Amirite?

Tag questions were the topic of a recent edition of "The word" in the Boston Globe, a column shared by Jan Freeman and Erin McKean, now of Wordnik. Writes McKean,

Some days it seems that the most common kind of understanding is misunderstanding: Every conversation — not to mention each e-mail, IM, or text message — is rife with opportunities for crossed circuits and hurt feelings. There’s no end of advice about how to avoid miscommunication: Keep things simple. Take your time. Be aware of cultural differences. But missing from all these communication-helper lists is a little linguistic tic that most people use every day: the tag question.

You know what tag questions are, don’t you? Tag questions are those little questioning upticks, usually found at the end of a sentence — like that don’t you? — that grease the conversational wheels.

Tag questions often feature in English language classes too. The Azar series treats them in their Fundamentals (black) book. Other topics addressed there are such things as: using how, prepositions of time, and expressing ability with can and could. So clearly this is considered a fairly basic topic.

The frequency of negative tags such as isn't it? in the google books corpus shows some interesting patterns. (Note that the scales differ from graph to graph.) The most common are those with be:

Followed by do:

Not surprisingly, present tense are more common than past tense:

The most common modal is won't:

And a few laggards:

• In almost all cases, there's one pronoun that is far more common than than the others, usually it and you: isn't it (aren't you), don't you, didn't it, won't it, wasn't it, haven't you, can't you. Perhaps these should simply be taught as units without any internal or systematic analysis, at least at the beginning.
• Tag questions seem to have peaked in popularity around 1940. It would be interesting to see how things have changed since 2000. (The google data isn't reliable past then.)
• Altogether, negative tags seem to occur perhaps 700 times per million words in published books. The Longman Grammar of Spoken and Written English shows that they are about four times as common in conversation, and positive tags (e.g., is it?) are about one fourth as common as negative tags (p. 212).
• While that's common, it's not all that common, nothing like can/could which, together occur about 240,000 times per million words, or how, which occurs about 50,000 times per million words (not all occurrences of these are the basic usages dealt with in the Black Azar.)