Tag questions and gender: Project

Introduction

The following is an outline of a number of questions that could be asked while putting together an investigation into tag questions.

Read the extract by Robin Lakoff in Language and gender: an advanced resource book (J.Sunderland, Routledge, 2006) which is reproduced in the handout at the bottom of this page.

In this extract, Lakoff makes a number of points about how men and women use tag questions. There seems little dispute about the grammatical form of a tag question - an interrogative tagged onto the end of a declarative – but what is more open to interpretation is the meaning and use of this feature.

Project

To begin with, we can use data from a corpus, i.e. a large database of authetic spoken and written text materials. One such corpus is the British Component of the International Corpus of English (ICE-GB; for more information, click here). We can use software to find tag questions in ICE-GB and then ask the question 'Do women use more tag questions than men'?

Step 1: A search for TAGQ (tag question) in the British Component of the International Corpus of English yields:

Total hits: 757

Spoken: 722

Written: 35

• Why would there be tag questions in written texts?
• What reasons might there be for discarding the written results?

Tag questions in written texts are likely to occur in passages of dialogue in fiction. Fictional dialogue can often represent stereotypes of male and female speech – even naturalistic dialogue can be quite inaccurate.

Step 2: Discarding written examples and narrowing the search criteria down by looking for examples uttered by females and males, we get:

Spoken tag questions uttered by females = 296 (41%)

Spoken tag questions uttered by males = 426 (59%)

So, according to these numbers, men use more tag questions than women in ICE-GB.

• How can we find out if this is true? In other words, do we perhaps need more information to establish that men use more tag questions than women? What information would that be?
• What else do we need to know?

We would need to know how many utterances were spoken by men and women in the entire corpus. We’d need to know the relative proportion of tag questions to utterances where a tag question might have been used. We’d also need to think about subclasses of spoken data, for example if particular types of conversation encourage more tag question use (classroom talk, etc).

Step 3: Narrowing down the variables. Can we search for different examples uttered by men and women taking their level of education into account? If we do so, the results are as follows:

Number of tag questions uttered by females with secondary education = 85

Number of tag questions uttered by females with university education = 170

What about the missing 41? Perhaps they are mixed conversations.

Step 4: Zoom in to one extract to find out more detail.

An interesting example might be S1A-009:

2 speakers: 1 male, 1 female.

Male uses no tags; female uses 10.

What does this tell us about male/female conversation styles?

Can we generalise from this?

Examples of those 10 tag questions are here (click on the '+'-symbol to enlarge):

Extension project

Read text S1A-020 from ICE-GB, shown below (click on the '+'-symbol to enlarge).

Here we’ve got a 4 way conversation (3 male, 1 female).

The female uses 1 tag, and the three males combined use 7.

Break this down further and we have:

• speaker A: 1
• speaker B: 6
• speaker C: 0

Are there any differences between the tag questions used in this transcript?