Monday 5 January 2015

The truth about British tag questions

why do we use tag questions?

I bet you`ve heard about tag questions (TQs) before, haven`t you? But, apart from knowing what kind of question they are, have you actually thought about what these TQs actually do? You might think that the primary goal of the person who utters a question is to ask for some knowledge that they previously didn’t have – and you would be quite right. However, is there something more to TQs than just questions?

The thought that there might be more to questions than just a request for information has previously crossed a number of minds, some of them belonging to professional linguists. For instance, Ditte Kimps, Kristin Davidse and Bert Cornillie set out to create a typology of the basic functions of English TQs in speech. They trawled through two corpora of spoken British English – because TQs typically occur in spoken language. The corpora were the London-Lund Corpus of Spoken English and the Bergen Corpus of London Teenage English.

Overall, the researchers came up with five functions that a TQ can perform. Apart from being a question i.e. a linguistic strategy for seeking information, TQs can function as:

·       a statement
·       a statement-question blend
·       a response
·       a command or an offer (e.g. a TQ can negotiate a desired action)

The criteria used in the categorization procedures were the intonation on the tag (whether the last part of the TQ is pronounced with rising or falling intonation), the polarity of the tag (for example, in he`s fine, is he? there is constant polarity, as opposed to in he`s fine, isn`t he? where there is regular polarity, meaning that the tag is negative but the preceding clause is not). The authors also considered whether the TQ was the last item preceding a response (so-called turn-final TQ) or whether the same speaker continued his or her turn in the discussion after asking a TQ (so-called turn-medial TQ).

Surprisingly enough, only 20% of all occurrences were categorized as questions. The examples include (the TQs are underlined):

B:  he says the contraction makes it quite normal, but the other doesn’t
A:  you’re sure of that, are you?

Here you see all the typical ‘textbook’ features of questions: rising intonation, expectation of a response, lack of knowledge on the part of A. Interestingly, constant polarity is more frequent in questions than in any other function types of TQ.

What about the other function types? 21% of the data consisted of TQs that were statements not usually expecting a response. As a rule, these are turn-medial and uttered with falling intonation on the tag. For example:

A: er heˈs not gonna give it to you twice, though, is he? cos, I donˈt reckon he would give it to you twice 
B:  he donˈt, he donˈt give it to you twice

The most common function (44%) of the TQs was a statement-question blend; that is, the TQ states a specific proposition, but the speaker expects a response. 88% of this type of TQ did get a response, usually confirming the proposition. The speakers usually positioned themselves as more knowledgeable, hence the statement part. A typical example of a statement-question blend is:

B:  and he makes this hideous giggle, doesn’t he
A:  yes, he does

Finally, a tiny (3%), but peculiar part of the data contains TQs initiating an exchange where the speaker is demanding or offering a desired action, as in this example:

you know, Pat, don’t say that, will you?

These are usually pronounced with rising intonation on the tag, which indicates uncertainty and thus softens the request that the addressee complies with the command.

So although these forms are known as tag questions, question’ doesn’t seem the right word to use, does it?

Kimps, Ditte; Davidse, Kristin; and Cornillie Bert (2014) A speech function analysis of tag questions in British English spontaneous dialogue. Journal of Pragmatics 66: 64-85.

This summary was written by Marina Myntsykovska