Introduction
In several (colloquial) varieties of English around the world, the tag question is reduced to a single generalized invariant tag ‘isn't it?’. This is used irrespective of the general structure and semantic content of the matrix or main, superordinate clause that embeds it. Linguists started debating seriously on the actual use of tag questions in English since the early 1960s and 70s. The debates then were particularly focused on the use of tags on (elliptical) imperatives, with many concluding that the auxiliary in the imperative main clause is an elliptical will that serves as the source of input for the tag that is appended (Katz & Postal, Reference Katz and Postal1964; Postal, Reference Postal and Dinneen1966). Arbini (Reference Arbini1969) discussed the putative parallel between tag imperatives and tag questions, and other early discussions on tags include Bolinger (Reference Bolinger1967), Huddleston (Reference Huddleston1970), and Klima (Reference Klima, Fodor and Katz1964). The present discussion looks to a different direction. As already implied, it focuses on the emergence of isn't it? as a generalized invariant tag.
Background
In Standard varieties of English, a question tag functions as a subordinate (or embedded) clause that is standardly adjoined or appended to another clause (a matrix or main superordinate one) that may be positive or negative in polarity. As Kolln & Funk (Reference Kolln and Funk1998), for example, point out, the use of the tag is a common way to turn a statement (i.e., the matrix, superordinate clause) into a question (usually to get a confirmation). Thus, a question tag is a elliptical yes/no question – reduced to just an auxiliary verb and a subject pronoun – used as a common device for seeking confirmation of what one says (Huddleston & Pullum, Reference Huddleston and Pullum2005). It is used to obtain a positive or negative confirmation of the assertion of the proposition made in a preceding clause based on the expectation of the questioner.
A question tag consists of an operator, followed by a pronoun (Quirk et al., Reference Quirk, Greenbaum, Leech and Svartvik1973; Biber et al., Reference Biber, Johansson, Leech, Conrad and Finegan1999), the term ‘operator’ referring to the auxiliary verb (in any form) with or without a (contracted) negative particle that may mark a contrast in polarity between the embedded tag itself and the superordinate main clause. It is called an operator because it literally undergoes an inversion with the subject pronoun in which it syntactically repositions itself from its canonical position after the pronoun to the front of the latter. The subject pronoun in question is a personal pronoun, and following Huddleston (Reference Huddleston1988), personal pronouns are distinguished from many other pronouns by their ability to occur as the subject in a tag question. It is obligatorily co-referential with the subject of the superordinate clause, while the tense and choice of tense of the operator depend on the tense in the main clause (be it on the lexical or grammatical verb). Thus, as Huddleston (Reference Huddleston1988) observes, the form of the tag is largely predictable from that of the main clause. If no auxiliary is used in the main clause, do is introduced (Peters, Reference Peters2004).
It is the negative polarity of tags that usually expresses a need for confirmation of the statement expressed in the main clause (Huddleston & Pullum, Reference Huddleston and Pullum2005). This is so when the operator takes a low tone (Quirk et al., Reference Quirk, Greenbaum, Leech and Svartvik1973), in which case, the questioner expresses no doubt. Thus, another function of the tag is that it allows the possibility of another view (Barry, Reference Barry2003), when the operator takes a high tone, in which case, the questioner expresses a doubt.
Operators, polarity, pronouns, and position of tag
Consider the paradigm in (1) that follows – following Downing & Locke (Reference Downing and Locke2006: 187), a question tag may be appended to a declarative clause (1a), an exclamative clause (1b), or to an imperative clause (1c), of which the declarative is the most common.
1
a. The boat hasn't arrived, has it?
b. How quiet it was in the hall, wasn't it?
c. Be quiet for a moment, will you?
As indicated above, tag questions may occur with a wide variety of auxiliary verbs in different morpho-polarized forms that show person, number, and tense as in (2).
2
a. The children go to school, don't they?
b. John eats fish, doesn't he?
c. You are a man, aren't you?
d. Paul is intelligent, isn't he?
e. It doesn't rain everyday, does it?
f. The children don't play games, do they?
g. We are not taking out the books, are we?
h. The children had to be at school yesterday, didn't they?
i. John has eaten fish, hasn't he?
j. I have completed the assignment, haven't I?
k. Peter was not afraid of us, was he?
l. We were the true children of God, weren't we?
Note that the contracted negative first person singular form, supposedly ain't, is not used in standard speech purely for reasons of linguistic etiquette, having been stigmatized sometime during the nineteenth century for public speech, and reserved for uneducated and ignorant speakers (Kolln & Funk, Reference Kolln and Funk1998). The word ain't is a contraction of am not following from the same internal rule that produces other contractions like aren't and isn't, and written texts from the seventeenth and eighteenth centuries testify to it as having been part of standard varieties of English.
Tag questions may also occur with a wide variety of modal verbs, in different morpho-polarized forms as shown in (3).
3
a. We can now sing and rejoice, can't we?
b. We shall go home in the coming days, shan't we?
c. Man won't live forever, will he?
d. You ought to be ashamed of yourself, oughtn't you?
e. John could not be here last week, could he?
f. Mary should be defending the course, shouldn't she?
g. You will give me the files, won't you?
There exist words including seldom, rarely, scarcely, hardly, barely, little, few, and only, that have negative meanings but do not appear negative (Quirk et al., Reference Quirk, Greenbaum, Leech and Svartvik1973: 186). In this regard, Huddleston & Pullum (Reference Huddleston and Pullum2005: 154) term them approximate negators. In other words, as they specifically point out, ‘few of them comes close in meaning to none of them: none indicates absolutely zero, while few puts the number within a small part of the scale down at the end close to zero.’ Positive tag questions are therefore used when such words appear in the main clause. The examples are as follows:
4
a. We seldom go to the fish market, do we(/*don't we)?
b. The children hardly ever visit us, do they(/*don't they)?
c. The boys rarely spoke to the girls, did they(/*didn't they)?
d. Few books ever make it to their buyers, do they(/*don't they)?
Note the contrast with positive a few and a little, however, as in A few books have made it to their buyers, haven't they (/*have they)?.
But there are other words that are also negative in meaning and duly appear so, but have a different ramification from those exemplified in (4) when used in the main clause before a question tag. Examples are deny, without, dishonest, unaware, and against which are loosely paraphrased as ‘not accept’, ‘not with’, ‘not honest’, ‘not aware’, and ‘not for’, respectively. Thus, they take negative tags as in the examples in (5), although the use of the positive counterparts of the same words triggers the same negative question tag as in (6).
5
a. Obama denied that he had played a role in the war, didn't he (/*did he)?
b. He emerged without a bullet proof vest to address us, didn't he (/*did he)?
c. The senators were dishonest, weren't they (/*were they)?
6
a. Bush accepted that he had played a role in the war, didn't he (/*did he)?
b. He emerged with a bullet proof vest to address us, didn't he (/*did he)?
c. The senators were honest, weren't they (/*were they)?
That is, while the words exemplified in (4) go with positive question tags, those exemplified in (5) rather go with negative question tags, though both sets of words have negative meanings. Observe however that the asymmetry between (4) and (5) follows from the fact that the words that carry the negative meanings in (5) are predicates, whereas those in (4) are rather modifiers – compare The books are few, aren't they? where few is a predicate, to (4d) where few is rather a modifier. One thing that is suggested by this observation is that, in a given sense, the negative particle as used in examples like (1a), (2e, f, g), and (3c, e) above is to be viewed as a modifier – a negative modifier that modifies the predicate in the main clause that embeds the tag – otherwise the tag is negative. (Compare the figurative The world is nothing, isn't it/*is it? where the negative word nothing is a predicate. Also see (7c) below.)
As said earlier, the form of the pronoun in the tag is determined by the subject of the main clause to which reference is made. However, it occurs as a default they when the subject of the main clause is an indefinite human singular pronoun (Downing & Locke, Reference Downing and Locke2006) as in the following examples:
7
a. Everyone seems to enjoy the party, don't they?
b. Somebody should be at home, shouldn't they?
c. No one will agree to that, will they?
d. Anybody could figure out the problem, couldn't they?
e. Anyone who speaks out will be blackmailed, won't they?
As the facts show, question tags are usually placed at the end, after the main clause that embeds them, but they may also occur in the middle (Biber et al., Reference Biber, Johansson, Leech, Conrad and Finegan1999: 208) as shown in (8).
8
a. It's easy, isn't it, to get into the habit?
b. The girls lied, didn't they, when they were asked about their relationship with Philips?
Invariant tags
One thing that is shown to be very important about tag questions is polarity. The tags that we have seen so far are all opposite to the main clause in polarity, that is, if the main clause is negative (i.e., the predicate is negatively modified), the tag is positive, and vice versa. Some early grammarians rule out non-polarized tags in tag questions as ungrammatical (see Hintikka, Reference Hintikka1982 for discussion). However, as Downing & Locke (Reference Downing and Locke2006: 187) point out, a second type of tag is actually a non-polarized one, that is, if the main clause is positive, the tag is also positive, and vice versa. They point out that this type occurs mostly in the positive form, and that the whole sentence is often preceded by a discourse marker like Oh and So, to indicate that the speaker is inferring or drawing a conclusion from context; in this case, the effect is often emotive as in Oh, you are the new secretary then, are you?, with either an agreeable surprise or perjury depending on the implication. In fact, as Huddleston & Pullum (Reference Huddleston and Pullum2005: 150) point out,
… reversed polarity tags have to be distinguished from those with constant polarity. These [i.e., constant polarity tags] don't ask for confirmation [and indeed, they do not allow the possibility of other views], but suggest an attitude such as surprise, disbelief, disapproval or the like: an author might say So they've read my book, have they? Amazing!'
They also point out that
For many speakers constant polarity tags aren't used with negative clauses: So they haven't read my book, haven't they? will be rejected by many speakers of Standard English. For those speakers, if a negative tag is acceptable on a clause, the clause must be positive.
Besides, as Huddleston (Reference Huddleston1988) asserts, exclamative main clauses only take opposite polarity tags as in How quiet it was in the hall, wasn't it/*was it?, while imperative ones take a slightly wider range of tags as in Open the door, can you/can't you/could you/will you/won't you/would you?.
Maintaining the same polarity asides, observation is increasingly being made (as in Downing & Locke, Reference Downing and Locke2006: 189 for example) of a generalized invariant tag, namely, isn't it?, often simplified to innit, used in environments other than those involving the operator is and the third person inanimate pronoun it. In fact, in many non-standard varieties of English and in informal or popular speech situations, isn't it? is used as a generalized invariant tag irrespective of the structure, content, and polarity of the main clause to which it is appended. This is commonly the case, for example, in my native Cameroon, among a majority of Cameroonian English speakers, including even some very educated ones. Selected examples are listed in (9).
9
a. John doesn't eat fish, isn't it?
b. You are a man, isn't it?
c. The boat hasn't arrived, isn't it?
d. Paul is intelligent, isn't it?
e. We can now sing and rejoice, isn't it?
f. Everyone seems to enjoy the party, isn't it?
g. Anybody could figure out the problem, isn't it?
With the exception of exclamatives and imperatives as illustrated in (10) below, this invariant tag is used in a generalized fashion irrespective of the form of its main declarative clause.
10
a. How quiet it was in the hall, wasn't it/*isn't it?
b. Be quiet for a moment, will you/*isn't it?
In fact, cases like those in (10) with exclamative and imperative main clauses are rarely used in Cameroonian English, being only sometimes used by educated speakers.
Examples like those in (9) are indeed not restricted to Cameroon. They have been reported for other English-speaking communities around the world as in Trudgill & Hannah (Reference Trudgill and Hannah1994) for example. Thus, Trudgill & Hannah (p. 35) state that the generalized invariant tag isn't it? ‘... can be observed in the speech of even some educated WEng [i.e., Welsh English] speakers’, their examples cited as (11a/b) below, though not usually the case in the written form. They make similar observations for West African English (p. 125), their examples cited as (11c/d); Indian English (p. 133), their examples cited as (11e/f/g); and for Singaporean English (p. 136), where they state that ‘again as in other non-native varieties, is it?/isn't it? are used as invariant tags’, their example cited as (11h).
11
a. You're going now, isn't it?
b. They do a lot of work, isn't it?
c. We should leave now, is it?
d. She has gone home, is it?
e. You are going home soon, isn't it?
f. They said they will be here, isn't it?
g. We could finish this tomorrow, isn't it?
h. He is going to buy a car, isn't it?
Although these observations by Trudgill & Hannah are not recent, they are a reflection of an ongoing trend, judging from my observations in Cameroonian English. A few testimonies from individuals around the world make the claim well established. On the online forum, http://forum.wordreference.com, a member, Allifathima, a Tamil native and Tamil English speaker, posted the following question on 6 April 2011 – members are known by their usernames:
I have constructed a question and answer as below.
Q: He went to school. isn't it?
A: Yes. He went.
Will it satisfy the tag-question and answer?
Here are some of the ensuing reactions in response (on isn't it), on the same day:
Languagethinkerlover, native English speaker from the UK:
I've heard ‘isn't it’ used before to confirm that the question being asked is correct so if this is the reason you wrote ‘isn't it’ instead of, “He went to school, didn't he?” then it's fine.
Askalon, native English Speaker from the USA:
I've never heard that sort of usage before. Are you describing British English? Maybe it doesn't exist so much in American English. “You're from Essex, isn't it?” sounds very ungrammatical to me, colloquial or not.
George French, native English Speaker from the UK:
It is used in UK-EN spoken English, at least. And it is understood.
Natkretep from Singapore:
… I think isn't it is very well established in Indian English (and can be found in some other varieties like Malaysian English and Singaporean English). …
Loob, native English Speaker from the UK:
I disagree with George, here: I don't think it is used in standard British English. (I don't even think it's used in non-standard British English – though “innit” is.) That said, I agree with Nat (post 14): I have certainly heard invariant question tags from speakers of – standard – Indian and South African English.
Languagethinkerlover, native English speaker from the UK:
Askalon, I do hear its use more among the British. I'm sorry it doesn't sound grammatical to you but it's said nonetheless. ‘Innit’ is a shorter version of ‘isn't it’ and I have also heard that. I think it's important to know whether Allifathima is using ‘isn't it’ to mean ‘Right?’ (see my first response if confused) or if it's a case of incorrect grammar.
Alxmrphi, native English speaker from the UK:
Yeah innit/isn't it is used in colloquial London dialect. It's an invariant tag as Natkretep pointed out. It's not part of the standard language. It's usage is spreading more north I believe, it's often parodied on TV (Lauren Cooper ;p).
These reactions to Allifathima's question in that web page conversation in effect speak for themselves: the use of isn't it? as a generalized invariant tag is actually happening on a global scale. Also see the website http://english.stackexchange.com for the usage of isn't it as an invariant tag, with similar claims.
Take note that, on the face of it, the examples with the invariant tag do not appeal to principles of grammaticality, since the tag can be seen purely from the perspective of a growing linguistic decorum. Standard examples of invariant tags which are not derived from the structure of the main clause to which they are appended as illustrated in (12) include right? and okay? (Downing & Locke, Reference Downing and Locke2006).
12
a. It's getting warmer, isn't it/right?
b. Leave the cat alone, will you/okay?
c. How quiet it was in there, wasn't it/right?
d. Let's go home now, shall we/okay?
e. John will go with you, won't he/okay/right?
But even then, these invariant tags do not go in all situations.
13
a. It's getting warmer, isn't it/*okay?
b. Leave the cat alone, will you/*right?
c. How quiet it was in there, wasn't it/*okay?
d. Let's go home now, shall we/*right?
So one may well be left to wonder how the invariant isn't it? gets to be applicable in a wide variety of situations – in fact, after all declarative main clauses, whatever their subjects, operators, and polarity.
It is important to take note though, that such an invariant tag question is not strange in natural language. As Ferrar (Reference Ferrar1955: 263) points out, French exhibits the same invariant tag, and the French equivalent of it is used in exactly the same way, featuring as n'est-ce pas? after all declarative main clauses, whatever their subjects, operators, and polarity. Some examples are as follows:
14
a. Jean mange beaucoup, n'est-ce pas?
Jean eats much, doesn't he?
b. Il n'était pas le premier d'arriver, n'est-ce pas?
He wasn't the first to arrive, was he?
c. Le train suivre bien son chemin, n'est-ce pas?
The train is well on course, isn't it?
d. Mr Obama a fait de beaux discours, n'est-ce pas?
Mr. Obama made pretty speeches, didn't he?
e. Tu ne tardera pas, n'est-ce pas ?
You won't be long, will you?
In comparing the facts in (9/11) to those in (10), one can observe that in each of the examples in (9/11), there is an assertion of the truth value of the propositional content of the declarative main clause, which is not the case in (10). What this actually means is that, in (9/11), the speaker has the choice to conceive and entertain a declarative sentence like (15a) in mind, but only utter a portion of it as in (15b) by truncating the initial assertive clause it is true.
15
a. It is true that John doesn't eat fish.
b. It is true (that) John doesn't eat fish.
Sentences like (15a) are not frequently used, but they are heard once in a while, often in contexts where the speaker needs to emphasise the truth of the relevant proposition. The availability of such uses is confirmed by Languagethinkerlover in that web site conversation quoted above. Thus, the assertive portion it is true as used in (15a) is always available on declaratives, entertained in the speaker's mind, even though it may stay unuttered as in (15b) John doesn't eat fish. In this respect, the speaker of English has a free choice between two different tags to append to (15b), isn't it? to correspond to the truncated initial assertive clause It is true as in (16a) and does he? to correspond to the actually mentioned portion of the sentence as in (16b).
16
a. It is true (that) John doesn't eat fish, isn't it?
b. John doesn't eat fish, does he?
In other words, the paradigm in (16) shows that, in the case of (16a), the assertive introductory clause It is true is conceived with the rest of the sentence, but is truncated upon utterance so that it only stays in the mind from where it is recoverable as the source for the invariant tag isn't it?. Considering therefore that there is no assertion of the truth value of the propositional content of exclamatives and imperatives, it is clear why examples like those in (10) do not allow the invariant tag, as there is no source for the invariant tag.
Tag questions and ellipsis
Tag questions in a dialogue/discourse provide one condition under which ellipsis takes place, that is, repetition of what has earlier been said by the first speaker; here, ellipsis takes place both on the part of the first speaker and the second speaker. In a tag question, the speaker avoids repeating information, in this case, the predicate, already mentioned by the latter in the preceding main clause (see Huddleston, Reference Huddleston1988: 139 for discussions on elliptical tag imperatives). This is illustrated in (17), the elliptical portion shown in parentheses.
17
a. John ate the fish, didn't he (eat the fish)?
b. John didn't visit the doctor, did he (visit the doctor)?
Here, repetition of the predicate is avoided in order to focus on the new material, the positive or negative operator that requires confirmation of the previous statement from the listener. Quirk et al. (Reference Quirk, Greenbaum, Leech and Svartvik1973: 306) write:
The usual function of a question in discourse is to request the listener to respond verbally with information that the questioner seeks. The link between question and respond is often reinforced by ellipsis in the response, thereby avoiding repetition of material from the question and focusing attention on what is new.
In responding to tag questions like those in (17), therefore, the answerer equally elliptically avoids repeating the same predicate material that the questioner has provided, in order to focus attention on the confirmation as in (18).
18
a. Yes, he did (eat the fish).
b. No, he didn't (visit the doctor).
However, this is not the same ellipsis that takes place in the case of the generalized tag question we have been discussing in this paper. Note that, as illustrated in (19) below, with the generalized tag question, a chunk of materials has also been avoided. But, in what can be called ‘pseudo/shadow ellipsis’, part of the avoided material (underlined) has not hitherto been mentioned, following truncation, as indicated by the strikethrough.
19
a. It is true (that) John ate the fish, isn't it (true (that) John ate the fish)?
That is, (19) involves pseudoellipsis in that some of the elliptical material in the tag has not hitherto been mentioned in the discourse.
Conclusion
This paper has focused on the use of tag questions in English. Tag questions are appendages to declarative, imperative, and exclamative sentences that literally turn the latter into questions, and mainly seek to get confirmation for an assertion. They generally depend on the structure and semantic content of the main clause that embeds them for their own structure and semantic content, and therefore import heavily from such main clauses. In other words, a tag question includes an auxiliary and a pronoun determined by the main clause, and since they usually seek confirmation for what their speakers assert, they are often placed on the opposite polarity of the main clause. Tags are therefore expected to vary very widely in their form and content. In this paper, however, we have examined the emergence of the generalized invariant tag question isn't it? that is increasingly being used in several (colloquial) varieties of English around the world irrespective of the general structure and semantic content of the main clause that embeds it. Looking at the facts, it would seem that a possible source of this generalized invariant tag is a truncated assertive clause conceived at the beginning of the declarative main clause.
Acknowledgements
I have benefitted from a semestral research allowance from the Faculty of Arts at the University of Buea, Cameroon, and a quarterly research modernization allowance from the Cameroonian Ministry of Higher Education for which I am grateful.
BLASIUS ACHIRI-TABOH obtained a PhD from the Johann Wolfgang Goethe University in Frankfurt am Main, Germany. He is currently a Senior Lecturer of Linguistics at the State University of Buea in Cameroon, where he teaches undergraduate and postgraduate courses in Generative Syntax, Semantics, and in English and Bantu Morpho-Syntax. His research affiliations lie in current world English and usage, English Syntax, the syntax of the clausal left periphery, and Bantu Morpho-Syntax. Email: achiri.blasius@ubuea.cm.
