A long time ago I read that someone applied a statistical test for language to human DNA. They found that DNA which codes for proteins failed the test, but "junk" DNA passed.
I'm sure there's a mundane explanation, but it was fun speculating about what it might say.