August 25, 2006

New planetary definition a "linguistic catastrophe"!

Owen Gingerich, chairman of the International Astronomical Union's Planet Definition Committee, is quite distressed about the resolution passed by the IAU's General Assembly in Prague yesterday, wherein "planets" (encompassing Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, and Neptune) are distinguished from "dwarf planets" (the newly demoted Pluto, along with Xena and Ceres, with many more on the way). Gingerich told the BBC that the resolution was reworked after a "revolt" by planetary dynamicists, who felt "terribly insulted" that the original definition of a planet was drafted with a focus more on geology than dynamics. What's more, only 424 of the 2,700 attendees in Prague ended up voting (Gingerich himself didn't vote since he had to catch a plane back to the U.S.), or about 4 percent of the IAU's total membership of 10,000.

Gingerich laid out his objections in stark terms to the Guardian:

"We now have dwarf planets which are in fact not planets. I consider this a linguistic catastrophe. I think the union is going to get a lot of flak for this, in doing it in such a muddy way."

As I mentioned in my post yesterday, distinguishing "dwarf planets" as non-planetary runs counter to our expectations of hyponymy in English. In a compound noun of the form A-B, we generally assume that the compound is composed as a hyponym, a particular type of a more general category B, which in turn is called a hyper(o)nym. So alley cats are types of cats, rocking chairs are types of chairs, bay windows are types of windows, and so forth. The impulse towards hyponymic compounding is so strong that we sometimes form such compounds out of thin air, just to make sense out of a semantically opaque word — consider such folk etymologies as sparrowgrass, reinterpreting asparagus as a type of grass, or crayfish, reinterpreting French (é)crevisse as a type of fish (or at least a fishy thing).

So the fact that the IAU would like us to think of dwarf planets as distinct from "real" planets lumps the lexical item dwarf planet in with such oddities as Welsh rabbit (not really rabbit) and Rocky Mountain oysters (not really oysters). In a 2004 article in American Speech, Larry Horn dubbed such formations ironyms, since they "represent lexical irony."

Should we think of dwarf planet as the latest ironym, then? I doubt the astronomers in Prague really had lexical irony in mind — rather, the planet vs. dwarf planet distinction emerged as a somewhat messy compromise between different stances on planetary redefinition. Though the result may not be quite the "linguistic catastrophe" that Gingerich envisions, it still opens the door to some unintended possibilities. For instance, if Dutch people get offended at pejorative ironyms like Dutch courage meaning 'false courage (brought on by drunkenness)', then will dwarf humans take offense at the dwarf planet distinction, on the grounds that it implies that dwarfs are somehow not quite human? I haven't seen any such outcry yet, though some famous fictional dwarfs did release their own statement yesterday.

[Update #1: I see Gingerich himself raised the question about dwarf humans, as quoted in the Washington Post:

"Pluto is a dwarf planet, but we are now faced with the absurdity that a dwarf planet is not a planet,'' Gingerich retorted. ''Is a human dwarf not a human?'' ]

[Update #2: Betsy McCall writes:

Even worse than defining a "dwarf planet" as not-a-planet, the difference between a dwarf planet and a planet has NOTHING to do with its size! We could - and are likely to - find objects in the Kuiper Belt the size of Mars and they would be called dwarf planets even though they are significantly bigger than Mercury.

Indeed — what makes a dwarf planet a non-planet according to the IAU resolution is that it "has not cleared the neighbourhood around its orbit" (something of a contentious point among astronomers right now). This is apparently what happens when you let dynamicists define the terms.

And Robert Cumming of the Stockholm Observatory writes:

One thing that no one seems to have mentioned yet is that hyponymy in the solar system is nothing new. The name 'minor planet' been more or less synonymous with 'asteroid' for a very long time. So it seems to me pretty insane to complain about any ambiguity or risk for confusion with the introduction of 'dwarf planet'.

I've actually seen this point raised in some of the coverage of the IAU vote — for instance, in this New Scientist report. As I understand the history, astronomers of the early 19th century wanted to classify the first discovered asteroids — such as Ceres, the largest — as planets, and the minor planet label simply stuck around even after asteroids were classified as distinct from planets. The difference with "dwarf planets" is that they're being defined as non-planets from the outset, rather than reflecting a bygone terminological system. In any case, it looks like minor planet has been eliminated by the IAU. Ceres now joins the dwarf planets Pluto and Xena, while all other former "minor planets" are to be known as "small solar system bodies."]

[Update #3: Anatoly Vorobey objects to the hyponymy argument:

A sea cow is not a cow. A guinea pig is not a pig (nor is it from Guinea). A sea lion is not a lion, and a sea horse is not a horse.
A koala bear is not a bear. A buffalo bison is not a buffalo. An aardvark is not a pig (vark).
All of these examples are perfectly ordinary and aren't really perceived as oddities by anyone except when using them to make a joke. In fact, some of them are formed rather regularly, like the "sea" compounds. Hyponymy in compounds may be the usual case in English, but that doesn't mean that non-hyponimic compounds are somehow problematic or unintuitive - we're living with many of them and using many of them daily without giving them a second thought. Insisting that they must in some way be ironic seems fatuous and unnecessary.
Is it really a problem that "dwarf planets" aren't really planets? Of course not, no more than the fact that a sea lion is not a lion. Is it going to confuse people? Well, are they regularly confused by guinea pigs?
Why make a mountain out of a molehill? Especially considering that it's not really a hill at all.

This is a fine and valid point. But I think we're able to recognize that, say, sea lion and sea cow are not compositionally typical hyponyms because we know that actual lions and cows aren't sea creatures. Thus we have sufficient knowledge to interpret those compounds metaphorically. (Except, perhaps, for Jessica Simpson, who was notoriously stumped by "Chicken of the Sea.") But how are laypeople supposed to grasp that dwarf planets are not to be considered planets at all, despite seeming to fall in the same natural class (which can't be said for pigs and guinea pigs, and so forth)? I don't think minor planet was known well enough outside of astronomical circles for this to have become an issue with that term, but now dwarf planet is getting massive exposure because of the demotion of Pluto. So I agree with Owen Gingerich that confusion will continue to reign over this terminological distinction (even if that confusion isn't "catastrophic").]

Posted by Benjamin Zimmer at August 25, 2006 08:12 AM