Introduction
In the study of linguistics, the phenomenon of "meaning transfer"—where words shift to describe related concepts—has traditionally been viewed through the lens of metonymy. When a speaker says, "I am parked out back," conventional wisdom suggests a referential shift: the pronoun "I" is simply standing in for "my car." However, in his invited talk, Geoffrey Nunberg challenges this assumption. He argues that such expressions are not instances of referential metonymy, but rather predicate transfers, where the meaning of the property being ascribed shifts to describe a new attribute of the subject.
The Problem with Referential Shifts
Linguists have historically categorized meaning transfers into disparate camps, including lexical rules (e.g., "grinding" an animal name into a meat name), semantic composition, and pragmatics. The primary challenge has been the lack of a unified account that explains why certain transfers are linguistically permissible while others feel unnatural.
Nunberg addresses this gap by analyzing how these shifts interact with grammatical constraints. If "I" in "I am parked out back" truly referred to a car, the sentence should behave like a sentence about a car. Nunberg uses "linguistic probes" to demonstrate that this is not the case.
The Evidence for Predicate Transfer
The Number Test
One of Nunberg’s most compelling arguments against referential shift is the "Number Test." Consider a scenario where two people share a single car. If they are asked about their parking situation, they would say:
"We are parked out back."
If the subject "We" had undergone a referential shift to mean "the car," the plural form would be ungrammatical because there is only one car. The fact that the subject remains plural proves that "We" still refers to the people. Consequently, it is the predicate—"parked out back"—that has shifted its meaning from "being a vehicle located in a spot" to "being the owner of a vehicle located in a spot."
The Reflexive Puzzle
Nunberg further validates this through the behavior of reflexives. In the sentence, "Ringo squeezed himself into a narrow parking space," the reflexive pronoun "himself" must match its antecedent, "Ringo," in gender and number. If "himself" referred to a car, it would violate standard English syntactic rules. By treating this as a predicate transfer, the syntactic identity remains intact: "himself" still refers to Ringo, even though the action described involves the maneuvering of his vehicle.
Conditions for Transfer
Nunberg proposes that for a predicate transfer to be successful, it must satisfy two primary conditions:
- Salient Correspondence: There must be a clear, functional mapping (an injective function) between the two domains. In the parking example, there is a direct correspondence between a car's location and its owner's status.
- Noteworthiness: The transferred property must be "criterial" or useful to the conversation. We can say "I am reading a Picasso" because the artist's style is a unique, noteworthy property of the work. Conversely, a transfer might fail if the relationship between the two entities is not sufficiently salient or relevant to the context.
The Three-Layered Rule System
To unify the various approaches to meaning transfer, Nunberg suggests a hierarchical framework consisting of three levels:
- Nonlinguistic: General cognitive associations between entities, such as the relationship between a creator and their work.
- General Linguistic: The universal "machinery" of language that allows for type-shifting and category changes.
- Language-Specific: The specific "licenses" granted by a particular language. For instance, some languages may allow the "grinding" of tree names into wood names, while others require distinct terms.
Conclusion
By reframing metonymy as a shift in property-assignment rather than a shift in reference, Nunberg provides a more robust explanation for how language maintains syntactic integrity during meaning transfers. This framework demonstrates that "syntactic identity"—the agreement between pronouns and verbs—is preserved even when the underlying meaning becomes complex. While the formal mathematical details of the injective functions involved remain a subject for future elaboration, Nunberg’s work offers a vital bridge between cognitive associations and formal linguistic rules.
Summary updated