January 26, 2006

Surprising crocodile kin

It's great having a brother who's a noted science writer, especially one who's a fellow blogger. Today Carl Zimmer's blog ("The Loom") has an entry about his New York Times article describing a fascinating new paleontological discovery: the fossil remains of an ancient reptile related to modern crocodiles and alligators, with a body much like a dinosaur. What's surprising is that the fossil, named Effigia okeeffeae, dates to 210 million years ago, or about 80 million years before dinosaurs evolved similar bodily structures.

In the comments section for the blog entry one can find knowledgeable discussion about whether Effigia should be considered an "ostrich mimic mimic mimic." (This relies on the peculiar sense of "mimic" used by paleontologists, which is evidently applied when a newly discovered fossil resembles a previous discovery — thus when early ostrich-like dinosaurs were found, they were dubbed "ornithomimids," or 'ostrich mimics.') But what caught my eye was a comment about the headline of the Times article: "Fossil Yields Surprise Kin of Crocodiles." A commenter known as "Clueless" wrote:

When I saw the headline, I was wondering how a fossil yield could surprise crocodiles (or their kin), and it took a few moments to figure out what it was intended to mean. Does the author have any control over the headline, or is it completely up to the editors at the newspaper?

To answer the commenter's question, journalists rarely if ever have control over the headlines that are put on their articles, much to the chagrin of writers who wake up to find their painstaking work undercut by a misleading headline. In this case, the headline wasn't factually misleading, only syntactically so. It's a great example of the kind of ambiguous sentence that teachers of introductory syntax classes often present to their students (like the old standby, "I hate visiting relatives"). If this were a diagramming exercise in Syntax 101, the students would have to come up with phrase-structure trees to account for the structural ambiguity:

The ambiguous reading hinges on whether "yields" is understood as a noun or a verb. Once a reader decides to parse "yields" as a plural noun (with "fossil" understood as an attributive modifier), then the garden path has been established. The unusual headlinese of "surprise kin" further encourages the alternate parsing.

A similar ambiguous headline occasionally gets hauled out for the amusement of linguistics classes: "British Push Bottles Up German Rear." Again, the key to the battling interpretations is whether a single word (in this case "push") is parsed as a noun or a verb. I always figured that this headline was apocryphal (one also sometimes sees "French" in place of "British"). But I've seen two references online that say there was an actual headline from World War II along these lines, evidently reproduced in Fritz Spiegl's What The Papers Didn't Mean to Say (1965). The headline given in Spiegl's book reads: "Eighth Army Push Bottles Up German Rear." For American readers this isn't quite as elegant as using "British" or "French," since the ambiguity of Spiegl's headline requires construing "Eighth Army" as plural. That's not a problem for British readers, but in American usage so-called "collective nouns" typically take singular verbs. The ethnonyms "British" and "French," much like "Chinese," can be construed as plural and thus lend themselves to ambiguous readings.

(Another variant on the headline offered by the author Terry Pratchett is "Russian Push Bottles Up German Rear." That doesn't work nearly as well, since the noun "Russian" can only be construed as singular and thus doesn't agree with the verb "push" — unless, of course, one reads "Russian" as a vocative and "Push Bottles Up German Rear" as an imperative. Ouch.)

Posted by Benjamin Zimmer at January 26, 2006 07:05 PM