The is-ought problem

The is-ought problem — sometimes called Hume’s guillotine or Hume’s law — is the philosophical thesis that no set of purely descriptive premises about how the world is can, by deductive logic alone, entail a normative conclusion about how the world ought to be. First articulated by David Hume in Book III of A Treatise of Human Nature (1739–1740), the problem identifies a logical gap between statements of fact and statements of value that has shaped every major debate in metaethics since.¹ The gap poses a challenge to any moral theory that purports to derive ethical obligations from observations about nature, human psychology, evolutionary biology, or divine decree: if descriptive premises alone cannot yield prescriptive conclusions, then an additional normative premise is always required, and the question of where that premise comes from becomes the central problem of moral philosophy.

The is-ought problem intersects with some of the deepest questions in the philosophy of religion. Proponents of the moral argument for God’s existence contend that the irreducibility of normative facts to natural facts points toward a transcendent moral lawgiver — that if morality cannot be grounded in nature alone, it must be grounded in something beyond nature. Critics respond that divine command theory faces the same logical gap: the statement that God commands a particular action is itself a descriptive claim, and the inference that one therefore ought to obey requires a suppressed normative premise. The is-ought problem thus bears on the Euthyphro dilemma, the viability of secular ethics, and the relationship between evolution and morality.^{1, 14}

Hume’s original formulation

The passage that gave rise to the is-ought problem appears at the close of Hume’s argument against moral rationalism in Book III, Part I, Section I of A Treatise of Human Nature. Hume observes that in every system of morality he has encountered, the author proceeds “for some time in the ordinary way of reasoning” — establishing claims about God, human nature, or the state of society using “is” and “is not” — and then, without explanation, shifts to propositions connected by “ought” and “ought not.” This transition, Hume writes, “is imperceptible; but is, however, of the last consequence.” He continues: “For as this ought, or ought not, expresses some new relation or affirmation, ’tis necessary that it should be observed and explained; 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.”¹

Hume’s language is carefully hedged. He does not state flatly that an ought can never be derived from an is; he says that such a derivation “seems altogether inconceivable” and demands that those who attempt it provide an explanation. This ambiguity has generated an enormous interpretive literature. The dominant twentieth-century reading, championed by R. M. Hare and A. N. Prior, takes Hume to be stating a strict logical thesis: that no argument whose premises are entirely descriptive can have a normative conclusion, because the conclusion would contain normative vocabulary not present in the premises, violating the principle that nothing can appear in the conclusion of a valid deduction that is not at least implicitly contained in the premises.^{5, 6}

Other interpreters read the passage more modestly. Some take Hume to be making an epistemological rather than a logical point: that the initial discovery of moral distinctions requires an input from sentiment, not from reason alone, and that therefore one cannot arrive at moral knowledge through purely theoretical inference from non-moral premises.¹² On this reading, Hume is not denying that moral conclusions can figure in cogent probable arguments alongside factual premises; he is denying only that moral properties are discoverable by demonstrative reason operating on non-moral evidence. Charles Pigden has argued that regardless of which interpretation is correct, the passage establishes a constraint that any moral theory must acknowledge: the transition from “is” to “ought” is never logically trivial and always requires justification.¹³

The context of the passage matters. Hume’s argument in Book III is directed against moral rationalists — philosophers such as Samuel Clarke and William Wollaston — who held that moral distinctions are discoverable by reason in the same way that mathematical truths are. Hume is arguing that moral judgments are grounded in sentiment rather than reason: we call actions virtuous or vicious because of the feelings of approbation or disapprobation they produce in us, not because we perceive some objective moral relation through rational intuition. The is-ought passage functions as one piece of this broader argument, highlighting the gap that moral rationalists must (but cannot) bridge between their factual premises and their normative conclusions.^{1, 12}

The logical structure of the gap

The is-ought problem can be formulated with precision in terms of formal logic. Consider a standard deductive argument. For such an argument to be valid, the conclusion must be entailed by the premises: there must be no possible situation in which the premises are all true and the conclusion false. This means that no term or predicate can appear in the conclusion that is not already present, at least implicitly, in the premises. If all premises are purely descriptive — containing only terms like “is,” “exists,” “causes,” “produces” — then a conclusion containing the normative term “ought” introduces new content not warranted by the premises. The argument is therefore invalid.

P1. Torturing children causes suffering. (descriptive)

P2. Suffering diminishes well-being. (descriptive)

C. Therefore, one ought not to torture children. (normative)

This argument, despite its intuitive force, is logically invalid as stated. The conclusion introduces the normative concept “ought not” which is absent from both premises. To make the argument valid, a normative premise must be supplied:

P1. Torturing children causes suffering. (descriptive)

P2. Suffering diminishes well-being. (descriptive)

P3. One ought not to perform actions that diminish well-being. (normative)

C. Therefore, one ought not to torture children. (normative)

The revised argument is valid. But the normative work is now performed by P3, which is itself an “ought” statement. The is-ought problem has not been solved; it has merely been pushed back one step. Where does P3 come from? If from further descriptive premises, the same gap reappears. If P3 is taken as a self-evident moral axiom, then the theory rests on a normative foundation that is not itself derived from “is” statements. If P3 is grounded in God’s commands, then the question becomes why one ought to obey God — a further normative claim.^{5, 13}

A. N. Prior demonstrated in 1960 that in standard formal logic, if a set of premises contains no evaluative terms, no conclusion containing evaluative terms can be validly derived from those premises alone. Prior also noted a technical complication: in certain non-standard logics, trivial counterexamples can be constructed (for instance, from the premise “Tea-drinking is common in England,” one can derive “Either tea-drinking is common in England or one ought to commit murder” — which is technically a disjunction containing an “ought”). Pigden and others have responded by refining the thesis to exclude such trivial cases, specifying that the is-ought gap holds for conclusions in which the normative content is essential to the conclusion’s argumentative force.¹³

Moore’s open question argument and the naturalistic fallacy

In Principia Ethica (1903), G. E. Moore identified what he called the naturalistic fallacy: the error of defining the property of moral goodness in terms of any natural property, such as pleasure, evolutionary fitness, or desire-satisfaction. Moore’s central argument — the open question argument — proceeds as follows. Suppose someone proposes that “good” means “pleasant.” If this definition were correct, then the question “This is pleasant, but is it good?” would be trivially equivalent to “This is pleasant, but is it pleasant?” — a closed question admitting of no genuine inquiry. But the first question is plainly meaningful and non-trivial; a competent speaker can coherently wonder whether all pleasant things are good. This shows that “good” and “pleasant” do not mean the same thing.²

Moore applied the argument against every reductive account of goodness available to him. Against hedonists like Jeremy Bentham, he argued that the openness of the question demonstrates the non-identity of good and pleasant. Against evolutionary ethicists like Herbert Spencer, who identified goodness with what promotes evolutionary fitness, Moore pointed out that one can coherently ask whether what promotes survival really is good. Against metaphysical ethicists who defined goodness in terms of conformity to God’s will, Moore held that the same challenge applies: it remains a meaningful question whether what God commands is good. Moore thus treated the fallacy as encompassing both naturalistic and supernaturalistic reductions of moral goodness.²

The open question argument and the is-ought gap are related but distinct. The is-ought gap is a claim about logical entailment: descriptive premises do not logically entail normative conclusions. The naturalistic fallacy is a claim about property identity: moral goodness is not identical to any natural property. A philosopher might accept the is-ought gap while rejecting Moore’s diagnosis — holding, for example, that goodness is a natural property that cannot be captured by analytic definitions but that is nonetheless empirically discoverable. Conversely, one might accept Moore’s non-naturalism while interpreting Hume’s original passage more modestly. Together, however, the two arguments constitute a formidable challenge to any project of reducing morality to empirical fact.^{2, 5}

Attempts to bridge the gap

The is-ought gap has provoked numerous attempts at a bridge, ranging from institutional facts to speech-act theory to neo-Aristotelian teleology. These attempts share a common structure: each identifies a class of facts that, the proponent claims, already contain normative content, so that deriving an “ought” from them does not constitute a genuine leap from pure description to prescription.

John Searle offered the most widely discussed counterexample in his 1964 article “How to Derive ‘Ought’ from ‘Is.’” Searle argued that the institution of promising creates a class of facts that are simultaneously descriptive and normative. The sequence of reasoning proceeds from the purely descriptive premise “Jones uttered the words ‘I hereby promise to pay you, Smith, five dollars’” through a series of steps involving the constitutive rules of the institution of promising to the normative conclusion “Jones ought to pay Smith five dollars.” The key move is the claim that certain “is” statements — specifically, those that describe the performance of institutional speech acts — carry normative implications as part of their constitutive structure.³

Searle’s argument generated extensive debate. R. M. Hare and others objected that the derivation succeeds only by smuggling a normative premise into the description of promising: to say that someone has made a promise is already to invoke a normative institution, one that includes the rule that promises ought to be kept. On this reading, Searle has not derived an “ought” from a purely descriptive “is”; he has derived an “ought” from a set of premises that includes, implicitly, a normative commitment to the institution of promising. Defenders of Searle respond that the objection misses the point: the existence of institutions whose constitutive rules are simultaneously factual and normative shows that the sharp distinction between “is” and “ought” cannot be maintained in practice, even if it holds in the abstract logic of purely descriptive statements.^{3, 5}

Max Black, also writing in 1964, challenged what he called “Hume’s Guillotine” — a term Black himself coined — arguing that the principle should be regarded as “a dogma which ought by now to have been finally exploded.” Black contended that in ordinary reasoning, we routinely and legitimately move from factual claims to evaluative conclusions: from “this act will cause unnecessary suffering” to “this act should not be performed.” The transition is not a logical fallacy but a rational inference that depends on the established meaning of moral terms in natural language.⁴

Philippa Foot, in Natural Goodness (2001), offered a different strategy rooted in neo-Aristotelian virtue ethics. Foot argued that moral evaluations of human beings are logically continuous with evaluations of other living organisms. Just as a defective wolf is one that fails to function as wolves characteristically do — a wolf that cannot hunt in a pack or does not nurture its young — a morally defective human is one who fails to exhibit the characteristic excellences of the human life form. On this view, the distinction between “is” and “ought” dissolves when we recognise that certain descriptions of living things are inherently evaluative: to say what a creature is (what kind of life form it belongs to) already implies norms about what it ought to do or be.⁷

Alasdair MacIntyre, in After Virtue (1981), argued that the is-ought problem is a specifically modern problem — the product of the Enlightenment’s abandonment of Aristotelian teleology. In the classical tradition, human beings were understood to have a telos (an end or purpose), and moral language functioned to evaluate how well a person was progressing toward that end. Within this framework, the transition from “is” to “ought” is unproblematic: from the fact that a human being has a certain nature and a certain end, it follows that certain actions are conducive to that end and therefore good. The is-ought gap opened, MacIntyre argued, precisely when Enlightenment thinkers rejected teleology while retaining the moral vocabulary that had been intelligible only within a teleological framework.⁸

Implications for theistic ethics and the moral argument

The is-ought problem occupies a pivotal position in debates about the relationship between God and morality. Proponents of the moral argument for God’s existence have argued that the irreducibility of normative facts to natural facts supports a theistic worldview. William Lane Craig’s formulation of the moral argument proceeds from the premise that “if God does not exist, objective moral values and duties do not exist” to the conclusion that God exists. The is-ought gap functions as supporting evidence for this conditional: if purely descriptive facts about the natural world cannot ground normative obligations, then the source of moral normativity must lie outside nature — in a transcendent moral lawgiver whose commands constitute moral duties and whose nature constitutes the paradigm of moral goodness.¹⁴

This line of reasoning faces a structural challenge, however. The statement “God commands X” is itself a descriptive claim — a statement about what is the case. The inference “therefore one ought to do X” requires a suppressed normative premise: “one ought to obey God’s commands.” If that premise is treated as a brute normative truth, then the theist has not eliminated the is-ought gap but merely relocated it. If the premise is grounded in some further fact about God — that God is perfectly good, or supremely powerful, or the creator of all things — then the same gap reappears at the next level: why does God’s goodness or power generate an obligation to obey? This challenge connects directly to the Euthyphro dilemma: if one ought to obey God because God’s commands are good, then goodness is independent of God; if God’s commands are good simply because God issues them, then morality is arbitrary.^{14, 15}

Theistic philosophers have responded in several ways. Robert Adams and William Alston have argued that God’s nature — which is essentially loving, just, and good — serves as the metaphysical foundation of moral goodness, and that divine commands, flowing necessarily from that nature, constitute moral obligations. On this view, the normative premise “one ought to obey God’s commands” is not a free-standing principle but is grounded in the identity between God’s nature and the good itself. The is-ought gap is bridged not by a bare inference from a descriptive fact about God to a normative conclusion, but by the metaphysical thesis that God’s nature is the ultimate standard of value — that goodness is what God’s nature is.¹⁴

Critics question whether this response succeeds. If the claim “God’s nature is the standard of goodness” is itself a normative claim, then it does not derive from descriptive premises and the is-ought gap remains unbridged. If it is a descriptive claim — a metaphysical thesis about the identity of goodness with a particular being’s nature — then Moore’s open question argument applies: it remains meaningful to ask whether God’s nature really is good, which suggests that goodness and God’s nature are not analytically identical. The debate remains unresolved, with both sides acknowledging that the is-ought gap poses a genuine constraint on moral theorising that no framework — theistic or secular — can simply ignore.^{2, 15}

Divine command theory and natural law

Divine command theory holds that moral obligations are constituted by the commands of God: an action is morally required because God commands it, morally forbidden because God prohibits it. The theory appears to offer a direct bridge across the is-ought gap, grounding normative conclusions in a particular kind of fact — namely, the fact of divine command. If God commands that one keep one’s promises, and if divine commands constitute moral obligations, then from the descriptive fact “God commands promise-keeping” the normative conclusion “one ought to keep one’s promises” follows directly.¹⁴

The difficulty is that the bridge depends on a linking principle — “divine commands constitute moral obligations” — that is itself a normative claim. Why do God’s commands generate obligations? One answer is that God has the authority to command because God is the creator and sustainer of all things. But authority-based arguments face the same structural problem: from the fact that God created the universe, it does not follow, without a further normative premise, that creatures ought to obey their creator. Another answer is that God’s commands track independently existing moral truths, but this concedes that morality is independent of God and abandons the distinctively theistic claim that God is the source of moral obligation. These are, in essence, the two horns of the Euthyphro dilemma as applied to the is-ought problem.^{14, 15}

The natural law tradition, rooted in the work of Thomas Aquinas, offers a distinct approach. Aquinas held that human beings, by virtue of their rational nature, can discern through practical reason the basic goods toward which human nature is directed: life, knowledge, sociability, procreation, and others. The first principle of natural law — “good is to be done and pursued, and evil is to be avoided” — is held to be self-evident to practical reason, not derived from any prior descriptive premise. Natural law theory thus claims to avoid the is-ought gap by locating moral knowledge in the deliverances of practical reason rather than in inferences from theoretical descriptions of human nature.

Critics have argued that natural law theory either commits the naturalistic fallacy (by identifying goodness with what human nature tends toward) or covertly relies on a theological premise (that human nature was designed by God to pursue genuine goods). If the theory grounds the normativity of basic goods in God’s creative intent, it faces the same is-ought challenge as divine command theory. If it treats basic goods as self-evident to practical reason without theological support, it concedes that morality has a foundation independent of God — a concession that undermines the moral argument for God’s existence. Contemporary natural law theorists such as John Finnis have argued that the theory requires neither a prior proof of God’s existence nor an inference from human nature to moral norms, but rather rests on the intrinsic intelligibility of basic goods as grasped by practical reason — a response that sidesteps the is-ought gap by denying that moral knowledge is ever a matter of theoretical inference from descriptive facts.^{7, 8}

Evolutionary debunking and the is-ought gap

The development of evolutionary biology has added a new dimension to the is-ought problem. If human moral intuitions are the product of natural selection, shaped over millions of years to promote reproductive fitness rather than to track objective moral truth, then the reliability of those intuitions is called into question. This line of argument — the evolutionary debunking argument — was developed most influentially by Sharon Street in her 2006 paper “A Darwinian Dilemma for Realist Theories of Value.”⁹

Street’s argument poses a dilemma for moral realists. Either evolutionary forces have influenced the content of human moral judgments in a way that tracks objective moral truth, or they have not. If they have, then there must be some explanation of why natural selection would produce beliefs that are both fitness-enhancing and true — and no such explanation has been offered. If they have not, then there is a massive coincidence between our evolved moral beliefs and the objective moral truth, a coincidence for which the moral realist has no account. Street concludes that antirealist theories of value — which hold that moral truths are constructed by, rather than independent of, human evaluative attitudes — are better positioned to accommodate the evolutionary evidence.⁹

The evolutionary debunking argument intersects with the is-ought problem in a specific way. Even if evolutionary biology can fully explain why human beings have the moral intuitions they do — why we feel that fairness is good, that cruelty is wrong, that we owe special obligations to kin — this causal explanation is entirely descriptive. It tells us what is the case about the origins of our moral sentiments. It does not tell us what ought to be the case. The is-ought gap thus functions as a constraint on evolutionary ethics: the fact that evolution produced a particular set of moral intuitions does not, by itself, vindicate or undermine those intuitions. To move from “evolution shaped us to believe X is wrong” to “X is (or is not) wrong” requires a normative premise that evolution itself cannot supply.^{9, 15}

Theists have used the evolutionary debunking argument in support of the moral argument for God. Mark Linville, for example, has argued that on a naturalistic worldview, evolutionary debunking undermines the reliability of moral intuitions, leaving naturalists without a basis for moral knowledge. Theism, by contrast, can explain the reliability of moral intuitions by appealing to a God who designed the evolutionary process to produce beings with a capacity for moral truth. This response invokes the is-ought gap as a problem specifically for naturalism: nature alone cannot bridge the gap, but God can, by ensuring that the natural processes that shaped our moral faculties were directed toward truth rather than mere fitness.¹⁴

The force of this response depends on whether the theist’s own bridge across the is-ought gap is any stronger than the naturalist’s. If the theistic bridge still requires a suppressed normative premise — “one ought to trust the moral faculties that God designed” — then the gap has not been eliminated but relocated. Moreover, Street’s dilemma can be reframed in theistic terms: if God designed human moral faculties to track moral truth, and if moral truth is constituted by God’s own nature, then the explanation is circular (God designed us to recognise as good what God defines as good). The debate thus continues to turn on the fundamental question of whether any framework — theistic, naturalistic, or otherwise — can fully close the logical gap between is and ought.^{9, 14}

Sam Harris and The Moral Landscape

In The Moral Landscape: How Science Can Determine Human Values (2010), Sam Harris argued that the is-ought problem has been overblown and that science can, in principle, determine moral truths. Harris’s central contention is that moral questions are really questions about the well-being of conscious creatures, and that the well-being of conscious creatures is an empirical matter that science can investigate. If we accept that the worst possible misery for everyone is bad, then — Harris claims — all other moral questions can be framed as empirical questions about how to move away from that nadir toward states of greater well-being.¹⁰

Harris is explicit that he regards the is-ought gap as a pseudo-problem. He argues that the distinction between facts and values is less sharp than philosophers assume, because values are themselves a kind of fact — facts about the well-being of conscious creatures. Just as we do not demand a non-scientific justification for the assumption that health is better than disease before we allow medicine to proceed, Harris contends, we should not demand a non-scientific justification for the assumption that well-being is better than misery before we allow science to inform morality.¹⁰

The philosophical response to Harris has been largely critical, even from thinkers sympathetic to his broader project. The central objection is that Harris’s argument presupposes precisely what the is-ought problem says cannot be derived: a normative starting point. If we accept that well-being is the fundamental moral value, then science can indeed help determine which actions promote well-being. But the premise that well-being is the fundamental moral value is itself a normative claim, not an empirical finding. It cannot be established by scientific investigation; it must be assumed or argued for on non-empirical grounds. Harris has, in effect, smuggled a normative premise into his argument and then declared that the is-ought gap has been closed.^{10, 11}

Hilary Putnam, whose work on the entanglement of facts and values in The Collapse of the Fact/Value Dichotomy (2002) might seem to support Harris’s project, actually drew a more nuanced conclusion. Putnam argued that the sharp dichotomy between facts and values — the idea that facts are objective and values are merely subjective — is untenable, because scientific inquiry itself relies on epistemic values (coherence, simplicity, explanatory power) that cannot be separated from the supposedly value-free facts they help to establish. But Putnam did not conclude that science can determine moral values; he concluded that the entanglement of facts and values runs in both directions, complicating any simple reduction of either to the other.¹¹

The fact-value distinction in philosophy of science

The is-ought problem is closely related to, but not identical with, the broader fact-value distinction in the philosophy of science. The fact-value distinction holds that there is a principled difference between descriptive claims about the world (facts) and evaluative claims about how things should be (values), and that science, as an enterprise concerned with facts, is or should be value-free. This distinction became a cornerstone of logical positivism in the early twentieth century, particularly through the work of the Vienna Circle and A. J. Ayer’s Language, Truth and Logic (1936), which classified moral statements as expressions of emotion (emotivism) rather than genuine propositions capable of truth or falsity.

The positivist fact-value distinction drew on Hume’s is-ought passage but extended it in ways that Hume himself might not have endorsed. Hume’s point was about the logical gap between descriptive and normative premises in moral arguments; the positivists generalised this into a wholesale division of meaningful language into factual statements (verifiable by observation) and value statements (neither true nor false). On this view, moral discourse is not just logically independent of factual discourse; it is cognitively meaningless — not the sort of thing that can be true or false at all.⁶

This strong version of the fact-value distinction has been extensively criticised. Putnam argued that the dichotomy collapses under scrutiny: scientific practice is pervaded by values (both epistemic and ethical), and the idea of a value-free description of the world is an illusion fostered by a bankrupt positivist metaphysics.¹¹ Thomas Kuhn’s work on scientific revolutions revealed that theory choice in science is influenced by values such as simplicity, fruitfulness, and explanatory scope — values that are not themselves empirically testable. If science cannot function without value judgments, then the sharp distinction between the factual and the evaluative is unsustainable even within science itself.

The implications for the is-ought problem are significant but limited. Even if the sharp fact-value dichotomy is untenable, and even if scientific practice is pervaded by epistemic values, it does not follow that moral values can be derived from empirical facts. The entanglement of facts and values in scientific practice concerns epistemic values (what counts as a good explanation, what standards of evidence to adopt), not moral values (what is right and wrong). A scientist can acknowledge that simplicity is an epistemic virtue without committing to any particular moral claim. The is-ought gap, understood as a logical constraint on moral reasoning, survives the collapse of the broader positivist fact-value dichotomy.^{11, 13}

Prescriptivism, non-cognitivism, and the gap

The is-ought problem played a central role in the development of non-cognitivist metaethics in the twentieth century. Non-cognitivism holds that moral statements do not express beliefs about objective facts but rather express attitudes, emotions, or prescriptions. On this view, the is-ought gap is explained by the fundamental difference in kind between descriptive statements (which express beliefs) and moral statements (which express something other than beliefs).

R. M. Hare’s universal prescriptivism, developed in The Language of Morals (1952) and Freedom and Reason (1963), accepted the is-ought gap as a corollary of the logical properties of moral language. Hare argued that moral judgments have two essential features: they are prescriptive (they tell someone what to do) and universalisable (if one judges a particular action to be wrong, one must judge any relevantly similar action to be wrong). Because moral judgments are prescriptions rather than descriptions, they cannot be derived from purely descriptive premises. The is-ought gap is, on Hare’s view, a consequence of the prescriptive character of moral language: you cannot get a prescription out of a set of descriptions, just as you cannot get an imperative (“shut the door!”) out of a set of indicatives (“the door is open”).⁶

Hare nonetheless sought to provide a foothold for moral reasoning within the constraints of the is-ought gap. He argued that the requirement of universalisability, combined with knowledge of the relevant facts, constrains moral judgments substantially: to prescribe that promise-breaking is permissible, one must be willing to prescribe that it is permissible for everyone in relevantly similar circumstances — including situations where one is the victim of broken promises. This constraint does not eliminate the is-ought gap, but it narrows the range of moral positions that a rational agent can consistently hold.⁶

J. L. Mackie, in Ethics: Inventing Right and Wrong (1977), took the is-ought gap as evidence for a more radical conclusion: moral error theory. Mackie argued that ordinary moral language purports to describe objective moral facts — facts about what is really right or really wrong, independently of human attitudes. But the is-ought gap, combined with the metaphysical “queerness” of objective prescriptive properties (properties that are simultaneously part of the fabric of the world and intrinsically action-guiding), shows that no such facts exist. Our moral judgments are systematically false: they claim objectivity, but there is nothing in the world that makes them true. Mackie’s position thus treats the is-ought gap not as a problem to be solved but as a symptom of the non-existence of objective moral facts.¹⁵

Contemporary analytical approaches

The is-ought problem continues to generate significant work in contemporary analytic philosophy. Several distinct research programmes address the gap from different angles.

The Cornell realists — including Richard Boyd, Nicholas Sturgeon, and David Brink — have argued that moral properties are natural properties, identifiable through a posteriori investigation rather than a priori definition. On this view, the open question argument shows only that moral terms cannot be given analytic definitions in natural terms, not that moral properties fail to be natural properties. Just as “water” and “H₂O” differ in meaning while denoting the same substance, “good” and some natural property might differ in meaning while referring to the same thing. If this is correct, then the is-ought gap as Moore understood it — as a reflection of the non-identity of moral and natural properties — does not hold. The gap between “is” and “ought” is no wider than the gap between “water” and “H₂O”: a gap in meaning, not in reference.²

Non-naturalist moral realists, including Derek Parfit and T. M. Scanlon, have accepted that moral properties are not identical to natural properties but have argued that they are nonetheless objective and knowable. On this view, the is-ought gap is real — moral truths are not derivable from natural facts — but this is not a problem, because moral truths are self-standing normative facts that do not need to be derived from anything else. Just as mathematical truths are not derivable from physical facts but are nonetheless objective, moral truths are not derivable from descriptive facts but are nonetheless mind-independent. This approach accepts the is-ought gap as a logical truth while denying that it undermines moral realism.

Constructivists, following Christine Korsgaard, have argued that moral normativity is explained not by the existence of independent moral facts but by the structure of practical reason itself. On Korsgaard’s view, the source of normativity lies in the reflective endorsement of principles by rational agents: we are bound by moral norms because, on reflection, we cannot coherently will to act otherwise. The is-ought gap is acknowledged but reinterpreted: morality does not derive from descriptions of the natural world, but from the internal requirements of rational agency. This approach echoes Kant’s original insight that moral obligation is autonomous — grounded in reason rather than in external facts about nature or God.

Major positions on the is-ought gap^{1, 2, 6, 7, 8, 15}

The persistence of the is-ought problem across more than two and a half centuries of philosophical inquiry suggests that it identifies something genuine about the structure of moral reasoning. Whether the gap is a deep logical truth, an artefact of a particular way of understanding language, or a problem that dissolves under the right metaphysical framework remains an open question — one that continues to shape debates in metaethics, the philosophy of religion, and the philosophy of science. What the debate has established, at a minimum, is a constraint that any adequate moral theory must address: the recognition that the transition from descriptive claims about the world to normative claims about how we ought to act is never trivial, never self-explanatory, and always in need of justification.^{1, 5, 13}