Disagreement and the Semantics of Normative and Evaluative Terms

January 9, 2026 | 896 words | 5min read

Paper Title: Disagreement and the Semantics of Normative and Evaluative Terms

Link to Paper: https://philpapers.org/rec/PLUDAT-2

Date: 2013

Paper Type: Conceptual Engineering, Meta-Ethics, Moral Expressivism, Moral Language

Short Abstract: This paper argues that genuine disagreement in normative terms (such as good) does not automatically imply that speakers share the same meaning. Instead, it claims that many disagreements are cases of metalinguistic negotiation, where speakers put forward different definitions or standards for a term because they want others to adopt them.

0. Introduction

The paper begins with Hare’s classic cannibals/missionary example from The Language of Morals. Both groups use the word good, but apply it to different kinds of actions:

“[T]hey know that when he uses the word he is commending the person or object he applies it to… They themselves are accustomed to commend people who are bold and burly and collect more scalps than average.”

At first glance, the dialogue between the cannibals and the missionary seems like a genuine disagreement. If meaning were not fixed, then they would merely be talking past each other, which clashes with our strong intuition that they disagree.

Hare uses this intuition to infer shared meaning: if there is a real disagreement, they must mean the same thing by good and disagree about what falls under it.

The author, Plunkett, argues that this move is flawed. Hare infers:

There is a genuine disagreement → therefore, the speakers mean the same thing.

Plunkett rejects this. He argues that speakers can genuinely disagree even when they do not mean the same thing by a normative term.

So how can the cannibals and the missionary disagree if they mean different things by good? According to Plunkett, the disagreement is a case of metalinguistic negotiation: a dispute about how a term should be used. When speakers put forward competing metalinguistic claims through ordinary usage, they negotiate the proper concept for the context.

Plunkett calls this conceptual ethics: normative disputes about concept choice and linguistic practice. These can be explicit or implicit.

Why does this matter?

• Traditional metaethical arguments that rely on disagreement intuitions (like Hare’s) lose their force.
• Contextual or speaker-relative semantic theories can still make sense of disagreement.
• Metaethics should stop using disagreement intuitions as quick refutations.

1. Disagreement-Based Arguments in Meta-Normative Theory

This section introduces the type of argument the paper criticizes, called disagreement-based arguments. These arguments move from the intuition of genuine disagreement to a conclusion about shared meaning.

Examples:

Horgan & Timmons — Moral Twin Earth (1993)

People on Earth and “Twin Earth” use moral terms regulated by different natural properties. To preserve the intuition that disagreements would be real, they claim both groups must share meanings.

Lasersohn — Taste Predicates (2005)

Lasersohn argues against contextualism for predicates like fun, because contextualism makes disagreements look trivial (e.g., “fun for me” vs. “not fun for you”) instead of genuine.

The Common Pattern

These arguments follow a common pattern:

1. There is a strong intuition of genuine disagreement.
2. Genuine disagreement involves conflicting literally-expressed contents.
3. Conflicting contents require shared meaning.
4. Therefore, we must posit shared meaning.

Plunkett and Sundell reject premise (2): the idea that the only explanation for genuine disagreement is literal conflict. They argue that many disagreements are non-canonical, especially metalinguistic ones.

2. Definitions

2.1 Meaning

The paper restricts meaning to semantics, not the full communicative effect (which includes pragmatics). Meta-normative theories combine both.

2.2 Canonical Disputes

Canonical disputes involve literal conflicts between expressed contents (propositions, plans, etc.).

2.3 Word Individuation

Metalinguistic disputes involve using the same word to express different concepts or negotiate usage.

2.4 Disagreement

Disagreement is a state involving rationally incompatible mental contents (beliefs, desires, plans, etc.). It need not be linguistic or explicit.

3. Non-Canonical Disputes

3.1 Non-Canonical Disputes and Genuine Disagreement

Genuine disagreement does not require literal incompatibility. Example:

(a) “There is one proton in a helium nucleus.” (b) “No, there are two.”

Literally compatible, but pragmatically incompatible (exactly-one vs. exactly-two).

3.2 Metalinguistic Disputes

These generalize non-canonical disputes to cases involving meaning variation.

Example: tall or spicy disputes.

Speakers do not disagree about facts, but about standards. Such disputes often serve practical coordination functions (e.g., how spicy is too spicy?).

Other examples:

• “Pluto is a planet.”
• “Missouri is in the Midwest.”
• “War on drugs.”

These disputes concern conceptual ethics, which concept is best to use.

4. Normative and Evaluative Disputes

The main claim: Many normative disputes are metalinguistic negotiations.

Implications:

• Genuine disagreement can coexist with meaning variation.
• Disagreement alone cannot justify inferring shared meaning.
• Contextualist or relativist theories can accommodate disagreement.
• The Shared Meaning Task in metaethics is misguided.

Examples:

“Torture” — Negotiation between UN-style broad concept vs. U.S. DOJ narrow concept. Stakes include legal and moral consequences.

Moral Twin Earth — Can be reinterpreted as negotiation over which concept of “morally right” should play key roles in guiding life, social relations, and responsibility.

4.1 Are Metalinguistic Negotiations Worth Having?

Yes. Conceptual ethics can track objective or practical stakes (e.g., biologist vs. chef on fruit). Even if both sides are “literally true,” the dispute matters for classification, policy, coordination, etc.

5. Methodological Implications

• Disagreement intuitions alone cannot establish shared meaning.
• Many normative and evaluative disputes are non-canonical.
• This rescues contextualist and relativist views from standard objections.
• The assumption of shared meaning in metaethics may be mistaken.
• Speakers can mean different things while genuinely disagreeing.

Metalinguistic analysis vindicates ordinary intuitions that such disagreements are genuine and worth having.

reply via email