Is-Ought Gap
Hume's observation that moral writers regularly slide from descriptive (is) claims to normative (ought) claims without justifying the inference — sometimes called Hume's Law or the naturalistic fallacy.
In practice
A report shows what people actually do, and the recommendation section says what should be done. Somewhere between the two a step was taken that the data does not supply.
The common mistake
Treating it as proof that values are arbitrary. Hume asks for the inference to be shown, not for it to be abandoned, and several serious attempts to supply it exist.
In every system of morality which I have hitherto met with, I have always remarked, 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 surprized 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.
Hume, A Treatise of Human Nature III.i.i, 1739. One paragraph, at the end of a section, offered as a curiosity. He adds that the change is imperceptible but of the last consequence, that a reason should be given for it, and that attending to it would subvert all the vulgar systems of morality.
Then he moves on. It is among the most consequential asides in philosophy.
What it does and does not claim
The claim is narrow and precise: no set of premises containing only descriptive statements validly entails a conclusion containing a normative one. Something has been added between the premises and the conclusion, and Hume's request is that whoever added it should say what it was.
Three things it does not claim, all of which are regularly attributed to it.
Not that values are arbitrary. Hume asks for the inference to be shown, not abandoned. Several serious attempts to supply it exist, and the demand is a standard of rigor rather than a verdict.
Not that facts are morally irrelevant. Almost nobody in the argument denies that what is the case bears on what should be done. The question is about the logical structure of the relation, not whether there is one.
Not that ethics is irrational. An argument can be rational without being a deduction from purely descriptive premises. Most ethical reasoning is not that shape and never claimed to be.
Where it bites
The gap is easiest to see in the places it is crossed most casually.
A report establishes what people actually do, and its recommendations say what should be done. Somewhere between the findings and the recommendations, a step was taken that no amount of data supplies.
Appeals to nature have the same shape. This arrangement is natural is a descriptive claim. Therefore it should be preserved requires a premise that natural arrangements should be preserved, which is not itself natural or given.
So does efficiency. That a process produces more output per hour is a measurement. That it should therefore be adopted needs a further claim about what is worth maximizing, and that claim is doing the actual work while looking like arithmetic.
The useful discipline is not suspicion of the conclusion but locating the missing premise and putting it on the table, where it can be argued about.
Moore's is a different argument
G. E. Moore's naturalistic fallacy, from Principia Ethica (1903), gets conflated with this constantly and is a separate claim.
Hume is about inference: you cannot validly get from is-premises to an ought-conclusion. Moore is about definition: good cannot be defined in terms of any natural property — pleasure, or what is desired, or what promotes survival. His test is the open question. For any proposed definition, asking "that may be pleasant, but is it good?" remains a live question rather than a confused one, which suggests the two were never the same concept.
The two arguments support each other and are not the same argument. Conflating them is the most common error made in this area, and it usually shows up as the phrase "naturalistic fallacy" applied to an ordinary is-ought slide.
The responses worth knowing
The gap has been attacked from several directions, and the interesting thing is that the attacks disagree with each other.
Deny it. Anscombe argued in 1958 that the gap is an artifact of having discarded a teleological picture. If a thing has a function, facts about how well it performs that function are already evaluative — a good knife cuts, and no separate premise is needed. That is the argument behind the revival of virtue ethics, and it is why telos matters to this dispute.
Accept it and relocate ethics. Expressivism takes moral claims to be doing something other than describing — expressing attitudes, or endorsing — in which case they were never supposed to follow deductively from descriptions.
Accept it and supply the premise. Most contemporary normative theories do exactly what Hume asked: state a basic normative commitment openly, and argue for it on its own terms rather than smuggling it in.
Concept web
Open the full webQuestions
What is the is-ought gap?
Hume's observation that writers slide from statements about what is the case to statements about what ought to be, without explaining how the second follows from the first.
Is the is-ought gap the same as the naturalistic fallacy?
They are related and distinct. Hume's point concerns inference from descriptive to normative claims. Moore's naturalistic fallacy concerns defining 'good' in terms of a natural property.
Can the gap be bridged?
Several attempts exist, including deriving norms from function or from the requirements of rational agency. None commands consensus, and the burden remains on anyone making the move to show their work.