Naturalistic fallacy

“There’s no rule against it”

At a community pool, a swimmer leaves towels on three lounge chairs at 7 a.m. and comes back at noon. When someone objects, he says: “I checked the posted rules. There’s nothing about saving chairs, so there’s nothing wrong with it.”

The premise is a fact about what the rules say, and it may be true. But whether saving chairs for five hours is fair to other swimmers is a different question, and the rules don’t answer it: they can simply have left it out. The argument works only with the unstated premise “anything the rules don’t forbid is fine”, which is exactly what the objector disagrees with. This is the clear-cut case: a fact about what’s permitted, presented as a verdict about what’s right.

“Studies show, so every parent should”

“Research shows that children who start music lessons young are more likely to reach an advanced level. So every parent ought to start their kids on an instrument by age five.”

This one looks respectable, because the factual premise is real evidence and it is relevant. The gap is in what it takes for granted: that reaching an advanced level is what matters most, and that it’s worth the cost, the time and the child’s other interests. Those are value judgments that parents reasonably make differently. The research can inform the decision; it can’t make it.

“That’s just normal”

“The average adult checks their phone dozens of times a day. So checking it every few minutes is perfectly healthy.”

Here the “is” is a statistic about what’s typical, and the “ought” hides inside an evaluative word: healthy. What’s average and what’s good for you are separate questions. An average can describe a common habit that most people would like to break. If “healthy” is meant as a factual claim (about sleep or attention, say), the statistic isn’t evidence for it either.

Spelled out, the argument has a premise it rarely states:

With the unstated premise added, the reasoning can be perfectly valid. The philosopher Michael Ridge makes this point in the Stanford Encyclopedia of Philosophy: charity means reading an argument like “X is pleasant, therefore X is good” as having a suppressed premise (“whatever is pleasant is good”), not as a logical blunder. So the real question is almost always the missing premise: is it true, and would the person arguing defend it if it were said out loud?

Two different ideas share this territory, and everyday usage runs them together.

• Hume’s is–ought gap. In A Treatise of Human Nature (1739–40), David Hume remarked that in every system of morality he had read, the author reasons for a while with the usual “is, and is not”, then suddenly every proposition is “connected with an ought, or an ought not”. Because “ought” expresses “some new relation or affirmation”, Hume wrote, 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.” This is the idea this page’s examples illustrate: facts alone don’t yield an “ought”.
• Moore’s naturalistic fallacy. G. E. Moore coined the name in Principia Ethica (1903) for a mistake about definitions: treating “good” as meaning the same thing as some other property, such as “pleasant” or “what we desire to desire”. His main argument, later called the open-question argument, was that for any proposed definition you can still sensibly ask, “it’s pleasant, but is it good?”, which you couldn’t do if the two words meant the same thing. Moore applied the name even to defining “good” in terms of a supposed reality beyond nature, so despite the name, it isn’t especially about nature.

The two are related but not identical. Ridge describes the intuitive idea behind Moore’s charge as the thought that evaluative conclusions need at least one evaluative premise, which is close to Hume’s point. But Moore’s target was a claim about what “good” means, and someone can commit it without making any inference at all (“’good’ just means ‘pleasant’”). When writers today say “naturalistic fallacy” for any move from facts to values, they’re using Moore’s name for Hume’s idea. Some treatments keep the two apart. The logic textbook Attacking Faulty Reasoning, by T. Edward Damer, names the everyday move the is–ought fallacy, in a narrow form: assuming that because something is now the practice, it ought to be.

Both ideas are philosophically disputed.

• Critics since W. K. Frankena (1939) have denied that Moore identified a fallacy at all. Ridge writes that the naturalistic fallacy “is very poorly named indeed” and, crediting Frankena, that it is “no fallacy at all”: even if Moore is right, the mistake is a false belief about what “good” means, not an error in reasoning. Frankena also argued that simply asserting that “good” can’t be defined begs the question against those who propose a definition. (Moore himself said he did “not care about the name”.) Thomas Hurka reports that more than a century of debate over the open-question argument “has not led to anything like a consensus on its merits.”
• Hume’s paragraph is contested too. Rachel Cohon writes that “few passages in Hume’s work have generated more interpretive controversy.” The dominant twentieth-century reading, sometimes called Hume’s Law, takes it to say that no evaluative conclusion can be validly inferred from purely factual premises. Other interpreters read it more narrowly, and point out that Hume himself often reaches moral conclusions from factual premises about how character traits affect the feelings of the people who observe them. Some read it only as saying that moral conclusions can’t be demonstrated by reason alone, which, Cohon notes, leaves open whether they can be conclusions of good probable arguments.

Not the same as the appeal to nature. The Appeal to nature reasons from one particular fact, that something is natural, to the conclusion that it’s good or safe. This page is about the general move from any fact to an “ought”. An appeal to nature in ethics (“it’s natural, so we should”) is one instance of it, and Moore himself treated arguments from what is natural as cases of his fallacy. But many appeals to nature are about safety or health, which are factual claims with no “ought”, and most examples of this error (“it’s legal, so it’s fine”) never mention nature.

Two neighbors are easy to mix up with it. The Appeal to tradition treats how things have long been done as how they should be done, and the Appeal to popularity treats what many people believe as true. Damer distinguishes both from the is–ought fallacy. A tradition is kept out of reverence for the past, while the is–ought fallacy argues from how things are done now. And the appeal to popularity argues that a claim is true, while the is–ought fallacy argues that a practice is right. “Most people do it, so you should” sits where they meet.

Facts are part of almost every good argument about what to do. The error is letting them do all the work.

• The value premise is stated, or genuinely shared. “The bridge is rated for ten tons and your truck weighs fourteen, so you shouldn’t drive over it” relies on a goal no one disputes (not collapsing the bridge). Nothing is being smuggled in.
• The “ought” is relative to a stated goal or rule. “If you want the bread to rise, you ought to let the dough rest” and “in chess, you must get out of check” draw their “ought” from a goal or a set of rules that the speaker names.
• The claim is only descriptive. “Most offices close on this holiday” says what’s common and draws no conclusion about what anyone should do.
• The facts change which values apply. Learning that a practice harms people is relevant to whether it’s acceptable, given the shared premise that causing harm matters.

The test: what value premise would make this conclusion follow, and would the speaker defend it if it were said out loud?

The load limit

“The sign says the footbridge holds four people at a time. There are already four of us on it, so you should wait until someone gets off.”

This moves from a fact (the load limit and the head count) to an “ought”. But the missing premise is one everyone on the bridge accepts: nobody wants it to fail. Because the goal is shared and obvious, spelling it out would add nothing. This is the stated or genuinely shared condition above.

The towels, argued properly

“Leaving towels on those chairs keeps everyone else from using them all morning while you’re not even here. Taking a shared thing away from other people for no reason isn’t fair. So you shouldn’t do it.”

This reaches the same kind of conclusion as the pool example, and it uses facts. The difference is that the moral premise is right there in the argument, where the swimmer can see it and dispute it. The facts matter (the chairs sit empty all morning), but a stated value turns them into a reason.

The value premise is often so widely shared that nobody notices it’s missing. “It’s legal, so it’s fine” sounds complete to anyone who, most of the time, agrees that legal things are fine. Hume called the shift from “is” to “ought” “imperceptible; but ... of the last consequence”: easy to miss, and decisive once it happens.

Facts also carry an authority that value judgments lack. Pointing to a law, a study or an average feels objective, and it moves a disagreement about values onto ground that seems more settled. The argument borrows that authority for a conclusion the facts can’t reach alone.

• Supply the missing premise. Say what would have to be true for the conclusion to follow (“so you’re saying anything that isn’t banned is fine?”) and discuss that.
• Don’t treat “you can’t get an ought from an is” as a refutation. The conclusion may still be right, for reasons nobody has stated yet. Pointing out the gap shows the argument is incomplete, not that its conclusion is wrong.
• Keep the facts. They usually matter. The aim is to put them next to an explicit value, not to throw them out.