Old and New Information in Amy Hempel's "Tonight is a Favor to Holly"
Hempel's narrator toys with expectations about what information can be taken as given.
Amy Hempel’s “Tonight is a Favor to Holly” is one of my favorite short stories, partly for its deep, dark humor. Some of the humor derives from the ways in which Hempel plays with the reader’s expectations about new and old information.
Information in a sentence comes in two basic kinds: given or new. (Inferred is a third kind that I will set aside.)
A simple example of old vs. new in noun phrases illustrates the distinction:
A woman entered the room. vs. The woman entered the room.
In noun phrases, the use of a definite determiner (the, this, that) implies that the noun that follows is given information—something that is already familiar to the reader. If the information were not given, then the indefinite determiner (a/an) would be used. In the case of “a woman,” that woman’s entrance carries no expectations. With “the woman,” however, the speaker assumes that the hearer accepts the woman as part of the “common ground” in the conversation—she is not new.
More complicated constructions also encode assumptions about the common ground. When the verb “to be” is used in a two-part construction called a “cleft,” expectations are baked into each part of the sentence about what is new to the conversation between narrator and reader. Wh-cleft sentences look like a broken-up version of a simple sentence where a single message is divided into two clauses.
In addition to the sentence’s main verb being “to be,” the subject consists of a clause introduced by a wh-term—usually, what. When employed conventionally, wh-clefts serve to connect what is already understood (old) to what is new to the listener/reader. That is, the first part provides “given” information, and what comes after the verb is “new”: What X (old) is Y (new).These examples of clefts reveal opinions of the narrator with what she chooses to cast as “given” in this description of her temporary home, an apartment in a complex with an aesthetic of “fake Spanish Colonial”:
What’s irritating is that the tiles were chemically treated to “age” them from the start. What you want to say is, “Look, relics are leftovers, you know?”
“Irritating” is in the position of “given” information—but we did not know that there was something irritating. That makes us slow down. The case is similar with “What you want to say is…” This wanting to say something is likewise not a given, so we have to adjust and put ourselves firmly in the mind of the narrator, who is worked up about the tiles and the décor of her current apartment complex. She is irritated. And she wants to say certain things, but does not. Both of these castings of phrases are character-developing in what they reveal about the narrator’s state of mind.
Clefts are wordier relative to their unclefted counterparts. Compare the first example, What’s irritating is that the tiles were chemically treated to “age” them from the start, with the alternative: That the tiles were chemically treated to “age” them from the start is irritating. The wh-cleft version forces the reader to linger, and the information structure is drawn into relief. Hempel capitalizes on the nuances of both focus and timing to get from these sentences the maximal humor, suspense, and heartache, adding a conspiratorial quality. The structure assumes that the reader already knows that something is irritating to the narrator; her irritation is given—per the conventions of information status within wh-clefts—and not new.
The next example also reveals character.
What you forget, living here, is that just because you have stopped sinking doesn’t mean you’re not still underwater.
By beginning with “what you forget,” the protagonist’s pessimism is revealed; she presumes forgetting, draws it into the foreground. In this example, consider the simple sentence that it is derived from: Living here, you forget that just because you have stopped sinking doesn’t mean you’re not still underwater. With the more complex construction, the reader’s attention is directed to the “new information,” the clause “just because you have stopped sinking doesn’t mean you’re not still underwater.” The light thus shone on this phrase gives us a better chance to appreciate its depth and despair.
While the wh-cleft typically functions to draw attention to the salient part of the speaker/writer’s message, Hempel uses the construction to draw attention to an unexpected part. The part of the sentence set in relief is the one that makes us pause, laugh, and/or wince. Ironically, the only thing the reader knows when encountering one of Hempel’s wh-clefts is that the sentence will not end in any way we could guess.
In this final example, Hempel presents old information as new information:
The beach is near the airport—so this town doesn’t even have the class L.A. lacks. What it has is airline personnel.
The first “is” is simply predicational (is near the airport). The second, however, is clefted (What it has is __ ). We expect what follows this “is” to be some sort of predicate, parallel to “class,” contrasting with it. Hempel primes us for the expectation of an attribute, along the lines of “class” (it doesn’t have class, but it has ___ ). By filling the blank with this surprisingly predictable (given the context) set of people, as opposed to an attribute, the reader has a small moment of “what just happened?” to figure out where their expectations were defied. Instead of a revelation in this position reserved for new information, we get information we could already infer from our knowledge about how airports work. But there is humor in the defiance. The inferable nature of the second phrase suggests an extreme frankness on the part of the narrator.
In this short story, Hempel takes full advantage of readers’ expectations about information newness (or oldness) built into the wh-cleft construction to subvert expectations, use humor, and reveal the mindset of her narrator.
Higgins, Francis Roger. 1979. The Pseudo-Cleft Construction in English of Outstanding Dissertations in Linguistics. Garland, New York.