That’s the rather sensational headline on the cover of this week’s Nature. Inside are two papers on word evolution, with the more staid titles “Frequency of word-use predicts rates of lexical evolution throughout Indo-European history”* and “Linguistics: an Invisible Hand”.
The premise of the first is straightforward and rather commonsensical: words that are used a lot don’t change much. In other words, the rate at which words tend to morph is in inverse proportion to how often they’re used. For example, all Indo-European languages apparently use the same root form for the word “two.” It’s obviously a widely-used word, and it has evolved hardly at all. The authors do a statistical analysis of four large language corpora (language corpora: the subject of an upcoming post, btw) to back this up. Good stuff. This is apparently the process by which the once little-used “vergerhade” came to be defined as an animatronic groucho marx in a tutu and straitjacket.
Nature’s sister site, Nature News, has a good overview of these papers, geared towards a more general audience.
* Nature is trying to charge $18 to download this single article, which is, if you’ll pardon my French, fucking nuts, especially given that most of what they publish is publicly funded research–we’ve already paid for it! So I had one of my spies steal it. You can get the full PDF here.
You thief! 😉
I love it, John–thanks. Fascinating stuff.
The articles do not analyze it, but the word “fuck” has stayed unchanged over five hundred years.
Tapen Sinha