The decline of ain't

The Lingua Franca blog over at the Chronicle of Higher Education has a nice post from Allan Metcalf, of OK fame, on "The decline and fall of a 4-letter word". He makes a nice point:

“Ain’t” had the advantage over other four-letter words in being pure defiance or pure sincerity, not carrying any baggage of obscenity, blasphemy, or indeed any other content at all, since it is just a form of “be” or “have.”

One basic point is that it has lost its power to shock, with the key date being 1961, with the publication of Webster’s Third New International Dictionary, Unabridged.

I didn't first read "decline and fall" as being about taboo status but rather about frequency of usage. While spoken usage would be more interesting in many words, Google Ngram Viewer provides a pretty interesting data on this. Below are the images for English, American English and British English. (And, do I need to write it anymore?, click to embiggen.*)

Pretty surprising decline and fall, actually, and maybe surprisingly early, especially in American English — decades before Webster's Third, in fact. (And what the heck is going on with the late resurgence?)

But it's useful to have some kind of crude basis of comparison here, so I checked ain't versus isn't. That provides another little surprise:

The two words track very close until the beginning of the 20th century, diverge pretty sharply and then show roughly parallel paths with isn't far more common.

Anybody have a smart story for what's going on here?

* Oh yes, of course, because it's a chance to use embiggen.