The Pragmatics of Spell-Casting in Harry Potter. This is what happens when linguists have too much time on their hands, and a perfect demonstration of why linguists should have too much time on their hands more often. The observations on wand deixis and spell syntax are delicious just on their own, but for me the real high point of this paper is the connection between fantasy magic and the boring everyday human magic that is language. I only wish it were longer and deeper. (It completely ignores non-verbal magic, for instance, which is understandable but still an unfortunate oversight.)

(Via Heidi Harley at Language Log, which I've started reading again so maybe I'll actually blog here occasionally.)

Determine the exact percentage of the time that people mean the phrase "to be led on a merry chase" non-sarcastically.

Reported without comment, other than "snrk": a new Maui TV station has, apparently inadvertantly, been given an unprintable call sign. Hee. (Wait, was that "hee" a comment?)

Idea for a code: Both parties are supplied with a huge corpus of English text, preferably one containing all the words that might be needed in a conversation. A computer program pulls out all of the words in the text and orders them in some fashion—perhaps alphabetically, or better yet by frequency first and then alphabetically. When encoding a message, the person looks up each word in this table and determines what number to give it. Let's say "antediluvian" is the 725^th word on the list, so every time the word "antediluvian" appears in the list, you replace it with the number 725. Then—and this is the clever part—you pad each number with a sequence of random digits, so that "725" now becomes "7251135" or "72514" or "72515869415". Finally, you add one last digit which tells the savvy reader how many of the preceding digits to pay attention to, in this case three. So, you get "72511353", "725143", and "725158694153". If a word is outside the corpus, you encode it using a more conventional system, and then append a short numeric code that flags it as being different. You could even have a series of prearranged alternate codes, each with its own number. So if you use code number 123 to encode the word "numbat" and get "16465625" (or whatever), "numbat" gets replaced with "16465625123". (Obviously, if your alternate code is called "123" you need to make sure that the random numbers you're adding to the normal words never end in "12" to prevent confusion with word number 164, but that's simple enough to do.)

I admit, much of the security is through obscurity—people just wouldn't expect such a system—but it's still kind of cute.

If the NSA steals this, I expect to see some royalties. Or, at the least, I expect them to stop listening to my phone calls illegally.

A disturbing thought came to me out of nowhere the other night—what if humans aren't actually particularly intelligent by animal standards, but only seem more intelligent than we are because we can share our discoveries with each other? I consider myself smarter than my mother's cat, but I've been absorbing unspeakable amounts of linguistic data for twenty-five years. (Especially since I discovered Wikipedia, of course.) I've mostly figured out how to use AutoCAD with a minimum of help and Mishki hasn't, sure, but somebody else told me to turn a computer on in the first place. Nobody tells Mishki anything, except "Mishki, stop untying my shoelaces!" (He does not listen.) But without anybody telling him how to open the bathroom door, he's sussed it out through observation and experimentation. If I didn't have language, would I do better?

I thought about non-linguistic humans—the isolated deaf, feral children, foreigners, and so on—but that's not really fair either, because even non-linguistic humans are part of human society, which is inherently linguistic, and it's much easier for a human to communicate non-verbally with another human than with a different animal, for a host of reasons. In any case, I don't really know enough about them to compare them to my mother's cat, intelligence-wise. But I'm not convinced that if you completely isolated a human infant and then pitted the kid against Mishki—or a dolphin, or raven, or elephant, or hell, maybe even a cephalopod (if you can find one that isn't too EVIL)—in a well-controlled battle of wits, the human would win.

(Of course, an experiment like that is fraught with problems well beyond the moral ones—namely, that raising a child without language means raising a child without human society, and that has broad implications for intelligence. Plus "intelligence" is a flawed concept anyway. Plus Mishki would never cooperate. Still, thought experiments are fun!)

The word "stupefy" dates from 1513. Could wizards not Stun people before then? And could they not find north before the English language had developed enough to say "point me"?

I mean, clearly just about every spell in the HP universe was invented some time after the rise of Rome (with the possible exception of Avada Kedavra) despite the fact that there are wizards all over the world and have been since at least ancient Egypt. No Chinese wizards have come up with useful spells lately? No useful little Sumerian spells survived the Dark Ages? Or is Dog Latin the scholarly language of choice worldwide? If pronunciation is so important that saying the wrong thing can leave you under a buffalo, why do they all pronounce the names of spells like English people reading funny foreign words? I mean, lax vowels? Diphthongs? I've heard amateur choirs pronounce Latin better than wizards do!

But that just raises the question: how do you invent a new spell, anyway? I can understand coming up with a new potion, but how do you discover that saying "waddiwasi!" while wiggling your wand a certain way makes chewing gum fly into a poltergeist's nose? (I always find it amusing when Lupin tells everybody how useful a spell that is. So many applications!) Is it all mental—is the incantation no more than a focus? Could you actually just go around saying "noamchomsky!" instead and have everything work? Because I'm going to try that.

Language Log wins again!

I do not know why there are not more Chomskian-themed advertising campaigns. And not just for cereal—for everything! I mean, Chomsky is one of the world's most famous academics, right? Surely he can finagle himself a totally sweet licensing deal!

Also: Man, have I ever been neglecting my ling blog.