Jeffrey Kaplan argues that the centuries-old debate over how names work ends not with a compromise, but with a synthesis that renders previous theories obsolete. He posits that Gareth Evans's 1973 breakthrough offers the only "one true philosophical theory of names" by merging the descriptive richness of earlier models with the causal necessity of modern ones. For the busy listener, this is not just a history lesson; it is a definitive resolution to a problem that has puzzled logicians for decades, suggesting that our understanding of language is finally complete.
The Duel of Giants
Kaplan sets the stage by contrasting two dominant but flawed frameworks that defined the field for years. He introduces the first, Frege's cluster theory (often attributed to Russell or Searle in this context, though Kaplan refers to "Sorl"), which suggests a name refers to whoever fits a specific set of descriptions. "According to Sorl's theory, you have a name and associated with that name, Napoleon Bonapart, is a bunch of information, a cluster of descriptions," Kaplan explains. The logic is intuitive: we know Napoleon because he conquered Europe and crowned himself. However, Kaplan notes that this approach crumbles when facts are wrong. If history books lied and David actually killed a different giant, the name "Goliath" would, under this theory, refer to the wrong person entirely.
The second challenger is Saul Kripke, who Kaplan describes as a "young prodigy" who dismantled the descriptive model in 1970. Kripke proposed a causal chain theory, where a name is passed down like a baton. "When Napoleon was a baby, his parents... dubbed him Napoleon," Kaplan paraphrases, "and then these other people... knew about the name and they thought to themselves something along the lines of, 'Now I'm going to use the name Napoleon and I'm going to use it to refer to whichever baby these people use their name Napoleon to refer to.'" This model prioritizes the origin of the name over the facts we know about the person. While this solves the "wrong description" problem, Kaplan argues it creates a new, equally absurd set of errors.
The absurdity in supposing that the denotation of our contemporary use of the name Aristotle could be some unknown item whose doings are causally isolated from our body of information is strictly parallel to the absurdity in supposing that one might be seeing something one has no causal contact with solely upon the ground that there is a splendid match between object and visual impression.
The Madagascar Paradox
Kaplan's most compelling evidence comes from Evans's famous "Madagascar" example, which exposes a fatal flaw in Kripke's rigid causal chain. Kaplan details how the explorer Marco Polo misunderstood locals who were referring to a piece of the African mainland as "Madagascar," while Polo applied the name to the island off the coast. Because Polo was influential, the causal chain of usage shifted to the island, even though the original "baptism" referred to the mainland. "The causal chain of Kripke's theory leads back to a piece of the African mainland," Kaplan writes, "so Kripke's theory has to say that Madagascar the name refers to this area in Mozambique but it doesn't."
This is where the theory breaks down. The name clearly refers to the island, not the mainland, yet the causal chain points to the mainland. Kaplan uses a second example, the "baby swap," to drive the point home. If a baby named Jaylen Brunson is swapped at birth and the other baby grows up to be the Knicks' point guard, Kripke's theory would insist the name belongs to the biological baby, ignoring the fact that the basketball player is the one everyone knows and describes. "The name Jaylen Brunson clearly refers to this guy who plays for the New York Knicks," Kaplan asserts, "and he answers to the name Jaylen Brunson and that's just his name obviously." Critics might argue that Kripke could adjust the theory to account for "dominant usage," but Kaplan suggests this is just patching a leaky boat; the fundamental mechanism is flawed.
The Evans Synthesis
Enter Gareth Evans, who Kaplan claims "just ends here with Evans" because he gets it right. Evans realizes that both previous theories captured half the truth. We need the causal link to ensure we aren't talking about a random object that happens to match our description, but we also need the descriptive cluster to ensure we aren't talking about a stranger we have no connection to. Kaplan illustrates this with a visual analogy: seeing a shoe. "The light bounces off the shoe and goes into the eyeball," he says, creating a causal link. Even if a shoe a thousand miles away looks exactly like the one in your head, you are seeing the shoe that caused the image, not the distant one.
However, Evans flips the script on how descriptions work. They don't define the reference; they are the result of the reference. "There is a cluster of descriptions or a body of information that's associated with a name and that's important," Kaplan notes, but the reference isn't determined by fitting those descriptions. Instead, the reference is determined by the causal source of those descriptions. "One way that someone could be related to this description is that they could fit it... However, there's this other way that someone could be related to this description. They could be the source, the causal source of that description." This elegant solution preserves the intuition that names are grounded in reality (causality) while acknowledging that our knowledge of the world (descriptions) is what sustains the name's usage.
The absurdity in supposing that one might be seeing something one has no causal contact with solely upon the ground that there is a splendid match between object and visual impression.
Bottom Line
Kaplan's strongest contribution is his ability to distill complex philosophical arguments into relatable narratives like the "Madagascar" mix-up and the "baby swap," making the abstract concrete. The argument's greatest vulnerability lies in its reliance on Evans's specific synthesis, which, while elegant, may still struggle with edge cases where the "dominant source" of a description is ambiguous or contested. Readers should watch for how this theory holds up against modern digital identities, where the causal chain is often broken or simulated.