1. Scope and evidence
Established: That Hume wrote the passage, that it flags an unexplained transition from "is" to "ought," and that it became the historical source of the fact/value distinction. That Moore's naturalistic fallacy (1903) is a later, separate argument by a different philosopher with a different target. Shared/Traditional: The "orthodox" twentieth-century reading of the passage as a universal logical thesis ("no ought from an is"). This is the standard textbook interpretation, but it is a reading, not an uncontested transcription of Hume's intent. Disputed: (1) The scope-of-reading debate — whether Hume asserted a universal logical law or a narrower, contextual critique of specific rationalist/religious moralists (MacIntyre's challenge). (2) The status of Searle's counterexample — whether it genuinely derives an "ought" from an "is" or covertly smuggles in a normative premise. Both are live, unresolved debates. Flagged error: The naturalistic-fallacy conflation — treating Hume's is-ought problem and Moore's naturalistic fallacy as the same thing. This is common in casual discourse and is, by the standards of careful scholarship, a mistake.
2. The exact passage and its immediate context
Hume's celebrated paragraph closes the first section of Book 3 of the Treatise — a section titled "Moral Distinctions Not deriv'd from Reason." In L. A. Selby-Bigge's standard edition (p. 469) it reads:
> "In every system of morality, which I have hitherto met with, I have always remark'd, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surpriz'd to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is, however, of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, 'tis necessary that it shou'd be observ'd and explain'd; and at the same time that a reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it. But as authors do not commonly use this precaution, I shall presume to recommend it to the readers; and am persuaded, that this small attention wou'd subvert all the vulgar systems of morality, and let us see, that the distinction of vice and virtue is not founded merely on the relations of objects, nor is perceiv'd by reason."
Read in place, the paragraph is the capstone of an extended attack on moral rationalism — the view, held by philosophers of Hume's era, that moral distinctions are discovered by reason alone and that moral truths are demonstrable in the way logical or mathematical truths are. Hume's targets included Samuel Clarke, whose doctrine of "eternal fitnesses and unfitnesses of things" grounded morality in necessary and eternal relations; William Wollaston, who held that vicious actions are those that express a falsehood; and the wider rationalist tradition including Ralph Cudworth and John Balguy. Clarke had argued that to deny the wrongness of murdering an innocent is as irrational as denying that three times three is nine. Hume's rebuttal, running through the section, is that vice and virtue are not among the four demonstrable philosophical relations (resemblance, contrariety, degrees in quality, proportions in quantity and number), that such relations can hold between inanimate things incapable of morality (trees, for example), and that reason alone cannot motivate action. The is-ought paragraph then adds a further, pointed challenge: even if rationalists established their factual and theological premises, they never explain how the "ought" gets into the conclusion. Hume's stated aim is a call for vigilance — a "small attention" that would "subvert all the vulgar systems of morality."
3. The is-ought problem explained precisely
Stripped to its logical core, the problem is this: an argument whose premises are all descriptive (statements about what is the case) cannot validly yield a normative conclusion (a statement about what ought to be the case) unless at least one normative premise is already present. The "ought" in the conclusion "expresses some new relation" not contained in the "is" premises, and a valid deductive argument cannot introduce in its conclusion content wholly absent from its premises. Consider: "Smoking harms your health; therefore you ought not to smoke." The conclusion does not follow from the factual premise alone; it requires a suppressed evaluative premise (roughly, "you ought not to do what harms your health"). That bridging premise does the real work — and it is itself a value judgement, not a fact.
This is why the doctrine sits at the heart of the fact/value distinction: descriptions and prescriptions occupy logically distinct territories, and no accumulation of facts, by itself, forces a value upon us. Crucially, Hume does not say moral reasoning is impossible or that "oughts" are baseless; he says the transition must be "observ'd and explain'd," and that the explanation, on his own account, lies not in reason's perception of objective relations but in human sentiment.
4. The naturalistic fallacy distinguished — the critical clarifying section
This is the section the site's "accuracy over patriotism" standard makes indispensable. Hume's is-ought problem and G. E. Moore's naturalistic fallacy are not the same thing, though they are routinely conflated in popular and even some academic usage.
Hume (1739–40) makes a point about inference: you cannot validly deduce a normative conclusion from purely factual premises. The error, if any, is a logical one — an invalid argument.
Moore (1903), in Principia Ethica, makes a point about definition and meaning: the property good is simple, unanalysable, and non-natural, and therefore cannot be defined in terms of any natural property (such as "pleasant," "desired," or "conducive to evolutionary fitness") or any supernatural/metaphysical property (such as "what God commands"). To attempt such a definition is to commit the "naturalistic fallacy."
Moore's diagnostic tool is the open question argument (Principia Ethica §13). For any proposed definition of "good" as some natural property X, the question "This is X, but is it good?" remains a meaningful, open question — not a trivial or self-answering one, as it would be if "good" simply meant X. ("It is pleasant, but is it good?" is a substantive question; "It is a bachelor, but is he unmarried?" is not.) The openness of the question shows the proposed definition has failed to capture the meaning of "good."
The distinction, then, is sharp: Hume's concern is logical (the gap in a chain of reasoning); Moore's is semantic (the indefinability of a concept). Notably, by Moore's own lights Hume would himself be a naturalist — Hume defines virtue in terms of the sentiments of approval it excites, which Moore would regard as committing the very fallacy. As the historian Lorraine Daston summarises William Frankena's landmark 1939 paper "The Naturalistic Fallacy" (Mind 48, p. 467), Hume and Moore made "strange bedfellows," sharing only a "bifurcationist" ontology of ethical judgments: just as Hume bifurcated "is" and "ought" propositions, intuitionists like Moore asserted that "ethical characteristics are different in kind from non-ethical ones." Frankena further argued that Moore's "fallacy" is not really a fallacy at all — better termed a "definist fallacy" — and that Moore erroneously lumped natural and metaphysical properties together. Moore himself acknowledged the label "naturalistic" was inapt, since his argument applied equally to supernatural definitions.
Why is the conflation so common? Several reasons: both doctrines police a boundary between the descriptive/natural and the evaluative/normative; both were enlisted by the same mid-century anti-naturalist movement (emotivists and prescriptivists cited Hume and Moore in the same breath); and the slogan "you can't get an ought from an is" is catchy enough to absorb Moore's separate point about defining "good." Careful scholars (e.g. in the Stanford Encyclopedia of Philosophy and Frankena's tradition) insist on keeping them apart: Hume targets inferences, Moore targets definitions, and unlike Hume's gap, Moore's non-naturalism was explicitly compatible with moral realism.
5. Interpretive debates about Hume's intended scope
How much did Hume mean to claim? Two broad camps:
The strict/universal reading. On the dominant twentieth-century interpretation — the one embraced by Hare, Prior, Nowell-Smith and many logicians — Hume is stating a sweeping logical law: no valid inference can ever move from purely descriptive premises to a normative conclusion, full stop. As Rachel Cohon notes in the Stanford Encyclopedia of Philosophy entry "Hume's Moral Philosophy," "a number of present-day philosophers, including R. M. Hare, endorse this putative thesis of logic, calling it 'Hume's Law.'" On this reading the passage articulates the "autonomy of ethics" and makes Hume, in this respect, an ally of Kant. This is the reading behind the names "Hume's Law" and "Hume's Guillotine."
Narrower/contextual readings. A substantial body of scholarship argues the strict reading over-reads a single paragraph. The most influential intervention is Alasdair MacIntyre's "Hume on 'Is' and 'Ought'" (Philosophical Review, 1959). MacIntyre argues that the orthodox interpretation is "inadequate and misleading" for several reasons. First, it is puzzling that the same commentators who approve of "Hume's Law" disapprove of Hume's parallel scepticism about induction — both rest on the assumption that "arguments must be either deductive or defective," an assumption MacIntyre (following Strawson on induction) rejects. Second, if the strict reading were right, "the first breach of Hume's law was committed by Hume" — for Hume's own moral theory repeatedly derives normative conclusions from facts about human interests, sympathy and social utility (his account of justice being the clearest case). Third, MacIntyre proposes reading the key words differently: taking "seems" literally (Hume finds the unexplained transition merely seemingly inconceivable) and reading "deduction" broadly as inference rather than strict entailment. On this account, Hume is not banning the move from facts to values but demanding that it be mediated and explained — for Hume, via bridging concepts such as human wants, needs and interests, which for Hume form part of what "ought" means, so that no additional free-standing evaluative premise is required.
MacIntyre's reading has been vigorously contested — R. F. Atkinson, Antony Flew, Geoffrey Hunter and W. D. Hudson all replied, and later critics argue MacIntyre himself fell into precisely the confusion Hume warned against. The debate is genuinely unresolved. As Rachel Cohon observes in the Stanford Encyclopedia of Philosophy, "few passages in Hume's work have generated more interpretive controversy." A defensible middle position: the strict logical claim (that ought-free premises cannot entail an ought-conclusion) is largely correct as a point of logic, but whether Hume intended to assert it as a universal meta-ethical principle — rather than as a targeted rebuke to particular rationalist and religious systems — is doubtful. (MacIntyre's mature development of this line, in After Virtue and beyond, is reserved for the final card.)
6. Hume's Law and twentieth-century meta-ethics
Hume's paragraph lay relatively dormant until the twentieth century, when it was elevated into a foundational principle of anti-naturalist meta-ethics:
A. J. Ayer's emotivism (Language, Truth and Logic, 1936). Applying the verification principle, Ayer held that value judgements are neither analytic nor empirically verifiable and so state no facts; they merely express emotion. To say "You acted wrongly in stealing that money" adds nothing factual to "You stole that money" — it merely evinces disapproval (the so-called "boo/hurrah" theory). Ayer explicitly saw himself as extending the British empiricist line from Hume.
C. L. Stevenson's Ethics and Language (1945) developed emotivism into a sophisticated account of the "dynamic" and persuasive uses of moral language.
R. M. Hare's prescriptivism (The Language of Morals, 1952). Hare argued that moral judgements are a species of prescription — universalisable imperatives — and that "it is impossible to derive any prescription from a set of purely descriptive sentences." Moral terms carry both prescriptive and descriptive meaning, but the prescriptive element cannot be squeezed out of facts alone. Hare was the writer who popularised the label "Hume's Law."
The informal names are twentieth-century coinages. "Hume's Law" was in currency by the mid-1950s (MacIntyre in 1959 attributes it to Hare). "Hume's Guillotine" was coined by Max Black in his 1964 paper "The Gap Between 'Is' and 'Should'" (Philosophical Review). Both labels post-date Hume by more than two centuries and should not be read back into his own vocabulary.
7. Searle's challenge and its reception
In "How to Derive 'Ought' from 'Is'" (Philosophical Review, vol. 73, 1964, pp. 43–58), John Searle offered what he called a counterexample to the thesis (which he explicitly links to "a famous passage in Hume's Treatise" and to "what has been called the naturalistic fallacy"). His derivation runs through five statements:
"Jones uttered the words 'I hereby promise to pay you, Smith, five dollars.'"
"Jones promised to pay Smith five dollars."
"Jones placed himself under (undertook) an obligation to pay Smith five dollars."
"Jones is under an obligation to pay Smith five dollars."
"Jones ought to pay Smith five dollars."
Searle argues that the relation between each statement and its successor, while "not in every case one of 'entailment,'" is not merely contingent, and that the extra premises needed to make each step deductive are not themselves evaluative — they are either statements of fact about English usage and circumstance (the "conditions C" under which uttering the words counts as promising), or tautologies unpacking the meaning of "promise," "obligation," and "ought" (e.g. "all promises are acts of placing oneself under an obligation"; "other things being equal, one ought to do what one is under an obligation to do"). The engine is the notion of an institutional fact governed by constitutive rules: within the institution of promising, to state that a promise occurred is already to invoke rules that generate an obligation. Having added the factual premises about utterance and conditions, Searle writes, "as far as I can see, no moral premises are lurking in the logical woodpile."
The reception has been sceptical, and the standard objection is that Searle's derivation smuggles in a normative premise. The most influential critique is Hare's "The Promising Game" (1964): the derivation, critics argue, at most yields the conclusion that Jones ought to pay Smith according to the rules of the institution of promising — an "ought" internal to the institution, comparable to "in chess, one ought to move one's king out of check." As Kurtis Hagen puts the Moore-style objection, "whatever obligation is derived from an institutional fact, it is still an open question whether it is a moral obligation. What is needed to complete the proof is a moral justification of the institution... Searle has at least not derived a moral 'ought' from an 'is.'" The "ceteris paribus" clauses at steps (4) and (5), critics add, may themselves conceal evaluative judgements. Searle continued to defend the derivation for decades (including in later work on social ontology), and some philosophers defend its validity: Frank Hindriks, in "Collective Acceptance and the Is-Ought Argument" (Ethical Theory and Moral Practice 16, 2013, 465–480), reconstructs the argument, establishes that "it is valid," and argues that "the soundness of the argument depends on whether collective acceptance as such can generate obligations," concluding that "the Is-Ought Argument deserves to be taken seriously once again." The upshot is a genuine, ongoing stand-off rather than a decisive refutation of Hume in either direction.
Alongside Searle, the logician A. N. Prior ("The Autonomy of Ethics," 1960) produced formal counterexamples exploiting the logic of disjunction — e.g. "Tea-drinking is common in England; therefore, either tea-drinking is common in England or all New Zealanders ought to be shot" — a valid inference from a factual premise to a conclusion containing an "ought." Defenders of the gap (such as Charles Pigden) reply that the ethical content of such conclusions is "vacuous" — the ethical term could be swapped for any other without affecting validity — so no substantive ought has been derived. These logical puzzles have sharpened, rather than settled, the formulation of "Hume's Law."
8. Connection to the Problem of Induction (Card #6) and forward to MacIntyre
The is-ought problem is naturally read as a "Humean gap argument" — one of a family of Humean arguments that dramatise the difficulty of bridging one logically distinct domain to another. The Problem of Induction (Card #6) exposes a gap between the observed and the unobserved: no accumulation of past instances logically guarantees the future. The is-ought problem exposes a gap between the descriptive and the normative: no accumulation of facts logically guarantees a value. Strikingly, MacIntyre's 1959 article draws exactly this parallel, noting that both rest on "the demand that induction shall be shown to be really a kind of deduction" (Strawson's phrase) — the assumption that arguments are "deductive or defective." The parallel is a structural resemblance in the form of the arguments; it is not a claim that Hume framed them as a single unified doctrine. They concern separate domains — the epistemology of causal and inductive inference on one side, the logic of moral inference on the other — and should be kept distinct even as their family likeness is noted.
The Hume cluster of this series (Cards #5–#7) thus opens onto its own conclusion. Alasdair MacIntyre's After Virtue and the revival of virtue ethics. MacIntyre is not merely a commentator on Hume in passing: his career-long engagement with the is-ought passage — from the 1959 article's narrower reading to After Virtue's claim that the gap opens only once Aristotelian teleology is jettisoned (for functional concepts such as "watch," "farmer" or "sea-captain," facts do entail evaluations: a watch that keeps poor time is a bad watch) — forms a thread running from the start of the philosophical cluster to the series' close. Hume's argument at Card #7 poses the question; MacIntyre's virtue ethics offers one of the most influential twentieth-century answers.
Timeline
Hume publishes the Treatise; the is-ought paragraph appears at the close of Book 3, Part 1, Section 1.
Hume recasts his moral philosophy in An Enquiry concerning the Principles of Morals, dropping the explicit is-ought remark but retaining the anti-rationalist thrust.
G. E. Moore publishes Principia Ethica, defining the "naturalistic fallacy" and deploying the "open question argument" — a distinct doctrine later confused with Hume's.
A. J. Ayer's Language, Truth and Logic advances emotivism, treating value judgements as expressions of feeling that are neither true nor false.
William K. Frankena's article "The Naturalistic Fallacy" (Mind) distinguishes Moore's fallacy from the is-ought point and argues it is better called a "definist fallacy."
C. L. Stevenson's Ethics and Language develops emotivism in detail.
R. M. Hare's The Language of Morals founds universal prescriptivism, using the is-ought gap as a cornerstone; "Hume's Law" enters currency.
Alasdair MacIntyre's "Hume on 'Is' and 'Ought'" (Philosophical Review) challenges the orthodox reading, provoking replies from Atkinson, Flew, Hunter, and Hudson.
A. N. Prior's "The Autonomy of Ethics" presents logical counterexamples (the "tea-drinking/New Zealanders" disjunction).
John Searle publishes "How to Derive 'Ought' from 'Is'" (Philosophical Review); Max Black coins "Hume's Guillotine" in "The Gap Between 'Is' and 'Should'"; Hare replies to Searle in "The Promising Game."
MacIntyre's After Virtue argues that the is-ought gap holds only once Aristotelian teleology is abandoned; for functional concepts, facts do entail evaluations.
Did You Know?
- Hume never used the words "is-ought problem," "Hume's Law," or "Hume's Guillotine." The whole edifice rests on a single paragraph; "Hume's Guillotine" was coined by Max Black in 1964, more than two centuries later, and "Hume's Law" was popularised by R. M. Hare in the 1950s.
- By G. E. Moore's own criteria, Hume commits the "naturalistic fallacy" — because Hume analyses virtue in terms of the sentiments of approval it produces, a natural-property definition of the kind Moore's open question argument attacks. The two doctrines are not just distinct; they partly cut against each other.
- Hume may have broken his own "law." Critics from MacIntyre onward point out that Hume's own theory of justice derives obligations from facts about human interests and social utility — which is why the strict universal reading sits awkwardly with Hume's actual practice.
- The famous paragraph directly follows one in which Hume seems to derive an "ought" from a feeling — his statement that to call an action vicious is to report "a sentiment of disapprobation which arises in you." Reconciling these two adjacent paragraphs is a central task of Hume interpretation.
- A. N. Prior showed the gap is logically slippery with the deliberately absurd "tea-drinking in England / New Zealanders ought to be shot" disjunction — a valid step from fact to a sentence containing "ought," which forces defenders to specify exactly what a "substantive" ought-conclusion is.
- The is-ought gap has migrated into artificial-intelligence research. Recent work on AI value alignment invokes Hume to argue that purely descriptive training data (records of what humans do and prefer) systematically under-determines the normative content the system is meant to capture. As the 2025 paper "The Specification Trap" (arXiv:2512.03048) puts it, Hume's observation that "no purely factual premises can entail a normative conclusion... has survived nearly three centuries of philosophical scrutiny, and no widely accepted solution exists."
- Emotivism was nicknamed the "boo/hurrah theory," because on Ayer's account "stealing is wrong" functions like "Stealing — boo!" rather than a statement that could be true or false.
Caveats: Myths and Disputed Claims
- Myth: The is-ought problem and the naturalistic fallacy are the same thing. They are not. Hume's point is about the invalidity of deducing normative conclusions from factual premises; Moore's is about the impossibility of defining "good" in natural terms. Different philosophers (Hume 1739–40; Moore 1903), different targets (inference vs. definition), and, by Moore's own criteria, Hume would count as a naturalist. Treating them as identical is a genuine and widespread error.
- Disputed: How broadly to read Hume's own paragraph. The strict/universal reading ("no ought from an is, ever") is the twentieth-century orthodoxy, but MacIntyre (1959) and others argue Hume aimed narrowly at the rationalist and religious moralists of his day and did not intend a sweeping meta-ethical law — pointing out that Hume himself derives oughts from facts about human interests. This is unresolved; present it as a real interpretive debate, not a settled fact.
- Disputed: Whether Searle refuted the gap. Searle's 1964 promising derivation is a genuine, live challenge, but the common and forceful objection is that it smuggles in a normative premise (that one ought to honour the institution of promising) and delivers only an institution-internal "ought." Defenders such as Frank Hindriks maintain the argument is valid and deserves to be taken seriously; no consensus verdict exists. Do not declare a winner.
- Caution: The names are anachronistic. "Hume's Law" (Hare, 1950s) and "Hume's Guillotine" (Black, 1964) are later labels, not Hume's own terminology, and the entire doctrine is extrapolated from one short paragraph.
- Caution against overreach. The is-ought problem does not by itself prove moral scepticism, nihilism, or the meaninglessness of ethics. It is a constraint on a certain kind of inference; how morality is grounded (sentiment, convention, teleology, non-natural properties) remains a separate, open question.