Tag Questions

Tag questions are concerned more with establishing interpersonal relations and suggesting whether some kind of response is required (or to close off any response), rather than requesting any specific information.

Andersen (Andersen, Gisle. “Are tag questions questions? Evidence from spoken data.” 19th ICAME Conference. Belfast. 1998.) identified eight main functions of tag question which differ mainly according to the intonation, pitch and stress patterns:

1. Confirming info (=I think so) – High falling tone; response required:

You ordered fish, didn’t you?       

2. Checking info (=Is it so?) – High rising; response required:

You like dogs, don’t you?  

3. Chatting (=Let’s chat) – Mid slight rising; response required:

Nice day, isn’t it?

4. Challenging (=You’re wrong!) – Low falling-rising; response required:

I told you so, didn’t I?   

5. Closing (=I don’t want to talk) – Low rising-falling; no response required:

Well, I forgot, didn’t I

6. Antagonizing (=I don’t like you!) – Low falling; no response required:

I’m not stupid, am I.  

7. Aggravating (=Do it!) – Low rising; no response required:

Just stop it, will you

8. Softening (=Let’s not argue) – Mid slight rising; response required:

Please don’t forget, will you

Personally though, I would also add two more in keeping with the full range of intonation patterns. So as well as 4. and 5. above, a high rising-falling and high falling-rising would also be possible:

10. Questioning (=I’m surprised) – High falling-rising; optional response:

That isn’t a cat, is it?

9. Acknowledging (=I’m grateful) – High rising-falling; optional response:

Oh, that’s wonderful, isn’t it?

The difference in function is also often recognisable through extra-linguistic factors, depending more on tone of voice, body language and the situation or context of the conversation. The same sentence may thus have different interpersonal functions depending on these factors.

If a response is required, there are a number of possible options:

1. Minimal responses: yeah, yes, mm, right etc

A: you have to get back by train won’t you, obviously.

B: yeah

2. Repetition of entire proposition

A: You’re not open on Saturday are you?

B: We’re closed Saturday.

 3. Elliptical repetition of proposition

A: She wouldn’t do that would she?

B: She would.

 4. Repetition of propositional element

A: You’re almost fluent in English aren’t you?

B: Almost.

 5. (Near-) synonymous expression

A: But that’s really quite quite bad, isn’t it?

B: Dreadful, poor parents.

 6. Implicature

A: Her father’s got money hasn’t he?

B: They’ve all got money.

(Implicature: yes, confirmation)

A: Never phone her do you?

B: Can’t be bothered.

(Implicature: no, confirmation)

A: You missed a lot did you?

B: Only the first lesson, which is …

(Implicature: no, rejection)

 7. Responses expressing reduced commitment/uncertainty

A: But you never used to hang around with her though, did you?

B: Well, sort of.

(Adapted from: Andersen, G. 1998. ‘Are tag questions questions? Evidence from spoken data’)


It is also possible to replace the question tag with an invariant one such as ‘OK?’, ‘right?’, ‘yeah?’, ‘correct?’, or ‘eh?’ among others. These can be dialectical or regional. There are some differences between them.

  • ‘Right’ often functions to check information and ask ‘Is this correct?’:

AD9 2214 ‘You’re the kid with Leila, right?’

ALJ 555 The compartment under the passenger seat in the front, right?

BN1 2508 So that last option is favourite, right?

G0N 2730 She’s your niece, right?’

G1W 2061 ‘It might just be possible though, right?’

G5E 28 I paid forty pound ninety five, right?

GV6 1922 ‘And they were married in Ireland, at Rathdrum in County Wicklow, right?’

H5K 80 You took it out the other night, right?

  • ‘OK’ often functions to close debate. It is often used with imperatives:

A0F 1101 ‘If this bounces, you’re out on your ear, OK?’

A0F 2901 Listen, you sit down, I’ll get a couple of coffees and we’ll have a chat, OK?’

C8E 3057 We all love you here, OK?’

C8T 346 Look, if I knew who he was I might know where he was, OK?’

CCW 70 Count me out, OK?

F9X 2691 I said I’ll do it, OK?

F9X 4049 ‘Look, I know what I’m doing, OK?’

FP7 342 ‘Let’s go, OK?

  • ‘Yeah?’ functions mainly to facilitate conversation, overtly indicating a response is required:

A0F 1560 ‘You were reckoning on trying your luck abroad, yeah?’

C8E 2182 ‘You liked it, yeah?’

ECT 2319 Maybe I’ll have one more shot at it, yeah?

FM7 728 Well done, that’s a good word, when we use our little circle of words you can use that word, yeah?

FM8 20 He was white slim and quite tall, yeah?

KB7 11755 Alright, yeah?

KBW 9719 You’re gonna have beans instead of tomatoes, yeah?

KCP 6739 Pat’s gone to theatre has she, yeah?

KPW 827 You’ve read this book called Roll Of Thunder, yeah?

(All examples from British National Corpus)



Polarity is, the “choice between positive and negative” (H & M, p.116). The concept of polarity in general, and specifically the negative, doesn’t seem to get much attention or specific textbook treatment (except perhaps in old audio-lingual drills) but, as Halliday says, “choosing positive is just as substantive and meaningful as choosing negative” (H & M, p.143). Polarity is one way that allows speech functions to be arguable in terms of exchange by setting up an opposition between ‘yes’ and ‘no’: “either ‘is’ or ‘isn’t’ (proposition), either ‘do’ or ‘don’t’ (proposal)” (p.116).

The polarity may be attached to either the Finite (temporal or modal) or the proposition*. Thus we may have:

  • I don’t have to go. (finite)
  • I have to not go. (proposition)

Which may also be combined:

  • I don’t have to not go.

The difference between them may be shown with a question tag, where the unmarked form of the tag reverses the polarity. As such, if the polarity is attached to the Finite we get:

  • She couldn’t have known, could she?

But if the polarity is attached to the proposition it becomes a distinct modal Adjunct and the tag is reversed:

  • She could have not known, couldn’t she?

As a modal Adjunct, the polarity then has the ability to form a mood element itself, often in non-finite clauses for example:

  • Not being funny but…
  • Never having been myself…

It is also seen with a corresponding change in intonation:


However, polarity is most commonly associated with the Finite which “reflects the systemic association of polarity with mood” (H & M, p.143). This is shown in English with contracted forms where the “negative marker may be reduced to the point where positive and negative are more or less equivalent in weight” (H & M, p.143) and it is only the Finite element that allows this to happen. From this we can also see that reduced forms are more than just ‘sounding natural’ but represent a meaningful distinction.

*The choice of one or the other does seem to be affected by context: I’ve noticed that English spoken in the North of England, such as Liverpool, tends to focus more on the proposition in negative question forms (Did you not want to come?) while that spoken in the South focuses more on the Finite (Didn’t you want to come?). As EFL textbooks are largely based on upper middle-class language of the South this distinction tends to get overlooked.

Process + Circumstance v Process

Here are some examples that came up in a business class that might cause problems for students but can be clearly explained:

1. I passed on the idea.

  • I (Participant) passed (Process) on the idea (Circumstance) because I didn’t think it would work.
  • I (Participant) passed on (Process) the idea (Participant) to management because I thought it was good.

2. We went over the bridge.

  • We (Participant) went (Process) over the bridge (Circumstance) driving from North Sydney.
  • We (Participant) went over (Process) the bridge (Participant) that we were designing.

3. I looked up the picture.

  • I (Participant) looked (Process) up the picture (Circumstance) to see how the frame was hung.
  • I (Participant) looked up (Process) the picture (Participant) to see who the artist was.

It might also be helpful to point out the phonological changes that can reflect the patterning of the ideational and textual meanings:

  • I passed / on the idea
  • I passed on the idea/



Interpersonal modal Adjunct v Experiential Circumstance

I always do that. (Adjunct)
I do that all the time. (Circumstance)

I usually do that. (Adjunct)
I do that almost everyday. (Circumstance)

I often do that. (Adjunct)
I do that at times. (Circumstance)

I sometimes do that. (Adjunct)
I do that now and then. (Circumstance)

I rarely do that. (Adjunct)
I don’t do that much. (Circumstance)

I never do that. (Adjunct)
I don’t do that at all. (Circumstance)


Context and Register

An important distinction can be made between the context of situation and register of a text. They often seem to be conflated but Halliday does distinguish between the two.

Here is a headline from the satirical magazine The Onion:

Coarse Sponge Excited To Join The Smith Family Dishwashing Team

The humour here works only because we can make a distinction between the context of situation and the language that typically accompanies it – by a mixing of the register of a company announcement (excited to join, the ~ team) with an unexpected field choice (sponge).

Context of situation, as the name suggests, is “the immediate environment in which a text is actually functioning ” (H & H, p.46), or the social and physical environment where the action takes place in terms of its field, tenor and mode. Importantly, this may or may not involve language. In fact, there are certain contexts in which the use of language is actively discouraged or even proscribed, such as Berstein’s (1971) concept of ‘resticted codes’ or the cultural value of silence in Japan (King, 2011). If you’ve ever been on a visit to the Sistine Chapel in Rome, with the attendants continually calling for silence, you’ll see one good example.

Register, on the other hand, is entirely semantic and is the “configuration of meanings that are typically associated with a particular situational configuration [and] include(s) the expressions, the lexico-grammatical and phonological features, that typically accompany or REALISE these meanings” (H & H, p.38-39). Register is a probabilistic tendency for certain items from the semantic and lexico-grammatical (and phonological/graphical) systems of ideational, interpersonal and textual meanings to co-occur in certain contexts. It is the semantic interface between the external context and the internal language of a particular text, and may range from restricted registers (the language of aviation, for example) to relatively open ones (casual conversation) although no registers are completely open as even casual conversation exists within certain boundaries and conventions.


Tenor of discourse

The second part of the context of situation is the tenor of discourse. Tenor refers to:

who is taking part, to the nature of the participants, their statuses and roles: what kinds of role relationship obtain among the participants, including permanent and temporary relationships of one kind or another, both the type of speech role that they are taking on in the dialogue and the whole cluster of of socially significant relationships in which they are involved?

(Halliday & Hasan, 1985, p.12)

There are three basic factors within tenor:

  1. agentive role, or the institutional (or not) roles of the participants, such as doctor/patient, teacher/student, etc.;
  2. social role, or the power relationship between them which may be hierarchic or nonhierarchic and includes expert/novice and also conferred social status and gender, etc.;
  3. social distance, or the amount or nature of contact the participants may have, which ranges from minimal (close friends) to maximal (formal settings).

Rather than an either/or situation, these tenor factors exist on a cline, as may be represented here:


It is also possible for these tenor relationships to change over time. A regular patient, for example, may have less social distance than one on a first-time visit. They may also be affected by field choices: an office-worker talking to their manager about football may use a different register than when requesting leave. This may also be affected by the context of culture with each factor given more or less value. In a Japanese work-place context (and in general) agentive and social roles have comparatively more prominence: even after years of close working contact (and even after retirement) many Japanese will continue to use formal work-place terms of address that encode these roles.


Field of discourse

Here is a recipe for Lemon Risotto from http://www.epicurious.com:

We know this is a recipe, and not say a research report, partly through the field of discourse.

Within the context of situation, field refers to:

what is happening, to the nature of the social action that is taking place: what is it that the participants are engaged in, in which language features as some essential component?

(Halliday & Hasan, 1985, p.12)

We can divide field into three areas:

1. experiential domain, or what the text is about. In the case of a recipe, it is about food and food preparation.

2. goal orientation, or what the text is for in terms of both short-term and long-term goals. The short-term goal is obviously to make the recipe but the long-term goal is, I think, a little more complicated. Why are we making this dish? For example, as some research shows, there are cultural differences between Japanese and English that affect the long-term goals of food preparation. In Japan, cooking is, to a large degree, a serious matter and the goal is to reproduce the recipe exactly as shown, as opposed to the life-style/aspirational aspect of more Western cooking. I think also there are differences in the idea of the home as a private space, as in Japan, or a more public space for entertaining, as in English. This in turn affects the last area:

3. social activity, or what the text is doing. The activity of cooking in Japan is essentially a private or family activity. There is no real equivalent of the English-speaking or European idea of entertaining in your home or the dinner party. You can see this difference most clearly when looking at portion sizes. Googling ‘lemon risotto recipe’ or ‘レモン リゾット レシピ’ brings up recipes that in English, such as the one above, nearly always serve 4-6, or more, whereas the Japanese ones are for 1-2 people.

These three areas together make up the field of discourse, which is expressed through the experiential metafunction: there are specialised lexical items, such as broth, arborio and peel as well as material Processes, like simmer and stir. To a large degree, I think we also know that the above is a recipe (and the field of discourse is recognised by) external features such as lay-out, pictures, and the fact it says ‘recipe’.

Although I think there is a lot more to it and Hasan, in particular, has gone much further into the system of field in more detail, for the purposes of EFL I’ve found discussing just these three to be useful.


Ideal L2 Self

I was reading an article recently by Yashima (2009) about the concept of an ‘ideal L2 self’. The article described the difficulties that many Japanese students in particular have in imagining an L2 Self that is separate from their L1 identities and that “Japanese students’ ideal selves are typically formed without an L2 component” (p148). Often Japanese learners want the language tools to be able to interact with members of the target community without necessarily becoming a part of it. As Yashima says, Japanese “are not particularly interested in identifying with [native speakers]” (p145). They often want merely the experience of speaking the target language for a variety of reasons but are generally unwilling, or even unable, to let go of their L1 identities and imagine an L2 self. In a globalised world, even the target language community can be hard for Japanese to identify (p.145). Students may want to learn an idealized ‘Queen’s English’ that does not exist or some vague notions of ‘for travel’.

I’m interested though in how this affects classroom language and interaction. I’ve noticed that quite a few of my students, far from L1 interference, are in fact speaking the L1 – with L2 lexical items. Here is an example. Asked to describe which the ‘classroom’ was from a set of pictures, a student produced:

Classroom is right side toilet under picture.

At first glance this seems a valid attempt by a beginner student at producing the target language but actually this student has been studying English about once a week for two years. If we look more closely, compared with the Japanese we can see that it is in fact merely a transliteration of the L2 with ‘is’ functioning as topic marker:

教室は 右側、 トイレ のした. 絵
kyoushitsu-wa migigawa toire no-shita e
Classroom-topic rightside toilet under picture

The grammar in terms of its ideational and textual features is entirely L1. It also features a distinct lack of any interpersonal features, which is entirely the point – it is a way of maintaining maximum distance between their ideal self and the L2 self. It could even be said that this student has not, in any real meaningful sense, learnt any English at all in two years but instead enough English lexical items which then can be mapped onto the L1 lexicogrammar sufficient to satisfy classroom communication. In fact, nearly all of the language produced by this student revolves around ‘is’. Another example:

Teacher: What are you doing this weekend?
Student: Weekend is play tennis

Again, the ‘is’ here is not functioning as a relational Process but as a topic marker. Similarly, ‘play tennis’ is not a material Process + Participant but is a single lexical item.

The classroom then, in the absence of any sense of an outside ideal L2 self or sense of a participation in a wider target community, becomes a language community of two – teacher and student – and, being a commercial relationship, the Vygotskian idea of Zone of Proximal Development becomes inverted and it is the teacher who must come down to the student’s level for communication to take place rather than the other way around. As nothing changes in this context from week to week, the student makes no real progress. While this particular student here might be an extreme case it is by no means unusual in a Japanese commercial EFL context. The question is how, or if, this can be overcome and help students create an L2 identity.

Yashima, T, (2009) “International Posture and the Ideal L2 Self in Japanese EFL Context” in Dornyei, E.Z & Ushioda, E (Eds.) Motivation, Language Identity, and the L2 Self, Multilingual Matters


Matching Pairs exercises

A student the other day had difficulty with a textbook matching exercise. The prompt was:

Are you watching the TV?

The required option was:

You can turn it off.

I think the difficulty here was due to the incongruence of the lexico-grammar (present-in present material Process) with the Interpersonal semantics (a request to turn off the TV rather than a request for information). The exercise was ostensibly practising the present progressive but the student was confused as the expected response (‘Yes, I am watching the TV) did not appear.

I must admit I’m not a big fan of these exercises. I can see why materials writers like them, they seem fairly straightforward, but they tend to assume that all forms of that tense are the same (‘Let’s practise the Present Progressive!’) and don’t take account of grammatical metaphor like the one above. Also there is an implicit assumption that practicing these will lead to native-like acquisition of the form but, personally, I feel that the opposite is the case – that native-speaker competence allows the interpretation of the form.

Recently then I’ve been taking a different approach by giving the students the completed matching pair dialogues and discussing which are congruent and which are not. For those inconguent, we try to guess what the grammatical metaphor is (this is related to Speech Act theory in traditional sociolinguistics).


Directions Game

I added a game to the Resources page that practises giving directions. I use it especially with Junior/Senior High School kids. Usually the language of commands is introducted, understandable enough, with giving directions around town (turn left at High Street). The problem, however, is that Japan doesn’t have any real street names. There is a conceptual problem as the activity doesn’t match their own schema of ‘giving directions’. I’ve found a way around this is to create a new schema of ‘game’ with a grid layout:

Student A: Where is the strawberry?

Student B: Go straight. Turn left at C. Turn right at A. Turn left at V. Go straight. It’s number 24.

The students enjoyed the ‘game’ element (I put them into teams and made it a race) and it seemed to make more sense to them.