Introduction

1Ain't is considered as an emblematic non-standard form in modern English (Poplack 228). Along with this “taboo word”, to refer to the rather exhaustive Ain’thology (Donaher) exists innit, another non-standard form that is even more marginal, in as much as it has received less academic attention so far (apart from the studies by Andersen 2001; Stenström 2002; Palacios Martinez 2011). Our purpose is to examine linguistic variation through the prism of these two variation markers.

2Sociolinguistic variability, or the comparison between different dialects or varieties (Donaher) as well as a corpus-based comparative quantitative study could seem at the core of a paper dealing with ain’t and innit. But this is not our purpose or work frame. We aim at anchoring our analysis in morphology and macro-syntax. Under the heading “margin within margin”, we will consider whether innit is morphologically any more marginal than ain’t. We will also consider whether those two non-standard markers are that marginal in the linguistic system. Is their predicative reach any more limited than that of standard forms?

1. Why consider ain’t versus innit?

1.1 Beyond the socio-linguistic variation

4Why consider innit and ain’t together? Why not also consider fused negative non-standard markers, such as dinnit (equivalent of didn’t it), dunnit (equivalent of doesn’t it), wunnit (equivalent of wasn’t it), intit (equivalent of isn’t it), however little academic concern they might have received? The answer is that the purpose of this paper is to put into perspective these two marginal, potentially non-grammatically dependent forms, with their standard equivalent. Dinnit, dunnit, wunnit, intit have kept some grammatical dependency. They are still grammatically dependent tags, be it for the tenses, the auxiliary they make up for or the personal pronouns involved. Thanks to the limited grammatical dependency of ain’t (its only dependency being the use of the pronoun) and the lack of grammatical dependency of innit, thanks to the formal invariability of those variation markers, we will be able to focus on the features of ain’t and innit and show that the two non-standard forms go beyond their non-variant counterparts.

5Why compare two variation markers known to belonging to two different dialects and varieties? The non-standard American coalesced negation ain’t is emblematic of African-American Vernacular English, even if its origin can be dated back to eighteenth century cockney. We will look further into its current use.

6Innit rather belongs to British English, if not London English. Stenström (167) suggests:

As may have become evident, the form innit is used in London adolescent speech as a tag that can be appended to any statement, regardless of the grammatical features of the statement, hence its description as an invariant tag. The tag innit consists historically of a third person singular neuter pronoun and a form of the verb BE (isn’t it).

7We could content ourselves with a socio-linguistic or dialectal approach and do away with any systematic comparison. Still, ain’t can be found in south-eastern British English, and is widely used by British Asian youth. But the use of ain’t has spread outwards and the British National Corpus does contain 3496 utterances. This can be illustrated by the following transcription of exchanges in the Houses of Parliament:

(1) Either a Tory or a Labour minority government will look at the next election down the track and say, ‘Hang on, we ain't going to win’. (Daily Telegraph, electronic edition of 1992 0405)

8So we cannot limit ourselves to considering ain’t as an American variation marker versus innit as a British variation marker. We also encounter utterances where they do not overlap but actually collocate.

(2) Right, it's nice, it's a nice neat one that innit? Alec ain't spent his yet. (BNC, conversation, Central Northern England)

(3) cos that ain't very heavy to carry, innit? Oh, I think they're about two pounds. (BNC, conversation, Scotland)

9Since both ain’t and innit appear in these utterances, they are likely to have distinct linguistic roles. It is relevant to go beyond socio-linguistic approaches and consider the morphological and syntactic aspects. In example 3, innit is a tag whose anchor is ain’t in an already set negative predicate. As mentioned by Gaudy-Campbell (“valeur invariante”, 174-177), ain’t corresponds to a set predication. It is set to such an extent that the predication cannot be questioned afterwards by another standard predicate. Other forms appear, such as ain’t or honest.

(4) I don’t work for nobody, I ain’t got a job, honest. (BNC, Sergeant Joe. Staples, Mary Jane. London: Corgi Books, 1992)

(5) I ain’t done nuffing, honest, I ain’t. (BNC, Wychwood. Thompson, E V. London: Headline Book Pub. plc, 1992)

1.2 Ain’t and innit: non-variability as an issue

10To consider ain’t and innit from the angle of their non-variability, also called invariability, we will abide by the definition given by Andersen (98) for invariant units.

The term “invariant” is meant to suggest that these particular forms are used across the inflectional paradigm, regardless of the syntactic-semantic features of the preceding proposition that they refer to.

11When Guillaume (8) or Kimps (28) speak of ain’t as a non-variable tag, or rather part of a tag since the pronoun still varies, they refer to the word pattern that would be just the same for innit. Is ain’t indeed that invariable? Could it be more invariable than innit?

Of the three uses to which ain’t is put in non-standard varieties of English, ain’t as a generic negator before a main verb […] is clearly least frequent; typically, ain’t is used as the negated form of be and/or have […].

13Krug (150) also reminds us that ain’t results from a historic coexistence:

[ain’t] is derived by regular sound changes from standard English forms of both have and be. It therefore substitutes for two verbs that – apart from their high frequency and primary verb status – perform rather different syntactic functions and are quite distinct in meaning.

14Thus ain’t reconciles two auxiliaries with opposite orientation. Indeed, be works as a main-verb copula and indicates the identification (Bouscaren 4) of two terms. It is thereby orientated to the right. It is the opposite of have, where the second term is localized (Bouscaren 33) with respect to the first one and is thereby orientated to the left. With ain’t, this dichotomy between identification and localisation is done away with (Gaudy-Campbell, “ain’t et got”, 2004 102-103). The following occurrence is ambiguous, playing with and neutralising the orientation of auxiliaries that have been elided:

(6) Gone back to the Lord ain't he. (BNC, Albert Gunter: sermon (Pub/instit). Rec. on 2 Jul 1989)

15Ain’t neutralises the opposite orientations of the two auxiliaries; it neutralises the tenses and the distinction between do (even if this is far less frequent), have and be. Therefore, we consider that ain’t becomes a versatile auxiliary, a very versatile operator, or even a multi-auxiliary.

17Finally, the use of ain’t can go beyond this complementarity role. Although standard tags can appear in the following contexts, ain’t tends to be resorted to in order to make up for the subject and auxiliary that have been elided:

(7) Getting quite low, ain’t we? (BNC, conversation, London)

(8) Got a can ain’t we? (BNC, conversation, South Midlands)

(9) Sweet, ain’t she? (BNC, Conversation, London)

18The hearer need not retroactively retrieve the unit that has been elided but can interpret ain’t as equivalent to any orientated auxiliary. What is at play is the capacity of ain’t to operate as an equivalent to any orientated auxiliary. Thus, ain’t is a multi-operator, collocating with all pronouns, substituting for both tenses, for all auxiliaries whatever their orientation.

19So rather than merely considering ain’t as a non-standard form beside standard auxiliaries, we would like to put forward its versatility. Ain’t can be the substitute for all auxiliaries, cancelling their orientation and bearing a negative polarity. Tenses are equally neutralised. In all these elements lies its invariability.

20As for innit, its invariability compared with standard tags has been a matter of concern. Andersen (104) considers innit as “a transition from the canonical to the invariant category” since a distinction is to be “made between paradigmatic and non-paradigmatic use of innit”. It is through the deviation from the norms of Standard English that Andersen tackles the invariability of innit:

[…] the use of innit (or is it) in contexts involving the present tense of BE and a third person singular neuter subject is considered paradigmatic, while use in other contexts is considered non-paradigmatic. (Andersen 104)

21The following examples illustrate the paradigmatic use where innit is the negative reassessment of an identification:

(10) It's getting colder now innit? I can't stand it. (BNC, conversation, Female, Central South-West England)

(11) It's a game of chance innit I mean you pay your money. (BNC, Six O’Clock News)

(12) That's what we're doing innit mummy? (BNC, conversation)

22As for the non-paradigmatic use of innit, in utterrances such as I might go innit, Andersen (105), says:

[it] involves a greater deviation from the norms of standard English, since this use implies non-adherence to the rules of subject-verb concord between proposition and tag […].

23A very limited number of non-paradigmatic innit can be found in the BNC:

(13) Alright then? Great! They will have to pay innit? It's too late! Just as well. (BNC, Conversation, female, Home counties)

(14) “We need to decide what to do about that now innit.” (http://www.bbc.co.uk/​voices/​yourvoice/​conversation3.shtml)

24Here, innit is more than the contracted form of isn’t it. It is in place of the modal will or in place of do.

25So how do ain’t and innit compare on an invariability scale? Innit, in its non-paradigmatic use, has more potential than ain’t in as much as it can stand for any predication in its negative polarity and can appear by itself in a tag (whereas ain’t needs an extra pronoun). Having a neutral element as a referent may not be the issue. It would be tempting to consider innit as a generic tag but we cannot ignore the constraint of tense. Innit, among the 2010 utterances in the BNC, always follows predications in the present tense:

(15) He's horrible! Well that's damn nice innit? It's not very nice though! (BNC, conversation, Female, Lancashire)

(16) Yeah. Never mind. It's difficult Pat innit? This is it, let's hope anyway something comes. (BNC, conversation, Female, Lancashire)

26This tense restriction limits the capacity of innit as a multi-operator. Like ain’t, it contains an embedded negation but it does not have the same potential. As for ain’t, even if it functions as a generic negative predicate, its personal pronoun restriction cannot be underestimated.

27In a word, both ain’t and innit are formally invariant tags. They still correspond to different uses, different systems of equivalence, and show different levels of predicative versatility.

2. Comparing two non-standard non-variable markers: the limit of a segmental approach

28To compare ain’t and innit, we could consider them both under the angle of cliticisation. We will adopt Crystal’s definition:

A term used in grammar to refer to a form which resembles a word, but which cannot stand on its own as a normal utterance, being structurally dependent upon a neighbouring word in a construction. […] Such clitic words (‘clitics’) can be classified into proclitics (i.e. they depend upon a following word, as in the case of the articles in English and French) and enclitics (i.e. they depend upon a preceding word, as in the attachment of some pronouns to the end of a verb form in Italian or Spanish). (Crystal 76)

29Merely considering ain’t and innit as clitics could be a possibility and would be in keeping with non-standard identity and their use in spontaneous speech. Also, in context, both ain’t and innit easily collocate with other phonologically reduced and fused forms, such as dunno/ gonna/ ’ve. Again, this has to do with the type of conversations and subjects dealt with as well as the spontaneity they bear in a rather non-standard context:

(17) Like his curly hair father? I dunno. Innit smart? Yeah. He's real smart! (BNC, conversation, Male, North-West Midlands)

(18) You're trying to pull the rest of us down with it, well it ain't gonna happen. The next steps, cos we aren't gonna be frightened by your tinpot threats about y'know, junketing etcetera. (BNC, Bradford Metropolitan Council: debate (Pub/instit). Rec. on 20 Jan 1994)

(19) We ain't gonna win it anyway. He's still only 21. (BNC, Bradford Metropolitan Council: debate (Pub/instit). Rec. on 20 Jan 1994)

30It might then be tempting to rank the two markers on a cliticisation scale. We could consider that innit is more of a clitic than ain’t in as much as it goes beyond a fused negation and predicate since it also contains the third person pronoun. With innit, we could indeed consider that cliticisation is recurrent, merging a predicative marker, a negation and a personal pronoun. It is such that only the personal pronoun can be recognised, the other markers having disappeared in one marker. This is indeed the first hypothesis made by Andersen when he argues that “the interrogative isn’t it has undergone phonological reduction, specifically loss of [t] and [z] (presumably in that order), and that the two morphemes (a negative verb and a pronoun) have merged” (Andersen 106).

31But the non-paradigmatic use of innit that has previously been mentioned with occurrences such as I might not go innit, should already make us aware that innit may not necessarily contain the pronoun it. Innit may not be the mere fusing of a predicate + negation + a personal pronoun (be it a referent or a whole proposition in the context). Andersen, in a phonetic and derivational morphology approach, explains that the third person singular neuter form may not be at stake. It leads him to consider that innit might have developed around ain’t (Andersen 98):

• 4 We are speaking here of phonetic realisation. This might overlap with the derivational morphologica (...)

The alternative hypothesis is that innit is not a direct reduction of isn’t it, but has developed via the non-standard negative verb form ain’t and may correspond to either isn’t or hasn’t. This assumption is supported in Cheshire (1982) (and upheld by Stenström & Andersen 1996), who shows that the form innit is one of the several realisations4 of ain’t. This realisation of isn’t/hasn’t, along with the less phonologically reduced form int, is also found in COLT [the Bergen Corpus of London Teenage Language]. Given these facts we may argue that innit consists of a reduced form of ain’t and the pronoun it. This development can be sketched as follows:

ain’t it [eint it] --> int it [int it] ­­--> in it [in it] (Andersen 106)

32It therefore comes as a bit of a surprise that innit and ain’t are not two different fused forms but that one most likely contains the other. We had questioned the relationship of these two marginal markers in our title and in the introduction. “Margin within margin” was questioning whether we should consider innit (a less studied non-standard form) as any more marginal than ain’t. The fact that innit is likely to contain ain’t comes as a counterargument. “Margin within margin” was also questioning whether both ain’t and innit are to be considered as sub-markers. We have already partly shown that they operate as multi-operators and thereby do more than standard markers.

33Our project was indeed to put into perspective those two non-standard forms. But the fact that innit might contain ain’t encourages us to suggest that limiting ourselves to derivational morphology and to a segmental approach, trying to assess what predicator and what pronoun have been reduced, may not lead us in the right direction. Instead, it might be worth considering these two non-standard forms on a wider scale, that of their positional flexibility, that of their macro-syntactic position. Considering them on a wider level might help us grasp their recurrent function.

3. Ain’t versus innit: toward standardised functions and invariabilisation

34In this third section, we are going to see that ain’t and innit both work as recurrent operators when considered beyond mere predicative scope. Ain’t, thanks to its predicative versatility, is a meta-operator that signals that what is said has previously been uttered orally.

35Innit, being more restrained and constrained, either starts an interro-negation or comes as a conclusion. Such uses reveal that innit tends to occur at the periphery of predications.

3.1 Ain’t as a meta-operator coding orality

36For ain’t, it is the lack of a fixed position on a macro-syntactic level that guarantees its versatility. It can be a generic multi-operator when it is a substitute for the predicate. It can also be a tool to make up for an elided predication when it is appended in a tag. Very much in keeping with what we have just said, it is most often used to make a comment at the end of the utterance and to reassert a predication that has already been validated.

37Indeed, ain’t can lend an oral quality to the words it affects (Gaudy-Campbell, “Ain’t métaopérateur” 247). It can appear in spontaneous speech, and is then the variation marker that we have considered until now. It can also appear in the more complex context that we have already quoted in 1.1:

(20) Either a Tory or a Labour minority government will look at the next election down the track and say, ‘Hang on, we ain't going to win’. (Daily Telegraph, electronic edition of 1992 0405)

38This is a written extract quoting discourse. Inverted commas appear to signal speech and ain’t corroborates the orality of the passage that is quoted. Such words have not been uttered but they are anticipated and ain’t contributes to their orality. They mimic speech, and anticipate what a member of the political opposition might say. The same type of context is to be found in the following occurrence. The orality is underlined by “in his own words”, which contributes both to quoting and to creating a distance with the speech that has been uttered and that may not be endorsed by the utterer:

(21) And, in the end, if much of the promise is still unfulfilled, in his own words, ‘I ain't done too bad’. (BNC, London: newspaper publication)

39Ain’t can go beyond this and can correspond to all types of marks signalling that speech is imported. It can make up for the lack of inverted commas or marks of subordination to signal imported speech and the orality of a passage. As such, ain’t is used to code orality. This can be seen in the following examples:

(22) He said they ain't no bother really. (BNC, conversation, East Anglia)

(23) “I tell you, you ain't leaving me here!” (BNC, Paper faces. Anderson, Rachel. Oxford: OUP, 1991)

40Both are extracts from spontaneous conversations. But within each, one notices heterogeneity, namely the introduction of speech within speech. In the first utterance, what the speaker has uttered is mimicked. Really confirms the orality of words that we are made to hear. The intonative contour is very much reconstructed. Ain’t appears to code the predication they-be no bother and signals that this is speech within speech. In the second utterance, it is not so much another speaker that is mimicked. The utterer distances himself from himself, “you ain’t leaving me here” being a threat or a plea that the utterer of “I tell you” has just worded in a different way (possibly uttering it more quickly, or with more intensity or with an intonative pattern that entices the co-utterer’s reaction). Ain’t signals here that different levels of orality are intertwined. What is common to both utterances is that ain’t signals that different speech acts and speech levels have been interwoven. As such, we consider that ain’t has an invariant function.

41Ain’t signals orality within written narrative, but it also signals orality within spontaneous speech, allowing the main utterer to give room to another speaker. This heterogeneity unfolds through different modes, the main utterer splitting himself from a second speaker, be it from another being or from himself. That distance is regularly backed by the presence of the –s agreement on the predicate, that puts speech in perspective, namely in the context of I says:

(24) And I said to me mother, I says look here, when I'm old enough to work I says, I ain't going to work for me dad I shall get a job on me own. (BNC, conversation)

(25) I says no I’m not! Cos I ain’t signed anything! (BNC, conversation, South Midlands)

42Regarding I says, Quayle, suggests the following analyses:

• 5 Le locuteur narrateur parle de lui-même, non pas comme de quelqu’un situé dans la sphère du Moi […] (...)

The utterer speaks for himself, not as somebody situated in the set of references for I […] but as somebody that exists outside or beyond that sphere.5 (Translated from Quayle 171-172)

44Hence, beyond its invariant function that signals the orality of the extract, ain’t points at the fact that the utterer distances himself from what he has said. It enables a modal distance on the predicative content, but is also a means to put into perspective the way the speech is appropriated. The utterer adopts two attitudes: that of the speaker reporting what the co-utterer has said and that of the reported speaker that he also incarnates. Thus, the function of ain’t goes far beyond the mere fusing of the elements that it contents. It is a meta-operator. It operates at a meta-discursive level and codes the discursive level of the words that it affects.

3.2 Innit: toward a standardisation of its use as a thetical

45As for innit, we consider that it is a macro-operator affecting the utterance as a whole. First it tends to have less syntactic mobility. We will still quote the following five counter-examples (out of 2020) that can be found in the BNC:

(26) Well don't move him. Oh innit sad? See it got all marked down there. (BNC, conversation, central South-West England)

(27) Oh innit sad, our Sally got a tank full of petrol. (BNC, conversation)

(28) It's a clear line isn't it? God innit clear? Int it clear! (BNC, conversation, Northern England)

(29) Paul Merton has the first post-modern catchphrase. ‘Innit marvellous!’ He's said it about three times, (BNC, New Musical Express. 1991)

(30) Five o'clock? I used to like that shift, oh innit alright for some? (BNC, conversation, Central Northern England)

46Those utterances merely correspond to 0,25% of utterances containing innit in the BNC. But as such, they are fully acceptable and quite often heard in very spontaneous conversations. In such utterances, innit appears at the start and works as an interro-negation. It hence pre-constructs the opposite of what is said. Innit is highly pre-supposing.

47Further examples are:

(31) Here you are there it is. That's beige though innit? it's the colour innit? (BNC, conversation, Central South-West England)

(32) He can't open the door well that's clever innit leaving the door open how funny, clever; (BNC, conversation, Midlands)

(33) That's round about thirty-seven per cent, innit, somewhere around there. Yeah, we paid. (BNC, conversation, Central Northern England)

48In these examples, innit may appear in a middle position, the linearity of the transcription may be confusing but innit actually corresponds here to tags. Another way of considering such cases could be to see them as discourse markers in the online processing of speech as the last two utterances seem to show.

49Further examples are:

(34) oh oh oh that's only four though innit you see? No, yeah. […] (BNC, conversation, North-East England)

(35) No I think this is the cheap one. Dear though innit? Well it is a dear game […]. (BNC, Conversation, Central south-west England)

(36) Yeah a cat's a little bit out on dashing upstairs now innit really? It's a shame […]. (BNC, Conversation, East Anglia)

(37) Get home now. It's boring life really, innit really? David's been away all week? (BNC, Lancashire)

50Innit collocates here with other discourse markers such as you see, though, really. We have noticed also another regular contextual pattern:

(38) Oh I heard that's what they had innit? You're favourite. Oh they're gone; (BNC, Conversation, East Anglia)

(39) Or five six seven and that's what you wanna do innit? Oh yeah. No no I wouldn't (BNC, East Anglia)

(40) and I said it's air conditioned and everything innit? He said, yeah! That's right, that's; (BNC, Central South-West England)

(41) There Oh yeah. It's like the tuner's dead innit? Oh It's getting colder; (BNC, Central South-West England)

51What is the scope of innit here? Is it on the main clause or on the subordinate clause? The fact that the tag can be formed on the embedded clause or on the main clause is a key issue (Charreyre 88-112; Quirk 811; Houck 32; Huddleston & Pullum 893; Mc Gregor 109; Kimps 273; Axelsson; Kim & Ann 9).

52If we consider that innit corresponds to a third person singular, then it targets the relation that-be. But we cannot ignore that in utterance 40, innit has an impact on the subordinate clause. In utterance 39, innit bears on the main clause and utterance 41 is most ambiguous. So rather than wondering what clause or predicate innit affects, it seems more adequate to consider that it affects the utterance as a whole.

53In this perspective, the issue is not to be found at a predicative level but at a wider level, at a level that is peripheral to the utterance, very much “outside the clause” (to quote a conference held in Wien in July 2014, https://otc-workshop.univie.ac.at/​). In other words, we could consider that innit is a thetical, to abide by the category put forward by Kaltenböck, Keizer and Lohmann:

[…] there has been increased interest in syntactic elements which do not fit neatly into the syntactic structure of the clause but are somehow external to it (e.g. Kaltenböck, Keizer, Lohmann forthc.). These elements are often subsumed under the term of macro syntax or extra-clausal elements but are also discussed under various subheadings such as parentheticals, disjuncts, pragmatic markers, interjections, tails and afterthoughts, vocatives, formulae of social exchange, etc. (translated from Kaltenböck 11)

54The mobility of innit might be limited within the utterance, but its scope is that of the utterance as a whole. Innit is not just a negative predicative marker, working in pair with other predicates inside the clause(s) in the complexity of interwoven clauses. What is its scope? The main clause? The subordinate clause? Its scope has a wider reach. It comes as a discursive comment, that can be at the end of the utterance or at the beginning where it pre-constructs from the start the opposite of the predication that appears then in the utterance. It works as a punctuation marker within an utterance at some stage of its processing more than it targets any predicate. Innit is therefore to be considered at a much larger macro-syntactic level.

55Another way to consider innit could be to follow Erman’s explanation (87), when he argues that innit (along with just) is “in the process of being grammaticalised in teenage language” and has “changed category membership from that of tag question … to that of pragmatic”.

56So both ain’t and innit operate beyond the mere predicative level. Ain’t encodes the whole utterance. It either signals its orality or the necessary distance that the utterer has toward his speech. Innit does not quite have the same meta-function. It does not code as such the content of the utterance. Situated at the periphery of discourse units, at the periphery of utterances, it has a strong global function in as much as it punctuates an utterance and works as a thetical.

57Ain’t and innit, however marginal they might be, both have a regular function and are undergoing an “invariabilisation (grammaticalisation) process” (Andersen 100). To be more precise, we could quote Andersen’s definition of invariabilisation (Andersen 98):

Invariabilisation is perceived as a type of grammaticalisation which involves structural analysis, specifically rebracketing of the verb and pronoun to a single morphemic unit ([in] + [it] --> [innit] and [is] + [it] --> [is it]), accompanied by loss of semantic features and increase in pragmatic significance.

58In other words, those two non-standard forms are on their way to taking on a recurrent function and that function surpasses that of their standard counterparts.

59To conclude this last part, we might want to reconsider the concept of invariability developed throughout this article. We first considered ain’t and innit for their invariable form. This conditions their marginal and non-standard status against the standard grammatically dependent tags. The very fact that they are formally invariable enables them to be less predicate dependent. In other words, the very fact that they are marginal against the standard predicative system enables them to have alternative influences, be it at the level of the clauses, or at the level of the whole utterance. It thus widens their mobility and macro-syntactic anchorage and gives them a recurrent function (with ain’t having a meta-dimension and innit working as a thetical). So rather than tackling the marginal forms for their formal invariability, it is more accurate to tackle them for their recurrent function and for their ability to undergo invariabilisation. A binary opposition between standard and non-standard forms is thus proved totally irrelevant and counterproductive.

Conclusion

60At the end of this paper, we can say that even if innit and ain’t are non-standard markers, their use is not marginal. Far from being a lesser version of the standard predicative markers, these so-called non-standard forms achieve more than the standard predicative markers and have specific roles.

61Ain’t operates at a predicative level. To this, one can add the meta-dimension that it bears, since it has the capacity to encode orality within speech.

62Innit cannot merely be considered for its limited impact. It can occur at the start or at the very end of an utterance and thus affect the whole unit of speech by inverting its polarity. It can also appear in a middle position, where it comes as a parenthetic comment. It thereby acts as a thetical. In parallel, innit can also have a non-paradigmatic use and be considered for its scope. Innit, this variant, has an impact on an utterance at a global level.

63Compared with each other, innit is no more marginal than ain’t. If we consider ain’t and innit simply as non-standard forms then we lose sight of their grammatical complexity. It has prevented us from identifying the function of those markers and from putting their relative functions in perspective. Far from belittling them as non-standard markers, one should value them for their non-standard identity. This enables them to function differently and to have more than just a micro-predicative impact.