Skip to content

Causal Theory of Reference

A semantic theory on which a name or natural-kind term refers through an appropriate historical-causal chain linking current use to an initial fixing or grounding in its referent, rather than through a uniquely identifying description held by each speaker.

Version
v2 · 2026-09-19 · History
Domain-specific #
1444
Domain group
Humanities
Origin domain
Philosophy
Subdomain
Philosophy of Language → Philosophy
Aliases
Causal-historical theory of reference, Historical-chain theory of reference, Causal chain theory

Core Idea

The Causal Theory of Reference is a family of philosophical accounts on which at least some expressions refer because present uses stand in an appropriate historical-causal relation to their referents. For a proper name, an initial baptism or grounding connects a name to an object; speakers then acquire the name from other speakers with an intention to use it as their source used it. The ability to refer can pass along a communicative chain even when later speakers lack a uniquely identifying description or hold important false beliefs about the bearer.

How would you explain it like I'm…

The Name-Passing Chain

When a baby is born, someone gives it a name. Then people pass the name along — you learn it from someone, who learned it from someone else, all the way back to the naming. The causal theory of reference says that chain is what makes the name point to that person, even if you don't know much about them or believe some wrong things about them.

How Names Get Handed Down

How does a word like a person's name end up meaning a certain person? The causal theory of reference says it's because of history. At first someone gives the name to a person or thing, like at a naming ceremony. Then people learn the name from each other, each one meaning to use it the same way as the person they learned it from. Because of that chain, you can use a name to talk about someone even if you only know a little about them, or even believe some wrong things about them. A similar idea works for words like 'water', where the word points to the real stuff in the world, not just to what's in your head.

Historical-Chain Reference

The causal theory of reference is a family of views about how words hook onto things. For names, the idea is that an initial 'baptism' links a name to an object, and afterward speakers pick up the name from others, intending to use it the way their source did. This communication chain lets later speakers refer to the object even without a description that picks it out uniquely, and even with false beliefs about it. Kripke and Putnam extended similar ideas to natural-kind words like 'water': the word was first tied to samples, and later speakers, deferring to experts and the community, keep referring to the same underlying kind. So what 'water' means isn't fixed just by the picture in your head; your environment and the word's history matter. Defenders admit it isn't simple: not every causal link counts, reference can shift, and fictional or empty names need extra treatment.

 

The causal theory of reference is a family of accounts on which at least some expressions refer in virtue of an appropriate historical-causal relation between current uses and their referents. For proper names, an initial baptism or grounding fixes the name to an object, and subsequent speakers acquire it from others with the intention of using it as their source did; reference is thereby transmitted along a communicative chain, even when later speakers lack any uniquely identifying description or hold important false beliefs about the bearer. Kripke and Putnam developed related externalist accounts for natural-kind terms: a reference-fixing event, ostensive or descriptive, connects the term to samples, and communal use with deference to experts preserves reference to the same underlying kind, so the content of 'water' is not fixed solely by a speaker's internal stereotype. The theory is not the claim that bare physical causation mechanically fixes meaning: accidental causal contact is insufficient, transmission must be of the right kind, reference can shift over time, names can have multiple groundings, descriptive material can help fix or disambiguate reference, and empty or fictional names require further treatment. What unites the family is historical grounding plus socially transmitted, deferential use with appropriate continuity, yielding reference not mediated by an identifying description in each speaker's mind.

Scope of Application

The theory is central to philosophy of language, modal semantics, metaphysics of natural kinds, semantic externalism, and debates over proper names. Kripke's examples challenge descriptivism by showing that a speaker can refer to Gödel or Aristotle despite ignorance or error, and can consider counterfactual situations in which the bearer lacks familiar descriptions. Putnam's Twin Earth cases extend externalist pressure to kind terms.

Applications must distinguish names from natural kinds. A name chain typically passes use of an expression for an individual. A natural-kind account also invokes samples, underlying structure, scientific discovery, and social deference.

Clarity

Separate reference fixing from reference transmission. A baptism can use a description—“let ‘Neptune’ name the planet causing these perturbations”—to establish reference. Later speakers need not associate that same description, because they defer to the established use. The description's fixing role does not make it the name's meaning in every later mouth.

Manages Complexity

Reference persists across large communities whose members possess radically different information. A purely descriptive theory seems to require each speaker to carry enough uniquely identifying content. The causal-historical account distributes that burden: some users have direct grounding, others borrow the established use, and experts may stabilize kind classifications. The chain explains how imperfect knowledge can coexist with coordinated reference.

Abstract Reasoning

  1. If a speaker learns a name from a community and intends to preserve its use, ignorance of the bearer need not break reference.
  2. If every associated description is false, successful reference can persist when the historical chain remains appropriate.
  3. An accidental causal path—hearing a noise caused by a person and inventing an unrelated name—does not automatically confer reference to that person.

Knowledge Transfer

The exact theory transfers among names, natural-kind terms, and some institutional or artifact terms only when grounding, historical transmission, and deference can be specified. The relevant form of grounding differs, so one should not assume the strongest natural-kind essentialism for every noun.

Database identifiers, provenance chains, and versioned handles offer useful analogies but are engineered reference systems with explicit rules. The general transferable structure is lineage-based identity and provenance. The philosophical node remains domain-specific because linguistic intention, community use, modal claims, and semantic reference are constitutive.

Relationships to Other Abstractions

Local relationship map for Causal Theory of ReferenceParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Causal Theoryof ReferenceDOMAINPrime abstraction: Causality — includesCausalityPRIME

Current abstraction Causal Theory of Reference Domain-specific

Parents (1) — outgoing typed relations

  • Causal Theory of Reference includes Causality Prime

    current reference depends on historical events and transmission.

Hierarchy path (1) — routes to 1 parentless root

Neighborhood in Abstraction Space

Causal Theory of Reference sits in a sparse region of the domain-specific corpus (98th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2769 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-25