The stuff that non-linguists ask linguists when finding out what we do covers a broad span. If you're a historical linguist, like a bunch of linguists here in Madison, one tidbit you get regularly is something about the Voynich manuscript, sample page from Wikipedia Commons shown on the right (click for larger image, of course). A basic introduction can be found here, but it has shown up in Scientific American and such and the Log has posted on it occasionally.
The story is pretty simple: It's a pretty big book and pretty old (15th c. C.E.?) in an unknown script that nobody has been able to decipher. Tons of effort has gone into it to no avail. When people ask me about it, I tell them basically that — we just don't know if it's a hoax, an invented alphabet to write an Asian language or a rendering of some European language. We just don't know. This is terribly unsatisfying to people who've read about it, but that's just how it is. Your average working historical linguist might find it a mystery wrapped into an enigma. That doesn't mean they'll stop working on projects that have a good likelihood of success to toy with this.
Anyway, today's kcxd presents a new theory on the manuscript:
For once, by the way, I won't urge you to check out the rollover.