Printer Friendly

La sintaxis de la particula pragmatica confirmatoria innit.

The Syntax of the Confirmatory Pragmatic Particle Innit


The particle innit has been defined in the literature as an invariant question and follow-up tag that behaves non-canonically (Sailor 2011; Palacios Martinez 2013). (1) Although it is frequently used in British English, it is also attested in other varieties of English such as Malaysian English, Hong-Kong English, Indian English and White South African English among others (Kortmann and Lunkenheimer 2011). In this article, I argue that when a speaker uses innit to request the addressee's confirmation on the asserted proposition, as in (1)--where Si, S2 and S3 indicate three different speakers--it should be analysed as a pragmatic particle that constrains a higher-level explicature (Espinal 2011, 65). That is, in (1), innit indicates that S2 expects the asserted proposition (It is a lot, which repeats the assertion made by Si That's a hell of a lot of people) to be the most relevant in the discourse context. In addition, by using innit, S2 seeks the addressees' confirmation of the asserted proposition, p, which s/he obtains, since S3 replies with a confirmatory answer. The claim that innit is a pragmatic particle has implications for its formal and pragmatic analysis.

(1) Si: That's a hell of a lot of people.

S2: It is a lot, innit.

S3: Yes. (bnc, JT3 367)

(Krug 1998, 155)

As argued in Krug (1998), innit has resulted from a process of lexicalisation (Traugott 1994): the invariable tag question isn't it has become monomorphemic and opaque, and is, therefore, no longer divisible into the three lexical items that it comprises, namely the third person singular present form is, the negative marker not and the third person singular neuter subject pronoun it. In this respect, innit is different from other invariant tags or pragmatic particles such as right, yes, no and eh.

Before lexicalising into innit, however, the tag isn't it extended to persons other than the third person singular, to subjects in the main clause other than it, and to other tenses in the main clause other than the present. In fact, this change is, according to Krug (1998, 149), not exclusive to the tag isn't it, as the same phenomenon is also attested in other tag questions in different non-standard varieties of English. For instance, the forms don't and doesn't often merge into don't for all persons in non-standard English, whereas ain't may replace several negated auxiliaries (e.g., be, do and have) (cf. Anderwald 2002). There would therefore seem to be a tendency towards the simplification ofthe English question tag system.

Different pragmatic functions have been identified for innit in the literature (cf. Krug 1998; Andersen 2001; Stenstrom et al. 2002; Palacios Martinez 2011, 2012 and 2013); the most common are listed in (2a-f) for the so-called question tag innit and in (2g and 2h) for the follow-up tag innit. Each of the listed functions is illustrated in (3).

(2) a. Facilitative: it asks for the speaker's agreement.

b. Epistemic: it asks for the speaker's verification.

c. Challenging: it expresses aggression.

d. Filler: to keep conversation going.

e. Dramatic effect: it introduces something important in the narration.

f. Softening device: it functions as a request.

g. Emphatic: it emphasises something.

h. Emotional: it expresses surprise and incredulity.

(adapted from Palacios Martinez 2012)

(3) a. Those old games, they're so shit, innit? (Andersen 2001, 119)

b. You told your mum yesterday, innit? (Andersen 2001, 119)

c. A: Got any new games for your computer?

B: No, it's fucked, innit? You must have fucked it up. (Andersen 2001, 119)

d. Wi: yeah, but the insurance company are probably gonna pay erm, through me innit?

W2: oh yeah {unclear} {nv} laugh {/nv} COLT, 56-57

(Stenstrom and Jorgensen 2008, 647)

e. Susie: (...) But I was walking down <??> the street </> and this Turkishman, scratching his nose and, listen, <nv>mimicking bringing up phlegm and spitting </nv> in front of me, there's me, ah ah what are you doing! He started talking to me < uncle ar>

Ryan: <unclear> [<nv>laugh</nv>]

Susie: [Through his nose innit?] Listen, my cousin does, sometimes goes, watch this ... through his nose. (...) (Stenstrom et al. 2002, 171)

f. Grace: What Mothercare was it? What Mothercare ?

Samantha: All I'm looking for [<unclear>]

Dawn: [<unclear> it's wicked, there's only one Mothercare innit Grace.

(Andersen 2001, 124)

g. Lynne: He thought we were lying. <nv>laugh</nv>Can you imagine, <unclear>


Caroline: [Innit! Oh my god] I would just die! (140804/1:33)

(Andersen 2001, 146)

h. A: I've never, I've never heard Jim's voice before.

B: Innit?

A: Never. (COB132707/302) (Palacios Martinez 2011, 120)

In this paper, I only focus on (2a and 2b), the facilitative and epistemic functions of innit, which I will group under the label confirmatory innit, as they both have in common the fact that, by using innit, the speaker is ultimately double-checking on the truth of the expressed proposition. As Palacios Martinez (2012) puts it, in the case of epistemic innit, the speaker double-checks the truth of p or notp, whereas by using facilitative innit, the speaker is double-checking the addressee's agreement with the truth ofp or notp. Other so-called invariant question tags such as right,yes, no and eh fulfil a similar function. Hence, it seems highly plausible that the formal account developed in this work is also applicable to them.

The paper is organised as follows. In section two, I report three pieces of evidence discussed in Palacios Martinez (2011, 2012 and 2013) that show that innit departs from the usual behaviour of other question tags in English. Section three is devoted to introducing some theoretical background on pragmatic particles and speech acts that will be central to the account of the syntax of confirmatory innit that I put forward in section four. Finally, section five concludes the article.


By analysing the data in the Bergen Corpus of London Teenage Language (colt) and the London Linguistic Innovators Corpus, Palacios Martinez (2011, 2012 and 2013) describes a number of crucial respects in which innit does not behave like a canonical question tag. These are the following.

Firstly, innit may appear as a tag to clauses with be, as in (4a), have, as in (4b), or replacing do in the question tag, as in (4c). It can also occur as a tag to clauses with modal verbs like can, as in (4d), could, as in (4e), will, would, as in Qf), might, as in (4g), must, as in Qh), and should, as in (4i). (2)

(4) a. if as long as her clothes look alright underneath then it's not too bad innit? (CO/135201/67) (3)

b. Oh, she got A levels, innit? (CO/B133203/385)

c. told, you told mum [yesterday innit?] (CO/B139610/9)

d. you can have it for Friday, innit?. (CO/B138301/332)

e. could have got a bigger size innit Dawn? (CO/B134901/297)

f. [oh my god] I would just die, innit? (CO/B140802/46)

g. She might wear her shorts thing innit? (co/135201/67)

h. they must have the wrong place, innit? (CO/B135205/15)

i. should be in all day innit? (CO/B138301/20) 2 3

Secondly, innit occurs with clauses that have it as their main subject, as is the case in (4a), but it also occurs with other pronouns, such as I, in (4f),you, in (4c, d), s/he, in (4b, g), they, as in (4h), and also the expletive there. (4)

Thirdly, the particle innit may not agree with the tense of the main clause, as illustrated in (5), where the verb is in the past. (5)

(5) You forgot your book innit (CO/135201/67)

Finally, innit may optionally exhibit polarity reversal. That is, while innit can occur with positive sentences (thus showing reverse polarity with respect to the main clause it tags), it can also occur with negative sentences, as illustrated in (6).

(6) a. If as long as her clothes look alright underneath it's not too bad innit (CO/13 5201/67)

b. didn't have a hair cut then innit? (CO/B134803/141)

c. A: No, don't give anyone else the work then.

B: No, that you haven't marked, that you haven't marked innit. (CO/B134401/311)

d. You wouldn't think so innit? (CO/B132707/149)

In the present article I pursue the idea that confirmatory innit is not a question tag at all, although it is obvious that its use often yields an effect similar to that of a canonical question tag. This is shown in (7), where some of the sentences with innit from (4) have been paraphrased with canonical question tags.

(7) a. Oh, she got A levels, didn't she ?

b. told, you told mum [yesterday], didn't you?

c. you can have it for Friday, can't you?

d. [oh my god] I would just die, wouldn't I ?

Syntactically, however, innit is best analysed as a pragmatic particle (Espinal 2011) that encodes linguistic information that is relevant to the interpretation of the utterance, but that does not directly contribute to the truth-conditions of the proposition (Wilson and Sperber 1993, Espinal 2011).

As shown in (8), sentences with innit can also be paraphrased with what Krifka (forthcoming) labels as negated polarity questions, which are polar questions where negation is interpreted high in the clause (i.e. in a position outside the proposition) (Ladd 1981; Romero and Han 2004; Asher and Reese 2007; Krifka forthcoming; among others). Therefore, a formal analysis of innit should also be able to capture this parallelism.

(8) a. Oh, didn't she get A levels?

b. told, didn't you tell mum [yesterday?]

c. can't you have it for Friday?

d. [oh my god] wouldn't I just die ?

In the next section, I first present Espinal's (2011) account of pragmatic particles and then discuss Krifka's (forthcoming) analysis of negated polarity questions within a framework for the interpretation of speech acts. Reviewing these two particular pieces of research is crucial for the syntactic and pragmatic characterisation of confirmatory innit that is given in section four, where I defend the view that the confirmatory pragmatic particle innit does not behave like a canonical question tag because it is not a question tag syntactically speaking. Such a claim has an important implication for the study of question tags: namely, that invariant questions tags (e.g. innit, but also right?, eh?, yes?, cf. Sailor 2011), as they have been referred to in the literature (Tottie and Hoffmann 2006; Columbus 2009a, 2009b; Torgersen et al. 2011; among others), do not exist as a separate class of question tags. Rather they are pragmatic particles that fulfil a communicative function that is similar to that served by question tags.

3. THEORETICAL BACKGROUND: ESPINAL (2011) AND KRIFKA (FORTHCOMING) Addressing the syntactic and semantic nature of pragmatic particles in Catalan--a Romance language--Espinal (2011, 49) mentions a number of characteristics that are common to them. These characteristics are listed in (9), and can be used to evaluate whether innit qualifies as a pragmatic particle or not.

(9) a. They constitute a closed class.

b. They do not belong to any of the three basic classes of lexical categories:

Nouns, Verbs and Adjectives.

c. They are clearly functional.

d. They codify different types of linguistic information relevant for utterance interpretation.

e. They have the possibility of occurring as parenthetical expressions, or simply as isolated syntactic items.

f. Their meaning is underspecified by grammar.

g. They do not contribute to truth-conditional effects.

For a start, it is clear that innit fulfils the property in (9b): it is not a Noun, a Verb or an Adjective but a lexicalised expression. Palacios Martinez (2012) provides data, reproduced here as (10) and (11), which leads to the conclusion that innit can occur parenthetically (i.e. not as part of the host structure), as in (ioa-c), or in isolation, as in (11), thus aligning itself with pragmatic particles through meeting the condition (9e).

(10) a. Zack: everyone needs to calm down. Carved out

Alex: innit. Need to calm down a little. (lidhavy)

b. no I'm not that bad though but there is innit there's a scary one in the club (lidhavy)

c. Tina: #2 /er he goes after/ everyone innit every single person (lidhavy)

(11) Zack: but yeah when you're trying to do good man don't wanna see that though Alex: innit? (lidhavy)

Furthermore, innit also seems to fulfil the property in (9g): it clearly does not contribute to truth-conditional effects, as its presence/absence does not alter the proposition that is being asserted; and it also meets (9f), as illustrated by the fact that the same particle exhibits a wide range of functions or meanings, something that has been noted previously in the literature (Krug 1998; Andersen 2001; Stenstrom et al. 2002; Palacios Martinez 2011, 2012 and 2013).

Later in the section, it will become clear that innit is clearly functional (9c), and that it codifies linguistic information that is relevant to utterance interpretation, (9d): by using it the speaker is double-checking either the truth of the proposition asserted or doublechecking whether the addressee agrees with the proposition asserted by the speaker, both of which are functions related to procedural information that constrains the communicated proposition.

For all the reasons discussed above, and given the striking differences between the behaviour of canonical question tags and that of innit, I conclude that it is more desirable to put forward an analysis of innit within what we know about the syntax of pragmatic particles than doing so within the syntax of question tags. To do so, it is pertinent at this point to outline the main aspects of Espinal's (2011) analysis of pragmatic particles, with special emphasis on those that are defined as confirmatory.

Espinal (2011) makes use of Rizzi's (1997, 2001, 2004) and Cinque's (1999, 2002, 2006) cartographic projects to analyse the various pragmatic particles that exist in Catalan. The cartographic approach aims at establishing "the right structural maps for natural language syntax" (Cinque and Rizzi 2008, 42) and it endorses the view that the V(erb) P(hrase), the I(nflection) P(hrase) and the C(omplementizer) P(hrase) are split into a number of dedicated functional heads that allow different syntactic features to be appropriately checked. (6) Following these scholars, Espinal (2011) proposes the clausal structure in (12). (7)

(12) [[sub.GroundP] Ground [[sub.vocP] Voc [[sub.ExclP] Excl [[sub.porceP] Force [[sub.TopP] Top [[sub.FocP] Foc [[sub.TopP] Top [[sub.FinP] Fin [[sub.PolP] Pol [[sub.IP] Infl]]]]]]]]]]

(Espinal 2011, 50)

Pol (Laka 1990) encodes the polarity of the clause, whereas Fin expresses whether the clause is finite or non-finite. Foc hosts foci and wh-phrases, whereas Top hosts topics (either below or above Foc). Both Foc and Top are optional. Force determines the illocutionary force of the clause (declarative, interrogative, relative, imperative), and Ground (Bianchi 2004) is occupied by known or presupposed material.

In the specific case of confirmatory particles, these are assumed to occupy the Specifier of ForceP, in line with Hernanz and Rigau (2006) and Prieto and Rigau (2007). This is illustrated with the Spanish particle verdad lit. "truth" in (13) and with Catalan oi in (14). The complementizer que is the head of ForceP (Espinal 2011, 65).

(13) a. Verdadque Pedro es inteligente? (Spanish)

PART that Pedro is intelligent

'Isn't it true that Pedro is intelligent ? / 'Isn't Pedro intelligent ?' / 'Pedro is intelligent, isn't he ?'

b. [[sub.ForceP] Verdad [[sub.Force] que [[sub.IP] Pedro es inteligente]

(14) a. Oi que ha arribat en Pere ? (Catalan)

part that has arrived the Pere

'Isn't it true that Pere has arrived ?' / 'Hasn't Pere arrived ?' / 'Pere has arrived, hasn't he ?'

b. [[sub.ForceP] Oi [[sub.Force] que [[sub.IP] ha arribat en Pere]

Both (13) and (14) are also possible with the particle occurring in a sentence-final position. This is illustrated in (15).

(15) a. Pedro es inteligente, verdad? (Spanish)

Pedro is intelligent part

'Pedro is intelligent, isn't he ?'

b. Ha arribat en Pere, oi? (Catalan)

has arrived the Pere part

'Pere has arrived, hasn't he ?'

Espinal (2011,65) argues that by using these confirmatory particles, the speaker assumes thatp, the proposition asserted, is part of the common ground and seeks confirmation for the truth of p. Such particles may occur both with affirmative and negative sentences. With affirmative sentences, as in (13)^15), the speaker expects to confirm p, whereas with negative sentences s/he intends to confirm notp, as in (16), which contains the negative counterparts of (15).

(16) a. Pedro no es inteligente, verdad? (Spanish)

Pedro not is intelligent part

'Pedro is not intelligent, is he ?'

b. No ha arribat en Pere, oi? (Catalan)

not has arrived the Pere part

'Pere hasn't arrived, has he ?'

In the remaining part of this section, I summarise the main points of Krifka's (forthcoming) analysis of negated polarity questions within a theory of speech acts, with the aim of showing, in section four, that the combination of Espinal's (2011) account of pragmatic particles and Krifka's analysis of biased questions allows us to explain not only the non-canonical behaviour of innit as a question tag, but also the points that it has in common with both. Herein lies the novelty of the present account, which distinguishes it from other descriptive and/or purely pragmatic approaches to innit: not only does it constitute an attempt to offer new insights into the grammar of innit, but it also relates its syntax with that ofother structures with which innit has not previously been shown to relate.

Before looking at the details of Krifka's proposal, however, let us introduce the problem he focuses on. As previously noticed by Ladd (1981), a sentence like (17) is ambiguous in the sense that it allows negation to be interpreted outside the proposition, as in (17a), or inside the proposition, as in (17b). That is, the reading in (17a) would be understood to follow from the fact that negation is expressed in a position that is high in the clause and external to the proposition, whereas (17b) would be analysed as containing an instance of negation that is internal to the proposition.

Actually, the question in (17) can be easily disambiguated by means of the positive polarity item (ppi) too and the negative polarity item (npi) either. As the licensing requirements of ppis and NPis involve a very particular (anti-)licensing relationship with negation, testing which reading of (17) is associated with too and which with either can be taken as a diagnostic for the position of negation in the clause. As is well known, ppis cannot occur under the scope of negation, whereas npis are ungrammatical if they are not bound by negation. A sentence such as (18), therefore, can only be grammatical if negation is not part of the proposition and does not bind the ppi too. Hence, the only interpretation of (18) is (17a).

(17) Isn't there a vegetarian restaurant around here ?

a. Speaker wants confirmation that there is a vegetarian restaurant around here.

b. Speaker wants confirmation that there is no vegetarian restaurant here.

(Krifka forthcoming, 1)

(18) Isn't there a vegetarian restaurant around here, too?

A sentence such as (19), by contrast, can only be grammatical if either is bound by clause-mate negation, as required by npis. This disambiguates its meaning towards (17b).

(19) Isn't there a vegetarian restaurant around here, either?

After Ladd (1981), a number of scholars have developed other accounts that try to make sense of the ambiguity in (17) (cf. van Rooij and Safarova 2003; Romero and Han 2004; Reese 2007; Asher and Reese 2007; Repp 2013). However, as discussed in Krifka (forthcoming, 3 and ff.) they are all problematic to varying extents.

For van Rooij and Safarova (2003), for example, when a polarity question is based on a sentence denoting a propositionp, thenp is of more pragmatic utility than not p. Hence, in the case of a negated polarity question like (17), the interpretation in (17a) becomes possible if the question is assumed to be based on the negation of the sentence There is a vegetarian restaurant around here (i.e., [??]'there is a vegetarian restaurant around here'), for which notp is of more pragmatic utility thanp. While this view easily accommodates the bias of negated polarity questions, it cannot explain why the negative polarity item either cannot co-occur with (17) when the intended interpretation is (17a).

In Romero and Han's (2004) account, by contrast, the ambiguity in (17) is assumed to follow from the scope of a pragmatic operator, verum, with respect to negation. verum indicates how strongly a proposition can be taken to be part of the common ground (CG), and interacts with negation to yield the ambiguity in (17): while in (17a) negation takes scope over verum ([??]verum[[phi]]), thus expressing (20a), in (17b), verum takes scope over negation (verum[[??][phi]]), hence expressing (20b).

(20) a. [??]verum([phi]): 'It is not for sure that <[pounds sterling] should be added to the cg'

b. verum([??][phi]): 'It is for sure that [??][phi] should be added to the cg'

(Krifka forthcoming, 3)

Again, Krifka (forthcoming, 4) discusses a number of problems with Romero and Han's (2004) proposal: first, it is not clear how verum becomes available by proposing negation to an auxiliary; secondly, the authors do not explain why questions like (21) are biased (their interpretation is like [17b]); and finally, given the fact that verum is a pragmatic operator and not a proposition, it is not clear how or why it can be negated, as in (20a) ([??]VERUMf[[phi]]).

(21) Is there no vegetarian restaurant around here ?

In Repp (2013), another operator is postulated alongside verum. This is falsum, which, indicates, as does verum, whether the proposition should be considered part of the CG. Repp asserts that in (17a) high negation expresses falsum, thus indicating that the proposition should not be considered part of the CG. Since (17) is a question, the speaker is asking the addressee to consider the degree to which p should not be added to the CG. In other words, the speaker is asking the addressee whether s/he would deny p. One issue that is not, though, clarified in Repp's proposal is why it is possible for negation to be propositional--as in (17b)--or the pragmatic operator falsum, as in (17a).

Finally, in Reese (2007) and Asher and Reese (2007) negated polarity questions are claimed to be a combination of an assertion and a question that is a request for the addressee to acknowledge, contradict or confirm the speaker's assertion. This analysis has some advantages, such as, for instance, the fact that it correctly predicts that NPIs like either should not be possible in negated polarity questions (which are assumed to be nonnegated assertions), or the fact that negated polarity questions are compatible with an expression like after all, which occurs with assertions. However, Reese's and Asher and Reese's analysis has a crucial problem: it does not explain how syntactic high negation contributes to the assertion that is then tagged by the question speech act.

In order to explain the ambiguity in (17), Krifka proposes an analysis of outer negation within Cohen and Krifka's (2011) theory of speech acts as transitions between commitment spaces. While he builds on Ladd's (1981) intuition that high negation is syntactically high enough to scope over the whole proposition, he also aligns himself with Romero and Han (2004) and Repp (2013) in assuming that negation takes scope over a speech-act operator assert. Unlike previous accounts, however, Krifka explicitly discusses which mechanism is involved in the negation of a speech act. In line with Reese (2007) and Asher and Reese (2007), negated polarity questions are assumed to involve both an assertion and a question, with outer negation serving the purpose of allowing the speaker to ask whether the addressee would refrain from assertingp (Krifka forthcoming, 2). In the case of assertions, the type of speech act of interest in this paper, two commitments are expressed: first, a commitment to the proposition by the speaker, and second, a call on the addressee to share such commitment to the proposition, such that the propositionp becomes part of the CG (i.e., so that the addressee accepts the truth of the proposition).

In line with Krifka (forthcoming, 11), I assume propositions to syntactically correspond to Tense Phrases (tp). These are rendered speech acts thanks to the presence of illocutionary operators, which are assumed to be located in ForceP, in the left periphery of the clause. The syntax of an assertion like (22) would be (23), for instance, where ass stands for the speech-act operator assert. Notice that, in this particular respect, Krifka's account coincides with the assumptions made in Espinal (2011) (cf. (12) above) for clausal architecture.

(22) There is a vegetarian restaurant here.

(23) [porceP [porcePass [[sub.TP] there is a vegetarian restaurant here]]]

(Krifka forthcoming, 11)

Krifka (forthcoming, 24) introduces another speech-act operator when discussing the syntax of declarative questions like (24) (cf. Gunlogson 2002). This is REQUEST, which operates on assertp, as shown in (25), and constitutes an alternative to Qu(estion), an operator used in interrogatives that involves movement of the copula, as in (26). (8)

(24) There is a vegetarian restaurant here ?

(25) [[sub.ForceP] REQUEST [[sub.Forcep] ASS [TP there is a vegetarian restaurant here?]]]

(Krifka forthcoming, 25, example [49])

(26) [[sub.ForceP] QU+[Is.sub.i], [[sub.TP] there [t.sub.i] a vegetarian restaurant here?]]]

The declarative question in (24) is biased towards p being true. According to Krifka (forthcoming, 25) this is so because (24) involves REQUEST. By constructing the question with REQUEST and not by means of a qu operator, the speaker who utters (24) is indicating that s/he does not takep and notp to both be possible; rather s/he shows that only the option where p is asserted is considered.

Negated polarity questions like (17) (with the interpretation of [17a]) are analysed as "a request question, with a denegation that scopes over the assertion operator" (Krifka forthcoming, 29). This is shown in (27), where negation is a speech-act operator rather than an instance of propositional negation.

(27) [[sub.ForceP] request [NegP [is.sub.i]-n't [[sub.ForceP] ASS [[sub.TP] there [e.sub.i] a vegetarian restaurant here]]]]

(Krifka forthcoming, 29, example [60])

By using a negated polarity question like (17) (interpreted as [17a]), a speaker asks whether the addressee would reject the assertion ofp (Krifka forthcoming, 30). As discussed in the next section, I show that a declarative clause with the confirmatory particle innit achieves the same effect. I also show that an analysis along the lines of Krifka's makes it possible to explain why declarative sentences with innit can be interpreted not only as declaratives with canonical question tags, but also as negated polarity questions with high negation.


Given that, as has been noted earlier in the paper, most of the sentences in (4) (some of which have been repeated in [28] for convenience) can be paraphrased with declaratives followed by a canonical question tag, as in (29), and with negated polarity questions, as in (30), an analysis of the sentences in (28) that accommodates these similarities in a straightforward manner is highly desirable.

(28) a. Oh, she got A levels, innit? (CO/B133203/385)

b. told, you told mum [yesterday innit?] (CO/B139610/9)

c. you can have it for Friday, innit? (CO/B138301/332)

d. [oh my god] I would just die, innit? (CO/B140802/46)

(29) a. Oh, she got A levels, didn't she?

b. told, you told mum [yesterday], didn't you?

c. you can have it for Friday, can't you?

d. [oh my god] I would just die, wouldn't I ?

(30) a. Oh, didn't she get A levels?

b. told, didn't you tell mum [yesterday?]

c. can't you have it for Friday?

d. [oh my god] wouldn't I just die ?

Krifka (forthcoming, 27) argues that questions containing question tags, which are biased questions, involve the speaker's commitment to the truth of the proposition and restrict the addressee's reaction to either accepting p as part of the CG, or to asserting notp without having to reject treatingp as CG. As such, they are more conciliatory than assertions without the question tag. Krifka does not however include any syntactic analysis of questions with question tags, so I propose one here.

I will assume, in line with Sailor (2011) and Barros and Craenenbroeck (2013), that questions with question tags involve vp ellipsis. That is, question tags are adjoined to the host clause (an assertion), and then affected by vp ellipsis. The syntax of the question tag in (28a) above is assumed to be as in (31) below, where the strikethrough indicates ellipsis.

(31) [[sub.ForceP] REquEST [[sub.Negp] Didrn't [[sub.ForceP] ASS [[sub.TP] she [e.sub.i] [beginning strikethrough]get A levels[end strikethrough]]]]]

Notice that the analysis is very similar to the one Krifka (forthcoming) gives for negated polarity questions such as the one in (32a). As can be seen in (32b), ForceP contains a request operator that acts on the assert operator (i.e. it requires the addressee to consider whether s/he can also be committed to the truth ofp so thatp can be taken to be part of the cg). As negation scopes over assert, the speaker is actually asking the addressee whether the assertion of p (She got A levels) is to be excluded in favour of notp (She didn't get A levels).

(32) a. Didn't she get A levels?

b. [[sub.ForceP] REquEST [[sub.NegP] Didi-n't [[sub.ForceP] ASS [[sub.TP] she [e.sub.i] get A levels]]]]

Let us now provide the syntactic structure I assume for (28a), which was produced with innit. Recall that Espinal (2011) claims that confirmatory particles occur in ForceP.

Adopting the clausal structure that she proposes, (28a) could be initially represented as follows:

(33) [[sub.GroundP] Ground [[sub.ForceP] innit Force [[sub.TP] she got A-levels]]]

Since the tp she got A-levels is an assertion, an assert operator would also be present in ForceP (see footnote 8 above). This is illustrated in (34).

(34) [[sub.GroundP] Ground [[sub.ForceP] innit [[sub.ForceP] ASS Force [[sub.TP] she got A-levels]]]]

My claim is that the particle innit triggers the presence of the request operator, as it requires the addressee to consider the truth of the assertedp.9 The sentence in (34) is equivalent to a biased negative question such as (32a) due to the fact that negation (which is inherent to the particle innit, see section one) is in a scope position over the operator ASSERT. The same happens in (32b) after auxiliary movement.

(35) [[sub.GroundP] Ground [[sub.ForceP] Request [[sub.ForceP] innit [[sub.ForceP]ASS Force [" she got A-levels]]]]]

One last point to clarify is how the correct word order for a declarative with confirmatory innit is derived. As shown in the examples in (4), (5) and (6), confirmatory innit tends to occur word-finally. However, in the structures in (33)^35) it occupies a position to the left of the tp. After Spell-Out, however, innit surfaces to the right of the asserted proposition.

In what follows, I claim that the surface position of innit with respect to the tp it relates to crucially depends on the functional category GroundP, which, as can be seen in (12) above and in (33M35), is part of the clausal structure assumed in Espinal (2011). If one takes seriously the idea that the presence of innit indicates that the proposition (i.e., the tp) should be considered CG, then, it is possible to assume that a fourth step takes place in the derivation of a sentence like (28a) in addition to the three steps illustrated in (33)-(35). As shown in (36), the last step in the derivation of(28a) would be movement of tp to GroundP. Such movement would be motivated by an Edge Feature in the split CP, which attracts the tp to Spec, GroundP.10

9 Krifka (forthcoming, 27) states that "request can also be expressed syntactically, in a similar way as QU, by triggering head movement of auxiliary verbs or copulas," as in (31). He also assumes that request can be expressed prosodically, specifically, by using the prosodic contour H% (cf. Beckman et al. 2005 for details on the ToBI prosodic annotation system).

10 Remnant movement of IP to the Specifier of GroundP has also been proposed for other discourse-related phenomena that affect the surface word order of the sentence, such as right-dislocation of topics in Italian (cf. Frascarelli 2000 and ff., Cardinaletti 2001, 2002; Samek-Lodovici 2005), French stylistic inversion (Poletto and Pollock 2004) and instances of sentence-final sentential negative marker NO in Italian (Poletto 2010). Furthermore, Yang (2008) relates Edge Feature-movement to discourse-effects.

(36) [[sub.GroundP] She got A-levelSi Ground [[sub.ForceP] [[sub.ForceP] innit [[sub.ForceP]ASS Force [[sub.TP] [t.sub.i]]]]]]

Recall that by using the confirmation particle innit, the speaker is not only requesting the addressee to consider whether the assertion of p is to be excluded or not, but also whether p is to be taken as CG. I argue thatp is in fact considered to be part of the CG by the speaker. If this situation is translated into syntactic terms, the entire tp should occupy a position in GroundP, the functional projection reserved for linguistic material that qualifies as CG. Assuming a copy and deletion theory of movement (Chomsky 1995), the relevant structure would be (37).

(37) [[sub.GroundP] She got A-levels [sub.Ground] [[sub.ForceP] REQUEST [[sub.ForceP] innit [[sub.ForceP] ASS Force [[sub.TP] she got A-levels]]]]]

While the Phonetic Form interprets the highest copy, as in (38a), resulting in the confirmatory particle innit surfacing to the right of the TP, the Logical Form interprets the lowest copy, as in (38b), with request scoping over innit, assert and the tp, and innit scoping over ASSERT and the TP.

(3 8) a. [[sub.GroundP] She got'veh Ground [[sub.ForceP] REQUEST [[sub.ForceP] innit [[sub.ForceP]ASS Force [[sub.TP] [beginning strikethrough]she got A-levels][end strikethrough]]]]] (pf interpretation)

b. [[sub.GroundP] [beginning strikethrough]She got A-levels[end strikethrough] Ground [[sub.ForceP] REQUEST [[sub.ForceP] innit [[sub.ForceP]ASS Force [[sub.TP]

she got A-levels]]]]] (lf interpretation)

As shown in section one, there are three ways in which innit and canonical question tags clearly diverge. First, innit can occur with a main clause containing auxiliaries other than is. Second, innit can occur with a main clause that contains a subject other than it. Third, innit can occur with a main clause that contains sentential negation. While these characteristics of innit are problematic for its analysis as a question tag, they are perfectly compatible with the proposal presented in this article, namely that innit is a pragmatic particle. As such, it is invariable and does not involve any vp ellipsis mechanism (cf. (31) above), nor does it require a mechanism of polarity reversal.

When innit occurs with a negative clause, as in (6d), repeated here as (39) for convenience, the syntactic and interpretive mechanism that has been developed for affirmative clauses with innit is the same.

(39) You wouldn't think so innit?

In (39), what is asserted is notp, as the affirmative sentence I would think so is negated by means of the negative marker -n't, which occupies a syntactic position within the TP. The operator assert, therefore, scopes over the TP-internal negation, as shown in (40a). Innit is merged in ForceP, as shown in (40b).

(40) a. [[sub.GroundP] Ground [[sub.ForceP] ASS Force [[sub.TP] you wouldn't think so]]]

b. [[sub.GroundP] Ground [Forcep innit [Forcepass Force [[sub.TP] you wouldn't think so]]]]

As was the case in (35), the inherent negation in innit motivates the presence of a request operator, as shown in (41a). Finally, the tp moves to GroundP, as illustrated in (41b). By uttering the sentence in (6d)/(3 9) the speaker requires the addressee to consider whether not p (You wouldn't think so) is part of the CG and whether notp is to be excluded in favour ofp (You would think so).

(41) a. [[sub.GroundP] Ground [Forcep request [Forcep innit [Forcep ass Force [[sub.TP] you wouldn't think so]]]]]

b. [[sub.GroundP] you wouldn't think soi Ground [[sub.ForceP] REQUEsT [[sub.ForceP] innit [[sub.ForceP] ass

Force [TP ti]]]]]

5. Conclusion

In this article I have shown that innit can be analysed as a pragmatic particle that does not contribute to the truth conditions of the clause it occurs with. Rather, innit is a confirmatory particle that asks the addressee to provide confirmation of the proposition being expressed (eitherp or notp).

As innit is not a question tag involving vp ellipsis, it is not surprising that, as has been discussed in section one, it does not exhibit the restrictions that apply to these syntactic structures. Unlike canonical question tags, innit occurs with a variety of different auxiliaries (not just with the third person singular present form of be), with a variety of different subjects (not just with third person singular neuter pronoun it), and with affirmative and negative main clauses indistinctively. If one wished to maintain the idea that innit is a question tag, these facts would certainly force the characterisation of innit as non-canonical or invariant. However, it seems to me that separating canonical question tags from invariant question tags (Sailor 2011) is unnecessary, since invariant or non-canonical question tags can actually be best analysed as well-behaved pragmatic particles.

I have modelled the syntax of the confirmatory pragmatic particle innit along the lines of Espinal (2011), who, adopting Rizzi's (1997, 2001, 2004) and Cinque's (1999, 2002, 2006) cartography of the left-periphery of the clause, locates this kind of particle in the Specifier of the functional projection ForceP. This view is compatible with Krifka's (forthcoming) account of questions with question tags and of negated polarity questions in English, which are assumed to involve two speech-act operators known as assert and request, and which have much in common with sentences with innit. As I have shown, sentences with innit can be paraphrased by a question with a canonical question tag, but also by a negated polarity question where negation is interpreted high in the clause (i.e., in a position outside the proposition).

These parallelisms fall into place in the analysis presented here. On the one hand, both the speech-act operators request and assert have been argued to be involved not only in the syntax of a declarative sentence with innit, but also in the syntax of canonical question tags. The latter, though, are crucially different from declarative sentences with innit in that they involve vp ellipsis.

On the other hand, the REQUEST and ASSERT operators are also involved in the derivation of negated polarity questions. The effect of high negation that is obtained by means of auxiliary movement in negated polarity questions is also observed in declarative sentences with innit due to the interaction of two factors. First, as innit is a confirmation pragmatic particle, it is merged high in the syntactic structure, namely in Spec, ForceP, which necessarily outscopes tp. Second, innit is inherently negative as it lexicalised from a negative question tag structure. Hence, the presence of innit in Spec, ForceP is comparable to the presence of high speech-act negation (expressed by means of an -n't affix on the auxiliary) scoping over the tp.

English has other so-called invariant question tags like right, yes, no, eh to which the present account can be extended. However, innit is probably the most puzzling of all, as it has transformed from a genuine question tag to a pragmatic particle. My proposal does not invalidate previous (pragmatic) accounts of innit. Rather, it offers some new insights on the syntax of at least two of its functions, namely the facilitative and the epistemic, while preserving the intuition that the syntax of the particle innit is to a certain extent connected--though not identical--to the syntax not only of question tags, but also of negated polarity questions with high negation.

Received 16 November 2013

Revised version accepted 4 July 2014


ANDERSEN, Gisle. 2001. Pragmatic Markers and Sociolinguistic Variation: A Relevance-Theoretic Approach to the Language of Adolescents. Amsterdam: John Benjamins.

ANDERWALD, Lieselotte. 2002. Negation in Non-Standard British English. Gaps, Regularizations and Asymmetries. Routledge Studies in Germanic Linguistics. London and New York: Routledge.

ASHER, Nicholas and Brian Reese. 2007. "Intonation and Discourse: Biased Questions." Interdisciplinary Studies on Information Structure 8: 1-38.

BARROS, Matt and Jeroen van Craenenbroeck. 2013. "Tag Questions and Ellipsis." Paper presented at the Deutsche Gesellschaft fur Sprachwissenschaft. Potsdam, 12-15 March.

BECKMAN, Mary E.Julia Hirschberg and Stefanie Shattuck-Hufnagel. 2005. "The Original ToBI System and the Evolution of the ToBI Framework." In Prosodic Typology. The Phonology of Intonation and Phrasing, edited by Sun-Ah Jun, 9-54. Oxford: Oxford up.

BIANCHI, Valentina. 2004. "Resumptive Relatives and lf Chains." In The Structure of cp and IP. The Cartography of Syntactic Structures, Vol. 2, edited by Luigi Rizzi, 76-114. Oxford: Oxford up.

CARDINALETTI, Anna. 2001. "Subjects and wh-questions. A Reply to Uriagereka (1999)." Ms., u of Bologna and u of Venice.

--. 2002. "Against Optional and Null Clitics. Right Dislocation vs. Marginalization." StudiaLinguistica 56 (1): 29-57.

CHOMSKY, Noam. 1995. The Minimalist Program. Cambridge, ma: MIT.

CINQUE, Guglielmo. 1999. Adverbs and Functional Heads. A Cross-linguistic Perspective. Oxford: Oxford up.

--. 2002. Functional Structure in DP and IP. The Cartography of Syntactic Structures, Vol. 1. New York: Oxford UP.

--. 2006. Restructuring and Functional Heads. The Cartography of Syntactic Structures, Vol. 4. New York: Oxford up.

CINQUE, Guglielmo and Luigi Rizzi. 2008. "The Cartography of Syntactic Structures." CISCL Working Papers on Language and Cognition 2: 43-59.

COHEN, Ariel and Manfred Krifka. 2011. "Superlative Quantifiers as Modifiers of Meta-speech Acts." The Baltic International Yearbook of Cognition, Logic and Communication 6: 1-56.

COLUMBUS, Georgie. 2009a. "A Corpus-based Analysis of Invariant Tags in Five Varieties of English." In Corpus Linguistics: Refinements and Reassessments, edited by Antoinette Renouf and Andrew Kehoe, 401-14. Amsterdam: Rodopi.

--. 2009b. "'Ah lovely stuff, eh?'--Invariant Tag Meanings and Usage Across Three Varieties of English." In Corpus-linguistic Applications: Current Studies, New Directions, edited by Stephan Th. Gries, Stefanie Wulff and Mark Davies, 85-103. Amsterdam: Rodopi.

ESPINAL, Maria Teresa. 2011. "Pragmatic Particles at the Syntax-cognition Interface." In The Pragmatics of Catalan, edited by Lluis Payrato and Josep Maria Cots, 49-80. Berlin and Boston: De Gruyter Mouton.

FRASCARELLI, Mara. 2000. The Syntax-phonology Interface in Focus and Topic Constructions in Italian. New York: Springer.

GUNLOGSON, Christine. 2002. "Declarative Questions." In Proceedings of Semantics and Linguistic TheoryXII, edited by B. Jackson, 124-32. Ithaca, ny: Cornell up.

HERNANZ, M. Lluisa and Gemma Rigau. 2006. "Variacion dialectal y periferia izquierda." In Andolingogoan. Essays in Honour of ProfessorEguzkitza, edited by Beatriz Fernandez and Itziar Laka, 435-52. Bilbao: u del Pais Vasco.

KORTMANN, Bernd and Kerstin Lunkenheimer, eds. 2011. The Electronic World Atlas of Varieties of English [eWAVE]. Leipzig: Max Planck Institute for Evolutionary Antropology.

KRIFKA, Manfred. Forthcoming. "Negated Polarity Questions and Denegations of Assertions." In Contrastiveness and Scalar Implicatures, edited by Ferenc Kiefer and Chungmin Lee, New York: Springer.

KRUG, Manfred. 1998. "British English is Developing a New Discourse Marker, innit?: A Study in Lexicalisation Based on Social, Regional and Stylistic Variation." Arbeiten aus Anglistik undAmerikanistik 23: 145-97.

LADD, Robert. 1981. "A First Look at the Semantics and Pragmatics of Negative Questions and Tag Questions." In Proceedings of the Chicago Linguistic Society. Chicago, 164-71.

LAKA, Itziar. 1990. "Negation in Syntax: On the Nature of Functional Categories and Projections." PhD diss., mit, usa.

LARSON, Richard. 1988. "On the Double Object Construction." Linguistic Inquiry 19 (3): 335-91.

PALACIOS MARTINEZ, Ignacio. 2011. "The Language of British Teenagers. A Preliminary Study of its Main Grammatical Features." Atlantis. Journal of the Spanish Association of Anglo-American Studies 33 (1): 105-26.

--. 2012. "Variation and Pragmatic Uses of innit in the Language of British and Adult Teenagers." Paper presented at the XXXVI Congreso Internacional de aedean. Malaga, 14-16 November, 2012.

--. 2013. "Non-standard Negation in Modern English. A Corpus-based Study of Four Salient Features." ES Revista de Filologia Inglesa 33: 211-25.

POLETTO, Cecilia. 2010. "The Syntax of Focus Negation." Quaderni di Lavoro Asit 10: 39-61.

POLETTO, Cecilia and Jean-Yves Pollock. 2004. "On the Left Periphery of Some Romance Wh-questions." In The Structure of cp and IP. The Cartography of Syntactic Structures, Vol. 2, edited by Luigi Rizzi, 251-96. Oxford: Oxford UP.

Pollock, Jean-Yves. 1989. "Verb Movement, UG and the Structure of ip." Linguistic Inquiry 20: 365-424.

PRIETO, Pilar and Gemmar Rigau. 2007. "The Syntax-prosody Interface: Catalan Interrogative Sentences Headed by que." Journal of Portuguese Linguistics 6 (2): 29-59.

REESE, Brian. 2007. "Bias in Questions." Ph.D. diss. u of Texas, Austin, USA.

REPP, Sophie. 2013. "Common Ground Management: Modal Particles, Illocutionary Negation and VERUM." In Beyond Expressives-Explorations in Use-conditional Meaning, edited by Daniel Gutzmann and Hans-Martin Gartner, 231-74. Leiden and Boston: Emerald.

Rizzi, Luigi. 1997. "The Fine Structure of the Left Periphery." In Elements of Grammar, edited by Liliane Haegeman, 281-37. Amsterdam: Kluwer.

--. 2001. "On the Position 'Int(errogative)' in the Left Periphery of the Clause." In Current Studies in Italian Syntax. Essays offered to Lorenzo Renzi, edited by Guglielmo Cinque and Giampaolo Salvi, 287-96. Amsterdam: Elsevier North-Holland.

--. 2004. "Locality and Left Periphery." In Structures and Beyond. The Cartography of Syntactic Structures, Vol. 3, edited by Adriana Belletti, 223-51. New York: Oxford UP.

ROMERO, Maribel and Chung-hye Han. 2004. "On Negative Yes/No Questions." Linguistics and Philosophy 27: 609-58.

ROOIJ, Robert van and Marie Safarova. 2003. "On Polar Questions." Proceedings of Semantics and Linguistic Theory XIII. Ithaca, NY: Cornell U, CLC.

SAILOR, Craig. 2011. Tagged for Deletion: A Typological Approach to VP Ellipsis in Tag Questions. ma thesis. UCLA, USA.

SAMEK-LODOVICI, Vieri. 2005. "Prosody-syntax Interaction in the Expression of Focus." Natural Language and Linguistic Theory 23: 687-755.

STENSTROM, Anna-Brita, Gisle Andersen and Ingrid Kristine Hasund. 2002. Trends in Teenage Talk. Corpus Compilation, AnalysisandFindings. Amsterdam and Philadelphia: John Benjamins.

STENSTROM, Anna-Brita and Annette Myre Jorgensen. 2008. "A Matter of Politeness ? A Contrastive Study of Phatic Talk in Teenage Conversation." Pragmatics 18 (4): 635-57.

TORGERSEN, Eivind Nessa, Costas Gabrielatos, Sebastian Hoffmann and Susan Fox. 2011. "A corpus-based Study of Pragmatic Markers in London English." Corpus Linguistics and Linguistic Theory 7 (1): 93-118.

TOTTIE, Gunnel and Sebastian Hoffman. 2006. "Tag Questions in British and American English." Journal of English Linguistics 34 (4): 283-311.

TRAUGOTT, Elizabeth. 1994. "Grammaticalization and Lexicalization." In The Encyclopedia of Language and Linguistics, Vol. 3, edited by Ronald E. Ahser and J. M. Y. (Seumas) Simpson, 1481-86. Oxford: Pergamon.

WILSON, Deirdre and Dan Sperber. 1993. "Linguistic form and Relevance." Lingua 90: 1-25.

YANG, Dong-Whee. 2008. "On Edge Feature Movement." Ms. Cambridge, ma: mit.


Universitat Autonoma de Barcelona

(1) This research has been funded by a research grant awarded by the Spanish Ministerio de Ciencia e Innovacion (FFI2011-23356), and by a grant awarded by the Generalitat de Catalunya to the Centre de Linguistica Teorica (2014SGR1013). I thank the two anonymous reviewers for their comments, which have helped me improve the manuscript. All remaining errors are mine.

(2) Palacios Martinez (2012) analysed 305 tokens of innit in COLT and found that 171 (68.95%) occurred with be in the main clause, 42 with do (16.93%), 15 with have (6.05%), 6 with can (2.42%), 3 with will, 3 with might and 3 with would (1.21%), 2 with must (0.87%) and a single token each of should and could (0.4%). The category "Other" was also included, with only 1 token (0.4%).

(3) All the examples in (4) are taken from Palacios Martinez (2013), as are the examples in (5) and (6).

(4) According to Palacios Martinez (2012), out of a total of 252 clauses with an overt subject, 152 had it (60.32%), 34 had you (13.49%), 20 had he (7.94%), 15 had they (5.95%) and 15 had she (5.95%), I was the subject in 8 clauses (3.17%), 7 had we (2.78%) and 1 had expletive there (0.40%).

(5) 18 out of the 260 analysed tokens contained past tense.

(6) See Larson (1988) and Pollock (1989) for arguments on split VP and split IP, respectively. See also Cinque (1999) for a hierarchy of Adverb Phrases in the IP domain.

(7) The heads Voc(ative) and Excl(amative) are occupied by vocative and exclamative particles.

(8) Krifka (forthcoming, 25) assumes ForceP to be recursive in cases where a speech act needs to contain another speech act (e.g., request needs to embed another speech act, such as an assertion). Hence, the (im)possibility of a given speech act to contain another regulates the recursivity of ForceP, thereby preventing its over-generation.

Susagna Tubau obtained a PhD in English Philology at the Universitat Autonoma de Barcelona and Universiteit van Amsterdam in 2008 and an ma in Linguistics from University College London in 2004. She is interested in theoretical syntax and negation in natural languages. She is Assistant Lecturer in the Department of English Philology of the Universitat Autonoma de Barcelona. Address: Departament de Filologia Anglesa i de Germanistica. Facultat de Filosofia i Lletres (Edifici B). 08193 Bellaterra (Cerdanyola del Valles), Barcelona, Spain. Tel.: +34 93 581 1779. Fax: + 34 93 581 2001.
COPYRIGHT 2014 Spanish Association for Anglo-American Studies (AEDEAN)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Tubau, Susagna
Publication:Atlantis, revista de la Asociacion Espanola de Estudios Anglo-Norteamericanos
Article Type:Report
Date:Dec 1, 2014
Previous Article:Explorando la combinacion de lenguaje, imagen y sonido en las metaforas de anuncios televisivos.
Next Article:Los valores culturales y su correlacion con las estrategias del metadiscurso interaccional en las paginas web de negocios espanolas y estadounidenses.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |