are not nonsense; they are expressions of feeling. 'Stealing is wrong' does not state a fact but evinces disapproval — it is as if one said 'stealing!' in a tone of horror, or 'boo to stealing!'. Hence the 'boo/hurrah' theory: moral judgements express pro- or con-attitudes and function to influence others' feelings and conduct. Stevenson added that they also have 'emotive meaning' and are used to change attitudes through 'persuasive definitions'.

R. M. Hare's prescriptivism (The Language of Morals, 1952) refines non-cognitivism to answer the charge that emotivism reduces morality to mere venting. For Hare, moral judgements are not expressions of feeling but PRESCRIPTIONS — they tell people what to do — and they are UNIVERSALISABLE: to call an act wrong is to prescribe that no one, in relevantly similar circumstances, do it, oneself included. So 'you ought not to lie' entails 'do not lie' (a prescription) and commits me to holding that no one in like circumstances should lie (universalisability), on pain of logical inconsistency. This makes moral discourse rational and rule-governed — susceptible to consistency-checks and to reasoning about whether one can universalise one's prescriptions — while remaining non-cognitivist (prescriptions are not true or false). Prescriptivism thus captures the action-guiding and universal character of moral judgement that emotivism left obscure.

The decisive objection to non-cognitivism is the Frege-Geach (embedding) problem. Moral terms occur not only in simple assertions but in unasserted contexts — in the antecedents of conditionals, in questions, in disjunctions — where nothing is being expressed or prescribed. Consider the valid argument: 'If lying is wrong, then getting your little brother to lie is wrong; lying is wrong; so getting your little brother to lie is wrong.' This is a plain modus ponens, and its validity requires 'lying is wrong' to mean the SAME thing in the first (unasserted, embedded in the 'if') and second (asserted) premises. But if 'lying is wrong' just EXPRESSES disapproval, then in the conditional's antecedent — where no disapproval is expressed — it is doing something different, so the term equivocates and the inference should be invalid; yet it is obviously valid. Non-cognitivism thus seems unable to account for moral reasoning and the meaning of embedded moral claims. Sophisticated non-cognitivists (Blackburn's quasi-realism, Gibbard's norm-expressivism) have laboured to solve the problem by giving a logic of attitudes, with contested success. A balanced evaluation credits non-cognitivism with elegantly explaining moral motivation and avoiding queer facts, but finds the Frege-Geach problem a deep difficulty, and notes the further worry that non-cognitivism struggles to make sense of moral ERROR, PROGRESS and the felt OBJECTIVITY of moral conviction — we do not think 'slavery is wrong' is merely our 'boo', but that those who approved of slavery were mistaken.

Worked example

The Frege-Geach problem in action

Show how the embedding problem challenges emotivism, using a modus ponens argument.

The argument
P1: If lying is wrong, then getting your brother to lie is wrong. P2: Lying is wrong. C: So getting your brother to lie is wrong. (Valid modus ponens.)
Asserted vs embedded
In P2 'lying is wrong' is asserted; in P1 it sits in the antecedent of a conditional and is NOT asserted — no disapproval is being expressed there.
The problem
If the meaning of 'lying is wrong' is the expression of disapproval, it means something different in P1 (no disapproval expressed) and P2 (disapproval expressed), so the argument equivocates and should be invalid — but it is valid.
The demand
Non-cognitivism must give a single meaning that works in both contexts; quasi-realism tries via a logic of attitudes, but its success is disputed.

Result: The Frege-Geach problem shows non-cognitivism struggles to explain valid moral reasoning, because embedded moral claims cannot be expressions of attitude yet must be univocal — a deep difficulty that any credible non-cognitivism must solve.

Typical mistakes

Reducing emotivism to 'morality is just feelings, so anything goes'; Stevenson and Hare stress that moral discourse influences conduct and (for Hare) is rule-governed and universalisable.
Missing the Frege-Geach problem or misstating it; the point is that moral terms in UNASSERTED contexts (antecedents of conditionals) cannot be expressions, yet must mean the same for inferences to be valid.

Active revision

'Non-cognitivism is the best account of moral language.' Discuss. (25-mark essay)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Is-ought and the status of moral language#

AdvancedAQA 7172 3.2 Moral philosophy — Meta-ethicsDfE GCE Philosophy subject content

The is-ought gap and the theories

Fig. 6The gap is welcomed by non-cognitivism and error theory, a challenge for naturalism, and confirmation for non-naturalism.

A final evaluation is organised by two issues that run through the whole topic: the is-ought gap and the relation between moral judgement and motivation. Hume's is-ought observation (Treatise) notes that moralists slide, without argument, from premises joined by 'is' to conclusions joined by 'ought', and asks how this new relation can be deduced from premises that do not contain it. Read strictly, the point is logical: no valid deductive argument can have an evaluative conclusion from purely factual premises, since that would introduce content not in the premises. This poses a challenge to naturalism (which tries to derive or identify 'ought' with natural 'is'-facts) and is often paired with Moore's naturalistic fallacy, though the two are distinct — Hume's is about INFERENCE, Moore's about DEFINITION.

The is-ought gap cuts in different directions for different theories. For non-cognitivists it is welcome: it shows that moral conclusions are not factual conclusions at all, but expressions or prescriptions, which is exactly their view. For error theorists it supports the claim that morality's fact-stating pretensions are unfounded. For naturalists it is a challenge to be met, either by denying that morality requires DEDUCING ought from is (the naturalist identifies moral with natural properties rather than inferring across a gap) or by arguing, with Searle and others, that some institutional facts ('he promised') do entail evaluative conclusions ('he ought to keep it'), so the gap is not absolute. For non-naturalists the gap shows precisely that the moral (ought) is a distinct, non-natural realm not reducible to the natural (is).

The motivation issue (moral internalism vs externalism) is the other organising thread. Internalism holds that there is a NECESSARY connection between sincerely judging that one ought to do something and being motivated to do it — you cannot fully judge that you ought to give to charity and be entirely unmoved. Non-cognitivism explains internalism effortlessly: since moral judgements ARE (or express) pro/con-attitudes, they are motivating by their very nature. Cognitivism has more difficulty: if a moral judgement is a belief about a fact, and beliefs alone do not motivate (Hume: reason is 'the slave of the passions'), then how can moral judgement motivate? Realists respond either by externalism (the connection to motivation is contingent — the 'amoralist' who sincerely judges an act wrong yet is unmoved is possible) or by positing special motivating moral facts (which returns us to Mackie's queerness). So the very feature that makes non-cognitivism attractive (it explains motivation) is a pressure point for realism.

For an A-Level judgement, the defensible conclusion weighs the trade-offs each position makes. Realism (naturalist or non-naturalist) honours the objectivity, truth-aptness and rational discipline of moral thought — that we argue about morality, think we can be wrong, and take moral progress seriously — but must answer the metaphysical and epistemological queerness of moral facts (non-naturalism) or the open-question and is-ought worries (naturalism), and must explain moral motivation. Anti-realism (error theory or non-cognitivism) avoids queer facts and, in the non-cognitivist case, explains motivation and the is-ought gap elegantly, but must answer the Frege-Geach problem and account for moral error, disagreement-as-genuine-dispute, progress, and the phenomenology of objectivity (that 'cruelty is wrong' feels like a discovery, not a decision). A candidate might reasonably conclude for a sophisticated naturalism or a quasi-realist non-cognitivism, but what earns the marks is naming the criterion (best explains the whole practice of moral thought and talk), applying it, and defending a placement while acknowledging the strongest rival.

Meta-ethics closes the moral philosophy unit by turning the tools of the whole course on morality itself. The cognitivist/non-cognitivist question is the same shape as the debate over religious language (are religious statements truth-apt or non-cognitive expressions of a form of life?), and Ayer's verification principle is the shared engine of both. The is-ought gap and the analytic/synthetic distinction connect straight to the epistemology unit's a priori/a posteriori and analytic/synthetic distinctions. And the reliability of moral intuition mirrors the reliability of rational intuition in the rationalism debate. Meta-ethics is therefore not an isolated add-on but the point at which epistemology, philosophy of language and ethics converge.

Worked example

Model 25-mark plan: are moral judgements truth-apt?

Plan an essay defending a judgement on whether moral judgements can be true or false.

Thesis
Moral judgements are truth-apt (cognitivism is right), best defended by a sophisticated naturalism, though non-cognitivism's motivation point must be accommodated.
AO1
Map cognitivism/non-cognitivism and realism/anti-realism; state naturalism, non-naturalism, error theory, emotivism, prescriptivism.
For cognitivism
Moral thought behaves like fact-stating: we argue, think we can err, recognise progress; the Frege-Geach problem shows non-cognitivism struggles to explain moral reasoning.
Against
Non-cognitivism explains motivation and avoids queer facts; the is-ought gap and open question challenge naturalism.
Judgement
Favour synthetic naturalism (moral properties are natural properties, non-definitionally) or quasi-realism, on the criterion of best explaining the whole practice; defend the placement and concede the motivation worry.

Result: A defensible judgement reaches a defended placement (here, cognitivism via sophisticated naturalism or quasi-realism) on the criterion of best explaining moral thought, talk, reasoning and motivation — not a bare 'both have a point'.

Typical mistakes

Equating the is-ought gap with the naturalistic fallacy; Hume's gap is about deducing 'ought' from 'is' (inference), Moore's fallacy about defining 'good' in natural terms (definition).
Concluding 'it is all subjective' from anti-realism; non-cognitivists like Hare insist moral discourse is rule-governed and rational, and quasi-realists recover talk of moral truth and error.

Active revision

'Moral judgements cannot be true or false.' Discuss. (25-mark essay)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Key points

Classifying a meta-ethical claim

Key points

The water/H2O reply to Moore

Key points

The argument from disagreement against intuitionism

Key points

The 'proves too much' reply to queerness

Key points

The Frege-Geach problem in action

Key points

Model 25-mark plan: are moral judgements truth-apt?

Sources

Meta-ethics

Key points

Classifying a meta-ethical claim

Key points

The water/H2O reply to Moore

Key points

The argument from disagreement against intuitionism

Key points

The 'proves too much' reply to queerness

Key points

The Frege-Geach problem in action

Key points

Model 25-mark plan: are moral judgements truth-apt?

Sources